Gammarus Variability 
137 
Table 1. Scores of eye facet packing density (high, intermediate, low), lobe 
orientation (broad lobe up, equal, down), and lobe angle (more acute, 
intermediate, more obtuse). Localities are in rank order, with the most 
epigean at bottom. 
Packing density Broad lobe Lobe angle 
High 
Inter. Low 
BLU 
Equal 
BLD 
Acute 
Inter. 
Obtuse 
Emma Spring 'S 
11 
14 
0 
1 
6 
18 
4 
10 
11 
9 
7 
18 
0 
1 
7 
17 
2 
12 
11 
Smoke Hole 
0 
^25 
0 
4 
12 
*9 
5 
16 
4 
0 
25 
0 
4 
4 
17 
3 
14 
8 
Greenland Gap 
3 
22 
0 
3 
9 
13 
4 
17 
4 
8 
17 
0 
2 
7 
16 
2 
14 
9 
Cunninghams 
8 
17 
0 
2 
9 
14 
2 
13 
10 
6 
18 
1 
1 
9 
15 
3 
11 
11 
Marklesburg 
7 
18 
0 
3 
9 
13 
10 
12 
3 
10 
15 
0 
1 
11 
13 
4 
11 
10 
Church Camp 
3 
22 
0 
3 
9 
13 
4 
13 
8 
8 
17 
0 
1 
9 
15 
5 
13 
7 
Petersburg I 
4 
21 
0 
0 
11 
14 
4 
11 
10 
10 
13 
2 
1 
7 
17 
1 
10 
14 
Petersburg II 
7 
18 
0 
3 
8 
14 
5 
14 
6 
9 
16 
0 
3 
6 
16 
2 
12 
11 
James Creek 
6 
19 
0 
0 
8 
17 
4 
13 
8 
18 
7 
0 
0 
8 
17 
7 
12 
6 
Total 
49 
176 
0 
19 
81 
125 
42 
119 
64 
cfcf 
Total 9 9 
76 
146 
3 
14 
68 
143 
29 
109 
87 
be treated together. These characters, except for eye length, also were in- 
vestigated by Holsinger and Culver. They are correlated with body length 
by the relative growth equation T = a + c, in which Y is any one of the 
above characters as the dependent variable; X is the independent 
variable, body length; a is the slope of the regression line; b is the coef- 
ficient of allometry; and c is the intercept of the regression line on the or- 
dinate. If b is other than 1 there is nonlinear relative growth (allometry) 
of parts. If ^ = 1 there is no allometry and the equation simplifies to T = 
a X c, the linear regression equation. Allometry in populations of dif- 
fering size distributions would make meaningful comparisons more dif- 
