PRECIPITATION STRUCTURES SIMULATING ORGANIC GROWTH. 247 



scribed in a previous paper by the senior author. 1 In the present 

 paper description of methods will be confined to certain modi- 

 fications which have been found to be more suitable to the pro- 

 duction of a certain definite type of structure. Certain con- 

 ditions determining the type of structure produced have been 

 studied in fuller detail. Expecially interesting is the fact that 

 the process of formation is often periodic and accompanied by 

 rhythmic motion of the structures formed. Regularly segmented 

 or striated formations are thus often produced. Many interesting 

 biological comparisons are therefore suggested. 2 



II. Methods. 



If a piece of fine iron wire, wound about one end with a fine 

 copper wire, be dropped into a 2 per cent, egg-albumin solution 

 containing 4 per cent. K 3 FeCy 6 and 4 per cent, or more of NaCl, 

 the entire surface of the iron wire rapidly becomes covered with 

 fine filamentous growths (Fig. i). They are characteristically 

 regular in form; the majority are straight or slightly curved and 

 cease growth at a length of 200 microns or less. Very few, if 

 any, of the larger striated or branching structures are formed 

 under these conditions. A repetition of this experiment using 

 a less concentrated solution of the salts (2 per cent, egg albumin 

 containing 2 per cent. K 3 FeCy 6 and 0.5 per cent. NaCl) gives a 

 different result. The rate of growth is decreased, the form of 

 the filaments is more irregular and many larger structures are 

 formed (Fig. 2). In the stronger solution growth usually ceases 

 in about five minutes, while in the weaker solution it may con- 

 tinue for several hours. 



One method of controlling the rate of growth is to start the 

 reaction with a quantity of the solution just sufficient to cover 

 the surface of the metal and then to add from time to time with 

 a pipette a few drops of the fresh solution. All of the available 

 K.sFeCy 6 is soon transformed and the reaction ceases until 

 renewed by adding more solution. 



It is sometimes an advantage for the study of single filaments 

 or other structures to limit the number formed from the metal. 



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

 - Cf. Lillie, loc. cit., pp. 157-172. 



