Bibliographical Notices. 353 



On the subject of the formation of wood, the author adopts the 

 theory of Petit Thouars. The different kinds of inflorescence, and 

 the centripetal and centrifugal expansion of flowers, are explained in 

 such a manner as to render the subject easily intelligible. The same 

 may be said of the explanation given of the carpellary theory. 



In alluding to the formation of the placenta in different cases, Dr. 

 Gray notices the theory which has been recently advanced in Ger- 

 many, in which the placenta is considered an axile formation, or one 

 belonging to the axis and not to the carpellary leaves. "This theory," 

 he says, " offers the readiest explanation of free central placentation, 

 especially in such cases as Primula. It is also perfectly applicable 

 to ordinary central placentation, where we have only to suppose the 

 cohesion of the inflexed margins of the carpellary leaves with a cen- 

 tral prolongation of the axis or receptacle which bears the placentae. 

 But in the case of parietal placentation, the advocates of this theory 

 are obliged to suppose that the axis divides within the compound 

 ovary into twice as many branches as there are carpels in its compo- 

 sition, and that these branches regularly adhere in pairs, one to each 

 margin of all the carpellary leaves. Its application is attended with 

 still greater difficulties in the case of simple or uncombined pistils, 

 where the ovules occupy the whole inner suture, as in Sedum and 

 the Columbine, which are doubtless justly assumed as the regular and 

 typical state of the gyncecium ; but to which the new hypothesis can 

 be adapted only by supposing that an ovuliferous branch of the axis 

 enters each carpel, and separates into two parts, one cohering with 

 each margin of the metamorphosed leaf. This view, however, not 

 only appears very improbable, but may perhaps be disproved by di- 

 rect observation, as it has been most completely by those monstro- 

 sities in which an anther is changed into a pistil, or even one part of 

 the anther is thus transformed and bears ovules, while the other, as 

 well as the filament, remains unchanged ; a case where the forma- 

 tion of the placenta from a process of the axis is out of the question. 

 The hypothesis, therefore, is entirely untenable as a general theory ; 

 and whether it affords a correct explanation of any form of central 

 or basilar placentation, must be left for further observation to deter- 

 mine." 



In the systematic part of the ' Text-Book' a good view is given of 

 the principle of classification, and the Artificial and Natural methods 

 are well explained. In speaking of a natural method it is remarked, 

 that this term is applied because the method " expresses the natural 

 relations of plants as far as practicable ; for every form yet contrived, 

 or likely to be devised, is to a considerable extent artificial: 1. Be- 

 cause the affinities of a particular group cannot be fully estimated 

 until all its members are known ; and thus the progress of discovery 

 leads to changes, or modifies our views, as in every other department 

 of knowledge. 2. Because the boundaries of groups are not so ar- 

 bitrarily circumscribed in nature as they necessarily are in our clas- 

 sifications, but individuals depart from the assigned limits in various 

 directions (like rays from a centre), the edge of difference being as 

 it were softened down by an easy transition. 3. Because, even sup- 



Ann. # Mag. N. Hist. Vol. x. 2 A 



