Oil Abnormal Flowers of Solarium tuberosum. 79 
was developed in each anther lobe. Of the other four stamens, 
two only produced ovules on the free surface. All four, however, 
shewed cavities radially narrow, but considerably extended 
tangentially. Quite normal ovules were produced on the walls of 
these cavities. In the case of three of these stamens one of the 
anther lobes appeared to be wanting ; the other anther lobe formed 
one or two pollen-sacs. Basal sections (VII) shewed fusion of 
the tissues of the stamens and carpels, save in one instance where 
the right-hand stamen was united to a petal. Ovules were produced 
at this lower level, though the lower limit of the pollen-sac had 
been nearly reached. The small stamen on the left-hand side 
appeared to have become somewhat twisted, ovules coming off 
from its exterior surface. 
In none of the other flowers examined were any further 
variations seen in the method of developing ovules. As no 
endosperm could be distinguished in any embryo-sac, whether on 
stamen or carpel, it could not be determined whether the ovules 
appearing in these strange situations had been pollinated. 
A longitudinal section of another flower is diagramatically 
represented at V. One massive stamen shews two fertile 
pollen-sacs and many ovules springing from the free inner surface 
of the filament from the base upwards, and the bulging mentioned 
above is clearly seen. The ovules arose more or less in clusters, 
often slightly raised on platforms or ridges on the surface of the 
stamen. This flower shewed stamens with cavities running through 
the greater part of their length, on whose walls ovules were 
developed. It also shewed an outgrowth from the inner surface 
of a petal, which on closer inspection proved to be a shortly- 
stalked microsporangium. The pollen-sac contained pollen-mother- 
cells, which, to judge by the broken-down tapetum, were fully 
formed. The sac was surrounded by five or six layers of cells, 
and in this case shewed no signs of rupture. Pollen-sacs were 
found in the petals of another flower of which transverse sections 
were made (III). One end of the petal was normal and 
incurved, composed of fairly regularly-arranged cells ; the other 
end was several times as thick and was composed of cells 
comparable in size with those forming the main mass of a stamen. 
An attempt at an anther lobe was suggested by the curving 
outlines. A large crescentic pollen-sac was also present, whose 
outline followed the curve of the petal. This pollen-sac shewed 
distinct signs of rupture towards the inner side of the petal. Its 
