Dr. W. Hofmeister on the Fecundation of the Coniferee. 437 



strikingly resembles in the profile view young conditions of the 

 pro-embryos, of the Abietinece more particularly, in a lesser degree 

 those of Taxus itself. 



I found the behaviour of the pollen-tubes to the impreg- 

 nated corpusculum twofold, in Taxus as in Juniperus : either the 

 pollen-tube sends a short process into the narrow mouth of the 

 corpusculum J or it applies itself broadly over the summit of the 

 corpusculum, with a shallow convexity of its membrane projecting 

 a little way into it. In both cases the five or six cells of the 

 covering-rosette of the corpusculum were mostly only pushed 

 asunder and squeezed flat, not completely absorbed. The pollen- 

 tube was very firmly adherent to them. As they are very inti- 

 mately connected with their neighbouring cells, the pollen-tube 

 is usually torn at this place in the attempt to dissect it out from 

 the impregnated corpusculum. In rare cases the adhering cover- 

 cells of the corpusculum are separated from their connection and 

 lifted up with the pollen-tube. Then even the outline of the 

 pouch of the pollen-tube is frequently rendered indistinct by 

 adherent and contained granular mucilage, so that it is not often 

 to be clearly made out whether this pouch is closed or has a 

 minute opening. But so much the more certainly can we be 

 convinced of the closure of the flat protusions of the pollen-tube 

 squeezed in between the cells of the covering-rosette; the form 

 in which, as it appears to me, the impregnation most frequently 

 takes place. 



According to Schacht^s idea, one of these cell-rosettes, similar 

 to young pro-embryos, must be formed in the pollen-tube over 

 each corpusculum to be impregnated, into which, then, after the 

 destruction of the cover of the corpusculum, the large portion of 

 the pollen-tube lying above it has to bulge out and insert itself, 

 carrying the four- or many-celled pro-embryo to the bottom of 

 the corpusculum. Against this view speak not only the above- 

 stated, but the following reasons : — 



1. In all cases, I find the cell-rosette, which makes its appear- 

 ance in the pollen-tube, somewhat larger than the earliest rudi- 

 ment of the pro-embryo, and much too large to pass through the 

 ordinarily narrow mouth of the corpusculum. In particular, the 

 part of the pollen-tube which is torn ofi" in the attempt to extract 

 it from the just-impregnated corpusculum, is very small, quite 

 out of proportion to the size of the cell-rosettes in the pollen- 

 tube and of the young pro-embryo. 



I give a few of the measurements;— 



a. Transverse diameter of the young,^ four-celled pro-embryo, 

 0-753 miUim. 



