Flowers of Stachys sylvatica, Linn. 413 
lobe, was present (Fig. 3, f). These specimens resembled two of the 
semi-peloric forms mentioned above, and three of the four were recorded 
as side flowers ; the one with no record against it was probably a side flower 
also. There was one flower, a middle one, of a stalked side inflorescence 
which was pseudo-terminal owing to the destruction of the adjoining side 
inflorescence and of the main axis. The flower developed in an almost 
symmetrical manner and had an increased number of parts (Fig. 3, c). 
The remaining variant (Fig. $ } g) had its upper lip in line with the lower 
lip. The style was in line with the corolla tube, and the stamens, four in 
number, did not show above the tube. Flowers like this one were com- 
paratively common in the autumn. 
Fig. 3. For description see Table III and text. [Shoreham, Kent, 1919.] 
In the middle of September of 1919 (see Table IV), on loamy soils 
near West Wycombe, only two symmetrical (peloric) flowers were found, 
a few fused flowers, some with the upper lip half fused with the right or 
left side petal of the lower lip, a comparatively large number of the type 
of flower described last in the preceding paragraph, i. e. small flowers with 
stamens never growing beyond the tube (Fig. 4, d, c), and some of these 
with shortened styles as well ; an examination of these flowers preserved in 
alcohol showed either no signs of pollen grains or small, shrivelled, and 
probably abortive ones. An examination with a hand-lens in the field had 
suggested already that many of these anthers were aborted. The ovaries 
seemingly were not well developed, but I was unable to make out whether 
they were abortive as well. The evidence available seems to show that 
there was a tendency towards the development of gynomonoecism. The 
