130 THE PROTOZOA 



colony never varies, and the number of individuals is constant. 

 (In Gonium sociale there are 4 individuals ; in G. pectorale 16, while in 

 Eudorina there are from 16 to 32, and in Pandorina 32.) A Gonium 

 colony lies in one plane, but this arrangement is brought about by a 

 secondary shifting of the cells (Fig. 73), while a Eudorina colony 

 retains the spherical form. Pandorina, a similar compact and defi- 

 nite colony, is derived as in Eudorina, by the regular cleavage of a 

 single cell. 



In some colonies the individuals are connected in the centre by 

 protoplasmic strands, as in Synura, while in one genus (Uroglena) 

 connecting strands may or may not be present. Ehrenberg ('38) de- 

 scribed U. volvox as a colony form whose peripheral individuals are 

 connected in the centre by tail-like processes which, except for a much 

 greater length, are similar to those of Synitra. He was confirmed in 

 this by Zacharias ('95) and Kent ('81). Biitschli, however, regarded 

 this central attachment as extremely doubtful. In one form, U. vol- 

 vox, this connection does actually exist, but in another, U. americana, 

 the posterior ends of the cells are rounded and have no trace of a 

 central filament. The genus Uroglena may afford, therefore, a clue 

 to the phylogenetic relations of the relatively huge gelatinous colonies 

 which, save for the surrounding matrix, have no means of connection. 1 

 Proterospongia, in its general form and structure, agrees with U. 

 americana. In both cases, as far as known, there is an indefinite 

 number of individuals and no typical method of increase, as in Pan- 

 dorina, Eudorina, and Gonium. 



The most highly differentiated colonial forms are the genera Volvox 

 and Magosphcera, which should perhaps be considered simple multi- 

 cellular forms rather than Protozoa. 



In Volvox the monads (often as many as 12,000 in a single colony) 

 are arranged as in Uroglena, around the periphery of a gelatinous 

 mass, and no organized connections with the centre of the cell can be 

 traced, although they are connected with one another by definite 

 protoplasmic strands. 



In Magosphcera the individuals are connected not only by the jelly 

 matrix, but also, as in Synura, by protoplasmic stalks, and they are 

 in close contact at trie periphery. In both Magosphcera and Volvox 

 the appearance of the peripheral cells is strikingly similar to a pave- 

 ment epithelium, and the comparison which is so often made between 

 \ such colonies and the blastula stage in the development of Metazoa 

 is certainly justifiable. 



Stalked colonies have an entirely different mode of origin, being 

 formed by repeated longitudinal division, the daughter-cells remain- 



!Cf. Calkins ('91). 



