^{page 639 note *} Vegetable Kingdom, p. 686.

^{page 639 note †} Leçons sur les Fam. Nat. des Plantes, p. 14.

^{page 641 note *} Exceptions are sometimes met with. I have seen the posterior sepal overlapping only one of the lateral; or one, or both of the lateral sepals wholly external. A hasty observation of such an exception as the last, probably led Payer (Leçons, p. 14) to describe the æstivation of the calyx as quincuncial, which I can hardly believe it ever is.

^{page 643 note *} The slight bilabiation of the ovarian orifice seen in Plate XXVIII. fig. 9, though real, is doubtless in appearance considerably exaggerated by this antero-posterior narrowing.

^{page 644 note *} This flower was unfortunately detached before I had ascertained whether the sepals were antero-posterior or lateral.

^{page 645 note *} I have met with three instances of this bipartite condition of the posterior lip.

^{page 545 note †} To the left of an observer supposed to stand in the axis of inflorescence.

^{page 645 note ‡} I have in my possession a third example of an ovary with tripartite anterior lip, but as I have been unwilling to remove the stamens from the specimen, I cannot say what appearance is presented on an anterior view; its posterior aspect, however, is almost identical with that given in Plate XXX. fig. 27.

^{page 645 note §} This antero-lateral fissure is uncomplicated by any lobule.

^{page 646 note *} The variable and inconstant lobules at the base of the middle anterior lobe in this form of monstrosity I am, I think, justified in considering of secondary importance.

^{page 648 note *} Wight (Icones pl. Ind. Orient, t. 1621), Endlicher (Genera, p. 349), Lindley (Veget. Kingd. p. 652), and Payer (Leçons, p. 14) agree in describing Salvadora as having a unilocular ovary with solitary erect ovule. Professor Oliver has kindly examined for me flowers of S. persica, L., and S. Wightiana, Pl., from the Kew Herbarium, of which he reports in a letter as follows:—“In each of these I find a 1-celled ovary with a solitary basal ovule.” My own somewhat limited examination of the flowers of S. persica has led me to the same conclusion. On the other hand, Planchon (Sur les Salvadoracées, Ann. des Sc. Nat. 3^me serie x. p. 190), and more recently MM. Maout and Decaisne (Traité de Botanique, p. 453) describe the ovary here (Planchon in the genus Salvadora, Maout and Decaisne in the order Salvadoraceæ) as bilocular, with two collateral ascending ovules in each cell. The only explanation I can suggest for the statement in the “Traité de Botanique,” is that the authors have probably followed Planchon, for M. Decaisne had formerly described S. oleoides as having “ovarium … uniloculare, loculo uniovulato” (Jacquemont Voyage, p. 140, t. 144); while M. Planchon's description is so opposed to the results of other botanists, and so unlike anything I myself have been able to see, that I am constrained to believe that it was some other plant, and not Salvadora, that he examined. I should mention, however, that Decaisne (Jacquemont Voy. t. 144) gives a figure of a fruit of S. Madurensis containing three seeds.

^{page 649 note *} The observations on the embryo of this species were made after the paper had been submitted to the Society.

^{page 649 note †} Morphologie, pp. 755–6.

^{page 649 note ‡} In a communication to a meeting of naturalists, at Freyburg in Br., of which I have seen no report, but which is referred to by Treviranus in his subsequent paper in the Bot. Zeitung, 1848.

^{page 649 note §} Botanische Zeitung, 1848, p. 444.

^{page 650 note *} Treviranus' figure of the embryo from the seed is somewhat faulty, from the cotyledon being represented as considerably too short in proportion to the radicle, and from the absence of any indication of the rudimentary plumule. There is also no indication of the plumule in his figures of the earlier stages of germination, the result, doubtless, of imperfect observation (loc. cit. t. iv.). He also makes a curious blunder in describing the apex of the embryo as pointed towards the hilum of the seed (loc. cit. p. 442), the fact being that in this, as in all anatropal seeds, the apex of the embryo points away from the hilum, the radicle being directed towards it. This mistake is probably due to the circumstance that there is often a projecting portion of the testa at the chalazal extremity, which is apt to be mistaken for the somewhat similar projection at the hilum.

^{page 650 note †} I think it not improbable that back views of this embryo may have had something to do with the statement found in most of the books, that there are two “cotyledones brevissimæ” in Pinguicula.

^{page 650 note ‡} I should mention that a very brief statement, by me, of the differences between the embryos of P. vulgaris and P. grandiflora, has already appeared in the report of a meeting of the Dublin Microscopical Club (“Quarterly Journal of Microscopical Science,” viii. pp. 121–2). I now take this opportunity of describing them in greater detail, and with figures.

^{page 652 note *} In this figure, as also in fig. 31, the capitate hairs scattered over the surface of the ovary are not represented.