98 



small number, frequently but one or two (fig. 53, 55, PI. IX, 

 fig. 58 PL X). After these changes have taken place the 

 role of the fertil part of the embryosac has been played, 

 nothing more occurs in it, it is gradually obliterated owing to 

 the growth of the steril part; its retarded prothalliumcells 

 play no role whatever. 



CHANGES IN THE STERIL PART OF THE EMBRYOSAC. 



While studying the fertil part of the embryosac we have 

 lost sight of the steril one. Yet this has undergone important 

 changes in the mean time. When we saw it last (fig. 29 PI. 

 IV) it was very small and occupied by far the smallest part 

 of the embryosac, while in fig. 59 PI. X it has grown 

 enormously and occupies by far the largest part of it. We 

 notice that in fig, 29 PI. IV several cells contain as yet a 

 number of nuclei, while in fig. 34 PL V this has allready 

 been remedied, every cell containing but one. Soon after it is 

 seen that, owing to a lateral growth at about the middle of 

 the prothallium its shape has become flask-like and now an 

 enlarged cell may frequently be seen at the top (fig. 36 PL V). 

 In other cases nothing of such and enlarged cell can be seen. 

 Most frequently the tip of the prothallium is rather pointed 

 and sticks in the contracted part of the embryosac (fig. 62 

 PL XI) this point can be very prominent (fig. 63 PL XI) it 

 also can be barely visible (fig. 54 PL IX). In all cases, though, 

 a. kind of conus is seen, whose top is immersed in the con- 

 striction and whose base is planted on the broadest part of 

 the prothallium (fig. 61 PL XI and others). 



It is in the top of this conus that organs like those drawn 

 in fig. 62 PL XI and fig. 56 PL X are found, organs which 

 one is rather compelled to consider as rudimentary archegonia 

 though certainty as to their nature can of course not be ob- 

 tained. Yet the presence of a body like a canalcell in both of 

 those pictured here, the regularity of the tissue around them, 

 especially striking in fig. 62 PL XI must, I think, make us 



