122 STUDIES IN THE EVOLUTION OF ANIMALS 



compose the groups on its shoulders (Fig. 4) ; and not improbably 

 the stretching of the skin might partially account for the charac- 

 teristic spotting on the haunch of the Cheetah shown in Fig. 59, 

 Nos. 33-35. 



In Fig. 70 (a) I have endeavoured to restore some of the plates of 

 the carapace of the Jaguar's immediate ancestor. Of course this 

 restoration is imaginary, but it is suggested by five contiguous 

 rosettes on the Jaguar's left flank, behind the shoulder (Fig. 4). I 



a. 



S 



yCKj 



Fig. 70. — {a) Restoration of the rosettes of an imaginary Glyptodontoid ancestor of the 

 Jaguar, with closely interlocking plates ; {d) The same in subsequent forms of mammals, 

 when the plates may have become dissociated, with intervening flexible skin. 



have shown the plate-rosettes as closely fitting in (a) ; but imagine 

 them to have become somehow dissociated as in (d), from causes 

 which I have already discussed, and you have intervening channels 

 or commissures of flexible skin, such as we see in the Crocodiles, 

 and also between the bands of the Armadillos. These skin com- 

 missures are now represented in the Jaguar by the paler ground 

 colour reticulated between the rosettes. I have given the imagined 

 plates a trapezoid shape, to make them similar to the Jaguar rosettes, 

 but the shape of the latter may have become much distorted, as 

 indeed we see it in various parts of the Jaguar skin itself I have 



