McCollum and Davis appreciated the fact that the animal body is capable of producing synthetically the complex lipins, or fat-like substances such as lecithin, which in addition to glycerol and fatty acids contains phosphoric acid and a nitrogenous base called cholin. McCollum, Halpin and Drescher (6) had shown that young hens could grow for a considerable time on a diet free from this class of substances, and while confined to it, could produce many eggs in which were contained much more of the complex lipins than the bodies of the hens could furnish, they remaining in good health at the end of the experiment. Butter fat is free from lecithin and related substances, so their data led McCollum and Davis to the conclusion that the stimulating effect of butter fat on growth could not be attributed to any known lipin.

In 1912, Osborne and Mendel (7) published a paper describing successful nutrition over a considerable period of time with diets which contained no fats and but traces of substances soluble in fat solvents such as ether. This paper tended greatly to confuse the entire subject. They fed a diet consisting of purified protein, cane sugar, starch, and "protein-free milk," and in some of their experiments with similar diets "artificial protein-free milk" was used instead of the natural product. The artificial product was composed of milk sugar and a salt mixture from reagent bottles. It was made in close imitation of the mineral content which analysis showed to be present in the natural product. These diets, which contained no fats, were capable of inducing normal growth for a period of sixty days.

Commenting on these results Osborne and Mendel stated: "Employing the methods which were adopted in our earlier feeding experiments with isolated food substances, we have succeeded in inducing a normal rate of growth in white rats with dietaries devoid of fat throughout almost the entire period during which growth ordinarily continues." They stated that "McCollum has demonstrated that the phosphorus needed by an animal for phos-phatid formation can be drawn from inorganic phosphates, and that phosphatids can be synthesized anew in the animal body. Rohmann asserts the possibility of lecithin synthesis in mice which were maintained into the second generation on lecithin-free food. Our own experiments point in the same direction with regard to the lipoids in general, and they give positive evidence of the dispensableness of true fats for growth."