sented as long-stalked, and the aspect of the plant would appear to be 
somewhat different. I am aware that the characters drawn from 
the tubers are said by some authorities not to be reliable. Dr. Welwitsch 
has appended a note to some specimens in the Kew Herbarium of S. 
cordigera (which is usually described as producing only two tubers), in 
which he states that the greater number of individuals of this species in a 
certain part of Portugal habitually produce three tubers, and more rarely 
only two. Here I may observe, however, that it is possible that in certain 
districts races may be found characterized by peculiarities in their tubers, so 
that we might distinguish the Serapias cordigera, L., as it appears in some 
parts of Portugal, from that of Southern France, where I have taken up 
large numbers of the roots of this species without finding a single example 
with more than two tubers. I am indebted to Dr. Welwitsch for some 
interesting notes on the Serapias occultata, Gay, which he recognises as 
identical with his Serapias strictiflora, Welw., specimens of which may be 
found inthe Kew Herbarium. Dr. Welwitsch says, ‘‘ I met with this plant 
first in the month of May, 1840, and more frequently in the month of 
May, 1841, and nearly always in pratis brevius herbidis humidiusculis 
juxta rivulos, nunc in sylvis lucidis nunc in ericetis apricis prope Lumiar et 
Bellas in agro Olisiponensi nec non ad pedem de ‘Serra d’Arrabida’ trans 
Tagum. At that time I was preparing my Flora lusitanica exsiccata for 
the Unio itineraria Wiirtemburgensis, and shortly after the distribution 
of my Portuguese plants by the said Union, Professor Hochstetter 
informed me that my S. strictiflora had been described by J. Gay under 
the name of S. occultata. In the following years, from 1848 to 1850, 
when extending my botanical excursions to the other districts of Estrema- 
dura and the neighbouring provinces of Portugal, I met with the same 
Serapias nearly everywhere in similar localities, but also sometimes on 
dry and sometimes on nearly inundated spots, and so great was the number 
of varieties that I became very doubtful about the real specific difference 
of the said Serapias from S. lingua. By this circumstance I have been 
induced to retain my first manuscript name for the primitive form in 
nearly all my latter distribution of Portuguese plants. . . . I have never 
published a description of my plant... .” We have here then high authority 
in favour of the doubts as to whether Serapias occultata is really a distinct 
species. In all the specimens of Serapias occultata which I have examined, 
the flowers were capable of self-fertilization, on account of the loose 
structure of the pollen-masses, packets from which fall spontaneously on to 
the stigmatic surface, producing certain impregnation, as the quantities of 
regularly set capsules witness. Thisis quite a new feature for the genus, 
and one which, if it proves constant on extended investigation, will afford 
a specific character of great physiological importance. The condition of 
the pollen-masses seems therefore, to be quite analogous to that in Orchis 
intacta, Lk.,* these species forming two of the rare exceptions in which an 
orchid is found to be capable of self-fertilization. Serapias occultata, Gay, 
* See Darwin, ‘ Notes on Fert. of Orchids’ in Ann, Nat. Hist. for Sept. 1869, p. 3. 
