EVOLUTIONARY TENDENCIES AMONG GYMXOSPERMS 419 



features of the primitive ovule, and its association with fertihzation 

 by ciliated sperms is so evident that it is natural that the two disap- 

 peared simultaneously with the establishment of Coniferales, but 

 the abruptness of the disappearance is evidently more apparent 

 than real. The presence of an extraordinarily deep pollen chamber 

 in Ephedra can hardly be regarded as a contradiction to this general 

 statement, for in that case it is evidently of secondary origin, associated 

 with a remarkably massive archegonium neck. 



THE FEMALE GAMETOPHYTE 



The female gametophyte of gymnosperms exhibits a progressive 

 series of changes which is significant because it leads toward the 

 angiosperm condition. At this point the very important historical 

 record fails, and the entire testimony must be obtained from living 

 forms, which do not represent a series, but the ends of many series. 

 For this reason, and also because such progress is always very unequal 

 in different forms, various stages of advancement may be expected 

 to be found in forms grouped in a single alliance. The series, there- 

 fore, is not so much one which conforms to recognized groups, as a 

 series of stages each of which may be exhibited by members of various 

 groups. 



The general development of the gametophyte is quite uniform; 

 and since the same sequence of events occurs among the hetero- 

 sporous pteridophytes, it may be inferred that a knowledge of the 

 ancient gymnosperms would not materially change this situation. 

 The general sequence referred to in the development of the gameto- 

 phyte is as follows: free nuclear division, usually accompanied by 

 vacuolation which results at some stage in parietal placing; the for- 

 mation of w^alls, resulting in a parietal tissue; the centripetal growth 

 of this tissue until it reaches the center of the embryo sac; and the 

 final growth of the gametophyte until it reaches its mature size. In 

 certain cases there may be no parietal placing, the free nuclei remain- 

 ing distributed throughout the embryo sac; and therefore there is 

 no centripetal growth, but general wall-formation throughout the 

 sac. The details of the formation of permanent endosperm tissue 

 from the primary walled cells are variable and perhaps very important 

 from the evolutionary point of view, but the range of forms from which 



