33o LAW AND FREEDOM IN HISTORY

by interpreting the four-years cycle as an 'inventory cycle' whose 'charac-
ter stems from the nature of the merchant's trade', and 'which one
could, almost certainly, trace back into the eighteenth century, and per-
haps even back to mediaeval times'.1 He points out 'that, until about
the [eighteen-] fifties, the principal British exports were consumers' rather
than capital goods',a but that, 'from the late lySo's at least, , . . this
rhythm is woven into the longer and deeper rhythm of fluctuations in
long-term investment',3 and that the sixth decade of the nineteenth cen-
tury marks the date at which, in the economic life of Great Britain, long-
term investment in capital goods supplanted the production of con-
sumers' goods as the country's major economic activity.4 'The two types
of fluctuations', however, 'did not pursue their course in separate and
discrete channels. They were linked in at least four ways.'5

Students of the economic history of the United States in the Industrial
Age had here descried a cycle with a wave-length of 43 or 43 months on
the average.6

'The average (and the model) American cycle seems to be made up of
two unequal segments, a two-year period of gradually increasing activity, and
a period, four to six months shorter, of less gradually shrinking activity.'7

The same school of American investigators had seen in this threc-and-a-
half-years cycle in the United States one local variant of a kind, likewise
exemplified in the four-years cycle in Great Britain, of which there were
other local variants to be found in the contemporary economic histories
of France, Germany, and Austria8—the chronological locus of this short-
wave cycle lying within miminum and maximum limits of three and six
years.9 On the other hand, in the opinion of a Belgian scholar,

Tour qui dcpasse le cadre de 1'histoire des HJtats-Unis, la distinction
entre le cycle de sept a dix ans et le cycle court de quarante mois ne trouve
plus aucun semblant de confirmation dans les faits; le cycle de quarante
mois n'existe sirnplement pas.'10

A British scholar,11 who likewise discounts, as being dubious, the evi-
dence for the existence 'of a purely commercial short wave of about 3 J
years', also makes the same reservation of judgement about the evidence
for the existence 'of longer waves of from 20 to 80 years' duration'.
'Long cycles of remarkably regular duration*—with a chronological locus
between minimum and maximum limits of fifteen and twenty years-
were descried nevertheless, by some investigators, in the history of
building construction in the United States;12 and, in the general econo-
mic history of Great Britain between A.D, 1790 and A.D. 1914, five phases
of an average wave-length of just under twenty-five years were descried
by one of the investigators already quoted,13

1 Rostow, op. cit, p. 41. Cp. pp. 39-40, * Ibid., p, 41, n. i,

s Ibid., p, 41 •* Ibid., pp, 42-43. s Enumerated ibid,, p. 43,

6 Mitchell,, Business Cycles, the Problem and its Setting p, 341, See also Huntington,
op. cit, pp. 463-8.

^ Mitchell, ibid., p. 337. *> See ibid., pp, 385 and 390-1.

' See ibid., p. 45?- xo Dupriez, op, cit., vol. u, p. 280,

« Professor T. 8. Ashton, in the letter quoted above.

» See Burns, A. P,, and Mitchell, W. C.: Measuring Business Cycles (New York 1946,
National Bureau of Economic Research, Inc*),p, 418, » See Rostow. on. He. n n