336 
Merism mid Sex in " Spin ax Niger" 
TABLE 6 b. 
B. Anterior Spine. 
Number of segments 
Number of specimens 
Calculated values for position] 
of anterior spine J 
Then 0-= ±-284. 
Observed values for position of anterior spine : 
14 15 16 
Number of specimens ... 2 56 256 
Whence o-=±-624. 
63 64 
65 
66 
67 68 
8 58 
140 
113 
52 9 
15-5 15-75 
16 
16-25 
16-5 16-75 
c. 
TABLE 6 c. 
First girdle-piercing Nerve. 
?s 
(296) 
/Number of segments 
Number of specimens 
Calculated values for position! 
j of first girdle-piercing nerve J 
Mean number of segments = 65-5 ; mean of first girdle-piercing nerve = 29-0. 
63 64 
65 
66 
67 
68 
4 44 
120 
80 
42 
8 
27-89 28-34 
28-78 
29-22 
29-66 
30-11 
(168) 
Whence o-= ±-461. 
Whilst observed value of 
-760. 
28-06 28-50 28-94 29-38 
Nrmiber of segments ... 63 64 65 
Number of specimens ... 10 38 74 
Calculated values for position] ^, 
of first girdle-piercing nerve/ 
Mean number of segments = 65-0 ; mean of first girdle-piercing nerve = 
Whence o-= ±-458. 
Whilst observed value of o- -= + -879. 
29-82 
=28-5. 
If a number of specimens of Spinax are taken, in each of which the number 
of total segments is known, we may, knowing also the mean number of total 
segments and the mean position of the posterior spine, calculate the theoretical 
value of the variability (ct) of the posterior spine on this hypothesis. This has 
been done for the characters anterior and posterior spine and 1st g. p. nerve in ^/s 
as well as $s, and the results are given in Table 6*. For the posterior spine this 
theoretical value works out to + "678, which is considerably less than the observed 
value + •942. The variability of the posterior spine is actually much greater than 
it should be on the excalation hypothesis. This is also true for the anterior spine 
where the observed value, + -624, is very much greater than the calculated value 
± •284. The same is also true for the 1st g. p. nerve where the calculated values 
* These correlations were worked out in 1902 on the material oollected in 1901 and 1902. It has 
been thought unnecessary to recalculate these figures by incorporating the material collected in 1903. 
The differences that would be introduced into the above values would probably be very slight. 
