NOTES ON SAXIFRAGES. 



31 



tube. In S. tridactylites the ovary is entirely inferior, there is 

 no free tube, and the stamens are epigynous, as in Ribes and 

 Philadelphus. Saxifraga is the only genus!I remember in which 

 there is every transition between a superior and an inferior 

 ovary. In other respects the structure of the flower is very uni- 

 form, except that in a few species the petals are unequal, and in 

 florulenta the carpels are three instead of two. The stamens 

 are always ten, five opposite the petals and five opposite the 

 calyx-segments, the filiform filaments exceeding in length the 

 two-celled anthers, which dehisce longitudinally ; but this differ- 

 ence in the cohesion of calyx-tube with ovary is a character 

 which in systematic Botany is usually regarded as so important 

 that plants that differ in this way are not classed in the same 

 natural order. 



Petals. — We may roughly classify the Saxifrages under three 

 groups, according to the colour of the petals : pure white, as in 

 liypnoides and granulata ; white dotted with small spots of red, 

 as in umbrosa and stellaris ; and bright red or bright purple 

 or bright yellow, as in aizoides, Hirculus, and oppositifolia. I 

 should like to ask whether this difference in the colour of the 

 flower is at all correlated with its attractiveness to insects. 

 According to the observations of Hermann Miiller, some of the 

 bright-coloured species, in their native homes amongst the 

 mountains, are visited by an extraordinary number and variety 

 of insects. He gives a catalogue of 126 species which he has 

 observed to visit the bright yellow flowers of S. aizoides. Eight 

 of these are Coleoptera (beetles), 85 Diptera (flies), 20 Hymeno- 

 ptera (bees), and 13 Lepidoptera (butterflies and moths). To 

 what extent are the flowers of the garden Saxifrages visited by 

 insects ? What relation, if any, have the red dots on the white 

 petals of several of the species to insect fertilisation. The 

 glandulosity of the calyx must interpose a barrier to small wing- 

 less insects creeping up into the flower from below, 



Fertilisation and Hybriclity. — There are not nearly so many 

 hybrids in Saxifraga as in Primula. I give in an appendix a list 

 of the principal hybrids known in cultivation, with their par- 

 entage. In some of these — as, for instance, between media and 

 aretioides, there exists a series of forms leading gradually from 

 one specific type to the other. In other cases — for instance, 

 Andrewsii — the hybrid type is uniform and equidistant between 



