242 TRANSACTIONS AND PROCEEDINGS OF [Sess. lvi. 



cease to secrete honey {Alliaria, Cramhe). The pollen of the 

 short stamen is wholly devoted to cross-fertilisation by 

 insects seeking honey, while that of the long stamens seems 

 in part used by pollen-eating flies and in part employed in 

 self-fertilisation, hence the- minor importance of the nec- 

 taries of the long stamens is obvious enough. The nectaries 

 of the short stamens sometimes consist of two separate 

 peg-like processes, in other cases they are confluent with one 

 another behind the filament, and form a single pulley-like or 

 kidney-shaped mass {Barbarea, Nasturtium officinale, Carda- 

 minc ficxuosa, Lepiclm-m Smithii, CalHle, Cramhe). In Arahis 

 Thaliana and Alliaria, however, the whole of the tissue 

 about the insertion of the short stamen has become nectareal, 

 and the filament is seated on a sort of cushion of honey 

 tissue. 



3. Sepals. — The sepals are in many Crucifera? saccate 

 at the base, so as to form a sort of pouch to hold the honey. 

 The degree of this pouching seems to depend upon the 

 development of the corresponding nectary, as has been already 

 pointed out by Mliller. Thus in RaijTianiis maritimus, 

 Brassica monensis, Cardamine ficxuosa, C.pratensis, and Arahis 

 Thaliana, the sepals opposite the short stamens {i.e., the 

 largest, most important nectary), possess distinct pouches, 

 while the other sepals do not. Where the nectaries are 

 nearly equal in size, the sepals are either not pouched at all 

 or are equally saccate {Nastiirtium palustre, Cardamine 

 amara, Brassica Sincqns). 



In Cakile, Cranibe, Zepidium, and other Siliculoste, the 

 honey is contained within the filament, and the same sort 

 of pouching does not, so far as I know, occur. 



Of course, the mere fact of the existence of a prominent 

 swelling at the base of a long and narrow pod, must tend 

 to produce a corresponding hollow at the base of the sepal 

 which covers it in the bud ; in fact, it seems to me that this 

 is quite sufficient to account for the presence of the pouch, 

 and for its usually accompanying a well-developed nectary. 

 In the Siliculosu!, the pod being broader upwards, prevents 

 a similar formation (as the sepaline midrib usually runs in 

 a straigiit line from the broadest part of the pod to the base), 

 but even in these forms a well-developed nectary sometimes 



