136 Dr. Hancock on the Mullets of Guiana^ 



of the lung, which has capacious air-cells, lined with a strong 

 capsular membrane, are of firm texture, unlike those of qua- 

 drupeds, and invested by a strong tunic which enables this 

 viscus to support the requisite hydrostatic pressure. — In the 

 manati of the interior of Guiana, the lung is one-third the 

 length of the whole body — in its texture similar to the swim- 

 bladder in the gilbagre. M. Humboldt, reflecting on the size 

 of the lung in the manati, thought it strange the animal should 

 be compelled to rise so often to blow or spout water ; vide 

 Pers. Nar., v. iv., p. 448. But it seems to me a fact sufficiently 

 evident, that the lungs in those animals answer the double 

 or rather triple purpose of respiration and oxygenising the 

 blood, and as that peculiar kind of gasometer or buoyant blad- 

 der found in the true fishes. This will, I presume, be ad- 

 mitted by competent judges as the most rational view of the 

 subject ; and in which there is nothing strange, unless it be, 

 that so many sagacious physiologists should have passed the 

 affair quite unheeded. 



This fish (the grey mullet) appears to be much more nearly 

 allied to the mugil albula, as given by Linnaeus, than to the 

 cephalus ; and it may be observed, that the different species of 

 this genus resemble one another so closely, that they are with 

 difficulty distinguished by the external characters, unless we 

 except the number of rays in the fins. ^ Inattention to such 

 circumstances has led naturalists into frequent errors. 



Another character which has been recommended as afford- 

 ing the most certain distinction, I find to fail here, that is, the 

 number of bones in the vertebral column; in several cases, 

 distinct species are found to have the same number. However 

 valuable this may be to the anatomist, as a specific character, 

 it is not applicable to general use, as, for the progress of natural 

 history, more ready and obvious external characters are re- 

 quisite. *' i^>.'- -'^'^ --^- ■- ■■''-''■■''■■ ■'' ' 



Pendfcllt'S'figui^'(;BW Zop^^^'. iii., p. 436) appears to be 



♦ With respect to the numerous ailuri of the tropics sdsO, it is quite impossible 

 to distinguish them by other means. I have found the number of rays in the 

 fins to be remarkably constant, both in these two and in other genera, although 

 some assert the contrary. — The method of those, therefore, Avho neglect such 

 specific characters must be very defective, especially for genera consisting of 

 numerous approximating species. 



