HOWE: NEW ENGLAND ODONATA 109 
difficult as it is interesting. Erosion, the elevation and submer- 
gence of coastal lands, soil conditions, and glacial invasion are ex- 
planatory of change, though in the last case the discovery of for- 
ests supported by glacial waste on the surface of the Malaspina 
Glacier in Alaska complicates the hypothesis even more. INIr. R. 
J. Tillj^ard in his distributional studies of Odonata in Australia has 
advanced some interesting geologic explanations. The extensions 
of lowland faunas to uplands by base-leveling, and valley segrega- 
tions are very illustrative of physiographic influence (Walker, 
Can. Ent., 40 : 171, 1917). The proximity of the ocean and ocean 
currents exert definite controls that must not be ignored. INIan's 
influence also through irrigation and land-making is, and will be, 
of very considerable consequence. 
Temperature and Moisture 
That temperature and moisture play a very important part in 
distribution is clearly evident. Temperature, however, as governed 
by latitude or elevation from the point of view of zoogeography is 
a misleading factor for the question of exposure is the mere im- 
portant. Our planet for example reaches perihelion in July, but 
it is nevertheless our hottest month. The stability of temperature 
perhaps not only explains the overlapping of boreal and austral 
species in our so-called "cold-bog islands," but along the seacoast 
as well. A nice correlation in temperatui'e is to be noted in the 
fact that typically boreal species, or restricted boreal species oc- 
cupying middle elevations, e.g., WiUiamsonia lintneri, Agrion 
amatuni, when inhabiting stations near or beyond the general 
southern limits of their range during their imaginal life appear in 
the cool months of April, May, and early June, and that the most 
widely distributed species have the longest flight season, e.g., 
Anax Junius. Moisture seems not to govern distribution as di- 
rectly as it does size in development and pigmentation (Scuddei', 
/oc. al, 86-89, 1889). 
Haloi'hile Control 
It seems to have been fairly well ascertaincMl that no species of 
Odonata are halophilc. 'V\\v niarinc httoral (listril)iition of sc^veral 
species is pr()ba])ly more accurately exi)hiine(l by th(> necessity of 
a stable temperature condition for both larval and imaginal states, 
