14 BAHI A— BRAZIL. [chap. i. 



time, generally expelled the air and water with considerable 

 force from the branchial apertures and mouth.- It could emit, 

 at will, a certain portion of the water ; and it appears, therefore, 

 probable that this fluid is taken in partly for the sake of regu- 

 lating its specific gravity. This Diodon possessed several means 

 of defence. It could give a severe bite, and could eject water 

 from its mouth to some distance, at the same time making a curious 

 noise by the movement of its jaws. By the inflation of its body, 

 the papillae, with which the skin is covered, become erect and 

 pointed. But the most curious circumstance is, that it secretes 

 from the skin of its belly, when handled, a most beautiful car- 

 mine-red fibrous matter, which stains ivory and paper in so 

 permanent a manner, that the tint is retained with all its bright- 

 ness to the present day : I am quite ignorant of the nature and 

 use of this secretion. I have heard from Dr. Allan of Forres, 

 that he has frequently found a Diodon, floating alive and dis- 

 tended, in the stomach of the shark ; and that on several occa- 

 sions he has known it eat its way, not only through the coats of 

 the stomach, but through the sides of the monster, which has thus 

 been killed. "Who would ever have imagined that a little soft 

 fish could have destroyed the great and savage shark ? 



March 18/A. — We sailed from Bahia. A few days afterwards, 

 when not far distant from the Abrolhos Islets, my attention was 

 called to a reddish-brown appearance in the sea. The whole 

 surface of the water, as it appeared under a weak lens, seemed as 

 if covered by chopped bits of hay, with their ends jagged. These 

 are minute cylindrical conferv£e, in bundles or rafts of from 

 twenty to sixty in each. Mr. Berkeley informs me that they are 

 the same species (Trichodesmium ery thraeum) with that found over 

 large spaces in the Red Sea, and whence its name of Red Sea is de- 

 rived.* Their numbers must be infinite : the ship passed through 

 several bands of them, one of which was about ten yards wide, 

 and, judging from the mud-like colour of the water, at least two 

 and a half miles long. In almost every long voyage some 

 account is given of these confervas. They appear especially com- 

 mon in the sea near Australia ; and off Cape Leeuwin I found an 



* M. Montagne, in Comptes Eendus, &c., Juillet, 1844; and Annal. des 

 Scienc. Nat., Dec. 1844. 



