Birds, 6839 



is performed by Acanthylis without any apparent motion of the wings 

 — a simple downward plunge owing its rapidity to impetus and gravi- 

 tation alone. Whilst the first piramidig is pursuing this erratic course 

 its cries will be answered by a second and perhaps a third, which will 

 gradually approach with similar movements. If they do this they will 

 generally take a wide sweep and disappear, to be succeeded by others 

 or return after a time. But very often a much larger number of birds 

 may be seen steadily moving over a much more circumscribed space. 

 I have seen them amount to as many as about thirty, but generally 

 less. Not a sound is uttered ; the birds beat over their chosen ground 

 unceasingly ; they are evidently busy feeding, and it is then some of 

 the most remarkable points of their flight may be best observed. It 

 was out of a flock of this sort I shot one some time after sunset, whilst 

 they were beating over a steep narrow valley peculiar to the porphy- 

 ritic conglomerate, close to the ' Bull's Head,' in Clarendon. On 

 dissection I found the capacious stomach stuffed to protuberance with 

 the winged portion of a community of small red ants. Though there 

 were a very great number of insects, and all in a perfectly uninjured 

 state, I could not detect a single individual of another species among 

 them. It would of course not be safe to lay too much stress on two 

 solitary examples, but it is very remarkable that in the one dissected 

 by yourself (' Birds of Jamaica,' p. 40) the contents of the stomach 

 should prove 'almost (if not quite)' composed of a single species of 

 beetle, whereas in mine they were a single species of ant. The only 

 inference which it seems possible to draw from this, as we know many 

 species of insects are appropriate food to these birds, is, that (like 

 Acanthylis) they are swarm-destroyers : hence their beating over cir- 

 cumscribed spaces, as over the steep little valley I observed, or the 

 clump of flowering trees noted by Mr. Hill (p. 39), or, as it may be 

 often also remarked, over a pond or water in some shape. Now, as 

 far as I am able to observe, a swarm of insects, though often quite 

 irregular in shape, frequently assume a spherical or ellipsoid form, the 

 major axis vertical ; and this, it is obvious, will frequently be the case 

 where a common object of attraction, as a pond or particular tree, 

 keeps the insects together, and still more where, as in the swarming of 

 an immense formicary, the attraction of the insects is for each other. 

 A section of such a swarm may then be represented by a circular line. 

 In the instance I have above alluded to there were with Chordeiles, 

 even at that late hour, two or three Acanthylis feeding on the same 

 prey, and it becomes very interesting thus to compare, on the spot, the 

 motions of the two species. On arriving at the edge of the swarm, as 



