Over a wide tract of country there must have been millions of these plants, and out of the enormous number which came under observation during a month and more no individuals were seen which could not be at once allocated (even from horse-back) to one or other of the two types. Now if these types were intra-specific, then their contrasting characters would be allelomorphs. I regret that it did not at the time occur to me to observe the methods of fertilisation ; but there is reason to expect that a typical papilionaceous flower with diadelphous stamens is liable to cross-fertilisation. If so, and if the contrasting characters were allelomorphs, then the occurrence of two separate types each of them exhibiting the whole set of contrasting characters, without the occurrence of even one individual which could be classed as heterozygotic in even one pair of characters, would, I suppose, be an impossibility. However, even if Alysicarpus were shown always to be self-fertilised, still the absolute constancy and uniformity of these two sets of characters in the same environment would, I fancy, have convinced anyone who had seen them that we have here two species.
In a single homogeneous tract like the black soil referred to the habit of a species is likely to remain fairly constant. But there is little doubt that most of the species of Alysicarpus are highly plastic and react strongly to environment ; and this plasticity is probably the main cause of the confusion which now exists as regards the species with the rugose lomentum, since the superficial (somatic) differences between individuals of the same species are often more obvious than the morphological (germinal) differences between the species.
The systematic notes which follow are based on the field observations referred to as well as on the material in the herbarium of Mr. T. R. D. Bell, CLE., and myself, and the material in the herbarium of the St. Xavier's College, Bombay, collected by Father Blatter and Mr. Hallberg, and very kindly placed at my disposal by those gentlemen. This material is all from Western India at various points from Mt. Abu to Madura. It will be seen that I discriminate three certain and one probable species. But I should be far from suggesting that a study of material from the whole of India would not necessitate the recognition of further species. This paper is intended to be suggestive only, and is published in the hope that other workers may find it of use in a further and more complete study of this particular systematic problem. Especially has it been quite out of the question to attempt the unravelling of the intricate synonymy of these plants ; indeed it is doubtful whether this work could be done anywhere but in Europe, where alone the numerous type sheets may still be available.
