CENSUS OF SPECIES. 275 



for any of tbe species), and then arranging the names under each group 

 in accordance with the number of subprovinces. Still a different set of 

 groups or series of names would have resulted, if the 18 primary pro- 

 vinces had beeu resorted to as the leading test, followed in succession 

 by the secondary or subprovinces, and then by the counties and vice- 

 counties. After actually drawing out lists in the three different methods, 

 that in which the subprovinces was made the leading test was adopted, 

 because appareutly giving the most correct sequence of the names. The 

 comital and vicecomital floras are yet incompletely ascertained ; and it 

 was found that the use of these more numerous divisions tended too 

 much to place plants high in the census on account of partial frequency. 

 On the contrary, the provincial floras are most of ihem nearly complete ; 

 and so far they are best adapted to give a conrect census. But 18 such 

 lists are not sufficiently numerous; plants very thinly scattered over 

 Britain, if they chance to be found in all the 18 provinces, being placed 

 above other more frequent plants which are abseut or simply unascer- 

 tained in one or two of the provinces. 



Whatever may be the particular mode of applying it, the adopted test 

 involves a combination of the extent of area along with frequency of 

 repetition. By adopting the more numerous comital divisions, a greater 

 importance is given to the latter condition, that of frequent repetition. 

 The larger, and therefore fewer, divisions assign a predominant value to 

 extent of area. On the whole, the 38 subprovinces afford a better test 

 than the 18 jjrovinces or the 112 counties and vicecounties ; a copious 

 list of species having been made out for each of them, and their floras 

 being nearly complete for thirty of them. False positions of the specie.^ 

 in the census list may be attributed much more usually to the imperfect 

 state of botanical knowledge, than to any inadequacy of the adopted 

 test towards meeting the object in view. A census list so made out is 

 really a summary of the records left by several generations of British 

 botanists, augmented by manuscript contributions from fully one half of 

 the most competent local botanists of the present age. Comparatively 

 with a census so obtained, the personal experience of any individual 

 botanist, or any score of iudividual botanists, sinks into insignificance, 

 as being unavoidably imperfect and partial, too local and unei|ua]. 



But in using the general list as a test or measure of rarity and 

 frequency, it is to be recollected that botanists do not record all plants 

 equally and impartially. Old records are often found to be unavailable 

 in tracing the area and frequency of modern segregate species, as has 



