J. P. Hall— Short Cycle in Weather. 229 



March 12, 13. Following B, much as C did A, another warm 

 wave, D, developed in the third curve on Nov. 2 and 3, and 

 was reproduced 'Nov. 27 and 28, Dec. 25 and 26, Jan. 20, Feb. 

 18 and weakly March 18. The second considerable decline of 

 temperature in the first period, F, culminated Sept. 11. This 

 was distinctly paralleled on Oct. 8, Nov. 5-6, Dec. 4, Dec. 31, 

 Jan. 28, Feb. 22, and ineffectively March 19, when it had 

 almost merged with H, a weak fall which separated B and T> 

 in the last six curves. 



In Series II, all within 1892, we have the warm wave A 

 culminating Jan. 2 and 29 and Feb. 26. In the fourth, fifth 

 and sixth curves it practically disappears ; in the seventh, 

 eighth and ninth, there is a hint of recurrence. Another 

 warm wave, C, weak in the first curve, is strong enough in the 

 second to check the reaction after A. It recurs weakly in the 

 third curve, but has good parallels in the rest of the series. 

 The first thermal depression after A in the first curve of this 

 second set is really made up of two marked cold waves, reach- 

 ing their greatest severity on Jan. 4 and 10. The former is 

 not well reproduced ; but the second, F, is apparent right 

 through the whole nine traces. In this series, as in the pre- 

 ceding one, there is a second conspicuous rise in the curve, B, 

 manifest at least five times in succession. Whether or not the 

 less distinct elevations B 2 , in the other four traces, should be 

 regarded as repetitions of it, these latter are at least recogniz- 

 able. Another warm wave, E>, is in nearly every instance 

 inferior in rank to the two or three preceding it in the same 

 cycle (A, C and B) ; yet it has a certain weak individuality 

 that entitles it to notice. A fifth tendency to warmth, shown 

 in the two peaks grouped as E in the first curve, finds a paral- 

 lel in the next, but none in the third and fourth ; and the sub- 

 sequent imitations are too feeble to count. The warm waves 

 B and D are at first fairly well separated by a fall in tempera- 

 ture, Gr, but this gradually fails as D grows in importance. 

 The reaction, H, after both B and D is in most cases greater 

 than Gr in the first four curves ; but in the latter half of this 

 series it fades out of existence gradually, and cannot fairly be 

 identified with H 2 in the eighth and ninth traces. 



In Series III the same tendency of prominent features to 

 recur is at least equally obvious. 



In all three sets of curves, and in those for other years, this 

 apparent disposition to repeat an incident is more conspicuous 

 at some times than at others. At best, it is never very strik- 

 ing; often a stretch of the imagination is required to perceive 

 it. Seldom are all parts of a curve reproduced equally well in 

 the next ; some one member, perhaps, failing altogether to re- 

 turn, or even being represented by a reversal. Occasionally a 



