638 
BULLETIN OF THE BUREAU OF FISHERIES 
Pacific razor clam. In their work the relative growth rate, as initial rate and rate 
at two years, was correlated with geographical position, age, and length. The 
factors involved in geographical position are numerous and largely unknown, but 
include all of the physical and chemical features of the environment. 
Geographical position was given a numerical value for correlation by counting 
the distance along the coast from Pismo, Calif., the most southern beach from which 
statistically valuable data were gathered. The sign of the correlation, then, is 
entirely arbitrary. 
In order to compare the variations of growth in different localities, as observed 
in the cockle, with that mentioned above for the razor clam, similar constants were 
AGE IN YEARS 
Figure 7. — Growth curves of Cardium corbis from three localities 
calculated fof each locality. The values for geographical position were determined 
as in the razor clam, the distance being measured from Tillamook. Maximum age 
and maximum length were obtained as follows : On a survival curve for each locality 
the age was located at which 5 per cent of the clams passing through the first winter 
were still alive. This was taken as the maximum age. Owing to the small numbers of 
cockles available for some localities, this value is only a rough approximation, but 
the errors involved are much less than the differences in length of life. The maximum 
length is that length read from the absolute growth curve at the maximum age as just 
defined. 
The relative growth rate and acceleration were derived from the natural loga- 
rithms of the length as discussed earlier in the paper. These five constants for the 
eight localities are given in Table 2. It should be mentioned that Constantine Harbor 
and Port Moller do not represent normal habitats for Cardium. The beach at the 
former place is of coarse gravel, and Port Moller is near the northern limit of the 
species. 
