12 



NATURE 



[March 2, 191 1 



searches. He and I are in agreement that the real origin 

 of our magnetic disturbances is to be referred primarily 

 to an outside electric system situated somewhere in our 

 own atmospheric regions. I go one step further, and place 

 in the same region the chief origin of the earth's own 

 magnetisation. 



In conclusion, it will be well to point out that the 

 method used in my Table II. above to get the average rate 

 of progression of sudden disturbances does not accentuate 

 the actual time differences between distant stations, but 

 tends rather to diminish them, as was the case with storms 

 Nos. 24 and 29. In brief, as I have already hinted, th<' 

 precise method of grouping of stations cannot be a fixed 

 one, but must vary with the region in which the disturb- 

 ance originated. When the data from the remaining parts 

 of the globe have been received, this matter will become 

 more evident. L. A. B. 



Washington, D.C., February 13. 



Colliery Warnings. 



May I say a word about colliery warnings to point out 

 that it is not the high barometer that is of any import- 

 ance, but the dryness of the air? It happens in our 

 country that the high barometer and dry air generally 

 come together. 



Gas explosions in coal mines are trivial, and they occur 

 on an average of more than one per day. It is when 

 there is sufficient dust to make the explosion spread over 

 a big area that an explosion 'S serious, almost without 

 exception. Dry air is the danger, and should be the basis 

 for "colliery warnings." John Harger. 



University Club, Liverpool, February 21. 



In all collieries more than 600 or 700 feet in depth the 

 air is always dry — somewhat drier in cold than in warm 

 weather — and, consequently, coal-dust is always present 

 in the workings provided there is no natural " seepage " 

 of water into them. In these circumstances, the one 

 essential element of a great explosion is always present. 



Blasting shot, when fired under certain conditions, and 

 comparatively small volumes of explosive gas when ignited, 

 will always raise and ignite coal-dust the quality of which, 

 as regards its contents in volatile matter and ash, lies 

 between certain upper and lower limits. On the other 

 hand, shots are always being fired, and larger or smaller 

 accumulations of explosive gas are always being formed 

 here and there in mines of this class, quite irrespective of 

 weather conditions. 



All that can be said as regards the influence of weather 

 is that, other things being equal, a coal-dust explosion is 

 more likely to occur in cold weather, when the mines are 

 driest, than in warm weather, when they are not so dry, 

 and with a falling rather than with a steady or rising 

 barometer. 



These subjects were fully discussed at the very inception 

 of the coal-dust question, and Mr. Harger might do well 

 to study what was then said about them. 



As the issuer of " Colliery Warnings " so frequently 

 advocates that special attention be paid to the condition of 

 mines when the barometer is rising rather than when it is 

 falling, I may perhaps be allowed, in this place, to 

 correct a statement which lately appeared in a letter to 

 Nature, written by "The Author of the Warnings," to 

 the effect that Mr. R. H. Scott and I were amazed (sic) 

 to find that fire-damp was frequently reported to have 

 been found in mines even when the barometer was steady 

 and rising. We were not amazed, for we knew by the 

 actual experience of one of us that, in consequence of falls 

 of roof, damage to trap-doors, stoppings, brattices, and 

 so on, which are amongst the commonest incidents in 

 mining, the ventilation often becomes so stagnant at 

 certain critical points that the air becomes explosive at or 

 near these points before the defects can be rectified. 



As a matter of fact, the principal province of the fire- 

 men is to guard against this very contingency. Conse- 

 quently, when we saw appearances of fire-damp reported 

 time after time with a steady or rising barometer, we 

 experienced no surprise, but, perhaps rather unfortunately, 



NO. 2157, VOL. 86] 



considering the use the " Author of the Warnings " 

 made of the sentence in which we merely recorded a fa^ 

 without further comment, we did not stop to explain what 

 passed in our own minds at the time, but proceeded in the 

 next following sentence to show that, when the records of 

 a large number of mines were compared, the effect of 

 these casual irregularities was practically eliminated 

 altogether. It would have been fairer if the " Author of 

 the Warnings " had quoted one more sentence (Quarterly 

 Journal of the Meteorological Society, October, 1874). 



W. Galloway. 



The Hydrogen Spectrum. 



The colour of the light which is emitted from the 

 capillary of a vacuum tube containing pure hydrogen is 

 the familiar vivid, deep pink, the prevailing tint being due 

 to the predominating brilliance of the red line (Ha). 

 When the same capillary is viewed from one end, how- 

 ever, the colour is a very pale pink, indeed, nearly white.- 



This interesting effect appears to be due to the different 

 relative intensities of the lines Ha and H/3 in the two 

 cases, for while H/3 (and probably each of the other lines 

 in the primary series) has an intensity appropriate to the 

 length of the column of gas in the capillary, Ha appears 

 to suffer reduction. 



In the course of some observations on the secondary 

 spectrum of hydrogen, during which I had occasion to use 

 the tube when placed against the slit of the spectroscope 

 in the usual way, and also end on, I was greatly interested 

 in this apparent variation in the intensities of the lines 

 mentioned. The explanation which suggests itself is that 

 Ho is weakened by absorption in traversing the column 

 of gas (although this distance is only about 5 or 6 cm. in 

 my tube), but it is difficult to understand why this absorp- 

 tion should be specially selective for Ha and not equally 

 effective in the case of Hj8, which is also a very brilliant 

 line, and to which one would imagine the same argument 

 would apply. There appears to be no reason for attribu- 

 ting the effect to polarisation. 



I have seen no notice of this effect, but I imagine it j 

 must be quite familiar to spectroscopists, and perhaps one 

 of them who has devoted special attention to the hydr<^en ij 

 spectrum may be able to throw some light on the matter. 



Charles W. Raffety. 



Beechcroft, 2 Park Hill Road, East Croydon, 

 Surrey, February 22. 



Life and Habit. 



On p. 505 of Nature for February 16, in a review of 

 a new edition of one of Samuel Butler's books, these 

 words appear- — "therefore the apparently unpractised but 

 perfect pecking of a newly-hatched chick proves that the 

 chick has done it before," &c. 



Now, I have tried many experiments with chicks hatched 

 out singly and away from a hen, but never has any chick 

 attempted to peck until shown how to or made to walk 

 over food which tickled its toes, and my opinion is that a 

 chick might die of starvation while surrounded by food 

 unless taught what to do. They are quick to take a hint, 

 and will imitate the motion and action at once if a bent 

 finger be worked up and down like the head of a bird 

 when pecking ; and, if once they feel a bit of food within 

 their beaks, they know what to do with it as well as a 

 new-born babe knows how to suck when anything is 

 placed in its mouth, though they have never done it 

 before. W. H. M. 



February 20. 



Your correspondent's remark is interesting, but, of 

 course — as he himself indicates in connection with the 

 babe — it does not invalidate Butler's argument. It can 

 hardly be doubted that observation and imitation do not 

 cover the ground, and that there is something calling for 

 explanation in what is called " instinct " — a word which, 

 it must be admitted, only disguises our ignorance of what 

 it is. Butler's theory that " heredity is memory " is at 

 least worth consideration. The Reviewer. 



