238 RALPH S. LILLIE AND EARL N. JOHNSTON. 



But a detailed knowledge of the constitution of these proteins, 

 i. e., of the chemical nature and serial (or other) arrangement of 

 the amino-acids united to form the protein molecule, would not 

 at present if ever enable us to predict the course of develop- 

 ment of either egg. 



We may assume, however, that the essential reason for such a 

 ailure would be that the conditions in the living system are too 

 complex (for historical and other reasons) to be analyzable in 

 the requisite detail, and not that a sufficiently complete knowledge 

 of the initial state of the system would never enable us to predict 

 the general course and outcome of the developmental transfor- 

 mation. Under these circumstances the study of some relatively 

 simple inorganic process, which in certain respects may serve 

 as a model for the organic formative process, may throw light 

 upon the general features of such a problem and indicate the 

 type of solution which it is practicable to seek. The purely 

 scientific aim is not to follow the developmental sequence through 

 all of its stages and account causally for each separate trans- 

 formation, but rather to determine what the essential and con- 

 stantly operative factors are in all cases of development, i. e., 

 the general conditions which make possible a specific type of 

 development in any particular case. Each case presents its 

 own special problems; but there are factors common to all cases 

 of development, and any general theory of development requires 

 that the nature of these should first of all be clearly indicated. 



In the inorganic growth-structures described in the present 

 and preceding papers 1 certain definite relations can be recog- 

 nized between the type of structure formed and the chemical 

 and physical character of the precipitate which is deposited to 

 form that structure. Thus filaments of zinc ferricyanide are 

 coarser and less coherent, as well as more variable and irregular 

 in form, than filaments of iron ferricyanide formed in the same 

 solution (see Figs. 1-5, 12-15). The metal may be compared 

 with the germ; its specific chemical nature determines the special 

 type of structure to which it gives rise in a given solution of 

 potassium ferricyanide i. e., under definite external conditions; 

 the reason for this is that each insoluble ferricyanide has its 



1 R. S. Lillie, BIOL. BULL., 1917, Vol. 33, p. 135. 



