I3 8 SCIENCE PROGRESS 



molecule of water ; the compound must therefore be excessively 

 unstable, and it is hardly to be wondered at that the formulation 

 of such a complex has come to be regarded almost as a reductio 

 ad absiirdwn of the hydrate theory. Similar considerations have 

 led to the non-recognition of hydrates with 51, 131, 508, and 

 1,600 molecules of water; the prevailing feeling appears to be 

 that such compounds, even if present in the solution, would be 

 so largely dissociated as to be incapable of producing any 

 abrupt change in the properties of the liquid ; the small 

 irregularities on the strength of which these hydrates have 

 been postulated appear to have been regarded as due to ex- 

 perimental errors, the effects of which would be most pronounced 

 in dilute solutions. 



Similar conclusions follow from a consideration of the ex- 

 perimental data. The best justification of Mendeleef s method 

 of analysis is to found in the study of the simple hydrates 

 S0 3 .£HO, S0 3 .H 2 0, and S0 3 . 2H0O, whilst its limitations are 

 most clearly seen in the case of the more dilute solutions. Of the 

 above compounds the two former are so stable that they are not 

 usually considered as hydrates at all, but are referred to as 

 *■ pyrosulphuric " and " sulphuric " acids respectively, although 

 the acid containing 100 per cent. H 2 S0 4 is known commercially 

 as " monohydrate." The dihydrate of sulphur trioxide is less 

 stable, but produces such marked effects in the solutions that its 

 existence is scarcely a matter of controversy. 



The conductivity curves are those in which the presence of 

 the three hydrates is most clearly indicated. Water and sulphur 

 trioxide are both excellent insulators when pure, but a mixture 

 of one part of sulphur trioxide with four parts of water is one of 

 the best electrolytic conductors known, the specific conductivity 

 rising at this concentration to 07388 reciprocal ohms per centi- 

 metre cube at 18° C. The curve is, however, smoothly rounded, 

 and since in the case of electrolytes it is usual for mixtures to 

 show a higher conductivity than pure compounds, there is no 

 reason for attributing the maximum to the formation of a 

 hydrate, although the concentration at which it occurs would agree 

 very fairly with that of a hydrate of the formula SO s i8H 2 0. 



In the case of the converse mixture of 4 parts of sulphur 

 trioxide with 1 part of water precisely opposite conditions 

 prevail. The conductivity falls at 81*44 per cent. S0 3 to 

 0*0080 ohm cm. -1 — a minimum value which is barely 1 per 



