Parnassia and its bearing on the Affinities of the Genus. 501 
embryo in Parnassia , Saxifraga , and Drosera , she concludes that Parnassia 
is much more closely related to the Droseraceae than to the Saxifragaceae. 
Eichinger, however, who deals with the subject on broader grounds, recom- 
mends that Parnassia shall not be removed to the Droseraceae, but shall 
be retained in the Saxifragaceae, an order which, as at present constituted, 
can claim little unity. The conclusion to be drawn from previous work on 
this point seems to be that there is an undeniable affinity between Parnassia 
and the Droseraceae. As the subject has been so fully dealt with by other 
writers, I shall not enter upon it here, and in the following discussion, which 
is merely supplementary to Drude’s memoir, I shall confine myself to the 
subject of the connexion between Parnassia and the Saxifragaceae and 
Hypericineae respectively. 
I am inclined to think that too much stress has often been laid by 
systematists upon the affinity of Parnassia with the Saxifragaceae, with the 
result that the other relationships of the genus are apt to be overlooked. 
Both Bentham and Hooker in the ‘ Genera Plantarum ’, and Engler in 
the ‘ Pflanzenfamilien \ place Parnassia in the Saxifragaceae, so that 
these botanists do not convey, by means of its systematic position, the 
idea that the genus is an isolated one with affinities with several families. 
I am far from wishing to deny that Parnassia is related to the Saxifragaceae, 
but at the same time I think it should be realized that certain of the argu- 
ments, which appear to have been most powerful in convincing botanists 
of the validity of this relationship, are rather plausible than sound. 
Hooker and Thomson, 1 in 1858, described several new species of 
Parnassia from India, and after discussing various reasons for relating them 
to the Saxifrages, they add that certain alpine species of these two genera 
have ‘a habit so similar, that when in the Himalaya, their close affinity 
appeared to us self-evident \ This argument from agreement of habit is 
a somewhat dangerous one ; the similarity in form might perhaps equally 
well be regarded as due to similar modifications in vegetative structure 
brought about in response to alpine conditions. There are analogous cases 
in the European Alps ; for example, .certain 4 cushion plants ’ belonging to 
widely different families (such as Draba pyrenaica , L., of the Cruciferae and 
Androsace glacialis , Hopp., of the Primulaceae) resemble one another so 
closely as to be easily confused at first sight. It is a mere commonplace to 
point out that there are many other cases of deceptive resemblance in 
vegetative structure between plants of unrelated orders, such as between 
certain Cactaceae and the succulent Euphorbias of South Africa, and again, 
between the shoots of Veronica cupressoides , Hook, f., and the twigs of 
Cypress. 
The abbreviated stem of the Parnassias bears petiolate ‘ radical ’ leaves, 
e 
1 Hooker, J. D., and Thomson, T. : Praecursores ad Floram Indicam. Journ. Linn. Soc. Bot., 
vol. ii, 1858, p. 54. 
Mm3 
