apparently established. As said above; geographical changes of an abrupt type are probably caused by isolation of lines. The result of isolating lines is well seen in the cultivated cereals. In natural conditions isolation is much less frequent; in fact it would not be too much to say that Nature abhors a pure line, just as she abhors the inter-specific cross. And it is these two facts alone which secure the stability and continuity of the Linna^an species. Nevertheless varieties due to isolation of lines without geographical isolation almost certainly do occur,—for instance Scirdus supinus var. uniondis Cke, and Cyperus lria var. panicifonais Cke.
Sixthly there are the not infrequent cases of species and genera which are in an active state of evolution, producing a maze of forms among which clear specific lines of demarcation are not discernible. These forms may be the De Vriesian mutants. It would be beyond the scope of this paper to discuss this point; but it is necessary for the argument to say that whereas "lines" seem to be the product of isolation and combination of already existing, stable allelomorphs, the changing evolutionary forms now referred to would seem to be caused by instability of the allelomorphs themselves, which in such cases are disappearing and reappearing with changed properties.
Seventhly (although they stand on quite a different footing) it is necessary to refer to interspecific hybrids, which though normally abhorred by Nature, do occur, and will ultimately be detected in India as they have been in Europe.
Now, omitting hybrids, our floras have for the most part in the past called all these varied forms of whatever origin by the one word variety. It is true that in cases of species with a very great range of variability the forms have sometimes not been given a varietal name, but simply listed under numbers or letters. But it would appear that this has been due more to convenience than to a clear discernment of the different phenomena.
Before considering the use of terms for units below the species it is desirable to consider whether we are justified in using the same term "species", undifferentiated, for the Linnaean group of individuals in which the allelomorphs are stable as well as for those in which they are unstable. The writer would suggest that the two phenomena be distinguished as "species (constaus)", and "species inconstans", omitting in practice the word constans.
For units below the species the following scheme is suggested:—
A. For freaks or sports no special term, each one to be separately described.
B. (1) For geographical forms, where sharply distinguishable, the trinomial system now used by ornithologists.
