ON THE SELF-FERTILIZATION OF PLANTS. 267 



ble of self-fertilization, and is not improbable in the sense 

 that they " can and do fertilize themselves habitually."' But 

 his inference that the majority of flowers, or that any flowers, 

 actually propagate for a series of generations by self-fecunda- 

 tion, or that a cross if it occur is " exceptional," and of no , 

 account, is surely unwarranted by the evidence which he has 

 adduced. 



Occasionally the reported facts will not bear scrutiny. 

 Gentiana Andrewsii, it is said, never opens at all in America. 

 It opens in sunshine in the middle of the day here in New 

 England. And while looking at closed flowers we have seen 

 a humble-bee emerge from one. We have in this Journal 

 shown how it is that self-fertilization is impossible during the 

 first three or four days of anthesis, but neatly practicable after- 

 wards. It is rash to infer (as on p. 330) that papilionaceous 

 flowers which shed their pollen early in proximity to the 

 stigma are therefore self-fertilized. In most of the cases ad- 

 duced the pollen is not lodged upon the stigma, but upon the 

 style below it, and the adaptations for intercrossing, though 

 the mechanism be different, are as explicit as in the analogous 

 case of Campanula. " Fremont pathetically describes the 

 solitary bee that rested on his shoulder at the top of Pike's 

 Peak." The pathos is wasted as respects all but this particu- 

 lar bee ; for the entomologists find the alpine region of the 

 Rocky Mountains to be as well stocked with flying insects as 

 are alpine regions in other parts of the world. They do not 

 super-abound, but if from the alpine flora we subtract the evi- 

 dently entomophilous and the anemophilous blossoms, the re- 

 mainder will be nearly nil. And as to the correlation of this 

 comparative scarcity of insects with the marked conspicuous- 

 ness of blossoms, this is the way the lesson is read by a most 

 eminent physiologist : " Even the glowing hue of alpine flow- 

 ers is accounted for by the attraction which brighter-colored 

 individuals exercise upon the insects, scarce in those heights 

 and necessary for fertilization." 



One or two of the author's own observations are perhaps to 

 be revised. " Gaura parviflora . . . has no corolla and is 

 cleistogamous, in that it is self-fertilizing in bud, as I found 



