the Dillenian Herbarium. 115 



idea which most forcibly occurred to my mind was that of their 

 necessarily imperfect nature ; but, on reading over what I have 

 written, I am far more struck with the apparent presumption of 

 undertaking thus freely to comment upon the works of the most 

 celebrated naturalists that ever lived. I trust, however, the So- 

 ciety will agree with me, that this presumption exists only in ap- 

 pearance ; for I have done nothing more than observed what, had 

 they enjoyed a similar advantage, they would themselves have 

 observed long since, and the difficulty* of conveying an adequate 

 idea of a plant by descriptions, or sometimes even by figures, is too 

 great for it to be possible for the most accurate observer always 

 to avoid mistakes in references. I have indeed one satisfaction, 

 which is the consciousness that I incur no risk of a charge of de- 

 traction ; for Dr. Roth, the author whose works have most fre- 

 quently fallen under my notice, ranks too highly as a botanist, and 

 has too many claims upon my esteem as an individual, to leave 

 any apprehension of such a suspicion. I will only add that, spite 

 of the progress made in our knowledge of the class Cryptogamia 

 since the days of Dillenius, the excellence of his Historia Musco- 

 rum is so great, and its utility to every student so indispensable, 

 that, as the work itself is become extremely scarce, although im- 

 pressions of the plates may far more easily be procured, 1 do not 

 think any botanist, who has sufficient leisure for the undertaking, 

 could engage in a more necessary task than the re-publication of 

 the letter-press, adding specific names and descriptions from the 

 writings of Linnzeus, Roth, Hedwig, and Acharius. 



Yarmouth, 

 6th April, 1803. 



* This difficulty is far greater among the Conferva than among the Muscl, or any other 

 part of the Algce, unless high magnifying powers are employed: for want of these, and 

 of colouring to his plates, they are in many instances absolutely useless. 



q 2 XI. Description 



