MB. A. W. BEIST^ETT Otf CLEISTOGAMIC FLOWERS. 



271 



Fig. 6. 



orange-yellow extension of the connective (fig. 2), the number is 

 reduced to two, or the merest rudiments are discernible of the 

 other three ; these two (fig. 6) have long, brown, strap-shaped 

 filaments, ending in a large brown spoon-shaped pro- 

 longation of the connective, at the base of which are 

 placed the two anther-lobes at some distance from 

 one another and opening by terminal pores. These 

 anther-lobes attach themselves with such strength 

 to the recurved stigma that they adhere to it with 

 the growth of the ovary, the filaments being rup- 

 tured at their base (figs. 7, 8). Here they may fre- 

 quently be seen even when the capsule is nearly V.jucullata. 

 mature. Finally, the pollen-grains are very few in Fertile sta- 

 number, of very small size, and their cell-wall evi- 

 dently very thin and transparent. The structure is, it will be 

 seen, in all important points identical with that described by 



men. 



Fig. 7. 



Fig. 8. 



Viola cucullata. 7. Side view of pistil, with stamen attached to stigma. 

 8. Front view of pistil, with two stamens attached to stigma. 



Darwin in the case of V. canina ; and there is also one other point 

 of resemblance. Darwin describes, in the case of the species just 

 named, and Michalet in that of V. alba, an open passage conduct- 

 ln g from the funnel-shaped stigmatic cavity to the upper part of 



Th ,> - -» m . . • *W^ 17 t _ / ^ 



the 

 8ho 



ovary 



o 



What, however, the purpose of this passage may be is more diffi- 

 cult to determine. Is it for the passage of the pollen-tubes ? 

 Darwin states that he " was able to trace the tubes from the grains 

 ^me way down the stigma," but does not say that they made 

 their way down this passage. In none of the various descriptions 



of cleistogamic flowers have I been able to find that the actual 



x 



