Chap. II. THE TRIONYCHID^E. 331 



till late in lifo. The rourtli ]r.ur oxteiul backward under the pelviw ; their front 

 edges extend pretty directly inwai-d, their hind edges more backward, so that they 

 are broad wdiere ihoy meet under the pelvis. The odd bone is long and slender, 

 and arches forward and overlaps the second paii*. The bones of the fir.st pair are 

 small, and Ijent nesirly to a ri";ht angle; one of their limbs rests against the odd 

 bone, while the other reaches almost directly forward. A thick derm miderlies all 

 this bony framework, and spreads out before it, under the shoulders, as far as the 

 end of the body, and, in Trionyx proper, behind it, under the hind legs. A con- 

 siderable portion of this derm, on and immediately around the l)ony frame, is ossi- 

 fied ; but the larger part, including a space in the middle, is not. There is, on that 

 account, some mobility in the plastron, so that when the animal takes breath it 

 yields and expands. The microscopic structure of the unossified derm has already 

 been illustrated above.^ 



As stated before, the ossification of the shield is very irregular, as it undergoes a 

 great variety of changes during its growth. There is, however, a regular alternation 

 between its growth and that of the true skeleton, with which it is connected, now 

 the one advancing,^ now the other. The ossification is much less fixed and deter- 

 mined, both as to extent and position, than in the other families of the sub-order. 

 These peculiarities, and their relation to the general form, ai'e still subjects of inves- 

 tigation, and consequently their value as family characters is not fully determined. 

 This much however is certain, that the ossification goes on more slowly, is not 

 carried so far, is much less intimately connected with the true skeleton, and is 

 more varj^ing, than in the other famihes of the sub-order. 



As shown above, the vertebral column is nearly at the same level in the sacral 

 region as within the scapular arch. The pelvis and shoulder apparatus have nearly 

 the same height ; they take the proportions of a cross section of the body, that 

 is, they are low and wide spread. The scapula is long, as also are the coracoid 

 and tlie acromion ; but the scapula reaches far outward, and the acromion from 

 thence inward, so that the arch is stretched out, as it were, sidewise, and the 

 shoulder joints are carried close to the edges of the body. The sacrum is broad, 

 the iliac bones are nearly parallel, and the pelvis is as broad across the hip joints 



> See Chap. 1, Sect. 5, p. 2G3. Trionyx, (Transact. Pateont. Society, 1840,) as far 



° The regular allernation wliich is observed in as they relate to the extension of tiio ribs licyond 



the increase and enlargement of the bony derm and the solid carapace and to tiie form of its rim, are 



of the true skeleton, especially at the ends of the not specific, but may be observed in a series of speci- 



ribs, is an additional proof lliMt tlic sliirld is not to m(>ns of the same species, in different stages of ossi- 



be considered as formed by a dilalation of the ribs fication. I have satisfied myself of this by a careful 



only, but chiefly by the ossificiition of the derm. Tlic comparison of fourlcen skeletons of Aspidonectes 



differences noticed by Owen, in his paper on the fossil spinifer, and mulicus, of all ages. 



