548 
BULLETIN OF THE BUREAU OF FISHERIES 
Table 3. — Position of the umbo at each centimeter of length for Siliqua patula and Siliqua alia 
Length 
Position of umbo 
Length 
Position of umbo 
Length 
Position of umbo 
S. patula 
S. alta 
S. patula 
S. alta 
S. patula 
S. alta 
Centimeters 
0.0-0. 9 
71.43 
70. 75 
72. 83 
72. 67 
71. 67 
71.00 
Centimeters 
6.0 
70. 38 
70. 25 
69.20 
68. 00 
67. 86 
67. 17 
Centimeters 
12.0 
67. 33 
67. 73 
68. 65 
67. 70 
68. 25 
67. 25 
74. 44 
74.67 
76.00 
1.0 
7.0 
74.50 
72. 50 
72.00 
74. 25 
74. 29 
13.0 
2.0 
8.0 
14.0 
3.0 
9.0 
15.0 
4.0 . 
10.0 
16.0 
5.0 
11.0 
17.0 
DIRECTION OF RIB 
One very noticeable character of the shell is the direction of the rib. If the shell 
is opened until the two valves lie in the same plane, the two ribs of Siliqua alia form 
nearly a straight line, but those of S. patula lie at a distinct angle. The direction of 
the rib from the umbo was measured by placing the dorsal margin of the shell on the 
table and reading from a protractor the angle between the table anterior to the umbo 
and the posterior margin of the rib. The residts show a small individual variation 
which can not be correlated with age or size. In S. patula the angle is between 69° 
and 84°, in S. alia between 84° and 90°. 
From this study we are convinced that there are four species of the genus Siliqua 
on the west coast of North America. These include, in addition to S. media and 
S. lucida, generally recognized, S. alta and S. patula. The only species considered 
in the present growth study is S. patula. 
An additional conclusion applicable to other animals may be drawn from this 
study of body proportions. It is useless for comparative purposes to state the ratio 
between the measurements of any two parts of the body, such as length and width 
of the shell in the present case, or the head and the body length of a fish, unless the 
total size or age is also given. The variation of these ratios with size and age can 
not be foretold but must be determined for each species; they are, in the razor clam, 
so considerable that they can not be ignored. The greatest variation occurs in clams 
below 5 centimeters in length ; but if we disregard these, the larger ones still are widely 
variable. If we deal with what may justly be considered “adult” specimens, the 
change in proportional width of either S. alta or S. patula with size is as great as 
the differences between the species. In other words, while specimens of the same 
length show an average difference in width which no one would hesitate to call 
specific, large specimens of S. patula may be selected giving the same average width 
as those of small S. alta. This fact should be borne in mind by systematists in fram- 
ing specific descriptions. 
SILIQUA PATULA 
VARIABILITY 
The variability of each age class of clams has been measured in terms of “D ” or 
interdecile range (Kelley, 1921) as the most appropriate measure to accompany the 
median. This may readily be visualized as the range in length of the central 80 
per cent of each frequency distribution. 
This value for all ages and localities is given in Table 8. For four typical 
localities Figure 3 shows “D” plotted on length. It will be seen that the absolute 
variability rises to a maximum after which it again declines to become fairly constant 
