204 Prof. H. E. Armstrong and Mr. E. H. Podd .
[ June 6 , ( 1 ) The necessity for duly fatiguing the balance beam prior to effecting any final weighing .
( 2 ) The ensuring , by means of an inner and auxiliary case , of a practically perfect uniform temperature throughout the beam of the balance .
( 3 ) The extinction of air streams in the immediate vicinity of the balance pans .
( 4 ) The plan for continuously supplying very dry and carbon-dioxide-free air to the interior of the balance case , with the consequent removal of water skins from any objects that are to be weighed .
( 5 ) The employment of a tilt indicator whereby the horizontality of the balance shelf can , at any time , be almost instantly tested .
Finally , we may state that Landolt found the mean apparent change in mass during chemical reaction to be not greater than 1 in 1 x 107 .
With the aid of our additional devices and precautions we show that in one very fully investigated case in which barium chloride and sodium sulphate were the reacting bodies , the apparent change in mass was not greater than 1 in 1 x 108 .
Morphological Studies of Benzene Derivatives.\#151 ; III .
Para-dibromobenzenesulphonates ( .
) of the " Rare Earth " Elements\#151 ; a means of determining the Directions of Valency in Tervalent Elements .
By Henry E. Armstrong , F.R.S. , and E. H. Rodd , B.Sc. ( Received June 6 , \#151 ; Read June 27 , 1912 .
) In a communication on the " Classification of the Elements " brought under the notice of the Society by one of us in 1902 , * a method of ordering the elements was advocated which was different , in essential respects , from that we owe to Mendeleeff , more especially in that it involved the systematic application of a principle dominant in organic chemistry\#151 ; the principle of homology\#151 ; in a manner and to an extent not generally recognised as necessary at that time .
The subject has been further discussed in a recent lecture to the Royal Philosophical Society of Glasgow , f various issues being considered which have been brought into prominence in the meantime\#151 ; e.g. the atomic * 'Roy .
Soc. Proc. , ' 1902 , vol. 70 , p. 86 .
t 'Glasgow Roy .
Phil. Soc. Proc. , ' 1912 ; 'Science Progress , ' 1912 , No. 24 .
I Morphological Studies of Benzene Derivatives .
205 .
1H 2 3 4 5 6 7L ; 8 9Be 10 It B 1* C 13 14 N 15 16 o'* !
17 18 19 F 20 *1 *2 *3 Na 24M\#171 ; 25 *6 27 Au *8 Si 29 30 31 P 32 S 33 34 35 Ol 36 37 38 39 K 40 Ca 41 4* 43 Sc 44 48 Ti 49 50 51 V 52 Cn 53 54 55 Mn 56 Ft 60 Cl 62 63 C\#187 ; 65 Zn 66 67 68 69 Ca 70 72 Ge 73 74 75 As 76 77 78 79 80 Bn 8 !
78 8* 83 84 85 Ra 86 79 Sc 87Sn 88 89 90 Yr 91 Zn 02 93 94 96 Na 96 Mo 97 98 99 100 10* Rw 103 Rh 105 106 107 108 Ac 109 ( 107 Po ifaco 113 114 115 In 116 119 Sn 120 1*1 Sb 122 123 124 1*5 1*6 1*7 I 1*8 = 130 131 13* 133 Cs 134 ) 128 Tc 137 8a j i 139 La 140 Cc 141 Pn 144 No 5 llll 150 Sm 15* Eu 157 Go 159 Tb I6*DY = 186 188 168 Er 169 Tu 170 171 17* 173 Ya 174 180 i 181 182 183 Ta 184 W 185 187 19 ] Os 1 193 1a 194 Pt 195 196 197Au 198 *00 Ha *01 *0* *03 204Tl *05 *1* *13 *14 215 207 Pb 208 Bi 209 *10 211 E 220 a*i a** **3 *24 2X6 Ra *28 229 *30 *31 232"Th *34 *35 *36 *37 *38 U 206 Prof. H. E. Armstrong and Mr. E. H. Rod .
[ June 6 , weights of argon and the allied inert elements , the nature of the radioactive , elements and the position of tellurium in the scheme of elements .
An amended table of the elements put forward in the lecture referred to is reproduced on p. 205 .
In this table , most of the elements derived from the so-called rare earths are arranged in a vertical series , as members of a single family , in a manner corresponding with that adopted in the case of homologous organic compounds .
Thus arranged they form the most conspicuous group in the table .
If the plan put forward be in any way permissible , it is obvious that the record of the elements must be far from complete , especially in the family of the rare earths : it is noteworthy that in the case of this family , since 1902 , doubt has been cast upon the existence of several reputed members of the group whilst others have been definitely foreshadowed in their place .
In any case , the group is one in which , more than in almost any other , it is likely that opportunities will occur of " proving " the nature of the relationship between closely allied elements .
It is therefore desirable that the methods of separating the rare earths should be perfected , also that the exact manner in which properties alter throughout the series should be ascertained .
At present our knowledge of the group is remarkably defective in many particulars .
When work was commenced , soon after 1902 , few salts had been examined other than those derived from the inorganic acids and a few simple organic acids ; it was therefore to be expected that many new salts characterised by desirable properties might be found .
As sulphonic acids have the advantage of being strong acids and it is possible to find among them acids giving salts of every degree of solubility , it was decided to make systematic use of such acids .
It was also decided to study the behaviour of the salts at various temperatures .
It is customary in separating substances by fractional crystallisation to allow more or less concentrated solutions to cool to the atmospheric temperature , so that crystallisation takes place finally at this temperature .
Bearing in mind the fact that hydrated salts often undergo changes in the degree of hydration on heating , it is to be expected that , by effecting crystallisation at definite temperatures above the transition temperature , it may be possible , in some cases , to establish differences greater than those the salts present at ordinary temperatures and , in this way , to effect separations otherwise difficult to secure .
For various reasons one of the acids first used was paradibromobenzene-sulphonic acid .
The didymium salt prepared from this acid separated from a hot concentrated solution , when this was allowed to cool to atmospheric 1912 .
] Morphological Studies of Benzene Derivatives .
temperature , in ill defined , faintly coloured , striated prisms containing a high proportion of water ; when , however , the magma was maintained at about 37 ' during a day or so , the pale striated crystals were converted into more deeply coloured , massive , beautifully defined prisms of a less hydrated salt .
The inquiry was laid aside until recently on account of the pressure of other work and especially because of the difficulty of obtaining the various earths in a purified state .
Meanwhile Holmberg* has published an account of a number of sulphonates of rare earth elements from which it is clear that sulphonic acids are likely to be of special value in separating these elements .
It is pointed out in the first of these studiesf that an inquiry , commenced about 20 years ago by one of us , carried on in the hope that it would be possible to correlate crystalline form eventually with internal molecular structure had been successful in that it was possible to discuss the data from the point of view of the remarkable and comprehensive theory introduced by Barlow and Pope .
It was shown in the second communication of the series* that the measurements recorded of a large number of derivatives of benzenesulphonie acid could all be interpreted with the aid of the theory , clear proof being given that it was possible to demonstrate the existence of a benzene framework in the molecules of the compounds examined .
In view of this result , it was desirable that metallic salts of the acids studied should be examined crystallographically : the crystallographic character of the framework of the acid being known , it was not improbable that it would be possible to draw useful conclusions from the crystalline form as to the disposition of the metal and also of water of crystallisation in the salt .
We have therefore resumed the study of sulphonates of the rare earth metals for the reasons given above and also because of the special opportunity these elements afford of investigating the relationship between a series of closely allied metals which , on various grounds , are of particular interest .
he results we have obtained are noteworthy in several respects .
Lanthanum , neodymium , praseodymium and cerium form paradibromo-benzenesulphonates of the formula ( C6H3Br3-S03)a(La , Nd , Pr , Ce),18Ha0 when crystallised from water at ordinary room temperatures .
The crystals are monosymmetric but unfortunately they are not measure-able : apparently , the four salts are isomorphous .
At 35'\#151 ; 37 ' they pass * 'Chem .
Soc. Journ. , ' 1907 , Abs .
II , p. 90 .
t 'Chem .
Soc. Trans./ 1910 , vol. 97 , pp. 1580\#151 ; 1584 .
X 'Chem .
Soc. Trans./ 1910 , vol. 97 , pp. 1585\#151 ; 1605 .
Prof. H. E. Armstrong and Mr. E. H. Rod .
[ June 6 , over into less hydrated salts containing nine molecular proportions of water , which crystallise in very beautiful orthorhombic prisms ; these four salts are closely isomorphous , the cerium salt , as might be expected , differing more from the others than these do among themselves:\#151 ; a. b. c. La ... ... ... ... 1-3965 : 1 : 0-8753 m ... ... ... ... .
1-3990 : 1 : 0-8789 Pr ... ... ... ... 1-3964 : 1 : 0*8798 Ce ... ... ... ... 1-4106 : 1 : 0-8873 In the case of samarium , only the hydrate containing 18H20 has been obtained as yet ; this apparently is isomorphous with the corresponding salts of the other four metals .
Judging from the few not very satisfactory measurements that have been made , the crystals are monosymmetric , the dimensions indicating a close approximation to orthorhombic symmetry .
Two hydrates of the gadolinium salt have been obtained : the higher of these contains only 12H20 , the lower the abnormal proportion 7H20.* Both hydrates crystallise out from the solution together at 37 ' ; at 50 ' only the lower hydrate is obtained .
It is therefore obvious that as the atomic weight of the metal increases peculiarities become manifest which are not noticeable in the case of the lower terms of the series .
We have not had sufficient material at our disposal to complete the determination of the solubility relationships of the various salts but have ascertained , in the case of neo- and praseo-dymium , that it is possible to extend the solubility curve of each hydrate many degrees beyond the transition temperature ; on this account the conversion of the higher into the well defined lower hydrate can be effected within a reasonable time only by raising the temperature considerably above the transition point .
This subject will be discussed fully in a later communication .
We are now studying salts formed from a considerable number of sulphonic acids in order to ascertain the manner in which degree of hydration at various temperatures and crystalline form are influenced by variations of the structure of the acid .
We may mention here that the paradiiodobenzene-sulphonates are very sparingly soluble salts .
Crystallographic Data .
The salts examined were prepared usually by neutralising the acid with the sesquioxide ; when praseodymia containing peroxide is dissolved , * Holmberg has obtained samarium and gadolinium metanitrobenzenesulphonates containing 7HsO ; the majority of the benzenesulphonates he prepared contained either 9H20 or 6H20 .
1912 .
] Morphological Studies of Benzene Derivatives .
209 oxygen is evolved .
The only case in which we experienced any difficulty was in preparing the cerium salt from the dioxide ; this is slowly dissolved and reduced when digested with the sulphonic acid and hydrogen peroxide but the method is unsatisfactory and the cerous salt is perhaps contaminated with ceric salt .
Neodymium Paradibromobenzenesulphonate , Nd(C6H3Br2S03)3,18H20.\#151 ; This hydrate is formed when a hot saturated solution of the salt is cooled or when a solution is allowed to evaporate spontaneously .
It separates in bunches of pale pink prisms which have not yet been obtained in form suitable for goniometric examination .
When viewed under ' the polarising microscope , the prisms show extinctions parallel with their length and they appear to be monosymmetrie .
Water : Found 22*41 ; calculated 22*93 per cent. Nd(C\#171 ; H3Br2S03)3,9H20.\#151 ; When a mass of crystals of the above hydrate is maintained in contact with the saturated solution at about 37 ' C. , the ill-defined forms change into massive amethystine prisms such as are depicted in fig. 1 .
The water of crystallisation is lost below 180 ' 0 .
Water : Found 12*58 ; calculated 12*95 per cent. The prisms belong to the orthorhombic system .
The forms observed were { 100 } , { 301 } , { 102 } , { 110 } , { 120 } and { 010 } .
The faces in the zone [ 100\#151 ; 010 ] are badly developed usually and excepting those of { 100 } are much striated in the direction of the c axis .
The form { 120 } has sometimes given very good reflections , but { 010 } is always bad .
The crystals generally grow on the form { 010 } , sometimes on { 100 } .
The cleavage is perfect parallel to { 100 } but indifferent in the direction perpendicular to the c axis .
Axial ratios : a:b:c \#151 ; 1*3990:1 : 0*8789 .
Prof. H. E. Armstrong and Mr. E. H. Rod .
[ June 6 , The following angular measurements were made:\#151 ; Angle .
No. of observations .
Limits .
Mean observed .
Calculated .
100 : 301 25 46 ' 11'\#151 ; 47 ' 0 ' % 42 ' 301 : : 102 25 25 10\#151 ; 26 1 25 41 25 ' 52 ' 102 : : 102 10 34 51 \#151 ; 35 18 35 7 34 52 100 : : 110 15 53 55\#151 ; 54 58 54 24 54 27 110 : : 120 15 15 44\#151 ; 16 3 15 55 15 53 100 : : 120 17 70 13 \#151 ; 70 30 70 20 \#151 ; 120 : : 120 9 39 17\#151 ; 39 24 39 21 39 20 120 : : 301 12 76 24\#151 ; 76 57 76 39 76 39 301 : : 120 12 102 55 \#151 ; 103 35 103 20 103 21 010 : : 301 2 89 57\#151 ; 90 0 89 58* 90 0 301 : : 010 2 90 1\#151 ; 90 3 90 2 90 * 0 Praseodymium Paradibromobenzenesxdphonate , Pr(C6H3Br2S03)3,18H20.\#151 ; This salt corresponds in all respects with that of neodymium , excepting that it has a pale green colour .
Water : Found 22*68 ; calculated 22*93 per cent. Pr(C6H3Br2S03)3,9H20.\#151 ; Except that it has a pronounced green colour , this hydrate is exactly similar to the corresponding hydrate of the neodymium salt .
System : orthorhombic .
Forms observed:\#151 ; { 100 } , { 301 } , { 102 } , { 110 } , { 120 } and { 010 } .
Water : Found 12*65 ; calculated 12*98 per cent. Axial ratios : a:b:c ==1*3964:1 : 0*8798 .
The following angular measurements were made :\#151 ; Angle .
No. of observations .
Limits .
Mean observed .
1 Calculated .
100 : 301 32 46\#174 ; 10'\#151 ; 47 ' 5 ' 46 ' 37 ' r 301:102 15 25 31 \#151 ; 26 26 25 48 25 ' 54 ' 102 :102 6 34 35\#151 ; 35 23 34 56 34 58 100:120 15 70 9\#151 ; 70 29 70 18 \#151 ; 120:120 3 39 14\#151 ; 39 33 39 23 39 24 100:110 3 54 24\#151 ; 54 29 54 26 54 24 110:120 3 15 49 - 15 56 15 51 15 54 120:010 2 19 34\#151 ; 19 40 19 37 19 42 120 : 301 11 76 24\#151 ; 76 49 76 39 !
76 40 301:120 10 103 6 \#151 ; 103 36 103 21 103 20 Lanthanum Paradibromobenzenesulphonate , La(C6H3Br2S03)3 , l8H20.\#151 ; This hydrate appears to be isomorphous with the corresponding salts of neo-and praseo-dymium ; like these salts , it effloresces when exposed to the air .
Water : Found 22*73 ; calculated 23*03 per cent. La(C6H3Br2S03)3,9H20.\#151 ; The crystals of this hydrate are very closely Morphological Studies of Benzene Derivatives .
isomorphous with those of the corresponding neo- and praseo-dymium salt , though the habit of development is generally slightly different from that of the other two ( fig. 2 ) .
The forms are the same , excepting that { 010 } rarely appears .
1IO * I O O Fig. 2 .
Fig. 3 .
Water : Found 12*95 ; calculated 13*0 per cent. System : orthorhombic .
Axial ratios : a : b : c \#151 ; 1*3965 :1 : 0*8753 .
The following angular measurements were made:\#151 ; Angle .
| No. of observations .
Limits .
Mean obsorved .
Calculated .
100 : 301 43 46 ' 32'\#151 ; 47 ' 3 ' 46 ' 46 ' 301 ; 102 23 25 32\#151 ; 26 5 25 48 25 ' 50 ' 102 :102 8 34 35\#151 ; 35 5 34 52 34 48 100 : 110 15 54 5\#151 ; 54 36 54 21 54 24 110 :120 9 15 41\#151 ; 16 7 15 55 15 54 100 :120 27 69 55\#151 ; 70 35 70 18 \#151 ; 120 :120 6 39 14\#151 ; 39 28 39 22 39 24 120 : 301 8 76 31\#151 ; 76 53 76 40\#163 ; 76 39 301 : 120 7 103 5\#151 ; 103 26 103 19 103 21 Cerium Paradibromobenzenesulphonate.\#151 ; Two hydrates of the cerium salt have been obtained containing respectively 18 and 9 molecules of water of Prof. H. E. Armstrong and Mr. E. H. Rod .
[ June 6 , v * crystallisation ; the latter was found to be closely isomorphous with the corresponding Nd , Pr , and La salts but as the purity of the salt is doubtful , the measurements are not quoted .
Gadolinium Paradibromobenzenesulphonate.\#151 ; By crystallising this salt at different temperatures , two different hydrates have been obtained , one separating at 25 ' with 12 molecular proportions of water , the other at 50 ' with seven ; at 37 ' both hydrates crystallised out together .
Gd(C6H3Br2S03)3,12H20.\#151 ; This hydrate crystallises in the form of fairly large , well-developed rhombs belonging to the monosymmetric system ( fig. 3 ) .
The salt effloresces very slowly in air at ordinary temperatures .
Water : Found 1644 ; calculated 16*39 per cent. The crystals are rhombic plates , the most prominent form being { 010 } ; other forms present are { 110 } , { Oil } , { 101 } and occasionally { 130 } .
No definite cleavage was detected .
System : Monosymmetric .
Axial ratios : 0*5952:1 : 0*3817 .
/ 3 = 76 ' 48 ' .
The following angular measurements were made:\#151 ; Angle .
No. of observations .
Limits .
Mean observed .
Calculated .
010 :011 15 69 ' 29'\#151 ; 70 ' 1 ' 69 ' 37 ' Oil : Oil 6 40 32\#151 ; 40 42 40 39 40 ' 46 ' 010:110 15 59 11\#151 ; 60 18 59 50 59 55 110:110 7 60 2\#151 ; 60 50 60 26 60 10 110:011 11 68 45 \#151 ; 69 7 68 56 \#151 ; Oil : 101 6 40 44 \#151 ; 40 57 40 48 : 40 51 101 :110 7 70 8\#151 ; 70 19 70 13 \#151 ; 010 : 101 4 89 55 \#151 ; 90 5 90 0 90 0 110 : Oil 6 89 22\#151 ; 89 29 89 25 89 28 Oil : 110 6 90 28 \#151 ; 90 40 90 35 90 32 Gd(C6H3Br2SO3)3,7H20.\#151 ; Three estimations of water in this salt gave 9*88 , 9*95 , 9*97 per cent , of water , the mean of these corresponding to 6*75 molecular proportions ; therefore the salt is probably the heptahydrate .
It forms very massive crystals ; when seen beside the colourless duodeca-hydrate , these appear to have a yellow tinge .
The crystals belong to the monosymmetric system but the forms developed do not allow of a complete determination being made of the axial ratios .
Samarium Paradibromobenzenesulphonate.\#151 ; The only hydrate of this salt prepared contains 18 molecules of water .
Water : Found 22*85 ; calculated 22*83 per cent. .
This salt is probably isomorphous with the corresponding hydrates of the Pr , Nd , La and Ce salts but has not been obtained in crystals good enough 1912 .
] Morphological Studies of Benzene Derivatives .
for goniometric measurement .
It separates in pale straw-coloured prisms , which effloresce rapidly in air .
Discussion of Results , On inspecting the series of axial ratios several facts of interest appear .
System .
a : b : Nd(C\#171 ; H3BiyS03)3 , 9H20 ... Orthorhombic ... 1-3990 : 1 : 0-8789 Pr \#187 ; . . .
\#187 ; . . .
1-3964 : 1 : 0-8798 La " .
) ) . . .
1-3965 : 1 : 0-8753 Ce . . . . . .
1-4106 : 1 : 0*8873 Gd 12H20 ... , Monosymmetric 0-5952 : 1 : 0-3817 , \amp ; = 76 ' The relationship between the three salts of Nd , Pr and La , as indicated by the approximate equality of the axial ratios of the crystalline substances , is much closer than is ordinarily observed between members of an orthorhombic isomorphous series ; a further indication is thus afforded of the correctness of the view that the metallic elements concerned form a very closely related group of the kind contemplated in the modified periodic table now given .
The axial ratios of the cerium salt diverge somewhat from those of the other salts , probably owing to the material being impure ; the axial ratios of the Nd , Pr and La salts are , however , identical within the limits of error of measurement .
It is next to be observed that the number 3 is frequently involved in connexion with the compositions of the orthorhombic salts ; these contain in the molecule 3 phenyl groups ( 3x3x2 atoms of carbon ) , 3 sul-phonic groups , each containing 3 atoms of sulphur and 3x3 atoms of sulphonic oxygen , 3x3 molecules of water , etc. , whilst the metallic atom has a valency of 3 .
The gadolinium salt is of lower crystalline symmetry and , at the same time , the frequency of repetition of the factor 3 in the composition is less marked in that the salt contains 3x4 molecules of water in place of 3 x 3 molecules ; the sums of the valency units in the molecular units , 150 and 162 , are also multiples of 3 .
The present is thus a suitable case to which to apply the conclusion drawn from the work of Barlow and Pope that the frequent repetition of the factor 3 in the manner indicated should lead to the presence in the crystal structure of a trigonal or pseudo-trigonal axis of symmetry.* The presence of a trigonal axis of symmetry involves the presence of a value 1 : v/ 3/ 2 = 1 : 0*8660 , amongst the axial ratios , if the crystal structure under discussion be referred to a rectangular system of co-ordinates ; the * Compare Jerusalem , 'Chem .
Soc. Trans. , ' 1910 , vol. 97 , p. 2190 .
Prof. H. E. Armstrong and Mr. E. H. Rod .
[ June 6 , presence of a pseudo-trigonal axis is revealed by the presence of a ratio approximating to the above value amongst axial ratios obtained by reference to approximately rectangular co-ordinates .
Since , in the orthorhombic series now described , the axial ratio , c:b , in each case approximates to the value quoted , it is clear that the axial direction a is that of a pseudo-trigonal axis .
Similarly , in the monosymmetric gadolinium salt , the axial direction c is that of a pseudo-trigonal axis , because the axial ratio b:a = 1-6801 : 1 = 2x0-8400:1 again involves a quantity approximating to ^/ 3/ 2 .
If , in the latter case , the axis c were truly trigonal , the angles in the zone [ 010:110 ] should all be 60 ' ; actually they have the values 60 ' 10 ' and 59 ' 55 ' , so that no doubt can exist concerning the pseudo-trigonal nature of the axis in question .
In discussing the crystalline structure of benzene derivatives , Barlow and Pope showed that one particular class of such substances could be represented as derived from a cubic closest packed homogeneous assemblage and that the derived assemblages should possess , in the ideal condition , a trigonal axis of symmetry .
On referring the assemblages in question to a rectangular axial system in which the axis c is the trigonal axis , the ratio of the other two , say a and b , should be , a:b = 1 : ^3/ 2 = 1 : 0*8660 ; further , on calculating the equivalence parameters , and of a substance of this class , it should be found that whilst : = 1:0-8660 = the third parameter , z , should have a somewhat lower value than z = 2-780 , the value calculated for crystalline benzene itself .
In order to exemplify these somewhat complex numerical relationships the case of y\gt ; -diiodobenzene may be quoted ; this substance is orthorhombic and the axial ratios of the crystals are 0*4342 : 1 : 0*3653 .
On dividing unit length along the axis b by two , the axis c is revealed as a pseudo-trigonal axis , because the axial ratios become a : b : c =\#166 ; 0*8684 : 1 : 0*7306 and the ratio a/ b is nearly .y/ 3/ 2 = 0*8660 .
Since c is the pseudo-trigonal axis , the equivalence parameter , z , corresponding to this direction , should be rather less than 2*780 , the value for benzene .
The sum of the valencies in the molecule is W = 30 and the equivalence parameters are x :y:z \#151 ; 3*1402 : 3*6161 : 2*6419 , The specified conditions are thus all fulfilled.* The markedly pseudo-trigonal character of the neodymium , praseodymium , lanthanum and cerium salts now described , the axis a being the pseudo* Barlow and Pope , 'Trans .
Cbem .
Soc. , ' 1906 , vol. 89 , p. 1702 .
1912 .
] Morphological Studies of Benzene Derivatives .
trigonal axis , indicates that the crystal structure is derived from the benzene structure of cubic origin typified by p-diiodobenzene ; the manner in which the salts are derived from this compound consisting in the opening out of the zigzag columns of atomic domains representing carbon atoms in the assemblage depicted by Barlow and Pope* to a sufficient extent to allow of the insertion , in homogeneous close-packing , of the domains representing the substituting groups in the salts .
Inasmuch as the structure must necessarily remain pseudo-trigonal in symmetry , the opening-out process must occur to an approximately equal extent in all directions perpendicular to the pseudo-trigonal axis .
The conditions which have been specified above enable a connexion of a far-reaching and quantitative character to be established between the crystalline forms of p-diiodobenzene and of the complex salts now described in the light of the Barlow-Pope theory ; this may be stated as follows :\#151 ; The axial ratios , the equivalence parameters and the valency volume of p-diiodobenzene are taken as , a !
\ V \ c , x ' : y ' : z ' and W ' and the corresponding values for the metallic salts as a " : V ' : c " , a " : y " : z'\ and W " .
W ' = 30 and W " = 150 .
The directions a ' and c " being pseudo-trigonal axes , the following conditions hold in the manner essential to the theory :\#151 ; a ' : V = 0 8684 : 1 , c " : 5 " = 0-8789 : 1 , : 1 = 0-8660 approx. Further , c ' : 6 ' = 0*7306 : 1 = 1 : 1-3687 , 5 " : = 1 : 1-3990 approx. The close correspondence between the theoretical equalities just stated affords convincing proof of the existence of the quantitative relationship which the theory indicates between p-diiodobenzene and the salts now described .
It remains to find simple multiples of the axial ratios of the orthorhombic salts which lead to the calculation of equivalence parameters such that the value of x " for the salts is approximately equal to the value z ' \#151 ; 2*6419 for p-diiodobenzene ; this last condition is fulfilled by taking a"/ 3 and 2c " .
The equivalence parameters then assume the following values :\#151 ; W. x " .
z " .
Nd(C6H3Br2S03)3,9H20 150 2-6475 : 5-6773 9-9796 Pr A A \#187 ; \#187 ; .
150 2-6434 : 5-6788 9-9925 \gt ; \gt ; )y . .
150 2-6480 : 5'6885 9-9582 Ce " " 150 2-6537 : 5-6438 10-0154 * Figs. 7 to 9 , loc. cit. VOL. LXXXVII.\#151 ; A. Q Morphological Studies of Benzene Derivatives .
The requisite correspondence between the value for diiodobenzene , z \#151 ; 2*6419 , and the mean value of x"for the first three salts which have been accurately measured , namely x " = 2*646 , is quantitatively complete to within about one part in one thousand .
It is now necessary to find simple fractions of the axial ratios for the gadolinium salt , a : b:c = 0*5952 :1 : 0*3817 , / 3 = 76 ' 48 ' , such that the equivalence parameter measured in the direction of the pseudo-trigonal axis c has the same value as the parameter z ' \#151 ; 2*6419 for diiodobenzene and x ' = 2*646 for the three orthorhombic salts ; all the other conditions specified in connexion with these latter salts are fulfilled .
On calculating the equivalence parameters of the gadolinium salt , taking 2c/ 3 , the following values are obtained : / 3 = 76 ' 48 ' , W. x. y. z. Gd(C6H3Br2.S03)3,12H20 . .
162 6*1415 : 10*3185 : 2*6257 .
In this case also the parameter z is identical within very narrow limits with the z value for diiodobenzene .
The mass of quantitative correspondence which has now been adduced between the crystal structure assigned to ^\gt ; -diiodobenzene and that possessed by the pseudo-trigonal salts now described , is a conclusive demonstration that we are in possession of a very powerful method of correlating crystalline form with chemical composition and constitution .
The crystal structure of the metallic salts under discussion must be regarded as one in which the zigzag columns of carbon domains depicted in the diagrams referred to remain intact but are merely pushed apart so as to allow of the introduction of the substituting groups by means of which the salts are derived from benzene.* Under these conditions the plane ( 100 ) perpendicular to the pseudo-trigonal axis a in the orthorhombic salts should exhibit differentiating physical properties because it is the plane which separates layers of aggregated molecules of the salt ; in accordance with this it is observed that ( 100 ) is a plane of perfect cleavage in the crystals .
One other conclusion of importance remains to be mentioned .
Hitherto it has been possible to determine the directions in which valency acts only in the case of carbon and the allied quadrivalent elements , the method used being one which involves the application of geometrical and optical but not of crystallographic considerations .
The proof arrived at in this communication , by a crystallographic method , that the three directions of valency in the tervalent elements La , Nd , Ce , Pr , Sm , Gd are symmetrically disposed , the metal occupying the central position in a plane containing three benzene groups , is therefore of special interest , as the first example of a novel method * Loc .
citp .
1700 .
Series Involving the Fourier Constants of a Function .
217 of determining the directions in which valency acts .
It is well known that this is a problem which has attracted much attention of late years in the case of nitrogen .
We are now studying the salts of tervalent metals generally , in order to ascertain if the method be applicable to all such elements : it is proposed also to apply the method to the elements of the nitrogen group .
In conclusion , we cannot refrain from calling attention explicitly to the proof that is again given of the service crystallography can render to chemistry and of the value of the Barlow-Pope theory as a means of correlating external form with internal molecular structure .
O71 the Convergence of Certain Series Involving the Fourier Constants of a Function .
By Prof. W. H. Young , Sc. D. , F.RS .
( Received May 16 , \#151 ; Read June 13 , 1912 .
) S1 .
In a paper " On the Fourier Constants of a Function , " published in the ' Proceedings ' of this Society , * I showed how certain theorems , previously given by myself , might be employed to obtain formulae for the sums of certain series involving the Fourier constants of a function .
In the case in which the function is bounded , it was only proved that the formulaef hold whenever the series converge , or , more generally , when the summation is performed in the Cesaro manner ( index unity ) .
In a more recent paper , also published in the ' Proceedings ' of this Society , !
" On a Mode of Generating Fourier Series , " I showed incidentally that the formulae are still applicable when the function has every power of index less than 1 +p summable , provided only the index q , which occurs in the formulae in question , is greater than 1/ ( 1 +p)\gt ; Here again the theorem , as stated , contained the restriction that the summation was to be made in the Cesaro way .
The main object of the papers in question was , in fact , to explain how certain methods might be employed , and these methods were in themselves inadequate for the purpose of removing the restriction .
The theorems used involve the general theory of the integration of the Fourier series of * 1911 , A , vol. 85 , pp. 14\#151 ; 24 .
t Loc .
cit. , S 4 .
t 19H , A , vol. 85 , p. 422 .

