REVIEWS. 237 



it may attract some minds, it is not too much to say that it is not yet 

 satisfactorily proved, and that it labours under some great difficulties 

 and serious objections. A really good classification, independently 

 formed, might greatly assist our judgment on the question, but the 

 assumption of the hypothesis, and the attempt to trace the order of 

 descent of the multiplied forms occurring in nature, can only encour- 

 age fanciful analogies, and interfere with our application to the great 

 questions. What are the parts ? and. What are the circumstances re- 

 specting them ? which, in themselves, and from their connection with 

 general structure, have most real importance in determining affinities 

 and leading us to a natural grouping of objects. Mr. Clarke's ar- 

 rangement is certainly not rendered more plausible by the fancy that 

 Endogens (Monocotyledonese) are derived through Riccia and Lemna 

 from Marchantiacese ; Epigynose Dicotyledonege through Balanophor- 

 aceee from Bryacese ; the Chloranthal division through Gnetacese and 

 Lycopodiacese from the same source, and so in other instances. We 

 cannot but be struck with the insufficiency of the analogies by which, in 

 one or two cases, it is af tempted to justify these speculations. Take for 

 example the relation of Endogens to Marchantiacege. The author 

 expresses himself thus : — " I am aware that there is no evidence to 

 show that any near affinity exists between Lemnacece and Ricciacece, 

 further than that the habit of some species of Riccia is so much like 

 that of Lemna, that they are stated to have been mistaken for species 

 of that genus by some authors, while species of the latter genus have 

 in fact been described as belonging to the former ; and although habit 

 is often of little value as a character among phanerogamous plants, 

 it is acknowledged to be of more importance among the cryptogamous. 

 But as the present arrangement includes the placing of all the crypto- 

 gamous families in relation with the phanerogamous, the Endogens 

 should in all probability be compared to one, at least, of their lowest 

 forms ; and as the Ricciacese have no affinity with the apetalous forms 

 of Exogens, a negative evidence is afforded of their being the crypto- 

 gamous form of the Lemnacece.'" What this amounts to is, that grant- 

 ing the truth of the system, and that each race of higher plants is 

 derived from, or immediately related to, some form of the lower, there 

 seems some probability that Ricciacese afford the form nearest to Lem- 

 nacese, there being cases in which the foliage, when the flowers or 

 fruit cannot be found, may be mistaken one for the other. The sys- 

 tem being by no means proved, we might leave any reader to judge 



