• KNOWLEDGE ♦ 



[Nov. 17, 1882. 



From experiments, he was of opinion that gas was not 

 dangerous so long as it occupies less than r> pvr cent of a 

 given space ; between 5 and 1 2 per cent it begins to be 

 ^jniUble. and at 1'9 per cent the danger of explosion is 

 present in high degree. He has improved Anselm's 

 apparatus (wliich, acting on the principle of diffusion, 

 causes an electric Wll to ring as soon as a certain quantity 

 of gas is prest nt), so that even }, per cent of escaped gas 

 causes loud ringing. Another ingenious arrangement 

 ('hows the actual percentage quantity of escaped gas, while 

 a control-manometer, which Professor Than has used 

 several years in his laboratory, reveals the place of escape 

 of the gas. 



Tub liailrimd (•'<iz-ll'- record of American train accidents 

 for Auijust shows for tliat month a total of 139 accidents, 

 by which 4G persons were killed and 218 injured. There 

 were 6r> collisions, in which 27 persons were killed and 117 

 injured; 70 derailments, with 18 persons killed and 100 

 injured, and four other accidents, in which one person was 

 killed and one injured. Twenty-six accidents caused the 

 deatli of one or more persons each ; 2S caused injury, but 

 not death: while in 8.'>, or 61-2 per cent of the whole 

 number, no serious injury to persons is recorded. These 

 accidents may be classed as to their nature and causes as 

 follows: — Collisions: Rear collisions, 41; butting colli- 

 sions, 20 : crossing collisions, 4 ; total, 6.5. Derailments : 

 Broken rail, 2 ; broken switch road, 1 ; broken bridge, 4 ; 

 spreading of rails, 7 ; broken axle, 4 ; broken wheel, 2 ; 

 broken truck, 1 ; wash-out, ■') ; accidental obstruction, 4 ; 

 cattle on track, 8 : misplaced switch, 7 ; neglect to use 

 signals, 1 ; purposely misplaced switch, 2 ; malicious ob- 

 struction, 1 : unexplained, 21 ; total, 70. Boiler explosion, 

 1 ; broken connecting-rod, 1 ; broken crank-pin, 1 ; broken 

 axle, not causing derailment, 1 ; total, 139. Derailments 

 of the long American cars are thus the most frequent form 

 of accident 



A SEW use has been discovered for potatoes. They can 

 lie convert<!d into a substance resembling celluloid by 

 peeling them, and after soaking in water impregnated 

 with eight parts of sulphuric acid, drying and pressing 

 between sheets of blotting paper. In France pipes are 

 made of this substance, scarcely distinguishable from nieer- 

 achaum. By subjecting the mass to great pressure, billiard 

 balls can be made of it rivalling ivory in hardness. 



We pro[X)80 in the next two weeks to give a very much 

 simpler account of Venus's transits than in the com- 

 paratively difficult papers now nearly completed. 



Sevekal physicists, chemists, spectroscopists, and others 

 who chance to l<e not also mathematicians or astronomers, 

 h»Te rwently advanced, advocated, or adopted the theory 

 of an interst<-llar and interplanetary atmosphere, to which 

 — being fn-e from the restraints imposed by a knowledge 

 of astmriomical and mathematical laws — they assign 

 various d<-gree« of rarity, according to the requirements of 

 their own or their friends' pet theories. Astronomers have 

 looked on with some amust^ment at the innocence with 

 which enormous volumes of carlionic acid gas, the vapour 

 of watiT, nitrogen, oxygen, hydrogen, and so forth, have 

 b»*n add>-<l to the int<Tstellar universe, in calm uncon- 

 •ciouine>.s that if science is to accept these twlditions, all 

 that Galibo, Kepler, and Newton have taught (and wlii.li 

 we had xupposed provwl) must be abandoned. While 

 enjoying the joke, astronomers have not \x:(tn very anxious 

 to advance the overwhelming arguments which have always 



been available. But as these invaders of the domain of 

 the exact science have lately seemed to mistake this reti- 

 cence for acquiescence, it seems desirable to show at least 

 the .i/iiipe of the rock-masses which overhang the preci- 

 pitous and trackless region over which they Imve been for 

 some time cheerfully adventuring. For this reason wo give 

 this week M. Faye's miisterly though short demolition of 

 these theories, which we would recommend in a special 

 manner to the attention of Dr. Siemens, Mr. W. M. 

 Williams, the Editor of Xahnr. and others, who appear 

 to regard mathematics as the bano of science, because it 

 kills wild theories. 



Ek;iit full columns of Mr. Jlurr.ay's publications ap- 

 peared iu the .Siumiard of the 14tli inst 



A.MONO the many unexpected developments of electrical 

 science is an application in the hiving of bees when they 

 swarm, successfully tried by German experimenters. It 

 was thought that by utilising the electric force the bees 

 might be stupefied for the necessary period of time without 

 being injured, and the result proved the correctness of the 

 idea. The first attempt was made upon bees that had 

 gathered upon trees, the insects falling upon the ground in 

 a kind of trance, which admitted of their being safely 

 handled. The next stage in the experiment was to capture 

 the bees when they were about to swarm. By introducing 

 the ends of two connecting wires into a fully-occupied 

 honeycomb and turning on the current, the bees were 

 rendered inactive for about thirty minutes, while no bad 

 results appeared to follow their awakening. 



FitOM Sydney is reported the discovery of a new building 

 niateri.al at Suva, on one of the Fiji Islands. It is known 

 as fossil coral, and when cut from a mass is soft, but on 

 exposure to the air it looks and is very much like brick. 

 During the short time it has been known it has satisfac- 

 torily stood the tests that have been applied to it. Orders 

 have already been received for quantities of it to be used 

 in building. 



It is a remarkable fact that there are no rats in the 

 islands of the Pacific Ocean. Repeated attempts have 

 been made to acclimatise the rodents there, as the flesh is 

 much esteemed by the natives as an article of food. But 

 the attempts thus far have failed, as they [th<! rats] invari 

 ably die of consumption. 



In the hour columns on p. .373 relating to the comet's 

 path, there are some obvious mistakes, relating, however, 

 only to the part of the motion still to be observed. Thus 

 in the first table the hourn on No\. 2.'), 29, and Dec. 2, 

 should bo 2, not 3, while those in the second table, for 

 Nov. 21, 2.0, 29, and Dec. 2, should be 4, not .3. Circum- 

 stances caused our time for correcting those colunms to 

 be very short indeed. 



Nkxt week wc shall give some pictures of the comet, 

 and our promised discussion of th'i probable effect of its 

 destruction by the sun. 



With No. 01 the Exchange and Sale Advertisement 

 Columns will be discontinued, as we find wc frequently 

 have to defer them for want of space. 



