IlEV. O, HENSLOW OX THE STRUCTURE OF IlVDIGOrERA. 357 



over, although the staminal tube may perhaps rise up sponta- 

 neously, yet many flowers wither without the carina having fallerv 

 at all. In Ind'ujofera the petals are more fugacious, and the ala^ 

 are not so firmly attached to the carina as in Medicago. 



Both DeCandolle and Treviranus (Bot. Zeit. 18G3,p. 3) think 

 that this falling of tlie carina takes place spontaneously at the 

 full development of the flower; and the latter authority calls; 

 in question tlie necessity of insects, now proved by Darwin in 

 JPItascohis, TrifoUum, and Lvpinus^ and even goes so far as to 

 say, '' but all these movements occur In the natural develop- 

 ment of the parts, and only after self-fecundation has taken 

 place." Dr. Hildebrand thereupon justly remarks that "both 

 these naturalists are right in saying that this movement of the 

 carina may take place spontaneously in the natural development 

 of the several pax*ts of the flower; for specimens {Indigofcrd) com- 

 pletely secured from insects showed the carina lapped back in the 

 oldest flowers. It is quite possible that by this violent motion, 

 produced entirely by the development of the flower, pollen may 

 fly on the stigma from the shock ; but on the other hand avc must 

 allow that In most cases the flowers will probably be visitt^d by 

 insects and fecundated by their aid before they have, in the pro- 

 cess of their development, advanced as far as the lapping back of 

 the carina. This is the point which is my ground fur the con- 

 jecture that in these cases, although self-fecundation may be pos- 

 sible, yet, doubtless, in nature, insects mostly anticipate the na- 

 tural course in the development of the flower, by causing the 

 above-mentioned movement, in tbe search after honey." 



But it seems to me that Dr. Hildebrand does not quite reach 

 the probable motive ; and that is, as I have already said, the inter- 

 crossing of distinct jlowers : for in all these cases, although self- 

 impregnation may be possible, and the aid of insects may enhance 

 that end, yet at the same time by the flowers being so contrived 

 as to specially load tbe insects that visit them with poPen, the 

 inference cannot be avoided, that the intercrossing is the must 

 important end. 



Dr. Hildebrand also describes the structure o^ ^ledicago saliva 

 somewhat similarly to the present writer, a notice of which will 

 be found at page 327, 



Lastly, he describes an apparently very feeble imitation iu^- 



^^^us caiiarlensts and alhiis. 



*'The anthers/* he "observes, "open in this species in the bud, 

 and the stigma is already developed at that time; but it prq- 



