JO- 



NATURE 



[Al'C.LST I, 1S95 



i6-iS" 



We i;..^ .....> minutely all the determination of density of 

 such >amplcs, because, although they refer to an imperfectly 

 |>uriliei.l sample, yet they show that the density is very low, and 

 they trace, moreover, the gradual change as one ingredient 

 after another is removed. 



The bn'^erite which had been heateil in a vacuum was next 

 fused in successive portions with hydrogen potassium stdphate. 

 A large quantity of gas was evolved, consisting of sulphur dioxide, 

 carlion dioxide, nitrogen, and helium. The sulphur dioxide was 

 removed with chromic mixture, and the carbon dioxide with 

 caustic soda : the yield was 45 c.c. The density was then 

 detcrmine<l. 



Volume of bull' 

 Temperature ... 

 Pressure (corr. ) .. ... 75J'3 mm. 



Weight ... ... ... ... 001035 gra'" 



Density (0= 16) 374S 



No alteration in volume occurred on passing the gas for 

 several hours over red-hot cop|>er oxide. Hence no hydrogen 

 was present in the free state ; and if combined, passage over 

 cop|)er oxide does not decompose the hydride, as was seen 

 Iwfore, when the water produced was weighed. It may be 

 remarked that every known hydride would yield its hydrogen 

 on such treatment. 



This sample of gas was ne.\t circulated over red-hot magnesium 

 for several hours. It is hardly necessary to state that the mag- 

 nesium was first heated to redness in a vacuum so as to remove 

 hydrogen. In ca.se any shoidd escape removal, however, a red- 

 hot tube of copper oxide formed |»art of the circuit, as well as 

 a tube filled with phosphoric anhydride. .Some caustic soda 

 solution was present in the reser\oir above the niercur)', which 

 would have absorlied the products of combustion of any hydro- 

 carlwn present. The density of this gas was calculated from the 

 data appended. 



Volume of bulb ... .. ... 33'023 c.c. 



Temperature ... ... ... I4'SS' 



Pressure (corr.) ... ... ... 7560 mm. 



Weight 0-00845 gram 



Density (O = 16) yoyi 



On examining the magnesium tube, after it had cooled, it was 

 found that on moistening it ammonia was evolved. The gas 

 was, therefore, again circulated over magnesium, at a somewhat 

 higher tem|x;raturc, so high, indeed, that the gas must have 

 [M-ssed repeatedly through magnesium va|K)ur. On pumping out 

 the tubes, an accident led to the loss of a few c.c. of gas ; hence 

 the weighing bulb hail to be filled at a somewhat reduced pressure. 

 The density is given t>elow. 



\'oluine of bull' 

 Temperature 

 f'res.sure (corr.) 



Weight 



Density (O =. 16) .. 



33023 c.c. 



615-8 mm. 

 0-0049 gfam 

 2-187 



Again, on moistening the broken magnesium lube, ammonia 

 was evolved ; it was recognised by its odour and by its turning 

 rc<l litmus paper blue. 



A further experiment was made with broggerile. 30-8 grams 

 were heated in a vacuum and the gas was collected over mercur)-, 

 on to the surface of which a few c.c. of caustic .smla solution 

 were intrcxluced. The yield of gas was 65 c.c. It wascirculated 

 over copper oxide to remove hydrogen, and its density was then 

 dctcrmmcd. 



Volume of bull. 

 Temperature 

 I'rcssurc (corr.) 

 Weight 

 DetLsity 



33023 c.c. 

 • "970 

 .. 756-7 mm. 



0-0068 gram 



2-481 



The <lensily of this samiilc is almost coincident with that of a 



f>rcviiiii- -.11111.1. . idryi ..l.i.^incd in the same way, after it had 

 "ccn I This gas was next circulated over 



v-ri- to remove nitrogen. Again, it is 



' I'l juaiiy li.iui.i the gas must have been mixed with 



r vaiHiur, for the magnesium had been completely 



'it part of the combuslion-lulie, and had 

 rid. Again, the priHlucI, when moistened 



I ammonia, proving that nitrogen had been 



removed. The density of this sample was next taken. 



\'olume of bulb 

 Temperature 

 Pressure (corr.) 

 Weight 

 Density 



33-023 cc. 

 19-17- 

 756-7 mm. 

 0-0056 gram 

 2-044 



The copper oxide tuln; was omitted during this circulation ; 

 hence the density was low, 2 044. The spectrum of this gas 

 showed hydrogen lines and feeble nitrogen bands. A second 

 determination of density, in which the bulb was freshly filled, 

 gave, at the same pres.sure and at a temperature dift'cring by 

 only I' from the previous one, an identical weight. I'urther 

 circulation for a whole day over red-liot magnesium, raised to 

 the highest temperature which the lube could stand, gave a 

 specimen from which hydrogen and nitrogen were absent ; at 

 least, the Ixirest trace was visible in a vacuum-tube filled at a 

 fairly high i)ressure ; and care was taken to interpose a red-hot 

 copper oxide tube, and, as usual, a tube containing phosphorus 

 pentoxide. The e(fecl of this circulation was to raise the 

 density. 



Volume of bulb ... ... ... 33-023 c.c. 



Temperature I7'l" 



Pressure (corr.) ... ... ... 763-2 mm. 



Weight 0-0060 gram 



Density (O = 16) 2-152 



It is of interest to note that this .sample, procured by heating 

 broggerite in a v,acuun), has a density practically identical willi 

 that of gas obtained by fusing bniggerite with hydrogen pot;is- 

 .siuni sulphate ; that sample had density 2-187. 



We next proceeded to extract the gas from 6-96 grams of 

 Swedish cleveite. When heated in a vacuum, the gas was rajiidly 

 evolved at first, more quickly than from broggerile. Almul 

 60 c.c. were obtained, and, after treatment with soda, the residue 

 occupied 26-3 c.c. As this was not surticieiit for our purpose, 

 and as we had already by density and spectrum proved the 

 identity of gas evolved from bniggerite on healing, and on fusion 

 wilh acid sulphate, the remaining cleveite was mixed withaliout 

 five limes its weight of fused and dried hydrogen potassium 

 sulphate, placed in a lube, and healed in a vacuum. A further 

 quantity of gas was evolved, which was at once treated with 

 caustic soda solution. Both quantities of gas were mixeil. This 

 samjile was then circulated over copper oxide for several hours, 

 and the density was then determined with the following result. 



Volume of bulb 

 Temperature... 

 Pressure (corr.) 

 Weight 

 Density 



33-023 c.c. 



19 43" 

 763-2 mm. 

 00061 gram 

 2-205 



NO. 1344, VOL. 52] 



The spectrum of this gas showed the merest trace of nitrogen, 

 but no hydrogen. The density, it will be seen, is practically 

 coincident wilh thai of the gas from bniggerite. It is note- 

 worthy that the gas from cleveite contains no nitrogen. We 

 are absolutely certain that the presence of nitrogen in the gas 

 from broggerile is not to be explained by leakage of air, for 

 the tightness of the apparatus was frequently tested during each 

 operation. 



We have therefore three determinations of density, and the 

 mean may be taken as approximately correct to within 0-05. 

 They are : 



(l.as from bniggerite by heating ... ... 2152 



(las from briiggerite wilh II K.SOj ... 2187 

 Gas from cleveite 2-205 



Mean 



2-181 



All these samples of gas were now mixed and p.issed thmugh 

 the u.sual alworbents vn nitmgen and fir hydrogen, namely 

 magnesium, copper oxide, so<la-lime, ami phosphoric anhy- 

 dride. The riensily of this sample was ihen determined with 

 the larger bulb. The ernir due lo error in weighing; cannot in 

 this ca.se amount to more than 0-3 |)er cent., and is pnibably 

 less. Of course, the purity of the gas would aflect the result. 

 The (lata are as follows. 



Volume of bulb 



Temperature... 



Pressure (corr.) 



Weight 



Density (O = 16) 



162-843 c.c. 



1707" 

 764-9 mm. 

 003057 gram 



2-2lS 



