RELATION OF CLIMATE TO SPREAD. 6.) 
esting to see how accurately this distribution has been followed. Pro- 
fessor Smith in 104 called attention to the fact that the spread of the 
insect in New Jersey seemed to be limited on the north by the so-called 
'•red shale" line, extending, approximately, from Perth Amboy on the 
east to Trenton on the west. If this limitation had been absolute it 
would have been curious and unusual, since the transition enters New 
Jersey only at the northwestern corner. Professor Smith, however, 
writes us under date of December 29, 1895, that the scale has been 
found at Washington, Warren County, north of the State line, and at 
the borders of the transition belt. The more northern occurrence of the 
scale in Columbia County. N. Y.. is similarly significant, since the upper 
austral zone extends up the Hudson River. The occurrences at Lewis- 
burg'. Bristol, and Atglen, Pa., are all within the extension of the upper 
austral into the southeastern one-fifth of Pennsylvania. The two West 
Virginia occurrences are both in the upper austral strip in the western 
portion of the State. The three Idaho occurrences are all in the narrow 
upper Sonorau or upper austral band along the Snake Eiver, with the 
exception of the one at Lewiston, which is the only locality in the pan- 
handle of Idaho where the Sonoran dips in from the west. The occur- 
rence at Amherst. Mass.. may be explained by a hitherto not thoroughly 
realized extension of the upper austral up the Connecticut Valley, similar 
to the extension up the Hudson, and which the distribution of certain 
other insects seems to justify. 1 Those about Boston are probably to be 
explained by a more or less definite coast law which finds its extreme 
on the Pacific Slope, and which causes considerable overlapping of 
northern and southern forms. Should future observations support the 
apparent significance of the occurrences so far known, the scale will 
uot establish itself to any serious extent in transition regions. This 
fact will relieve the fruit growers of much of New England; those 
inhabiting the greater portion of Pennsylvania, except in the south- 
eastern one-fifth and a western strip; those in Xew York, except for 
the strip up the Hudson Eiver and the loop which comes in from the 
northwest and includes the counties bordering Lake Ontario- on the 
south, as well as those inhabiting the northern portion of the lower 
peninsula of Michigan (except for a strip along the east border of Lake 
Michigan) and all of northern Wisconsin, from any fear of this insect. 
Such a condition of affairs would seem almost too good to be true, but 
its possibility is suggested by what we know up to the present time. 
Against its probability may be urged the fact that, in general, scale 
insects belong to the group of potential cosmopolites, and that they are 
seldom restricted by geographical limitations which hold with other 
insects. 
1 This extension has recently been accepted n.\ Dr. Merriam. 
