STRUCTURES RESEMBLING ORGANIC GROWTHS. 143 



larger filaments are typically hollow or tubular; it is only the 

 shorter and slenderer structures that appear to be solid, and even 

 here the case is doubtful, since filaments that appear solid on 

 ordinary examination often show a distinct double contour when 

 examined more closely. The probability is that all filaments are 

 hollow, at least when first formed, and that each consists essen- 

 tially of a solution containing a salt of the metal and enclosed 

 by a semi-permeable precipitation-membrane. The mode of 

 formation about to be described indicates this clearly. The 

 appearance of the mass of filaments covering a piece of metal 

 that has lain for some time in the solution is strikingly vegetative 

 or fungoid in character. The special character or "habit" of 

 the growth varies with the different metals; in the case of iron 

 and copper the filaments are usually more uniform in appearance 

 and have a smoother contour and straighter course than in the 

 case of zinc; with the latter metal the growths are more irregular 

 and varied and have a characteristic granular appearance due 

 to adhering coarser particles of precipitate. The appearance 

 of such growths is often remarkably hypha-like in character; 

 thus a thin strip of zinc about two centimeters long with a 

 copper wire bound about one end, placed in a watch-glass con- 

 taining a 2 per cent, solution of K 3 FeCy 6 in dilute egg-white, 

 developed in the course of eighteen hours a most complicated 

 feltwork of winding and interlacing hypha-like filaments, covering 

 the whole strip from end to end. Numerous tubular filaments 

 extended to the surface of the solution and along the surface, in 

 some cases for three or four centimeters; many such surface- 

 filaments had undergone curious modifications suggesting efflores- 

 cences or spore-capsules, e. g., becoming enlarged at their termina- 

 tions to form stalked vesicular or bladder-like formations with 

 thin walls. Such a mass of modified filaments irresistibly sug- 

 gests a vegetative growth. 



Study of the mode of formation of these tubular filaments 

 under the microscope reveals a somewhat unexpected complexity 

 of conditions. Elongation is not by basal growth, i. e., by 

 deposition of precipitate at the surface of the metal where the 

 metallic ions enter solution, but is always terminal, the tube 

 advancing through the solution by the deposition of precipitate 



