THE CELL AND ITS CONSTITUENT STRUCTURES. 163 



ments were present, and that these were split lengthwise 

 into four quarter chromosomes, by two longitudinal planes 

 at right angles to each other. Thus the number of the 

 chromosomes was already reduced, but the daughter chro- 

 mosomes for this and the next division were provided for. 

 According to Boveri, who is supported by other investi- 

 gators, the same thing probably occurs also in connection 

 with the formation of the ova in the same animal. Again, 

 in the Salamander, as Flemming showed, the first division is 

 a longitudinal one, although owing to the complexity intro- 

 duced by the formation of closed rings, it at first sight 

 might seem to be a transverse fission. That the second of 

 the two divisions is longitudinal and not transverse is clear, 

 inasmuch as it occurs while the daughter chromosomes of 

 the first divisions are retreating to the poles. In this re- 

 spect, Salamander offers a fine transitional stage between 

 a normal case in which the splitting takes place at the 

 commencement of the second division, and the instance 

 quoted from Brauer, where the chromosomes undergo a 

 simttltaneotLS double fission which is thus effected for the 

 two mitoses at one instead of at two stages. Finally, in the 

 pollen mother cells of such plants as lilies, in which the 

 details of the process have been most carefully worked out 

 by a number of observers, it is clear that the chromo- 

 somes which appear in the first division spindle are reduced 

 in number to one half, that these undergo longitudinal 

 fission, and that in the next, and final division which follows 

 rapidly on the first, a second longitudinal division also 

 occurs. These observations have also been repeated on 

 the division of the embryosac, and they have lately been 

 again made the subject of an extremely careful investiga- 

 tion by Miss Sargant,^ with the result that there is not the 

 slightest evidence in favour of the reduction being in any 

 way brought about by a sorting of the chromosomes as a 

 whole ; that is, the reduction as it occurs in these plants 

 does 7iot conform to the type as exhibited in the oogenesis 

 and spermatogenesis of Copepods. But if it does not 

 signify whether the chromosomes divide longitudinally or 



^ Loc. cit. 



