128 Trans. Acad. Sci. of St. Louis. 
change which we may imagine has been thus produced in 
Ocymum. But exposure of ‘the anthers and stigma in irregu- 
lar crowded flowers is equally degrading both to those origi- 
nally nototribe or ster ikea proposition which I have 
already propounded in the Bot. Gazette XIII, 230. 
In addition to being a most important organ in protecting 
the anthers and stigma and in preserving the original Labiate 
character of the flower, the galea, from its more or less hori- 
zontal position, also plays an important part in rendering the 
nectar less accessible. When, therefore, we find the flowers 
crowded in a flat-topped inflorescence, the upper lip reduced or 
reflexed in such a way as to expose the anthers and to render 
the nectar more accessible, we have the most degraded of the 
family and the form farthest removed from the type. 
The fact that Mentha and Pycnanthemum have a more 
regular form than Scutellaria is no reason for supposing that 
it approaches the form from which the Scufellaria was de- 
rived. If a form like Mentha had its flowers exposed hori- 
zontally and separated so that each flower must be visited 
separately, I think it would result in sternotribe zy gomorphism, 
like that of the Papilionacew,—a form almost always char- 
acterizing lateral flowers with exposed stamens. But I could 
hardly imagine any process by which closely crowded flowers 
with exposed organs would give rise to a group of flowers 
characterized by having the anthers protected under the upper 
wall of a deep tube. In the paper above referred to, I have 
expressed the view that small closely crowded regular flowers 
do not tend towards zygomorphism. 
The modifications which we have considered have a most 
important influence in determining the kinds of insects fre- 
quenting the flowers of different species of the Labiate. 
There remains to be mentioned the varying length of the tube. 
In some species in which the general characters of the flower 
are of a low grade, a rather highly specialized set of visitors 
is retained as a result of a great lengthening of the tube, as 
in the case of Monarda. 
In the table I have arranged the flowers as nearly as practi- 
cable in the order of their departure from the type, beginning 
with those which have a more perfect Labiate form, simpler 
