338 



NATURE 



iFeb. 7, 1884 



assistants, observed, at 9.45 p.m., an aurora appearing as 

 a lustrous green light behind the nearest hills. The top 

 of the mountain, " Sadlen," 1200 metres in height, was 

 distinctly seen aboiv the lustrous plane. The phenomenon 

 rapidly disappeared. M 10.45 a light was seen in the 

 south, which resembled that of dawn, and contracted 

 into a faint shining cloud, oblong in shape, which oscillated 

 slowly before the mountains " Hjortetakken" (1200 metres) 

 and " Store Malene" (800 metres) at a distance of 8 to 12 

 kilometres from the station. The tops of the two moun- 

 tains were distinctly seen a/wve the luminous cloud, while 

 at times small but intense spots of light developed them- 

 selves in it. When the cloud, at 11.45 p.m., had moved in 

 front of the mountain " Lille Malene," the light became 

 more intense, and had the appearance of a lustrous white 

 cloud of smoke rolling up the hill to north-east. When 

 the cloud travelled over the hill, the light became yellow, 

 and was bordered by a coloured rim. At 11. 10 it shot 

 three faint red streamers up towards the zenith, and then 

 the whole disippeared. 



The following phenomenon was observed by the writer 

 of these lines: — Dn November 14, 1882, at 6 a.m., I 

 observed an auroral band without streamers through 

 Vega, the Great Bear, and the Twins, while another 

 stood parallel with this further west. From the " Store 

 Malene" I now saw a peculiar shining white cloud 

 descend into the fjord below. It descended in long, 

 straight, shining bands. In a few minutes the mountain 

 in question, as well as " Hjortetakken" were completely 

 hi Iden in the cloud. A little further cast the cloud pos- 

 sessed greater intensity, while on the plain at the foot of 

 the hill on which the observatory stands, two luminous 

 gatherings were seen, which seemed to rest on the snow, 

 with a strongly phosphorescent light. These two gather- 

 ings, which were at first isolated, now came in contact 

 with the above described cloud with long luminous bands 

 radiating from the latter. By opening the slit of the 

 spectroscope as much as possible and simultaneously 

 keeping foreign light from the eye, I beheld the auroral 

 line faintly but clearly definei. The cloud now beg.in to 

 disappear without oscillation, when suddenly a number of 

 horizontal openings formed in it, through which the 

 mountain stood forth. In the next second all had 

 di, appeared. 



I admit that, as regards the last described phenomenon 

 of the lustrous cloud, it might be explained as being 

 caused by the reflex of the aurorx which were simul- 

 taneously visible ; but such an explanation is not appli- 

 cable to the one first described. It would be very 

 interesting to learn if other observers have noticed this 

 form of the aurora. Ada.m P.\ULSen 



Copenhagen 



TffE EFFECTS OF THE WEATHER UPON 



DEATH RATE AND CRIME IN INDIA 

 t^OME time ago a very interesting series of articles, by 

 •^ Mr. Buchan, upon the connection between certain 

 meteorological conditions and the zymotic diseases, as 

 illustrated by the mortuary returns of the London district, 

 appeared in .Nature. Happening to have undertaken, 

 at the request of the provinci.d superintendent of census 

 operations, certain investigations concerning the life 

 statistics of the population of the North-West Provinces 

 and Uudh, just about the time when Mr. Buchan's articles 

 ajjpeared, it occurred to me that it would be worth while 

 to see whether any similar concomitant variations of 

 meteorological conditions and causes of death could be 

 detected m India. The results arrived at are so curious, 

 and at the same lime so definite, that I think they may 

 be of interest to readers of this journal. 



At starting, however, it should be observed that, though 

 the mortuary returns of the province with which I 

 am connected are probably the best in India, they are 



still very far from complete. The agency employed for 

 registration is that of the village chaiikiddr or watchman, 

 who is supposed to take note of all births and deaths 

 which occur in his village (aiding his memory, if neces- 

 sary, by variously cut notches on a stick) and to report 

 these weekly at the nearest police station. From such 

 an agency nothing like an exact account of the causes of 

 death can be expected ; hence in the detailed tables given 

 below I have confined my attention to the four most 

 obvious causes — sniall-pox, cholera, suicide, and wounds. 

 Even as regards the number of deaths registered a con- 

 siderable defect may confidently be anticipated, owing to 

 lapses of memory on the part of the chaukidAr. This 

 defect has been found by Dr. Plauck, the Sanitary Com- 

 missioner, to amount to about 20 per cent, of the whole 

 on the average of a large number of cases personally 

 examined by him in various parts of the province. The 

 proportion thus obtained is confirmed by a comparison of 

 the deaths actually registered with the death rate arrived 

 at in the last census report. During the five years, 

 1878-82 (the only ones for which complete returns are 

 obtainable), the deaths registered appear, from figures 

 supplied by Dr. Plauck, to have numbered 7,311,013. 

 The average population during the five years having been 

 about 45,000,000, this gives an annual death rate of 32^5 

 per thousand. Now in Mr. White's report on the census 

 of 1 88 1 it is shown that the distribution of the population 

 according to age, and the observed death rate among 

 certain tribes and castes suspected of practising infanti- 

 cide, and therefore placed under strict police surveillance, 

 point to 40 per mille as the probable rate of mortality for 

 the general population. The unrecorded deaths therefore 

 amount on the average to 7 5 out of 40, or 19 per cent, of 

 the total — almost exactly the same defect as Dr. Plauck 

 arrived at by his personal investigation of special cases. 



It follows that, though the returns collected by the rude 

 illiterate agency cinployed are not strictly accurate, the 

 totals arrived at probably on the whole bear a nearly con- 

 stant proportion to the true number of deaths, the popu- 

 lation dealt with being sul'ficiently numerous to eliminate 

 any individual peculiarities of the agents. 



The death rate varies enormously from year to year, as 

 may be seen from the table of the total number of deaths 

 recorded, here given in full : — 



Number of Deal/is from all Causes Re^^isterci in the North- West 

 Provinces and Oudh during the Five Years 1878-82 



1S7S 1137,161 140,173 143,760 



1S79 75.3''7; 6^,837, 71.874 

 1880,116,366 72,030, 69,250 



1851 I 95,226 



1852 114,220 



91,011' 97,829 

 92,4721 96,596 



Total 538,360458,523-479,309 



April 



157,326 136,867 120,767 91,677 

 87.302100,040 83,802: 73,120 

 72,534 76,622, 78,200 56,502 

 124,831 115,683 86,083 81,609 

 107,628 119,714 114,382 122,110 



549,621 548,926483,234425,018 



Annual total 



1S78 113,701 120,6071138,997 127,6561 93,032 1,521,724 



1879 1131,702 I96,I35'429,IIS369,390233,795 1,914.499 



1880 74,127 87,618' 91,218 99,459 93,264! 987,190 

 i88i 86,316 io9,837|iS.', 519 180,683 151,846! i,402,473 

 1882 151,779 159,604 i56,o65'i28,040 122,5171 1,485,127 



Total 557,625673,8011996,914 905,228^694,454 7,311,013 



The deaths recorded average a little under a million 

 and a half per annum, but in 1880 they were less than 

 a million, and in 1879 nearly two millions. In that 

 disastrous year one district or county, that of Aligarh, 

 lost nearly half a million of its po l.ttion. The chief 

 difierence between 1879 and 1880, from the meteorological 



