5'2 



NA TURE 



[March 29, 1906 



From observations which I have recently made it seems 

 •evident to me that the cause for seeing the motion is 

 entirely different. 



In the first place, you can always see the motion a 

 fraction of a second before you begin to feel it. In the 

 second place, you cannot see a perfectly horizontal motion 

 •or a gentle vertical (heaving) motion. In the third place, 

 watching a fixed point close to you, such as a pattern on 

 a carpet, when the ship is pitching and rolling, is far more 

 tiring to the eyesight than when the ship is motionless or 

 running perfectly steadily. All this points to the appearance 

 ■being due to a true relative motion of the eyes to the ship. 



The eyes are suspended in their muscular settings, much 

 in the same way as are ships' compasses in their binnacles. 

 The eyes are, furthermore, perfectly balanced, so as to 

 make their muscular displacements as little tiring as 

 possible. In their normal position, the pull of gravity is 

 ■exerted vertically through their centres, and the "muscular 

 mei nanism is compensated for gravity. 



Any angular change of position will displace the eyes 

 just as it displaces the stomach, excepting that the eyes, 

 being a great deal more sensitively suspended, will register 

 •the displacements more quickly. It is not, however, the 

 motion of the eyes which strains the eyesight, but the act 

 lit resisting this motion. 



If, with your eyes shut, you attempt to fix the mental 

 representation of a point, which a moment previously you 

 were watching with eyes wide open, you will find that, 

 after one or two motions of the ship, the bodily feeling 

 will precede any visual sensation which your imagination 

 can conjure up. The imaginary point is no longer fixed, 

 but follows the eyes as they let themselves go to the motions 

 of the ship. No strain of the eyesight is caused by a 

 muscular resistance, and the displacements, while felt, can 

 no longer be seen. Alfred Sang. 



Pittsburg, U.S.A., February 26. 



Production of an Electrically Conductive Glass. 



Experiments have from time to time been made, both 

 in England and abroad, to ascertain what ingredients are 

 best for the purpose of producing glasses of very high 

 electrical resistance. 



The utility of a vitreous substance which would conduct 

 electricity comparatively well does not appear, however, to 

 have so far claimed any consideration. 



I beg therefore to direct attention to a glass which has 

 recently been made in my laboratory. Its chief feature is 

 that it readily conducts electricity. 



For the windows or cases of electroscopes and all high- 

 tension apparatus requiring a transparent cover capable 

 "I screening off external electrical fields, this material offers 

 many advantages. A conducting varnish is no longer re- 

 quired for glass which conducts electricity itself. In addi- 

 tion to these practical considerations, there arises the 

 interesting question as to the process by which electricity 

 passes through this substance— whether it is electrolytic. 

 Its resistance varies very markedly with temperature 

 changes. I hope later to give more precise details. The 

 'basis of the glass is sodium silicate. 



Charles E. S. Phillips. 



Shooters Hill, Kent, March 12. 



Interpretation of Meteorological Records. 

 Is discussing the records of the meteorological instru- 

 ments at Canterbury (Nature, March 15), Dr. Aitken 

 suggests that the heavy rain which fell dragged down the 

 higher air, and so caused the fall of 12 indicated on the 

 thermograph curve, and he very clearly and convincingly 

 shows the consequent effect on the barometric pressure and 

 wind velocity. If. however, the air had been in a state of 

 Stable equilibrium previous to the thunderstorm, the effect 

 ■ ■I such a mechanical dragging down of the higher air 

 would have been to heat by compression that air so 

 much that the temperature would have been raised rather 

 than lowered at the ground-level. But if, previous to the 

 in in, the upper air had from any cause become very much 

 colder than the lower air, the atmosphere would be in a 

 sta 1 unstable equilibrium, that is to say, the rate of 



^O- 1900, VOL. 7$] 



change of temperature with height would be greater than 

 the adiabatic rate of change due to heating by compression 

 of descending air. In such a case the changes recorded by 

 the various curves may have been initiated by this heavy 

 cold air suddenly descending and displacing the lower air, 

 which by its sudden uprising would be cooled, the moisture 

 in it condensed, and a heavy fall of rain caused. 



The lightning which accompanied this storm introduces 

 an element of uncertainty into any attempted explanation, 

 for we do not know yet the manner in which electric 

 charges are generated in the atmosphere. But it seems 

 probable that a great cooling of the higher air is an 

 accompaniment of a state of electric tension, for it is 

 difficult to see otherwise why a thunderstorm should be 

 followed by a lowering of the temperature near the ground- 

 level. R. T. OMOND. 



Edinburgh. 



Oscillation of Flame Cones. 



I should be glad if any of your readers could give an 

 explanation of the cause of the following flame pheno- 

 menon, produced while experimenting with a modification 

 of Prof. Smithells's apparatus for the separation of the 

 cones of a Bunsen flame. 



A mixture of gas and air is burned at the top of a 

 vertical tube (made preferably of combustion tubing) about 

 4 feet long and J inch to 1 inch in diameter, having a 

 delicate screw adjustment for regulating the proportions of 

 gas and air. 



The air supply is carefully and slowly increased, until 

 an almost explosive mixture is reached, and the inner cone 

 is very short and sharp and of a light green colour. On 

 admitting a very slight increase of air after this point, the 

 innei cone (sometimes the two cones) descends the tube 

 to a distance of about 2 feet, and then pauses and goes 

 up again, re-joining the outer cone. The flame then 



sharpens " again and repeats the process, and will con- 

 tinue to do so for several hours without further adjustment 

 of the gas or air being made. 



There is every appearance of an explosion wave being 

 propagated, as shown by the increasing velocity of the 

 descending flame and by the occasional emission of a note 

 as it reaches the end of the travel. 



'I he length of travel can be regulated by the amount of 

 air admitted, varying from 1 or 2 inches to about 2 feet 

 in the same tube. If it be allowed to exceed a certain 

 limit the inner cone is extinguished at its lowest point, but 

 immediately re-lights at the top of the tube, and then re- 

 turns as before. The periodicity can be varied from about 

 once in five seconds to once per second. 



The gas pressure does not need any special regulation, 

 llie ordinary variations from a town supply not affecting 

 the results. 



The following are the points requiring explanation : — 



(1) As the proportions of gas and air are constant, what 

 is the cause of the periodic " sharpening " of the cones 

 after meeting at the top of the tube? 



(21 What prevents the explosion wave being completed, 

 and the consequent firing back of the mixture? 



(;,) What causes the inner cone to return and travel up 

 the tube, re-joining the outer one at the top? 



(4) The alteration in the character of the flame (in view 

 of the fact that the proportions and pressure of gas and 

 air are constant) points to some form of wave motion 

 bringing the molecules into closer contact. If this be so, 

 what are the conditions which set up this wave motion 

 and what determines its periodicity? 



Harold E. Temple. 



Olton, Warwickshire. 



The phenomenon described in the foregoing letter is in 

 part dealt with in a paper by Dr. Ingle and myself in the 

 Transactions of the Chemical Society for 1S92 (vol. lxi., 

 p. 204). The continued oscillation of the inner cone is, I 

 think, explained by the fact that the mixture of gas and 

 air in the tube is not uniform. We have, indeed, found it 

 necessary to use elaborate mixing appliances to make it 

 uniform. When a portion of the mixture rich in air reaches 



