THE EMBRYOLOGY OF THE HONEY BEE 35 



ment between Polistes and Apis in the particulars relating to the 

 entrance of the cleavage cells into the cortical layer and their 

 subsequent transformation into blastoderm. The chief points 

 of difference are: the narrow, almost pointed shape of the 

 cleavage cells of Polistes when entering the cortical layer, and 

 the relatively scanty amount of basal (central) material left to 

 form an inner cortical layer. 



The further changes in the history of the blastoderm can best 

 be followed by taking them up in the order of their age. This 

 is in fact the only safe method of determining the order of 

 development during this period. 



Seventeen to twenty hours. The formation of the blastoderm, 

 it is estimated, is completed in from 14 to 16 hours, since the 

 stage under consideration, identified by means of eggs whose age 

 is actually known, immediately follows and is directly connected 

 with the stages just described by changes which can occupy only 

 a brief interval. 



While the cleavage cells are entering the cortical layer, or im- 

 mediately after this event, they divide mitotically, with the spin- 

 dles parallel to the egg's surface, as already described. This may 

 possibly be followed by another similar division, but whether 

 this occurs is uncertain ; at all events not more than two divisions 

 of this character intervene between the entrance of the cleavage 

 cells into the cortical layer and the stage of 17-20 hours. Sub- 

 sequent mitotic divisions however do occur, probably in every 

 blastoderm cell, but the spindles of the mitotic figures are no 

 longer uniformly parallel to the surface of the blastoderm; they 

 now more frequently are inclined at an angle to it, in other 

 words, they are oblique (Fig. I2B and C). As these illustrations 

 show, the angle formed by the spindles of different cells with 

 the surface of the blastoderm is not uniform; some spindles are 

 parallel to the surface, as in the left hand cell of figure I2B ; 

 some form an angle of 45 degrees with it, as in the middle cell 

 of the same figure, w r hile others are inclined at a lesser angle. 

 The effect of these oblique divisions is shown in figures I2D and 

 15. At first glance the blastoderm appears now to consist of 

 two layers of cells, superimposed one above the other, since two 

 rows of nuclei are seen, except along the dorsal mid-line. That 

 two layers of cells are actually present is rendered improbable 



