INVERTEBRATA, CUYPTOQAMIA, MICROSCOPY, E'JO. 51 



two methods of division which have been described by various observers, 

 the direct and the indirect. 



In the direct method, which was, until recently, supposed to be 

 the normal one, the nucleolus first divides, then the nucleus, and 

 finally the cell. 



In the indirect method, the nucleus first of all undergoes metamor- 

 phosis, separating into a network of highly refracting filaments (kern- 

 figur), which take up colouring matters strongly, and an intermediate 

 substance not aftected by staining fluids. The nuclear network goes 

 through a definite series of changes,* and finally divides into two 

 equal or subequal masses, which retreat from one another and go 

 through, in inverse order, the changes undergone by the mother 

 nucleus, finally forming the nuclei of the two daughter-cells. The 

 cell-body divides after the young nuclei have separated from one 

 another, but before they have assumed the characteristics of quiescent 

 nuclei. 



It is this indirect cell-division which recent researches have shown 

 to be of such wide occurrence.f As to the direct method, Flemming 

 states his disbelief that it ever takes place. The chief facts in its 

 favour are the occurrence of bi- or multi-nucleate nuclei, of renifurm 

 or biscuit-shaped nuclei and of bi- or multi-nucleate cells. He con- 

 siders that multinucleate cells are cells in which indirect division 

 has begun, but has undergone arrest ; that the formation of several 

 nuclei in one cell by direct division has never been observed, except 

 in one doubtful case, and that the steps of the apparently direct division 

 in motile cells (leucocytes, &c.), are probably homologous with, but 

 simpler than, those which characterize the indirect division of fixed 

 cells. 



Professor Flemming remarks on the fact that the indirect method 

 has -only been discovered of late years, and considers that this is 

 due to the tissues not having been examined in the living state ; in 

 the case of preserved tissues, to their not having been placed living into 

 the preservative solution ; to the fact that cell-division probably 

 exhibits periodicity, going on vigorously at certain times, and but 

 little, or not at all, in the intervals ; and, lastly, to the circumstance of 

 the most suitable reagent for preserving and staining not having been 

 used. 



Ho is quite convinced that all cases in which cell-division has 

 been thoroughly and certainly followed out, show the ■•irocess to be an 

 indirect one ; that is, to bo duo to a filamentous metamorphosis of the 

 nucleus. 



As to reagents, Flemming states that Miiller's fluid and potassium 

 bichromate are unsuitcd for observing tlie division-figures of tho 

 nucleus, and recommends a O'l to 0*2 per cent, solution of chromic 

 acid, or a saturated solution of picric acid. After treatment with ono 

 of these fluids tho tissues shouhl bo thoroughly washed, and stained 

 either witli luematoxylin or witli Hermann's aniline staining fluid — 

 the latter not answering well for picric acid preiiarations. 



* See this Journal, ii. (187!)) p. V37. 

 t Ibid., pp. 137, 273, G'J2. 



E 12 



