422 



NATURE 



{March 3, 1881 



In the table is also given the measurements of the repulsion 

 exerteu on the blackened end of the mica plate by a candle-flame 

 placed 500 millims. off. The repulsion due to radiation com- 

 mences just at about the same degree of exhaustion where the 

 viscosity begins to decline rapidly, and it principally comes in at 

 the exhaustions above looo M. 



The close agreement between the loss of viscosity and the in- 

 creased action of radiation is very striking up to the 35 millionth, 

 wiien the repulsion curve turns round and falls away as rapidly 

 as the viscosity. 



Experiments are next described on the resistance of air to the 

 passage of an induction spark. 



Since the publication of the author's researches on the pheno- 

 mena presented by the passage of the induction discharge through 

 high vacua, the present results — which, although never published, 

 precede by a year or two those just mentioned — have lost much 

 of their interest. 



The phenomena at the very high exhaustion of o"02 M may be 

 of interest. With a coil giving a spark 85 millims. long, no 

 discharge whatever passes. On increasing the battery power 

 till the striking distance in air was 100 millims. the spark 

 occasionally parsed through as an intermittent flash, bringing 

 out faint green phosphorescence on the glass round the end of 

 the — pole. 



On one occasion the author obtrined a much higher exhaustion 

 than o'02 M. It could not be measured, but from the repulsion 

 by radiation and the low log dec. it was probably about O'oi M. 

 The terminals of the vacuum tube and wires leading to them 

 were well insulated, and the full power of a coil giving a 20-inch 

 spark was put on to it. At first nothing was to be seen. Then 

 a brilliant green Ight flashed through the tube, getting more and 

 more frequent. Suddenly a spark passed from a wire to the 

 glass tube, and pierced it, terminating the experiment. 



Since these experiments vacua have frequently been got as 

 high, and even higher, but the author has never seen one that 

 would long resist a 20-inch spark from his large coil. 



Viscosity of Oxygen. — The series of experiments with air show 

 a complete history of its behaviour between very wide limits of 

 pressure. It became interesting to see how the tw-o components 

 of air, oxygen and nitrogen, would behave under similar 

 circumstances. Experiments were therefore instituted exactly 

 as in the case of dry air, but with the appai-atus filled with piu'e 

 o.xygen. 



The results are given iu the form of tables and plotted as 

 ciu-ves on diagrams. 



The figures show a great similarity to the air curve. Like it 

 the log dec. sinks somewhat rapidly between pressures from 760 

 millims. to about 75 millims. It then remains almo>t steady, 

 not varying much till a pressure of 16 millims. is reached. Here 

 however it turns in the opposite direction, and increases up to 

 I'S millim. It then diminishes again, and at higher exhaustions 

 it rapidly sinks. This increase of viscosity at pressures of a few 

 millimetres has been observed in other gases, but only to so small 

 an extent as to be scarcely beyond the limits of experimental 

 error. In the case of oxygen hov\ ever the increase is too great 

 to be entirely attributable to this cause. 



Oxygen has more viscosity than any gas yet examined. The 

 viscosity of air at 760 millims, being o'Il24, the proportion 

 between that of air and o.\ygen, according to these results, is 

 II1S5. 



1 his proportion of I'll85 holds good (allowing for experi- 

 mental errors) up to a pressure of about 20 millims. Between 

 that point and 1 millim. variations occur, which have not been 

 traced to any assignable cause : they seem large to be put 

 down to "experimental errors." The discrepancies disappear 

 again at an exhaustion of about 1 millim., and from that point 

 to the highest hitherto reached the proportion of l'llS5 is fairly 

 well maintained. 



Viscosity of Nitrogen. — The proportion between the viscosities 

 of nitrogen and air at a pressure of 760 millims is, according to 

 these experiments, o'97i5. 



A comparison of the air curves with those given by oxygen and 

 nitrogen gives some interesting results. The composition of the 

 atmosphere is, by bulk. 



Oxygen 20-8 



Nitrogen...' 79-2 



The viscosity of the two gases is almost exactly in the same 

 proportion : thus at 760 millims — 



20'8 vis. O -H 79'2 vis. N 



—^-2 = vis. air, 



100 



20-8 (0-1257) -f 79"2 (0-1092) _ 



100 " 



2-61456 -fS-64072 _^.^,, 



100 ^^' 



a result closely coinciding with 0-1124, the experimental result 

 for air. Up to an exhaustion of about 30 M the same proportion 

 between the vi-cosities of air, oxygen, and nitrogen is preserved 

 with but little variation. From that point divergence occurs 

 between the individual curves of the three gases. 



Ol'sei-jations on the Spectrum of Nitrogen. — Spectrum observa- 

 tions during exhaustion give the following results : — 



At 55 millims. pressure the band spectrum of nitrogen com- 

 mences to be visible. The red and yellow bands are easily seen, 

 and the green and blue are exceedingly faint. As the pressure 

 grows less the bands become more distinct, until at I '14 millim. 

 the band spectrum is at its brightest. 



At a little higher exhaustion a change comes over the spectrum, 

 and traces of the line spectrum are observed. 



At 812 M both the band and the line spectrum can be seen 

 very brilliantly. 



At 450 M the line spectrum is seen in great purity. As the 

 exhaustion becomes higher the lines commence to disappear at" 

 the two ends of the spectrum. 



At 18S M the lines below A 610 ms. of m.m. at the red end, 

 and above A 400, cease to be visible. 



At 94 M a bright greenish yellow line is visible at^about 

 A 567. 



At 55 M this greenish yellow line is still very prominent. The 

 red lines have di-appeared altogether, and the highest blue line 

 visible is one at \ 419. The line 567 varies much in visibility ; 

 sometimes it cannot be seen, whilst at others it is very visible. 

 Thus — 



At 40 M the line 567 has quite disappeared. 



At 17 M line 567 is visible again, being the most prominent 

 line left. 



At 12 M line 567 is not seen, although several other green and 

 blue lines are left. 



At 3 M only three lines are visible in the green, and these are 

 very faint. 



At 2-8 M line 567 is detected again. 



At 2 M only traces of one or two lines can be seen, the faint 

 light of the lines being overpowered by the green phosphores- 

 cence of the glass. 



Line 567 has been seen on several occasions at high exhaus- 

 tions when the gas under examination has been mixed with a little 

 air. It is probably a nitrogen line, for one of the most brilliant 

 nitrogen lines has a wave-length of 567-8 (Thalen), 568-0 

 (Iluggins), or 56S-1 (Pliicker), and the author's interpolation 

 curve is not sufficiently accurate to enable him to say that the 

 line entered in as being at 567 may not in reality be a trifle 

 higher. The reason of its being only sometimes visible may 

 be accounted for by a difference in the sensitiveness of the eye 

 at different times, or by a difference in ibattery power. This 

 however cannot be the whole explanation, for other lines are not 

 found to vary in the same manner. 



The curve of Repulsion exerted by Radiation is much lower 

 than in oxygen or air, and sinks rapidly after the maximum is 

 passed. 



/ 'iscosity of Carbonic Anhydride. — The curves of this gas are 

 given in diagrams plotted from the observations. At first the 

 curve seems to follow the same du-ection as the air curve. But 

 at a pressure of about 620 millims. it slopes more rapidly till the 

 pressure is reduced to about 50 millims., when the curve again 

 takes the direction of the air curve. The total diminution 

 between 760 millims. and 1 millim. is nearly double that of air. 



Observations have also been taken with the spectroscope 

 during the exliaustion of carbonic anhydride. The nnximum 

 brilliancy of the spectrum occurs at an exhaustion of about 

 300 M. After that it ge's fainter ; at al out 75 M the blue band 

 (A 409 to 408 ms. of mm.) disappears ; as the exhaustion gets 

 higher the other bands vanish until, at a vacuum of about 40 M, 

 nothing is visible but the two lines A 519 and A 560. At higher 

 exhaustions these lines disappear, and the phenomena of 

 " Radiant Matter' commence. 



The proportion betw een the viscosity of carbonic anhydride 

 and an- at 760 millims. is 09208. 



