The present invention relates to a composition which is advantageous for inerting harmful impurities, such as clays, in hydraulic compos, it ions. It is sometimes difficult to control the properties of hydraulic compositions in a consistent manner. The quality of the raw materials is often the cause of these variations. In particular, it has been found that sands, or more particularly the impurities which are contained in sands, such as clays, can bring about fluctuations in the properties of hydraulic compositions. In some cases, these fluctuations are due to a decrease in the efficacy of plasticisers of the polymer type having a comb structure, also referred to as superpLasticisers. According to document WO 98/58887, the absorption of these polymers by swelling clays of the 2:1 type present in sands is the cause of this decrease in efficacy. It is known to eliminate clays and other impurities from sands by means of washing in water. The dirty water is then processed using a flocculatlon agent which allows the liquid/solid separation to be accelerated and thus allows clean water to be recovered. This water is then recycled in order' to wash the sand again. However, this solution is costly in terms of equipment and requires a substantial supply of water. Furthermore, the washed sands generally contain residual flocculant which has negative effects on the properties of the hydraulic compositions. The application JP 9-221346 proposes neutralising the negative effect of the residual flocculant of the anionic polymer type by adding a cationic polymer. Another solution involves not separating the clays from the sands, but instead iherting the clays in the sands. Document WO 98/58687 proposes the use of agents which modify the activity of the clay, for example, by reducing the absorption capacity thereof or by carrying out a pre-absorption operation. The document proposes, inter alia, the use of organic or inorganic cations, including quaternary (poly)amines which may be alkoxylated. However, the agents described are still not entirely satisfactory in terms of efficiency. It is thus necessary to add a large measure of agents in order to process materials which have a significant content in terms of impurities. A large measure results in costs which render the method disadvantageous on an industrial scale. Furthermore, in contrast to the teaching of document WO 98/58887, different clays, of the type 1:1, for example, can also bring about a decrease in the quality of the hydraulic compositions. The object of the present invention is therefore to provide a composition useful for reducing the undesirable effects linked to the presence of harmful impurities, such as clays. In particular, such a composition is sought which is robust, that is to say, effective for different hydraulic compositions. Furthermore, a composition is sought which is stable during storage, inexpensive and whose use does not require sophisticated equipment. This object is achieved by a, composition comprising a specific eationic polymer. Preferably, the specific cationic polymer is associated with an anionic polymer. In this manner, the invention relates mainly to a composition useful for inerting clays in hydraulic compositions comprising at least 50% by weight of a catioiiic polymer which has a density of cationic charges which is greater than 0.5 me<3/9 s-hd an intrinsic viscosity which is less than 1 dl/g as an active substance. In the context of this description, the term "polymer" is intended to refer to a compound which comprises more than two monomeric units which may be identical or different and which may or may not have a specific order. The term "cationicity" is intended to refer to the density of positive charges which are carried by a compound. The cationicity can be measured by means of colloidal titration. The term "intrinsic viscosity" is intended to refer to the limit value of the reduced viscosity r\i / c with infinite dilution of the polymer. This value is linked to the mean molecular weight of a polymer. The term "hydraulic composition" is intended to define any composition which has hydraulic setting, and quite particularly mortars and concretes which are intended for all construction markets (building works, civil engineering or prefabrication sites). The terra "sand" is intended to define granulates having a mean gra'nulometric size of between 0 and 6 mm, preferably between 0 and 4 mm. They may be of any mineral, calciferous, siliceous or siliceous limestone type, or any other type. This definition also includes fillers or other specific inorganic substances which are likely to be present in hydraulic compositions. The term "clays" is intended to. refer to aluminium and/or magnesium silicates, in particular phyllosilicates having a lamellar structure, which are typically spaced with approximately from 7 to approximately 14 Angstr6ms. However, this term also refers to clays of other types, in particular amorphous clays. Clays which are commonly found in sands include in particular montmorillonite, illite, kaoliriite, muscovite and chlorite. Clays may be of the 2:1 type but also of the 1:1 type (kaolinite) or 2:1:1 type (chlorite). The term "active substance" is intended to refer to the constituents of the composition which have, an effect on the properties of the hydraulic composition, excluding constituents which form the carrier, in particular the solvents or dispersing agents, emulsifiers and other additives. The invention is based on the finding that eationic polymers which have a specific density of cationic charge and molecular weight, preferably in association with anionic polymers, are particularly effective for inerting the impurities in hydraulic compositions. According to the invention, the cationic polymer which is advantageous for optimising the inerting effect has & cationieity greater than 0.5 meg/g, preferably greater than meg/g, and in particular greater than 2 meg/g. According to the invention, the cationic polymer further has a molecular weight which is expressed by an intrinsic viscosity less than 1 dl/g, preferably leas than 0.8 dl/g, and in particular less than 0.6 dl/g. Advantageously, the composition comprises, at least 50% by weight, and in particular at least 70%, and guite particularly from 70 to 80% by weight of cationic polymer. The cationic polymer may have a linear, comb or branched structure. Preferably, the cationic polymer is linear. Cationic groups may be in particular phosphdnium, pyridinium, sulphonium and quaternary amine groups, the last ones being preferred. These cationic groups may be located in the chain of the polymer or as a pendant group. A large number of cationic polymers are known per se. Polymers of this type may be obtained directly using one of the known methods of polymerisation, such as radical polymerisation, polycondensation or polyaddition. They may also be prepared by means of post-synthetic modification of a polymer, for example, by means of grafting groups which carry one or moTe cationic functions to a polymer chain which carries appropriate reactive groups. The polymerisation is carried out starting with at least one monomer which carries a cationic group or an appropriate precursor thereof. Polymers obtained from monomers carrying amine and inline groups are particularly advantageous. Nitrogen can be quatcruised after polymerisation in known manner, for example, by means of alkylation using an alkylating compound, for example, with methyl chloride, or in an acid medium, by means of prbtonation. Cationic polymers containing cationic groups of quaternary amine are particularly appropriate. Monomers which already carry a cationic quaternary amine function may include in particular the salts of diallyldialkyl ammonium, quaternised (meth)acrylates of dialkylaminoalkyl and {meth)acrylamides N-substicuted by a quaternised dialkylaminoalkyl. The polymerisation can be carried out with non-ionic monomers, preferably having a short chain, comprising from 2 to 6 atoms of carbon. Anionic monomers may also be present so long as they do not affect the cationic groups. In the context of the modification of polymers by means of grafting, it is possible to mention grafted natural polymers, such as cationic starches. Advantageously, the cationic polymer contains groups whose cationic character is present only in an acid medium. Tertiary amine groups, which are cationic by means of protonation in an acid medium, are particularly preferred. The absence of ionic nature in hydraulic compositions of the concrete or mortar type having an alkaline pH further renders them robust with regard to other ionie, in particular anionic, compounds. By way of example, it is possible to mention cationic polymers from the polyvinylamine family which can be obtained by means of polymerisation of N-vinylformamide, followed by hydrolysis. Quaternised polyvinylamines may be prepared as described in the patent US 5,292,441. Polymers of the polyethyleneimine type are also suitable. The latter are quaternised by means of protonation. Particularly preferred are the cationic polymers obtained by means of polycondensation of epichlorohydrin with a mono or dialkylamine, in particular methylamine or dimethylamine. Their preparation is described, for example, in patents US 3,738,945 and US 3,725,3.12. The cationic polymer unit obtained by means of polycondensation of dimethylamine and epichlorohydrin can be represented as follows: ""Me Me HO cie Also appropriate are polymers of the polyacrylamide type modified by means of a Mannich reaction, such as polyacrylamide N- substituted, by a dimethylaminomethyl group. Also appropriate are cationic polymers which are obtained by means of polycondensation of dicyandiamide and formaldehyde These polymers and the method for obtaining them are described in the patent FR 1 042 084 . The invention thus also relates to a method for inerting clays in sands which are intended for the preparation of hydraulic compositions, wherein during the production of the sand, the sand is placed in contact with a polymer obtainable by condensation of dicyandiamide with formaldehyde, optionally in the presence of other compounds. According to a preferred embodimenti the polymer is obtainable by condensation of dicyandiamide with formaldehyde in the presence of: A) a polyalkylene glycol, and/or B) a polyalkoxylated polycarboxylate; and/or C) an ammonium derivative. The exact chemical constitution of the polymer obtained in this manner is not known precisely. It will therefore be described below substantially using the preparation method thereof. Method of preparation The polymer is obtainable by condensation of dicyandiamide with formaldehyde, optionally in the presence of other compounds, in particular a polyalkylene glycol (A) , a polyalkoxylated polycarboxylate (B) and/or a qu.aternisation agent (C>. The condensation reaction between dicyandiamide and formaldehyde requires 2 moles of formaldehyde for 1 mole of dicyandiamide, according to the following reaction diagram (Figure Removed) In this manner, the molar ratio between formaldehyde and dicyandiamide is preferably in the range of from 0.8:1 to 4:1, in particular from 1:1 to 3:1. A molar excess greater than 4 does not provide any additional advantage, but may lead to an undesirable setting of the reaction mixture. It ia particularly preferred to carry out the reaction with a Slight stdichiometric excess of formaldehyde, with a molar ratio between formaldehyde and dicyandiamide in the range of 2.2:1 to 2.8:1. Preferably, the polymer is obtained by means of condensation of formaldehyde with dicyandiamide in the presence of additional compounds. This allows the properties of the polymer to be adjusted, in particular the solubility thereof in water and the affinity thereof with respect to clays. The polyalkylene glycol (compound A) is preferably a compound having the formula (I): R2-0-[R1-0]n-R3 where: R1 is an alkyl group of Ci to C4, preferably an ethyl and/or propyl group; R2 and R3 are, independently of each other, a hydrogen atom or an alkyl group of :Ci to C4, preferably a methyl group,- and n is a number from 25 to 1000. By way of example, it may be polyethylene glycol, polypropylene glycol, a copolymer of ethylene oxide/propylene oxide or an admixture of these different compounds. Preferably, it is polyethylene glycol. The molecular weight of the compound A is preferably from 1000 to 35000. viscosity measurements have shown that the presence of the compound A modifies the structure of the polymer formed and the efficiency levels thereof. The quantity of compound A optionally used in the reaction is in principle leas than that of the main react ants dicyandiamide and formaldehyde. In this manner, the reaction admixture generally contains from 0 to 10%, preferably from 0.5 to 3, and quite particularly from 0.8. to 1% by weight of compound A. The polyalkoxylated polycarboxylate (compound B) is a comb polymer which comprises a main hydrocarbon chain to which both lateral carboxylic groups and alkoxylated groups are connected, in particular groups of propylene oxide (PO) , ethylene oxide groups (EG) and/or combinations thereof. The lateral groups may be ionic or non-ionic. It is preferably a compound having the following formula {ID : (Figure Removed) where: R1 and R2 are, independently of each other, a hydrogen atom or a methyl group; R3 and R4 are, independently of each other, an alkylene group of Ci to C