EXPERIMENTS IN HYBRIDISATION, &c. 97 
P. caudatum, for a flower once appeared on a typical P. caudatum 
midway between this curious variety and the type.* This curious 
variety, P. c. Lindenii, is interesting from another point of view. In its 
native home it apparently breeds true to character, yet when crossed 
with other species its curious characters do not appear to be reproduced 
at all. For instance, it has been crossed with P. longifoliam, giving P. 
x grande macrochilum (fig.14); with P. x conchiferum, giving P. x Clonius; 
with P. x Ainsworthii calurum, giving P. x Penelaus (fig, 13) ; and with 
P. x grande, giving P. x macrochilum, and in no case has its 
curious lipless character been transmitted. It would be interesting if 
someone would fertilise this curious variety with its own pollen, or cross 
it back again with one of the four hybrids of which it has been a parent, 
and record the result. 
Many more instances of inheritance and non-inheritance of varietal 
characters in the Orchidew might be recorded, but space will not permit 
any more. Suffice it to say that a general survey of the whole of the 
facts has brought me to the following conclusions :— 
(1) Distinct varieties tend to transmit their qualities, especially if 
fertilised with their own pollen, though exceptions are not rare. 
(2) The chief exceptions seem to arise where the parents or the 
ancestors of the variety have been variable. 
(8) Slight variations are seldom hereditary. 
(4) Abnormal sports are generally transmitted wholly or not at all. 
(5) Distinet varieties, as a general rule, transmit their qualities in 
different degrees, sometimes wholly, sometimes partly, and at other times 
not at all. 
(6) Varietal characters can seldom be traced in the second or follow- 
ing generations, unless they happen to recur on both sides of the 
pedigree. 
(7) The law of Partial Prepotency, elaborated further on in this 
paper, may possibly account for these varied results in the inheritance 
of varietal characters. 
Tue INHERITANCE OF SPECIFIC CHARACTERS. 
Those who have studied hybrids between distinct species must be 
impressed with the undoubted inheritance of specific characters. 
Varietal characters, while perhaps of more practical importance, are 
yet so indefinable, so uncertain, and so fleeting that in the second 
generation they are with difficulty traced at all. On the other hand, 
specific characters are more definable, more certain, and more lasting, 
and can be traced through several generations. For instance, in 
Paphiopedilum x triumphans, a hybrid of the third generation raised by 
M. Jules Hye-Leysen, of Ghent, the crimson veining in the upper sepal 
of the flower can be traced through the parent, P. x enanthum superbum, 
and the grandparent, P. x Harrisianum, back to the great-grandparent, 
P. barbatum. 
In studying the inheritance of specific characters, I have found it a 
great advantage to take up a special group of plants and to study their 
* See Dr. Maxwell Masters, in Veitch’s ‘‘ Manual of Orchids,” x. 1894, p. 45, figs. 6, 7. 
H 
