PRINCIPLES AND METHODS OF PALAEONTOLOGY. 385 



iiidieate tbe g-eneral affinities of tlie animal to wiiieli it belonged; and to 

 justify luui in making those wonderful predictions of what woukl bo the 

 nature of its other parts, which were so often to bo verified in the course 

 of future investigations. 



One of the most remarkable examples of such successful prediction is 

 that which Cuvier himself mentions as "a very singular monument of 

 the force of zoological laws, and of the use which maybe made of them." 

 From the famous gypsum quarries which furnished so many occasions 

 for the display of his genius and knowledge a block was brought con- 

 taining the imperfect remains of the skeleton of a small animal; the 

 shape of the lower jaw and the characters of the teeth were such as are 

 alone known to exist in the order of marsupial animals, of which the 

 opossum and the kangaroo are the most familiar examples. But all 

 known MarsKpialia possess two remarkable appendages to tlie "pelvis" 

 or bony girdle of the hips, which are termed the "marsui)ial bones," be- 

 cause they are connected with the pouch in the female. Here was a law 

 of invariable correlation of anatomical peculiarities (certain teeth and 

 certain forms of jaw being always associated with the presence of these 

 bones) of universal application to living animals; would the law hold i 

 good for the fossil ? Cuvier was so confident that it would, that he in- 

 vited some friends to witness the picking- away of the stone from the 

 region where he believed the marsupial bones woukl be found, and the 

 result verified his expectation, for the bones were discovered just in that, 

 very situation. 



3. It will be easily understood, however, that the whole of this train, 

 of reasoning is only valid on the assumption that a certain uniformity 

 ■Las prevailed in organic nature; that the structures which vre find in- 

 variably' associated now were invariably associated in earlier times; that, 

 in short, the great laws Avhich are expressed by our conceptions of com- 

 mon plans have always remained the same. We know of no reason, save 

 the invariable occurrence of the co-existence, v,'hy a jieculiar form of jaw 

 shoidd always be accompanied by the existence of marsupial bones; and 

 just as certain animals now exist in which the marsupial bones are 

 IH'esent, while the peculiar structure of jaw is absent, so it is quite within 

 the limits of possibility- that, at an earlier period of the earth's history, 

 animals might have existed possessing the peculiar jaw, but deprived of 

 the marsupial bones. Of course, in this case Cuvier's reasoning would 

 not have been conclusive, and his prophecy might not have bcien verified. 

 In point of fact it would not be safe in all cases to regard the laws of 

 invariable anatomical correlation, deduced from the observation of the 

 existing animal world, as applicable, without reservation, to the members, 

 of extinct faunas. Xo generalization from the structure of existing ani- 

 mals could be better established than that biconcave vertebra? are found, , 

 throughout the spinal column, only in fishes and perennibranchiate 

 amphibia, or hollow bones of a certain form are characteristic of birds,. 

 and yet Ave should be led into most erroneous conclusions by reasoning 

 without hesitation from these data, to the structure and aflinities of the- 

 animals to which certain vertebrae and certain bird-like bones found in; 

 the mesozoic strata belong. In fact, while experience shows, with a con- 

 stantly increasing weight of proof, that the great laws of the construction' 

 of animals have been identical throughout all recorded time, and while, 

 therefore, when we possess any clear indication that a fossil animal 

 belongs to any one of the great groups, we may safely predict that it 

 will exhibit all the othar characteristic peculiarities of that group; we 

 must be careful to remember that in many of the smaller groups combi- 

 nations of organic peculiarities have existed of a very ditferent nature' 

 2o S 



