Zoology. 665 



disposition. We therefore at last adopted the plan of keeping 

 a large patch of frogbit (Hydrocharis morsus ranas L.) con- 

 stantly vegetating on the surface. This tended also to keep 

 the water from corrupting. When this floating patch was 

 slowly moved from side to side of the reservoir, the siren 

 kept most accurately beneath it, endeavouring to avoid ob- 

 servation ; and in this exercise it showed great alertness and 

 sagacity. 



" On the morning of the 22d October, 1831, the siren was 

 found by the gardener lying dead on the paved footpath of 

 the hot-house, not far from the reservoir. It did not appear 

 to have met with any injury in its fall from the reservoir, 

 which was placed on a trellis about 3 ft. above the pathway. 

 The fall, indeed, must have been broken by intervening flower- 

 pots, and no external marks of lesion [hurt by bruising] 

 appeared on the body. The fine fimbriae of the branchial 

 apparatus, however, were completely dried and shrivelled 

 up ; and 1 have no doubt in my own mind that the death of 

 the animal arose from this cause. The older naturalists 

 supposed these fimbriae or fringes to be opercula or gill- 

 covers, and regarded the vertical clefts or perforations in 

 front of the fringes as the true gills; but Mr. Wilson is 

 certainly right in considering the fringes as the true gills. 



" In the New York Medical and Physical Journal for June, 

 1824, Dr. Samuel L. Mitchell gives an account of the 

 examination of several specimens, dead and alive, which had 

 been transmitted to New York by the same Dr. Farmer of 

 Charleston who sent the living specimen to Edinburgh. Dr. 

 Mitchell seems to think it most probable that the air-sacs, 

 called lungs, do not perform any direct respiratory function, 

 but are mere receptacles of air, performing only an auxiliary 

 service, in occasionally furnishing the branchiae or gills with 

 the atmospheric air which the animal from time to time in- 

 hales, and which is detained in these receptacles till wanted. 

 I cannot help thinking that this view acquires additional 

 probability from the circumstance of the Canonmills siren 

 having died without any other apparent cause than the exsic- 

 cation of the extreme fimbriae or fringes, that is, the true gills." 



During the six years and four months through which Mr. 

 Neill kept the *Siren Zacertina, he states that " no structural 

 change took place, nor was the slightest tendency to any such 

 change discernible. The animal had evidently increased in 

 size. When it arrived, it was nearly a foot and a half long. 

 When it died, it was fully twenty inches in length ; and it had 

 also perceptibly increased in grossness. It may, therefore, I 

 think, be pretty confidently concluded, that it is no larva, 



