112 THE VITAMINES 



have a tendency to scatter their food. This difficulty may be over- 

 come if we use such a cage as has been described explicitly by our 

 former co-worker, A. B. Macallum (311). The cage is equipped 

 with funnels of various sizes through which the rat may reach its 

 food without scattering it. The cage is arranged for two rats and 

 permits of separating the urine from the feces, so that an exact metabo- 

 lism experiment may be carried out. 9 Great care is necessary in the 

 preparation of the diet; above all. it is important that the basal 

 diet is vitamine-free. This precaution was not sufficiently observed 

 in the older investigations, thus giving rise to differences of opinion. 

 We need only quote one example, lactose, widely used as a constituent 

 of an artificial diet. Regarding this, we showed 10 that lactose, which 

 does not undergo any particular purification process, contains traces 

 of nitrogen. These traces may come from the milk and permit of 

 the growth of rats, because of its vitamine content. This was 

 demonstrated experimentally by McCollum and Davis (I.e. 97) and 

 also by Drummond (313). This is likewise true of the protein of 

 protein-free milk, as we have already seen. It is necessary to subject 

 the protein to be used to a purification process before it is fed. In 

 most feeding experiments casein is used, and it is rendered vitamine- 

 free by extracting with alcohol a number of times, according to the 

 procedure of Funk and Macallum (I.e. 81); we found that this 

 procedure did not decrease the nutritve value of casein, in spite of 

 statements by McCollum and Davis (I.e. 97) to the contrary. It is, 

 of course, possible to purify the casein by dissolving it in alkali and 

 precipitating with acid, and also by continued washing with water. 

 In using meat, according to Osborne, Wakeman and Ferry (I.e. 102), 

 it suffices to extract it repeatedly with water. From edestin, as we 

 have seen, this procedure removes the vitamine only with great 

 difficulty. Of the other commonly used dietary constituents, cane 

 sugar is vitamine-free, as are also most of the varieties of starch. 

 Regarding fat, we have used ordinary lard and under certain con- 

 ditions it can be entirely dispensed with. The purification of lard is 

 made in our laboratory by autoclaving at 30 pounds for 3 hours; we 

 shall have occasion to show later that this lard is then vitamine-free. 

 For the entire dietary mixture, about 3 per cent of agar is added to 

 diminish the possibility of constipation. If it is desired to use a sub- 



' Ferry (312) describes the technique of rat feeding experiments. 

 10 The Vitamines, 1st ed., p. 159. 



