78 DB. HOOKEE AND DB. THOMSON'S PE^OUESOBES 



than to any other Order except Droseracecs. In one curious phy- 

 siological point it agrees with Saxifraga itself, viz. the stamens 

 approaching the stigma by pairs, shedding their pollen and re- 

 tiring, an economy common to all the species we have examined 

 in a living state ; it also agrees so remarkably with that genus in 

 habit, that it is difficult to distinguish some of the smaU alpine 

 species in their native places of growth at first sight ; the petals 

 in both are persistent, of a singularly thick and coriaceous or 

 fleshy texture with transparent veins. Another point of resem- 

 blance with BaxifragecB is the variable amount of adhesion of the 

 calyx-tube and ovary : in all the species the ovary is almost free 

 in a very young state, and it is more during the after-development 

 of the fruit than of the ovary that the former becomes imbedded 

 in the tube of the calyx. The stigmas are always free, and the 

 very short styles also in the species with dehiscent capsules : in 

 the half-superior fruited species with dehiscent capsules, the 

 dehiscence takes place only above the calyx-tube, and the fruit of 

 P. nuhicola strongly resembles that of a Wahlenbergia in general 

 appearance, though not in internal structure. 



There are three principal modifications of the fruit in Parnassia: 

 a superior ovoid 3-4-valved capsule, which is the structure in 

 P. palustris, P.Jbliata, and several American species ; — a superior 

 trigonous coriaceous indehiscent capsule, in which the placentae 

 are sometimes confined to the lower part of the cell, and which 

 has a longer style than either of the other forms of capsule ; this 

 characterizes the * Peninsular' Khasia, and a few of the Himalayan 

 species ; — and a semi-superior fruit, as in the P. nuhicola, P. qffi- 

 nis, P. pusilla, and P. ovata ; species that may possibly prove 

 forms of one. 



The structure of the seeds and ovules of aU is very similar, and 

 has been well described : in being exalbuminous they differ from 

 those of any of the other groups of Saxifragece, as well as from 

 DroseracecB, amongst which they are most frequently placed in 

 systematic works. 



The placentation in all the species is decidedly parietal, as in 

 BroseracecB ; nor have I, in the earliest-examined stages, detected 

 any evidence of this being a deviation from the axile type : this 

 has been held as an argument against allying Parnassia with 

 Hypericinecd, but which the truly parietal placentation of Elodea 

 invalidates ; the loculicidal dehiscence of Parnassia, on the other 

 hand, is against its Hypericineous alliance, as are its truly peri- 

 gynous stamens and petals. 



