102 IX.- On the Xpecijic Volume of ?Vuter of CyysLallistx.tion. By T. E. TEORPE F.R.S. and JOHN I. WATTS. PLAYFAIR and JOULE have pointed ont that the volumes of certain highly hydrated salts e.g. sodium carbonate with 10 mols. of water, and the alkaline arsenates and phosphates with 12 mols. are exactly equal to that of the water considered as ice which they respectively contain. The molecules of the salt proper would thus seem to exist in the interstitial spaces of the ice since they exert no appareut in-fluence on the bulk. I n many of these as for example borax with 5 mols. of water sodium pyro-phosphate and normal aluminium sulphate the volume seems to be made up of the water considered as ice together with that of the base as existing in the free state. Schiff some years since showed that the members of certain classes of hydrated salts have practically the same specific volume.Thus all the alums have a specific volume of about 277 ; double sulphates of the form M,.M"( SO,),SHi,O have a common volume of 207 ; arid all the vitriols that is the salts o l the form M"S04.7H20 whether isomor-phous or not have the specific volume 146. It is of course well known that many salts can unite with water in different proportions to form perfectly dcfinite combinations. Thus, according to various authorities i t is possible to obtain ferrous sulphate combined a t ordinary temperatures with 1 2 3 4 5 6 OF 7 mols. of water. Indeed the so-called magnesian sulphates give rise to a larger number of variously hydrated compounds of definite character than any other group of salts.It appeared to us of interest especially in view of the investigations of Playfair and Joule and of Schiff t o determine the precise relation of the specific volume of a salt to its degree of hydration ; and in this communication we beg to lay before the Society the results of our observations on such of the variously hydrated sulphates of copper, magnesium zinc nickel cobalt iron and manganese as we could obtain in a definite form. The investigation of this subject was undertaken many years ago by Dr. Playfair in continuation of the series of researches published by him in conjunction with Dr. Joule but although considerable progress was made in the work various circumstances prevented its completion. On learning that one of us was engaged on the question, Dr.Playfair kindly placed the very ample notes of his investigation a t In salts less highly hydrated this law does not hold good TBORPE ASD M'-jTTS ON THE SPECIFIC VOLUME ETC. 103 O u r disposal and although we have a t his suggestion gone over the whole of the ground again much of the experimental matter of the present communication is confirmed by his previous observations. OUY investigation was made partly in the laboratory of the Yorkshire College and partly in that of the Owens College under Professor Enwoe's supervision. We give whenever necessary the mode of preparation of the various hydrates ; the identity of these was in all cases ascertained by analysis usually by tlie estimation of the water.The determination of the specific gravity was effected by weighing in benzene and the finai rcsults are the means of several concordant determinations made by tile aid of different bottles. I n the case of the anhydrous or other liygroscopic salts the bottles were heated to the proper temperatures :ifher the introduction of the salt until its weight became constant. 111 filling t8he bottles tthe salt was placed in a small specimen tube tfhe iieck of which was closed by a perforated caoutchouc stopper through which was inserted the neck of the specific gravity bottle. By shaking tlie tube when in an inverted position the salt was made t o enter the bottle without undue exposure to air. The weight having been deter-mined the salt was covered with benzene the bottle was placed within the receiver of an air-pump and the exhaustion continued until the benzene boiled.The bottle was then filled up with benzene and placed in a water-bath havizg a constant temperature of 15". The level of the benzene was then adjusted to the mark and the bottle and its contents again weighed. The benzene had been crystal!ised, and boiled at about 80" and was thoroughlydried over sodium. Four determinations of its specific gravity at 15" compared with water at, tlie same ti:mpcrature gave-I 0.8856 I1 0.8860 I11 0.8859 IV 0.8357 Mean. . 0.8858 -I. Copper SuQihate. 1. Air hydrous Copper Sulphate CuS04.-Prepared by heating the Iwntahydrate to 280" until i t ceased to lose water. Three determinations of specific gravity gave-I 3.608 I1 .. . . . . . . . . 3.606 r n . . . . . . . . . . 3.60 Other observations on record are-o r Filhol . . . . . . . . . . . . t7.30 Joule and Playfair. . 3*G:31 Ksrsten 3.5 72 Playfair 3.560 2. iMonohydrated Copper Subhate CuSOJ.H,O.-PI.epttred by heating A Theory The specific gravity bottles were heated to 110' after the powdered pentahydrate to l l O o until it ceased to lose weight. determination of the amount of sulphur gave 18.04 per cent. requires 18.06. the introduction of the salt until the weight was constant. The determinations of specific gravity gave-1 . . . . . . . . . . 3-289 r1 . . . . . . . . . . 3.287 111 . . . . . . . . . . 3.289 Mean. . 3.289 PIayfair (communicated) 3.996. 3. Dihydrated Copper Xulphate CuS04.2H,0.-This hydrate which was first obtained by Playfair was prepared by pouring a cold satu-rated solution of copper sulphate into concentrated oil of vitriol with constant stirring and washing the precipitated salt with absolute alcohol until free from acid.A determination of water gave 18.24 per cent. Theoiy 18.45. --Two detci*minations of specific gravity gave-I 2.952 I1 . . . . . . . . . . 2.954 Mean 2.953 Playfair found that a similar prepara'tion had a specific gravity W E 2.891. He also obtained the dihydrate by boiling the finely powdereti pentahydrate with absolute alcohol. A determinatim of water gave 18.26 per cent ; its specific gravit'y was 2.858 (mean of two determi-nations). 4. Trihydratecl Copper Sulphate CuSOb.SH,O.-This salt whicl I appears not to have been made before was obtained by us by pouring a cold saturated solution of copper.snlphate into an equal volume OE sul-phuric acid of sp. gr. 1.7. The precipitate was washed with small quail-tities of absolute alcohol until free from uncombined salphuric acid. Analysis 1.3952 gram gave 0.349 gram water = 25.01 per cent. Calculated 25-34. 0.5913 gram gave 0.6568 gram Bas04 = 15-25 S per cent. Cal-culated 15.U2 Two citterniinations of specific gravity gave-I . . . . . . . . . . 2.Gii3 I1 . . . . . . . . . . 2.ci6,l A salt of the composition CuSO4.3&H,O l~eptl~en,hlldrated copper sulyhatr, was however obtained by placing the finely powdered pentahydrate over concentrated oil of vitriol u n t i l it ceased to lose weight even a f t e r prolonged exposure.An estimation of copper by standard potassium cyanide gas-e 28.60 per cent. Two determinations of specific gravity gave-All attempts to prepare the tetrahydrate were ansuccessful. Theory requires 28-40. I . . . . . . . . . . 964.5 II . . . . . . . . . . 2.645 Mean . . . . 2.645 A second series of observations on another preparation gave-I . . . . . . . . . . 2650 11 . . . . . . . . . . 2.652 n1ca11 . . . . i.(j,51 -A-This salt also appears not to have been obtained beforc. 5. Pe?ituIi~yd~cttsd C’opper X d p l r a t e CuSOi.5H,O. -Prepared hv repeatedly crystallising the commercial salt. 1,6523 gram pi- e 0-593‘3 gram water = 25-96 per cent. Theory 36.13. ‘1 wo detcrrriinations of specific gravity gave--7.-rij.> 1 .. . . . . . . . . - L-c -I1 . . . . . . . . . . - -CL 3lleaii 2.283 L ) . . ) I t --‘Yhe specific gravit1 of a second preparation which was found to give off 3ti.33 per ccnt. of witer was-C > . b l 9 & I . . . . . . . . . . - i( . I1 . . . . . . . . . . 2.288 Nem . . . . 2.086 _-.-0 t:ier observations on record are-Joule and Playfair 2.278 . . . . . . . . 2.2130 Filhol 2.286 Stolba . . . . . . . . . . . . . . . . . . 2.2i0 3 > > Kopp 2.27 Hydrate . . . . . . . . 0 1 2 Kolecular weight. 159.1 177.1 193.1 Specific vol. according to-T. and W. I'layfair . . . . . . . . $4.7 53.7 67.8 { ""'"> 66.1 . . . . . . 44.1 53.8 Joule and Playf'air 43.8 - -Ram t en - - . . . . . . . . 44.5 t'il iiol . . . . . . . . . . 44.9 - -Stolba - -I - Ropp - .. . . . . . . . . - . . . . . . . . . . I I. JIiy 9 )mix i ) ~ S 11 Iphu te. 1. Anhydrous Hagnesiurn Xulyliate MgS04.-PPreparecl by h e a h g The determinations of specific gravity gave -the heptahgdrate to 28a0 until it ceased to iose weight. I . . . . . . . . . . 2*i08 I1 . . . . . . . . . . 2.T10 Mean 2.709 -0 ther ohservers have found-Pilhol . . . . . . . . . . . . . . . . 2.628 Pape . . . . . . . . . . . . . . . . 2.675 Joule and Playfair . . . . 2.706 h~~lti11g the heptahydrate to 130-140" until it ceased to lose weight. 2. Monohydrated Xagizesium Bulplrnte MgSO4.H20.-Obtained by Two determinations of specific gravity gave-I 2.4-42 II . . . . . . . . . . 2.447 &lean . 2.445 -Playfair found that the monohydrate prepared in the same manner, and which lost 13.24 per cent.water (calculated 13.04) had a specific gravity of 2.478. (mean of two observations). Kieserite (MgSO,.H,O), according to Bischof has a specific gravity of 2,517. 3. Uih ydrated 31agwesiurn Sulphate Mg S 04.2H,0 .-Obtained by boiling the finely powdered heptahy,irate with ab3olute alcohol SPEOIFIC VOLUI\IE OF WATER OF CRPSTIILLISATIOS. 1t)'i 2.0175 grams lost 0.4735 gram water = 23.40 per cent. 23-08. Calculat,ed, Two determinations of specific gravity gave-I . . . . . . 2-374 2.372 I1 . . . . . . . . . . Mean . . 2.373 Playfair who also obtained this salt in the manner described above, and likewise br heating the heptahydrate to loo" found its specific gravity to be 2.279. 4.Penfaluydrated Magnesiz~~z~ Xulphate MgS04,5H,0.-This salt was obtained by Playfair by drFing the heptahydrate over strong sulphuric acid in air until it ceased to lose weight. It contained 43.05 per cent. of water. Its specific gravity (mean of two determinations) was 1.869. 5. Hex11 ydi-ated Magnes iwjn Sdphate Mg SO,. 6H,O .-Prepared by wystsllising a solution of the ordiiiary salt at above 40". 2.7717 gram lost 1.309 gram water = 47.2 per cent. Theory, 4 ; :i 7. The determinations of specific gravity gave-I1 Theory requires 42.8. I . . . . 1.734 1.734 Jlc:in. . . . 1.734 . . . . . . . . -Playfair who prepared the salt in the same manner found its specific 6. Hepttshydruted iiagnesiunz S d p h n t e MgS0,.7H20.-Prepared by 1.415 gram lost on heating 0.7255 gram gravity to be 1.751.rccrystallising Epsom salts. water = 51.2 per cent. Calculated 51.2. A determination of specific gravity gave 11676. A second preparation containing 51.23 per cent. of water gave t,lie I . . . . . . . . 1.678 II . . . . 1.678 Mean . . 1.678 following niimbers :--Other observations on record are-Hassenfratz . . . . . . . . . . 1-660 Kopp . . . . . . . . . . . . . . . . 1.674 Joule and PlnyEair . . . . 1.683 Schiff . . . . . . . . . . . . . . 1.685 Buignet . . . . . . . . . . . . . . 1.67 108 THORPE AXD WATTS ON THE Summary of Results qf Obseruations on Hydrated Mugnesiim Xulphate. Hydrate . . . . . . . . . . . . 0 1 2 5 6 7 Bxolecular weight . . . . 1.20 138 156 210 228 2446 Specific volume according to-T.zlllcl w. . . . . . . . . 44.3 56.4 E'ilhol . . . . . . . . . . . . 45.6 -Pape . . . . . . . . . . . . 44.9 -Joule and Playlair. . 44.4 -Playf air . . . . . . . . . . - 55.7 Rischof . . . . . . . . . . - 54.8 Kopp . . . . . . . . . . . . - -Schiff . . . . . . . . . . . . Hnigne t . . . . . . . . . . __ I - -146% { 146.6 - 146.2 130.2 -- -- 146.9 - 146.0 146.9 -Mean 44.8 55.6 67.0 112.4 130.8 146.6 111. Zinc Sdphate. 1. Atahydrtlus .%PIC Xulpliats ZnS04. - Obtained by heating the Ueterminations of specific gravity-lieptahydrate to 280-300" until it ceased to lose water. I . . . . . . . . . . 3ti29. 11 3-62 7 Mean,. 3.628 A second preparation gave the following numbers -I . . . . . . . . . . 3-621 I1 3.617 Mean 3.61'3 Other observations on record are :-Filhol .. . . . . . . . . . . . . . . 3.400 Playfnir . . . . . . . . . . . . . . 3.413 Joule and P1:Ljfaiim. . . . . . 3.681 Pape . . . . . . . . . . . . . . . . 3.435 2. Il.lonohyclrnted Zinc S i d p h t e ZiiS04.€f,0.-Prepaied by hcating tlie heptahydrate to 100-110" until it ceased to lose weight. 1.2433 gram lost 0.1275 gram water = 10.26 per cent. 1.3623 , 0.1395 , = 10.24 ,, Theory = 10.04 per cent Two determinations of specific gravity gave-I . . . . . . . . 3.283 3.278 I1 . . . . . . . . . . Mean . . 3.280 A second preparation afforded the following numbers :-1 . . . . . . 3.287 3.291 I1 . . . . . . . . hlean . . 3.289 A determination of specific gravity made by Playfair on a salt con-taining 10.08 per cent.of water gave 3.259. 3. Diliydrated Zinc XirZphate ZnS04.2H20.-Obtained by pouring a cold saturated solution of zinc sulphnte into strong sulphnric acid, and washing the precipitate with absolute alcohol until free from un-combined acid. 1.371 gram salt gave 0.2495 gram water = 18.1 per cent. Theory = 18.2 ,, Two determinations of specific gravity gave-I . . . . . . 2.958 2.957 11 . . . . . . . . Mean . . 2.958 4. Penta7qdrated Zinc Xukhatc ZnS04.5B,0.-This salt was ob-tained by boiling the finely powdered heptahydrate with alcohol of sp. gr. 0.825 (Playfair) ; 7.174 grains lost 2.549 grains water or Xi.5 per cent. Theory 36.0. Two determinations of specific gravity gave-I I11 . . . . . . . . . . 2.197 I1 . - . . . . .. 2.205 2.217 . . . . . . . . . . Mean . . 2.206 5. Bexhydrated Zinc Su@hatr? ZnS0,.6H20. - Obtained by crys-Determinations of specific gravit.y-tallising a solution of zinc sulphate at about 40". I . . . . . . . . 2.071 2.075 I1 . . . . . . . . . . Mean . . 2.07 1 10 THORPE AXD WATTS ON THE Playfair also obtained this salt by crystallising a solution a t about 10.35 grains gave 4.175 grains water or 40.30 per cent. Theory Its specific gravity was found to be 2.056. 6. Beptahydrnted Zinc Sidphate ZnSOp.7Hz0. - Ordinary white vitriol after repeated crystallisation contained the following amounts of water :-30". requires 40.29. 1.6963 gram gave 0.7441 gram water or 43.87 per cent. 1.5160 ) 0.6648 ) 7 43.85 7 7 Theory 43-87' per cent. Its sp citic gravity was found to be-I .. . . . . . . . . 1.963 I1 . . . . . . . . . . 1.963 A second sample gave the following numbers :-I . . . . . . . . . . 1.ncx TI . . . . . . . . . . 1.965 Other observations on record are :-Joule and Playfair. . Schiff . . . . . . . . . . . . 1.953 Stolba . . . . . . . . . . . . 1.9534 Holker 1.976 1,931 (mean of 4 observations). Buignet . . . . . . . . . . 1.95 7 Sunarnary of Results of Observations o n Hgdrated Zinc Sulphates. Hydrate . . . . . . . . . . 0 1 2 5 6 7 Molecularweight . . 161 179 197 251 269 287 Specific volume according to-- - Filhol . . . . . . . . . . 47.2 -Playfair . . . . . . . . 47.1 55-33 - 113.7 Jonle and Playfair. 43.7 - - -Pape . . . . . . . . . . . . 46.8 -Buignet . . . . . . . . .. - -- - - - Schiff . . . . . . . . . . . . S to1 ba . . . . . . . . . . Holker . . . . . . . . . . - - - -- - - -- - - -130.8 148.G - 148.5 - 146.9 - 146.6 - 146-9 - 145.3 - -Mean . . . . 45.6 54.7 66.6 113.7 130.2 146. SPECIFIC VOLUME OF WATER OF CRfYThLLISAITION. 11 I ITr. K i c k e l Xltlp7rde. 1. Adlyclmus Nicliel Sdphnte NiSO1. - Prepared by heating 3 weighed quantity of the heptahydrated salt a t 250° until it ceased to lose weight. Calculated loss = 1.33 gmms. 2.964 grams lost 1.330 grams H,O. Two determinations of specific gravity gave :-I . . . . . . . . . . 3.419 I1 . . . . . . . . . . 3.417 Mean . . 3.418 -. -Playfair obtained 3,526. 2. Ahnohydrated Nickel Sdpliate NiSO4.H20.-Prepared by heating 0.3585 gram lost 0.03CiO gram H20 = 10.03 per cent.Theory, 3. Hezhydrated Nickel Sdphate NiS0,.6H20.-Prepared by crys-0.4682 gram lost on heating 0.191 gram = 40.8 per cent. Theory, A determination of specific gravity gave 2.031. TopsoB obtained values varying from 2.042 to 2.074. 4. Heptahydrated Nickel Sulphafe NiSOd.7H20. - So-called pure nickel snlphate of commerce was freed from cobalt by treatment with chlorine adding barium carbonate in excess filtering precipitating the nickel in the filtrate with potash and after well washing dis-solving the oxide in sulphuric acid and recrystsllising. Theory, 44.9. the heptahydrate at loo" until its weight was constant. 10.40. tallising a solution of nickel sulphate at a temperature above 40". 41.2. 2.964 gram lost 1.33 gram on heating = 44.9 per cent.Two determinations of specific gravity gave :-I . . . . . . . . . . 1.949 I1 . . . . . . . . . . 1.948 Mean 1.949 -A determination on a second preparation gave 1.945. Schiff observed 1.931 112 THORPE AND WATTS ON THE S u n m a y of Results of Observations on Hydrated Nickel Sulplzate. Hydrate 0 1 6 7 Molecular weight 154.7 172.7 262.7 280.7 Specific volume according to-T. and W. 4rj.2 56.6 - { :::: Playfair . . . . . . . . . . 43.9 56.4 129.3 -- 128.6 - TopsoS -- 145.4 Schiff - -Mean 44.6 56.5 129.0 144.6 V. Cobalt Sulpkate. 1. Anhydrous Cobalt Sulphate CoS04. -Obtained by drying the Eieptahydrate at 250". A determination of specific gravity gave 3.472. Playfair obtained 3.444. 2. Monohydrated Cobalt Sulphate CaS04.H,0.-Prepared by heating the heptahydrate a t 100" until it ceased to lose weight.Its specific gravity was found to be 3.125. 3. Bihydmted Cobalt Sulpha,te CoS04.2H,0.-Prepared by boiling the finely powdered heptahydrate with absolute alcohol. 0.229 gram lost 0-0435 gram water = 18.9 per cent.; calculated 18.9. Playfair also obtained the salt in the same manner and found the specific gravity to be 2.712. 4. Tetrahydrated Cobalt Xulpphnte CoSO4.4H2O.-Obtained by ex-posing the finely powdered heptahydrate over oil of vitriol until its weight was constant. 5. Pendahydrated Cobalt S.dphate CoS04.5H20.-This hydrate was obtained by Playfair by drying the heptahydrated salt over sulphuric acid. 3.475 grams gave 1.290 gram water = 37.1 per cent. Calculated, 36.7.6. Hex hydrated Cobalt Suly hnte Co S 04. 6 H,O .-Prepared by cr ys-tallising a solution of cobalt sulphate a t about 25"-0.5400 gram lost 0.2190 gram H20 = 40.6. Theory 41.2 The specific gravity was found to be 2.019. 7. Reptahydraied Cobalt Sulyhate CoSOd.7H20.-Prepared by re-crystallising pure cobalt sulphate. The salt was freed from nickcl by treatment with chlorine and barium carbonate solution in acid repre-cipitation with potash and re-solution in sulphuric acid. Schiff observed 1.924. The specific gravity was 2.668. Sp. gr. 2.327. A determination of specific gravity gave 2.134. A determination of specific gravity gave 1.918 SPECIFIC VOLUME OF WATER OF CRYSTALLISATION. 1 13 Summary of Eesdts of Observations o n Cob& Szclphate.Molecular weight. 154.7 172.7 190.7 226.7 244.7 262.7 280.7 Hydrate 0 1 2 4 5 6 7 Specific volume according to-T. and W. . . . . . . 44.5 5.5.2 71.5 97.4 - 130-1 146.4 - - - - 145.5 Schiff Playfair . . . . . . . . 44.9 - 70.3 - 114.6 - -- I ~ Mean 44.7 55.2 70.9 97.4 114.6 130.1 146.0 VI. Nangnnous Sidplmte. 1. Anhydrous Mangmous Sz@hate MnS04.-Obtained by heating the pentahydrate to 280" until it ceased to lose weight. Its specific gravity was found to be 3.282. Playfair observed 3.386. 2 . ,VorLohycl.r.ated Maiiganous Xulplza3te blnS04.H20.-Prepared by drying the pentahydrate at 100" until no further loss of weight mas observed. Theory 10.74. A determination of specific gravity gave 2,845. Playfair ob-served 3.210. 3. D ihyldrated Man g itn ozcs S d p ha t e S 04.2H20 .-0 btained b JT boiling t'he finely powdered pentahydrate with absolute alcohol. This hydra+e may also be prepared by pouring a satuyatecl solution of the pentaliydrate into oil of vitriol. Theory 19.35. 1.1.315 gram lost 0.1250 gram HzO = 11.04 per cent. 1.183 gram lost 0.2275 gram water = 19.20 per cent. Two determinations of specific gravity gave :-I 2.526 IT 2-525 Mean 2.526 4. T d u ~ dr ated JIan gal ious Szc lp h ate Mn S 0,. 3H,O. -0 bt ained by Playfair by evaporating a solution of the sulphate a t a boiling heat until a pellicle formed on the Surface removing this and drying it rapidly between hot filter-paper. 5.875 grains gave 1.535 grains H,O = 26.1 per cent. Theory 26.2. Sp. gr. '2.356. 5. Tet rak y clyated AIa tapzotLs Sui$ h i t te Jf nS 04.4H20.-This hydrate, according to TopsoG has a specific gravity of 2.261, 6. Pentahy drat ed Mmgm OZLS S idp A CI te Mn SO4 :5H20. - 0 b tained by repeatedly crystallising the ordinary sulphate. Two concordant deter-VOL. XXXVIT. 114 TEIORPE AXD KATTS OK THE minations of specific gravit,y gave 2.103. 2.087. Kopp obtained 2.095 and 8um.mary of Results of Ohservat ions on Manylxnous Xulpliates. Hydrate 0 1 2 3 4 5 Molecular weight 150 168 186 204 222 240 Specific volume according to-T. and W. 45.7 59.1 73.6 - - 114.1 Playfair 44.3 52.3 - 86.6 -Tops05 --95.2 - - -- - - - Kopp Mean 45.0 55.7 73.6 86.6 98.2 114.4 VII . F e w o us Xu lp hate. 1. Anlydrous Ferrous Sulp!iate FeS04.-Obtained by heating the powdered heptahydrate in a current of hydrogen.Analysis showed that i t was anhydrous. Its specific gravity was 3.346. Playfair observed 5-48. 2. A l ~ n o h ~ ~ d r a t e d Ferrous Xup7zate FeSOl.H,O.-Prepared b y heating the powdered heptahydrate in a current of hydrogen a t 120". An estimation of iron by potassium permanganate solution gave 32.6 per cent. Theory 32.9. Sp. gr. 2.994. Playfail. found 3.047. 3. Di7iydmted Ferrous Sulpyhafe FeS04.2H,0.-Obt8ained by boiling the heptahydrate with successive quantities of absolute alcohol. Sp. gr. 2.773. 4. T e t rahydi-ated Ferrous Sulyli ate Fe S 04.4H,0 .-OW ained by ex-posing the finely powdered heptahydrate over oil of vibriol in an atmosphere of carbon dioxide until it ceased to lose weight. An estimation of iron by standard potassium permanganate solut,ion gave 25.3 per cent.Calculated 25.0. Sp. gr. 2.227. 5 . Heptahydrated Perrous Sulphate FeSOp.7H20.-It has been thought unnecessary to add to the determinations of the specific gravitS of this salt already on record. Joule and Playfair 1.889 Buignet 1.902 Filhol 1.904 Schiff 1.88 SPECIFIC VOLUME OF JTATER OF CRYSTALLISATION. 11 5 130.8 130'2 199.0 130.1 - --Xumnzary of Results of 0bsermtion.c 011 Fewous Xulphntes. Molecularweight. 152 170 183 224 2 i 8 Hydrate 0 1 2 4 7 - -146.6 146.8 144% 146.0 146 -7 Specific volume according to-44.4 44.8 45.6 44.6 44.7 45.0 44.5 44-8 T. and W. 45.4 56.7 67.7 100.5 -Playfair 43.6 55.7 -Joule and Playfair - - -Buignet I - -- -- 147.2 - 146.0 - 147.5 - 146-1 - I - Filhol Schiff - - -54.3 55.6 54 7 56.5 55'2 55.7 56.2 55-5 -___ Mean,.. 44.5 56.2 67.7 100.5 146.7 On comparing the foregoing observations ,it is evident that the anhydrous sulphates of copper magnesium zinc nickel cobalt man-ganese and iron have identical specific volumes ; or since we may define specific volume as the volume in cubic centimetres occupied by the equivalent of the salt in grams it follows that equivalent quantities of these different sulphates occupy the same bulk in space. A like con-clusion as regards certain of these anhydrous sulphates has already been drawn by Playfair and Joule. A further consideration of the numbers warrants us we believe in supposing that this conclusion may be extended to the various hy-drates of these salts.The evidence on which this assumption is grounded is contained in the following table which is merely a synopsis of the experimental results detailed in the foregoing pages :-Hydrate, Copper sulphate Magnesium sulphate . Zinc sulphate Nickel sulphate . Cobalt sulphate . Manganous sulphate . Ferrous sulphate Mean of means I -2 67 '0 67 '0 66.6 70 .Y 73 -6 67.7 --5 109 -1 112 '4 113.7 114 *G 114.4 ---112 -9 -Some of these numbers are not so precise as we could wish it would have been more satisfactory for example to have obtained a larger number of observations on the trilsydratcs. But these interme-diate hydrates are obtained with great difficulty owing in great part to the indefinite character of the methods of preparation ; indecd b 116 THORPE AND WATTS OK T6E far the greater portion of the time occupied in what has proved to bc a very tedious research was spent in attempts in many cases fntile, to procure them in a state of purity.An examination of our experi-mental data will make it clear that the disparities between t)he values obtained for the different members of each series are almost entirely due to accidental variations in the degree of hydration of the salts, and not to errors in the determinations of the specific gravities ; errors of this kind such as are likely to occur would exercise only a com-paratively small influence on the result. That the disparities are actually accidental would seem to be proved by the factl that no r e p -layity or symptom of order can be discovered in the mode of the variation.We think therefore we are justified in concluding that what Playfair and Joule have shown to be true of certain of the anhjdrons sulphates of the form M"SO4 and what Schiff has shown to hold good in the case of the hydrated sulphates of the form ;RJI"SO,. '?H,O-and their results are confirmed by our own independent observations-is equally applicable to the case of the intermediate hydrates. The final mean then of each series will afford us an ap-proximation to the true specific volume of the particular hydrate arid enable us therefore to trace the influence of the varying degree of hydrat'ion on the bulk of the salt.The first and main conclusion we draw from our observations is that in the case of the so-called magnesian sulphntes the volume occupied by the several molecules of water varies with the degree of hydration. The first molecule of water the constitutional water or " water of halhydration '' of Graham occupies considerably less bulk than the remaining molecules ; its mean relative value is 10.7. Each additional molecule appears to occupy a gradually increasing volume. The difference between the monohydrate and dihydrate is 13.3 ; be-tween the dihydrate and trihydrate it is 14.5 ; between the trihydrate and tetrahydrate it is 15.4. There is a break in the continuity of the rate between the tetrahydrate and pentahydrate due to the low value found for copper sulphate.The difference between the hexhydrate and heptahydrate is 16.2. We give these numbers merely as first approximations since a far larger number of observations made upon hydrates of definite composition will be required to fix the exact values. We would point out however that our main conclusion is in harmony with experience. The general fact that the different molecules of water in a hydrat,ed salt' are held with varying degrees of tenacity as shown by the difference in the intensity of heat needed to expel them, is of course well known. Now as has already been pointed out by Muller chemical combination is generally attended by contraction and the stability of the resulting compound corresponds with if it is not dependent on the diminution of bulk. It would be a problem of con SPECIFIC VOLUME OF WATER OF CRTSTALLISXTION.11 'i siderable interest in the light of these results to determine tlre different quantities of heat evolved in the combination of successivc molecules of water with these anhydrous sulphates. Graham nearly forty years since made the beginning of such a research. From com-parative experiments made with the anhydrous sulphates of manganese, copper zinc and magnesium and the monohydrates of these salts, Graham found that far more heat was developed in the act of com-bination of the first molecule-that in which the greatest contraction occurs-than is evolved in the combination of any of the remaining molecules; the amount in the case of the first molecule mas from a fourth to a third of that needed for the complete hydration of the salt, These observations have been greatlp extended by Favre and Valson (Compt.rend. 73,1144; 74 2016 1065 ; 75 798 925 1000 1066, 1071) who have determined the amount of heat produced or abstracted by the solution of sulphates of various degrees of hydration in water. The observations were made on a number of the sulphates mentioned in the foregoing pages on certain double sulphates and on the alums. The results so far as they are applicable to the salts under considera-tion are as follows :*-cuso CuSO,.HO CuS045HO MgSOa MgSO4.HO MgS 0,7H 0 . . . . . . . . . . . . . . . . ZnSO ZnS04.H0 . . . . . . . . . . . . . . . . ZnS047H0 NiS047H0 CoS047HO lSiInSOa MnS04H0. . Mn S 0 ,5 H 0 FeS047H0 Heat-units. 8149 4734 - 1216 10152 5493 - 1860 9289 481 2 - 20'74 - 1944 - 1680 7085 4216 235 - 2182 These numbers which express the amount of hcat evolved by the solution of 1 equiv. of the salt in grams in 1 litre of water are not exactly comparable with those of Graham but they serve to confirm the general conclusion that the heat due to the combination of the first molecule of water is much greater than that evolved in the com-bination of the succeeding molecules. * 0 = 8 S = 16