MECHANISM OF PLUMAGE IN WATER BIRDS. 205 



those made with other birds. The absolute specific gravity of 

 this specimen was '86. While dry and clean, with plumage 

 unruffled, and the aerial envelope kept as large as possible, 

 it was -66, and with the feathers carefully bound down with 

 fine yarn in as natural a manner as possible it was *84. The 

 actual loss of bulk between the two conditions — with feathers 

 held out and with them depressed — was 2*24 cubic inches. Of 

 course, the whole of this air is not available for lifting the 

 bird in the water, but I see no way of measuring the exact 

 amount of the effective air, and must be allowed to guess 

 that at least one full inch can be used. The increase in buoy- 

 ancy due to the addition of one cubic inch of air to a body 

 six ounces in weight must be very considerable, and quite 

 enough to cause the alterations in draught that we observe in 

 the living bird. 



Some observers have thought that the air-sacs play a part in 

 the submergence of water birds. I cannot think this is so. In 

 a dead bird the amount of air in these rarely examined reservoirs 

 is very small, as may be seen by opening them under water, and 

 the structure of the body prevents the bird increasing this 

 quantity to any useful extent ; and, it will be remembered, the 

 volume of air would bear but a small proportion to that of the 

 aerial envelope outside the epidermis and within the " feather 

 film." Observations seem to show that the syrinx of a Little 

 Grebe is too small to allow the escape of a sufficient quantity 

 of air to cause the sudden difference of draught in the alarmed 

 bird. 



There are many minor points that will be noticed by a 

 thoughtful observer, but I do not think it is necessary to detail 

 them here. An alarmed bird lifts its heavy head and neck right 

 out of the water, thus destroying the large reservoir of air 

 round the crook of the neck, and throwing additional weight 

 on the floating body — just as an uplifted arm will press a 

 swimmer deeper in the water. The actual shape of the bird 

 is also important, but I am sorry to say that I have not been 

 able to make any reliable experiments. When a diving Duck 

 is standing on the land (a Pochard was an excellent example) 

 it exhibits a broad and flat keel, and is obviously farther through 

 from side to side than when it is diving. This is an important 



