On the Atomic Weight of Graphite .
By B. C. BRODIE , F.R.S. , Professor of Chemistry in the University of Oxford , and President of the Chemical Society .
Received May 12 , -Read May 12 , 1859 .
THE term Graphite has been indiscriminately applied to many varieties of native carbon of very different properties .
The graphite of New Brunswick differs but little in appearance from anthracite coal .
The graphite of Greenland is not very dissimilar , but possesses rather more metallic lustre .
However , among these varieties of carbon , two may be especially distinguished , -by a superior degree of metallic lustre , by their structure , and other well-defined properties .
In the following paper , the term Graphite is limited to these two varieties , which may be further distinguished as c ; lamellar " and amorphous .
" The lamellar graphite is found in great abundance in Ceylon , whence large quantities are annually imported into this country .
It appears in commerce in masses , sometimes of the weight of many pounds , of a brilliant metallic lustre , and possessing a distinct fibrous structure .
It is very difficult by mechanical processes to bring this graphite to a fine state of division ; however , by a prolonged grinding in water , it may be reduced to minute flat plates .
This graphite is associated with quartz .
A deposit of the same variety of graphite has recently been discovered at Travancore .
The graphite from Travancore has no fibrous structure , but is in the form of slightly coherent , minute plates .
Specimens of a similar graphite have also been given to me , from Moreton Bay in Australia , in a matrix of quartz , and from Ticonderoga , in the State of New York , associated with olivine and sphene .
When cast iron is dissolved in acid , a residue is left of about 4 per cent. of carbon in the form of graphite .
This graphite also consists of minute brilliant plates , and is perfectly similar in its appearance and properties to the lamellar variety of native carbon .
Amorphous graphite is found in Borrowdale in Cumberland , and is also largely imported into this country from Germany , probably from Griesbach near Passau , but I am unable to speak with certainty as to the locality whence it comes .
It appears as a powder of a silvery grey colour , soft to the touch , and which rubbed on paper gives a brilliant metallic streak .
This graphite is much softer than the other variety , and therefore better adapted for the manufacture of pencils , The graphite used in my experiments was in all cases carefully purified by boiling with acids , and by fusion with hydrate of potash in a silver crucible ; unless otherwise mentioned , it is to be understood to be the variety from Ceylon .
The graphite thus prepared leaves an almost inappreciable residue , and gave to analysis 99-96 of carbon .
Its specific gravity was ascertained in two determinations as 2 25 and 2-26 .
The following investigation was undertaken with the view of ascertaining the weight with which carbon in the form of graphite enters into combination .
With this end I examined its reactions .
The first experiment in which I succeeded in eliciting a difference in the chemical reactions of the different forms of carbon , was , in the case of the action upon them , of a mixture of concentrated nitric and sulphuric acids .
When finely-divided carbon , in the form of lampblack , or charcoal from the decomposition of sugar , is heated with a mixture of 1 of nitric and 4 of sulphuric acids , the carbon is rapidly oxidized , and a black substance is formed , soluble in the concentrated acid , but precipitated on the addition of water .
This substance is insoluble in acids and saline solutions , but is soluble in pure water and in alkalies .
It is accompanied with other products which render its purification difficult .
When the graphite of Ceylon is treated in a similar manner , the result is very different : the graphite becomes of a beautiful purple colour , and falls to pieces in the fluid .
The substance , after the acid has been washed from it by water , has much the appearance of the graphite itself , but is darker in colour .
It was found on analysis to contain the elements of sulphuric acid combined with oxygen , with hydrogen , and with a large amount of carbon .
My efforts to procure this substance of a constant composition have been unavailing ; it is insoluble in all reagents ; it may be boiled with a strong solution of potash , without separation of sulphuric acid , and with slight , if any , alteration of weight ; when heated it undergoes a remarkable change gases are given off in the interior of the substance , which swells up in a most singular manner and is reduced to the minutest state of division .
The residue consists of carbon , which has the appearance and the structure of the lamellar graphite .
A similar oxidation takes place when the nitric acid in the above experiment is replaced by other oxidizing agents , such as bichromate of potash and chlorate of potash .
These experiments established one point of importance , the existence of a peculiar compound of carbon in the form of graphite , The discovery of this substance led me to turn my attention to the oxidation of graphite .
I found that graphite , when heated with a mixture of nitric acid and chlorate of potash , increased in weight , and that the substance formed was , on the application of heat , disintegrated with evolution of gas .
The disintegrated substance differed but little in appearance from the original graphite .
The analysis of the body resulting from different preparations , and with a difference in the time of oxidation , gave no constant result , These difficulties led me to further experiments .
I found that when the substance formed , by the treatment of the graphite with the oxidizing mixture was washed free from the salts produced in the reaction , and dried at 1000 and again oxidized , it gradually under 'went a change in appearance , until , after the fourth or fifth repetition of the process , the whole of the graphite was converted into a substance of a light yellow colour , consisting of minute transparent and brilliant plates .
Analysis showed that this change was attended with a gradual alteration of the constitution of the substance , but that , finally , a time arrived when treatments with the oxidizing mixture produced no further change .
It is remarkable that this result cannot be produced by one prolonged treatment ; before the oxidation can proceed , the original conditions must be restored , The details of this process are as follows:-A portion of graphite is intimately mixed with three times its weight of chlorate of potash , and the mixture placed in a retort .
A sufficient quantity of the strongest fuming nitric acid is added , to render the whole fluid .
The retort is placed in a water-bath , and kept for three or four days at a temperature of 60 ?
C. until yellow vapours cease to be evolved .
The substance is then thrown into a large quantity of water , and washed by decantation nearly free from acid and salts .
It is dried in a water-bath , and the oxidizing operation repeated with the same proportion of nitric acid and of chlorate of potash until no further change is observed : this is usually after the fourth time of oxidation .
The substance is ultimately dried , first in vacuo , and then at 1000 C. A modification of the process which may be advantageously adopted , consists in placing the substance with the oxidizing mixture in flasks exposed to sunlight .
Under these circumstances the change takes place more rapidly , and without the application of heat .
These crystals , when examined with the microscope , are perfectly transparent , and exhibit beautiful colours by the agency of polarized light .
Professor MILLER of Cambridge , who was good enough to examine them , has communicated to me the following observations:- " The crystals , though not absolutely too small to be measured , are too thin and too imperfect to admit of measurement with the reflective goniometer .
I have examined them under a microscope , for the purpose , if possible , of making out at least the system of crystallization to which they belong .
Their system appears to be either the prismatic or the oblique , -most likely the former .
Sometimes the crystals appear broken , as in the annexed figure .
The straightness of the fractured edges ( in the direction --___ OF ) evidently indicated the existence of aA cleavage in that direction .
The crystals are so extremely thin in a direction perpendicular to a. the paper on which the above figure is traced , GU that it is impossible to obtain any reflexion , except from the faces parallel to the plane of the paper .
" The crystals , on the application of heat , are decomposed with ignition .
Gases are evolved , and a black residue is left of a substance resembling in appearance finely divided carbon .
I have not discovered any reagent by which these crystals may be dissolved , and they admit of no process of purification .
The explosion which the crystals undergo on the application of heat render special methods of analysis necessary , and I have taken , with regard to the analysis , every precaution which suggested itself .
Neither chlorine nor nitrogen could be detected .
The substance was placed in a platinum boat in a porcelain tube and burnt in a* Measurements of the crystals , by means of the microscopic goniometer , which have subsequently been made by Mr. A. H. Ciauln , of Lincoln College , have determined with certainty that the crystals belong to the one or other of these systems .
current of oxygen , the gases evolved being passed over heated oxide of copper .
At the extremity of the tube , where the oxygen entered , was placed a stopper of asbestos , and a second stopper between the substance and the oxide of copper .
The ordinary potashbulbs were replaced by two tubes filled with pumice moistened with concentrated solution of potash .
In order to ascertain in the mostrigorous manner the point at which the constitution of the compound became fixed , it was analysed after each successive oxidation .
It will be seen that the analyses became constant upon the fourth treatment ; after which , the most prolonged action of the oxidizing mixture produced no further change .
The following are the results per cent. of the analyses of the same preparation of graphite , treated in the first instance once , in the second , twice , with the oxidizing mixture : I. II .
Carbon ... 69 67 67-79 Hydrogen ... 148 1'84 Oxygen ... 2885 30 37 100-00 100'00 The following are the results of analysis after the constitution of the body may be regarded as fixed .
Analyses 1 and 2 are of a substance prepared by four treatments ; analyses3 and 4 , of a substance prepared by five treatments ; analysis 5 , of a substance prepared by six treatments ; analyses 6 and 7 , of a substance prepared by seven treatments ; analysis 8 is of a substance dried in vacuo , the previous preparations having been dried at 1000 ; analysis 9 is a preparation from the amorphous graphite of Cumberland , the previous preparations having been from the graphite of Ceylon .
I have also prepared the same compound from the graphite of cast iron .
Substance taken .
Carbonic acid formed .
Water formed .
1 .
0-23375 0-52025 0'02925 2 .
0-233 0-51875 003775 3 .
0.24475 0'545 0'04175 4 .
0-27275 0-609 0-0465 5 .
0-209 0-46 0-036 6 .
025275 0-5635 0'042 7 .
0-23825 0o5325 0'04025 8 .
0-2555 0-56875 0 0435 9 .
0-216 0 48 0 0355 These analyses correspond to the following percentage composition:1 .
2 .
3 .
4 .
5 .
6 .
7 .
8 9 .
Carbon 60-70 60 74 60-73 60-88 60-47 60680 60-94 60-71 60-69 Hydrogen 1-37 1 80 1P89 1-88 1 91 1'80 1-87 1-89 1-75 Oxygen 37-93 37-46 37-38 37-24 37-62 37-49 37419 37 40 37-56 100 00 100100 100 00 100 00 10000100 100I00 100-00 100I00 The mean of these analyses ( omitting the hydrogen of the first analysis , which is evidently faulty ) isCarbon ... 60'74 Hydrogen. .
185 Oxygen ... 37T41 100 00 A correction has yet to be made in these numbers .
The graphite from which the substance is prepared leaves on combustion a slight ash , and during the prolonged treatment of the substance a further portion of incombustible matter is introduced , probably from the glass of the vessels in which the operations were conducted , which , as I have ascertained , may be estimated at 05 per cent. When this correction is made , we have for the percentage constitution of the substanceCarbon ... 6104 Hydrogen. .
185 Oxygen ... 37411 100*00 The formula C , , H4 0 corresponds to this result .
Yft requiresCHl ... 132 61411 114 ... 4 l85 05 ... 80 37-04 216 100-00 This substance has the following properties .
It is insoluble in water containing acids or salts , but is very slightly soluble in pure water .
The crystals , placed upon litmus paper , have a feeble acid reaction .
It combines with alkalies .
When agitated with dilute ammonia , it is converted into a transparent jelly , but is not dissolved .
On the addition of acids it is separated unaltered from this combination in the form of a gelatinous mass resembling silica , which , dried under the air-pump , appears as a slightly yellow and spongy body of the same weight as the substance originally treated with ammonia .
The crystals treated with deoxidizing agents are readily decomposed .
When a solution of sulphide of ammonium or of potassium is poured upon the dry substance , a crackling sound is heard , and a body is ultimately formed possessing the metallic lustre and general appearance of graphite itself .
Changes similar in appearance take place on boiling the substance with an acid solution of protochloride of copper and of protochloride of tin .
The substances formed in these processes admit of no processof purification , and I have not been able to procure them in a state of purity .
This body belongs very distinctly to the class of acids , but from the insolubility of its salts and the facility with which they are decomposed , I have not been able to procure them pure .
The following determinations , however , indicate it to be bibasic .
A portion of the moist substance shaken up with baryta-water , washed out , and dried at 1000 , gave a compound containing 21-19 per cent. of barium .
The same substance suspended in water , and decomposed by a current of carbonic acid , gave a substance Mf9cccLIx .
2 M~ which , dried at 1000 , contained 13'30 per cent. of barium .
The formula CQ , H3 BaO5 requires 24-13 per cent. of barium .
The formula C22 H7 BaQ0 requires 13-73 per cent. of barium .
These substances are excessively hygroscopic , and when heated , explode with yet greater violence than the original body .
The compound itself may be termed graphic acid .
The residue which is left upon the decomposition by heat of graphic acid , although resembling carbon in appearance , is not that element , but contains a considerable amount of oxygen .
I made many experiments with a view of procuring this substance in a pure state , -both by decomposing the compound , alone , in an atmosphere of nitrogen , and by decomposing it mixed with a large quantity of chloride of sodium , with the view of moderating the violence of the decomposition , and causing it to take place at the lowest possible temperature .
My endeavours were unsuccessful , until it occurred to me to effect the decomposition in a fluid medium , by which the particles of the substance would be separated from one another and elevation of temperature precluded .
The fluid which I selected for the experiment was the mixture of hydrocarbons of high boiling-points from the Rangoon naphtha .
It was purified by rectification over sodium , and boiled at about 270'C .
When graphic acid is heated in this fluid , a considerable quantity of water is formed , which distils over between 100 ?
and 2000 , and which is accompanied at first by a slight , and ultimately by a brisk evolution of gas .
This gas was in one experiment collected and examined : it consisted solely of carbonic acid .
But water and carbonic acid are not the exclusive products of this reaction .
The hydrocarbon becomes of a deep red colour , and when filtered and distilled , leaves a black carbonaceous residue .
I have found no means of separating the body formed from the naphtha in which it is dissolved .
In the following experiments the substance was heated in the naphtha for three or four hours , at about 2500 C. , until all perceptible evolution of gas had ceased .
The graphic acid employed was weighed , and also the residue of its decomposition .
The residue was collected on a weighed filter , washed out with ether and with alcohol , and dried at 1000 .
Expt. I. 1-024 grrm .
of graphic acid gave 0-684 grm. of residue .
Expt. II .
1-0854 grm. of graphic acid gave 0-7248 grm. of residue .
Expi .
III .
2'1805 grms. of graphic acid gave 14607 grm. of residue .
This corresponds to a loss on 100 parts of the substance taken-in Experiment I. of 6-6 78 , in Experiment IL of 66T77 , in Experiment III .
of 66-98 parts .
These substances , burned in a current of oxygen , gave the following results .
In these and all following analyses the residual ash was weighed , and its amount deducted from the substance taken .
Substance taken .
Carbonic acid formed .
Water formed .
1 .
Expt. I ... .0-311 0-9147 0 0148 2 .
Expt. II ... .
0-4237 1'2372 0-022 3 .
Expt. 111 ... .
03573 1P0404 00178 4 .
The same e. 0'3391 0*9857 0 0185 These results correspond to the following per-centage composition : 1 .
2 .
3 .
4 .
Carbon.80-19 79-64 79 40 79-27 Hydrogen ... 0-52 057 0.55 0-60 Oxygen.1929 19-79 20-05 20-13 100 00 100o00 100.00 100-00 When this substance was carefully examined under the microscope , traces of transparent crystals could still be perceived , similar in appearance to the original substance .
To preclude the possibility of an error from the imperfect decomposition of the substance , the following experiments were made , in which the substance was submitted in the naphtha to a prolonged action of heat , for fourteen hours .
Expt. IV .
3-9185 grms. of substance gave 2'577 grms. of residue .
Expt. V. 17T175 grms. of substance gave 11'2965 grms. of residue .
In Experiment IV .
the temperature was raised to 2400 C. , in Experiment V. to 2200 C. These results correspond to a loss on 100 parts of the substance-in Experiment IV .
of 65-76 , in Experiment V. of 65-77 parts .
The analyses of these substances gave the following result : Substance taken .
Carbonic acid formed .
Water formed .
1 .
Expt. IV ... .
0 377 1P1126 00210 2 .
The same ... 04766 1-4068 0-0255 3 .
Expt. V ... .
0-3835 1-1327 0-022 These analyses give the following per-centage composition of the substance:1 .
2 .
3 Carbon.8048 80-50 80-56 Hydrogen ... .
062 0.59 0-63 Oxygen ... . .
18-90 18-91 -18-81 100 00 100 00 -100r00 In this substance , placed under the microscope , no traces of the transparent plates of the original body could any longer be perceived .
The result , therefore , of heating the substance for the ten additional hours had been to increase the per-centage of carbon in the residue from about 0-5 to 1 per cent. It is evident that the decomposition is arrested at about this point ; and if we take the mean , as may reasonably be done , of three analyses of the first and the three of the second series of experiments , as representing the constitution of the body , we have the following as its per-centage composition : Carbon ... . .
80 13 Hydrogen ... .
0'58 Oxygen. .
19'29 1 00 00 c2 This corresponds to the formula C , H , O , , which has the following composition : C22. .
264 80-00 H2. .
2 0-60 04. .
64 19-40 330 100-00 There is every reason to believe that , although the decomposition is difficult to regulate , this same body is formed by the simple application of heat to the substance without the intervention of the naphtha .
The result of one experiment , in which the graphic acid had been carefully heated in an air-bath to 2800 until it ceased to lose weight , gave to analysis-carbon 80-36 , hydrogen 0 71 .
In other experiments numbers approximating to these were obtained .
We have not the data to determine with certainty the precise mode of the formation of this body .
The simplest hypothesis which coincides with the ascertained facts , is that 7 equivalents of the graphic acid give by their decomposition 6 of the new body .
We have 7(C,1H405)-=1512 and 3(C221H2 4)=-990 and 1512:990::100:65'48 , 65'48 being the amount of residue found .
On this view it would be possible to form the new substance by the elimination of water , carbonic acid , and carbonic oxide from the original body , according to the equation 7(Cl H4 05 ) 3(=C22 112 04)+ 61H20+6 CO2+5 CO. The slight excess of carbon in the last series of analyses indicates the presence of some compound containing a higher percentage of carbon .
It will be rendered evident , by the following experiments , that we have only to heat this substance under slightly different circumstances to determine a further decomposition .
Portions of the substances last analysed were placed in a platinum boat in a glass tube , through which was passed a current of nitrogen , and which was heated in an air-bath to a temperature of 2500 .
The substance was weighed before and after the experiment .
Expt. 1 .
'9087 grm. of the substance resulting from Expt. IV .
gave a residue of 0-883 grm. Expt. 2 .
0 9209 grm. of the substance resulting from Expt. V. gave a residue of 0 8968 grm. This corresponds to a residue on 100 parts of the substance taken-in Expt. 1 of 97'16 , in Expt. 2 .
of 97-38 parts .
Water is given off from the substance during the process , which was collected in a sulphuric acid tube and estimated .
It corresponded , in the two experiments , to 2 26 and 2-30 per cent. The quantity formed was , however , too small for correct estimation , and these numbers are undoubtedly too high .
Traces of probably carbonic oxide , amounting to something under 1 per cent. , are also given off .
The action is perfectly definite .
The experiment was continued until the substance ceased to lose weight ; in the last two hours the loss in weight was not above 1 inilligramme .
The following are the results of the analysis of the bodies Substance taken .
Carbonic acid formed .
Water formed .
1 .
Expt. I ... .
0-367 11051 0-0163 2 .
The same ... 0 3783 1 1366 0 0139 3 .
Expt. II ... 0 3979 1 1938 0*O158 4 .
The same. .
, 0-3927 1-1724 0.0155 These numbers give the following percentage composition : 1 .
2 .
3 .
4 .
Carbon ... 8212 81'95 81L80 8141 Hydrogen. .
049 0-41 044 0-43 Oxygen ... 17-39 1764 17'76 1816 100(00 100'00 100^00 100^00 The substance derived from the preceding body , .
by the elimination of I atom of water from 3 atoms of the substance , would have the formula Cd H14 0 , .
3(C22 H2 04 ) -4 104+H112 0 .
The calculated composition of this body is as follows:066 .
792 81-48 H14 .
4 0'41 011 * 176 18111 972 100'00 Every 100 parts of the substance taken would leave a residue of 98-18 parts , and the water formed would amount to 1'80 part .
The amount of carbonic oxide formed in the preceding experiments is so small , that it can hardly be regarded as an integral part of the , decomposition ; and if this amount be added to the residue , we arrive at precisely the theoretical numbers .
By a greater elevation of temperature the substance undergoes further change , with the loss of carbonic acid and carbonic oxide .
But it may be exposed to a red heat for several hours in a current of nitrogen and only undergo a very partial decomposition , the residual substance containing a considerable portion both of hydrogen and oxygen .
BUFF and W6HLER , in their researches " 1 On the Graphitoidal Form of Silicon* , " discovered a remarkable series of compounds derived from it .
When hydrochloric acid gas is led over this form of silicon at a low red heat , a volatile liquid is formed containing silicon , chlorine , and hydrogen , to which they assign the formula Si2 H12 Cl. With hydriodic acid gas a similar reaction takes place , with the formation of the corresponding compound Si2 H12 1 .
In water , these substances decompose with the production of a compound of silicon , hydrogen , and oxygen , of the formula Si4 114 05 The analyses are unfortunately neither so concordant nor so exact as might be desired ; but the formulae are derived from the consideration of a system of decompositions , which hardly leaves room for any other hypothesis .
The general properties of the substance Si4 H1140 , correspond very closely with those of the graphite compound as separated from its combinations by an acid .
It is described as a white and voluminous body , which floats upon water , in which it is very slightly soluble .
Heated to a temperature of above 300 , it is decomposed with ignition and evolution of hydrogen .
The ratio of the number of equivalents of hydrogen to those of oxygen is as 4 to 5 .
It is formed exclusively from graphitoidal silicon , as the graphite compound from graphitoidal carbon .
From these considerations we may infer that the graphite compound is the same term in the system of carbon as the silicon compound in the system of silicon .
When we proceed to state this analogy in the formula of the substance , we are led to very remarkable conclusions .
The total weight of graphite , which in this compound is combined with the 4 atoms of hydrogen and the 5 of oxygen , is 132 .
If we assume that this weight is , like the corresponding weight , 84 of silicon , to be divided into 4 parts , we arrive at the number 33 as the atomic weight of graphite .
Representing this weight by the letters Gr , the formula , of the substances CQ , 11405 and C.2 H2 04 , and C66 H4 011 become Gr H14 05 Gr8 H2 04 Gr24H4114 O We are acquainted with a property of graphite by which this theory may be tested .
According to the law of DULONG and PETIT , which has been confirmed by the exact researches of REGNAULT , the specific heats of the elemental bodies vary inversely with their atomic weights .
With the atomic weights of the elemental bodies , as generally received , the elements are divided into two classes , the one in which the product of the specific heat into the atomic weight is approximately 3 3 , the other in which this product is approximately 6-6 .
This law expresses the only common physical property byr which these weights are characterized .
To this law , however , there is one remarkable exception , namely , the specific heat of carbon in its various forms .
The following are the specific heats of carbon as determined by REGNAULT : Specific heats .
Diamond .
014687 Native graphite .
020187 Graphite from the blast furnace. .
019702 Carbon from gas retorts.02036 Carbon from anthracite 0-201 Wood charcoal ... .
0.2415 Animal charcoal ... .
026085 In no single case , whether the atomic weight of carbon be assumed as 6 or as 12 , does carbon conform to the law .
The product of the specific heat of graphite into the atomic weight is with the weight 6 , 12 , with the weight 12 , 2A .
If , however , xee assume the atomic weight of graphite as 33 , we have for the product of the specific heat into the atomic weight the number 6-6 , which is according to the law ; this product being the same as the product of the specific heats into the atomic weights of the elements phosphorus , antimony , arsenic , bismuth , and iodine .
This form of carbon should be characterized by a name marking it as a distinct element .
I propose to term it Graphon .
In the formulia assigned by BUFF and W`iIILER to the compounds of silicon , the atomic weight has been assumed as 21 .
If the atomic weight of silicon be assumed as 28 , these relations no longer appear ; the atomic weight of graphite must be assumed as 44 , and the product of the specific heat into the atomic weight would be 8-8 .
We are thus enabled to decide upon the atomic weight of the graphitoidal silicon .
It is greatly to be regretted that those chemists who have had a sufficient quantity of this substance in their possession have omitted to determine its specific heat .
There is much reason to believe that in its other forms silicon , like carbon , has a distinct atomic weight .
The vapour-density of the chloride of silicon leads us to the weight 28 ; and it is worthy of remark , that the atomic weight 44 , which in the system of carbon corresponds to the weight 28 of silicon , would , assumed as the atomic weight of diamond , render its specific heat conformable to the law , since 44 x *147 =6 46 .
The relation is also singular which exists between the atomic weight of graphon and the atomic weights of the elements boron , silicon , and zircon , it being precisely the kind of numerical relation which exists between the weights of analogous elements .
Boron .
11 Silicon. .
21 Graphon. .
33 Zircon .
66 The weight 21 of silicon must be considered as still open to correction .
The results arrived at by DEVILLE * , from the oxidation of graphitoidal silicon , indicate a higher number .
Finally , these considerations lead us to the remarkable inference that carbon in the form of graphite functions as a distinct element ; that it forms a distinct system of combinations , into which it enters with a distinct atomic weight , the weight 33 .
Analogy would lead us to a similar conclusion with regard to the elements boron and silicon .
How far this inference is to be extended to the allotropic forms of other elements experiment alone can decide .

