﻿104 



PROCEEDINGS OF THE NATIONAL MUSEUM 



M, with the histograms in the right column and the means, to their 

 left, represented by straight lines. The upper line is the mean for the 

 entire population sample, and, the line below, the mean for the top 10 

 percent. These two lines could never have the same length, but the 

 nearer they come together, the more natural will be the curve. Any 

 preponderance of young would lower the mean for the entire popula- 

 tion. It will be noticed, from a comparison of the proximity of the 

 two means and the configuration of the histogram, that colonies such 

 as A, B, E, and K, in which the two means are closest, have more 

 compact histograms approaching the normal curve. (L is based on 

 only 8 specimens and cannot be considered significant for this purpose.) 

 On the other hand, samples D, F, and especially G and J, whose two 

 means are relatively far apart, possess histograms that are consider- 

 ably distorted and drawn out by a high percentage of young. A 

 resume of the statistics is given in table 4. 



Table 4. — Shell lengths of Guam colonies 



Although this method of comparing colonies lacks the usefulness 

 and accuracy of presenting the true nature of each population, i. e., 

 the percentage of young, the modes of one or more growth stages, and 

 so forth, it seems to be the only recourse in problems confronting us 

 in those invertebrates that show no morphological signs of maturity, 

 and that grow throughout their life span. In reality, it answers only 

 the question "how large does this species grow in this environment?" 



A map of the southern half of Guam Island has been drawn and is 

 presented in figure 40. The collecting localities for colonies A to M 

 have been spotted, and the relative size of the top 10 percent of the 

 individuals entered in the form of outlined shells. It will be noted 

 that there is no geographical cline evident in their distribution. The 

 only possible correlation noted is that between size of the large streams 

 and size of large shells, but unfortunately insufficient studies of the 



