of the Desmidica; and Diatomca;. 3 



spore). No intcnnocliatc stages could l)e found between this 

 and the previously-described condition. J^Ixperinienta, in whieli 

 an attempt was made to obtain a comph'tion of the less-advanced 

 conjugation under the microscope, all failed. Apparently the 

 conjugation-cell is exceedingly sensitive to any external injury, 

 especially to contact with foreign bodies. Very probably the 

 contents, in the above-described cases, were already abnormally 

 altered, and incapable of further development. 



In other conjugation-cells the young spore displayed a still 

 thicker membrane, covered on the outside with truncate-conical 

 elevations, in which membrane could be detected a composition 

 of two colourless layers. The outer of these layers remained 

 clear and transparent even in the advance to maturity. Its 

 elevations became developed into rather long spines, which forked 

 at the apices into two or four branches*. The deeper-seated 

 layer of the spore-membrane meanwhUe assumed a dark -brown 

 colour. By rolling under the covering-glass, the tough, colour- 

 less, outer layer may be readily stripped from the inner, more 

 brittle, brown layer ; then the latter appears covered on its outer 

 surface with slight elevations, similar to those which first ap- 

 peared upon the young spore (PI. I. fig. 3). The brown layer 

 of the spore-coat encloses a third, delicate, colourless layer (per- 

 haps the ])rimary membrane of the spore) which immediately 

 envelopes the cell-contents. 



At the beginning of July, the green contents of all the spores 

 appeared conglobated into a spherical mass with sharp outlines, 

 which, lying free in the middle part of the cell, nowhere touched 

 its internal wall. Three weeks later, in many of the spores these 

 contents appeared separated into two flattened ellipsoidal masses 

 (PI. I. fig. 3) ; when I cracked the cell by careful pressure, I was 

 sometimes successful in driving out one or both of the masses of 

 contents in an uninjured condition. They could then be recog- 

 nized, beyond all doubt, as primordial cells ; bodies destitute of 

 a solid cell-membrane, having a thin coat of protoplasm which 

 ' bubbled' out in water, to which adhered a thick investment, 

 coloured bright green by numerous imbedded chlorophyll- 

 granules, surrounding a central cavity filled with transparent 

 tiuid. The fluid contained in the spore in which the two pri- 

 mordial cells were immersed, was not colourless, but rendered 

 turbid by numerous immeasurably small granules exhibiting 

 molecular motion. In August each of the ellipsoidal primordial 

 cells had divided into two globular cells, of similar character to 

 the mother-cell. Towards the end of September, some of the 

 spores exhibited another such division, so that they then con- 



* ' Spines finally branched.' llalfs, /. c. p. 98. The figiu-c represents a 

 spore not quite ripe, with unbranched spines. 



1* 



