418 MARINE BIOLOGY OF THE SUDANESE RED SEA. 
The first number in each column of the Table on p. 417 gives the number 
of species recorded from the source in question, the number in brackets gives 
the number of these which are known to occur also in the Red Sea. 
Table II. is based upon Borradaile’s classification (Borradaile, 1907, 
p. 477 et seg.). Elsewhere in this paper I have arranged the material in 
Nobili’s order, to facilitate comparison with his ‘ Faune Carcinologique de la 
Mer Rouge.’ , 
It will be seen that the affinity of the Red Sea crabs is strongest with 
those of the Persian Gulf, 68 per cent. of the species recorded from the 
jatter locality having also been recorded from the Red Sea. 
The homogeneity of the Indo-Pacific region is illustrated by the fact that 
in places so far apart as Seychelles and the Hawaiian Islands the percentage 
of crabs common to the Red Sea is very similar, approximately 33 per cent. 
in each case, that this per cent. occurs at Ceylon and a fairly similar one at 
the Maldives and Laceadives. India is below and Torres Straits distinctly 
above this average figure. In estimating the significance of these per cents. 
one notes that the Red Sea, having been very fuirly explored by a variety of 
collectors, is a good standard ; but, on the other hand, one must bear in mind 
that other populations may have been sampled under different conditions, and 
it will be noted that the Torres Straits result, which appears somewhat 
anomalous, is associated with a comparatively short list. If allowance were 
made for these possible sources of error, the similarity might be even more 
striking. But taking the figures at face value, certain points stand out 
clearly : (1) the comparative similarity of the crabs over the whole Indo- 
Pacific area ; (2) the almost entire lack of specific resemblance between the 
erabs of the Red Sea and those of West Africa ; (3) omitting Persian Gulf 
as being in the immediate vicinity of the Red Sea, the Indo-Pacific figures 
suggest that one may prophesy with a probable error of + 5 or 6 that 35 is 
the most likely percentage of species common to the Red Sea which will be 
found in a collection of say 200 species of crabs from any hitherto insufii- 
ciently investigated portion of the Indo-Pacific Region ; the odds are about 
20 to 1 against the error in such a prophecy being greater than + 15. One 
may also prophesy with a probable error of + 2 or 3 that 35 is the most 
likely average percentage of species common to the Red Sea which will be 
found on examining six further samples of Indo-Pacific crabs of similar 
numbers to those which I have considered ; the odds are about 20 to 1 
against the probable error of such a prophecy being greater than + 7. The 
number of samples I have considered is of course small, and the attempt to 
give numerical precision to the degree of homogeneity of the Brachyuran 
fauna of the coral zone presents certain intricacies not considered here, but 
the figures I give are, I believe, not without some value and may form the 
basis of further investigation upon similar lines. 
The West African forms recorded from the Red Sea comprise only 
