THE CELL AND ITS CONSTITUENT STRUCTURES. 149 



the degree of the lobing varies in different species. The 

 important fact for our purpose is that in all these four 

 lobed cells, at the commencement of division, four, and not 

 two, centrospheres appear simultaneously, placed sym- 

 metrically at the periphery of the nucleus. F"urther, in 

 those spore mother cells in which the lobing is not very 

 prominent, the four centrospheres ultimately coalesce in 

 pairs, so that a normal bipolar spindle is the result as soon 

 as the first division is completed, the daughter nuclei begin 

 to divide in a perfectly normal, that is, bipolar, fashion. 

 But in the cases of extreme lobing, as in Pallavicinia} there 

 is no room for a symmetrical bipolar spindle, and the four 

 poles persist throughout the entire division. This is modi- 

 fied, inasmuch as two divisions of the chromosomes occur 

 without an intervening period of rest, and thus the four 

 daughter nuclei are formed simultaneously ; transitional 

 stages, however, connect the two extremes. The general 

 conclusion impressed on the mind of the observer is, that it 

 is the shape of the cell which in some way determines the 

 proximate character of the division. 



Now, these facts seem to me to militate strongly against 

 the permanent organ — and individuality — hypothesis of the 

 centrosome. How can one imagine that the centrosome 

 could in these cases become divided into four, and direct the 

 protoplasm to transmit each quarter to its right place, with- 

 out betraying any evidence of the process '^ For it is hardly 

 conceivable, that in a large mass of material, all fixed at the 

 right time, such stages, if present, could have escaped ob- 

 servation when they were being directly searched for. To 

 my own mind a more natural explanation seems to be that 

 the centrosome in this case originated, de novo, at the spots 

 at which they were found, and that their singular mode of 

 distribution is to be regarded as the result of the relative 

 degree of independence associated with the four masses of 

 protoplasm, an independence due primarily to the greater 

 or less complete severance of the living contents of any 

 one lobe from those of the rest. And, furthermore, the 



^ Annals of Botany, vol. viii., p. 47 et seq. 



