48 
NATURE 
[January ay 1928 

includes eight families. One of these families, the 
Asteropectinide, contains species with non-attaching 
larve. Two other families (the Asterinide and the 
Gymnasteriidz) have species with attaching larve 
(Asterina, Q.J.M.S., 1896, and Porania, Q.J.M.S., 
1915). It is true that the Asterinide approach the 
Cryptozonia in some respects, but taken by itself this 
fact might rather lead us to look on the Asterinide as 
“ primitive ’’—an annectant family between the two 
great orders of starfish. The Gymnasteriide are 
frankly Phanerozonate. 
While Asteroid classification is admittedly perplex- 
ing, we are on fairly safe ground when dealing with 
the recognised families. At present it is known that 
members of five different starfish families (Gymna- 
steriide, Asterinide, -Echinasteride, Solasteride, 
Asteriide) have attaching larve, while members of 
only one family (Asteropectinide) have larve without 
a sucker. 
Dr, Mortensen’s virtual narrowing down of Phan- 
erozonia to Asteropectinide renders valueless his 
citations of Sladen, Ludwig, Hamann, and Gregory in 
support of the arguments in his letter. I yield to no 
one in appreciation of Dr. Mortensen’s work, but even 
if the adult Asteropectinide were in some respects a 
primitive family (I believe the opposite), still to draw 
the conclusion which he says “ inevitably ’’ follows 
from this premise, in defiance of the direct data of 
comparative Asteroid ontogeny, not to speak of other 
considerations, would surely be one of those strangely 
naive misuses of the Recapitulation theory which 
have done much to obscure its essential truth. 
JAMEs F. GEMMILL. 
University College, Dundee, 
December 22. 

Age and Area in Biology. 
In his recent book, ‘““ Age and Area,” Dr. Willis 
gives (p. 114) the following: ‘‘ Table showing in 
the horizontal lines the average number of vice- 
counties in Britain reached by the most widely 
distributed species in each genus of different sizes, 
and by the second, third, fourth, and fifth most 
widely distributed species.” 
Average No. of vice-counties reached by the 
istsp. e2ndsp. 3rdsp. 4thsp. 5th sp. 
Genus of over Io sp.. 108 104 96 86 79 
6 IO sp.. 103 84 64 49 33 
5:Sps - 98 76 39 22 16 
4\sp- 5 89 61 35 13 of 
3 sp. . 89 48 27 Se 
2 Sp. . 73 33 -- 
TESPs ce y5O 
Dr. Willis is convinced that the only explanation 
of the gradual diminution in average distribution 
from top to bottom of the table is that the average 
age of the species in the upper rows is greater and 
that they owe their wider distribution to their age. 
The gradation can, however, be explained without 
the help of either the principle of ‘‘Age and Area” 
or that of ‘‘Size and Space,’’ as will become obvious 
if the method of constructing the table be con- 
sidered. The average distribution of all the species 
in each of Dr. Willis’s classes, in part obtained from 
the table above and in part from the London Cata- 
logue, tenth edition, proves to be as follows : 
No. of spp.\ over 
in Genus { be) 
Average No. of 
vice-counties | 40 48 
reached 
10-6 5 4 3 2 I 
50°25 49°5. 54°6 53 50 
The averages thus vary somewhat irregularly. In 
NO.2776;°VOL: t11 | 

taking the average of the most widely distributed 
species in the first class, more than 90 per cent. of 
the lower numbers are rejected, in the second more 
than 84 per cent. are rejected, in the third class 
80 per cent., in the fourth 75 peg cent., in the fifth 
66 per cent., in the sixth 50 per cent., and in the 
seventh none at all. Naturally this changes an 
approximately equal set of numbers into a falling 
series. 
It is now possible to deduce the converse of Dr. 
Willis’s theorem; for by reversing his process and 
rejecting the higher numbers it can be shown that 
the age of the ‘‘ youngest’’ species decreases with the 
size of the genus. 
The average distribution in vice-counties of the /east 
widely distributed species in each genus according 
to size of genus, in part from Dr. Willis’s table and 
in part from the London Catalogue, is as follows: | 
No. of spp. over 
in Genus 10 
Average No. of 
vice-counties }-3 5°4 16 (“ARB © sagas 50 
reached 
10-6 5 4 3 2 1 
The regularity continues for the next “ youngest ”” 
species, as can be seen from the original table. In 
neither case would the regularity suffer if the vice- 
comital numbers were redistributed to the species 
by any random method, for the chance of a genus 
receiving both a very high and a very low number 
would increase proportionately with its size. 
W. C. F. NEwTon. 
The John Innes Horticultural Institution, 
Merton, October 31. 
Soaring Flight and the “ Olfactory ’’ Organs of 
Birds. 
THE note on page 784 of NaturE, December 9, 
misses the point of the theory I wish to be tested. 
The theory is that the well-developed “ olfactory ” 
nerves and apparatus of those birds which are 
capable of soaring flight has the function, not of 
smell, but of a delicate tactile sense whereby the 
bird is able to detect and take instant advantage 
of those upward air currents which recent experi- 
ments with gliding machines have shown to be so 
important in soaring flight. 
It has been shown by Darwin and others that 
vultures do not smell with their well-developed 
olfactory apparatus. The experiments referred to 
in NaturE of December 9 show that this nervous 
apparatus is not necessary to give the bird its home- 
ward direction or to enable it to indulge in flapping 
flight. So well-developed an apparatus is almost 
sure to have some function. It is obvious that 
soaring birds are in constant need of a means to 
detect the direction and strength of wind currents, 
especially those in. an upward direction, and to 
adjust their balance and their wings accordingly. 
When soaring, the eyes and bill of the bird are 
directed downwards and the mucosa of the nostrils 
is exposed to any upward currents of air. I think it 
very likely, therefore, that the well-developed “ ol- 
factory ’’ apparatus of these birds is a mechanism 
for detecting the direction and quality of air currents, 
and that the central “olfactory ’’ ganglia enable 
the requisite adjustments of balance and direction 
of wing and tail planes to be made. The fact that 
birds whose nostrils have been plugged have been 
able to fly home by flapping in no way contradicts 
this theory. 
To test it, I suggest in the first place that the 

