GREATKR Pll'KFlSH. 353 



that there had been three separate deposits, so that while 

 some were in one portion of the pouch ahnost fully developed, 

 the latest barely displayed the existence of the eye and the 

 snout. Yet such is not always the case, and perhaps not 

 often; and the following are the notes of my own examination; 

 with the additional remark, that the pregnancy has been found 

 so late as the month of September; although it has not extended 

 into October, When, in April, the pouch was found filled with 

 ova, the edges of the slit or opening were united together by 

 thin fibres, as they were before any had been received. The 

 grains were then all closely fastened together by a covering 

 membrane, and also attached to the walls of the pouch both at 

 the sides and br.ck, but not in front; and each one lay in a 

 cell, the borders of which passed across, with scarcely a mark 

 longitudinally. Each egg was formed of a transparent fluid, 

 and on one side was a red mark in distinct grains of different 

 sizes, in small proportion to the whole bulk. As the grains lie 

 in the pouch, this red spot is in every instance directed towards 

 the opening; and on a short exposiire to the warm sunshine 

 the whole substance became solid without shrinking. When 

 further grown the body of the fish is seen curved into a circle, 

 and the head projecting with a short snout in the middle. At 

 a still further growth the ovum remains attached to the body; 

 but even when fully developed a kind of attachment still 

 continues between the parent and the young, for in case of 

 alarm they fly again to the shelter of the pouch, and are readily 

 received into it. 



While searching for food among the overhanging weeds and 

 crevices of rocks and stones which they frequent, every attitude 

 is adopted, with the head up or down in each kind of the 

 perpendicular, and with much contortion; while the snout is 

 thrust into the chinks where the prey is likely to be met with. 

 The food generally appears to be the smaller kinds of crustacean 

 animals; but not unfrequently shrimps of comparatively no small 

 size are swallowed; and there have been found in the stomach 

 some so large as to raise our wonder how they could have 

 been made to pass between the jaws and through the gullet; 

 and it is only the remarkable structure observed in these parts 

 that will explain the possibility. This structure is complex in 

 a very high degree, by an arrangement of jointed bones, 

 VOL. IV. 2 Z 



