490 Dr. Falconer on Cuvier's Laws of Correlation, 



8. The number of the mcisors excluded Dasyurus, leaving 

 only Didelphijs. 



9. The sum of all the characters throughout the skeleton, 

 and each of them taken separately, indicated Didelphys. 



10. Therefore the fossil animal was a Didelphys, like the non- 

 prehensile tailed Opossums, which are now restricted to the 

 American continent. 



If, in turn, we analyse the process, it is obvious that the 

 result was obtained, first by determining the class, and then 

 eliminating, by a series of successive steps, every differential 

 condition, down to a single residual form ; and if we examine 

 the nature of the correlations upon which the successive steps 

 were founded, it will be seen that most of them were of the 

 necessary order, and but few of the empirical. Cuvier was con- 

 fident, upon the evidence, that the conclusion was sound : but 

 a crucial instance remained, by which to verify it. If the 

 extinct form was an Opossum, it must have had a marsupial 

 pouch, and to sustain the pouch, marsupial bones were neces- 

 sary. He summoned some competent friends to witness the 

 expected verification. A portion of gypsum was cleared away 

 from the slab by the graver, at a sacrifice of some of the ver- 

 tebrse, and a pair of marsupial bones, concealed in the matrix, 

 were brought to light, resting in their natural position above 

 the edge of the pubis. Thus, after determining the class, the 

 first step in the further analytic deduction rested upon a rational 

 or necessary correlation, and so also did the last, crowning the 

 identification. AYhen referring, afterwards, to this signal 

 triumph, the great anatomist quietly remarked : " Je laisse cet 

 article tel qu^il a paru d'abord, dans les annales du Museum, 

 comme un monument, selon moi assez curieux, de la force des 

 lois zoologiques, et du parti que Ton pent en tirer." 



Let us next examine what the true principle is, according to 

 Mr. Huxley. It is not denied, that in palaeontology, legitimate 

 consequences may be deduced from the laws of living form : on 

 the contrary, the whole science is admitted to be built on them. 

 But the process of restoration depends, "not on the physiological 

 correlation or coadaptation of organs -j" but, " first, on the validity 

 of a law of the invariable coincidence of certain organic pecu- 

 liarities established by induction ; secondly, on the accuracy of 

 the logical process of deduction from this law.^' Now, the ability 

 to demonstrate a proposition, or to infer a legitimate deduction, 

 may be a measure of the capacity of the individual, but it is no 

 criterion of the abstract truth of either. The second clause 

 may therefore be struck out, as self-evident and superfluous. 

 The principle is thus limited to " the invariable coincidence of 

 certain organic peculiarities.'^ This invariable coincidence may 



