214 
LORD RAVLEIOH OX A XKW .MANOMETER. 
into accmiiit. Passing over this (jnestion for the present, vve may consider how far 
the results conformed to Boyle’>s law. The agreement of the ratios, except, perhaps, 
at the highest pressures of alxait iT milliins. of mercury, was sulficiently good, and of 
itself goes a long way tr) conlirm Boyle’s law. Iji strictness, all tliat the constancy 
of the ratio can proye is that tlie relation between pressure (/>) and density (p) is of 
the fo]'ni 
•p = 'xp’‘ .( 1 ), 
where n is some numerical quantity. To limit n to the yalue unity, the constancy of 
ilie ratios might ])e fidlowed up into the region of pressure for which Boyle’.s law is 
knoYcn to hold, l)ut this can scarcely be said to haye been done here. Otherwise, vre 
need to know what the ratio of densities in the two positions of the mercury really is, 
and iiot merely that it remains constant. 
In the case of the original yolume cTiam])er the first was the method emvjloyed. 
O X V 
’file smaller volume, defined by the upper mark in the yolume tube and by the 
'■ 2 >oint ” in the manometer, yras fitted with dry air at a known atmospheric iiressure. 
The included air was theii isolated and exiianded until it occiqiied the larger 
(ajijn'oximately double) yolume, and the new piressure determined by obseiwation of 
the difference of leyels in the tube and in a mercury reseryoir similarly fashioned, 
'fhe operation was rather a difficult one, and the result was only liarely accurate 
enough. The ratio of yolumes thus determined by use of Boyle’s law, as ajijilied to 
air at atmospheric and lialf atmos|)heric pressures, agreed sufficiently well with the 
ratio of jiressures found by the manometer for rare hydrogen and nitrogen: and thus 
BoimE’s law may be considered to be extended to these rare mises. The rarefaction 
*j O 
Y'as carried doivii to a total jiressnre of only ’02 millim. At this stage discrepancies 
of the order of 5 per cent, are to be ex^iected. 
Having obtained fairly satisfactory results with iiydrogen and nitrogen, I returned 
to oxygen, fully ex})ecting to yerify the anomalous behaviour described by Bohr. In 
this I have totally failed. The gas was 2 )re 2 jared Iw heating ])ermanganate of 2 )otash, 
dried by 2 )lios|)horic anhydride, and may be regarded as fairly 2 )ure. The region oi 
2 )i'essure round '7 millim. was carefully examined, use being made of the inter¬ 
mediate diyisions of the 50 cub. centims. rano-e of volume. No unsteadiness of the 
kind indicated by Bonn, or apiirecialile dejiarture from Boyle’s law, was detected. 
And when the jiressures were diminished down to a few hundredths of a millimetre, 
there was no falling off in the product of ])ressure and volume. The observations 
YY're rejieated a second time with a fresh su 2 ) 2 )lv of oxvgen, 
I’he exiierience gained up to this date (August, 1900) showed that the manometer 
worked well, and that there yns no difficulty about the vacuum, hut I was uot alto¬ 
gether satisfied Avith the Avay in AA'hich the A'olumes had been determined. There 
was some AAOint of elegance, to say the least, in using Boyle’s Liav for tliis purpose, 
and barely adecpiate accuracy in the a])])lication itself The latter olijection might 
