DESCRIPTION METHOD OF PRODUCTION OF ULTRALOW CARBON CAST SLAB TECHNICAL FIELD The present invention relates to a method of producing an ultralow carbon cast slab by continuous casting. BACKGROUND ART The dissolved oxygen in molten steel refined by a converter or vacuum treatment container is generally removed by the deoxidizing element Al. However, if deoxidizing ultralow carbon steel with a large content of dissolved oxygen by Al, alumina (Al203) will form. These agglomerate and merge to form large amounts of coarse clusters of hundreds of microns size or more. Part of these alumina clusters enter the immersion nozzle from the tundish at the time of continuous casting. If sticking at the inner bore of the immersion nozzle, they cause nozzle clogging and obstruct operation. Further, if the alumina clusters enter the casting mold and remain at the surface layer of the cast slab, they become causes of formation of surface flaws of thin-gauge steel sheet and have a detrimental effect on quality. As a countermeasure against this, in general there is the method of blowing in Ar gas from a tundish upper nozzle, sliding nozzle, or immersion nozzle to cause the Al203-based inclusions to stick to the surfaces of bubbles, prevent them from sticking at the inner bore of the immersion nozzle, and cause them to float up in the casting mold for removal. However, with this method, there were the problems that the Ar bubbles blown in became the cause of pinhole defects in the cast slab and that, further, the Ar bubbles floating up in the casting mold disturbed the meniscus and caused powder entrainment, with the entrained powder particles becoming the cause of surface flaws at the thin-gauge steel sheet. Further, when continuously casting molten steel, from the viewpoint of ease of production, usually, as shown in FIG. 7, an immersion nozzle 1 of a straight shape with a fixed inside diameter from the top end to the bottom end of the inner bore 10 is used. However, in the case of an immersion nozzle with an inner bore of a straight shape, as shown in FIG. 8, the opening part 12 of the sliding nozzle 11 is offset from the center of the immersion nozzle 1, so when the molten steel in the tundish (not shown) passes through the sliding nozzle 11 and flows into the immersion nozzle 1, similarly, as shown in FIG. 8, an uneven distribution of the molten steel flow rate inevitably occurs in the immersion nozzle 1 (in the figure, see downward arrows at the center). Due to this, there was the problem that uneven flows 13a, 13b with different flow rates occurred at the left and right discharge ports, the state of fluid motion in the casting mold was disturbed, and powder or bubbles were carried to deep positions in the unsolidified part of the molten steel and remained in the cast slab. To solve these problems, two methods of solution have been disclosed based on prior discoveries. The first is the method of using an immersion nozzle having an orifice in the inner bore for molten steel killed by Al. This method has as its object the prevention of sticking of alumina in the inner bore and suppression of uneven flow. For example, Japanese Patent Publication (A) No. 2001-239351 discloses an immersion nozzle having a plurality of step differences in its inner bore. Further, Japanese Patent Publication (A) No. 2004-255407 discloses an immersion nozzle having a plurality of discontinuous projections in the inner bore. Further, in these patent documents, if providing an orifice in the inner bore (step differences or projections), the part in the immersion nozzle where the molten steel flow rate becomes remarkably slow is eliminated and the flow rate is made uniform. As a result, it is disclosed, the effects of suppression of uneven flow and prevention of sticking of alumina are obtained. Further, Japanese Patent Publication (A) No. 2001-239351 discloses that an inert gas flow rate is suitably INI (normal liter)/min to 40N1 (normal liter)/min. Note that below, "normal liter" will sometimes be simply expressed as "Nl". The second is the method of preventing the formation of alumina clusters. For example, Japanese Patent Publication (A) No. 2002-88412, Japanese Patent Publication (A) No. 2003-49218, Japanese Patent Publication (A) No. 2003-268435, Japanese Patent Publication (A) No. 2005-60734, and Japanese Patent Publication (A) No. 2005-139492 disclose the method of using Ti and rare earth metals for deoxidation (hereinafter referred to as "deoxidation by Ti-rare earth metals". This method deoxidizes the molten steel by Ti to form Ti oxides, then adds rare earth metals to change the Ti oxides to spherical inclusions resistant to agglomeration and merger. According to this method, it is possible to prevent sticking of inclusions on the immersion nozzle, clogging of the immersion nozzle, and formation of surface flaws due to alumina clusters. Further, Japanese Patent Publication (A) No. 11-343516 discloses the method of adding one or both of Ca and a rare earth metal after deoxidation by Ti and continuously casting without blowing in Ar gas. This method as well is a method suppressing the formation of cluster-like inclusions and making the inclusions finely disperse. By this method, it is possible to obtained titanium-killed steel excellent in surface properties. DISCLOSURE OF THE INVENTION First, the method disclosed in Japanese Patent Publication (A) No. 2001-239351 and Japanese Patent Publication (A) No. 2004-255407, that is, the method of using an immersion nozzle having an orifice in its inner bore to continuously cast Al-killed molten steel will be explained. In the technology disclosed in the above patent documents, even if using an immersion nozzle having an orifice in its inner bore, the effect of making the distribution of the flow rate uniform is hard to obtain. This is because in the above art, there is the problem of nozzle clogging. This reason is that at the part below the bottom end of the orifice, an eddy flow occurs and stirring is caused, so the alumina will not stick, but at the part above the top end of the orifice, no eddy flow occurs, so sticking of alumina cannot be avoided. In particular, the top end of the orifice is the location where sticking of alumina progresses the most. If a large amount of alumina-based inclusions stick here, the nozzle will clog. As disclosed in Japanese Patent Publication (A) No. 2001-239351, if blowing Ar gas into the molten steel, nozzle clogging can be prevented. However, part of the Ar gas blown into the molten steel fills the immersion nozzle and pushes the melt surface in the immersion nozzle (secondary meniscus) down in position. The molten steel flowing in from the tundish to the immersion nozzle drops freely from the position of the sliding nozzle to the secondary meniscus, but if the secondary meniscus is pushed down, the dropping distance of the molten steel will become longer, so a strong downward flow will easily occur right below the? dropping position of the molten steel. In some cases, as a reaction to this, as shown in FIG. 9, a reverse rising flow 14 (in the figure, see dotted line arrows) occurs whereby an uneven distribution of the molten steel flow rate occurs in the immersion nozzle (in the figure, see solid line arrows). The uneven distribution of the molten steel flow rate occurring due to the dropping flow of the molten steel is eased if making the secondary meniscus close to the position of the sliding nozzle. However, to obtain the effect of prevention of sticking of alumina, at least a predetermined level of Ar gas flow rate is required. Normally, the Ar gas flow rate is 5 to 20Nl/min, but with this Ar gas flow rate, it is difficult to make the secondary meniscus close to the position of the sliding nozzle. If the secondary meniscus is low and the distance to the top end of the orifice (in the figure, see "21a"i is short, there is the possibility that the molten steel will not stop once at the part above the orifice but will pass through the orifice with the distribution of the flow rate still not made uniform. Accordingly, in the state of a low secondary meniscus, it is difficult to suppress uneven flow caused by a dropping flow by just the orifice. Next, the method disclosed in Japanese Patent Publication (A) No. 2002-88412, Japanese Patent Publication (A) No. 2003-49218, Japanese Patent Publication (A) No. 2003-268435, Japanese Patent Publication (A) No. 2005-60734, Japanese Patent Publication (A) No. 2005-139492, and Japanese Patent Publication (A) No. 11-343516, that is, the method of continuously casting molten steel not forming any alumina clusters will be explained. In the art disclosed in the above patent documents, the inclusions are resistant to agglomeration and merger, so no coarse clusters are formed and the nozzle is resistant to clogging. However, in the above patent documents, the shape of the inner bore of the immersion nozzle is not defined. No technical matter relating to the secondary meniscus is described. In the above art, no means is devised for suppressing uneven distribution of the flow rate and uneven flow, so there is a high possibility of powder or bubbles being carried to deep positions of the unsolidified part of the molten steel and of powder or bubbles remaining in the cast slab becoming a cause of surface flaws formed when worked to a thin-gauge steel sheet. In this way, in the prior art, there was the problem that it was difficult to achieve both prevention of nozzle clogging and securing of the cast slab quality. Note that the "securing of the cast slab quality" here means stably producing a cast slab free of surface flaws even if worked to a thin-gauge steel sheet. The present invention, in consideration of these problems, has as its object the provision of a method of production of an ultralow carbon cast slab enabling achievement of both efficiency of continuous casting aic; cast slab quality. The inventors engaged in repeated research to solve the above-mentioned problems and as a result discovered that if deoxidizing ultralow carbon molten steel by Ti-rare earth metals (Ce, La, Nd) and using an immersion nozzle having an orifice in its inner bore, it is possible to prevent nozzle clogging and continuously cast an ultralow carbon cast steel free of surface flaws even if worked to a thin-gauge steel sheet. The present invention was made based on the above discovery and has as its gist the following constitutions: (1) A method of production of an ultralow carbon cast slab characterized by adding Ti to molten steel decarburized to a carbon concentration of 0.01 mass% or less, further adding at least one of Ce, La, and Nd, and using an immersion nozzle to inject the above molten steel from a tundish to a casting mold for continuous casting while maintaining a flow rate of Ar gas blown 8from any location in a range from a tundish upper nozzle to discharge ports of said immersion nozzle at 3N1 (normal liter)/min or less. (2) A method of production of an ultralow carbon cast slab as set forth in (1), characterized in that said immersion nozzle has an orifice in its inner bore. (3) A method of production of an ultralow carbon cast slab as set forth in (2), characterized in that said inner bore has a circular cross-sectional shape and (i) a relation 3s • K)] is calculated from the amount of heat removal q of the casting mold [J/(nr-s)] and the temperature difference (Tco-TMD) between the molten steel and the casting mold surface [K]. The amount of heat removal C[ of the casting mold is calculated from the change in temperature (tout-tin) before the casting mold cooling water passes over the copper plates and after it passes over it [K], the cooling water flow rate Qw [kg/s], the specific heat Cw of water [J/(kg*K)], and the copper plate surface area S [m2] . h=q/(Too-TMD) (1) q=QwCw(tout-tin)/S (2) Below, the present invention will be explained in detail. The present invention covers ultralow carbon steel. The upper limit value of the carbon concentration is not particularly limited, but thin-gauge steel sheet of ultralow carbon steel is used for steel sheet for automobiles etc. subjected to severe working, so has to be provided with superior workability. For this reason, the carbon concentration is preferably 0.01 mass% or less. Note that the lower limit value of the carbon concentration is not particularly defined. In the present invention, in the secondary refining, the steel is decarburized to a carbon concentration of 0.01 mass% or less, then Ti is added to the molten steel for deoxidation. The amount of addition of the Ti is preferably 0.04 mass% or more. If less than 0.04 mass%, the deoxidation does not sufficiently occur and there is a high possibility of dissolved oxygen remaining in the molten steel. In secondary refining, when decarburizing steel to a carbon concentration of 0.01 mass% or less, it is also possible to deoxidize the steel by Al preliminarily before the decarburization in the refining apparatus, for example, at the stage of refining by the converter. In this case, the Al concentration after deoxidation is made 0.01 mass% or less, preferably 0.008 massl or less, more preferably 0.006 massl or less. If the Al concentration after deoxidation is 0.01 mass% or less, the deoxidation product, that is, alumina, floats up to the surface of the molten steel in the interval until continuously casting the molten steel d■;■:, can be removed, so the amount of alumina remaining in the molten steel during casting becomes smaller and the problems of nozzle clogging etc. do not occur. Further, if the Al concentration after deoxidation is 0.008 massl or less, the amount of alumina remaining in the molten steel during casting becomes smaller, so this is preferable. Further, if the Al concentration after deoxidation is 0.006 mass% or less, the amount of alumina remaining in the molten steel during casting becomes even smaller, so this is more preferable. On the other hand, the upper limit of the amount of addition of Ti is not particularly defined. The Ti oxides produced by addition of Ti are harder to aggregate and merge than Al203-based inclusions, but easily stick to the refractory, so nozzle clogging is a concern. Therefore, after deoxidizing the molten steel by Ti, at least one of Ce, La, and Nd is added. By this addition, the Ti oxides become hard to agglomerate and merge and are converted to spherical inclusions hard to stick to the refractory. The total amount of addition of Ce, La, and Nd is preferably 0.001 mass% to 0.01 mass%. If the above total amount of addition is less than 0.001 mass%, the modification of the Ti oxides becomes insufficient and spherical inclusions hard to agglomerate and merge become hard to form. Further, if over 0.01 mass%, the modification of the Ti oxide becomes excessive, the Ti-based inclusions become heavier in specific gravity and become harder to float up, and the cleanliness of the molten steel deteriorates. Rare earth metals other than Ce, La, and Nd (for example, Pr, Sm, etc.) do not have effects of modification equal to chose of Ce, La, and Nd, so for modification of the Ti oxides, addition of one or more of Ce, La, and Nd is effective. Here, FIG. 1 shows an immersion nozzle having an orifice used in the present invention. In the present invention, the portion of the inner bore 10 smaller in inside diameter than the top end of the inner bore is defined as the "orifice 21", while any portion with an inside diameter equal to the top end of the inner bore or with an inside diameter larger than the same is defined as a "non-orifice part 21z". At the interfaces of the orifice 21 and the non-orifice parts 21z, the boundary at the upstream side is referred to as the "orifice top end 21a" and the boundary at the downstream side is referred to as the "orifice bottom end 21b". As explained above, in the present invention, the Ar gas flow rate can be greatly reduced from the time of the conventional Al deoxidation. If reducing the Ar gas flow rate, the secondary meniscus rises. When reaching 3Nl/min or less, the secondary meniscus rises to a position of about 100 to 120 mm from the sliding nozzle. If the secondary meniscus rises to the above position, almost no strong downward flow occurs and the distance to the orifice top end can be sufficiently secured, so it is possible to reliably eliminate uneven distribution of the molten steel flow rate. This fact was discovered by the inventors. Note that as the value of the Ar gas flow rate in the present invention, it is possible to use a value measured using a commercially available flowmeter. The smaller the Ar gas flow rate, the more the secondary meniscus rises, so the Ar gas flow rate is preferably 2Nl/min or less, more preferably less than lNl/min. If the Ti oxides are all suitably modified, almost no sticking of Ti oxides to the immersion nozzle will occur, so the lower limit value of the Ar gas flow rate also includes ONl/min. Ar gas is generally blown in from one or more of a tundish upper nozzle, sliding nozzle, or immersion nozzle, but if in the range from the tundish upper nozzle to immersion nozzle discharge ports, the positions of the locations where the Ar gas is blown in and the number of those locations may be freely selected. In the immersion nozzle used in the present invention, to more remarkably secure the effect, of suppression of an uneven flow, there is a preferable range of orifice size. As the cross-sectional shape of the immersion nozzle, normally a circular or elliptical shape is used. FIG. 2 shows an immersion nozzle with a circular cross-sectional shape, while FIG. 3 shows an immersion nozzle with an elliptical cross-sectional shape. When the cross-sectional shape is circular, the radius of the top end of the inner bore is defined as R [mm] and the smallest radius of the orifice is defined as £ [mm]. On the other hand, when the cross-sectional shape is elliptical, the radius of the top end of the inner bore in the long axis direction is defined as A [mm] and the smallest radius of the orifice in the long axis direction is defined as a [mm]. Here, for the orifice, the smallest radius is used because in the present invention, an orifice is defined as a "portion smaller in inside diameter than the top end of the inner bore". As the shape of the orifice, even a shape with an inside diameter not fixed from the top end to the bottom end of the orifice may be envisioned, so the smallest radius was used to define it so that the invention can be applied even to such a shape. Next, the difference between the radius of the non-orifice parts and the smallest radius of the orifice is defined as the "height of the orifice". Normally, the radius of the non-orifice parts is equal to the radius of the top end of the inner bore, so the "height cf the orifice" can be alternately referred to for the difference between the radius of the top end of the inner bore and the smallest radius of the orifice. This being so, the "height of the orifice" is expressed by "R-r" when the immersion nozzle has a circular cross-sectional shape and "A-a" when it has an elliptical one. Further, the distance from the top end to the bottom end of the orifice is defined as the "length of the orifice" and labeled L [mm]. When the immersion nozzle has a circular cross-sectional shape, the "height of the orifice" preferably satisfies the relation "330 results in a remarkably larger flow rate of the molten steel passing through the orifice and the fluid motion in. the casting mold is easily detrimentally affected. Next, the length L of the orifice [mm] preferably satisfies the relation 50150, the part of the small inside diameter becomes longer, so the flow rate of the molten steel becomes remarkably greater and the fluid motion in the casting mold is easily detrimentally affected. When the immersion nozzle has an elliptical cross-sectional shape, the "height of the orifice" preferably satisfies the relation "3 small, a problem occurs in the strength of the material forming the immersion nozzle, but Rz can be made larger, so this is advantageous for prevention of divergence of the discharge flow. For example, in the case where =45° or so, the radius of curvature Rz is preferably 50 to 100 mm. Further, as the immersion nozzle used in the present invention, as shown in FIG. 18, one having an orifice 21 with an opening cross-sectional area smaller than the opening cross-sectional area of the cylinder between the top end of the nozzle and the discharge ports 2 is preferable. In an immersion nozzle having a slit at the cylinder bottom part, if the flow rate of the melt increases, the force of the flow of melt pushing the slit wider increases. For this reason, as shown in FIG. 19, it is preferable to provide the slit 3 with ribs 22 connecting the slit side walls 9. By providing the ribs 22, even if the force pushing the slit 3 wider increases, deformation or breakage of the immersion nozzle 1 can be prevented. As the refractory used for the immersio-n nozzle, alumina graphite, alumina spinel, or another conventionally used refractory may be used. However, depending on the ingredients of the molten steel, an alumina graphite immersion nozzle will sometimes be corroded and dissolve during casting, so it is preferable to make the cylinder side walls and cylinderottom part of the immersion nozzle or part or all of the surfaces of the discharge ports and slit contiguous with the molten steel out of any of carbonless spinel, low-carbon spinel, magnesia graphite, zirconia graphite, or silica-less alumina graphite (high melt loss resistance refractories). FIG. 20 shows an immersion nozzle where the cylinder side walls 5 and cylinder bottom part 6 and all of the surfaces of the discharge ports 2 and slit 3 contiguous with the molten steel are made of a high melt loss resistance refractory 23 and the other parts are made of an ordinary refractory 24. In the present invention, there is extremely little sticking of nonmetallic inclusions to the immersion nozzle and no nozzle clogging, so the surface of the cast slab is resistant to formation of surface defects due to cluster-like inclusions. Further, the cast slab obtained by the present invention is free of surface flaws even if made into thin-gauge steel sheet by hot rolling, cold rolling, or other ordinary methods since the penetration of bubbles and powder causing surface flaws is suppressed at the time of production. Further, the present invention exhibits similar effects not only when applied to continuous casting of a slab of a usual thickness of 250 mm, but also in continuous casting of a thin slab with a casting mold thickness thinner than this, for example, 150 mm, so a cast slab of an extremely good quality can be obtained. [Examples 1] Below, examples and comparative examples (see Table 1) will be given to explain the present invention using immersion nozzles of various shapes (see FIGS. 1 to 6). Table 1 (Table Removed) Example 1-1 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in a converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.3. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method without Ar (flow rate ONl/min) to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, and the inner bore was straight in shape. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 200 J/(m2 •s•K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 250J/ (m2-s•K), so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, the cast slab did not have any pinhole defects. Further, the surface of the cold rolled steel sheet had 0.2 surface flaw/coil. However, the surface flaws occurred at positions in the range of the edge trim, so did not become a problem in product quality. Example 1-2 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La ;-' oy of a mass ratio Ce/La=1.3. The result was refluxec. ' 5 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 2.8Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: R-r=5 mm, length: L=90 mm (orifice). The top end of this step difference was at a position 400 mm from the top end of the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 150J/(m2 •s•K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 250J/(m2 ■ s • K) , so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, there were neither pinhole defects in the cast slab nor surface flaws on the steel sheet. Example 1-3 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce, La, and Nd were added by a Ce-La-Nd alloy of a mass ratio Ce/La=1.3 and La/Nd=3.5. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce, La, and Nd of 0.01 mass%, and a Ld concentration/Nd concentration of 3.5. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 0.5Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 mm and an inside diameter of 85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: R-r=5 mm, length: L=90 mm (orifice). The top end of this step difference was at a position 400 mm from the top end of the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 100J/ (m2- s•K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 250J/(m2-s•K) , so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast: slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, there were neither pinhole defects in the cast slab nor surface flaws on the steel sheet. Example 1-4 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.3. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method with no Ar .(flow rate ONl/min) to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 mm and inside diameter of 85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: R-r=5 mm, length: L=90 mm (orifice). The top end of this step difference was at a position 400 mm from the top end of the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 50 J/(ITT • s • K) difference of coefficient of heat transfer occurred. This was over the uneven flow judgment criteria of 250J/(m2 •s•K), so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, there were neither pinhole defects in the cast slab nor surface flaws on the steel sheet. Example 1-5 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce, La, and Nd were added by a Ce-La-Nd alloy of a mass ratio Ce/La=1.3 and La/Nd=3.5. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce, La, and Nd of 0.01 mass%, a Ce concentration/La concentration of 1.3, and a Ld concentration/Nd concentration of 3.5. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 0.5Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The cross-sectional shape of the immersion nozzle was an outer shape of an ellipse with a long axis of 170 mm and a short axis of 120 mm and an inner shape of an ellipse with a long axis of 105 mm and a short axis of 7 5 mm. The material of the inner bore was alumina graphits;. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: A-a=5 mm, length: L=90 mm (orifice). The top end of this step difference was at a position 400 mm from the top end of the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 100J/ (m2-s-K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 250J/ (m2-s • K) , so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet,of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, there were neither pinhole defects in the cast slab nor surface flaws on the steel sheet. Example 1-6 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.3. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 0.5Nl/min to obtain a cast slab of a thickness of 130 mm and a width of 1600 mm. Here, as the thin slab continuous casting method, a casting mold of a thickness of 150 mm or less was used for the casting. The cast slab was held at a temperature of 1000 to 1200°C by a holding furnace placed at the downstream side of the machine end and hot rolled without cooling to near ordinary temperature. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 70 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: R-r=5 mm, length: L=90 mm (orifice). The top end of this step difference was at a position 400 mm from the top end of the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 100J/(m2-s•K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 250J/(m2 • s • K) , so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, there were neither pinhole defects in the cast slab nor surface flaws on the steel sheet. Example 1-7 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% and a concentration of Al of 0.005 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce, La, and Nd were added by a Ce-La-Nd alloy of a mass ratio Ce/La=1.3 and La/Nd=3.5. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce, La, and Nd of 0.01 mass%, a Ce concentration/La concentration of 1.3, and a Ld concentration/Nd concentration of 3.5. This molten steel was cast by the continuous casting method into a cast slab of a thickness of 250 mm and a width of 1600 mm with a flow rate of Ar gas blown in from a tundish upper nozzle of 2.8Nl/min. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: R-r=5 mm, length: L=90 mm (orifice). The top end of this step difference was at a position 400 mm from the top end of the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 150J7 (m2-s•K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 2 50 J/(m2 •s•K) , so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 ram to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, there were neither pinhole defects in the cast slab nor surface flaws on the steel sheet. Comparative Example 1-1 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Al was added for deoxidation. This was refluxed for 5 minutes to produce molten steel with an Al concentration of 0.04 mass%. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish 35upper nozzle of 7Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, and the inner bore was straight in shape. During casting, starting from when a cast mass of 150 tons is passed, the opening degree of the sliding nozzle of the tundish gradually became larger, so it was judged that inclusions had stuck to the immersion nozzle. The speed was reduced to secure the supply of molten steel to the casting mold and the casting was completed. At the casting mold short side copper plates, a maximum 300J/ (m2-s•K) difference of coefficient of heat transfer occurred. This was over the uneven flow judgment criteria of 250J/(m2•s•K) , so it was judged that uneven flow occurred. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, the cast slab had 15 pinhole defects/m", while the steel sheet had 10 surface flaws/coil. Comparative Example 1-2 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, 36Al was added for deoxidation. This was refluxed for 5 minutes to produce molten steel with an Al concentration of 0.04 mass%. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 2Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height:: R-r=5 mm and a length: L=90 mm (orifice) . The top end of this step difference was at a position 400 mm from the top end of the inner bore. Soon after the start of casting, the opening degree of the sliding nozzle of the tundish gradually began to rise. At the point of time of a casting mass of 150 tons, even if making the opening degree full open, the supply of molten steel into the casting mold became insufficient. At this point of time, it was judged that the immersion nozzle was clogged. The casting was suspended in a state with a total of 130 tons of molten steel left in the ladle and tundish (cast mass 170 tons). During casting, at the casting mold short side copper plates, a maximum 330J/ (m2-s•K) difference of coefficient of heat transfer occurred. This was over the uneven flow judgment criteria of 250 J/ (m2 •s•K), so it was judged that uneven flow occurred. Along with the progression of the clogged state of the immersion nozzle, it is believed that the effect of the step difference (orifice) was lost, an uneven flow rate distribution occurred, and an uneven flow resulted. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, the cast slab had no pinhole defects, but the steel sheet had 10 surface flaws/coil. Comparative Example 1-3 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment oy a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.3. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 7Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, and the inner bore was straight in shape. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. However, at the casting mold short side copper plates, a maximum 300 J/(m2 •s•K) difference of coefficient of heat transfer occurred. This was over the uneven flow judgment criteria of 250J/(m2-s • K) , so it was38 judged that uneven flow occurred. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, the cast slab had 15 pinhole defects/m' , while the steel sheet had 3 surface flaws/coil. Comparative Example 1-4 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.2. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.2. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 4.5Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, and the inner bore was straight in shape. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. However, at the casting mold short side copper plates, a maximum 280J/(m2-s•K) difference of coefficient of heat transfer occurred. This was over the uneven flow judgment criteria of 250J/(m2 •s•K), so it was judged that uneven flow occurred. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 160 0 mm. The steel sheet quality was visually inspected o:\ the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, the cast slab had 10 pinhole defects/rp/, while the steel sheet had 3 surface flaws/coil. Comparative Example 1-5 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.2. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.2. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 3.5Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of.85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: R-r=5 mm, length: L=90 mm (orifice). The top end of this step difference was at a position 400 mm from the top end of the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. However, at the casting mold short side copper plates, a maximum 260J/(m2.s.K) difference of coefficient of heat transfer occurred. This was over the uneven flow judgment criteria of 250J/(m2.s.K), so it was judged that uneven flow occurred. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, the cast slab had 5 pinhole defects/m2, while the steel sheet had 1 surface flaw/coil. Here, FIG. 10 shows the relation between the Ar gas flow rate and cast slab pinhole defects based on Examples 1-1 to 1-6 and Comparative Examples 1-3 to 1-5 (deoxidization by Ti-rare earth metals). In the figure, the black dots show the case of use of an immersion nozzle with an inner bore of a straight shape, while the black squares show the case of an immersion nozzle having an orifice in its inner bore. Further, FIG. 11 shows the relation between the Ar gas flow rate and occurrence of steel sheet surface flaws based on Examples 1-2 to 1-6 and Comparative Examples 1-3 to 1-5 (deoxidization by Ti-rare earth metals). In the figure, the black dots show the case of use of an immersion nozzle with an inner bore of a straight shape, while the black squares show the case of an immersion nozzle having an orifice in its inner bore. In the present invention, it is learned that if using an immersion nozzle having an orifice in its inner bore or an immersion nozzle with an inner bore of a straight shape and casting with an Ar gas flow rate of 3Nl/min or less (region A in the figure), a good quality cast slab free of both cast slab pinhole defects and steel sheet surface flaws can be obtained. Examples 2 Below, examples and comparative example (see Table 2) will be given to explain the present invention using an immersion nozzle having a slit at its bottom (see FIG. 12 to FIG. 20). Table 2 (Table Removed) Example 2-1 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.3. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 2.8Nl/min to obtain a cast slab c" a thickness of 250 mm and a width of 1600 mm. The immersion nozzle used was the immersion nozzle shown in FIG. 13. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The cross-sectional area of the opening of the discharge ports was 2829 mm2, the slit width was 15 mm, and the angles 0i and 62 were both 45°. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, and the inner bore was straight in shape. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 100J7 (m2-s•K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 250J/(m2 •s■K), so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, there were neither pinhole defects in the cast slab nor surface flaws on the steel sheet. Example To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.3. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0*. 01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 0.5Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The immersion nozzle used was the immersion nozzle shown in FIG. 17 and FIG. 18. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm, the cross-sectional area of the opening of the discharge ports was 2829 mm2, the slit width was 15 mm, and the angles 9i and 62 were both 45°. The radius of curvature Rz of the curved surface 11 was 60 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: R-r=5 mm, length: L=90 mm (orifice). The top end of this step difference was at a position 400 mm from the top end of the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 80J/(m2-s-K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 250J/ (m2-s•K) , so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and ccla rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, there were neither pinhole defects in the cast slab nor surface flaws on the steel sheet. Example 2-3 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.3. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of ONl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The immersion nozzle used was the immersion nozzle shown in FIG. 17 and FIG. 18. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The cross-sectional area of the opening of the discharge port was 2829 mm2, the slit width was 15 mm, and the angles 0i and 92 were both 45°. The discharge angle § was 45°, while the radius of curvature Rz of the curved surface 11 was 60 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: R-r=5 mm, length: L=90 mm (orifice!. The top end of this step difference was at a position 400 mm from the top end c: the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 90J/(m2-s-K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 250J/(m2 • s•K), so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, the cast slab did not have any pinhole defects. Further, the surface of the cold rolled steel sheet had 0.2 surface flaw/coil. However, the surface flaws occurred at positions in the range of the edge trim, so did not become a problem in product quality. Example 2-4 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.3. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a fundish upper nozzle of 2.8Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The immersion nozzle used was the immersion nozzle shown in FIG. 19. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The cross-sectional area of the opening of the discharge ports was 2829 mm2, the slit width was 15 mm, and the angles 0i and 92 were both 45°. The discharge angle <|> was 45°, while the radius of curvature Rz of the curved surface 11 was 60 mm. The vertex angle 03 of the rib was 30°, the width Wr of the bottom surface of the rib was 15 mm, and the distance Dr between the side walls of the rib was 85 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: R-r=5 mm, length: L=90 mm (orifice). The top end of this step difference is at a position 400 mm from the top end of the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 110J/(m2 • s • K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 250J/(m2 • s • K) , so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, there were neither pinhole defects in the cast slab nor surface flaws on the steel sheet. Example 2-5 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% and a concentration of Al of 0.005 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.3. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method into a cast slab of a thickness of 250 mm and a width of 1600 mm with a flow rate of Ar gas blown in from a tundish upper nozzle of 2.8Nl/min. The immersion nozzle used was the immersion nozzle shown in FIG. 13. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The cross-sectional area of the opening of the discharge ports was 2829 mm2, the slit width was 15 mm, and the angles 9i and G2 were both 45°. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, and the inner bore was straight in shape. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 120J/(m2- s • K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 2 50 J/ (rrr • s • K) , so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, there were neither pinhole defects in the cast slab nor surface flaws on the steel sheet. Example 2-6 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% and a concentration of Al of 0.005 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Ti was added for deoxidation. The result was refluxed for 6 minutes, then Ce and La were added by a Ce-La alloy of a mass ratio Ce/La=1.3. The result was refluxed for 3 minutes to produce molten steel having a Ti concentration of 0.03 mass%, a total concentration of Ce and La of 0.01 mass%, and a Ce concentration/La concentration of 1.3. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of ONl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The immersion nozzle used was the immersion nozzle shown in FIG. 17 and FIG. 18. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The cross-sectional area of the opening of the discharge ports was 2829 mm2, the slit width was 15 mm, and the angles 0-; and 6? were both 45°. The discharge angle § was 20°, while the radius of curvature Rz of the curved surface 11 was 51 mm. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, while the inner bore had a step difference of a height: R-r=5 mm, length: L=90 mm (orifice). The top end of this step difference was at a position 400 mm from the top end of the inner bore. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. Further, at the casting mold short side copper plates, a maximum 50J/(m2-s•K) difference of coefficient of heat transfer occurred. This was under the uneven flow judgment criteria of 250 J/ (m2 •s•K), so it was judged that uneven flow did not occur. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, the cast slab did not have any pinhole defects. Further, the surface of the cold rolled steel sheet had 0.2 surface flaw/coil. However, the surface flaws occurred at positions in the range of the edge trim, so did not become a problem in product quality. Comparative Example 2-1 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 massl by refining in the converter and treatment by a vacuum degassing apparatus, Ai was added for deoxidation. The result was refluxed for 5 minutes to produce molten steel with an Al concentration of 0.04 rnass%. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 2.8Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The immersion nozzle used was the immersion nozzle shown in FIG. 13. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm. The cross-sectional area of the opening of the discharge ports was 2829 mm', the slit width was 15 mm, and the angles 0i and 02 were both 15°. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, and the inner-bore was straight in shape. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. However, at the casting mold short side copper plates, a maximum 280 J/{m2 •s•K) difference of coefficient of heat transfer occurred. This was over the uneven flow judgment criteria of 250J/(m2 • s • K) , so it was judged that uneven flow occurred. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per ceil. As a result, the cast slab had 10 pinhole defects/m", while the steel sheet had 3 surface flaws/coil. Comparative Example 2-2 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Al was added for deoxidation. The result was refluxed for 5 minutes to produce molten steel with an Al concentration of 0.04 mass%. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 2.8Nl/min to obtain a cast slab of a thickness of 250 mm and a width of 1600 mm. The immersion nozzle used was the immersion nozzle shown in FIG. 13. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm.. The cross-sectional area of the opening of the discharge ports was 2829 mm", the slit width was 5 mm, and the angles 0! and 92 were both 45°. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, and the inner bore was straight in shape. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. However, at the casting mold short side copper plates, a maximum 260J/(m2 • s • K) difference of coefficient of heat transfer occurred. This was over the uneven flow judgment criteria of 250J/(m2 •s•K), so it was judged that uneven flow occurred. The cast cast slab was cut to a length of 10000 mm to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold roller by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, the cast slab had 5 pinhole defects/m2, while the steel sheet had 1 surface flaw/coil. Comparative Example 2-3 To 300 tons of molten steel in a ladle given a concentration of carbon of 0.002 mass% by refining in the converter and treatment by a vacuum degassing apparatus, Al was added for deoxidation. The result was refluxed for 5 minutes to produce molten steel with an Al concentration of 0.04 mass%. This molten steel was cast by the continuous casting method with a flow rate of Ar gas blown in from a tundish upper nozzle of 2.8Nl/min to obtain a cast slab of a thickness of 2 50 mm and a width of 1600 mm. The immersion nozzle used when, casting was the immersion nozzle shown in FIG. 13. The cross-sectional shape of the immersion nozzle was a circle with an outside diameter of 150 and an inside diameter of 85 mm, the cross-sectional area of the opening of the discharge ports was 2829 mm2, the slit width was 25 mm, and the angles 01 and 02 were both 45°. The material of the inner bore was alumina graphite. The length from the top end of the inner bore to the top ends of the discharge ports was 590 mm, and the inner bore was straight in shape. During casting, there was no sticking of inclusions on the immersion nozzle and there was no problem with operability. However, at the casting mold short side copper plates, a maximum 300J/ (m2-s•K) difference of coefficient of heat transfer occurred. This was over the uneven flow judgment criteria of 250J/(m2 • s • K) , so it was judged that uneven flow occurred. The cast cast slab was cut to a length of 10000 nun to obtain 1 coil unit. The surface of this cast slab was examined by a CCD camera. The pinhole defects were evaluated by the number present per square meter cast slab surface area. Next, this cast slab was hot rolled and cold rolled by ordinary methods to finally obtain a coil of cold rolled steel sheet of a thickness of 0.8 mm and a width of 1600 mm. The steel sheet quality was visually inspected on the inspection line after cold rolling and was evaluated by the number of surface flaws occurring per coil. As a result, the cast slab had 15 pinhole defects/m2, while the steel sheet had 3 surface flaws/coil. INDUSTRIAL APPLICABILITY As explained above, according to the present invention, it is possible to prevent nozzle clogging and possible to produce an ultralow carbon cast slab not giving rise to surface flaws even if processed into thin-gauge steel sheet. Accordingly, the present invention has a high applicability in the steelmaking industry. " ^ We claim: 2 8 OCT 2013 1. A method of production of an ultralow carbon cast slab characterized by adding Ti to molten steel decarburized to a carbon concentration of 0.01 mass% or less, further adding at least one of Ce, La, and Nd, and using an immersion nozzle to inject the above molten steel from a tundish to a casting mold for continuous casting while maintaining a flow rate of Ar gas blown from any location in a range from a tundish upper nozzle to discharge ports of said immersion nozzle at 3N1 (normal liter)/min or less and said immersion nozzle has an orifice in its inner bore . 2. A method of production of an ultralow carbon cast slab as claimed in claim 1, wherein said inner bore has a circular cross-sectional shape and (i) a relation 35R-r<30 stands between a radius R of a top end of the inner bore [mm] and a smallest radius r of the orifice [mm] and (ii) a length L from a top end to a bottom end of the orifice [mm] is 50'^