REPRODUCTION OF THE ZYGNEMACE®. 437 
number of conspicuous starch-grains are observed in the proto- 
plasm, usually from 4 to 6. In the same filament the plates of 
endochrome in the various cells appear always all to lie in one 
plane; and the conjugating processes are put out from the 
side of the cell in contact with the axile plate of protoplasm. 
Secondly, the zygospore is not formed in one of the conjugating 
cells, but in the conjugating tube, into which the endochrome 
passes from both the conjugating cells (see figs. 6 & 7). 
Two other points of difference should also be noted. As 
soon as the zygospore is fully formed, it is always sharply 
marked off from the rest of the conjugating tube, or from the 
mother-cell, by a cell-wall on either side (figs. 6, 7). "This is well 
shown by Cleve, Z. e. pl. ix. figs. 8, 9. Cooke (7. c. pls. xli. & xlii.) 
represents this septum in the case of M. recurvus, but not in those 
of M.scalaris or M. parvulus. The other point, so far as Iam aware, 
has not been noticed by previous observers, but it may have an 
important physiological significance. In Spirogyra and Zygnema 
the whole of the protoplasm of the conjugating cells is used up 
in the production of the zygospore, and they are left com- 
pletely empty. In AMesocarpus, on the contrary (fig. 7), a por- 
tion of the endochrome appears to be always left behind in both 
the conjugating cells. 
p. 
Mesocarpus scalaris. Conjugating filaments, showing relative size of male 
(upper) and female (lower) cells. 100. 
Although the probability of a differentiation between the two 
conjugating cells appears at first sight to be much less in the case 
of Mesocarpus than of the two genera already described, yet a 
careful observation shows that here also it is not impossible. My 
observations were made on JM. scalaris, Hass., which I found in 
conjugation, though rather sparingly, in September, forming dense 
light-green mats on the surface of moor-ditches and pools in 
Yorkshire. The first point that struck me was that the zygospore 
does not usually, although so represented by Cooke (l. c. pl. xlii.) 
