s44
G E O R G E R.ICPH.IIL S N l T H
iveiglits of sulphur corresponding to the several atoinic Iveights of silver. A T O M I WEIGHT ~ O F SLXPHL'K. I f Ag=i07.930 and cl ...-3j,4;j, S = j z , i 1 3 Ii.ig--Io7.8go and C1 - j j . 4 6 0 , S ~ 3 2 . o ; S If Xg-ro7.SSo and C1 - ~ 5 . 4 5 ; , S - 32.069
The lowest value in this case, as \vel1 as in the case ot' nitrogen, is the oiie supported by the recent work 011 the densities of gases. T h e unaniiiious verdict of these veri different resnlts is interesting, and possibly significant. I n conclusioii, it is a pleasure to aclmx\~letlgethe generous assistance of the Carnegie Institution of l17ashington, uithout ivhich the present work could not h ave been perf or met1.
Summary T h e most important results of the research may be briefly suiiiiiied up as fo1lon.s: ( I ) ,i method for the preparation of pure silver sulphate ivas devised. ( 2 ) T h e specific gravity of silver sulphate (previously fused) was foulid to be j . 4 j . ( 3 ) Indication \vas obtaiiied that Stas ivas unable \vIiolly to reduce silver sulphate in hydrogen. ( 3 ) Silver sulphate was found to Le occluded by silver chloride from sol~itionscontaining an excess of sulphuric acid. ( j) It \vas proved that silver sulphate can be completely converted into silver cliloritle by lieatiiig in a current of hydrochloric acid gas. ( ( j ) 10o.ooo parts of silver sulphate \\'ere thus found to yield 91.933 parts of silver chloride. ( 7 ) The atomic weigh: of sulphur as calculated from this ratio, if oxygen is take11 as 1fi.000, with several assurlied values for silver is : .~g=107.9j AgZio7.S9 .A~=IO~.SS
S
32.1
S
32
S
32.~69
13
0;s
(CONTRIBUTION FRO31 T H E C H E M I C A L I , A B O R A T O R S
OF THE USIVkRSITY OF
ILLINOIS ).
THE CONSTITUTION OF AMMONIUM AMALGAM. BY G E O R G EMc.PH.+ILS M I T H .
T h e substance known as ammoniuni amalgam was first obtained by Seebeck', immediately following Davy's announcement of the discovery of potassium and sodium. Seebeck prepared t h e amalgam by the electrolysis of moistened ammonium carbonate, with a mercury cathode. About the same time, BerzeIius and Pontin obtained a like result with a .inn. 66, 191 (1808).
A M M O N I U M AMALGAM
845
solution of ammonia. This discovery they coiiimunicated to Davy, declaring their conviction that ammonia, like potash and soda, must be an oxide, and that the new substance was a compound of its metallic constituent with mercury. Davy’ immediately began a series of elaborate experiments on the preparation and properties of the anialgam ; and, in a n account of these experiments laid before the Royal Society, he first used the name “ammonium)’ to indicate the supposed metallic basis of ammonia. Davy discovered that ammonium amalgam could also be readily prepared by the action of moistened sal ammoniac on potassiuni amalgam. I n the following year Gay-Lussac and ThCnard investigated the properties of the amalgam, and were led to regard it as a triple compound of mercury, ammonia, and hydrogen. ThCnard afterwards described the amalgam, in his ‘‘ Traifk de Chintie,” under the name of “ammoniacal hydride of mercury.” I n 1816, Ampbe’, in the passage where the now universally received views on t h e constitution of ammoniacal compounds are first propounded, refers to the amalgam. Speaking of the difficulty of assimilating the constitution of ammoniacal to metallic salts, he says that this difficulty would disappear if it was admitted that, just as cyanogen,although acompound body, exhibits all the properties of the simple bodies which are capable of acidifying hydrogen, so the combination of one volume of nitrogen and four volumes of hydrogen which is united to mercury in the amalgam discovered by Seebeck, and to chlorine in the hydrochlorate of ammonia, behaves i n all the compounds which it forms like the simple metallic substances. This theory was more fully developed by Berzelius, and was soon generally received, excepting as regards the amalgam. Concerning this, various conflicting opinions have been entertained. Daniell, for exaniple, speaks of it a s a mere mixture of mercury and gases, resulting from t h e cohesion of the mercury and the adhesion of it to the gases ; and he cites the absorption of oxygen by melted silver as a similar case. Grove, i n 1841, made a few experiments on the amalgain and advanced the idea t h a t it is a chemical compound of mercury and nitrogen, merely swelled u p with hydrogen. Weatherills investigated the amalgam, and came to t h e conclusion t h a t it is not an alloy of nierciiry and ammonium. Landolt‘ showed that the electrolytically prepared anialgam contains approximately 2 vo!umes (2.15-2.4) of ammonia to I volume of hydrogen. According to Landolt, however, the amalgam differs from those of potassium arid sodium in not giving double decompositions with aqueous soluPhil. Trans. 1808, 353, Ann. chirn. phys. 1Sr6, 11, 16 ( N o t e ) . ’ Silliman’s Attier. J. ( 2 ) 40, 160 (1864). Ann. (Suppl.) 6, 346 (1868).
‘
846
GEORGE NCPHAIL SMITH
tions of copper, silver and iron salts; although it contains the group NH, in combination with mercury, it should not be considered a s a veritable metallic amalgam. I n 1 8 7 2 , Routledge’, who regarded Landolt’s volume deterniinations a s unsatisfactory, repeated his experiments and obtained as the mean of four determinations the values : vol. NH, : vol. H, = 1 , 9 6 : I . Routledge also determined the effect of pressure on the inflated mass ; its compressibility agrees fairly well with the supposition of its being a gas and mercury, though it is somewhat less compressible. Routledge concludes that ammonia and hydrogen are here coniliiied, because of tlie evolution of the gases in atomic proportions’; and tligt the compressibility of the mass proves t h a t t h e enlarged volume or swelling u p is due to free gases entangled in i t . T h a t the S H , is combined with mercury he thinks probable, because of the apparently riniform diffusion of NH, throughout the mass, and from the fact that such a union would be only one additional instance of the innunierable cases i n which this radicle plays the part of a metal. H e says : “ W e may adniit that such a compound is originally formed, and decomposes rapidly into mercury, ammonia, and hydrogen, while the gases, becoming entangled iii the mass, impart to it that remarkable turgescence, which is not, however, a property of tlie original compound, but merely an accidetital result of its decomposition,’’ T h e next important addition to the solution of the problem was furnished by LeBlanc,:’in 1890. LeBlanc considers that 1,andolt’s failure to obtain double deconipositioiis with amnionium amalgam and the salts of copper, etc., did not disprove the metallic nature of ammonium in its amalgam. I t is rather possible that. owing to its instability. ammonium has not the force necessary to displace metals from their compounds, but that it immediately decomposes on severing its union with mercury. Employiiig niercury cathodes, LeBlanc electrolyzed solutions of alkali, and alkaline-earth salts, and measured the polarizations between the res d t i n g amalgams and amalgamated ziiic. H e found them greater by about I volt than the polarization arising from the electrolysis of hydrochloric acid, H e then electrolyzed amnioniuni salt solutions under the same conditions, and compared the resulting polarization on tlie one hand with those given by the metallic amalgams, and 011 the other hand with that arising froni the electrolysis of hydrochloric acid. ,iccording to whether the value obtained with ammonium amalgam corresponded with the.former or with the latter of these values, the conclusiou could h e drawn that here i11 the mercury a body analogous to tlie alkali metals Clierri. S e w s , 26, 2 1 0 . T h e writer acknowledges his indebte,liiess to this paper for much of the historical data in t h e preceding. I t is. however. liard to see how this fact alone call prove coilibination. Z. phy-sik;. Chtin. 5 , 467. ~
A M M O S I U M AMALGAM
847
was present, or that the polarization was due only to hydrogen.’ T h e values obtained led to the conclusion that a metallic body analogous to potassium and sodium was combined with the mercury to form ammonium amalgam. A difference between it and sodium and potassium amalgams was noticed in the greater fall in potential, in the case of ammonium amalgam, on breaking the primary circuit. This LeBlanc ascribes to the greater instability of this amalgam, as well as to its low KH, content. Some ten years later, Coehn and D a m e n b e r g * compared the phenomena of depolarization, which were observed in the case of alkali salt solutions with a mercury cathode, with those observed with ammonium salts under the same conditions. T h e potential necessary for the continuous evolution of hydrogen on a mercury cathode is I . 5 2 volts. Coehn and Dannenberg found that if alkali metals are present in the solution hydrogen is evolved a t a .point below I 3 2 volts. This point is characteristic of the metal, and corresponds to the formation of its amalgam. I n the case of ammonium such a point was found also to exist ( a t 1.24volts), in full analogy with the alkali metals. In consequence of the results of LeBlanc and of himself and Dannenberg, Coehn’ repeated Landolt’s experiments with ammonium amalgam and copper sulphate solution, only to meet with the same negative result. But, on treating ammonium amalgam, prepared at oo by the electrolysis of a solution of ammonium sulphate, with a solution of copper sulphate, previously cooled to o o , h e found t h a t copper was displaced from the solution. I t is conceivable, however, that this reduction might be due to nascent hydrogen, resulting from the decomposition of the amalgam ; and, in order to remove this objection, Coehn sought to reduce metals by means of ammonium amalgam which are not reducible by hydrogen. On bringing t h e amalgam into a solution of cadmium sulphate, cadmium was precipitated, and, i n order to remove the still possible objection that hydrogen might be liberated from the amalgam a t a sufficiently high pressure to reduce the cadmium, i t was finally shown that zinc, ‘‘ the most electropositive metal’ that can be deposited from aqueous solution,’’ is also reduced from its sulphate solution by ammonium amalgam. Coehn says this should remove t h e last objection as to the metallic nature of ammonium. But Coehn has overlooked the fact that, according to LeBlanc,’ atomic Not being a n ion, NH, would have no influence. 2. physik. Chem., 38, 609 (1901). a 2. anorg. Chem., 2 5 , 430 (1900). *Zinc, however, is not in reality the most positive metal that can be deposited from aqueous solutions in presence of mercury. Kraut and Popp ( A n n , , 159, 188 (1871) ) discovered t h a t even potassium is precipitated from its salt solutions by sodium amalgam. Cf. Smith : J. Physic. Chem., 8 , ZIZ (1904);9, 13 (1905). Am. Ch. J. 37. 506 (1907). loc. cit., 476 f.
848
GEORGE YCI'HAIL
SNIT13
hydrogen is capable of precipitating zinc under strikingly similar conditions. T o 1 0 cc. S I HC1 Le Blanc added n siiiall amount of N / r S a C l o r N, I BaCI, etc., an:l on e!ectrolyziiig the solution 110 increase indicain the voltage of t h e po1arizatio:i current was observed,-an tion that amalgam forination did not take place. This he explains as t h e result of the greater migration velocity of hydrogen a s compared witli the nietals; relatively few metallic atoms would arrive a t the cathode, and, owing to the liigii concentration of hydrogen, eveu they would be prevented froiii amalgamating. On the addition ot ZnCI,, hoiverer. in place of iYaC1 or BaCI,, zinc anialgaiii was formed and tiie evolution of hydrogen 011 the cathode ceased altogether. LeBlanc says this process cannot be primary, for against tliis tlie experinieiits with S a C i , etc. soiutions speak, arid furthermore it is not apparent how the evolution of hydrogen could cease i n that case. T h e phenoinenon is doubtless secondary : hydrogen is separated, aiid this precipitates the zinc froiii t h e zinc chloride solution nit11 the forniation of hydrochloric acid." I n a recent inr-estigatioii, hloissan prepareti aiiiiiioiiiiini amalgam b y the action of pssty sodium anialgniii on a solution of ammonium ioditie in liquid ammonia.' At 39' the reaction proceeded smoothly, and tlie amalgam became inore fluid without iiiflatioii or evolution of gas. Tlie product was washed with liquid :Inirnonia, and then, at -Soo, with ether. but on coniT h e aiiialgain was found to be stable i n a v:icuuni a t --80°. iiig to the rooiii temperature i t gradually decoiiipmed : i t began to swell Durat ---30°, and at - - I j oits voluine \vas 1 5 - 2 0 times that a t -Soo. ing its decompositioii the temperature of the sealed tube containing t h e amalgam rose j o - G o above the surroundiiig temperature. As the niean of seveii e x per i iii en t s , 11oi s w 11 01) t ai 11 ed t ii e v ol Li in e r el a t i oii N H :, :H = 1 . 9 9 : ~ . H e s a y s : " 111tliis series of aiialyses the radicle S H , , corresponding to two volumes of an1nioiii:i arid one volume of hydrogen. would seem to exist i i i the iiietailic ina$s prepared a t -39'. T h a t hon.ever can hardly be ; in reality the hydiogeii of li! drioclic acid apFears to furnish a simple or double aninioniacal metallic h!.dride, and it is tlie decomposition of this hydride which produces the iii~flatioi~." Still more receiitly, Rich anti Travers' have made a series of cryoscopic determinations n-ith aninioriium ainalgam. similar to those of TaminannJ with + o d i u m amalgam. From their results they conclude that : " T h e resenililance in the behavior. of sodium amalgain and ammon i u m arnalgani is very marked, and leaves 110 room for doubt that t h e latter is reallj- a solution of aninioiiiuni i i i iilercury. T h i s view is sup'
I
I3ull. 9oc. chim. ( 3 ) 2 7 , 714 ( 1 9 0 2 ) . The liquitl arxiiiionia alone had n o action whatever on the sodium amalgam. " J. C h e m . %c. (1,ondon~89, 1172 1906). %. pliysik. C l i e m . , 3, 440. I
AMMONIUM AMALGAM
849
ported by the fact that the amalgam continues to evolve ammonia and hydrogen after it has been allowed t o stand for many hours a t the temperature of t h e experiment. Further, if the temperature of the amalgam is allowed to rise until the volume begins to increase, and the anialgam is then cooled, it appezrs to return to its original state ; hence it is not impossible that ammoninm ’ can exist transitorily in the free state.” It is seen from the foregoing that a t the present time there are three theories in regard to the constitution of ainmoniom amalgam : ( I ) the German theory of LeBlanc ( 1 8 g o ) , according to which it is a compound of the metallic radicle NH, with mercury; ( 2 ) the French theory of , in harmony with t h e old view of Gay-Lussac and hloissan ( ~ g o z ) mho, T h h a r d , regards the substance as a simple or double ainmoniacal hydride of mercury ; and ( 3 ) the theory of Rich and Travers (1906), according to which it is simply a solutiou of free Xu“, in mercury. T h e experimental results of the present paper are in complete accord Furthermore, they shorn the views of with the theory of LeBlanc. Moissan and of Rich and T r a r e r s to be untenable.
Experimental Part’ Moissan’ supports his opinion that ammonium amalgam is in reality an ammoniacal hydride of mercury, not coutaining the radicle NH,, with t h e sole fact that wheii a so!ution of sodium hydride in sodium amalgaill is allowed to stand with aqueous ammonia, an intense swelling takes place, with formation of a buttery mass which maintains itself for two or three days in the animoniacal soltition ; while, if sodium amalgam alone is treated with the same liquid, hydrogen is slowly evolved, with no itiflation of the amalgam. AIoissan’s experiments with sodium amalgam and ammonia have been extended by the writer to potassium, lithiuin, barium, stroiliturn aud calcium amalgams. T h e amalgams were prepared by the electrolysis of salt solutions of the corresponding metals with niercury cathodes. T h e mercury was contained in a small beaker, under the solution to be electrolyzed, and was connected with the circuit by means of a platinuin wire fused through t h e closed end of a small glass tube. T h e tube in turn contained mercury, and into this dipped the copper wire of the circuit. T h e anode was of platinum foil. T h e mercury was stirred from time to time during t h e electrolysis, in order to distribute the amalgam throughout t h e mass. T h e amalgams prepared were approximately of T h e mercury used in t h e following experiments was carefully purified by shakiiig it with a solution of potassium dichromate acidified with the sulphuric acid, and then causing i t t o run several times in succession, through a capillary tube, into a tall, upright cylinder containing a solution of mercuric nitrate, strongly acidified with nitric acid. Its purity was finally confirmed by analysis. T h e salts employed were of Kahlbaum’s manufacture. Iloc. cit.
850
G E O R G E NCPHXIL SMITH
tlie coinpositions : 0 .j per cent. E( : 0 .j per cent. S a ; 0.04 per cent. Li : O . 3 j per cent. Ba ; 1 . 0 per cent. S r : and 0.0; per cent. Ca. 111 six test tubes, mixtures of i cc. each of the respective aiiialgams with j cc. of saturated aniinoiiiuni chloride solution were allowed to stand at 20'. T h e potassium ~ninlga1iigradually swelled u p , until at the end of one hour its volume had increased to g cc. I t remained stationary a t this point for many hours. T h e sodinin amalgam overflowed the test tube and reached a volunie of about 20 cc. in one minute. T h e lithium ainalgain gradually expanded to 3 cc., but did not exceed that volume. T h i s was apparently owing to tlie size of the gas bubbles, which quickly e:caped from the amalgam. T h e barium amalgam expanded very slowly, attaining a maximum volume of 2 . 5 cc. in one hour and 2.j niiiiutes. T h e strontiuni amalgam expanded to I O cc. in 5 minutes, and after 40 minutes it had regained its original ~ ~ o l u m e T. h e calcium amalgam expanded to nearly 2 cc. in x hour ar:d I j minutes. I t then gradually receded to its original volume. After having thus ascertained that all tlie amalgains in question gave ainnionium amalgam (recognizable by the phenoinena accompanying its decomposition) with animoniuni chloride solution, the experiments were repeated in t h e same wa). with aqueous ammonia. T h e amalgams were treated, i n portions of I cc., with 5 cc. of chemically pure ammonia water (sp. gr. o . g o ) , taken from a freshly opened bottle. 111 order to prevent the formation of aniriioiiiuni salts, clue to the absorption of acid vapors by the aninionia, the test tubes were loosely stoppered and then covered ivitli larger, inverted test tubes. I n this way tlie auinionia was completely protected from any acid fumes preseiit i l l the air. Without this precaution the sodium aiiialgalii began to inflate after I O or 1 2 hours. Iii this solution tlie potassium amalgam attained a volume of 4 cc. in 24 hours. This volume remained constant for eight days, and on tlie ninth day it had shrunk back to I cc. Hydrogen was very feebly evoived all along, but much more slowly than from sodium amalgam and t h e same solution. T h e sodium amalgam did not become inflated at all, and hydrogen was evolved from it with about the same rapidity as in a blank experiment with I cc. of thesame amalgam and j cc. of distilled water. T h e lithium amalgam expanded to 3 cc. in 20 minutes, and then i t gradually went back to I cc. in about six hours. T h e barium, strontium and calcium anialgamsslowly swelled to maxima of 1 . 2 cc. in 2 j minutes, 1.3 cc. in one hour, and 1.4 cc. in three and one-half hours, respectively. T h e inflation of the amalgams is plotted graphically on the accompanying charts. Although the sodi,um amalgam, for some reason, does not become inflated when treated with aqueous ammonia, all the others do ; and, taken in this connection, tlie experiments of Moissan with sodium amalgam alone, and with sodium amalgam containiiig sodium
851
AMMONIUM A M A L G A M
They furnish no evidence that am-
hydride are seen to be irrelevant.
CHART i. Inflation with iVH*Cl Solution at Room Temperature.
monium amalgam does not contain the radicle IVH,, or that it is an ani-
1
I
20
.e
$0
*Q
3-
ilD
110
(60
C H A R T 2.
Inflation with Aqueous Ammonia at Room Temperature.
inoniacal hydride of mercury.
It0
GEORGE MCPHAIL S X I T H
8j2
Moreover, a very slight tendency towards inflation was noticed in t h e case of t h e sodium amalgam and ammonia ; the action of the water appeared, however, to predominate over that of the slightly dissociated ammonium hydroxide present in the ammonia. T o obviate this difficulty the experiment was repeated several times in duplicate as follows, -and always with similar results. One of the tubes, cotitainirig I cc. of amalgam and j cc. of ammonia, was allowed to stand a t the ordinary temperature, as before, while the other was kept for about one hour in a freezing mixture a t from 3' to ioo below zero. T h e amalgam iii the first tube evolved hydrogen continually a t constant volume, while that in t h e cold ammonia showed at 'its surface a slight, but distinct inflation. LVhen taken from the freezing mixture and allowed to stand a t room temperature, it attained in 4j minutes a volume of from 1 . 3 to 1.7 cc., depending on the Concentration of the sodium ama1,oani used, and the degree and time of the cooling. T h i s volume it niaititaiiied for a coup!e of days. T h e question was next investigated whether aninioniurn amalgam contains the metallic radicle XH,. Coehn's' work with ammonium amalgam aiid solutions of copper, cadmium and zinc sulphates would be direct evidence of this, if it were not for t h e possibility that the zinc, etc., might have been reduced by nascent hydrogen arising from the decomposition of a m m o n i u n amalgam. In addition to showing that nascent hydrogen is capable of reducing zinc, LeRlanc' at the same time showed that bariam, potassium, etc., are not reduced in this way, even i l l the presence of mercury-. Therefore, if ammonium aiiialgani should he found capable of di.;placiiig barium, potassium, etc., from their salt soltitions, this tlisplacement conld only be the result of an exchange with S H , iu the amalgam. T h e writer:' has found that the metals of the alkali and alkaline-earth groups are reversibly displaceable in the form of their anialgams, and if amnioniuni is analogous in its arrialgatii to potassiuni in potassium amalgam, it should also possess this characteristic. T h i s has actually been found to be t h e case. Aiiimoniutn amalgam was prepared by the electrol of a solution of 2 j g. of animonium carbonate [volatile without residue:) in 50 cc. of water. T h e solution was previously heated to 7 5 O , in order to remove any anirnoniuin carhainate present in it. T h e electrolysis was carried out, a t -3' t o o o , i n a hoo cc. beaker, the cathode coirsisting of 2 kg. of mercury, connected with the circuit a s already described, and the anode of a Iiorizontally disposed strip of platinum foil. T h e current strength was 2 . 3 - 2 . 5 amperes a t 7 yolts, and on breaking the primary circuit, a polarization current of about three volts was noted. T h e electrolysis was con-
'
loc. cit. " l o c . cit. .' .Im. C11. J., 37, jo6-j42 \rgo;).
AMMONIUM AMALGAM
853
tinued for five minutes (=- about 0.13 g. NH,). during which the cathode was frequently stirred with the thermometer. T h e solation was a t once poured off, and the amalgam washed once by decantation with ice'water. I t was then rapidly poured through ice water contained in a series of five beakers, the wash water being decanted each time. T h e freshly prepared ammonium amalgam was immediately added, in portions of about 500 grams, to four solutions which had been previously cooled to - T '. These were : ( I ) a saturated solution of potassium chloride; ( 2 ) a solution of 25 g. of potassium hydroxide in 5 0 cc. of water; ( 3 ) a saturated solution of barium chloride ; and (4)a saturated solutio11 of barium hydroxide. T h e mixtures were allowed to stand for 30, 45, 6 0 and 75 miiiutes respectively, the temperature being kept between oo and -3'. T h e solutions were then decanted, and the amalgams washed a s already described in ice water. They still contained ammoniuni, as was seen from their tendency to become slightly inflated. Especial care w a s taken t o prevent any of t h e original solution from being carried along; the amalgams were poured through the water in a thin stream, and eight beakers, each containing IOO cc. of ice water, were used in the washing. T h e barium-ammonium amalgams displayed a tendency to form bubbles a t the surface after pouring, and in every case these were broken with a blunt glass rod before the amalgam was poured into the next beaker. T h e wash-water from the seventh beaker was tested for barium, of which not a trace was present. Moreover the experiments were performed five or six times with identical results. T h e amalgams were finally extracted with hydrochloric acid, and the extracts analyzed. In making the analyses the extracts were first evaporated to dryness. I n every case some ammonium chloride was present,-especially in the case of the potassium residues. T h e barium residues consisted mainly of well formed crystals of barium chloride. T h e potassium residues were freed from ammonium chloride by volatilizing the latter, after which the potassium was weighed a s K,PtCl,. T h e barium residues were taken u p in water and the barium determined as BaSO,. T h e results were as follomrs : K,PtCI, K,PtCl, BaSO, (3) RaSO, (4)
(I) (2)
= 0.0192g. ( = 0.0014g. NH,). = 0.0445 g. ( = 0.0033 g. NH,). = 0.0348 g. ( = o . o o j j g. NH,). = o.oogo g. ( =2 0.0014g. XH,).
T h e foregoing experiments demonstrate positively that the radicle NH, in ammonium amalgam has the properties of a n alkali metal ; it behaves analogously t o potassium in potassium amalgam. I t remains to show that there is no evidence that the amalgam is a solution of free animonium in mercury, while there is every indication that it is a compound of ammonium with mercury, dissolved in an excess of mercury.
8%
GEORGE MCPHAIL SMITH
Rich and Travers' conclude, solely on the ground of their cryoscopic investigation of ammonium amalgam, that the substance is undoubtedly a solution of ammonium in mercury. Their results are given in modified form in the accompanying table. T.\BLE I. g.PI", in 100 gHg Solute
",
VI
Molecules iiH4i n loo atoms Hg.
os4
0.027
0.9297 0.6309 4.593 0.8744 0.298s
0.0094 0.0117
0.
('
O.Oj7
0.415 0.079
