PERIODICITY OF GOOD AND BAD SEASONS. 101 
ora bad one? He has apparently thrown over his rainfall and 
temperature statistics, and his bald statement that such-and-such 
a year was good or bad seems far from convincing. I am afraid 
that the diagram which he shews us to-night, based upon these 
statements, is useless, if not misleading. I object also to another 
detail in the diagram. Mr. Russell ‘draws a vertical red line 
through A between the first and second years.” He says that the 
interval “ was regular and exactly nineteen years.” Also this 
so-called A drought varies in its length, ‘lasting from three to 
seven years.” It follows that a line drawn so as to mark the 
middle of an A drought would not have recurred at equal intervals 
of nineteen years. Yet the middle seems to me a more natural 
datum line. 
Again, periodicity cannot ‘be shewn by picking out a drought 
in Egypt, another in India, and a third in Australia, all happen- 
ing at different times. Mr. Russell distinctly lays down as his 
thesis “the salient points (in the weather) in our century are 
repetitions of the salient points in all past times, and probably in 
all countries.” To prove this he must produce evidence of a con- 
stantly recurring period in Egypt, of an identical period in 
Australia, and so on, independently. This he has not done. 
I myself have examined certain weather statistics, published in 
Vols. xv. and xx1., of the “Smithsonian Contributions to 
Knowledge,” which refer, in one case, to the ‘ Precipitation of 
Rain and Snow,’ and in the other, to the ‘ Average Temperature, 
over a considerable part of the United States of America. These 
volumes I lay before you to-night. The first table shews minima 
in 1818-9, 1836-7, 1856—just enough agreement with a nineteen 
years’ cycle to tantalize—but the other minima do not conform to 
such a cycle, nor do the maxima. In the second table minimum 
temperatures occur in 1785, 1797, 1816, 1835, 1857, 1868, and 
maxima in 1793, 1802, 1826, 1845, 1865. Mr. C. A. Schott, 
who compiled these tables says, “the average of the longer waves 
is about twenty-two years.” I would point out that such an 
Average does not assist Mr. Russell in any way, but quite ee. © 
~ 
