48 Selmar gchonland, [May 28, 



meaning of this structure ? I believe that all these arrangements are 

 there to ensure crossfertilisation. It will be seen from Fig. it. and v., 

 that the pollengrains deposited on the top of the style can hardly 

 reach the entrance of the stigmatic cavity. There remains further the 

 fact that the pollen is frequently removed, which can only be done by 

 insects (although I must admit that with the exception of a little 

 butterfly I have not seen any insect visiting the flowers), and insects, 

 laden with pollen from other flowers, will probably leave some of it on 

 the slime round the entrance of the stigmatic cavity. However, this 

 has to be confirmed by actual observation before it can be accepted as 

 representing the full truth ; besides as the male and female conditions 

 of the flowers are not so sharply separated as in most Lobelias, it is 

 quite possible that self-fertilisation may also sometimes be effected 

 either with or without the aid of insects. 



One curious fact remains to be mentioned. Pollengrains are not 

 only found at the mouth of the " stigmatic cavity," but also inside it. 

 I have found them there so frequently that it is almost impossible to 

 imagine that they got there accidentally when my sections were 

 prepared. I may suggest that the slime dries up a little occasionally, 

 and that the pollengrains are thus drawn in by the slime reducing its 

 volume. Fresh slime being formed other pollengrains are caught, and 

 the same game may be repeated. To quote a somewhat analogous case 

 I may mention that in many Goniferae the pollengrains are also caught 

 by a drop of fluid at the mouth of the micropyle and drawn down on to 

 the top of the nucellus when the fluid dries up. 



The structure of the style besides giving a hint as to how pollin- 

 ation is effected in Cyphia volubilis, is also very interesting from 

 another point of view. Harvey, in his " Genera of South African 

 plants " ( x ), and Sonder in " Flora Capensis " ( 2 ), say that the 

 '* stigma " of Cyphia has an " obsolete ciliate indusium " and this 

 character probably induced them to place the tribe of " Cyphieae " 

 next to the " Goodenoviece " their 4th tribe of Campanulaceae. The 

 latter tribe is only represented in South Africa by a species of the 

 genus Scaevola, a widely spread sea-shore plant. Most species of the 

 " Goodenoviece " are Australian. Bentham and Hooker separated the 

 Goodeniacece from the Campanulaeece as an order, following in this 

 respect almost all previous botanical authors, who had had occasion to 

 treat of them. Baillon ( 3 ) united them again without, however, 

 recognizing that the " Cyphiece " possessed any structure corres- 

 ponding to the " indusium " of ithe " Goodenie<z." In his description 

 of the style of Cyphia he simply says " stylo apice stigmatoso obliquo 

 vel 2-lobo." Now as long as no plant was known among the 



Q. 2nd ed» p. 215. ( 2 ). iii. p. 597. ( 8 ). " Histoire des plantes," viii p. 869. 



