ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 485 



Insects and the Fertilization of Heterostylous Flowers.* — 

 Primula elatior is the plant selected by J. MacLeod for his observa- 

 tions. Here the anthers of the longistyle flowers, which are at the 

 same level as the stigmata of the brevistyle ones, are placed a little 

 above the extremities of the sepals, and in consequence, at some 

 distance from the mouth of the corolla ; those of the brevistyle flowers 

 and the stigmata of the longistyle ones are very near this mouth. 

 These arrangements are utilized by the fertilizing insects, two species 

 of Bomhus, in two different ways. The one Bomhus (sp. ?) grasps the 

 corolla with its feet, and inserts its head, but not beyond the mouth 

 of the corolla ; in order to reach the nectaries at the bottom of the 

 tube, it unrolls its proboscis. Thus, on coming to a brevistyle flower, 

 its head comes into contact with the anthers and carries off pollen from 

 them ; but the proboscis, which is projected beyond, does not touch 

 them at all. The same result, substituting stigma for anthers, takes 

 place with a longistyle flower, and pollen is deposited from the bee's 

 head on the stigma, while pollen belonging to the flower itself is 

 carried off by the proboscis in this case, to be deposited on the stigma 

 of the next brevistyle flower which it sucks. 



The other Bomhus {B. viuscorum) acts very differently ; it grasps 

 the side of the flower and pierces it laterally with its mandibles. 

 Only fully-blown flowers are observed to be thus pierced, and the 

 perforation is- always effected at the same level, namely, that of the 

 stigma of the short pistil or the anther of the short stamen (according 

 as it is a brevistyle or longistyle flower). The head is pushed into 

 the opening thus made, and the proboscis is directed towards the 

 nectaries ; thus no part of the head or proboscis touches the long 

 stigma or anthers, and it is only brevistyle flowers which are 

 fertilized by this species of bee. Thus, as far as the agency of the 

 Bomhi is concerned, the brevistyle flowers have two chances of fertiliza- 

 tion to one which the longistyle flowers possess, seeing that the latter 

 can be fertilized by but one of the two species of bee ; if these insects 

 were the only ones which fertilize these flowers, it seems probable 

 that natural selection would favour the brevistyle flowers in the 

 struggle for existence, at the expense of those with long styles. 



Contrivances for Insect-pollination in Erodium.f — F. Ludwig 

 points out that, besides the ordinary form of Erodium cicufarium, 

 there is another, var. pim^nnellifolium Willd., esj)ecially adapted for 

 cross-fertilization by insects. In the ordinary form the petals are 

 all coloured uniformly, the upper shorter ones having only occasion- 

 ally a deeper tint ; while in this variety the two upper petals are 

 shorter, broader, and of a much deeper red, with an oval dark spot at 

 the base. This form is also distinctly proterandrous, and the position 

 of the stamens prevents insects entering the flower while the pollen 

 is being discharged and before the stigmas are mature ; wliile the 

 original form is homogamous or slightly proterogynous, and to all 

 pp earance autogamous. 



* Bull. Acad, Roy. Belg., 1. (1880) pp. 27-33. 



t Kosmos, viii. (1881). See Bot. Centralbl., ii. (1881) p. 298. 



