28 MEMOIRS OP THE NATIONAL ACADEMY OP SCIENCES. 



II is similarly constructed for the six summer months. Figs, iii and iv each represents the elec- 

 trical composite of sixty days also, covered by the two years used in Figs, i and ii upon which 

 the neuralgia prevailed. Comiiarison of these with Pigs, i and ii shows a lower mean value of 

 positive electricity and a decided increase in the negative product. Fig. iv has a larger per- 

 centage of summer neuralgia in it than has Fig. iii. 



Comparison, then, of one hundred and twenty curves of electricity taken on neuralgic periods 

 with one hundred and twenty curves taken as already stated shows that pain seeks identity with 

 low positive and negative curves and with fluctuating values of great amplitude. 



MAGNETIC STORMS ACCOMPANIED BY AURORA. 



In our original paper (p. 23, April, 1877,)* attention was first drawn to the association of neu- 

 ralgia and aurora during the decennial sun spot or auroral pgriod of 1868-1873. No record was 

 made of neuralgia at the time of this supposed relationship, and it became necessary to await the 

 next recurring period, which reached its maximum in 1882 and 1883. We repeated this belief of 

 relationship in our next paper (p. 18, June, 1883),t and our record then orought us to the middle of 

 the auroral period. Having passed through that period, and having the records nearly complete 

 of the magnetic and auroral displays, it is now possible to contribute additional evidence to what 

 originally seemed to be a relation or a coincidence. 



To cover the period of greatest auroral activity we have selected seven of the greatest electrical 

 storms from April, 1882, to November, 1883, and one isolated storm of March, 1885. In describ- 

 ing these eight storms we have used the monthly weather review, Prof McAdie's professional 

 note of the Signal Bureau, and the monthly reports of the Toronto Observatory and our own notes 

 made at the time. Referring to Fig. V the last column is the auroral record — the stars indicate 

 the visible display and the dots the suspected. Taking an exact copy of the contemporaneous 

 curve of pain and weather made at time and from day to day, and placing in jiositiou according 

 to date of auroras, and there is at once an obvious relation, and each aurora throws some coloring 

 on the pain and storm condition. But the greatest surprise of all was, having selected the mag- 

 netic storms and placing the other curves in position — to find that from April, 1882, to March, 

 1885, there were but seven attacks of pain of the greatest intensity, such as have already been 

 discussed in previous papers — and that of these seven, four of the greatest of them should turn 

 up in this diagram of a record of only eight magnetic storms. 



Corresponding to the attack of November, 1882, we have probably the greatest magnetic storm 

 of which any record is made. This and its relationship to the pain attack was described in paper 

 of June, 1883 (pp. 17, 18). As stated then, no cyclone storm was present. 



Passing to the next, of September, 1883, we find the pain curve from the 10th to the 14th was 

 prol)ably under the influence of the great hurricane storm No. iv, as traced in the monthly re- 

 view of this date, with falling barometer over nearly the whole country. The auroral display did 

 not occur till the 15th. According to the Toronto repoi't covering this time it seems while there 

 was some declination trouble on the 11th and 12th, it was not until the evening of the 15th, ac- 

 companied by the aurora, that the heaviest magnetic storm since the previous November began 

 and lasted all night and was renewed with cousiderable intensity the next night. We find jiain 

 of the usual intensity corres[)onding to this display. A cyclone storm followed on the 16th and 

 17th. 



Taking up the third (November, 1883,) long attack, as given in the diagram, we find a record 

 of four days' pain out of the week — 18th to 24th. The magnetic storm, as given by the Toronto 

 rec^ord, was active from the 19th to 22d, inclusive. Three cyclone storms also i)assed during same 

 period. 



The magnetic storm of March, 1885, was active from the 12th to the 16th, and especially so 

 during the brilliant aurora of the 15th, which was also reported as disturbing the magnetograph 

 at Kew (England) all day. Three cyclone storms moved within neuralgic range from the 12th to 

 the 16th. 



The aurora of the night of August 4, 1882, placed as on diagram, extended clear across the 

 continent and as far south as Florida. A cyclone depression on the 4th was within neuralgic 

 range. 



• American Journal Medical Sciences. t Vol. VI Transactions of College of Physicians, Philadelphia. 



