Kastle and Loevenhart demonstrated the synthesis of a trace of ethylbutyrate by lipase if the latter enzyme was added to the products of the hydrolysis of ethylbutyrate, ethyl alcohol, and butyric acid by the same enzyme.[19] Taylor[20] obtained the synthesis of a slight amount of triolein by the addition of the dried fat-free residue of the castor bean to a mixture of oleinic acid and glycerine.... No synthesis occurred with acetic, butyric, palmitic, and stearic acids with glycerine, mannite, and dulcite, and the experiments with the last two alcohols and oleinic acid likewise yielded no synthesis. [19] Kastle, J. H., and Loevenhart, A. S., _Am. Chem. Jour._, 1900, xxiv., 491. [20] Taylor, A. E., _Univ. Cal. Pub._, 1904, _Pathology_, i., 33; _Jour. Biol. Chem._, 1906, ii., 87. This suggests possibly a specific action of the enzyme. If this slight reversible action had any biological significance (which might be possible, since in the organism secondary favourable conditions might be at work which are lacking _in vitro_) there should be a parallelism between masses of lipase in different kinds of tissues and fat synthesis. Loevenhart indicated that this might be a fact, but a more extensive investigation by H. C. Bradley has made this very dubious.[21] [21] Bradley, H. C., _Jour. Biol. Chem._, 1913, xiii., 407. Very little is known concerning the reversible action of the hydrolytic protein enzymes. A. E. Taylor digested protamine sulphate with trypsin and found that after adding trypsin to the products of digestion a precipitate was formed after long standing; and we may also refer to experiments of Robertson with pepsin on the products of caseinogen to which we shall return in the next chapter. It therefore looks at present as if van’t Hoff’s idea of reversible enzyme action might hold in the modification offered by Armstrong. It remains doubtful, however, whether this reversibility can explain all the synthetic processes in the cell. No objection can be offered at present if any one makes the assumption that each cell has specific synthetic enzymes or some other synthetic mechanisms which are still unknown. The mechanisms for the synthesis of proteins must have one other peculiarity: they must be specific in their action. We shall see in the next chapter that each species seems to possess one or more proteins not found in any other but closely related species. Each organism develops from a tiny microscopic germ and grows by synthetizing the non-specific building stones (amino acids) into the specific proteins of the species. This must be the work of the yet unknown synthetic enzymes or mechanisms. The elucidation of their character would seem one of the main problems of biology. Needless to say crystallography is not confronted with problems of such a nature.