118 Correspondence. 
In his discussion as to the relation between the perianth and 
the bracteoles in the Coryleae and Betuleae, the writer seems almost 
driven into a corner in his attempt to account for the origin of the 
perianth in the former group. Finding a correlation in their protec¬ 
tive function to exist between the bracteoles and the perianth he 
feels bound to associate the two sets of organs together in their origin, 
and, discovering the presence of both perianth and bracteoles in the 
female flowers of the Coryleae, he concludes that “ if it is not possible 
to regard the perianth as a new foliar organ sui generis we must seek 
its origin in additional bracteoles.” This mode of origin of the 
perianth by the intercalation ex nihilo of additional foliar organs 
between those already present at different levels of the axis would be, 
to my mind, a quite unnatural and impossible one. For whence did 
these extra bracteoles arrive on the scene? Why not recognise the 
far more natural, and, to my mind, far less forced method of derivation 
by means of sterilisation of the stamens in the ancestral hermaphro¬ 
dite flower ? 
In treating of the Fagaceae, the writer remarks : “ In the male 
flower the numbers in the two series, perianth and androecium, vary 
in the same direction, an increase in the number of perianth-leaves is 
associated with an increase, not with a decrease, in the number of the 
stamens.” Here lie obviously refers to the case, mentioned by him, 
of Nothofagus obliqua , in whose flowers both stamens and perianth- 
segments exist to the number of 30 — 40 , which is in contrast to other 
genera, in which they are often only from 4—7 in number. I regard, 
however, with Celakovsky, such forms as Nothofagus, possessing a 
large number of stamens, as primitive, while the 4—7 stamens and 
perianth-segments of other genera must be regarded as the result of 
reduction from the larger number; hence it is clear that in a case 
such as Nothofagus where an increase in the number of perianth- 
segments obtains, there can be no possible question of an increase in 
the number of stamens; on the contrary, I hold that very consider¬ 
able decrease in their number must have taken place in order to pro¬ 
vide the numerous perianth-leaves which are present. 
With regard to the Juglandaceae, Dr. Rendle says: “There is a 
strong suggestion of homology of perianth and bracteoles, and it is 
difficult to imagine a different origin for the two sets of leaves.” This 
is precisely my own view, and I may as well here state once for all 
my firm conviction that all foliar organs (even such curious, inde¬ 
pendent-looking structures as the cup substending the androecium in 
the Poplar) which are in close proximity to the flower, have, within a 
comparatively recent period, been derived from sporophylls. 
Dr. Rendle himself discovers and acknowledges the production 
of unisexual flowers by abortion of the stamens in the hermaphrodite 
flowers of the Saurureae. But he remarks that: “ Abortion of the 
stamens . . . never results in the formation of a perianth.” 
Abortion of any organ could surely hardly result in anything but its 
annihilation ! And the absence of a perianth is clearly correlated 
with the development of the bract as a protective structure for the 
androecium and ovary. 
In the case of the Ulmaceae, &c., the difficulty raised may be 
easily and quite naturally surmounted by assuming, as in the case of 
Primula, the abortion of an outer whorl of stamens. 
Turning now to the point raised in the penultimate paragraph of 
