262 DR J. M. MACFARLANE ON THE 



trees is taken, the amount of wood formed by C. purpureus in the first year is small, 

 while that of the second year is nearly twice as deep. As shown by the figures (Plate VIII. 

 figs. 2, 3), the opposite is true, of C. Laburnum, and C Adami is tolerably intermediate, 

 though it decidedly approaches nearer to C. Laburnum. The fibre elements of the 

 xylem in C. purpureus are open and moderately thick walled ; dense, and with 

 indurated shining walls in the others. From the figures it will be seen that the vasa 

 of C. Adami essentially agree with those of C. Laburnum. 



The pith-cells of C. purpureus are very slightly thickened in their walls, with few 

 pore areas, and accordingly show very delicate markings of shallow pits, while their 

 outline is sharply polygonal. The cells of the other two are rounded and thick walled, 

 so that they show correspondingly deep and well-marked pore canals. 



The above description proves that the hybrid part is most nearly like C. Laburnum 

 in its stem configuration and upbuilding, but a very remarkable feature is revealed 

 when careful preparations are made from the young epidermis of the three. We 

 have already indicated that the epidermal cells of the hybrid approach most nearly 

 to C purpureus in size and amount of cuticular thickening, but the manner in which the 

 cell-nuclei of the two agree is quite striking, when we compare them with the nuclei of 

 C. Laburnum. Plate VIII. figs. 10, 11, 12 illustrate these as perfectly as figures can. 

 Alike after staining in aqueous eosin solution from the fresh living state, as after careful 

 hardening in picric alcohol, the large spherical nuclei of the two former stain deeply, while 

 they have a finely granular and spongy aspect. Those of C. Laburnum are small, shining, 

 and pretty homogeneous in texture, even under very high powers, as if the chromatic 

 substance of the nuclear membrane were specially abundant. The above description 

 applies equally to the nuclei of the leaf epidermis. 



After satisfying myself that this held true of the epidermis, the idea occurred of com- 

 paring the nuclei of internal elements, such as those of cortex, phloem, &c. So far this 

 has been without result, as the nuclei vary greatly even in adjoining elements. But still 

 it is very significant and suggestive that we should have tissues of one plant — no matter 

 what its origin — with at least two very distinct types of nuclei in cells which work 

 harmoniously together, so far as vegetative growth is concerned. This fact alone com- 

 pletely demonstrates the extremely elastic adaptability of protoplasm and its modifications 

 in the formation of vegetative structures. 



Leaf. — The three leaflets of the compound leaf of C. purpureus are small, elliptic- 

 lanceolate, glabrous, and somewhat fleshy. Darwin (op. cit., p. 414) states that a twig 

 of the purple form from a composite tree has " the leaflets a little broader, and the 

 flowers slightly shorter, with the corolla and calyx less brightly purple" as compared 

 with the ordinary species. There is a strong probability that this is a variation condition 

 of some of the plants, just as the Dunkeld specimen has varied by becoming double in its 

 " Adami " flowers, for while most that I have studied agree with Darwin's account, a 

 branch bearing all three forms which was brought to me during the past summer by 

 Dr Scott of Melrose agreed in the above, as in other features, with specimens of C. pur- 





