1904] LAND—EPHEDRA TRIFURCA 5 
THE REDUCTION DIVISION AND MALE GAMETOPHYTE. 
As has been said, the microspore mother-cell remains in the rest- 
ing stage about one month. The mother-cells up to the late pro- 
phase are filled with starch, which now quickly disappears. ‘There 
does not at all times seem to be uniformity in the stages of division 
in the cells of a loculus. Instances were observed in which all the 
cells of a loculus were in the same phase of division. Again, those 
in the upper part of a loculus were in synapsis, those at the bottom 
had formed tetrads, while all intermediate stages were between. 
The spirem segments into twelve chromosomes (not all of which 
are shown in jig. 8), which as they come to lie in the equatorial plane 
of the nucleus are short and thick, closely massed, and can be counted 
only with extreme difficulty. The result of repeated countings made 
in various stages of the first division, as well as in the second, leaves 
no doubt that the gametophyte number is twelve. Jaccard reports 
eight in E. helvetica. 
Twelve appears to be the prevailing number of chromosomes in 
the gametophytes of gymnosperms. Three exceptions to this state- 
ment are to be noted: Overton (’93) reports eight for Ceratozamia 
mexicana; Strasburger (’04) finds eight in Taxus baccata; the other is 
E. helvetica, with eight according to Jaccard. Dixon (’98) reported 
eight chromosomes for Pinus sylvestris, but Blackman (’98) and Miss 
Ferguson (’or) have shown beyond doubt that the number in this 
species is also twelve. 
Each chromosome apparently consists of four rods lying in 
extremely close contact. In the heterotypic division (fig. 9) the 
chromosome divides longitudinally, the ends opening out to form the 
X, Y, V, and 0 forms which are characteristic of the heterotypic 
division in higher plants. A membrane (figs. 17, 12) begins to form 
between the daughter-nuclei, as if spores of the bilateral type are to 
result, but in the great majority of cases the membrane wholly dis- 
appears, and the spores are of the tetrahedral type; although instances 
were noted in which they are probably bilateral. ‘The second division 
in the pollen mother-cell is homotypic (fig. 73) and immediately 
follows the heterotypic division. In this second division longitu- 
’ dinal splitting of the chromosomes can be seen with little difficulty. 
The J form is quite conspicuous. As the chromosomes separate, in 
