406 REPORT OF THE CONFERENCE ON GENETICS. 
typical descendants. Such a plant with its descendants would constitute 
a ‘‘good”’ species, such as earlier botanists believed in, but which we 
now know has no existence outside the imagination of some. No! Quite 
clearly is it borne in upon us that abnormal or variant cells exist in 
every organism, and it is to the union of such cells inter se, or with those 
which are normal, that we must trace every variation from the parent 
occurring in self-fertilised plants. In such cases the union of normal 
cells produces typical offspring ; the union of abnormal cells with normal 
cells produces non-typical or variant offspring (instances of mutability) ; 
and the union of two abnormal cells (where such union can bear fruit) 
may reasonably be supposed to produce deformed, erratic, and often sterile 
progeny. 
Perhaps the comparative fixity of some species may be due to the fact 
that the abnormal cells are, in such cases, not only few but also unable 
to mate with those that are normal, or that such unions are sterile. 
Again, the extreme mutability of florists’ plants may be due to the great 
relative number of abnormal cells rendering almost every class of variant 
possible among the offspring. 
It is the task of the exponents of Mendelism to discern whether 
cell-union mutability in its widest sense is really governed by any law 
which can be made clear to us by mathematical expression. 
Now, it seems to me that the exponents of the Mendelian thesis 
have established a prima facie case, but i do not think that up to the 
present they have done more than this. The majority of their experi- 
ments have unfortunately been carried out among extremely mutable 
plants of garden origin, rather than among comparatively stable species. 
In my own experiments I have dealt with this latter class of plants, 
and I cannot say that my results have done much to establish their 
contention. In general I have found that inter-specific hybrids are 
fairly equipoised between their parents, without showing a complete 
dominance of one over the other in any one character. In cross-bred 
plants I have noted the tendency of a certain percentage to revert to 
the ancestral (specific) types. This is notably the case in the Garden 
Hippeastrums, in which reversion to H. vittatwm, or at least to the 
colour-markings of H. vittatwm,is common. If any Garden Hippeastrums 
in which white is the predominating colour be self-fertilised, several 
individuals will probably revert to H. vittatwm as far as colour and 
markings are concerned, although retaining the size, and often the width 
and regularity, noticeable in the perianth segments of the parent. Such 
reversion is not to any H. vittatwm which I have ever seen wild, but rather 
to a glorified form of the species, such as might have been produced in 
gardens by many generations of careful selection. In fact, what I have 
noticed is a colour reversion rather than a general reversion, and does not 
bear out the Mendelian law in its entirety. 
In general I have noticed that cross-bred (mongrel) plants which 
possess great colour-yariety continually produce from seed a fair percentage 
of individuals bearing the ancestral colour or colours. These plants are 
always the best seed-bearers, and those most widely removed therefrom in 
colour are generally infertile or nearly so. ‘This is specially noticeable 
in Cinerari. 
