FRESH-WATER ALG.K OF Till: f M T K D STATES. 26 



but increase of tin- dot - -. rtainly occur by means of these so-called spores. 



The growth of tin; plant takes place in tin- same way as in tin- true nostocs. 

 The filaments increase in length by tran>\i rse division and consequent multiplica- 

 tion of the cells, whil>t new tilaincnts an- formed by the consentaneous longitudinal 

 division of all the cells of a filament. 



The spores of a Cylidroj>ennutn have the power of germinating after prolonged 

 desiccation, they having been successfully cultivated even from specimens long 

 preser\ed in the herbarium. Their development lias b.-rn carefully and success- 

 fully studied by M. Thuret. According to this authority the first change consists 

 in an elongation of the spore, which ruptures the wall of the sporangium, pushing 

 a portion of it before it. Directly after this the spore undergoes division, so that 

 out of it is formed a little torulose filament, composed of four or five cells. Growth 

 takes place at both ends, but more rapidly at the free one. The new cells formed 

 are smaller than those which arise directly from the spore, but, finally, all the arti- 

 <!.-< assimilate. The wall of the sporangium remains attached for a long time to 

 the end of the filament forming a little cap to it. The heterocysts, according to 

 Thuret, at lirst are indistinguishable from the ordinary cells, but after awhile the 

 grannies in them begin to disappear, the color to pale, the outer wall to become 

 apparent and grow thicker, until at last a perfect "connecting cell" is educed. 

 I have, myself, carefully watched the early development of the spores of a cylin- 

 (Irospermum, and can confirm, in all essential particulars, the description Thuret 

 has yiven of the process. Fig. 10, pi. 2, represents a partially formed filament, 

 to which the empty sporangium is still attached. 



As no sexual reproduction has as yet been shown to exist among the XostocJia- 

 it is very e\ident that their whole life-history is not comprised within the 

 changes which have been detailed. It has long been known that the gonidia of 

 many lichens have the power of independent existence, i.e. that when they are 

 discharged from their thallus they can continue to live and multiply, if circum- 

 stances favor them, without giving origin to a new thallus. This, and the great 

 similarity of structure between the nostocs and the lichen genus Collema, has 

 suggested a possibly close relation between the two. The first observer, I believe, 

 who asserted that they were different stages of the same plant was Dr. Hermann 

 It/igsohn. 



His observations are, however, rendered of so little value by his own statements 

 that it is not necessary to review them here. Thus, he says, that after seven years' 

 rvation he had yet to see a true one called algtc, that the Desmidia; are, at 

 least, two-celled, &c. &c. The most weighty observations upon this subject are 

 those of Professor Julius Sachs and of J. Baranetzky the former published in the 

 Botanische Zeituny for 18.55, the other in the Bulletin of the St. Petersburg 

 Academy for 1867. 



Professor Sachs states that he watched a whole bed of Noafoc commune deve- 

 loping into Collema bulbosum. He says that the peculiar Collemoid threads first 

 appeared as little lateral offshoots or prolongations from the cells of the nostoc 

 filament, and rapidly developed into well-formed collemoid filaments. Every 

 possible stage from the typical nostoc to the typical collema was seen repeatedly. 



4 rbrury, 1872. 



