6



which the water-routes carry, and, for the conduct of local
business, stimulated every possible means and form of trans-
portation.  Nevertheless, railway construction -is not always
attempted judiciously, or to the best advantage of all con-
cerned. Lines which are little needed, and render no service
commensurate with the cost of their construction, are some-
times built with money which would have sufficed to complete
roads that could be operated successfully. It is no easy matter
to suggest criteria by which the necessity or benefit of addi-
tional railroad service to a territory already provided with it
may be determined. Very many considerations must influence
the solution of such a question, and the data are not always
the same in all cases.
  In his exceedingly interesting paper, published in October,
1881, entitled the " Standard of Railway Service," Mr. Edward
Atkinson has essayed to compute the railway extension that
will be necessary within the next twenty years from that date,
to meet the demands of the whole country. Assuming, as a
standard of comparison, the railway mileage of Massachusetts
at "1,950 miles in a territory of 7,800 square miles, or one
linear mile to each four square miles of territory," and defining
this as 100 per cent., he divides, for the purposes of such com-
parison, the States and Territories into five classes, and esti-
mates the increase of mileage in each, within the period
indicated, which their growth in population and general devel-
opment will probably require. It will be observed that he does
not base his comparison upon Massachusetts' ratio of railway
mileage to her entire area of 8,315 square miles, for the reason,
perhaps, that, to use his own words, " a large part of the State
is mountainous or sterile, and does not need railway communi-
cation to one-half the extent in which such service will be called
for in the near future in many other States." It is to be
regretted that Mr. Atkinson, deterred, doubtless, by the magni-
tude of the subject, and the apprehension that what was meant
to be a brief paper might swell into a volume, has merely
stated conclusions, with barely a hint of the data and reasoning
upon which they are predicated. He computes, that by the date
of the expiration of these twenty years, there will have been
needed and built 117,500 miles of new railroads; basing his cal-