7o8 SCIENCE PROGRESS 



\ 

 In the table on the opposite page a series of examples are | 

 given showing that very frequently the relationship between 

 allied substances, if not at once apparent in the parameter 

 values deduced directly from the axial ratios, becomes so after 

 but very simple modification, such as, e.g. the interchange of 

 X and y in the three sulphonamides. All the compounds 

 included in the table on page 709 exhibit pseudo-cubic mar- 

 shalling and may be regarded as "derived from rhombohedral 

 benzene. 



Many of the data from which the above results have been 

 deduced had been known for a long time but publication had 

 been withheld as there appeared to be no point in publishing 

 mere figures. But now, since the advent of the new theory, 

 we are beginning to see how the data which have been 

 accumulated may be applied to the solution of the problem of 

 correlating crystalline form and chemical constitution. One 

 thing is certain from the above results, namely that crystallo- 

 graphically hydrogen and the halogens are approximately 

 equivalent. But not precisely so ; and it is hoped that when 

 we see our way more clearly it will be possible to determine 

 the differences in volume between the spheres of atomic influence 

 of hydrogen, chlorine, bromine and iodine which cause the 

 small differences observed between the equivalence parameters 

 of such related substances as those given in the above tables. 

 These differences, be it noted, though small, are beyond the 

 limits of experimental error. Barlow and Pope have already 

 concluded from a study of the crystallographic constants of 

 the trihalogen compounds of the alkali metals that iodine 

 possesses a larger sphere of atomic influence than either 

 bromine or chlorine. However, the mathematical problems 

 involved in the determination of the magnitudes of these 

 diff'erences would appear to be of a high order of difficulty. 



The benzenesulphonic chlorides and bromides which have 

 been examined present many cases of polymorphism. In a 

 few cases two forms of one substance have been measured 

 and the results obtained for these bear out the statement 

 made during the discussion of silver iodide that the different 

 crystallographic modifications of a substance are probably 

 very intimately related. In the two following cases it can be 

 seen that there is very little diff'erence between the equivalence 

 parameters of the two forms ol each substance and it can 



