2 12 



NATURE 



[December 27, 1894 



carbon monoxide will do exactly the s.im: thing when the air 

 is laden wi'.h coal-dust, whilst the temperature of ignition is 

 slightly lower than with methane, so that in the cise of the 

 Camerton Colliery, it being perfecily well ascertained that the 

 air was charged with coal-du-t, the probabilities are that not 

 10 feet, but a lar larger volume of explosive mixture was formed 

 by the rapid escape of the products of c imbusiiin into the ooal- 

 laden air ; and this Ixring ignited, ei'her by the tlame or red hot 

 solid products driven out into it by the blown-out shot, would 

 initiate a considerable area of expIo>ion. 



The classical researches of Prof. H. Dixon have shown that 

 hydrocarbons and, probably, carbon burn in air lo carbon 

 monoxid>-, and that this carbon monoxide will not form explo- 

 sive mixtures with air, or even with oxygen, if they are abso- 

 lutely dry ; but if water vapour is presen', they explode owing 

 to the oxidation of the carbon uionoxide to dioxide, causing the 

 propagation of an explosive wa%e, which reaches its maximum 

 velocity when the percentage of water vapour, between $ and 

 6 per cent., and inasmuch as the air of the mines would always 

 contain some moisture, and as the products of combustion also 

 would give a large volume of water vapour, these requirements 

 would be amply fulfilled. 



Still more conclusive on this point were Mr. Hall's experi- 

 ments. In these a charge of blasting powder was fired 

 from a cannon suspended in a shaft, the air oi which was proved 

 by careful chemical analy^is to be absolutely free from any trace 

 of cou^busiible gas. 



In order lo get some idea of the condition of the air inside 

 the pit during the explosion, samples of air were l.aken and 

 were analysed. Two brass tubes were fastened to the rope that 

 was used to lower the cannon, one twenty yards from the 

 bottom, the other forty yards from the bottom. 



These tubes were so arranged and constructed that the explo- 

 sion, as it passed the tubes, unsealed the outlet pipe, and the 

 escaping water sucked in a sample of air which was (rapped by 

 a special arrangement, and kept in the tu ^e until the rope could 

 be wound up. By this method it was intended that the sample 

 of gas taken should represent that state of the air whilst the 

 flame was passing, or directly afterwards. 



The tube nearest the bottom, as the analysis will show, did 

 partly collect the gas in the above condition. The tube at the 

 top, however, commenced to act premature y, an I was probably 

 started by the sound wave which preceded the explosion. This 

 tube simply contained ordinary air. 



The following is an analysis of the gases found in the lowest 

 tube : — 



Per csnt. 



Oxygen ... 3 9 



Nitrogen ... 75-9 



Carbon dioxide ... ... ... iz'i 



Carbon monoxide ii'l 



This ingenious arrangement was due to Mr. W.J. Orsman' 

 and it isprobablyihefirst successful attempt which has been mad 

 10 get a sample of gas during the pro rcss of explosion, an 

 there is not the slightest doubt that the presence of such an 

 amount of carbon monoxide converts mixtures of coal dust and 

 air into a highly explosive body. 



As ihe explosion takes place, and as the carbon monoxide 

 ready produced is oxidised to carbon dioxide by the action 

 upon it of water vapour present, a.rd also by its direct com- 

 bustion with oxygen, the hydrogen of the water vapour is set 

 frte, whilst ihe heated coal-dust also yields certain inflammable 

 products of disiillaiion to the air, and partial combustion also 

 of ihe coal du-t gives a considerable proporii n of carbon 

 monoxide once more, and these driven r.ipidly ahead of the 

 explosion form with more coal-du'.t and air a new explosive 

 zone, and so by waves and throbs the explosion is carried 

 through the dust-laden galleric-> of the min-r. 



The experiments made by Mr. Hall, and invcsiigitions in 

 various colliery explosions, mikc It ariundanily m.inilcit that no 

 explirtive should be licensed for use in mines unless it can be 

 ab-olutrly proved that it give- off no imflamm iiile products of 

 combu lion. The following table will show ihe results given by 

 some of ihe explosives most largely ucd, which point very 

 clearly to the fact thai, wiih therxccp'ion of the Sprengcl ex- 

 plosives, such as roburlle and nitr.iKlyccrinc, none of the 

 bodies in use conform to this iniporiani rer| Jireln^:nt. 



Products of Combustion of Blasting Explosives. 



Combustibles. 



Powder. 



Carbon dioxide. Crirbon monoxide. 



Gunpowder ... 506 



Blasting power ... 321 

 Sprengel explosives — 



Roburite ... ... 320 



Ammonite ... 330 



Nitroglycerine explosives— 



NitrogUcerine 



Gelinnite ... 



Carbonite ... 



Blasting gelatine 



63 o 

 25 'O 

 190 

 365 



105 



33 '7 



nil 



nil 



nil 



7-0 



"SO 



Hydrosen and 

 marsh gas. 



3'i 

 7'9 



nil 

 nil 



nil 



nil 



26 'o 



S-6 



Whilst not only these considerations, but Mr. Hall's experi- 

 ments, point to the absolute necessity of legislative enactments 

 at once forbidding the use of blasting powder in any coal mines, 

 no matter how free they may appear to be from fire-damp or 

 from dust, if the returns made as to deaths ciu^ed by gun- 

 powder and o'her explosives in mines for the year 1S93 are 

 examined, it will be clearly seen that theexcbision of gunpowder, 

 in handling alone, would do a'vay with So per cent, of the 

 accidents. So that if explosives of iti-f Sprengel class were em- 

 ployed, accidents due to the explosives used would be prac- 

 tically eliminated from the mining death loil ; and it is only a 

 question of time as to when England will icllo* the action of 

 France and Germany in altogether prohibiting the use of blast- 

 ing power in dusty mines. 



THE POSSrB/LlTIES OF LONG-RANGE 

 WEATHER FORECASTS.'^ 



T F we had perfect command of this subject, we should be able 

 to trace the motion of a panicle of aqne us vapour from 

 point to point over the whole earth, and could predict whether 

 at any lime in the future it will fall a-s rain, or rise and lly away 

 as an invisible gas. In the absence of this higher knowledge 

 the only long-range forecasts that we are at present able to 

 make are based upon empirical and very imperfect rules deduced 

 from our study of the accumulated clim.ilological sl.itistics. Of 

 course, sech predictions do not imply any special knowledge of 

 meteorology. Among the methods adopted in long-range 

 forecasts are the following : 



(a) The average rainfall, temperature, &c., for any period, 

 such as a month, and deduced from many yeais of observation, 

 is called the normal. The excess or deficiency of this month in 

 any given year is called llic departure for that year. A general 

 prediction may be made to the elTecl that the raii'fall for a given 

 month and place may be expected to lie within the range of the 

 values indicated by these known departures. 



(/') The scries of annual or monthly values just mentioned 

 gives us the means of finding out whether iheie is any simple 

 sequence or connection between them and (he .ipparently un- 

 connected values that occur (roni year to year. Thus, it some- 

 times happens that rainy seasons come alter two or three dry 

 seasons, or ihat after the same month has been dry in three 

 successive years, one is then justified in pr< dieting a wet month. 

 Thus, Governor Raw~on elaborated a sysiem lor ihe prediction 

 of rain and the sugar crop in liarbados. 



{<■) Slight but appreciable widespread, rather regular fluc- 

 tuations of temperature, pre-sure, and rain have been revealed 

 in the climate ol Europe by Dr. Biiickni-r, who linds that a de- 

 ficient temperature and an excess of r.iin have alternated with 

 excess of temperature and deficiency of rain in periods of thirty- 

 six or thirty-seven years during the past two or tlnee centuries ; 

 the glaciers increase and dimmish m volum- , ur advance and 

 retreat, in correspondingly regular but somewhat retarded 

 intervals. Predictions may be based on these well-establisb.ed 

 periods. 



((0 Droughts are sometimes due lo what happens in distant 

 regions: thus, if there is a heavy sno« on the llii'alnyas during 

 the winter, there is a special liability to diought in lower India 

 in the follo«ing summer, so that the putliciionof a drought 

 may be basid i.pon the reports of snow tall in a rlisiant region 

 several months beloie the drought occurs ; but • ther droughts 

 may occur wiihout this preliminary sno«-lall. This connection 



' Keprini of .in pnicl? contributed by Prof. Cleveland Abbe to the 

 U.S. l^toHthty tt'titllicr Kevitw. 



NO. 1313, VOL. 51] 



