AKKANGEMENT OF FOSSIL VEGETABLES. 175 



A remarkable fact mentioned by my son throws some light as to the comparatively recent 

 extirpation of the Moa. In one spot the natives pointed out some little mounds covered with 

 herbage, as consisting of heaps of ashes and bones, the refuse of the fires and feasts left by their 

 remote ancestors. Upon digging into them, a quantity of burnt bones was discovered : these 

 belonged to Man, Moa, and Dog, and were promiscuously intermingled. These calcined bones 

 present no traces whatever either of the earthy powder or manaccanite sand which the cells 

 and pores of the fossil bones invariably contain. If, as the natives affirm, these are the 

 rejectamenta of the feasts of the aborigines, the practice of cannibalism by the New Zealanders 

 must have been of very ancient date, and could not have originated, as Professor Owen 

 supposed, from the want of animal food in consequence of the extirpation of the colossal birds. 

 (See ante, p. xi.) 



IV. Botanical arrangement of Fossil Vegetables. — Mr. Artis, in the Introduction of 

 his work, offers some judicious observations as to the proper method in which the study of Fossil 

 Botany should be pursued. He remarks, "that from the imperfect state in which fossil vegetables 

 are generally found, the ordinary characters by which recent plants are referred to their con- 

 geners, can scarcely ever be detected in them. The sexual organs on which the systems of 

 Linnffius and Jussleu are founded, and even the integuments of those organs while in the state 

 of flowering, have uniformly perished. The external parts of the seed or fruit exist, indeed, 

 in a fossil state, but they are almost always insulated from the other organs. If leaves are found, 

 it is almost certain that scarcely any portion of the stem will be attached to them ; if the 

 external parts of a trunk, then very rarely any vestiges of the branches and foliage. And when 

 traces of the internal structure can be discovered, it is seldom that the external character of the 

 stem remains. 



" In consequence of this deficiency of the essential characters on which the determinations 

 of the botanist are founded, there exists a necessity for examining more minutely and accurately 

 than has yet been done, the internal structure of recent plants ; their habits of growth, the 

 cicatrices or scars left on the stem by the leaves that are spontaneously shed, the different 

 appearances which their fruits exhibit in their various stages of development— all these points 

 must be minutely studied before we can obtain any certainty as to tlie identity of ibssil and 

 living species of plants. 



" It is not by publishing detached and unconnected delineations and descriptions of fossil 

 plants, as they occasionally occur, that the knowledge of them can be considerably promoted. 

 A systematic arrangement must be formed; and the first step to this is the accurate deter- 

 mination of the species. Hoc opus, hie labor est." 



" It will be seen," he observes, " in the course of this work, how easy it would be to 

 imagine parts of the same specimen to be different species, when they happen to be broken 

 and dispersed. I can confidently assert, that in at least a thousand different specimens whicli 

 I have had in my possession, not more than a hundred distinct species can be recognised. 

 Furthermore, still fewer indeed can be referred to any living species ; for it is not the fern- 

 like leaf of a plant, the palm-like cicatrix, or tlie cane-like joint of a stem, that will suffice 

 to identify them with those tribes of the vegetable kingdom. The whole anatomy of the 

 plant must be studied. The subject has, indeed, been begun by Professor Martins, in his 

 comparison of certain fossil stems of plants with those of the living plants growing in the Brazils, 

 but the study is as yet too new to afford certain results. Accordingly, several of that professor's 

 opinions are at variance with those of M. Adolphe Brongniart, who has also compared the recent 



