116 GEIS^ERAL HISTORY OF THE INFUSORIA. 



According to Prof. Henfrey, the primordial cells or gonidia of Pandoinna 

 (XIX. 59-63), and also, in the opinion of many, the Euglence (XVIII. 45-48), 

 are similarly undefended. 



The internal globular coloured body of the motile form of Protococcus is in 

 the same state. Thus Cohn {R. S. p. 531) points out that although this 

 body has a sharply defined outline, yet, " either by mechanical means, or by 

 chemical reagents, the internal globular mass may suddenly be made to lose 

 its contour, and to sjjread so as entirely to fill the ca\ity of the colourless 

 envelope. From which it would appear that the internal globular body is 

 not surrounded by any special cellulose membrane, but only by one readily 

 destroyed by chemical or physical agency — probably nothing more than a 

 dense layer of protoplasm." 



In the case of Volvox the cells originate without an enclosing membrane ; 

 but after the appearance of the red spot, a dehcate one shows itself, and 

 extends at difierent points into the connecting thread-like processes (XX. 37, 

 39, 41). So in Gon'ium we may presimie the jDrimordial cells to be originally 

 naked, although Cohn has not remarked this fact, but confined himself to 

 describing the matui^e ceUs (XIX. 32, 34), which have an enclosing wall of 

 cellulose (Enhv. pp. 175, 1 76). Lastly, in the ' still' form of Protococcus a special 

 membrane invests the protoplasmic gonidium. In Goniiini (XIX. 34), and in 

 Volvox (XX. 37, 39, 40), filiform prolongations extend between the several 

 cells in the compound organism ; in Stephanospluera similar processes are 

 given ofi" at the opposite poles of the cells, and are consequently not inter- 

 current (XIX. 39). Prof. Williamson has in the case of Volvox offered the 

 best explanation of these threads, wliich have by some been supposed inter- 

 communicating canals. He first makes good his opinion that the green cell- 

 Kke organism represents the nucleus of a cell, the wall of which is separated 

 from it by a greater or less space ; and then he compares the processes in 

 question with the fihform extensions fi'om the nucleus which are met with 

 in many vegetable cells, suspending that organ in the centre. In the early 

 stage of the cell, the protoplasmic substance fills up more or less completely 

 the cell- wall (XX. 42, 44) : by-and-by the latter becomes outstretched from 

 it by a sort of di'opsical efiusion within it (XX. 37) ; but as the protoplasmic 

 nucleus has contracted adhesions at difii'erent parts, it becomes di'awn out from 

 the adherent points into thi^ead-hke processes (XX. 39, 40, 45), which grow 

 more and more filiform in proportion as the cell- wall expands. This expla- 

 nation (agreeing in every particular with the observed phenomena of cell- 

 growth) being accepted, it follows that these elongations are bounded by the 

 particular cell- wall to which they belong, and are not continuous with those 

 of adjoining cells. The processes of Volvox are therefore ofi'-shoots of the 

 protoplasm of which each cell or gonidium consists ; they are given ofi" before 

 any enclosing wall or peUicle appears, and wliilst that substance is still duc- 

 tile, and they disappear on the commencement of the process of development, 

 whether of macrogonidia or of microgonidia, and whether mth or without the 

 process of encysting. 



In the case of Gonium, Cohn gives {Entw. p. 176) a different account of 

 the connecting bands. It will be remembered that in this genus that ob- 

 server indicates an enclosing cellulose membrane to each cell or gonidimn. 

 Now this cell does not closely invest the protoplasmic substance at all points, 

 but is so separated as to produce a hexagonal cell-wall aroimd it, from each 

 angle of wliich the membrane is produced in a tubular form, and joins mth 

 a similar process coming from the angle of an adjoining cell (XIX. 32, 34). 

 Hence each process of the membrane has a double outline, and is in fact a 

 tube, only that its interior must be presumed to be shut-off from that vdi\\ 



