300 Beer. Pevelnpiiieiit of tlie j)itllen grain and auther of some Onagraceae. 



their measiirements witli those described for fresh material. I 

 will not however, lay great weiglit on the evidence of the 

 inicrotome soctions as. in spite of every precaution, I never suc- 

 coeded in ontirely avoiding shrinkago of the pollon-protoplast 

 even when all the other cells of the anther wevc un-contracted. 

 I will add here a comparison between the measurements of the 

 })ollen from a living anther with those of microtome sections: 



1. Fresh material of Ocnoiliora hiennis examined in the 

 Juice squeezed from the anther: 



Pollen grain =- 74 //, 



„ cavity = 40 //, 

 „ protoplast = 30 ,«. 



2. Sections of anther of about same age fixed in Flem- 

 mings Solution: 



Pollen grain = 72 ft . 70 // . 70 // . 70 (f. 



„ cavity = 42 '« . 40 '// . 40 // . 40 '//, 

 „ protoplast = 26 // . 28 // . 24 // . 22 /,. 



Diu'ing the Avhole time that the protoplast is separated from 

 the membrane in this way the latter continues to grow botli in 

 extent and in thickness. We are at present quite in the dark 

 regarding the manner in which this growth takes place but a 

 very brief theoretical consideration of the subject will be found 

 among the conlusions at the end of tliis paper. 



We must now enquire whence is derived tlie material 

 necessary for tliis growth. 



There are two sources from which the plastic material of 

 the membrane miglit be derived, viz the protoplast of the 

 pollen grain itself or the tapetum. 



That metabolic processes of no mean order are taking place 

 in the former is evident from a study of the changes which 

 can be observed in it during this period. 



Starch appears and disappears in the pollen grain in a 

 manner which shows that carbohydrates are being used up in 

 the cell; the cytoplasm continually grows less and less in amount 

 Avhilst a liquid, apparently the direct consequence of the fore- 

 going processes, gradually forms in the protoplasm. 



This liquid iii'st occupies small vacuoles in the cytoplasm, 

 these continue to increase in size and run together (Fig. 24) 

 until w^e find nothing left of the protoplasm but a hollow shell 

 consisting of a plasmoderma (Hautschicht) and a nucleus, sm-- 

 rounded by a trace of granulär cytoj)lasm (see Fig. 25 which 

 gives a rather later stage). The centre of the shell is occupied 

 by one enormous vacuole^). 



There is no reason to doubt that this liquid diffuses out 

 from the protoplast into the space which is forming between 



1) Strasbiirger in his work of 1882 already v/rote of Gaura hiennis. 

 „In meinen Alkoholjjräparaten bildet der nacii Anlage der Wand erschöpfte 

 Inhalt der Pollenzelle nur noch ein unscheinbares Klünipchen" cf. his 

 Figs. 47, 48 and 49, Tafel VI. 



