60 THE VITAMINS 



passed into solution. The precipitate, freed from barium, was active in 

 amounts furnishing about 0.00017 gram of nitrogen per day. 



In testing the activity of the various fractions, rats about six weeks 

 old which had been on a vitamin B-free diet (McCollum) for three 

 weeks and were losing weight were given definite quantities, as deter- 

 mined by nitrogen content, of the solution to be tested, carefully neu- 

 tralized and preserved in 30 per cent alcohol. After 3 or 4 days the 

 rats were again weighed and the strength of the solution being tested 

 was calculated from the weight gains by a method which is described 

 as follows : "From 'normal growth' curves of several rats the average 

 increase in weight during 3 days for a body weight of 40 to 70 grams 

 is found (5 to 7 grams). Considering this increase normal, the amount 

 of the substance (measured by the quantity of nitrogen) was deter- 

 mined which would be required to give a normal increase in weight in 

 the test animals during the feeding period. It is realized that the adopted 

 unit may be insufficient to bring about a normal course of the entire 

 period of growth up to maturity. However, it offers a rapid and reliable 

 method for standardizing the relative strength of various products." 



Levene and van der Hoeven called attention to Seidell's crystal- 

 line picrate (described in the preceding section), with the comment 

 that : "The standards of testing employed by Seidell and by us are so 

 different that it is difficult to compare the results of the two investiga- 

 tions. It seems unlikely that the picrate of Seidell and our concentrate 

 have potencies of different orders of magnitude. Yet we are certain 

 that our material is not an individual substance, but a complex mixture. 

 The pure vitamin, if ever isolated, will prove many times more potent 

 than the picrate of Seidell or than our present concentrate." 



In their next paper Levene and van der Hoeven (1925, 1925a) 

 reported a further concentration of the Osborne- Wakeman fraction 

 involving a preliminary precipitation with basic lead acetate, with re- 

 moval of the lead by sulfuric acid instead of hydrogen sulfide as had 

 generally been the custom. By this step alone the potency of the product 

 was increased three or four times. Precipitation of this material by 

 barium hydroxide as in the original method yielded a product approxi- 

 mately as active as the one previously prepared by the silica gel adsorp- 

 tion process. Quantities of from 2 to 4 milligrams, containing from 

 0.1 to 0.2 milligram of nitrogen (5.2 per cent on an ash-free basis) 

 sufficed in the usual feeding tests. Adsorption of the material by silica 

 gel and subsequent extraction twice with alcohol at pH 3 yielded 

 products of only moderate potency. A third extraction with alcohol at 

 pH 9 gave the most potent material thus far prepared. This was active 



