[ 507 ] 
VI. On the Development øf the Flowers of Welwitschia mirabilis, Hook. fil. By Wir- 
LIAM Ramsay McNab, M.D. Edin., Professor of Botany, Royal College of Science 
for Ireland. Communicated by J. D. HOOKER, M.D., V.P.L.S., &c. 
(Plate XL.) 
Read December 19, 1872. 
IN Dr. Hooker’s memoir on Welwitschia (Trans. Linn. Soc. vol. xxiv.) but little notice 
is taken of the development of the flowers of this remarkable plant. This was due to 
imperfection of the material then in Dr. Hooker's hands for examination. Since the 
memoir above mentioned was published, abundance of materials for the purpose was 
obtained by Dr. Hooker; but as opportunity for their dissection failed, the necessary 
specimens were kindly placed by him in my hands; and I now wish to lay before the 
Linnean Society the results of my examination of the development of the flower. In 
some points I differ from Dr. Hooker; and in stating my opinion I do so with the 
greatest diffidence, and only after most careful examination of the material at my 
disposal. 
The development of both flowers has been examined and described separately, as they 
are essentially different in structure. The chief point in which I differ from Dr. Hooker 
is that I consider the peculiar structure in the hermaphrodite (male) flower, which 
Dr. Hooker describes as an ovular integument, to be carpellary. From this, however, 
it must not be supposed that I consider the covering of the ovule in the female flower to 
be carpellary ; on the contrary, I consider it to be a true ovular integument, and the 
plant thus truly gymnospermous. From this it will be apparent that the male and 
female flowers are very differently constructed. As this, however, is not very uncommon 
in the gymnosperms, it is not to be regarded as of special importance in Welwitschia. In 
the male flower the parts are opposite and decussate, two outer parts of the perianth, two 
inner parts, two primordial stamens and two carpels. After the formation of the carpels 
an arrest of development takes place, and the floral axis forms a conical projection 
which enlarges but slowly, and, instead of forming a single terminal and axial ovule, no 
differentiation of the tissue takes place. The female flower is much more simple in its 
construction, as only two lateral parts of the perianth are formed, while at a considerable 
distance from the perianth, and near the end of the axis, a cireular cushion forms round 
the punctum vegetationis, now the naked nucleus of the ovule. No structure resembling 
the carpels of the male flower is formed ; and it seems unlikely that while in the male 
two carpellary cushions form, a carpellary annulus should be formed in the female. 
Development of the Male Flower. 
There is à marked difference between the earliest stages of the male and female 
flowers. In the female flower there is a short basal portion about :004 of an inch long, 
and which is constricted above. Immediately above this constricted part the two por- 
* VOL. XXVIII. 3z 
