384 MR. С. Е. М. SWYNNERTON ON 
“Оп the outside of some of the flowers thus probed (but not of all) were 
small green aphides, but I had been near enough to the bird to be able to 
judge quite definitely from its actions that it was not these that it was taking. 
The point of the bill had in every case been applied in a decided and un- 
hesitating manner toa single point in the base of the flower and been pressed 
well in, the flower sometimes quivering with the pressure. I examined many 
other Cannas, and found that in not less than one in three the same small 
hole was present. In two cases its place had been badly chosen, immediately 
above the ovaries, and therefore just too low for the bird's purpose ; but in 
neither of these had the bill penetrated deeply. Such flowers as I opened 
contained no insects, and there was nothing to indicate that the holes had 
been made by any agency other than the bird's bill." 
* I have frequently, since making the above note, watched the sunbirds 
utilising or making such holes: on a few occasions when I have noted the 
exact spot at which the bill was inserted and gone to inspect, I have found a 
wet, glistening, obviously freshly-made hole. Tt has usually been near the 
point of the angle formed by two segments of the outer whorl. I have also, 
as in the above observation, broken up many such flowers without finding a 
trace of any insects inside, nor have the holes seemed to me such as might 
have been made by any of our usual honey-eating insects, unless by Xylocopa, 
and I have not seen this at the flowers, There has been throughout a fairly 
marked difference between individual birds in this matter of proneness to 
take short cuts: thus on August 6th (1911) Т noticed that of а male and two 
female Cinnyris chalybous that I watched simultaneously at the Cannas for 
about ten minutes, one female never used the flower’s natural opening, while 
the other seldom failed to, and the male used sometimes the one and some- 
times the other. Тат unable to say at the moment which mode of entering 
the flowers is the most used. I have т a few eases seen indirect evidence of 
а failure to pierce.” 
The above passage was probably written in July or August, 1912, when, 
at Professor Poulton's suggestion, I commenced to shape my notes into a 
paper for publication. 
“Мау 15th, 1913. The Canna bed las been popular of late, and on 
examining to-day the bases of a large number of flowers of five varieties, 
I found that nearly all of them had been pierced in the manner described 
already." 
“May 20th, 1913. Watched a female С. chalybwus enter two, I think 
three, flowers of (A), a Canna sp. with small orange-yellow flowers that had 
grown out far from its original spathe, all Ъу previously made artificial 
openings, then go on to another (В) (Canna indica, Linn., var. orientalis, 
Rosc.), the flowers of which were closely backed by the краће. Here she 
probed three flowers, all by their natural openings, using the spathe as perch. 
I at once drew both plants (Pl. 81. fig. 1), and a reference to the figures 
