100 DUBLIN UNITEESITT 



8tricted form of the cells ; while it does not appear improbable that other 

 Algae not Desmidian may have a like mode of cell-increase. 



The mode of true reproduction in the Desmidiaceae is by conjugation, 

 very like the same process in the Zygnemaceous group of Confervoids. 

 This process consists essentially in the pouring out and amalgamating 

 of the contents of two distinct cells, resulting in a definitely formed 

 spore or sporangium ; and in the different genera of Desmidiaceae pre- 

 sents some slight modifications. In the filamentous forms, such as Hy- 

 alotheca, Didymoprium, &c, conjugation does not take place till the 

 constituent joints of the filament become disunited, apparently any two 

 of which meeting in the water, couple and blend together their contents, 

 the old cell- wall of each pair of conjugated joints remaining, as it were, 

 fused together, and surrounding the sporangium. In Closterium, Penium, 

 Cosmarium, Staurastrum, &c, the free cells conjugate in pairs. In them 

 the process differs somewhat from that which takes place in the Zygne- 

 maceae. In that group, the cells of two filaments lying side-by- side, 

 bud-out, as it were, pouch-like protrusions towards each other, which, 

 meeting half-way, become inosculated, forming ladder-like transverse 

 tubes between the connected filaments ; the contents of each conjugated 

 cell of one side either pass over by the transverse tube into the cavity 

 of the opposite cell, or the contents of each meet half-way in the centre 

 of the connecting tube, and in either case become ultimately massed to- 

 gether into a rounded compact spore. In the most of the Desmidiaceae 

 there is no transverse tube formed ; but the fronds about to conjugate 

 approach and dehisce by the transverse suture, and each pours out its 

 contents in one mass into the water, which, gradually becoming conso- 

 lidated with the contents of the other conjugating individual, assumes a 

 membrane, and becomes a perfect sporangium, surrounded by gelatine, 

 and destined for reproduction. The sporangium, which is usually sphe- 

 rical (sometimes quadrate) either remains smooth or ultimately by de- 

 grees produces, in every direction from its surface, variously formed 

 spines or forked processes, and bears, in this state, no resemblance to 

 the parent species. Plate XI., Fig. 13, is an example. Any observations 

 that have been made in regard to the subsequent development of the 

 sporangium go to prove that its contents ultimately give birth, by binary 

 segmentation into a definite number of portions, to forms which at first 

 are very minute, but when set free by the solution of the wall of the spo- 

 rangium, afterwards attain the size of the parent species. We have, in an 



