402 МВ. С. Е. М. SWYNNERTON ON 
examined their neighbours too, with the same result as in the other case. 
These three racemes, I ought to add, showed no disturbance of the bracts. 
“In the first of the two second-clump racemes (examined row by row, 
each individual flower being removed for separate examination) the first row 
consisted of 16 just open flowers and 4 buds. There were 23 flowers in the 
second row. In the yet lower rows a large proportion of the flowers were 
already becoming bruised and flattened down, and some оЁ the slits may 
possibly, therefore, have been the work of claws. 
Slit Slit Basal 
Intact. 
above. below. punctures. 
Ist row ............ 10 3 9 5 
2nd row ,.,...,...... 14 6 3 — 
Lower rows ,,...... 9 23 95 — 
The basal punctures were confined to the exposed surface of the exposed row 
and honey was exuding from them. 
“ The second raceme had wider-mouthed flowers than the first. Its injuries 
were as follows :— 
Slit Slit Basal 
Intact. . 
above. below. punctures. 
Ist row ............ 10 6 3 2 
Lower rows ........ 73 7 36 — 
* Of the three third-clump racemes, two, of much the same type as the 
second raceme above, showed slits (3 and 1 respectively) in. the upper sur- 
face of flowers of the top row, and their injuries generally were about on 
а par with those of that raceme. I did not make an actual count here, but 
ran carefully through all the flowers to gain a correct general impression. 
In addition, one raceme had four non-continuous injuries in its top row, 
three of them at the bases of flowers. They were capable of inflietion by a 
bird perehed below, owing to the scantiness of the intervening flowers. None 
of the lower ones showed any such injuries. 
“The third raceme’s flowers possessed the same constricted neck as those 
of the first raceme deseribed. In this case there were no basal injuries, but 
a number of. upper surface slits not penetrating far, both in the topmost row 
and those below it, as well as the usual slits elsewhere. So far as ] could 
tell without actually counting, its injuries of all rows were very much the 
same in quantity and character as the lower-row injuries of the other con- 
stricted raceme, and I now see the probable reason. The neck is in a large 
proportion of cases so narrow that a bulbul’s bill could not be inserted 
without splitting the flower. Damage of this kind would probably not 
prejudice the flower's chances of pollination, but the basal slits and punctures 
distinctly would.” 
“ Later: evening.—l ran over to Chipete at sunset and partly stripped 
al more flowers—10 in one place and 21 in another—as well as 9, of which 
