table units-species-may at once be discriminated from those diverged forms-varieties-to which they have given rise, and with which, from the important structural differences they frequently assume, they might be hopelessly confounded. Such, at least, is the opinion of those naturalists who regard species as the result of distinct creatiye acts. On the other hand, those naturalists who believe in derivative hypotheses, and look upon all existing organisms as the genealogical connections of other and earlier kinds, entertain the directly opposite view, and maintain that no such essential differences as those above stated exist between the results of hybridism and mongrelism; though they readily admit a difference in degree. This point has been ably and philosophically discussed by Mr. Darwin, who, after a careful and impartial examination of all the evidence he could collate, considers himself justified in concluding, that "first crosses between forms known to be varieties, or sufficiently alike to be considered as varieties, and their mongrel offspring, are very generally, but not, as is so often falsely stated, universally fertile.........consequently that neither fertility nor sterility afford any clear distinction between species and varieties; but that the evidence from this source graduates away, and is doubtful in the same degree, as is the evidence derived from other constitutional and structural differences."

Though Mr. Darwin thus clearly anticipates an essential accordance between the result of hybridism and mongrelism, it is to be observed that the extreme paucity of experimental observations on the latter phenomena prevents his illustrating the subject so fully and satisfactorily as its importance demands. The want of such observations, and the importance of their bearing on that theory of the "origin of species" proposed by Mr. Darwin, has been frequently and strongly insisted on by Professor Huxley. Thus in his "Essay on Man's Place in Nature," p. 106, we find the following remarks: "Our acceptance of the Darwinian hypothesis must be provisional so long as one link in the chain of evidence is wanting, and so long as all the animals and plants certainly produced by selective breeding from a common stock are fertile, and their progeny are fertile one with another, that link will be wanting." Again in his Lectures on our knowledge of the cause of the phenomena of organic nature, Lecture VI. p. 147, after

* Darwin's "Origin of Species," 3rd Edition, pp. 271 and 300.

discussing the obligations of a hypothesis, he remarks, that "Mr. Darwin, in order to place his views beyond the reach of all possible doubt, ought to be able to demonstrate the possibility of developing from a particular stock, by selective breeding, two forms which should either be unable to cross one with another, or whose cross-bred offspring should be infertile with one another," "Now it is admitted on all hands that at present so far as experiments have gone, it has not been found possible to produce their complete physiological divergence by selective breeding..........If it should be proved, not only that this has not been done, but that it could not be done, I hold that Mr. Darwin's hypothesis would be utterly shattered." Professor Huxley, however, though thus strongly insisting upon the absence of facts showing that any degree of sterility has resulted from the crossing of varieties known to have originated from a common stock, states that he does not know a single fact which would justify the assertion that such sterility could not be produced by proper experiment, expressing his belief that it may and will be produced.

Considering then the as yet positively equivocal nature of the relations between the phenomena of hybridism and mongrelism, together with its important bearings on the converse theories which now divide the scientific world, I trust the reader will bear with me, while giving a somewhat detailed statement of my own experiments on the above phenomena. I venture to premise that they show pretty clearly the relative claims of the two views now held by naturalists on our acceptance, and illustrate also one or two other points of high interest in theoretical natural science. First, for the union of V. phoeniceum vars. roseum and album and V. nigrum.

The following descriptive notice of the plants in Tab. 1, will show their close morphological relations. First, V. phoeniceum; stem somewhat downy, simple, producing upwards a racemose panicle. Leaves crenate, oblong-ovate, nearly glabrous above, deep green. Radical subcordate, ovately-acuminate, petiolate. Upper cauline crenulated, semi-amplexicaul. Bracteas lanceolate. Raceme elongated. Flowers lax, solitary; pedicels longer than the bracteas. Corolla purplish-violet, beset with violet hairs at its base. Stamens; filaments of the three shorter stamens covered with long glandular purplish hairs, these of the two longer naked, except on the upper side, where there are a few similarly characterised hairs. Anthers of the three longer stamens nearly circular, and covered with purple and white glandulose hairs, these of the shorter stamens, reniform and nearly naked. Pollen copper-coloured. Second, V. phoeniceum, roseum differs from the above only in the less elongated raceme and the rose-coloured flowers. Third, V. phoeniceum, album is of a more robust habit than the other two.

Radical leaves ovate-lanceolate, light green. Flowers white and rather larger than the others, with a few whitish glandulose hairs near the base of petals. Filaments and form of anthers similar to these of V. phoeniceum, but beset with white instead of purple, glandular hairs. Pollen similarly copper-coloured in each.

Thus, judging from the characters of these three forms alone, there can be no doubt as to their being other than conspecific. In addition to this I may add, on the authority of Mr. Stirling of Edinburgh, that they have been raised from pure seed of the V. phaniceum, the rose-coloured variety frequently appearing amongst the seedlings of V. phoeniceum, the white presenting itself more rarely.

In the first part of Tab. 1, the number of flowers fertilised, and the simple results are shown, and in the right hand, for the sake of comparison, the calculated produce of the number of seeds from 20 capsules of each is given.* If we compare the results, we see that reciprocal unions may be effected becween the V. phoeniceum and varieties, with one exception, viz., V. phæniceum, roseum, by pollen of V. phoeniceum, album, in which case I have found that though the pollen tubes are abundantly developed and freely penetrate the stigmatic tissues, the capsules nevertheless drop prematurely. The goodness, however, of both the male and female elements of the above varieties is nevertheless shown by their reciprocal unions with V. phoeniceum. The individual potency of the respective sexual elements of these varieties, in their reciprocal relations, is clearly shown; whereas by those experiments given in the three last lines of the table, in which the stigmas of each variety were covered by their own good pollen, no unions were effected, each proving utterly self-sterile!

This absolute, or conditional, sterility of the three varieties of V. phoeniceum, when treated by their own good pollen, led me to examine

* From Mr. Darwin's suggestion in "The Origin of species" that the decreased fertility of mixed unions, as compared with that of the pure unions, might possibly be increased by the fact, that for perfectly satisfactory results, castration is necessary in the cross-unions; whereas in the latter, in pure unions, this not being necessary, we may have indiscriminate comparisons, of the two results though clearly castration may have a direct sterilising influence. In view of this prudent suggestion, I took the precaution to castrate every flower both of the pure and mixed unions, from which I intended to draw results. The sole exception to this is that given in the first line of Table 2 of V. phoniceum as I was unable to get any of the plants under me to produce sced by their own pollen. Whatever be the effects of castration then on the fertility of the plants so treated, in the present cases, all having undergone it, the results will be mutual.

into the apparent cause, as in certain cases we find it arising from the non-emission or non-penetration of the pollen tubes; the pollen through some mysterious cause being thus utterly impotent on its own stigma. The results of my present examination will, I trust, be found of sufficient interest to permit of my stating them here. They are as follows: first, I applied the pollen of each of the three varieties, reciprocally, to their stigmas; on dissecting these, I found them abundantly permeated by pollen tubes, many of which I distinctly traced into the ovary. Secondly, I fertilised several flowers in each variety, with its own pollen; on examining the stigmas of a few of these flowers, I found that many of the pollen grains had emitted tubes, but comparatively few had penetrated the stigmatic tissue, and of these still fewer permeated the conducting tissues of the styles. Several of the latter, however, I traced into the vascular bundles of the placenta, the pistillary cords, and in one or two instances, I believe that I detected them in the nucleus of the ovule. Nevertheless we have seen that, though these pollen tubes are developed, they most ineffectively perform their deputed function, inasmuch as not one of these matured even a single ovary! I have here to observe, however, that these pollen tubes do not seem utterly void of the fecundative influence, as many of the ovaries did undergo a certain degree of development; and on examination of these, as they dropped off, I found that the ovules also had undergone a partial and variable degree of development. In general, the fleshy albuminous envelope of the embryo was largely developed, whereas the embryo had undergone a very slight development, judging from a comparison of other good seeds of a similar stage, not at all proportionate to the size attained by the albuminous parts. In nearly all the embryos which came under my observation, the development had ceased ere they exhibited any distinct separation of parts; a few only had reached that stage in which the axial and lateral projections were visible.

We thus see, that whatever be the real cause of the inveterate selfsterility of the three varieties of the V. phoeniceum, it does not arise, as has been shown in other cases, from the non-emission of the pollen tubes. In these, as I have elsewhere noticed it, in certain individual plants of different species of Oncidia, Maxillaria, and Passifloræ, sterility apparently results from some slight differentiation of the male