IIt On the Application of the Conversion of Chlorates and Nitrates into Chlorides , and of Chlorides into Nitrates , to the determination of several equivalent numbers .
By FREDERICK PENNY , Esq. Communicated by H. HENNELL , Esq. F.R.S. Received January 17 , -Read January 24 , 1839 .
1 .
THE following researches originated from some experiments which were undertaken to discover an improved niethod for ascertaining the quantity of nitrate of potassa existing in crude saltpetre .
After several unsuccessful attempts the action of hydrochloric acid was tried .
The fact , that nitrates are decomposed by this acid , has been long known ; but the nature of the resulting compound of potassium has not , so far as I am aware , been hitherto determined .
I anticipated that the nitrate would be decomposed into chloride of potassium .
To decide the question some pure nitrate of potassa was mixed with hydrochloric acid , and the mixture heated ; while at comimon temperatures no perceptible action occurs , but inmmediately the acid becomes hot , decomposition commences .
Chlorine and nitrous acid are evolved with copious effervescence , and the nitrate slowly disappears .
The solution was gradually evaporated to dryness , and the dry salt treated with an additional quantity of acid until decomposition was no longer evident .
The resulting salt was then carefully examined , and it was found to be pure chloride of potassium .
This experiment was repeated several times , and all the results concurred in satisfactorily establishing the fact , that nitrate of potassa may be perfectly converted into chloride of potassiumn , provided a sufficient quantity of the acid be employed , and the temperature necessary to effect the decomposition be properly regulated .
2 .
So far the decomposition was admitably adapted for the object mentioned at the commencement .
The usual impurities , such as chlorides , sulphates , silica , &amp;c. which any sample of crude saltpetre might contain , would obviously remain unchanged , while the nitrate of potassa alone suffering decomposition , its quantity could easily be ascertained , by comparing the weight of the resulting salt with that obtained from a known quantity of absolutely pure nitrate .
Several experiments were therefore performed to determine the exact quantity of chloride of potassium corresponding to a known weight of nitrate .
The mean result of four experiments gave the ratio of 100 of nitrate to 73'730 chloride .
I was then naturally led to compare this result with the equivalent numbers of these two compounds .
In this country there are two series of equivalents in general use , one in which whole numbers are adopted , and the other in which fractional parts are admitted .
For example , according to the former , nitrate of potassa will be 102 , and to the latter 101-3 .
So chloride of potassium will be 76 and 74'6 .
Whence , according to the former , every 100 parts of nitrate should yield 74*51 of chloride : the latter gives the ratio of 100 to 73'613 .
But these results differ considerably from my experiments .
In the one case , we have a difference of '78 , and in the other '12 .
Whence therefore could they arise ?
Either the process must be defective , or the equivalent numbers , so generally considered as correct , must be erroneous .
3 .
To determine whether the process was inaccurate , I tried if chloride of potassium could be converted into nitrate of potassa by nitric acid .
A single experiment upon some pure chloride decided in the affirmative ; and the mean result of three experiments gave the ratio of 100 of nitrate to 73'727 of chloride .
The correctness , therefore , of the experiments upon the nitrate was satisfactorily confirmed .
Moreover in another experiment I converted a known weight of nitrate into chloride by hydrochloric acid , and then reconverted the resulting chloride into nitrate by nitric acid .
The quantity of nitrate obtained was very nearly the same as that originally employed ; there was not a hundredth of a grain difference .
The error was thus traced to the equivalent numbers , but whether it existed in one , or more , of the elementary bodies constituting the above compounds , remained still unknown .
The subject cannot be decided by the mere conversion of a nitrate into a chloride , or a chloride into a nitrate .
Additional data are required .
In order therefore to examine the question more scrupulously , the following experiments were undertaken ; and as the results have proved consistent and satisfactory , I have presumed to submit them to the Royal Society .
4 .
In the present communication I propose to examine the equivalent numbers of oxygen , chlorine , nitrogen , potassium , sodium , and silver ; and the following are the successive steps by which this examination has been conducted .
5 .
In the first place the equivalent of chloride of potassium will be determined , bf decomnposing the chlorate of potassa into oxygen , and chloride of potassium .
From the ratio which these results bear to each other , the number for chloride of potassium , as compared with oxygen , may be easily deduced .
The same results will also give the number for chlorate of potassa .
Secondly .
The equivalent of nitrate of potassa is ascertained .
This is effected by converting the chlorate of potassa and the chloride of potassium into nitrate : and as we have already established the equivalents of chlorate and chloride , we can easily calculate the number for nitrate from these two methods of analysis .
Moreover , from the equivalents of chlorate and nitrate of potassa , we can also learn the difference between the equivalents of chlorine and nitrogen , as the two salts agree exactly in coinposition , and only differ by nitrogen in the one being substituted for chlorine in the other .
To confirm the experiments upon the conversion of chloride of potassium into nitrate , I have detaled some upon the decomposition of nitrate into chloride .
The process , hereafter described , by which these experiments were performed , was 14 so exceedingly simple , that the preceding differences could not be referable to errors of manipulation .
Thirdly .
The equivalents of nitrate , chlorate , and chloride of sodium are resolved .
The method of investigation is similar to that adopted for the analogous salts of potassium .
The chlorate of soda has been decomposed into chloride , and into nitrate ; the nitrate of soda into chloride ; and the chloride of sodiu(n into nitrate of soda .
The results of all these experiments are perfectly consistent with each other , and they confirm in the most satisfactory manner the accuracy of the experiments upont the salts of potassium .
Fourthly .
The equivalent numbers of chlorine , nitrogen , potassium , and sodiumn have been ascertained .
For this purpose the metal silver has been selected .
The salts of potassium or sodium do not admit of application in this respect , as it is impracticable to ascertain with sufficient accuracy for the present purpose , the proportions in which either of these metals combine with chlorine or nitric acid .
Bismuth was first selected from the facility of converting it into oxide , but the difficulty of obtaining this metal in a state of absolute purity , and its liability to form subsalts , induced me to prefer silver .
I have accordingly determined the exact proportions in which this metal combines with chlorine and nitric acid ; and lastly , to prove the correctness of the relation which the resulting nitrate bears to the chloride , I have performed several experiments upon the conversion of the crystallized nitrate into chloride .
Perfect consistency exists throughout ; and from these results , conjoined with those from the salts of potassium and sodium , the equivalents of chlorine , nitrogen , potassium , sodium , and silver are easily deduced .
6 .
Such then is the arrangement of the evidence by which I propose to determine the equivalent numbers of these six elementary bodies ; and to show that a slight alteration is necessary in those which are generally employed .
Considering the number of distinguished chemists who have devoted their time and ability to the subject , it might naturally be expected that the truth had been satisfactorily established ; but the slightest observation of the discrepancies in the results obtained , will prove the urgency of more refined investigations .
In offering my feeble evidence , I am truly sensible of the responsible task I have undertaken .
The well-attested skill of those from whom I shall have occasion to differ has constantly urged me to seek every means calculated to attain truth ; and I trust , from the confirmatory nature of the evidence advanced , the numerous experiments performed , and the simplicity of the processes employed , that the results will be entitled to confidence .
7 .
In presenting this communication , I beg to acknowledge , with sinceire grlatitude , the constant assistance I have received from Mr. HIENNELL ; he has kindly watched the whole progress of the investigations , and by his advice has enabled me to overcome several difficulties .
Let it not , however , be inferred that he is responsible for any of the faults which this communication may contain : they must rest entirely on myself .
8 .
The violent effervescence which occurs during the action of nitric acid upon a 15 chloride , or of hydrochloric acid upon a nitrate , except in the case of silver , appeared , at first , to offer an insuperable difficulty to the application of these actions to so delicate a purpose as the determination of equivalent numbers .
It was requisite that some simple practical method should be devised to obviate this source of error .
The unavoidable loss occurring during effervescence has rendered many chemical processes , which promise the most favourable results in theory , inapplicable to important investigations .
In the experiments hereafter to be described , any error from this cause has been effectually prevented .
A loss by spirting cannot occur .
The substance is operated upon in a common flask placed in a horizontal position , with another attached to it as a receiver ; the resulting substance is dried ; and all the weighings are accomplished without removal from the flask .
The performance and completion of the operation in one vessel is an additional advantage , and one which renders the process especially applicable to minute chemical researches .
All those operations , such as transference , washing , filtering , &amp;c. which introduce so many errors , are avoided .
Little comment is required to show the advantages which the above process possesses in comparison to those hitherto applied to the determination of equivalent numbers .
Few improvements , I imagine , can more increase the precision of our analyses , than a reduction of the number of operations by which these analyses are accomplished , combined with the employment of vessels which will bear a high temperature , resist the action of acids , and admnit of being used with the most delicate balances .
All these conditions have been fulfilled by the process employed in these researches , and as it has been applied to all the experiments , I shall premise its details thus early , to avoid unnecessary repetition .
9 .
Into a Florence flask , previously washed , dried , and tared , with the precautions hereafter stated , the substance to be operated upon was introduced , and its weight carefully ascertained .
Another flask , with the whole of its neck removed , was also tared ; and , as this was intended to be adapted to the other as a receiver , it was termed the receiver flask .
The requisite quantity of acid was then poured upon the weighed substance , and the flask immediately placed in a horizontal position .
The receiver flask was also adapted in the same position .
If the decomposition would commence and continue without the assistance of heat , the flasks were allowed to remain as thus placed ; but if heat were necessary , they were transferred without breaking the connexion to a sand bath .
The arrangement is represented in the accompanying figure , and from an inspection of this , it can readily be perceived that a 16 loss by spirting cannot occur .
The drops of liquid , being ejected vertically by the effervescence , are thrown against the upper side of the flask .
Should any drops be thrown laterally , a circumstance of rare occurrence , they will be secured by the receiver .
When the decomposition is completed , and the action has entirely ceased , the receiver is detached , and the solution cautiously evaporated to dryness .
The receiver in the meantime is dried and re-weighed .
During the evaporation of the excess of acid , a portion of the vapour condenses upon the upper side of the flask , and , returning to the solution , carries with it any salt that may have been thrown to the top .
Moreover , from this internal washing of the flask , as it may not inaptly be termed , continuing during the whole of the evaporation , the dry salt , instead of being scattered over a considerable portion of the flask , is reduced to a small compact mass .
Indeed it is beautiful to observe the small compass which the dry nitrates occupy from this cause .
The flask and its contents are lastly heated in the flame of a spirit lamp , and the dry salt either fused or not according to its nature .
After a sufficient time has been allowed for the cooling of the flask , it is returned to the balance , and the weight of the resulting compound carefully ascertained .
10 .
From the preceding process it will be remarked , that flasks are the only vessels which have been employed .
Many precautions must be observed to ensure their successful employment .
Those that 1 have used are of the capacity of a pint , and the lightest were always-selected .
Their average weight was about 320 grains .
Perfect cleanliness was obtained by boiling nitric acid , and well rinsing with common , and afterwards with distilled water .
They were dried by heating on a sand bath , and then in the flame of a spirit lamnp .
I have found it advantageous to remove one half the necks ; for by this , their weights are materially lessened , and their dryness is more securely ensured .
While cooling they were placed inside a glass case , free from dust , and were allowed to remain for two hours , in order to restore the equilibrium of the internal and external air .
In ascertaining the weights of the flasks , it is better to weighl them against another flask , previously selected and dried for the purpose .
This latter flask may be termed the gauge flask .
It weighed 323'420 grains , and weights were of course added , on either side , to adjust the difference between it and the flask to be tared .
The gauge flask was fiequently reheated to insure accuracy , and when not in use was preserved from dust and damp under a bell glass placed inside a glass case .
Before the final observations of the weighings were noted , the flasks were allowed to remain in the balance for a quarter of an hour .
The success of the operations depends especially upon a strict observance of these precautions .
The following circumstance will serve as an ample proof of the care required .
Having observed in one of the experiments a small particle of dust upon the flask , I removed it from the balance , and wiped it as quickly as possible once round with a silk pocket handkerchief .
The flask was held during this operation by the very extremity of the neck .
Upon returning it to the scale it had lost *025 of a grain , and this was owing to a slight expansion of the internal air , produced by the heat of the hand .
It re gained its original weight in less than five minutes .
By repeating this operation several times the flask lost nearly two tenths .
The scales employed were made by Mr. ROBINSON , and would turn decidedly with the thousandth part of a grain , when loaded with five hundred grains .
In the following experiments they were never loaded with so much .
The weights were new , and by careful examination beforehand , were ascertained to be consistent .
Chlorate of Potassa .
ii .
It has already been stated , that the object with this salt , is its decomposition into oxygen and chloride of potassium .
The method of accomplishing this hitherto has been to submit the salt to a high temperature ; by which the oxygen is expelled and chloride of potassium remains .
There are , however , many difficulties which render this process objectionable , and therefore I was induced to inquire whether the same result could be obtained by the action of hydrochloric acid .
The changes occurring during the decomposition are well known .
When the chlorate is in excess , the euchlorine of Sir HUMPHRY DAVY , or , as since proved by SOUBEIRAN , a mixture of chlorine and chlorous acid , is evolved .
If , however , the acid be in excess , chlorine only will be given off .
All the oxygen of the chloric acid , and of the potassa , unites with the hydrogen of the hydrochloric acid , and forms water .
The potassium remains combined with chlorine .
I satisfied myself by numerous experiments that chloride of potassium is alone produced ; and being enabled , by the process already described ( 9 .
) , to counteract any error from the effervescence , I have adopted this method by hydrochloric acid in preference to that by heat .
12 .
The chlorate of potassa employed in the succeeding experiments was obtained pure by three successive crystallizations of ordinary commercial chlorate .
After each crystallization the crystals were dried , rubbed to powder , and washed with distilled water .
The salt which resulted from the third crystallization , dissolved without residue in water , and the solution was not affected by nitrate of silver or nitrate of baryta .
Two hundred grains were converted by heat into chloride of potassium ; the fused chloride was perfectly colourless , and when dissolved and tested not a trace of foreign matter could be detected .
13 .
The first experiments with this salt were rather discrepant , arising from my ignorance of the extreme caution which is requisite in drying it .
Having observed in some chemical works that it is fusible at a temperature between 400 ?
and 500 ?
FAHR. , and will bear a temperature of 6600 without undergoing decomposition , I conceived that very careful fusion would be the most secure method for ensuring dryness .
The want of uniformity in the results , however , indicated some source of error , and upon examination I found that a very notable quantity of chloride of potassium was produced by the process of fusion , although effected with the utmost care .
A temnperature just below that of fusion was next tried , but upon dissolving the dry salt in distilled water , and testing with nitrate of silver , a minute portion of chloride could 18 be detected .
Indeed it was not until after several trials that I succeeded in ascertaining the temperature at which this salt might be perfectly dried withouc risk of decomposition .
This temperature is about 220 ? .
The temperature of boiling water is quite safe : 1000 grains were kept in a copper pan , heated by steam , for fourteen hours without experiencing the slightest change .
I have found by several experiments that one hundred grains of perfectly dried chlorate will lose about '05 of a grain by fusion .
The precaution of testing a portion of the dry salt was repeated after every drying , to ensure accuracy .
14 .
When dry , the powder was transferred , while hot , into a small tube , and tightly corked .
As soon as cold , the quantity necessary for an experiment was transferred into a previously tared flask , and the weight of it ascertained .
The tube was sufficiently small to pass down the neck of the flask , and almost to the bottom , so that any adhesion of the powder to the sides of the flask was prevented .
The remainder of the process corresponded to the one already described , except that before adding the hydrochloric acid , it is expedient to fuse the chlorate in the flask .
By this the action is very much moderated , and the chlorate is slowly and quietly dissolved .
No heat is requisite if the acid be sufficiently strong .
15 .
the hydrochloric acid was prepared according to the general directions in a green glass retort .
It was twice redistilled , and had a specific gravity of 1'160 .
It was carefully examined for sulphuric , nitric , and sulphurous acids , but not a trace of them could be detected .
Two fluid ounces when evaporated left only '015 of a grain of solid matter* .
16 .
Owing to the compactness of the mass in which the resulting chloride is left in the flask , a minute portion of hydrochloric acid remains adherent to the salt when its quantity is large ; and this cannot be driven off without so high a temperature as to risk the cracking of the flask .
To expel this adventitious acid , I have always adopted the following process .
The mass , the weight of which has been ascertained , is detached from the side of the flask by gentle agitation , and broken into pieces sufficiently small to be removed .
These pieces are then rubbed to powder in a glass mortar , and the powder is transferred into a previously tared platina crucible .
Its weight is ascertained , and then it is heated to dull redness for half an hour .
Fusion was carefully avoided , as it cannot be effected without a minute loss by sublimation .
When cold the loss is easily determined , and as the weight of the salt in the crucible is known , it is easy to calculate the loss for the whole quantity originally in the flask .
The quantity which remained adherent to the flask generally amounted to '07 or '08 of a grain .
17 .
I subjoin the results of six experiments , and annex , for the convenience of comparison , the calculations to 100 parts .
Chlorate .
Chloride .
Chlorate .
Chloride .
76-626 ... 46-598 ... or as 100 to ... 60-825 82'048 ... 49'903 ... or as 100 to ... 60-822 75-200. .
45733 ... or as 100 to ... 60'815 63'114 ... 38'386. .
or as 100 to. .
60'820 61-164. .
37'202. .
or as 100 to. .
60'823 65-724. .
39-980 ... or as 100 to ... 60'830 The mean of these experiments indicates that 100 parts of chlorate of potassa contain 60-823 of chloride of potassium , and 39'177 of oxygen .
18 .
In another experiment I obtained the proportion of 100 of chlorate to 60-808 of chloride ; but in this case the chloride was partially fused , and while so I detected the evolution of a small quantity of white vapour .
It shows that 60'808 is too low .
BERZELIUS obtained by heat 60,850 of chloride , and 39*150 of oxygen from 100 of chlorate .
19 .
It being well known that many powders , although not deliquescent , are more or less hygrometric , and no precautions having been taken to obviate this in the preceding experiments , I thought an error on this account might influence them .
To decide the query , the following experiment was performed .
Two hundred grains of the chlorate in powder were dried with every care , and transferred while hot into a stoppered phial , just large enough to contain it .
When quite cold , the stopper was removed for a moment , and then the bottle and its contents , being placed in a small glass dish , the whole was counterpoised in the balance .
The powder was transferred very carefully into the dish , without removal from the balance , and the empty bottle placed on top of the powder .
The whole was left in this condition for two hours , and upon examination at the expiration of that time , not the slightest appreciable difference in weight could be detected .
The experiment was twice repeated with the same result .
20 .
Perfectly dry powdered chloride of potassium was submitted to a similar experiment , but there was no evidence of change .
If however the atmosphere be damp , both these salts when in powder rapidly absorb hygrometric moisture .
Conversion of Chlorate of Potassa into Nitrate .
21 .
Having determin2d , and I trust satisfactorily , that every 100 parts of chlorate correspond to 60,823 of chloride , the next object was to ascertain the quantity of nitrate resulting from 100 parts of the same salt .
Its conversion into nitrate cannot however be performed directly , for the action of nitric acid upon a chlorate , as I shall show at another opportunity , is not to decompose it wholly into nitrate , but into a definite mixture of nitrate and perchlorate .
The means by which the object now desired was accomplished , were by converting the chlorate into chloride , by hydrochloric acid , and then the resulting chloride into nitrate , by nitric acid .
The operations were performed without any removal from the flask .
The nitric acid was 20 added immediately the chloride was dry , without transferring and driving off the small quantity of adhering acid ( 16 .
) .
The minuteness of its quantity rendered it of no importance , as considerable excess of nitric acid was always employed .
The nitrate produced was always fused to ensure dryness , and its weight being ascertained , it was dissolved and tested for chlorate and chloride .
There was never the slightest indication of either .
The impurities introduced by the acids , namely , one ounce of hydrochloric , and half an ounce of nitric , were of course allowed for .
Only four experiments were performed , as their uniformity assured me of their close approximation to the truth .
22 .
The particulars of them are as follows : Chlorate .
Nitrate .
Chlorate .
Nitrate .
86'454 ... .
71-329 ... .
100 ... .
82-505 70-205 ... .
57'918 ... .
100 ... 82'497 62-244 ... .
51368 ... .
100 ... .
82-498 72-636 ... .
59-938 ... .
100. .
82-500 The mean will be as 100 chlorate to 82'500 of nitrate .
23 .
Hereafter I shall show that 100 parts of nitrate correspond to 73-726 of chloride , so that the above result would indicate that 100 of chlorate contain 60-825 of chloride .
This agrees very closely with the experiments by hydrochloric acid .
24 .
Moreover , this result , by comparison with that obtained in 17 , shows , that 100 parts of chloride correspond to 135'640 of nitrate and 100 parts of nitrate correspond to 73-724 chloride .
Chloride of Potassium .
25 .
Notwithstanding the experiments which have been made upon the action of nitric acid upon chlorides , much uncertainty still exists , as to the precise nature of the changes which occur .
Their explication would dispel the doubt respecting the changes which take place during the mutual decomposition of nitric and hydrochloric acid , in the formation of nitro-muriatic acid .
Sir II .
DAVY has stated that chlorine , nitrous acid , and water are produced .
According to Dr. JOHN DAVY , a compound of chlorine and binoxide of nitrogen , with chlorine , and water result .
Interesting as the decision of this question would be , I shall not now adduce the experiments I have made for this purpose , lest I should confuse the present inquiry .
It has reference only to the ultimate result , which , as before stated ( 3 .
) , is proved by the following experiment to be pure nitrate of potassa .
26 .
Two hundred grains of pure fused chloride of potassium were acted upon by one fluid ounce of nitric acid , sp. gr. 1425 ; effervescence began immediately on mixture : abundance of chlorine tinged with nitrous acid was evolved .
After the action had continued for two hours , a gentle heat was applied .
There was a more copious evolution of nitrous acid , but the action soon ceased , and the solution remained colourless .
Evaporation was carried to dryness , and the dry salt fused .
It was colourless , and weighed about 271-24 .
This corresponds closely to the quantity of nitrate which should be yielded according to the preceding experiments .
Part of the fused salt was then dissolved in distilled water , and tested with nitrate of silver : not the slightest turbidity was evident .
To another portion hydrochloric acid was added , to observe whether any chlorate of potassa had been formed , but not the slightest change was produced .
The remaining portion upon being crystallized proved to be all nitre .
27 .
The chloride employed was prepared by saturating pure bicarbonate of potassa with excess of pure hydrochloric acid .
The bicarbonate was purified by three crystallizations , and when tested no indication of foreign matter could be detected .
The solution of chloride was evaporated to dryness , and the dry salt heated to expel the excess of acid .
It was then dissolved in distilled water , and twice recrystallized .
The salt , thus purified , was examined for sulphate , nitrate , lime , &amp;c. , but no impurity could be discovered .
It was dried by careful fusion in the platina crucible .
The process has already been detailed .
Only half a fltlid ounce of the nitric acid was employed in the following experiments .
This nitric acid had been twice distilled and was perfectly colourless .
It had a specific gravity 1-430 .
Two fluid ounces when evaporated left *020 of a grain , and therefore not more than 005 of a grain of impurity would be introduced .
28 .
The following are the particulars of the experiments performed : Chloride .
Nitrate .
Chloride .
Nitrate .
84'090 ... .
113'058 ... .
100 ... .
135-639 61'998 ... .
84*092 ... .
100 ... .
135-637 51 823 ... .
70-293 ... .
100 ... .
135'640 67 145 ... .
91'072 ... .
100 ... .
135'635 55-350 ... .
75-071 ... .
100 ... .
135'630 whence , according to the mean , 100 of chloride correspond to 135'634 nitrate , and 100 of nitrate correspond to 73'726 chloride .
29 .
In confirmation of these results , I procured some chloride from pure chlorate , by heating in the platina crucible , and fusing the resulting salt .
I performed two experiments with this chloride and obtained the annexed results : Chloride .
Nitrate .
38'660 ... .
52'438 ... .
100 ... .
135-640 60-998 ... .
82732 ... 100 ... .
135-630 They correspond very closely to the preceding .
Nitrate of Potassa .
30 .
The conversion of this salt into chloride of potassium has been described , as well as the process by which it is effected .
In addition to the precautions mentioned 22 in the process , there are several others to be observed to ensure success .
As soon as the action commences , the flask should be removed from the sand bath , and the decomposition be permitted to continue without heat .
When the action slackens , the flask may be returned to the sand bath , and so on , to and from the source of heat , until it is so gentle that no lateral spirting can be feared .
The adaptation of a receiver throughout the whole of these operations must be strictly attended to ; for despite every care , a minute portion of the solution will sometimes be ejected during the evaporation .
This circumstance happened in three of the experiments I shall mention presently .
In one case the weight was '003 of a grain , in another *008 , and in a third '005 .
The receiver flask should not be detached from the other even while removing them to and from the sand bath , nor in fact until the resulting chloride is nearly dry , as slight action continues during the greater part of the evaporation .
31 .
Very particular attention was paid to the purity of the salt employed .
It was prepared from the common purified nitre of the shops , by four repeated crystallizations from distilled water .
After each crystallization the salt was dried , and finely powdered , and this powder well washed with distilled water .
When finally washed and dried , 200 grs. were fused in a platina crucible .
They yielded a perfectly colourless mass , soluble in water without a trace of insoluble matter , and the solution was not affected by nitrate of silver , chloride of barium , oxalate of ammonia , or hydrosulphuret of ammonia .
It was dried by careful fusion .
The resulting chloride was dried in the same way as the chloride from chlorate .
From one to two fluid ounces of hydrochloric acid are sufficient for the quantities of nitrate employed .
In adding this acid one important practical circumstance must be observed , namely to add it in two successive portions , and to expel the first quantity previously to the addition of the second .
This method is necessary from the small size of the flask , compared with the bulk of fluid .
The less the quantity of fluid in the flask at one time , the less chance of loss by spirting .
32 .
1 obtained , in seven experiments , the following very satisfactory results : Nitrate .
Chloride .
Nitrate .
Chloride .
60-290 ... 44'452. .
100 ... 73731 58-230. .
42-927 ... 100 ... .
73-726 63-040 ... 46-479 ... 100 ... .
73-730 51514 ... .
37'978 ... 100 ... .
73-724 56-036 ... .
41'315 ... .
100 ... .
73'730 80-517 ... .
59'362 ... .
100 ... .
73'726 78-643 ... .
57-983 ... .
100 ... 73'730 The mean will be as 100 of nitrate to 73'728 chloride , or 100 of chloride to 135'634 nitrate .
23 Thus then we have ascertained , by three different methods , the ratio which chloride of potassium bears to nitrate of potassa .
The results are as follows : from chlorate potassa ( 17 .
) , Nitrate .
Chloride .
Chloride .
100 corresponds to 73'724 ; or 100 to 135'640 .
From chloride potassium ( 28 .
) , 100 corresponds to 73'726 ; or 100 to I35'634 .
From nitrate of potassa ( 32 .
) , 100 corresponds to 73'728 ; or 100 to 135'633 .
Whence the mean will be , 100 corresponds to 73'726 ; or 100 to 135'636 .
34 .
Before quitting this salt , I cannot omit the opportunity of mentioning an instructive circumstance connected with my first experiments .
The hydrochloric acid employed was obtained pure ( as I then thought ) by the redistillation of some ordinary pure acid .
It was tested for nitric and sulphuric acids , but there was not a trace of either .
Two fluid ounces yielded by evaporation only '014 of impurities .
I performed four operations , and the results were in the ratio of 100 of nitrate of potassa to 7387 73S 7373877 73 73886 877of chloride .
The close approximation of these numbers inspired considerable confidence at the time , and in order to verify them , I converted a portion of pure chloride of potassiurn into nitrate .
The result was as 100 of nitrate to 73'728 chloride .
The difference between this and the above result indicated some source of error .
The chloride , resulting from a nitrate experiment , was therefore reconverted into nitrate by nitric acid , but instead of obtaining the original quantity , I obtained much less .
This circumstance suggested , that the error arose from some impurity in the materials , and suspecting the hydrochloric acid , a portion of it was submitted to another distillation .
With this new acid another portion of nitre was decomposed , but the result was worse instead of better .
The proportion was as 100 of nitrate to 74-04 of chloride .
It was evident that some adventitious matter , more volatile than hydrochloric acid , was passing over with the acid in distillation .
Sulphurous acid was suspected .
A portion of the hydrochloric acid was boiled with a little nitric acid , diluted , and then tested with muriate of baryta ; a precipitate of sulphate of baryta deposited .
Therefore the acid did contain sulphurous acid , and this , by the action of nitric acid or a nitrate , was converted into sulphuric .
The production of a higher result is therefore explained , and although vexatious at the time , it will now show the importance of obtaining the same result from different sources , and how admirably these reciprocal decompositions , of chlorides into nitrates and nitrates into chlorides , are adapted for this purpose .
24A Chlorate of Soda .
35 .
This salt has been very imperfectly examined .
It is recommended to be prepared by passing chlorine through a solution of soda , and separating the chlorate and chloride by crystallization , or by saturating chloric acid with carbonate of soda .
These processes , however , are very uncertain , and this may explain the contradictory descriptions given of this salt in chemical works .
The process bywhich I succeeded in obtaining it pure , is as follows : some bitartrate of soda was prepared from carbonate of soda and tartaric acid , and purified by recrystallization .
A quantity of pure chlorate of potassa was dissolved in a sufficient quantity of hot distilled water , and this solution while warmn was mixed with a saturated hot solution of the above bitartrate of soda .
Instant decomposition occurred ; cream of tartar was precipitated , and chlorate of soda remained in solution .
The whole was permitted to cool , and when quite cold the solution was decanted , evaporated , and crystallized .
It is better to employ excess of chlorate of potassa : the proportions which I used were 32 of chlorate and 40 of bitartrate in crystals .
The first crop of chlorate of soda generally contains a little bitartrate , but this may easily be separated by cold water , which dissolves the former and leaves the latter .
The salt employed for the present inquiry was dissolved in spirit of wine , and afterwards twice recrystallized from distilled water .
It was at last crystallized by spontaneous evaporation .
The crystals thus obtained were very large .
The diagonals of the largest face of many were not less than 4 of an inch in length .
They were either right square prisms or right rectangular prisms .
I have never obtained them of any other shape .
They contain no water of crystallization , and riemain unchanged by exposure to the air : they fuse at a temperature a little higher than that required by chlorate of potassa ; and , like this salt , they always undergo a slight decomposition by fusion .
36 .
Particular attention is required in concentrating the solutions of this salt ; for if the heat be carried too far decomposition will take place .
Oxygen in abundance is evolved , and chloride of sodium formed .
When this has happened the chlorate is spoilt , for the almost equal solubility of the chlorate and chloride renders it impossible to effect their perfect separation .
The criterion by which I judge of the proper strength of a solution , is blowing gently on the surface , and observing whether any crystals are formed ; if so the heat may be removed .
37 .
The process for decomposition is the same as that employed for chlorate of potassa .
In drying the resulting chloride of sodium , care must be taken to avoid a loss by decrepitation .
38 .
Isubjoin the results of four experimentswith the calculation to one hundred parts , Chlorate .
Chloride .
Chlorate .
Chloride .
81816 ... .
44-950 ... 100 ... .
54-940 75'010 ... .
41-199. .
100. .
54'925 116'655 ... 64'067 ... .
100 ... .
54920 130'705 ... 71800 ... .
100 ... .
54'933 The mean will be very nearly 100 of chlorate to 54'930 of chloride , The chlorate was next converted into nitrate by a process similar to that adopted for the analogous salt of potassa .
I performed three experiments , and the following are the particulars .
Chlorate .
Nitrate .
Chlorate .
Nitrate .
110'843 ... .
88-536 ... .
100 ... .
79-875 118'290 ... .
94-492 ... .
100 ... .
79-882 89-112 ... .
65-198 ... .
100 ... .
79'890 The mean will be as 100 of chlorate to 79'882 chloride .
And comparing this with the experiments on the conversion of chlorate into chloride , 100 parts of chloride will correspond to 145'425 nitrate , and 100 parts of nitrate will correspond to 68*762 chloride .
Nitrate of Soda 40 .
Was prepared from pure carbonate of soda and nitric acid .
The carbonate was obtained pure by repeated crystallizations of common carbonate .
It was obtained in fine large crystals , and 200 grains when dried and fused yielded a colourless mass .
This mass was then dissolved in distilled water and tested for chlorides , sulphates , lime , iron , &amp;c. , but no trace of impurity could be detected .
The crystals were dissolved and saturated with pure nitric acid in excess ; and the solution evaporated to dryness .
The dry salt after heating to expel the excess of acid , was recrystallized three times .
The resulting nitrate was quite pure , and was dried previously to use , by fusion in the platina crucible .
41 .
Its decomposition into chloride of sodium was effected by the same process as that applied to nitrate potassa .
In drying the resulting chloride care must be taken to prevent a loss from decrepitation .
Fusion must also be avoided , as the salt rapidly sublimes when fused .
The hydrochloric acid was added in two portions , and from one ounce to one ounce and a half was sufficient for the quantities of nitrate employed .
42 .
I subjoin the results of six experiments with the calculations to 100 parts .
Nitrate of Soda .
Chloride Sodium .
1 .
56'606 ... 38'926 ... .
100 ... 68-767 2 .
59-220 ... .
40-731 ... .
100 ... .
68'780 3 .
50'316 ... .
34'602 ... .
100 ... .
68'770 4 .
51'794 ... .
35618 ... .
100 ... .
68769 5 .
55-250 ... .
37-995 ... .
100 ... 68-770 6 .
68-430 ... 47-059 ... .
100 ... .
68770 The mean of these results give the following proportions .
100 parts of nitrate soda correspond to 681771 chloride sodium , and 100 parts of chloride sodium correspond to 145'410 nitrate .
26 Chloride Sodium .
This salt was prepared by supersaturating pure carbonate of soda with pure hydrochloric acid .
The solution was evaporated to dryness , and the dry salt , after heating to expel the excess of acid , was redissolved and recrystallized several times .
The crystals obtained were quite pure , and were fused in the platina crucible to ensure dryness .
It was converted into nitrate of soda by the same method as chloride of potassium .
The resulting nitrate was fused in the flask .
The following are the results of seven experiments .
Number 3 .
was performed upon some chloride of sodium obtained from chlorate , by heating until all the oxygen was expelled .
Number 7 .
was performed upon chloride in powder , dried in the platina crucible by the flame of a spirit lamp .
Half an ounce of the nitric acid ( 27 .
) was sufficient for all the following experiments .
Chloride .
Nitrate .
Chloride .
Nitrate .
1 .
43'758 ... 63630 ... .
as 100 to ... .
145'415 2 .
50-454. .
73-465 ... .
as 100 to ... .
145408 3 .
38-160 ... .
55'492 ... .
as 100 to ... .
145-420 4 .
48-914 ... .
71'112 ... .
as 100 to ... .
145'424 5 .
56-240 ... .
81'778. .
as 100 to ... .
145'410 6 .
58-942 ... .
74021 ... .
as 100 to ... .
145418 7 .
99 168 ... 144'210 ... .
as 100 to ... .
145'420 whence taking the mean 100 parts of chloride will correspond to 145'416 nitrate , or 100 parts of nitrate will correspond to 68'768 chloride .
43 .
The conversion of chloride of sodium into nitrate of soda , affords one of the best and easiest methods for testing the accuracy of the equivalent numbers in general use .
According to one series we have 60 as the number for chloride of sodium , and 86 for nitrate ; another series gives 58'72 , and 85-45 for the same compounds .
Now taking these two series as the data of separate calculations , we find , that , according to the former , a hundred grains of chloride should yield 143*333 grains of nitrate ; whereas the latter would give 145*504 grains of nitrate from the same quantity of chloride .
Here then is a difference of nearly 2-2 grains ; -a difference far exceeding any error of manipulation , and plainly indicating that one or other of these series must be incorrect .
Upon which series the greater error is chargeable , every chemist may satisfy himself .
The decomposition by nitric acid may be readily performed in flasks ; and the only requisite for success is purity of materials .
Silver .
44 .
The object for which the following experiments were made upon this metal has been already stated ( 5 .
) .
The conversion of silver into chloride has been so frequently repeated by the most skilful chemists , that I conceived the better plan would E2 27 be to collate their results , and to take the -mean as the basis of my calculation .
But the discrepancies existing between them , and the objectionable processes which have in many cases been applied , urged me to examine the subject for myself .
45 .
The purity of the silver employed was very rigorously tested .
One hundred and fifty grains were dissolved in pure dilute nitric acid ; not a trace of insoluble black matter indicative of gold could be recognised .
The nitric solution , being di.luted , was precipitated by excess of pure hydrochloric acid ; and the supernatant liquor , separated from the precipitate by filtration , was evaporated to dryness .
A very minute quantity of solid matter remained ; equivalent only to the impurities introduced by the acids and the water .
It was examined for copper , lead , and iron , but not the slightest indication of their presence was afforded .
The silver was in small pieces , and each piece was sufficient for an experiment .
Conversion into Nitrate .
46 .
The process for the conversion into nitrate was in all particulars identical to the one described in paragraph number 9 .
One fluid drachm of the pure nitric acid ( 27 .
) diluted with two drachms of distilled water was amply sufficient for the quantities of silver employed .
With acid thus diluted the action is very gentle , and , if time be allowed , the whole of the silver may be dissolved withlout the assistance of heat .
During the evaporation of the water and the excess of nitric acid , the receiver flask may be detached without any risk of loss .
The dry nitrate was in all cases fused ; but particular care is necessary to accomplish this operation successfully , for the salt being in a small cake , is , from the shape of the flask , much thinner at the edges than in the middle ; and therefore , unless the heat be cautiously applied , a slight reduction into oxide will happen .
This effect becomes evident by the fused nitrate being tinged with black .
It only happened in one of the following experiments , and then a very minute black film could be detected on one side of the fused mass .
During fusion and cooling , the flasks were always excluded from light .
47 .
The distilled water employed in these experiments had been twice redistilled in green glass , and four fluid ounces when evaporated left '024 of a grain of solid matter .
The quantity used , namely two drachms , would not introduce therefore a thousandth of a grain of impurity .
48 .
I subjoin the results of six from nine experiments , with the calculations to 100 parts .
Silver .
Nitrate .
Silver .
Nitrate .
1 .
59'087 ... 93*020 ... .
or as 100 to ... 157*430 2 .
60*311 ... 94-952. .
or as 00 to ... 157437 3 .
51-654 ... 81'333 ... .
or as 100 to ... 157'458 4 .
55'734 ... 86'765 ... .
or as 100 to ... 157-440 5,.45-622 ... 71'827 ... .
or as 100 to ... 157-430 6 .
64'726. .
101-913 ... .
or as 100 to ... 157'455 whence , taking the mean , 100 parts of silver correspond to 157,441 of nitrate .
28 Conversion into Chloride .
49 .
The quantity of chloride resulting from a known weight of silver was determined as follows .
To each of the fused nitrates from several of the foregoing experiments ( 48 .
) , half a fluid ounce of distilled water ( 47 .
) was added , and the salt dissolved with the assistance of a gentle heat .
The solution being slightly acidulated with nitric acid , was precipitated by an excess of pure hydrochloric acid , amounting to half an ounce .
These operations were performed without removal from the flask .
The precipitated chloride , notwithstanding constant agitation during the addition of the hydrochloric acid , forms into a cake , and lest this should retain any undecomposed nitrate of silver , a glass rod of a dark green colour was introduced and the mass broken up .
The rod , on removal from the flask , was always examined with a lens to see whether any chloride had adhered .
The flask was next placed upon a heated sand bath , and the water and excess of acid cautiously expelled .
The resulting chloride was beautifully white , and was always fused in the flask .
During this operation , and while cooling , light was carefully excluded .
50 .
In the Philosophical Transactions for 1833 , the late and inuch-lamented Dr. TURNER has stated that chloride of silver may be dried perfectly and without the risk of decomposition at a temperature of 300 ?
FAHR. ; but he adds that even after this a slight loss occurs during fusion .
With the latter part of these remarks my observations accord .
In several experiments in which the chloride has been dried for a considerable time with a spirit lamp , guarding carefully against fusion , I have found that thus dried it always decreased slightly in weight by fusion , amounting to about -006 of a grain for 70 grains .
Dr. TURNER attributed this loss to a slight decomposition .
I have repeatedly tried to satisfy myself on this point , but have hitherto been unsuccessful .
I have never seen any evidence of decomposition .
51 .
In five experiments performed as already stated , I obtained the following results .
Silver .
Chloride .
Silver .
Chloride .
1 .
59-087 ... 78489 ... or 100 to ... 132'836 2 .
60'311 ... 80'117. .
or 100 to ... 132-840 3 .
51-654 ... 68612 ... or 100 to ... 132830 4 .
55-734 ... 74037 ... or 100 to ... 132'840 5 .
64-726 ... 85-982 ... or 100 to ... 132840 52 .
Again , in two additional experiments I dissolved the silver in a sufficient quantity of dilute nitric acid , and precipitated at once with hydrochloric acid , without evaporating and procuring the nitrate .
These are the particulars of the experiments .
Silver .
Chloride .
57'882 ... .
76-884 ... .
100 ... .
132-830 51'380 ... .
68*252 ... 100 ... .
132'838 Taking the mean of the seven experiments , 100 parts of silver will correspond to 132-836 of chloride .
29 The mean of three experiments , described by Dr. TURNER in the Philosophlical Transactions for 1829 , gives the ratio of 100 of silver to 132'835 of chloride .
The following table , extracted from Professor BRANDE'S Manual of Chemistry , will show the results of different experimenters upon the same subject .
Dr. THOMPSON. .
100 of silver to 133'333 of chloride .
ROSE ... ... . .
100 of silver to 133*014 of chloride .
BERZELIUS ... ... .
100 of silver to 132'736 of chloride .
GAY LuSSA ... .
100 of silver to 132'890 of chloride .
MARCET and DAVY. .
100 of silver to 132'450 of chloride .
54 .
From the five experiments on the chloride , we can also ascertain the ratio of nitrate of silver to chloride .
For as the silver employed had been previously converted into nitrate , we can , by comparing the mean result from the nitrates with that from the chloride , obtain the ratio required .
Thus every 100 parts of silver correspond to 132'836 of chloride and to 157*441 of nitrate ; whence , 100 of chloride will correspond to 118'523 nitrate , and 100 of nitrate will correspond to 84'372 chloride .
Nitrate Silver .
55 .
That 100 parts of nitrate to 84 372 of chloride is near the truth , the following experiments on the crystallized nitrate satisfactorily confirm .
The nitrate of silver was obtained from pure commercial silver by solution in nitric acid and crystallyzing .
After three crystallizations it was obtained in fine bold crystals , and was quite pure .
Its perfect dryness was ensured by fusion in a small glass tube .
To convert it into chloride , the same process and precautions were adopted as have already been described .
The chloride was always fused .
56 .
I subjoin the results of five experiments .
Nitrate .
Chloride .
Nitrate .
Chloride .
93-452 ... .
78847 ... .
100 ... .
84370 115'414 ... .
97-395. .
100 ... .84'388 65500 ... .
55'267 .
100 ... .
84'377 93'034 ... .78'490 ... .
100 ... 84'367 108'645 ... 9 1664 ... .
100 .
84370 The mean of these experiments will be 100 of chloride to 1 18'520 nitrate , and 100 of nitrate to 84-374 chloride .
57 .
Dr. TURNER , in the Philosophical Transactions for 1829 , gives the details of two experiments upon the conversion of nitrate of silver into chloride .
According to one , 100 parts of nitrate correspond to 84'357 of chloride ; and to the other , 100 to 84'389 .
The mean will be 100 to 84'373 .
The close accordance of the above results 30 not only proves the correctness of the experiments , but also the purity of the materials .
58 .
It now remains to determine from the foregoing experimental data , the equivalent numbers of the several compounds described , and of the elementary bodies of which these compounds are constituted .
For the convenience of reference , I have arranged the results in a tabular form , and have included the results corresponding to the two series of numbers in general use in this country .
I have taken the liberty to name them according to their respective authors .
Calculated Calculated According to according to according to the foregoing Dr. THOMPSON .
Dr. TURNER .
experiments .
100 parts of chlorate of potassa yield of chloride ... .
61-290 60*838 60-825 100 parts of chlorate of potassa yield of nitrate ... ... 82-259 82-646 82'500 100 parts of chloride of potassium yield of nitrate ... .
134-210 135-845 135-636 100 parts of nitrate of potassa yield of chloride ... ... 74-510 73'613 73-726 100 parts of chlorate of soda yield of chloride ... ... 55-555 55*022 54-930 100 parts of chlorate of soda yield of nitrate ... ... . .
79-630 81'942 79-882 100 parts of chloride of sodium yield of nitrate ... ... 143-333 145-504 145-414 100 parts of nitrate of soda yield of chloride ... ... . .
69-767 68-666 68-771 100 parts of silver yield of nitrate ... ... ... ... . .
156-363 157'546 157-441 100 parts of silver yield of chloride ... ... ... ... ... 133'333 132*796 132*837 100 parts of nitrate of silver yield of chloride ... ... . .
85-272 84-290 84-374 100 parts of chloride of silver yield of nitrate ... . .
117'808 118-637 118-520 59 .
Of the six elementary substances included in the above compounds , there is only one upon the number of which chemists are agreed .
That is oxygen .
Upon a scale in which hydrogen is considered as unity , oxygen is stated to be eight ; and as this number is generally employed in this country , we may adopt it as the basis of the calculations in the present inquiry .
Should it be proved erroneous by future investigations , the numbers , presently deduced , must undergo a proportionate alteration .
Remembering , therefore , that I have merely assumed the number eight for oxygen to be correct , we may proceed according to the method already detailed ( 5 .
) to ascertain the several equivalents .
60 .
Chloride of Potassium .
According to the preceding table 100 parts of chlorate of potassa correspond to 60,825 of chloride , and therefore the same quantity of chlorate will contain 39'175 oxygen .
But in every proportional of chlorate , there are admitted to be six proportionals of oxygen ; whence as 39*175 : 60*825 : : 48 : to the equivalent of chloride of potassium .
By performing the operation we obtain the number 74'527 .
Chlorate of potassa will therefore be 74'527 + 48 = 122'527 .
Nitrate of potassa is determined by the ratio which chloride bears to nitrate , thus according to the table 100 of chloride correspond to 135,636 of nitrate ; and therefore as 100 : 135'636 : : 74'527 : 101'087 , which will be the equivalent of nitre .
Moreover from what has been stated ( 5 .
) we learn , that 122,527 , the equivalent of chlorate , minus 101-087 the equivalent of nitrate , equals the difference between chlorine , and nitrogen , namely 21"440 ; -and also that 101U087 , minus 74-527 the number for 31 chloride , is the difference between one proportional of nitrogen plus six of oxygen minus chlorine : namely 26*560 .
61 .
By performing the same method of calculation with the results from the salts of sodium , we have as follows : Chloride of sodium ... ... ... . .
58500 Chlorate of soda ... ... .
106'500 Nitrate of soda ... ... ... ... . .
85068 Chlorine nitrogen ... ... 21'432 Nitrogen +6 oxygen chlorine ... .
26'568 The close accordance of these latter differences with those from the salts of potassium cannot fail , I think , to inspire confidence in the accuracy of the experiments .
62 .
From these differences we may readily determine the numrnbers for the salts of silver as well as those for chlorine and nitrogen .
Thus , according to the table , 132 837 of chloride of silver correspond to 157*441 of nitrate : the difference being 24'604 .
But we have ascertained that the difference between a nitrate and a chloride is 26'565 , and therefore as 24,604 : 157,441 or 132'837 : : 26565 : to the equivalent of chloride or of nitrate of silver .
Whence the nitrate equals 169*989 and the chloride 143*424 ; and as 132*837 of chloride contain 100 of silver , therefore the number for silver is 107'970 .
63 .
Nitrogen will be 14'02 and chlorine 35'454 .
Potassium therefore is 39'073 , and sodium 23'046 .
64 .
For the convenience of reference I have subjoined the numbers of these elementary bodies in the following table ; with the numbers given by the two distinguished chemists whose names I have had occasion to mention .
THOMPSON .
TURNER .
PENNY .
Oxygen ... .
8 ... .
8 ... .
8 Chlorine .
36 ... .
3542 ... .
35'45 Nitrogen. .
14 ... 4 ... .
1415 ... .
1402 Potassium ... 40 ... .
39'15 ... .
39'08 Sodium .
24 ... .
23'3 ... .
23*05 Silver ... 110 ... .
108'0 ... .
107'97 65 .
These researches corroborate the conclusions to which Dr. TURNER was led by his experiimental inquiries , published in the Philosophical Trfansactions for 1833 .
They show that the estimates in general use among British chemists are not the strict representatives of chemical truth , founded on experiment ; and that the favourite hypothesis , of all equivalents being simple multiples of hydrogen , is no longer tenable .
My estimates of chlorine and silver correspond very closely to those of Dr. TURNER .
His number for nitrogen was deduced from experiments on the nitrates of silver , lead , and baryta .
He obtained the numbers 14*09 , 14'17 , 14'2 , and he adopted 14'15 as the mean .
I have carefully re-examined my experiments , but I cannot discover 32 anything to justify an alteration of the number 14*02 .
The experiments upon the conversion of silver into nitrate have satisfied me that 14*15 for nitrogen is inaccurate ; for by reference to the table in page 31 , it will be seen that according to this number I should have obtained 157*546 of nitrate from 100 of silver , whereas I obtained 157*441 .
The difference , namely , 0'1 , is too high to be referable to errors of manipulation , considering the simplicity of the process employed .
Dr. TURNER'S numbers for potassium and sodium are not founded upon any experiments of his own , but are extracted from BERZELIUS'S Tables of Equivalents .
66 .
In conclusion , I may mention the several additional subjects which the present inquiry has elicited for investigation .
I have extended the decomposition of nitrates into chlorides , and chlorides into nitrates , to the salts of baryta , lime , strontia , manganese , and lead ; but whether these salts are applicable to the deternination of equivalent numbers , must be decided by future experiments .
67 .
I have also satisfied myself , that the same method of investigation may be applied to several iodates and iodides , bromates and bromides .
They are converted into chlorides by hydrochloric acid ; and the iodides and bromides may be converted into nitrates by nitric acid .
These decompositions afford excellent means of determining the equivalents of iodine and bromine , and of corroborating the experiments already described upon the nitrates and chlorides .
For suppose that two equal quantities of pure iodate or bromate of potassa are converted into chloride and nitrate , it is obvious that the quantities of these resulting salts should bear the same ratio to each other as that established in the present paper .
The same ratio should result from the iodides and bromides , if the experiments be correctly performed .
Opportunity , however , will not permit me at present to complete the experiments undertaken upon these salts .
The special care required in the preparation of the materials , the number of experiments to be performed , and the extreme delicacy necessary to attain truth , would occupy so much time , that I have deemed it better to communicate the results already obtained , than to wait for the completion of the whole .
Moreover the details would far exceed the just limits of a single communication .
68 .
I have likewise examined the action of nitric acid upon chlorates and iodates , and have obtained some novel and interesting results .
The process so frequently referred to has also enabled me to make some satisfactory experiments upon the conversion of carbonate of soda into nitrate and chloride , and to obtain some important evidence respecting the equivalent of carbon .
69 .
As soon as opportunity will permit me to resunme and complete these several investigations , I shall do myself the honour of communicating them to the Royal Society .
lpothecaries ' Hall , January 10th , 1839 .

