sporophylls perfectly developed. It would seem, then, as if the greater development of the sporophylls occurred when the nutritive conditions were most perfect. May we infer that, if the nutritive con- ditions remained equally satisfactory throughout, the whole of the sporophylls would show repeated dichotomy ?
Professor Bower has observed that those abnormalities which occur most frequently in a given species will be those which are most worthy of consideration for morphological argument. If we accept this statement as a reasonable one, the variations now under dis- cussion appear to deserve serious attention. For they occur with considerable frequency, as I will show ; and though we may, in a sense, apply the term abnormality to them, it is in a sense with which nothing of the nature of a pathological variation can be asso- ciated. They are of healthy appearance, and occur in the most vigorous parts of the best shoots. I have found as many as five or six sporophylls with repeated dichotomy in a single fertile zone, whilst shoots with three such sporophylls are fairly numerous. In the course of half-an-hour's search in the forest I can always count on finding several shoots with such variations. But this applies only to those districts where favourable conditions for the growth of Tmesipteris are best realised ; in less favourable localities the varia- tions may certainly occur, but not in such numbers. It is probably not going too far when we assert that, when conditions are most favourable, the sporophylls of Tmesipteris normally show a repeated dichotomy.
2. In a second group of variations we find that the synangium, instead of being sessile on the petiole of the leaf just below the fork, that is, the point from which the leaf-lobes diverge, is raised up on a longer or shorter stalk. I have found a considerable number of these variations, and they show a good deal of diversity. Some- times the synangium is carried up to a height equal to half the length of the synangium, and lies transversely to the axis of the whole leaf. The two lobes of the synangium then appear to be balanced on the summit of the pedicel, hanging down somewhat, one on each side, so as to suggest a peltate sporangiophore. In other cases the synan- gium retains its direction between the leaf-lobes, but revolves on a transverse axis, so that the longitudinal groove of the synangium, by which dehiscence takes place, faces outwards between the leaf-lobes, instead of looking rather towards the axis of the shoot. One would think that the outward position would, on the whole, be more favour- able for the dispersal of the spores.
3. In the third group of variations no synangium appears, though the leaf has otherwise the character of a sporophyll, and is forked. Professor Bower has described, and I, too, have seen, many such sporophylls in which a minute scar appears in a position below the fork,
