DESMIDIACB^E. 279 



Reproduction both by conjugation and subdivision 

 variously modified, is common to all the families of 

 Desmidiacese ; and in the Zygnemaceae, which have a 

 close relation to them, the phenomena of conjugation are 

 very well known. In Staurocarpus we have those re- 

 markable quadrate spores formed in the cross branch, 

 produced by conjugation. In Spirogyra the union of two 

 cells belonging to the opposite filaments takes place by the 

 expansion of one side of each, so as to form a papilla or 

 short rounded-off tube (see fig. 145). The ends of the two 

 projections then come into contact, become slightly flattened, 

 then pressed together, and finally united. The double wall 

 formed by their union dissolves, or is broken through, so 

 that a free passage is established between the two cells. 

 Upon this, the whole of the chlorophyll, previously 

 arranged round the inside of each of tLe cells, becomes a 

 confused mass, which soon forms itself either in the cavity 

 of one of them, or in the connecting canal, into a globular 

 or oval spore invested with the colourless cellulose mem- 

 brane shown in one of our drawings of Penium (Fig. 155). 

 In Closterium conjugation takes place in a somewhat 

 similar manner, represented at No. 25, although it is 

 quite clear that if the formation of germs by conjugation 

 were the only provision for the reproduction of a species, 

 all must disappear, inasmuch as the conjugation and 

 consequent destruction of a pair of Closteria for the 

 formation of one new plant will ultimately destroy the 

 species. 1 Another mode, however, that of subdivision, 

 appears to be designed as an effectual safeguard against 

 such a possible extinction. Mr. Lobb has observed this 

 process take place in Micrasterias denticulata (Plate II. 

 fig. 30), in the course of three hours and a half. The small 

 hyaline hemisphere, put forth in the first instance from each 

 frustule, enlarges with the flowing in of the endochrome ; it 

 then undergoes progressive subdivision at its edges, first 

 into three lobes, then into five, then into seven, then into 

 thirteen, and finally at the time of its separation, acquires 

 the characteristic notched outline of its type, being only 

 distinguishable fron?. the older half by its smaller size. 2 



(1) In certain species of Closterium the act of conjugation gives origin to two 

 porangise. (2) E. G. Lobb, Trans. Micros. Soc. N.S. vol. i., 1861. 



