element with respect to its own female element. I have also to remark, that the ultimate conditional sterility of these plants is not, relatively considered, an absolute but a graduated quantum; this is shown by the different degrees of development the embryos had undergone, thus illustrating a most interesting, though as yet imperfectly known fact, namely, that the male element, even though reaching the female element, may nevertheless fail to communicate that amount of vital stimulus necessary to the complete development of the embryo. Furthermore, I may in passing briefly refer to the perfect parallelism between these phenomena, and those occasionally observed in hybridisation, at least in the zoological kingdom, for unfortunately we are as yet nearly void of information on this point in the vegetable kingdom, hybridists having, in most instances, satisfied themselves by attending to the ultimate results, without troubling themselves to examine into the nature or degree of embryonic sterilisation. From the published papers of the Hon'ble and Rev. W. Herbert, we find, as might indeed be expected, that this point did not escape observation: thus in one case he remarks, " It has, I believe, not been duly considered, that the fecundation of the ovules is not a simple, but a complicated process. There seems to me to be three or four several processes: viz., the quickening of the capsule of the fruit, of the outer coats of the seed itself, of the internal parts or kernel, and lastly, the quickening of the embryo."......... "It is further to be observed," he continues, "that there is frequently an imperfect hybrid fertilisation, which can give life, but not sustain it well. I obtained much good seed from Hibiscus palustris by H. speciosus, and sowed a little each year till it was all gone, the plants always sprouted, but I saved only one to the third leaf, and it perished then."

To recur, however, to the above parallelism, of which we have here additional and important illustrations: it has been stated by Mr. Darwin* on the authority of Mr. Hewitt, that in the hybridisation of gallinaceous birds a frequent cause of sterility in first crosses is the early death of the embryo. Again Mr. Salter records similar results from his experiments on the fertility inter se of several hybrid Galli,† thus concluding, "the one striking point of these experiments (which I believe has never been noticed before) is that a large proportion of loc. cit. p. 286.

+ Nat. Hist. Rev. 1863, p. 276.

these eggs from hybrid birds breeding inter se have failed to produce young, not from absolute sterility, but sterility in degree, from an amount of vitalization insufficient to carry out the whole result of reproduction, in which the young individual has been completed, leaving it with vital resistance insufficient to maintain life and cope with common and customary external influences." And thus in those curious cases of sterility of structurally hermaphrodite organisms, whose sexual elements have become differentiated with respect to their mutual fertile conjunctions, so in the phenomena of sterility from hybridism, we find, as Mr. Salter well remarks, with respect to the relations of hybridism and parthenogenesis, "that the sterility is not absolute but in degree, and that the stimulus, whatever it may be, which starts the embryonic changes is feeble and imperfect rather than wholly wanting."

I have now shown that a regular more or less early embryonic abortion results from the self-fertilisation of certain individual plants of V. phoniceum and vars. roseum, and album; whereas by their reciprocal fertilisation, highly fertile unions may in general be effected. By again consulting Table 1, however, it will be seen that besides a reciprocal fertilisation, these three plants are also susceptible of fertilisation by pollen of other species. Thus in lines 7, 8, 9, of Table 1, the male element of V. nigrum is singularly enough effective in the fertilisation of each, while in a succeeding Table-4-the goodness of the male elements is also similarly shown by each effectively fertilising the female element of the V. lychnitis, lutea. Again, we have fuller illustrations of these curious sexual phenomena in Table 2, in which one of the above plants, V. phoeniceum, yields a varying degree of fertility to four other distinct species; namely the V. ferrugineum, Blattaria lutea and alba; Lychnitis lutea and ovalifolia. These are indeed remarkable physiological revelations. How strange that an individual plant could be fertilised by the pollen of five distinct species, and yet not by its own good pollen: how singular also, as shown above, to see three hermaphrodite individuals incapable of self-fertilisation, yet having each sexual element reciprocally meeting and fertilising the opposite elements of other species. Thus, for example, the male element of V. phoniceum and vars. roseum and album fertilise the female element of V. lychnitis, while the female elements of the three

former are also susceptible of fertilisation by the male element of V. nigrum. The full explanation of these curious and complicated sexual relations, I leave for more sagacious and ingenious investigators, and simply confine myself to remarking on the apparent support that these and more especially those other cases which I have communicated to the Linnean Society,* on the fertilisation of certain species of Passiflora, -in which I showed that individual plants perfectly self-sterile readily effected reciprocal unions with other similarly characterised individuals of the same species-give to that view which Mr. Darwin has propounded regarding the existence of a law in nature necessitating "an occasional cross with another individual, or, that no hermaphrodite fertilises itself for a perpetuity of generations," but "that some unknown great good is derived from the union of individuals which have been kept distinct for many generations."+

In the following table, the results of the pure unions of V. phoeniceum given on the first line are taken from capsules on a specimen in the Edinburgh University Herbarium, as I have not yet been successful in getting good capsules from any of the plants which I have had an opportunity to experiment upon by their own pollen. The other plants of V. phoeniceum and varieties mentioned in the table are the same as those from which I had the results given in Table 1. Indeed, in one or two instances, the same experiments are re-stated, with a view to show more clearly the relative degrees of sterility resulting from the crossing of undoubted varieties of a species on the one hand, with those from the hybridisation of distinct species on the other.

In addition to the simple and calculated results given on Table 1, I have, in the above, given at the right hand, for the sake of comparison, the calculated product from an assumed 1,000 seeds of the pure unions relatively to those yielded by the cross and hybrid unions. By a further comparative study of these results, we find that the fertility of the pure unions of V. phœniceum, relatively to that of its cross-unions with the white and rose-coloured varieties, is, in the least differentiated or most highly fertile unions, viz., V. phoeniceum, rosea by pollen of V. phoeniceum, as 100: 95; whereas in the least fertile unions, V. phoeniceum by pollen of V. phœniceum, alba, the proportions are as 100: 56. The average fertility of the five cross-unions given in the table, relatively to the pure unions given in the first line, is as 100: 75; so that the pure unions thus exceed in fertility the cross-unions, in nearly the proportions of 4: 3. Again

by a similar comparative study of the relative fertility of the pure unions of V. phoeniceum and the different hybrid unions given in the Table, we find that the highest degree of fertility results from the union of V. ferrugineum (which perhaps is correctly regarded by De Candolle and others as a mere variety of V. phoeniceum) with V. phoeniceum, the proportions of the pure to the hybrid unions being as 100: 59, in favour of the former. The lowest degree of fertility results from the unions of V. ovalifolium, with V. phoeniceum, the proportion of the pure to the hybrid-unions in this case being as 100 24.) Lastly the average fertility of the five hybrid unions given in the latter lines of the Table, relatively to the pure unions of V. phoeniceum, is nearly as 100: 40, or as 2.5 seeds of the pure unions to one of the hybrid unions. Thus, the relative differences in the degree of sterilisation resulting from the hybridisation of distinct species, and that from the cross-impregnation of varieties of a species, relatively in either case to the pure unions, is in the former as 2.5: 1, and in the latter as 4: 3.