336 
Observations on Termites 
A careful comparison of these frequencies will, I think, lead bo the opinion 
that both of the suggested causes have a share in producing the seasonal variation. 
The apparent elimination of the small and unfit is best shown in nests " 630 " and 
" 672 " for both the small soldiers and large workers. 
(5) Ratios and Variation of the different Castes. 
Having thus noted the presence of a seasonal variability in a single nest, we 
may pass to the variability exhibited by a series of colonies. It would, of course, 
have been preferable to have avoided the disturbing element of the seasonal 
variation by collecting all material at one season of the year ; but this was not 
done as the presence of a seasonal variation was not suspected. 
In the following table (Table IV.) are given the arithmetical means of the 
means, standard deviations and coefficients of variation of the four castes in a 
number of nests (see Table I.). 
TABLE IV. Means. 
Caste 
Number 
of Nests* 
Mean 
Number 
of Nests t 
Standard 
Deviation 
Coefficient of 
Vaiiation 
Large Soldiers ... 
23 
434-01 + 
20 
13-64 + 
3-12 
Small Soldiers ... 
30 
245-00 
30 
7-37 
3-02 
Large Workers ... 
27 
242-21 
27 
5-94 
2-46 
Small Workers ... 
19 
156-90 
12 
4-17 
2-65 
Male Nymphs ... 
1 
300-05 
1 
4-71 
1-57 
There is no overlapping of the large soldiers and small soldiers or of the large 
workers and small workers ; in other words, there was never any question whether 
any individual should be classed amongst the " large " or " small " series. 
With reference to the particular dimension measured there is overlapping 
between the small soldiers and large workers ; but the soldiers and workers are 
totally distinct in general structure and shape. 
Ratios. 
Mean of Small Soldiers 
and 
Mean of Large Soldiers 
Mean of Small Workers 
245-001 . 
X 100 = TTrr^^i w X 100 = 50-4, 
434013 
1.56-901 
n-r ATT 1 X 100 = p^ T^— ^ X 100 
Mean ot Large Workers 242-206 
64-8. 
Thus, in the proportion in size that occurs between the " large " and " small " 
classes, the soldiers and workers approach each other ; the " small " class being 
more than half the size of the " large " class. 
and 
Mean of Large Workers 
Mean of Large Soldiers 
Mean of Small Workers 
X 100 
X 100 = 
242-206 
434013 
156-901 
X 100 = 55-S, 
X 100 = 64-0. 
Mean of Small Soldiers '~ 245-001 
* Nests from which less than 10 individuals were available are here excluded, 
t Only nests from which more than 23 individuals were available are included. 
+ Unit = 0-01 mm. 
