PRECIPITATION STRUCTURES SIMULATING ORGANIC GROWTH. 249 



very smallest of these structures are tubular. The walls of these 

 filaments are precipitation-membranes; in the interior of the 

 filament is a dissolved salt of the metal, which upon coming in 

 contact with the ferricyanide solution outside forms the pre- 

 cipitate upon which the growth of the structure depends. A 

 detailed description of this process as well as a very probable 

 explanation of the cause of the outflow of the solution has been 

 given by R. Lillie. 1 



The phenomenon of branching, although infrequent in iron 

 filaments, occurs under certain conditions (Fig. 6). No branch- 

 ing has-been observed except with the larger filaments and as a 

 rule only with those which run along the surface of the solution. 

 Some few cases, however, have been observed where the entire 

 structure was submerged during its formation. 



Slowness of growth is an essential condition for branching. 

 Large filaments like those shown in Fig. 6 grow at the rate of 30 

 to 60 microns per minute, while the small filaments shown in 

 the same photograph grow at the rate of 120 to 300 microns per 

 minute. It required about 2 hours for the large branching 

 structure shown in Fig. 6 to form. It is approximately 3 mm. 

 long. The branches are always formed within a short distance 

 of the growing end of the filament. It is probable that they are 

 given off at thin or weak places in the walls where rupture and 

 outflow of solution occur. 



In general it has been found that the slower the rate of growth 

 -that is, the nearer it approaches that of organic growth the 

 more resemblance there is to an organic structure. There is no 

 doubt that the rate of the local chemical reaction, i. e., the rate 

 of formation of the precipitate, is an essential factor in deter- 

 mining the final form of the structure. In every case where 

 there was a decided change in the rate of growth, there was a 

 modification in form. The subject of the relation of rates of 

 growth to form has been studied extensively in living organisms. 

 D'Arcy Thompson, 2 in his book "Growth and Form," maintains 

 that the final form which an organism assumes is entirely depen- 

 dent on rates of growth. He says, "the phenomenon of rate 



1 Lillie, loo. cii., pp. 144-146. 



2 D'Arcy Thompson, "Growth and Form," University Press, Cambridge, 1917. 



