CONTEIBUTIOXS TO THE THEOEY OF EVOLUTION. 
459 
This gives for the value 3892. 
Distance between mode and mean = 3'66106 yrs. 
Whence we find for ske’wness the value 
Sk. = • 5347 . 
The diagram (fig. iii.) shows that the fit may be considered a good one. 
^,000 
/ 
7 
f 
\\ 
7 
1 
irL 
\Lj 
ISU 
etc. 
1 
\ 
Fi 
.g.d 
UK 
ir 
\ 
\ 
1 
\ 
\| 
\\ 
\\ 
\\ 
\\ 
\\ 
I 
\\ 
\\ 
\\ 
\ 
\ 
V 
1 
1 
•V'- 
1 
1 
i 
1 1 
1 
SflOO 
r> 
(3 2,GOO 
I- 
I 
IfiOO 
10 
20 30 
A^e of Incidence. 
oO 
(9.) The conclusions of this paper are, I think, of some interest from tlie general 
standpoint of scientific investigation. A certain number of frequency distributions 
had been found, not only by my co-workers and myself here, but by biologists in 
America, not to fit into the general system of skew distributions dealt with by me in 
my original memoir. The first conclusion was that however wide-reaching that 
system ajipeared to be, it \vas a failure for a few" remarkably skew’ distributions. But 
on more careful investigation of tlie differential equation it appeared that tvvo types 
of solution had been left out of consideration, and that these were precisely those 
needed in the recorded cases of failure. 
I ow"e some apology to authors like Professor Davenport and Dr. Duncker, w"ho 
have recently issued text-books on the application of statistical methods to biological 
variation, because although w’e have kuowm and used these curves for some years past, 
no account has hitherto been published of them, and, consequently, biological 
investigators""' using their rhumh of my methods have been, and I fear still may be, 
occasionally puzzled. 
* E.g., Mr. A. G. Mayer in the paper on Medusse referred to above 
3 N 2 
