REPORT OF THE CONFERENCE ON GENETICS. 95 
kind more economically important than these, arising in ways not yet 
properly understood. I allude to the appearance of reversionary “ rogues ”’ 
among seed-crops, where circumstances preclude the idea that we have to 
deal with mere recessives, and make it prima facie unlikely that crossing 
is the provocative cause. For example, in the case of peas, such 
reversionary and wild-looking “rogues’’ with rvownd seeds have been 
shown me by my friend Mr. Arthur Sutton amongst crops of highly 
bred wrinkled peas. They are regarded as indications of that general 
degradation or degeneration which it is supposed would permeate all 
highly bred stock if selection were suspended. Now, though it is certain 
that in practice if the crops were neglected these hardy and productive 
“rogues ’’ would soon prevail and overwhelm the pure and more delicate 
strain, we are no longer content to regard their presence as inevitable. 
In order to cope with them we must find out exactly what they are. By 
the strict method of breeding from individuals under proper precautions 
we have now the means of doing this, and not till such investigations have 
béen made need it be regarded as the inevitable property of any high-class 
variety to produce “rogues.’’ Though as to this special case I make no 
prophecy, modern observations strongly suggest the paradoxical conclusion 
that there is no such thing as general degradation or degeneration. ‘These 
phenomena are due to specific causes, most commonly to nothing more 
obscure than insect-crossing, or to unsuspected mixture with an un- 
recognised variety. I mention these things simply to illustrate the fact 
that though the precise physiological nature of reversion may seem a 
matter remote from practical life, it is not remote at all, but closely 
bound up with very important industrial considerations. 
I have said that reversion on crossing is due to the meeting of long- 
parted factors. Conversely, variation is often due to the separation or 
elimination of factors. In other cases it is almost certainly, though 
perhaps not quite certainly, due to the addition of new factors. Genetic 
research has thus provided the first indication of the physiological process 
which results in the birth of a variation. The consequences of this 
knowledge to the systematist and to the science of evolution I will 
not now pursue. By following the clue which the discovery of unit- 
characters has provided, the long range of phenomena first grouped in an 
orderly fashion by Darwin in “ Animals and Plants under Domestication ”’ 
can at last be subjected to precise inquiry. The proximate significance of 
many of these mysteries is indeed already made out. Only those to 
whom that treatise has long been a kind of “De Occultis Nature 
Miraculis”’ are able to appreciate what the new knowledge means to 
biological science. 
Now once more as to the practical importance of all this. The 
breeder has two main objects in view: he wants to create novelties and 
to fiz them. In the second of these objects he can, as we have already 
seen, expect help from Genetics. As regards the creation of new forms 
I must not speak so confidently. Nevertheless, there is a valuable class 
of novelties which are really novel only in so far as they recombine 
pre-existing characters of known types. Such recombinations, say of 
hardiness with desirable qualities of colour or shape, or of size or free- 
flowering Labit with brilliancy, or of colours such as red and cream-yellow, 
