620 



NATURE 



[February b, 1913 



all the newly broken surfaces of glass gave the black 

 condensation, the next question which called tor con- 

 sideration was : What is the condition of a freshly 

 broken surface of glass? While what was now a 

 new surface was still in the interior of the solid it 

 would be in equilibrium with its surroundings; but 

 would it be in equilibrium with its new surroundings? 

 That is : Is a newly broken surface of glass in the 

 same condition as glass as we know it? To get an 

 answer to this question a number of samples of 

 different kinds of glass were broken and placed on 

 a sheet of glass and covered with a glass shade. All 

 the samples were placed with their test surfaces 

 facing downwards to prevent any dust settling on 

 them. When these were examined two days later it 

 was found they had entirely changed ; almost no black 

 or iridescent colours could now be seen ; nearly all 

 the surfaces gave the white deposit. 



It is not necessary to wait so long as two days 

 to find this change in the broken surfaces of glass. 

 If broken one day and examined the next the change 

 is quite evident, and so surprising did this seem that 

 I had to make a fresh break of the same piece to 

 compare the new with the old before I was satisfied 

 that the observation was correct. From this it will 

 be seen that it is possible to say whether or not a 

 piece of glass has been newly broken. The time 

 necessary for this change to take place has not been 

 determined, but probably depends upon the kind of 

 glass used ; possibly also the temperature and humidity 

 of the air may have some erfect. Wetting the surface 

 causes the change to take place more quickly. 



From the foregoing it would appear that the tests 

 made with newly broken surfaces of glass give an 

 uncertain answer as to the nature of the deposit on 

 clean glass. If we look on a newly broken surface as 

 clean glass, then the deposit is black. But if we wait 

 until this surface has acquired its equilibrium in its 

 new surroundings — that is, in the condition in which 

 glass as we know it always is — then the deposit is 

 white, and so far as these tests go the verdict tends 

 to the latter conclusion. 



Coming now to the experiments made with glass 

 tubes, in which the influence of the impurities of the 

 flame is eliminated, I found no difficulty in repeat- 

 ing Lord Rayleigh's experiments. In these tests I 

 used a number of different kinds of glass, and all of 

 them, after heating, gave the black deposit, but they 

 varied greatly ; some gave black but no colours, 

 whilst others gave good colours. The glass used for 

 ordinary test-tubes gave the best results. Something 

 of this may possibly be due to the composition of the 

 glass. K\\ easily fusible glasses are more soluble in 

 water than the others, and we may presume have a 

 greater affinity for it. In making these experiments 

 it was noticed that little or no effect was produced 

 unless the glass was heated high enough to cause the 

 bunsen flame to be coloured with the sodium of the 

 glass, showing that the glass was undergoing some 

 change. 



The test with the tubes having shown that after 

 beino- heated high enough they gave the black de- 

 posit, sheet-glass was again put under investigation 

 to see if higher temperatures would not make it also 

 give the black deposit. In the previous tests it was 

 not found possible to heat the glass verv highly ; owing 

 to their size and to the want of uniformitv in the 

 heating, the plates generally burst in pieces before 

 a high temperature was attained. Thin strips of gflass 

 of from J to J of an inch broad were now used. These 

 were placed on a piece of sheet-iron, about i inch 

 broad, and placed over a bunsen flame. With the 

 higher temperature now obtained it was found pos- 

 sible to change the glass and cause it to give the black 

 NO. 2258, VOL. 90] 



deposit, but verv little colour. As there might be some 

 hesitation in accepting this result, the method of 

 heatmg was changed. The strip of glass was put 

 inside a tube, and was thus thoroughly protected 

 from contact with the bunsen gases. Strips of glass 

 were heated in iron, brass, and glass tubes, which 

 had been previously highly heated to cleanse them. 

 Under these conditions the highly heated sheet-glass 

 gave the black deposit, but very little colour. 



What interpretation are we to put on these tests.' 

 One might at first sight say that the heat had cleansed 

 the surfaces of the tubes and sheet-glass, and that 

 purified surfaces gave the black deposit. The observa- 

 tion that the glass had to be heated to a temperature 

 high enough to cause a change to take place in it 

 first raised mv suspicions as to the correctness of the 

 above conclusion, so an examination of the surface of 

 the glass was made. Using in the first place the strips 

 of sheet-glass, as observations are more easily made on 

 them than on the less optically perfect glass of the 

 tubes, the first question was : Is the glass clean after 

 being highly heated? It proved to be far from it. 

 If a small piece of wood the thickness of a match be 

 covered with a piece of the polishing cloth and a small 

 surface of the glass be rubbed with it, it will be seen 

 that along th.' boundary of the rubbed part there is a 

 deposit of dust, rubbed off the plate. This deposit is 

 not seen when a clean part of the plate is rubbed in 

 the same manner ; further, with proper lighting and 

 a black background it will be seen that the rubbed 

 part is blacker than the surrounding glass. This 

 observation is much aided if a strong magnifying 

 glass be used. .Attention was now directed to the 

 condition of the surfaces in the tubes. Rubbing with 

 a small pad showed that their surfaces also were 

 covered with fine dust. It may also be mentioned that 

 plates that had previously been passed over a bunsen 

 flame or a blowpipe jet could be seen by the same test 

 to be covered with dust. 



The question may be asked : Where does this fine 

 dust come from in the tube experiments? The 

 simplest explanation seems to be that it is due to 

 some change produced in the glass by the high tem- 

 perature. The effect of the heat, therefore, does not 

 seem to be entirelv a cleansing one. Of course, it is 

 quite possible that the dust may not be the only cause 

 of the black deposit ; the surface of the glass may 

 have been clianged in some way by the heat, causing 

 it to form the black deposit. This supposition 

 does not seem to be at all unlikely, as the surface has 

 in all probability been changed by the heat, ^nd if it 

 has been changed it is not likely to be now in a state 

 of equilibrium. Plates and tubes which had been 

 heated were therefore put aside and protected from 

 dust to see if they changed with time like the broken 

 glass. All of them changed more or less towards 

 the white deposit. Sheet-glass was back in two or 

 three days to its usual white deposit. The tubes were 

 slower in- changing. 



In conclusion, I wish to reply to the two last para- 

 graphs in Lord Rayleigh's article, wherein he de- 

 scribes some experiments in which his experience 

 differs from mine. In the first he says that with an 

 alcohol flame and blowpipe he efot the black deposit, 

 while in my tests I only got slight indications. The 

 reason for this difference is very simple. In 

 the previous part of mv paper describing this experi- 

 ment there are noted the effects of the bunsen 

 flame, then those of the alcohol flame ; the effects of 

 the two flames are next compared, and, as stated, the 

 alcohol gave verv slight results. Lord Rayleigh in 

 his tests used a blowpipe with his alcohol flame, and 

 naturally got a much greater effect. 



In the last paragraph Lord Rayleigh says our 



