360 Prof. J. Dewar .
The Specific Volumes of Oxygen those of a Cygni was pointed out .
Later , Mr. McClean* confirmed this , and further stated that the bright lines of g Argus corresponded with the dark lines of the Nova ; while later still , Sir David Grillf has shown the similarity between the spectra of g Argus and Nova Aurigae .
Hence we are led to the conclusion that the temperatures reached in the outbursts of both these Novae were not greatly different from those of the stars named , and that in y Argus there are constant conditions which are similar to those temporary conditions which produce the appearance of Novae .
In the table the lines of Nova Persei are compared with those recorded by Campbell J and VogelS in Nova Aurigae .
The probably corresponding lines of a Cygni , chromospheric lines , and enhanced lines , are also given in separate columns .
The reduction to wave-lengths and discussion of the lines in relation to those of other celestial and terrestrial spectra , has been undertaken by Mr. F. E. Baxandall , to whom my best thanks are due .
" The Specific Volumes of Oxygen and Nitrogen Vapour at the Boiling-point of Oxygen .
" By James Dewar , M.A. , D.Sc .
, LL. D. , F.B.S. Received January 21 , \#151 ; Read January 30 , 1902 .
In my paper on " The Boiling-point of Liquid Hydrogen determined by Hydrogen and Helium Gas Thermometers " || it was pointed out that a constant-volume gas-thermometer filled with oxygen gas , having a pressure at 0 ' C. of about 800 mm. , gave a very accurate value of the boiling-point of liquid oxygen .
As it seemed advisable to confirm this result indirectly , an attempt was made to determine the vapour -density of oxygen at its boiling-point by direct weighing , the intention being , if the experimental results proved at all encouraging , to repeat the work on a larger scale and with greater precautions .
As at present there is no likelihood of my being able to undertake the more accurate determinations , the results of the preliminary enquiry are presented to the Society .
They give in any case , with considerable accuracy , the Specific Volumes which have never been directly determined .
In order to obviate any question of the buoyancy of the air , two iflasks A and B of as nearly as possible the same air displacement were counterpoised on an ( Ertling balance .
The B flask remained per- * * * S * ' M.N. , R.A.S. , ' toI .
61 , p. 387 .
f * Roy .
Soc. Proc. , ' vol. 68 , p. 457 .
$ * Ast .
and Ast .
Phys. , ' vol. 11 , p. 808 .
S * Ast .
and Ast .
Phys. , ' vol. 12 , p. 912 . .
|| * Roy .
Soc. Proc. , ' vol. 68,1901 .
and Nitrogen Vapour at the Boiling-point of Oxygen .
361 manently on one scale of the balance during all the weighings , while the A flask was weighed , either exhausted or filled with oxygen ( or nitrogen ) , under various circumstances according as the experiments required .
J As the flask cooled in the liquid oxygen or air had to stand an internal pressure of from three to four atmospheres , it was considered expedient to select a spherical vessel of about 300 c.c. capacity , to which was sealed a narrow tube having a very carefully ground stopcock at its end .
Preliminary experiments were made to determine the change of volume of the flask when subjected to internal pressure .
The flask , filled with air at five atmospheres pressure , was left for 24 hours with the stopcock closed , without showing any leakage .
To determine the effect of pressure on the capacity of the flask , it was filled with water under one atmosphere pressure , and again with water under three to four atmospheres pressure ( which included the range of the : observations ) and the weights noted .
From these the coefficient of Expansion of the flask was found to be 0*000306 per atmosphere excess I of internal over external pressure .
The temperature coefficient of ( expansion of glass ( cubical ) was taken to be 0-000025 , which for a ( variation of temperature of some 200 ' altered the capacity of the flask ( by about 1*5 c.c. The content of the flask up to a fixed mark on the ( neck below the stopcock was determined to be 315*973 c.c. at 17 ' C. ; ( the content between the mark and the stopcock was determined both ( by measurement and by the weight of mercury it contained , and was ( found to be 0*127 c.c. Before each experiment the A flask filled with the gas under observa-| tion was exhausted to a pressure of from 2 to 4 mm. of mercury ( which ( was afterwards involved as a correction in reducing the observations ) ( and weighed , the weight a which had to be added to its scale pan to | balance the B flask being noted .
The A flask was then filled with ( carefully purified oxygen ( or nitrogen ) , and the temperature of the 1 flask and contents ( still in communication with the gas reservoir ) was 1 lowered by immersing it up to the mark in liquid oxygen ( or air ) until ( the gas ceased passing into the flask , and the pressure was finally ( equalised to that of the atmosphere .
This is really the most important ( part of the manipulation , as the accuracy mainly depends on giving ( sufficient time to the cooling , while at the same time taking care to ( avoid any excess of pressure that would necessarily lead to liquid ( condensation on the glass surface .
During the rapid inrush of gas , it ft js advisable to keep the pressure well below that of the barometer at ( the time , and to finally adjust the pressure at the end of the absorption .
( The vessel containing the liquid oxygen had a cardboard cover which ( crossed the neck of the flask at the mark , thereby preventing the ( cooled vapour of the oxygen from freezing the stopcock .
The remain-| ing portion of the tube between the mark and the stopcock , namely , VOL LXIX .
2 C 362 Prof , J. Dewar , The Specific Volumes of Oxygen 0*127 c.c. , was therefore approximately at the temperature of the room .
When the temperature of the flask and its contents became stationary the stopcock was closed , the flask removed , and after heating up to the temperature of the room , weighed against the B flask and any further necessary counterpoise b. The weight of oxygen vapour in the flask at its boiling-point was thus equal to a + b , subject to the corrections which have been indicated .
To the five sources of error or correction indicated above , there remains to be added that due to additional buoyancy of the air during the weighing of the flask , when filled with oxygen at its boiling-point , for these weighings were made at the temperature of the room , which would cause a rise of pressure in the flask and therefore expansion of its volume .
But this error was so small that it could safely be neglected .
As the intention was not only to ascertain the density of oxygen and nitrogen at their respective boiling-points under atmospheric pressure , but also under diminished pressure , experiments were made with nitrogen at ordinary temperatures and at pressures varying from about one-sixth of an atmosphere to ordinary pressures , in order to find the range of variation in the results with the 316 c.c. flask to be used in the subsequent low-temperature experiments .
In the following table are details of these observations , and the results reduced according to the ordinary gaseous laws .
Table I.\#151 ; Density of Nitrogen , ordinary Temperature .
No. Pi- Tx .
P. a. b. y. d. d0 .
1 mm. 742*0 15*5 mm. 742 *0 gramme 0*37 gramme -0 *0022 315 *973 0*001168 0 *001265 ; 2 742*0 15*5 742*0 0 *3668 -0*001 315*973 0 *001162 0 *001258 3 742*0 16 *0 742*0 0 *367 -0*0015 315*973 0*001161 0 *001259 4 751 *0 16*5 281*3 0*371 -0 2328 315 *913 0 *000442 0 *001265 5 751*0 16*5 284*2 0*371 -0*231 315 *913 0 *000447 0 *001269 6 751 *0 17 *0 268 *5 0*371 -0*239 315 *911 0 *000422 0*001269 7 753 *0 17 *5 176*2 0 *3705 -0*2850 315 *899 0 *000275 0*001262 where p\ = barometric pressure at the time of observation , Tx = temperature of the room at the time of observation , p = pressure of the nitrogen vapour in the flask during the experiment , a , b \#151 ; as defined before , V = the volume of the flask corrected for temperature and compression , d = the calculated density at Tx ' and pressure p , do = the value of d reduced to 0 ' and 760 mm. by the ordinary gaseous laws .
1 and Nitrogen Vapour at the Boiling-point of Oxygen .
363 Sfc In reducing these observations the following corrections were involved : a correction of 0-0015 gramme , due to imperfect exhaustion of the flask A while being weighed as empty ; the correction due to the neck of the flask between the mark and the stopcock was negligible ; the volume of the flask when corrected for temperature was 315-973 c.c. , and when it was necessary to correct further for excess of external over internal pressure , the values were as given in Column Y ; in the last four experiments this correction varied from about 0*059 to 0-073 c.c. i The first three experiments give a mean value of 1*260 grammes , at standard temperature and pressure , as the weight of a litre of nitrogen .
This is about a quarter per cent , higher than the accepted value of 1*257 .
The extreme variation in the individual experiments is about half a per cent. The average value of the results under about one-third of an atmosphere is 1*266 grammes ; the tendency under the low pressures being to make the density half a per cent , higher .
Considering that in the actual low-temperature experiments the mass of gas to be weighed would be at least three times greater , it was inferred that in spite of difficulties of manipulation and corrections the results might be anticipated to lie within a half per cent , of the true value , jj The following table gives the results of six experiments made on the density of the vapour of oxygen at its boiling-point taken as 90'-5 Table II.\#151 ; Density of Oxygen Yapour at its Boiling-point .
| ; No. JP = Pi-1 T , .
a. l. d. V. I| 1 2 3 r 4 5 6 mm. 775*5 770-3 771 -0 771 -8 775 -o 774-2 15'-5 16 *5 17 -5 17*5 16 0 16 -5 gramme .
0-388 0 -3865 0 -3875 0 -3875 0-388 0-3875 gramme .
1 0225 1-015 1 -0245 1 -0260 1 1 -028 1 -028 0-004402 0-004403 0-004432 0 004432 0-004422 0 *004425 227 15 227 07 225 -60 225 -59 226 *12 225-97 absolute , and under atmospheric pressure , where p , b , are the same symbols as used in the previous table , d is the calculated density at 90'-5 absolute and 760 mm. , and v is the specific volume 1 In the above reductions the following correct ons were involved : 0-002 gramme due to imperfect exhaustion of A flask while being weighed as empty ; 0*0002 gramme due to the contents of neck of the flask between the mark and the stopcock not being a , the volume of the flask up to the mark contracted at 90 *5 to 314-398 c.c. ; no correction was necessary for compression .
Thus the mean weight of one litre of oxygen vapour at 760 mm. and 90'*5 absolute is 4-420 grammes , and the specific volume is 2 c 2 364 Prof. J. Dewar .
The Specific Volumes of Oxygen 226*25 c.c. If the first two experiments are eliminated on the assumption that the proper equilibrium of temperature had not been attained , the average weight per litre would become 4*428 grammes , and the specific volume 225*82 .
Taking Regnault 's density of oxygen at 0 ' and 760 mm. , the density at 90'*5 in the ordinary way would be 0*0043137 , and the specific volume 231*82 c.c. Thus the volume given by the ordinary gaseous laws is 1*0246 times the average observed volume ; or we may put it that pv is diminished at the boiling-point of oxygen by 2*46 per cent. Again , while the ratio of the absolute temperatures is 3*017 , the ratio of the densities is 3*091 .
Behn* has determined by an indirect method the specific volume of oxygen , and finds the value 358 , which is nearly 60 per cent , greater than the volume found by the direct method .
The mode of procedure he adopts is to ascertain directly three quantities out of four in the ordinary thermodynamic equation correlating latent heat , temperature , increment of pressure to temperature , and specific volume , thereby deducing the unknown quantity .
Now of the three experimental values required , one , viz. , increment of pressure to temperature , can only be got by calculation from the vapour-pressure curve , and much depends upon the accuracy of this value .
Accepting Estreicher 's vapour pressures for liquid oxygen below its boiling-point as the most reliable , a Gibbs equation gives the increment per degree near 7 40 mm. pressure as 78*67 mm. mercury pressure or 106*93 grammes per square centimetre .
This value , taken along with the latent heat found by Behn , and the boiling-point , gives , when inserted in the thermodynamic equation , a specific volume of 223*55 , which is within less than 2 per cent , of the value found by the direct-density determinations . !
Further experiments were made on oxygen vapour at 90 ' *5 and under reduced pressures .
These experiments and their results are given in the following table .
The same symbols are used as in the preceding tables , except that d is the calculated density at 90'*5 absolute and pressure p , and v is 1 The corrected volume of the flask is entered under V. * ' Ann. der Physik , ' vol. 1 , 1900 , " Sublimafcionswarme der Kolilensaure und die Verdampfungswarme der Luffc .
" t In the same way Behn 's specific volume of carbonic acid would be contradictory of my proof that a constant-volume gas-thermometer filled with carbonic acid at about atmospheric pressure gives a very accurate value of its own boiling-point .
Assuming the ordinary gaseous laws , the specific volume ought to be 361*6 instead of 423 given by Behn .
Now my value of the increment of pressure to temperature at the boiling-point is 62*84 mm. mercury pressure , and this , along with the values used by Behn in the thermodynamic equation , gives 363 as the specific volume .
This comes much nearer the anticipated value of the constant .
and Nitrogen Vapour at the of Oxygen .
Table III.\#151 ; Density of Oxygen Vapour at its Boiling-point under Diminished Pressure .
No. Pi- \ Tx .
P- a. b. V. d. V. 1 mm. ; 0 741 -8 : 18*5 mm. 287 *0 gramme 0 *3695 gramme 0 *1315 314 *340 0 *001588 625 *74 2 741 *8 18*5 281*5 0 *3695 0 *1345 314*339 0*001607 622 *01 3 746*5 !
20 *0 279*0 0 *3690 0*1290 314*338 0 *001588 629*41 4 746*5 20*0 310*2 0 *3690 0*1880 314 *342 0*001776 562 *95 5 746*5 | 19*0 159 *4 0 *3690 0 *0850 314*323 0 *000907 1101*46 I In reducing these observations the following corrections were involved : a correction of 0-0014 gramme to 0-0015 gramme due to imperfect exhaustion of the A flask while being weighed as empty ; the correction due to the neck of the flask between the mark and the stopcock not being at 90'-5 amounted to 0*00003 , 0'00007 gramme and was practically negligible ; the volume of the flask , which , when corrected for temperature , was 314-398 c.c. , had to be further corrected for excess of external pressure over internal pressure by amounts varying from 0*0551 c.c. to 0*0742 c.c. If the first three experiments are averaged ( the pressures being so near ) , the weight of a litre of oxygen at 90o,5 absolute under a pressure of 282*5 mm. would be 1*5982 grammes .
The ratio of this density to the value previously found for one atmosphere pressure , viz. , 4*42 grammes , is 2*765 , and the ratio of the pressures is 2*690 .
It appears that the ratio of the change of density of the vapour of oxygen at 90'*5 absolute , under variable pressure , is greater than the ratio of the change of pressure .
It is clear , however , that it would be necessary to work upon a larger scale in order to get satisfactory vapour densities at low temperatures under pressures below that of the atmosphere .
The following table gives the observations on the density of nitrogen vapour at the boiling-points of liquid oxygen and liquid air respectively ; the first two were made in oxygen , the last four in air .
The symbols used are the same as before , except that d is the calculate density at T ' absolute and 760 mm. and v is 1/ d. Table IV.\#151 ; Density of Nitrogen Vapour at the Boiling-point of Oxygen .
__ l 1 No. P = Tx .
a. j b. d. V. T. T i 2 3 4 5 6 mm. 771*8 773*0 777*2 777 *3 777*3 777*3 17 *5 17*0 16*0 16*5 16*5 16*5 gramme .
I 0*3875 0 *385 1 0 *3885 0*3885 0 *3885 0*3845 gramme .
0 *8555 1 0*8555 0 *9575 0 *9500 0*942 0 *9235 .0*0039021 0 *003885 0 *004192 0 *004168 0 *004143 0 *004073 1 256*27 257 *39 238*53 239 *90 241*34 245*49 ( 90'*5 ) !
( 90*5 ) 84*05 84 *54 85 *04 86 *50 366 Specific Volumes of Oxygen and Nitrogen Vapour .
In reducing these observations , the following corrections were involved : a correction of 0002 gramme due to imperfect exhaustion of the A flask , while being weighed as empty ; the correction due to the neck of the flask between the mark and the stopcock not being at temperature T amounted to 0-00014 gramme and was practically negligible ; the volume of the flask , which only required to be corrected for temperature , was 314-398 c-.c .
Experiments 1 , 2 were made with liquid oxygen taken to be at temperature 90'-5 absolute .
The experiments 3 , 4 , 5 , 6 were made in one and the same sample of liquid air , with rising temperature .
For these temperatures , obviously only a few degrees below the boiling-point of oxygen , the ordinary gaseous laws may be held to apply , in order to determine their values .
Thus we may employ the formula m 90-5 1 = V--------- , 256-833 ' where 256-833 is the mean of the volumes of Nos. 1 , 2 , to get the temperature of the last four experiments .
The values thence obtained are entered in column T. The first two experiments made with liquid oxygen give a ratio of the nitrogen densities from my own values of 3"088 , the absolute temperature ratio being 3-017 ; my values for the ratio of the oxygen densities for the same range of temperature being 3-091 as previously deduced .
We may safely assume that if the density of nitrogen were observed at its boiling-point it would deviate as much from the ordinary gaseous laws as oxygen .
Further , the specific volume of nitrogen at its boiling-point of 78 ' absolute , would from the above formula be 22T3 as compared with 226'2 , the similar value found for oxygen .
The general inference to be drawn from these preliminary experiments is that reliable vapour densities may be determined at very low temperatures .
There seems to be no reason why the vapour density of hydrogen at its boiling-point should not be accurately ascertained ; only , as in this case the internal pressure in the weighing flask would amount to nearly 15 atmospheres , it would be advisable to construct the flask of some metal or alloy .
A flask of the size used in the oxygen experiments filled with the vapour of hydrogen at its boiling-point would be equivalent in weight to between 4 and 5 litres of hydrogen at the ordinary temperature and pressure , and such an amount of material ought to give density results at the boiling-point of hydrogen of considerable exactness , notwithstanding the great manipulative difficulties that would necessarily be involved in the execution of such a determination at 21 ' absolute .

