1889.J On Isoetes lacustris, Linn. 307 



embedded in paraffin. A number of sections through each spore were 

 obtained by means of the Cambridge rocking microtome, and were 

 arranged in series, thus permitting of an examination of the internal 

 structure of the spores. The protoplasm, which is remarkably granu- 

 lar, is of a spongy texture (probably due to the extraction of the oil), 

 and contains a nucleus of very large size, in which bodies resembling 

 nucleoli were in some cases detected. The nucleus is sharply marked 

 off from the cytoplasm by a membrane, but of course it must be borne 

 in mind that this feature may be caused in part by the methods used 

 in embedding. When spores are examined in this stage, the proto- 

 plasm stains but slightly with heematoxylin, and the tint is inclined 

 to red, and even the nucleus is not deeply coloured. In somewhat 

 older spores, at the period immediately preceding germination, the 

 whole protoplasm stains far more readily and deeply in a given time, 

 but a nucleus is no longer differentiated by the haematoxylin, and the 

 colour now produced is of a deep blue. As I have frequently had 

 spores of different ages on the same slide, all of which were subjected 

 to exactly similar treatment, this difference in colour may probably be 

 taken to indicate an actual diffusion of the substance of the nucleus 

 through the cytoplasm, since the change is always confined to spores 

 in the condition referred to. 



This view receives some confirmation from the circumstances attend- 

 ing the formation of the prothallium, now to be described. The first 

 indication of cell-division occurs in a somewhat peculiar manner, but 

 its significance is rendered clear by what takes place subsequently. 

 Before entering upon a description of what actually happens, it may 

 be well, in order to avoid possible misconstruction, to state expressly 

 the opinion that the characters presented are made visible only by 

 the action of the means necessarily employed in embedding, but this 

 does not vitiate the conclusion that they may be taken as indications 

 of internal changes which actually occur in the protoplasm. In spores 

 in which cell formation is about to commence, the deeply stained 

 protoplasm is seen to be traversed by a few "cracks," which divide 

 the contents of the spore into large isolated masses. At this period 

 there is nothing to point to the existence of a membrane, except 

 the granular structure which is apparent on the surface of the 

 cracks, but at a subsequent stage in the development, one of the sur- 

 faces is seen to be bounded by a membrane of extreme tenuity. When 

 first formed it can only be distinguished in favourable places, but it 

 rapidly grows in thickness, and forms a limifing surface between the 

 two protoplasmic masses. From the mode of its formation it can 

 hardly arise otherwise than by the conversion of a layer already 

 present in the protoplasm directly into cellulose, and it appears to be 

 the presence of this substance arranged in a definite plate-like manner 

 which determines the splitting referred to. The first membrane cuts 



