Jan. 1892.] THE BOTANICAL SOCIETY OF EDINBURGH. 2-11 



any other cruciferous plant known to me, and by many 

 different species. Sisymbrium alliaria is much frequented 

 by Syrphidffi, though of few distinct kinds, and insect-visits 

 are very fairly common in the case of Capsella Bursa-pastoris 

 which secretes plenty of honey. One might, indeed, ask 

 why there are so many and such varied modifications 

 adapted to insect-fertilisation, and why cleistogamy should 

 not be universal if self-fertilised flowers are perfectly 

 fertile. 



In fact, the conclusion seems to me inevitable that 

 insect-visits are chiefly relied upon in every case, and 

 that self-fertilisation is only of occasional and additional 

 assistance. 



There are certain slight modifications even in the Cruci- 

 feraj only explicable by the advantage of insect-\'isits. 



1. The Flowers are Collected into Corymbs or 

 Corymb-like Heads. — This is the case in Ibe^^is amara, 

 Teesdalia nudicaulis, Nasturtium ojicinale, and some varieties 

 of Brassica campestris. Nasturtium officinale is very abun- 

 dantly visited probably for this reason ; so also is Teesdalia 

 nudicaulis (cf. Mtiller, loc. cit., p. lOG), in spite of its extremely 

 small flowers ; Ibcris amara has never, so far as I know, 

 been studied in this respect. Both in Iberis amara and 

 Teesdalia nudicaulis, the outer petals of the most external 

 flowers of the corymb are much longer and broader than 

 the inner, just as one finds very commonly in Umbellifer^e. 



2. Nectaries. — The nectaries seem to have been originally 

 six in number. One is found outside each pair of long 

 stamens, and a pair of nectaries occurs inside the insertion 

 of each short stamen. It seems almost certain that Mtiller 

 is right in considering these to be the remains of the lost 

 stamens of the flower. Whether, however, there were 

 originally twelve stamens, or four stamens, each with a pair 

 of stipular appendages at the base, does not seem to me very 

 clear. The nectaries are in some forms of almost equal 

 importance, e.g., Cakile maritima and Cardamine amara ; the 

 tendency, however, is for those at the base of the short 

 stamen to become more developed, while those of the long 

 stamens either disappear (Nasturtium officinale, N. palustre), or 



