60 EXPLANATIONS OF THE FORMULA. 



Hebrides, &c.) being reckoned as counties of themselves. 

 The intermediate numbers, between 15 and 80, are taken 

 in steps of 5 or 10, since the use of units would there be 

 only an affectation of exactness ; and the natm-e of the test 

 throws the larger proportion of the species towards one or 

 other extreme of the scale. The series of nmnbers actually 

 used in the census of counties nms thus : — 1, 2, 3, 4, 5, 6, 

 7, 8, 9, 10, 12, 15, 20, 25, 30, 40, 50, 60, 70, 75, 80, 81, 82. 



The fifth line indicates the range of latitude and the 

 geographic type. To show the limits in latitude, those 

 figures are used which correspond with the mathematical 

 lines on maps, between which all the known localities ai'e 

 situate. No fi-actions of degrees are used. Thus, in marking 

 the range of Clematis Vitalba, as 50 — 53, it is not intended 

 that localities for the species do certainly occur under the 

 mathematical' lines of 50° and 53°, but simjjly that they 

 occur more southerly than 51°, more northerly than 52°. 

 The native localities of the Clematis all occur under the 

 51st, 52nd, and 53rd degrees; and thus the first lines which 

 indicate whole degrees, south and north of the extreme lo- 

 calities, ai-e those of 50° and 53°. Some little uncertainty 

 will arise occasionally in applying the figures. For ex- 

 ample, the parallel of 59° crosses the Orkney Isles, and 

 there are no records adequate to show which of the Orkney 

 species, if any, do not pass to the northward of that line. 

 In this uncertainty, to distinguish those of Orkney fi'om the 

 species which cease in Sutherland or Caithness, all the 

 former are assumed to be bounded northward by the line of 

 60°. At the other end of Britain, the Coniish species are 

 assumed to be all on the north side of 50°, notwithstanding 

 that the Lizard Point is rather short of that line. Of the 

 Scilly flora, almost nothing is yet known ; and hence the 

 parallel of 49° is scarcely used in this work. 



The types of distribution have been explained in former 



