﻿I go BOTANICAL GAZETTE [march 



wall formation is indefinite and incomplete, and that beyond 

 this no walls are formed. 



The rapid elongation of the suspensor {^figs. 13-16) forces 



* 



the embryo down into the endosperm. The resistance is so 

 great that in later stages the base of the embryo is forced 

 upward into the archegonial chamber and is stopped only by 

 the hard seed coat. The suspensor is long and twisted. When 



{fig 



cm 



long. 



Certain features of the mature embryo may mentioned. The 

 two cotyledons are free at apex and base, but are more or 

 less completely fused in the middle region. The long cotyle- 



CA. 



the larger part of the embryo, 



« 



the whole axis being comparatively small. The plumule is 

 well developed, several leaf primordia being evident. Several 

 branching mucilage canals extend from the root to the tips ot 

 the cotyledons. According to Warming (4) the embryo of 

 Ceratozamia has only one cotyledon, a part of the meris- 

 tematic apex of the axis becoming differentiated into the single 

 cotyledon and the rest forming the stem tip. 



These results with Zamia have enabled us to arrange a sug- 

 gestive series of gymnosperms as shown by the accompanying 

 diagram [fig, ij). In the first series a gradual reduction in the 



number of free nuclei is shown. In Ginkgo, Cycas(?), and Zamia 

 it would seem that eight successive simultaneous nuclear divisions 

 precede wall formation, resulting in 256 free nuclei. Among 

 the Coniferales there is a sudden drop in the number of free 

 nuclei, but at the same time a diminishing number from 32 to 

 4, as follows: Taxus 32; Cephalotaxus, Podocarpus, and 

 Taxodium 16; Thuja 8; Pinus 4. While these numbers proba- 

 bly represent the usual condition, the higher numbers are not 

 entirely constant, wall-formation sometimes occurring at an 

 earlier division than usual. For example, in Podocarpus wall- 

 formation may occur after the third division, but usually only 

 after the fourth; while in Taxus wall-formation has never been 

 observed to occur until' after the fourth division, and often only 

 after the fifth. Further reduction in the number of free nuclei 

 occurs among the Gnetales, Ephedra sometimes showing only a 



