BIOLOGICAL SURVEY OF WOODS HOLE AND VICINITY. 187 
There are several ways in which two species may differ in respect to their dis- 
tribution patterns: (1) The species may occur throughout practically the same area, 
differing only in their relative abundance; (2) the range of one may be restricted to 
a portion of the area occupied by another; (3) they may have distributions which are 
in a certain degree complementary to one another. 
Referring to the genera named, it will be found that in few cases, if any, are the 
distributions of two members of the same genus practically identical, both as regards 
the area inhabited and the frequency of occurrence. In a number of instances, how- 
ever, they differ only in respect to frequency. 
The type of difference most often realized upon our charts is the second among 
those mentioned above. In such cases one species may occur throughout only a por- 
tion of the territory occupied by the other; or, at least, it may not be well established 
except in this portion. 
The third condition—that of two species of the same genus having distribution 
patterns which are complementary to one another—is realized in a surprisingly small 
number of cases upon our charts. It appears most clearly from a comparison of the 
distributions of Pagurus acadianus (chart 110) and P. annulipes (chart 112). We 
have seen that these are respectively northward-ranging and southward-ranging spe- 
cies; so that the habitat selected by each in local waters is not improbably determined 
by temperature. 
It is a familiar fact to field naturalists that the various members of the same genus 
frequently, if not generally, occupy somewhat different habitats. Obvious instances 
of this are not uncommon among our local littoral and shallow-water fauna, as for 
example the three familiar species of the genus Littorina. Now, our dredging charts 
are not adapted to revealing such slight differences of habitat as may occur within 
the limits of a single ‘‘station.” In the charts for Crepidula, for example, there is 
nothing to show that C. fornicata does not coincide in its habitat as well as its distri- 
bution, with C. plana; whereas we know that, in most cases, the latter occupies the 
inside of a hermit-crab shell, while the former may occupy the outside of the same 
shell, or may adhere to any solid object whatever. It is probable, likewise, that the 
drifting of shells and other lifeless remains may result in an apparent obliteration 
of actual distinctions in the distribution of species. Finally, it seems needless to 
remark that in no single case is the entire range of a species indicated upon one of our 
charts. Thus, even in cases where two species appear to coincide in their distribution 
locally, the range of one may extend into far deeper water, off the coast, than that of 
the other. It seems to us, therefore, that the differences in the distribution of closely 
related species have been minimized, rather than exaggerated, in our graphic represen- 
tations. 
Whether or not specific differentiation preceded or followed these changes of hab- 
itat, or whether they went on pari passu with such changes, is not even suggested by 
any of the facts which we have encountered. Who can say, for example, whether the 
tendency to restrict itself to muddy bottoms preceded or followed the differentiation 
of the amphipod Ampelisca macrocephala as a species distinct from A. spinipes? Never- 
theless, the bare fact that various closely related species do show decidedly different 
distribution patterns is one of great interest, for it shows that the slight morphological 
differences by which the species are distinguished from one another are oftentimes 
