238 DR J. M. MACFAELANE ON THE 



/. Menziesia and the cells of the anther wall are produced into long papillae, most of which are con- 

 va^ 6 B ' stricted above their point of origin. In 8 the anther tails are 3 to 5 mm. long, and the 



enetnB. 118 cells of the anther wall are beautifully intermediate in length of the papillae. A com- 

 ^h°ama«htus n parative study of these stamens is highly instructive. 



Pistil. — In 1 the ovary is finely hirsute, and at its base there are pouch-like nectaries, 

 joined at their orifices by secreting cells. In 3 the ovary is glabrous, and round its base 

 occur deep, isolated pouches, with large apertures. In ^ a few hairs occur over the ovary, 

 but greatly less than half as many as those in 1* The nectar pouches are smaller than 

 in 3, and, so far as I have observed, they agree with 3 in being isolated. 



(g) Bryanthus erectus, x . 



For several reasons I was induced to undertake the examination of this beautiful form. 

 First introduced to the notice of botanists by Mr Cunningham, of Messrs Cunningham 

 & Fraser's nurseries, it was handed by him to the late Professor Graham for examina- 

 tion, though nothing was said as to its origin. As it differed entirely from any known 

 species, Graham, ignorant of its true nature, described it as a new species. The 

 hybridiser, at that time a youth, stated positively that he had obtained it by crossing 

 Menziesia empetriformis, var. Drummondi, with pollen of Rhododendron Chamcecistus. 

 Great doubt has been expressed by competent authorities as to the accuracy of this 

 statement, and these doubts may have been strengthened, I believe, by the statement 

 having been circulated that Menziesia ccerulea — not empetriformis — was one parent, 

 and it has even been asserted that the plant has been found in a truly wild state. 

 Careful and continued observation of the three plants, hybrid and parents above given, 

 would furnish strong reasons for believing in the hybrid origin of Bryanthus, since the 

 stature, habit of growth, leaf form, flower numbers, and colour, as well as its failure in 

 my experience to produce any quantity of seed, while both reputed parents do, are all 

 in favour of Cunningham's statement. I shall first describe the stem shortly, though 

 the foliar parts are of greatest importance. 



Stem. — Selecting young first year's shoots of each, we find on longitudinal view of 3 

 that the epidermal cells are very irregular in outline, have secondary wall thickenings 

 with rather close-set pore canals, and the free surface exhibits fine ridges as in Lapageria. 

 Stomata are abundant, as many as seven to eight being the average under the Zeiss' D, 

 and still more abundant are short, curved, thick-walled hairs. Long-stalked glandular 

 hairs, each ending in a small terminal knob, occur sparingly. In 1 the cells are elongate 

 in outline, the pore canals of the walls are unevenly and often distantly placed, and the 

 free surface is quite smooth. Stomata are rarely if ever developed, and only a moderate 

 number of curved hairs are distributed over the surface. Glandular hairs, composed of a 



* It is possible that an explanation of this is got in lines of variability which 1 sometimes shows. I have pointed 

 out {Trans. Hot. Soc. Edin., 1891) that varieties of the species may be nearly glabrous. 



