209 
Willis .— Age and Area. 
Mr. J. T. Cunningham at Hull, that it shows chiefly in generic and family 
groups, rather than in species, so that any theory of evolution that begins 
with the species, as does the Darwinian, must fail in its explanation. But 
a separate paper is required to deal with so large a subject. 
Some Statistics of the British Flora, to show the Trifling 
Differences due to ‘ Lumfing ’ and ‘ Splitting ’, or to Con¬ 
tinued Work upon the same Flora. 
An objection often brought up (cf. 11 , p. 98) is that my figures, if not 
indeed accidental, depend upon the fact that the floras used are by 
a particular type of systematise It is contended, in other words, that the 
flora of a 4 lumper ’ will give a very different result from that of a ‘ splitter’. 
Others contend (cf. 11 , p. 84) that the figures will be vitiated by further 
work upon the floras concerned. 
I have already mentioned that in fact neither of these objections can be 
supported when the actual facts are examined, but in order to show their 
inapplicability in a striking way, I have taken at random ten British floras 
which I found in the University Library in Cambridge, ranging from 1782 
to 1908, or over a period of 126 years during which much work has been 
done, and including the work of such 4 lumpers ’ as Bentham, and such 
‘splitters 5 as Babington and the editor of the London Catalogue. 
Though in some cases hardly a genus or a species has the same limita¬ 
tion in two of these floras, it will be seen at once that the general result is 
one of the most extraordinary similarity. There cannot be the slightest 
doubt that the curve is the same curve in all cases, with only unimportant 
variations. All the floras show the same hollow curve. And from this 
I think that one may assume that a similar result would follow in any other 
case, as indeed I have found to be the case in other instances. 
The actual figures for the distribution of the genera in these various 
floras into sizes are given below T , commencing with the flora that shows the 
greatest number of genera represented in Britain by one species only. 
Flora. Genera cf 1. 
2. 
3 - 
4 - 
5 * 
10. 
Larger. 
Largest. 
A. 
Lindley, 1859 
271 
102 
52 
2 5 
18 
5 
3 /i 4 
■ 1/65 
B. 
W. J. Hooker, 1830 
254 
S3 
50 
2 7 
21 
6 
3/14 
1/66 
C. 
Bentham, 1878 
252 
84 
49 
22 
23 
4 
4 /i 5 
i /47 
t>. 
Babington, 9th ed., 1904 
249 
92 
43 
37 
20 
5 
2/14 
i /97 
1/69 
E. 
j. D. Hooker, 1870 
London Catalogue, 
10th ed., 1908 
228 
90 
53 
27 
18 
6 
3 /L 5 
F. 
227 
S3 
42 
27 
20 
7 
4/14 
I / I 3° 
G. 
London Catalogue, 
8th ed., 1890 
223 
90 
35 
3 2 
16 
3 
8/11 
1/72 
II. 
Hudson, 1798 
198 
76 
44 
20 
18 
4 
2/16 
i /35 
J- 
Sowerby (Smith), 1S14 
191 
74 
44 
38 
18 
6 
3 /L 5 
i /57 
K. 
Enchiridion Botanicum, 
1782 
177 
82 
54 
26 
1 r 
3 
3 /L 5 
1/38 
It is clear that though the ‘splitter’ 
pours 
contempt upon the ‘ lumpe 
