1 66 Analysis of Scientific Books and Memoirs. 



them, if they were not too hastily and indiscriminately appropriated. It 

 is true, they are generally given from a presumption that the plant so de- 

 nominated belongs exclusively to that country ; to which the objection is, 

 that it may afterwards prove to be a native of others. Even when this is 

 the case, there is still an important fact recorded, namely, that it was first 

 discovered in the district from which it derives its name, or that there its 

 characteristic marks were originally detected. 



A new Viola is given at p. 304, v. 1., V. Jlavicornis, found in Surry,, 

 and about Norwich. It is said to be allied to V. caniiia, and to have been 

 neglected for it. 



We have, under the genus Erythraa, a new species, E. latifolia ; dis- 

 covered by Mr Shepherd and Dr Bostock in sandy ground near the sea at 

 Liverpool. 



Gentiana acaulis should have been omitted altogether. It has assured- 

 ly never been found wild in Britain. 



If we were to point out one part in the present volumes, in which 

 the author has more successfully amended what has been done by former 

 writers than in another, we should fix upon what concerns the Umbelliferce, 

 a family of plants the most natural, and, at the same time, the most diffi- 

 cult of investigation of any in nature. Much, indeed, had previously been 

 effected by Hoffman and Sprengel, particularly the latter, in laying the 

 foundation of a new arrangement. Our able friend has adopted all that 

 is worthy of being adopted, from the latter author especially, and he has, 

 in many instances, corrected and altered him for the better. We have al- 

 ready had occasion, with specimens in our hands, to make the comparison, 

 and we have no hesitation in giving the most decided preference to Smith's 

 arrangement ; and in pronouncing it the most clear and simple that has 

 yet appeared. The bracteas are by him only considered as of secondary 

 importance, and he discards the term involucre, which, according to 

 Linnaeus, must properly be a part of the flower. The fruit, of course, af- 

 fords the most material points of distinction, but all the hard names which 

 Sprengel has applied to the ribs, furrows, and point of union of the double 

 fruit, (or naked seeds,) are rejected as worse than useless. The base of the 

 style and the floral receptacle are taken into account, and thus the 

 genera of this great natural family are characterized, like other genera, 

 solely by their flowers and fruit. Nor are these changes made to rest only 

 upon examination of the British species, the author has held all the exotic 

 kinds }n view, and carefully studied almost all that are known in other 

 countries. 



We could have wished to have seen Ligusticum cornubiense kept distinct 

 from that genus. We have stated in the Flora Scotica, under the head of 

 L. scoticum, that the L. cornubiense of English Authors is the Danae aqui- 

 legifolia of De Candolle. We did so from an examination of a Piedmon- 

 tese specimen, given us by Professor Balbis of Turin. This individual 

 now lies before us, by the side of our Cornish plant ; and again we must 

 observe that we cannot detect the slightest mark of specific distinction be- 

 tween them. In both there are many barren flowers, sometimes entire 



