30 MEMOIRS OF THE NATIOj^AL ACADEMY OE SGIEliOES. 



feature is in the storm column, wliicli gives an excess of tbii-ty-one storms, or an average of nearly 

 two and a half per mouth above the mean, and by analyzing we find ten mouths above the mean 

 and not one below. Thus it is manifest, by geueraliziug on the extremes of maximum months of 

 pain readings, that we add great potency to the law of relationship between pain and storm. 



By going to the other extreme and taking a similar number of minimum pain readings, Table 

 C, we find the values reversed — a slight deficiency in relative humidity, with large positive value 

 of force of vajjor. The number of days of raiu fall very short of the average, being twenty-six 

 days for the thirteen mouths below the mean. The dejith of rain falls short nearly 17 inches, 

 while the temjjerature was over 10° above the mean, or nearly 1° higher iier mouth than the average. 

 The number of storms is almost as far below the mean as they wereTound to be above in Table B. 



It will be seen in Table C that there are three small + values in the cohimu under "number of 

 storms;" but if we turn to the Signal Service chart for June, 1878, we read, " Four of the storms 

 too iudefinite to be computed," which if canceled would give us a —3 instead of +1, and by a 

 similar reference to May, 1880, chart we find three belong to the Pacific, and finally for the other 

 apparent excei)tion, the August, 1884, chart places most of the "lows " at^an unusually high lati- 

 tude, " above latitude 48°," and most of them beyond or at extreme neuralgic distance. From 

 these two tables (B and 0) of extremes we, then, are enabled to state the following law: That maxi- 

 mum pain hears a direct proportion to storm frcqueney and an inverse proportion to temperature and 

 elastic force of vapor, and minimum pain hears an inverse proportion to storm frequency and a direct 

 to temperature and elastic force of vapor; while depth of rain, etc., go along with numher of storms and 

 maximum pain. 



In this connection it is evident that meteorological charts of relative stor^n frequency would be of 

 service to the neuralgic subject in selecting a habitation. 



"spells" of weather. 



In the annual changes of weather, from January to July and fiom July to -lanuary, and, in- 

 deed, in regard to movement of storms in general, the law of mean change is uniform, but if we 

 differentiate the mean we can in any one year detect a departure from the law of uniformity, by 

 an injection here and there of an abnormal break which may last one to three weeks. These 

 breaks ai-e more likely to occur in late full or early spring, but may happen anytime, and these are 

 known as "spells," and may be assucintid with the cold or wai-m "types" of weather, or, indeed, 

 with any unusual condition probably brouglit about by the changing shapes and relative position 

 of the great anticyclone areas. Eaily in the history of this case the pain, both in amount and 

 intensity, became associated with these " si)ells," the intenser forms of pain occurring at beginning 

 of the " spells." So also when a uniform condition of weather prevails, as in period of drought, a 

 diminution of pain obtains. 



MONTHLY RATIOS OF PAIN. 



In determining the relative ratios of the monthly products of pain, fifteen j'ears were taken, 

 1875-1889, inclusive, and the following is the order, beginning with the highest: February, .176; 

 March, .175; January, .174; December, .168; November, .164; April, .161; May, .160; July, .159; 

 June, .154; September, .153, and October, .153, in which we fiiul January, February, and March 

 about equal; then follow December, November, April, May, and July, leaving June, August, Sep- 

 tember, and October the least harmful. 



By a reference to our last paiier (p. 7, 1883), the order there given for monthly products of 

 pain was corrected in proof, but the correction was omitted in some way. In that order as it 

 stands no allowance was made, iu determining the ratios, for the inequality in the number of days 

 for the several months. 



In the order as given above the ratio for February is obtained by dividing the total number 

 of hours of neuralgia for this month — for the period of fifteen years — by the zchole number of Feb 

 ruary days for the same period, and so on for each month. 



