August 8, 1895] 



NA TURE 



o4j 



Kor such systems A,, as we have considered, all projective 

 properties will he corresponding to each other, and all metrical 

 properties at least as far as they are dependent upon the inter- 

 |)retation of the constants employed. Emanuei, Laskkr. 



IlUley, July 9. 



P.S. — The same holds true, with slight modifications, for the 

 only curved space that contains no exceptional elements, that is 

 the surface of a globe of « manifoldness. — E. L. 



The Feigning of Death. 



TllIO discussion, a few months since, of the feigning of death in 

 reptiles (vols. li. pp. 107, 128, 223, and lii. p. 148), induced me 

 to exjjeriment on the Currant Moth, whose |xj\\ers of " sham- 

 ming" are so familiar. The moth was first seized l>y one wing, 

 and it at once feigned death ; thereupon I cut off Us head with a 

 pair of scissors, and the aitinial (Oitliiuied to feign death. I use 

 the expression advisedly, for absolute inunobility was maintained 

 for some seconds, anil then violent fluttering ensued, causing the 

 animal to rush wildly about the table, but failing to lift it into 

 the air. In this condition any impulse, .such as touching or 

 pinching, inducefl a repetition of " shamming." .\fter a strong 

 stimtdus the shamming was prolonged, and indeed a direct con- 

 nection was obvious between the strength of stimulus and the 

 length of period of (juiescence. This power of resjionse to 

 stimulus was maintained for two da)'s, and then weak fluttering 

 set in for some hours, followed by death. Our entire ignorance 

 of the jjhysiology of the nervous system of insects renders it 

 difficult to draw complete conclusions from these phenomena : 

 nevertheless, it is difticult to conceive that volition can persist for 

 fort)-eight hours in a tlecapitated animal. We are forced then 

 to conclude that here, at any rate, death-feigning is a purely reflex 

 phenomenon, and that the sensory stimulus received by the sur- 

 face of the body causes inhibitor)' impulses to arise reflexly from 

 the ganglia of the central nerve chain, and prevent all movement 

 of the locomotor inuscles. In confirmation of this, it may be 

 mentioned that denudingthe wing of its scales over any area caused 

 a marked diminution of sensiti\eness over the area so treated. 

 Since all stages between sensory^ hairs and ordinary scales occur 

 in Lepid<)ptera, it is not unreasonable to assume that the scales 

 still fiuiction as tactile eiul-organs, in spite of their modification 

 .subserving decorative purposes. Osw.vi.li H. I..\T1'F.R. 



Charterhouse, Godalming, July 31. 



Halley's Chart of Magnetic Declinations. 



In NaiurI'; for .May 23 and 30, 1895, are interesting com- 

 munications from Dr. Bauer and .Mr. Ward in reference to 

 Halley's old chart of magnetic declinations. 



I have a copy of this chart not referred to by either of these 

 gentlemen. 



It is bound in vol. i. of "Miscellanea Curiosa." This work 

 was eilited by Halley ; it consists of three volumes, containing, 

 in the main, rejjrints of papers read before the Royal Society. 

 \'ol. i. was published in 1705, and wa.s printed by J. B., for 

 jeftery Wale and John .Senex. 



The chart is 7A inches high and 13 inches long, and endiraces 

 just the circumference of the earth. 



The title in the u]>]ier left-hand corner reads : " A new and 

 correct Sea Chart of the Whole World, showing the Variations 

 of ye cf)m]')ass as the)' were iound .'\ii() 1700 with a view of the 

 ('■eneral and Coasting Trade Winds and Monsoons or shifting 

 Tr.ade Winds by the Direction of Capl. Kdm. Halley." 



In the lower left-hand corner is the note: " Capt. Halley's 

 map of the World in two large sheets is sold by R. Mount and T. 

 Page on Great Tower Hill, London." 



The name "I. Hairis, delin. iV scu.'' is in the lower right- 

 hand corner of the chart. CllARi.lis I,. Ci.arkic. 



New \'ork, July 27. 



THE ERUPTION OF VRSilliS, 



jri.y 3, 1895. 



'X' HIS recent distuiliance .it X'esitvitis is interesting in 

 ^ sc\cral ways, and at one time had all the appear- 

 ance or developing into as ^Tand a display as that of 

 187:;. 



The last eruptive cycle of \'esuvius commenced on 

 June 7, 1S91, when I had the good fortune to be but a 



NO. 1345, VOL. 52] 



few hundred yards distant at the time the main bursting 

 of the rift took place. The detads of that eruption, with 

 illustrations, can be referred to in my articles and reports.' 

 W'c may briefly state that cycle as follows : the splitting of 

 the whole of the great cone of X'esuvius by a radial rift 

 which extended beyond the base for some distance across 

 the Atrio del Cavallo. At the first moment a little lava 

 issued from the upper part of the rift, but after a {^\s 

 hours all came from its lowest extremity in the Atrio, and 

 continued to flow with practically no interruption for a 

 period of nearly three years, or, more correctly, from June 

 7, 1891, to February 7, 1894. During that period no 

 great quantity was given forth at any one time, so that 

 no stream could attain much length before cooling. 

 Though the amount etiiitted during that period is enor- 

 mous, and if vesicularised into pumice and scoria would, 

 I think, quite equal .Monte Nuovo in volume. The con- 

 sequence of this is, that a great and pure lava cone was 

 built up in the Atrio, of low inclination (14'), and adding 

 much to obliterate that interesting and characteristic 

 feature of the volcano. Coincident with the formation of 

 the rift, the central cone rapidly crumbled in, until a deep 

 crater was formed which eventually attained over 1 50 m. 



Fig. I. — Diagram sliowing the .ictu.-il sl.lle of \'esuviu\. from .t drawing Ijj- 

 M. A. liourdariat, after an earlier plan of mine (/-a i\'atiir(r, June 8, 

 1895). (a) Limit of the crater edge of 1S72: [he part represented by a 

 dotted line is that covered Ijy more recent lavas of different dates. The 

 parts a and a" are still uncovered. {/') Crater of June 1801. (^') Active 

 vent of the i8gi crater, {c) New cone in process of formation (May 

 1895). W) .Active vent of the cone on May 12. 1895. (rf) Rift and 

 vapour mouth of June 7, 1891, U) Fissure emitting acid vapours on 

 crater plain formed in the 1872 crater. (/) Very old hot-air passages 

 and fumaroles. {g) Fissure of May. 1889. _(//) Numerous fissures on the 

 south-east edge of the crater plain. (/") Guides' shelter. 



in depth and diameter. It was at its greatest dimensions 

 in February 1894. when the la\a stopped issuing by the 

 lateral outlet, and therefore commenced to rise in the 

 chimney. The immediate result of that stop|)agc was 

 that the formation of a cone was soon commencecl at the 

 bottom of the ciatcr by the ejection of lava cakes. The 

 growth of this new cone of eruption was so rapid that, when 

 I visited and photographed the interior of the 1891 crater 

 in Xovcmber last, this was not abov e 60 or 70 m. deep, 

 and the cone of eruption was rapidly increasing in height 

 w ithin it. 

 My friend M.Alex. Bouidariat has carefully observed the 



I.'Krtiption tlu 



1S91, 



upti 



1 " II Vesuvio." Corriert <ti Xa/'oli. June 

 V6suve," IS Itatte, Rome, June 13, 1891 ; /,(■ Figaro, Paris, June 17, 

 1891. "The Kruption of Vesuvius," Mcfittcfranean SaturaiUt, .Malta, 

 July I and .'\iigust i. 1891. " Lettre sur I'Kruption du Visuvc." L'ftatic. 

 Rome. July 18, 1891. " L'Kruption du VAsuve, visiles d'e.vplitration au 

 Volcan.'" La Xatmr, August 8, iSgl (itlastrated). " The Kruption of 

 Vesuvius," Natckk, vol. xliv. pp. 1^0-161, 320-322, _and 362 (illustrated). 

 " Report British As.sociation," 1891-92-93-94. '* L'Eriuipne del Vesuvio," 

 Kassfgna lielle Sticnzc Ccotoghtte^ vol. t. Rome, 1891 (illustrated). 



