492 Mr. C. E. Walker. [Aug. 2, 



joined together, forming in places a thick, deeply-staining thread, the axis 

 of which is continuous with the axis of the strings of separate granules 

 (fig. 3). In some such cells the separate granules and the thick thread 

 appear to be about equal in amount and continuous with each other (fig. 4), 

 while in others there are but very few granules, and the thread pre- 

 ponderates. From these it is possible to pass, by almost insensible 

 gradations, to cells where there are no granules, but only a thick thread 

 coiled up around the nucleus (fig. 6). In some of these cells it is practically 

 impossible to distinguish the nucleus, so dense and darkly-staining is the 

 coiled thread (fig. 5). So misleading are these figures that for several months 

 I took them for, and drew them as, the spireme stages in the prophase of 

 division in which the nuclear membrane had disappeared. It was only when 

 I found nuclei in a resting condition existing in the cells which also contained 

 this thread, that the present interpretation was arrived at. A careful 

 examination of the cells in this stage, however, will frequently, though not 

 always, demonstrate the presence of a resting nucleus. Where the nucleus 

 cannot be seen, the reason is generally very obvious : the thread is so deeply 

 stained with the basic stain, and is so closely coiled, that it entirely obscures 

 the similarly stained nucleus. Moreover, while the thread stains very deeply, 

 the nucleus, apparently, stains less readily than usual, thus adding to the 

 probability of its being overlooked. 



From this stage it is possible to pass again by practically insensible 

 gradations to cells where the coiled-up thread occupies a space slightly 

 larger than (fig. 7), equal to, or smaller than the nucleus, until we arrive 

 at some where it seems to be about the same size in proportion to the 

 nucleus as is the archoplasm in the case of the spermatid. 



Though it has not been possible as yet to trace the origin of this thread 

 farther, it is strongly suggested that it arises in the archoplasm, which is 

 often seen to be connected with it (fig. 9). 



During the whole of its existence the thread stains very deeply, and always 

 with a basic in preference to an acid stain. It seems to show best with 

 safrranin and basic fuchsin. When it has entirely broken up, the granules 

 formed from it still stain in the same manner ; but as they begin to lose their 

 regular arrangement, so they seem to lose their affinity for the basic stain. 

 It appears, however, that this happens but slowly in all such cells, and 

 probably in many the granules never stain with the acid dyes. In bone- 

 marrow large areas are often encountered where over 30 per cent, of the 

 cells contain granules staining with a basic dye. In other parts of the body 

 the cells containing such granules are not present in such large proportions, 

 and we meet with more in which the granules take the acid dye. 



