DE. E. A. GOELDI ON THE LEPIDOSLEEN OE THE AMAZONS. 419 



other vegetable matter. A vigorous growth of algae helps to increase this stratum. 

 The remainder is filled with fresh water, which is renewed every fourth day. As food, 

 small pieces of meat, fish, ampullarias, crustaceans, as well as mandioca-roots, are 

 constantly offered to it, but up to this time no one has seen our Lepidosiren actually 

 take any of these things. It is, however, in a better state of nutrition than when it 

 it arrived ; it is decidedly fat and round. Therefore it eats something, and it probably 

 does so when it burrows half the length of its body in the mud, as fi-equently seen. 



For respiration it comes from time to time to the surface of the water, and puts out 

 a portion of the head. The operation lasts several seconds. In the large aquarium 

 this act is only repeated at intervals of several hours, but when placed in a smaller 

 one — such as I used for my photographs of the living animal — the respirations at the 

 surface are much more frequent, the intervals being perhaps of a quarter of an hour's 

 duration. The respiration is sometimes singularly prolonged. When descending, a 

 series of air-bubbles is generally expelled from the branchial aperture on each side of 

 the anterior part of the body. The whole body is covered with a viscous or gluey 

 substance, which fills the water with whitish flakes, when the Dipnoan executes more 

 rapid and violent evolutions (see my photographs, PI. XXXVIIl.). 



During the day and when undisturbed our Lepidosiren is a quiet and passive 

 creature, not changing its curled position for hours. Only once we have noticed it 

 attempt to bite the finger of the keeper. It remains generally indifferent even when 

 small living animals are offered it with the pincers. When disturbed it executes most 

 elegant and varied evolutions, and shows itself to be a first-class swimmer, at least as 

 good as Anguilla and Murmna. 



I had the rare pleasure of exhibiting the living Lepidosiren paradoxa at a public 

 meeting of our Museum Society in Para on the night of June 3rd, 1897, which was 

 honoured with the presence of the State's Governor, Dr. Paes de Carvalho, a distin- 

 guished physician, and of a large number of persons interested in our licientific work. 

 Some days afterwards I received from Dr. O'Connor, of Oxley, Queensland, the news 

 that living specimens of Ceratodus (the Dipnoan of the opposite side of the southern 

 hemisphere) had been caught for the first time in Australia. 

 Pari, August 26th, 1897. 



EXPLANATION OF THE PLATES. 



PLATE XXX VII. 



Fig. 1. Lepidosiren paradoTa, half natural size. 



Fig. 2. Outline of the body, showing the lateral lines. 



Fig. 3. Head seen from below. 



Fig. 4. A portion of the body, side view, in order to show the exact aspect of the 



