86 RALPH S. LILLIE. 



ditions influencing their formation, to the structures arising by 

 normal growth in organisms. The two types of phenomena 

 exhibit many close and, as it seems to me, highly significant 

 parallels. Inorganic structures resembling vegetative growths 

 are seen not only in the formation of lead or tin "trees" from 

 metallic zinc immersed in solutions of salts of these metals, but 

 they are shown in an especially striking form in those cases where 

 the local process of electrolysis gives rise to precipitates which 

 form coherent membranes or otherwise exhibit colloidal character. 

 In a recent paper I have described the structure and conditions 

 of formation of such precipitation-structures in some detail, and 

 have discussed the reasons for their resemblance to organic 

 growths. 1 In solutions of potassium ferricyanide, especially 

 those containing egg-albumin or other protective colloid, pieces 

 of metallic iron, zinc, or copper produce characteristic filamentous 

 or quasi-cellular outgrowths, consisting of precipitation-mem- 

 branes of the ferricyanide of the metal used, w T hich resemble 

 strongly certain definite organic types of grow r th like mould- 

 hyphse. These structures grow out into the solution from the 

 anodic regions of the metallic surface; hence their formation may 

 be accelerated, retarded, or suppressed at will by varying the 

 character of the local circuits determining the rate at which the 

 ions of the metal enter solution. A region of (e. g.), iron which 

 is actively forming precipitation-filaments will at once cease this 

 action if it is rendered the cathode in another intercepting local 

 circuit, e. g., by the contact of zinc or other baser metal at a 

 neighboring point; or conversely it may be rendered still more 

 active i. e., more strongly anodic by increasing the intensity 

 of the local circuit in which it acts as anode, e. g., by the adja- 

 cent contact of a nobler metal (e. g., platinum) or carbon. 2 It is 

 especially to be noted that the relations between the different 

 electrode-regions of such local circuits are reciprocal, as regards 

 the character of the chemical changes there taking place; this is 

 inevitable, since in general the electrochemical processes at any 

 anode are the reverse of those at the cathode. Hence the for- 

 mation of filaments at one region of the metallic surface appears 

 to have the effect of inhibiting their formation at another ad- 



1 BIOLOGICAL BULLETIN, 1917, Vol. 33, p. 135. 



2 Loc. ci,., p. 148. 



