434 Dr. W. Hofmeister on the Fecundation of the Coniferse. 



smaller granules ; lastly, minute vesicles with very finely gra- 

 nular, almost transparent contents, of diameter about equalling 

 tlie length of the spindle-shaped bodies. 



It appears indispensable to the accomplishment of the impreg- 

 nation, that the pollen-tube should, in the Cupressinese also, 

 completely displace or push aside the covering rosette of the 

 corpusculam, so as to come into direct contact with the upper 

 convexity of the corpusculum. My former statement*, contrary 

 to this, seems to depend upon an erroneous interpretation of an 

 observation : either the large cells, apparently situated in the 

 lower concavity of two corpuscula, observed then (only in one 

 case) with a perfect preser\^ation of the rosette of cells closing 

 up the corpuscula, may have been abnormally enlarged sister- 

 cells of these corpuscula, beside, not in the corpuscula ; — or, the 

 pollen-tube may have reached the impregnated corpusculum late- 

 rally, by a protruded process removed in making the section. 

 In an uncommonly large number of examinations, I have 

 never met with anything similar in Juniperus. Cases correspond- 

 ing to the first hypothesis, I have observed repeatedly, as before f, 

 in CupressuSy in which the development of pollen-tubes and of 

 embryos was wholly arrested every year, while the albumen and 

 the corpuscula attained to their full size — as ordinarily happens in 

 our climate, probably resulting from the low temperature at the 

 period of the discharge of the pollen. Phsenomena analogous to 

 the second conjecture occur in the Abietinese, where it happens at 

 times, in the most varied species, that the pollen-tube makes its 

 way to the corpusculum by penetrating laterally through the 

 tissue of the albumen, and not by the appointed way, through 

 the covering-cells of the corpuscula. In Biotia orientalis also I 

 have observed a pollen-tube penetrate into the endosperm far to 

 one side of the group of corpuscula; it contained one of the 

 often-mentioned large cells. 



Ordinarily the pollen-tube sends out a short process into the 

 corpuscula to be impregnated, pushing inwards the softened, 

 apparently thicker membrane of their upper convexity. At the 

 same time it pushes the remains of the compressed covering- 

 cells before it, and penetrates into the interior of the corpus- 

 culum, through that gelatinous layer, if this has not been, as 

 often happens, already dissolved and destroyed. More rarely 

 the pollen-tube merely rests upon the summit of the corpus- 

 culum ; it is equally rare for it to penetrate into it more than 

 about Yo^^ ^^ ^^^ length. Once I found it (in Juniperus com- 

 munis) advanced as far as ^th of the longest diameter of a cor- 



* Vergleich. Unters. p. 131. pi. 33. fig. 12. 

 t Ibid. pi. 33. fig. 26. 



