524 THE GANNET 



minor is small (as also in Cormorants and Shags) for the 

 size of the bird and occupies only one-quarter of the 

 sternum." 



It is plain that the cells are not small, and that most 

 of them communicate with one another by a passage of 

 more or less size. It can be said in general terms that the 

 cells almost envelop the breast, belly, and sides, but there 

 are none on a Gannet's back, and this is another reason for 

 thinking that their function is to break the force of the 

 plunge. 



Some of the systa are nearly two inches in length, without 

 undue stretching, and are so firmly united to the pectoral 

 muscles that a knife is needed to part them. 



Down the centre of a Gannet's breast, and attached with 

 considerable tenacity to the keel of the sternum, there lies 

 a stout, tightly-drawn membrane, which forms a firm wall 

 of division between the cells on either side of the under-part 

 of the body. Yet this membrane may be perforated some- 

 where, for Professor Owen found on using a pipe, that air 

 could pass from one side of a Gannet's body to the other 

 {see " Proc. Zoological Soc," 1830-1, p. 91), but the passage 

 was very possibly internal. 



Mr. Wilson's drawings, excellent as they are, cover too 

 small a space to give an adequate idea of what the cells 

 are like. I regret that it is not possible to show 



