142 THE JOURNAL OF BOTANY 
r more spurs, these were often similar in size; when more than 
dene, generally, the lower spurs exceeded the others in size, The 
oO upper spurs were always smaller and sometimes were mere 
h 
of flowers with five spurs. Towards the end of April, 1900, I 
gathered indiscriminately 1000 flowers by the sides of, and adjacent 
to, the ancient highway known as the Ermyn Street, between 
Epsom Downs and Leatherhead Downs, of which 805, or more 
than 80 per cent., were nee The variations ranged from 
slight divergences in appearance’ to important m odifications in 
structure. Taking these latter ‘frst, 222 flowers showed, in addi- 
tion to the ordinary spur, saccate protuberances at the bases of the 
petals suggesting incipient spurs, and 65 flowers two, three, or four 
spurs each. Of these 65, three possessed four spurs, 18 three 
spurs, and 44 two spurs eac wo of the flowers with four 
petalline spurs showed traces of a fifth r rudimentary spur. With 
regard to the —o development of these auxiliary spurs, it may 
be noted that stam inal-processesy may or may not be developed 
inside. Thus, of the ee flowers in question, one showed two 
staminal- “processes in one spur Bet two of the supernumerary 
spurs bore each one stare: -proces In a second flower only 
i h 
foregoing instances it may be assumed that the spurs with two 
pss? -processes each were the normal spurs. The third flower 
characterized a the cea of a shed of staminal-processes 
spurs. Of the 18 pes with three ss each, eight possessed 
the normal pair of processes in one spur only, three were without 
the pair of celta in one spur, but had one process in each 
of two §] one flower two spurs Lt each a pair of 
staminal- process, and a: on one side, the entrance to the 
spur. ifferentiation of tissue, this curious organ 
resembied he pete Aiea with which only could it be 
compar 
The “chief variations itn among the 44 flowers with two spurs 
Am 
peseet within it two pri ae -processes. Hight others were fur- 
nished, in addition, with a third staminal-process which entered the 
second spur ;. two flowers had one process in each spur, and one 
showed two processes in each ae Five cna ‘processes were 
y 
tinct spur. A cu ae retanoxphosis was shown by one flower, in 
which one of the staminal-processes had become petaloid. One 
penetrated a spur, and in the sere of the flower was a curled 
petaloid organ showing on one side anther-cell and orange-coloured 
connective, the whole structure evidently representing a stamen 
