86 Prof. H. E. Armstrong .
[ Mar. 5 , in the adult bird the cones are most numerous .
Some of the bipolar cells from the internal nuclear layer emigrate through the external molecular layer , and develop into young cones .
These investigations on the retina led to an inquiry as to what relationship there might be between the rod and cone cells , and the ciliated epithelium of the cerebral ventricles , and it is found that these are homologous structures .
The germinal cells which line the cavities of the cerebral vesicles remain there after they cease to divide , and from them cilia develop .
The inner ends of the myelospongium fibres do not develop cilia as is usually described .
The processes of the hexagonal pigment cells appear on the same date and at the same spot as the rods and cones and do not at first contain pigment granules .
They resemble pseudopodia in their mode of development , and also , after full development , they behave like pseudopodia , for they elongate and retract under the influence of light and carry the pigment granules with them ( the granules having no inherent power of movement ) .
The inner wall of the retinal cup has been seen to have the same structure as the wall of the cerebral vesicle in the early stages , and even later ( as at the 8th day ) some resemblance can still be drawn between them .
" The Classification of the Elements .
" By Henry E. Armstrong , Y.P.RS .
Beceived March 5 , \#151 ; Bead March 20 , 1902 .
Although no direct evidence acceptable to chemists has been adduced which in any way justifies the belief that the elements are decomposible , it is impossible \#163 ; o resist the conclusion that they are genetically related \#151 ; so closely in many respects do they resemble a series of related compounds , especially when regarded from the point of view of the organic chemist .
The generalisation known as the Periodic Law is in itself a justification of this view : the manner in which interrelationship becomes manifest when they are classified in accordance with its canons , being probably the strongest of all the arguments which can be cited as tending to show that the elements are compounds\#151 ; but compounds very different from those with which we are accustomed to deal .
Even in the form in which it was put forward by Mendeleeff , however , the periodic generalisation is but a first approximation : and the great Russian has himself pointed out that it needs improvement and development.* As chemists are beginning to recognise this , t I venture to submit a scheme of classification which I have been led to draw up in * Faraday Lecture , ' Trans. Chem. Soo .
, ' 1889 , p. 656 .
+ Comp. Biltz , * Deut .
cbem .
G-es .
Ber .
, ' 1902 , p. 562 .
19Q2 .
] The Classification of the Elements .
87 writing an article for the forthcoming Supplement to the ' Encyclopedia Britannica .
' The article , I may say , was sent to press in May , 1900 and the first proof before me is dated November 20 , 1900 .
The suggestion of an octave of elements having been made m early days the tendency to adopt an octavo system has been irresistible .
It is difficult now to discover any real justification of such a practice .
As there is frequently a difference of ( about ) a single unit between the atomic weights of contiguous elements , and as the " homologous elements among those of lower atomic weight differ by ( about ) sixteen units\#151 ; I assume that the " elementary difference " maybe ( about ) a single unit , and that the first " horizontal period " comes to an end with oxygen : in other words , that there are sixteen " vertical series " of homologous elements .
In the table on pp. 88\#151 ; 89 , the elements are entered under whole numbers without regard to their exact atomic weights .
The dominant principle on which the arrangement is based is that of maintaining elements which belong to the same family in the appropriate columns .
The chemist cannot adopt any other mode of arrangement in view of the uncertainty which attaches to mauy of the atomic weights .
I have not hesitated to assume that the molecules of argon and similar elements are polyatomic like those of nitrogen their nearest ally in outward behaviour\#151 ; and have regarded them as , as the molecules of all the ordinarily gaseous elements are of this order of complexity. .
It has always been my opinion that they are elements of intense activity , the atoms of which so fully satisfy each other that no residual affinity is manifest in the molecules .
From this point of view the appearance of argon immediately after fluorine is quite in order .
Turning to the several periods , the first calls for little remark .
Helium is put next to hydrogen\#151 ; and it may well be that elements somewhat like helium and argon will be discovered which will fall into positions 3 , 4 , and 18 ; in fact , this region may well turn out to be the nesting-place of such elements .
Beryllium is shown separated from lithium and magnesium , but not to an extent which need give rise to objection .
It is perhaps not beyond the region of possibility , however , even in the case of an element with so low an atomic weight , that iso-morphous mixtures may have passed as pure substances : this possibility has not been sufficiently taken into account in preparing the material for determinations of atomic weight .
In the second period , although it does not come immediately below nitrogen , phosphorus is very near to that element\#151 ; and sufficiently so to correspond with their very close relationship .
In the third period , when scandium is passed , it is impossible to proceed by units : to bring titanium into position it is necessary to step down five units .
Proceeding from titanium , however , vanadium and chromium fall into appropriate positions .
Prof. H. E. Armstrong .
[ Mar. 5 , 1 H 2 He 3 4 5 6 7 Id 8 17 is 19 P 20 A 21 22 23 Na 24 Mg 33 34 35 Cl 36 37 38 39 K 40 Ca 63 54 55 Mn 56 Fe 57 58 59 CoNi 60 61 62 63 Cu 64 65 Zn 78 79 80 Br 81 82 83 84 85 Rb 86 87 Sr 98 99 100 101 102 Eu 103 Rh .
104 105 Pd 106 107 108 Ag 109 110 111 112 Cd 125 126 127 I 128 129 130 131 132 133 Cs 134 135 136 137 Ba ( S Ira 150 , Eu 151 , Gd 156 , 191 Os 192 193 Ir 194 195 Pt 196 197 198 Au 199 200 Hg 201 The Classification of the Elements .
1902 .
] .H*\#151 ; \#166 ; 9 Be ]0 No J 11B 12 C 13 14 N 25 26 27 A1 28 Si 29 30 41 42 Kr 43 Sc 44 B1 45 46 1 f 47 48 Ti 49 50 66 67 X 68 69 70 Ga 71 72 Ge I 73 74 75 As 88 89 90 Yt 91 Zr 92 93 94 95 113 114 115 In 118 Sn 116 119 117 Vc ( ?
) 120 121 Sb 122 138 La 139 140 Ce 141 Pr | .
142 143 144 Nd Tb 163 , Er ] L66 , Yb 1 73 .
) 202 203 204 T1 205 206 207 Pb 208 Bi 232 Th 51 Vd 96 Nb 183 Ta 52 Cr 77 Se 97 Mo 124 Te ___________I Prof. H. E. Armstrong .
Table II .
[ Mar. 5 , 1 H 2 He 3 4 5 6 7 Li 8 17 18 19 F 20 A 21 22 23 Na 24 Mg 33 34 35 Cl 36 37 38 39 K. 40 Ca 53 54 55 Mn 56 Fe 57 58 59 CoNx 60 61 62 63 Cu 64 65 Zn 66 78 79 80 Br 81 82 83 84 85 Rb 86 87 Sr 97 98 99 \#166 ; 100 101 102 Hu 103 Rh 104 105 Pd 106 107 1 108 Agr 109 110 111 112 Cd 125 126 127 I 128 129 130 131 ' 1 132 133 Cs 134 135 136 137 Ba 138 La 139 140 Ce 141 Pr 142 143 144 Nd 145 146 147 148 149 150 Sm 151 Eu 152 153 154 155 1902 ] The Glassification of the Elements .
Table II .
9 Be 10 No 11 B 12 C 13 14 N 15 25 26 27 A1 28 Si 29 30 31 P 41 42 Kr 43 Sc 44 45 46 47 48 Ti 1 49 1 \#163 ; 0 51 Vd 67 68 X 69 Ga 70 71 72 Ge 73 74 75 As 76 88 89 90 Yt 91 Zr 92 93 94 95 Nb 113 114 I 115 In 116 117 Vc ( ?
) 118 Sn 119 120 121 Sb 122 123 16 O 32 S | 52 Cr 77 Se 96 Mo 124 Te Prof. H. E. Armstrong .
[ Mar. 5 , 156 G-d 157 158 159 160 161 162 163 Tto 164 165 166 Er 167 168 169 170 Tu i 185 186 187 188 189 190 191 Os 192 193 Ir 194 Pt 195 196 197 Au i 198 199 200 Hgr 201 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 1 In the fourth period , a similar step down in the fourth column from iron to cobalt and nickel is necessary to bring copper into an appropriate position ; and for a like reason zinc is included in a group with copper .
There are similar drops in the eleventh and twelfth columns .
Selenium appears with 77 instead of 79 as its atomic weight , in order to preserve it in the oxygen-sulphur series : bearing in mind that the atomic weight of sulphur differs from that of chlorine by fully three units , whereas that of selenium is ordinarily supposed to differ 1902 .
] The Classification of the Elements . .
1 171 172 173 Yb 174 175 176 177 178 179 180 181 182 183 Ta 184 W 202 203 204 T1 205 206 207 Pb 208 Bi 209 210 211 * 229 230 231 | 232 Th. 233 234 235 236 237 238 239 240 TJ 1 from that of bromine by only a single unit , the change does not appear to be an unjustifiable one .
The same argument applies to tellurium in the following period .
Personally , I entertain no doubt that the atomic weight of this element will ultimately prove to be considerably lower than that of iodine .
It is unnecessary to go into further details\#151 ; the table is self-explanatory j'but it may not be superfluous to point out that the characteristic 94 Prof. H. E. Armstrong and Dr. T. M. Lowry .
[ Mar. 13 , and most remarkable feature brought out by this mode of classifying the elements is the existence in various parts of the system of groups or series of related elements from the highest term of which alone " progression " takes place .
The existence of such groups in the case of the iron and platinum metals and some of the rare earths , has long been recognised\#151 ; but not their significance .
It is not unlikely that some of the columns will be " cleared " of such groups when atomic weights generally are known with a closer approach to certainty .
Of their existence in column 4 , there can be no doubt ; but if in column 7 rubidium were put at 84 , 85 could be transferred to column 8 ; and if lower down in the same column , caesium were put at 131 , 132 and 133 could be transferred to Column 8 , or the necessary " correction " might be made perhaps with greater advantage by including in column 4 two terms ( 81 and 82 ) in the fifth , and three ( 128,129 , 130 ) in the seventh period : column 7 would then be free from " grouped " elements .
Such groups , however , are undoubtedly as characteristic of column 8 as of column 4 ; whether they are of column 9 is open to question\#151 ; the cerium group might well follow barium in column 8 ; but wherever they may come , it is clear that the elements of this group are a very numerous body , and that a remarkable expansion may be looked forward to in this part of the table .
Column 11 might also , by a similar process , be cleared of grouped elements .
If such a clearance turn out to be possible , grouped elements will be characteristic of only three of the families\#151 ; those in columns 4 , 8 and 12 .
Making allowances in the manner suggested , the " smoothed " scheme is arrived at which is embodied in Table II .
The possibilities disclosed by a system of classification such as that here suggested are remarkable , but they are on the surface and need not be dwelt upon .
Speculation on suet* a subject will be justified if it but lead to further appreciation of the rhythm which , undoubtedly underlies the relationships subsisting among the elements .
That work in plenty is left for the chemist to do is certain .
" Persulphurie Acids .
" By Henry E. Armstrong , V.P.K.S. , and T. Martin Lowry , D.Sc .
Eeceived March 13 , \#151 ; Bead March 20 , 1902 .
Although it was observed by Faraday in 1834 that " if the acid were very strong , a remarkable disappearance of oxygen took place " on electrolysing aqueous solutions of sulphuric acid , * it was not until 1878 that its disappearance was at all satisfactorily accounted for by the discovery of persulphuric acid by Berthelot .
* 1 Resell-dies in Electricity , ' series yii , S 728 .

