CULTURE OF THE GRAPE. 31 



facilitates the natural impregnation of the pistil, which, 

 of course, must be prevented where a cross is intended. 

 Another and greater difficulty arises from the cohesion of 

 the petals at the top. Fig. 1 illustrates a single 

 flower from a cluster ; the corolla being raised by 

 the stamens, the petals remaining united until 

 they fall. It is frequently the case that this little 

 cap remains over the stamens, holding them over Fl s- l - 

 the pistil, and letting in sufficient air to mature the 

 pollen for its work of impregnation; so that the work 

 is done before the flower is expanded. In many cases, 

 it would be difficult, if not impossible, to determine 

 whether this impregnation had actually taken place 

 or not; and hence the uncertainty in determining 

 whether seedlings are natural, or are artificial hybrids. 

 The only practical and satisfactory test is in the char- 

 acter of the offspring. It is but reasonable to expect 

 that many attempts at crossing will prove to be fail- 

 ures ; but it is also as reasonable to hope for success in 

 some cases. Indeed, it would seem to be indisputable 

 that success has already been attained in some instances. 

 In answer to some inquiries, I have received a letter from 

 Mr. Edward S. Rogers, who is widely known as having 

 made careful experiments in hybridizing. From this let- 

 ter I make an extract : 



