Elements and Atoms: Chapter 10
Prime Matter: Prout

William Prout (1785-1850; see portrait at John Park's ChemTeam site) was a medical doctor with a great interest in chemistry. He developed analytical instruments and published chemical analyses of organic material. His analyses of food classified components which correspond essentially to proteins, carbohydrates, and fats. [Prout 1827] Prout was also recognized as a clinical authority on urinary and digestive disease.

By the end of the 19th century, however, Prout was best known for the hypotheses that atomic weights were multiples of that of hydrogen and that hydrogen was a fundamental building block of matter.[1] These hypotheses, proposed in the two papers reproduced below, were tremendously influential. They inspired ever more accurate determinations of atomic weights to test the multiples hypothesis. As it became clear that the original multiples hypothesis was contradicted by experiment, various researchers modified rather than discarded it.[2] In other words, even though the hypothesis suggested by Prout was not correct, it was too attractive to be regarded as entirely baseless. And indeed, there was something to the idea of simple building blocks of atoms after all, as we will see in the concluding section of this work.

The selection presented below looks forward to material presented at the end of this work, and it looks back to one of the ancient notions mentioned in the first selection of this work, that of prime matter. While suggesting that atoms of the elements have structure, an idea developped in the book's final section, the present chapter does not advance the idea of discrete ultimate particles developed in the last three chapters. Yet there are also good reasons for including this chapter at the end of the present section on atoms. It is, after all, concerned with atomic weights, one of the main themes of this section. In addition, the time of its publication makes it part of the notions of atom contemporary with those of the preceding few chapters.

This selection is somewhat difficult to understand, and it is full of unwarranted assumptions. The papers are not models of clarity in exposition or in scientific method. As William Brock judged, in a volume devoted mostly sympathetically to Prout's work [Brock 1985, p. 99], "It was a certainly an understatement for Prout to admit [as he later did] that his ideas had not been clearly presented; in fact the whole effect of his anonymous paper of 1815 is one of intellectual confusion." I will not expend an inordinate amount of effort in explaining Prout's analyses or calculations.

On the Relation between the Specific Gravities of Bodies in their Gaseous State and the Weights of their Atoms

Annals of Philosophy 6, 321-330 (1815)[3]

The author of the following essay submits it to the public with the greatest diffidence; for though he has taken the utmost pains to arrive at the truth, yet he has not that confidence in his abilities as an experimentalist as to induce him to dictate to others far superior that its importance will be seen, and that some one will undertake to examine it, and thus verify or refute its conclusions. If these should be proved erroneous, still new facts may be brought to light, or old ones better established, by the investigation; but if they should be verified, a new and interesting light will be thrown upon the whole science of chemistry.[4]

It will perhaps be necessary to premise that the observations about to be offered are chiefly founded on the doctrine of volumes as first generalized by M. Gay-Lussac; and which, as far as the author is aware at least, is now universally admitted by chemists.

On the Specific Gravities of the Elementary Gases.

1. Oxygen and Azote.-- Chemists do not appear to have considered atmospheric air in the light of a compound formed upon chemical principles, or at least little stress has been laid upon this circumstance. It has, however, been long known to be constituted by bulk of four volumes of azote and one volume of oxygen; and if we consider the atom of oxygen as 10, and the atom of azote as 17.5, it will be found by weight to consist of one atom of oxygen and two atoms of azote, or per cent. of
Hence, then, it must be considered in the light of a pure chemical compound; and indeed nothing but this supposition will account for its uniformity all over the world, as demonstrated by numerous experiments.[5] From these data the specific gravities of oxygen and azote (atmospheric air being 1.000) will be found to be,[6]
2. Hydrogen.-- The specific gravity of hydrogen, on account of its great levity, and the obstinacy with which it retains water, has always been considered as the most difficult to take of any other gas. These obstacles made me (to speak in the first person) despair of arriving at a more just conclusion than had been before obtained by the usual process of weighing; and it occurred to me that its specific gravity might be much more accurately obtained by calculation from the specific gravity of a denser compound into which it entered in a known proportion.[7] Ammoniacal gas appeared to be the best suited to my purpose, as its specific gravity has been taken with great care by Sir H. Davy, and the chance of error had been much diminished from the slight difference between its sp. gr. and that of steam. Moreover, Biot and Arrago had obtained almost precisely the same result as Sir H. Davy. The sp. gr. of ammonia, according to sir H. Davy, is .590164, atmospheric air being 1.000. We shall consider it as .5902; and this we are authorized in doing, as Biot and Arrago state it somewhat higher than Sir H. Davy. Now ammonia consists of three volumes of hydrogen and one volume of azote condensed into two volumes. Hence the sp. gr. of hydrogen will be found to be .0694,[8] atmospheric air being 1.0000. It will be also observed that the sp. gr. of oxygen as obtained above is just 16 times that of hydrogen as now ascertained, and the sp. gr. of azote just 14 times.[9],[10]

3. Chlorine. The specific gravity of muriatic acid, according to Sir H. Davy's experiments, which coincide exactly with those of Biot and Arrago, is 1.278. Now if we suppose this sp. gr. to be erroneous in the same proportion that we found the sp. gr. of oxygen and azote to be above, (which, though not rigidly accurate, may yet be fairly done, since the experiments were conducted in a similar manner[11]), the sp. gr. of this gas will come out about 1.2845;[12] and since it is a compound of one volume chlorine and one volume hydrogen, the specific gravity of chlorine will be found by calculation to be 2.5.[13] Dr. Thomson[14] states, that he has found 2.483 to be near the truth,[15] and Gay-Lussac almost coincides with him.[16] Hence there is every reason for concluding that the sp. gr. of chlorine does not differ much from 2.5. On this supposition, the sp. gr. of chlorine will be found exactly 36 times that of hydrogen.

On the Specific Gravities of Elementary Substances in a Gaseous State that do not at ordinary Temperatures exist in that State.

1. Iodine. I had some reason to suspect that M. Gay-Lussac had in his excellent memoir rated the weight of an atom of this substance somewhat too high; and in order to prove this 50 grains of iodine, which had been distilled from lime, were digested with 30 grs. of very pure lamellated zinc.[17] The solution formed was transparent and colourless; and it was found that 12.9 grains of zinc had been dissolved. 100 parts of iodine, therefore, according to this experiment, will combine with 25.8 parts of zinc, and the weight of an atom of iodine will be 155,[18] zinc being supposed to be 40. From these data, the sp. gr. of iodine in a state of gas will be found by calculation to be 8.611111, or exactly 124 times that of hydrogen.[19]

2. Carbon. I assume the weight of an atom of carbon at 7.5.[20] Hence the sp. gr. of a volume of it in a state of gas will be found by calculation to be .4166, or exactly 12 times that of hydrogen.

3. Sulphur.-- The weight of an atom of sulphur is 20. Hence the specific gravity of its gas is the same as that of oxygen, or 1.1111, and consequently just 16 times that of hydrogen.

4. Phosphorus.-- I have made many experiments in order to ascertain the weight of an atom of this substance; but, after all, have not been able to satisfy myself, and want of leisure will not permit me to pursue the subject further at present. The results I have obtained approached nearly to those given by Dr. Wollaston, which I am therefore satisfied are correct, or nearly so, and which fix phosphorus at about 17.5, and phosphoric acid at 37.5,[21] and these numbers at present I adopt.

5. Calcium.-- Dr. Marcet found carbonate of lime composed of 43.9 carbonic acid and 56.1 lime.[22] Hence as 43.9:56.1::27.5:35.1, or 35 very nearly; and 35 - 10 = 25, for the atom of calcium. The sp. gr. of a volume of its gas will therefore be 1.3888, or exactly 20 times that of hydrogen.[23]

6. Sodium.-- 100 grains of dilute muriatic acid dissolved 18.6 grs. of carbonate of lime, and the same quantity of the same dilute acid dissolved only 8.2 grs. of carbonate of lime, after there had been previously added 30 grs. of a very pure crystallized subcarbonate of soda. Hence 30 grs. of crystallized subcarbonate of soda are equivalent to 10.4 grs. of carbonate of lime, and as 10.4:30::62.5:180. Now 100 grs. of crystallized subcarbonate of soda were found by application of heat to lose 62.5 of water. Hence 180 grs. of the same salt contain 112.5 of water, equal to 10 atoms, and 67.5 dry subcarbonate of soda, and 67.5 - 27.5 = 40 for the atom of soda, and 40 - 10 = 30 for the atom of sodium. Hence a volume of it in a gaseous state will weigh 1.6666, or exactly 24 times that of hydrogen.

7. Iron.-- 100 grs. of dilute muriatic acid dissolved as before 18.6 grs. of carbonate of lime, and the same quantity of the same acid dissolved 10.45 of iron. Hence as 18.6:10.45::62.5:35.1, or for the sake of analogy, 35, the weight of an atom of iron. The sp. gr. of a volume of this metal in a gaseous state will be 1.9444, or exactly 28 times that of hydrogen.

8. Zinc.-- 100 grs. of the same dilute acid dissolved, as before, 18.6 of carbonate of lime and 11.85 of zinc. Hence as 18.6:11.85::62.5:39.82, the weight of the atom of zinc, considered from analogy to be 40. Hence the sp. gr. of a volume of it in a gaseous state will be 2.222, or exactly 32 times that of hydrogen.

9. Potassium.-- 100 grs. of the same dilute acid dissolved, as before, 18.6 carbonate of lime; but after the addition of 20 grs. of super-carbonate of potash, only 8.7 carbonate of lime. Hence 20 grs. of super-carbonate of potash are equivalent to 9.9 carbonate of lime; and as 9.9:20::62.5:126.26, the weight of the atom of super-carbonate of potash. Now 126.26 - (55 + 11.25) = 60, the weight of the atom of potash, and 60 - 10 = 50, the weight of the atom of potassium. Hence a volume of it in a state of gas will weigh 2.7777, or exactly 40 times as much as hydrogen.

10. Barytium.-- 100 grs. of the same dilute acid dissolved exactly as much again of carbonate of barytes as of carbonate of lime. Hence the weight of the atom of carbonate of barytes is 125; and 125 - 27.5 = 97.5, the weight of the atom of barytium. The sp. gr. therefore, of a volume of its gas will be 4.8611, or exactly 70 times that of hydrogen.

With respect to the above experiments, I may add, that they were made with the greatest possible attention to accuracy, and most of them were many times repeated with almost precisely the same results.[24]

The following tables exhibit a general view of the above results, and at the same time the proportions, both in volume and weight, in which they unite with oxygen and hydrogen: also the weights of other substances, which have not been rigidly examined, are here stated from analogy.

TABLE I.-- Elementary Substances.[25]

Name.Sp. gr. hydr. being 1.Wt. of atom, 1[26] vol. hydr. being 1.Wt. of atom, oxygen being 10.Wt. of atom, oxygen being 10, from experiment.Sp. gr. atmospheric air being 1.Sp. gr. atmospheric air being 1, from experiment.Wt. in grs. of 100 cub. inches. Barom. 30, Therm. 60.Wt. in grs. of 100 cub. in. from exper.Observations.
Hydrogen111.251.32.06944.073(1)2.1182.23(1)Dr.Thomson. See Annals of Philosophy, i. 177.
Carbon667.57.54(2).4166--12.708--(2)Dr. Wollaston, from Biot and Arrago. Phil. Trans. civ. 20. Dr. Thomson makes it 7.51. Annals of Philosophy, ii. 42.
Azote141417.517.54.9722.969(3)29.65229.56(3)Dr. W. from Biot and Arrago.
Phosphorus141417.517.4(4).9722--29.652--(4)Dr. W. from Berzelius and Rose.
Oxygen16810101.11111.104(5)33.88833.672(5)Dr. Thomson, from a mean of several experiments.
Sulphur16162020(6)1.1111--33.888--(6)Dr. W. from Berzelius.
Calcium20202525.46(7)1.3888--42.36--(7)Dr. W. from experiment.
Sodium24243029.1(8)1.6666--50.832--(8)Dr. W. from Davy.
Iron28283534.5(9)1.9444--59.302--(9)Dr. W. from Thenard and Berzelius.
Zinc32324041(10)2.222--67.777--(10)Dr. W. from Gay-Lussac.
Chlorine36364544.1(11)2.52.483(12)76.248--(11)Dr. W. from Berzelius. (12)Quoted from Dr. Thomson, Annals of Philosophy, iv. 13.
Potassium40405049.1(13)2.7777--84.72--(13)Dr. W. from Berzelius
Barytium707087.587(14)4.8611--148.26--(14)Dr. W. from Berzelius and Klaproth.
Iodine124124155156.21(15)8.6111--262.632--(15)Gay-Lussac. Ann. de Chim. xci. 5.

TABLE II.-- Combinations with Oxygen.[27]

Name.Sp. gr. hydro. being 1.Wt. of atom, 1[26] vol. hydro. being 1.Wt. of atom, ox. being 10.Wt. of atom, ox. being 10, from exper.Sp. gr. atmos. air being 1Sp. gr. atmos. air being 1, from exper.Wt. of 100 cu. in. Bar. 30, Ther. 60.Wt. of 100 cu. in. from exp.Elements by volume.No. of vol. after combination.Elements by weight.Observations
Water9911.2511.32.625.6896(1)19.06221.033.5 ox + 1 hyd11 ox + 1 hy(1)Trales, Dr. Thomson, Annals, i. 177.
Carbonic oxyde141417.517.54.9722.956(2)29.65229.16.5 ox + 1 ca11 ox + 1 car(2)Cruikshanks, quoted by Thomson.
Nitrous oxyde222227.5--1.52771.614(3)46.59649.227.5 ox + 1 az11 ox + 1 az(3)Sir H. Davy.
Atmospheric air14.43645--1.0001.00030.530.5(4).5 ox + 2 az2.51 ox + 2 az(4)Sir G. S. Evelyn
Phosphorous acid.5 ox + 1 ph ?1 ox + 1 ph ?
Oxyde of sulphur?.5 ox + 1 sul ?1 ox + 1 sul ?
Euchlorine4444553.05552.409(5)93.19273.474.5 ox + 1 ch1 ?1 ox + 1 ch(5)Sir H. Davy.
.5 ox + 1 iod1 ox + 1 iod
Lime28283535.461.9444--59.304--.5 ox + 1 cal1 ox + 1 cal

1 ox + 1 hy(6)2 ox + 1 hy(6)This and all higher combinations of hydrogen with oxygen are unknown.
Carbonic acid222227.527.541.52771.518(7)46.59646.3131 ox + 1 car12 ox + 1 car(7)Saussure.
Nitrous gas153037.51.04161.0388(8)31.7731.6841 ox + 1 az22 ox + 1 az(8)Berard.
Phosphoric acid303037.537.42.0832--63.54--1 ox + 1 ph1 ox + 1 ph
Sulphurous acid3232402.22222.193(9)67.77766.891 ox + 1 sul12 ox + 1 sul(9)Sir H. Davy.
1 ox + 1 ch2 ox + 1 ch
1 ox + 1 iod2 ox + 1 iod

1.5 ox + 1 car3 ox + 1 car
Nitrous acid383847.52.63882.427(10)80.48474.02341.5 ox + 1 az13 ox + 1 az(10)Sir H. Davy.
1.5 ox + 1 ph3 ox + 1 ph
Sulphuric acid404050502.777784.721.5 ox + 1 sul13 ox + 1 sul
1.5 ox + 1 ch3 ox + 1 ch
1.5 ox + 1 iod3 ox + 1 iod

2.5 ox + 1 car5 ox + 1 carSee Gay-Lussac's memoir on iodine above referred to.
Nitric acid545467.567.543.75114.3722.5 ox + 1 az15 ox + 1 az
2.5 ox + 1 ph5 ox + 1 ph
2.5 ox + 1 sul5 ox + 1 sul
Chloric acid7676955.2777--160.9682.5 ox + 1 ch5 ox + 1 ch
Iodic acid16416420511.3883347.3522.5 ox + 1 iod5 ox + 1 iod

TABLE III.-- Compounds with Hydrogen.

Name.Sp. gr. hydro. being 1Wt. of atom, 1[26] vol. hydr. being 1.Wt. of atom, oxygen being 10.Wt. of atom, oxygen being 10, from experiment.Sp. gr. atmospheric air being 1.Sp. gr. atmospheric air being 1, from experiment.Wt. of 100 cub. inch. Bar. 30. Ther. 60.Wt. of 100 cub. inch. from exper.Elements by volume.No. of vol. after combination.Elements by weight.Observations.
Carbureted hydrogen878.758.86.5555.5555(1)16.99916.9992 hy + 1 car11 hy + 1 car(1)Dr. Thomson.
Olefiant gas141316.2516.4.9722.974(2)29.65229.722 hy + 2 car11 hy + 2 car(2)Ditto.
Hydro-phosphorus gasI have omitted these from the uncertainty that still hangs over phosphorus.
Phosphoreted hydrogen
1 hy + 1 az.5 hy + 1 azThis compound is at present unknown, but it probably exists in fulminating gold, silver, &c. united to these metals.
Ammonia8.515.519.37521.5(3).5902.59(3)18.00318.003 hy + 1 az21.5 hy + 1 az(3)Dr. Wollaston.
Sulphureted hydrogen1716.520.62520.661.18051.177(4)36.00635.891 hy + 1 sul1.5 hy + 1 sul(4)Sir H. Davy.
Muriatic acid18.536.545.62545.661.2841.278(5)39.18338.9791 hy + 1 ch2.5 hy + 1 ch(5)Ditto.
Hydriodic acid62.5124.5155.625155.664.34024.3463(6)132.3751 hy + iode2.5 hy + 1 iod(6)Gay-Lussac.

TABLE IV.-- Substances stated from Analogy, but of which we are yet uncertain.

Name.Sp. gr. hydro being 1.Wt. of atom, 1[27] vol. hydr. being 1.Wt. of atom, oxygen being 10.Wt. of atom, oxygen being 10, from exper.Observations.
Magnesium12121514.6(2)(2)Henry. Berzelius makes it 15.77.
Copper32324040(7)(7)As deduced by Dr. Thomson.
Molybdenum48486060.13(10)(10)Bucholz and Berzelius.
Antimony8888110111.11(14)(14)Ditto. Dr. Thomson makes it 112.49.
Mercury100100125125(19)(19)Fourcroy and Thenard.
Silver108108135135(21)(21)Wenzel and Davy.


Table I.-- This, as well as the other tables, will be easily understood. In the first column we have the specific gravities of the different substances in a gaseous state, hydrogen being 1: and if we suppose the volume to be 47.21435 cubic inches, the numbers will at the same time represent the number of grains which this quantity of each gas will weigh. In the third column are the corrected numbers, the atom of oxygen being supposed, according to Dr. Thomson, Dr. Wollaston, &c. to be 10: and in the fourth, the same, as obtained by experiment, are stated, to show how nearly they coincide. Of the individual substances mentioned, I have no remark to make, except with respect to iodine. I made but one experiment to ascertain the weight of the atom of this substance, and therefore the results stated may be justly considered as deserving but little confidence; and indeed this would be the case, did not all the experiments of Gay-Lussac nearly coincide in the same.

Table II.-- This table exhibits many striking instances of the near coincidence of theory and experiment. It will be seen that Gay-Lussac's views are adopted, or rather indeed anticipated, as a good deal of this table was drawn up before I had an opportunity of seeing the latter part of that chemist's memoir on iodine. That table also exhibits one or two striking examples of the errors that have arisen from not clearly understanding the relation between the doctrine of volumes and of atoms. Thus ammonia has been stated to be composed of one atom of azote and only 1.5 of hydrogen[28], which are condensed into two volumes, equal therefore to one atom; and this is the reason why this substance, like some others, apparently combine in double proportions.[29]

Table III.-- This table likewise exhibits some striking examples of the coincidences above noticed. Indeed, I had often observed the near approach to round numbers of many of the weights of the atoms, before I was led to investigate the subject. Dr. Thomson appears also to have made the same remark. It is also worthy of observation, that the three magnetic metals, as noticed by Dr. Thomson, have the same weight, which is exactly double that of azote. Substances in general of the same weight appear to combine readily, and somewhat resemble one another in their nature.[30]

On a general review of the tables, we may notice,

1. That all the elementary numbers, hydrogen being considered as 1, are divisible by 4, except carbon, azote, and barytium, and these are divisible by 2, appearing therefore to indicate that they are modified by a higher number than that of unity or hydrogen. Is the other number 16, or oxygen? And are all substances compounded of these two elements?[31]

2. That oxygen does not appear to enter into a compound in the ratio of two volumes or four atoms.

3. That all the gases, after having been dried as much as possible, still contain water, the quantity of which, supposing the present views are correct, may be ascertained with the greatest accuracy.

Others might doubtless be mentioned; but I submit the matter for the present to the consideration of the chemical world.

Correction of a Mistake in the Essay on the Relation between the Specific Gravities of Bodies in their Gaseous State and the Weights of their Atoms

Annals of Philosophy, 7, 111-113 (1816)

The author of the essay On the Relation between the Specific Gravities of Bodies in their Gaseous State and the Weights of their Atoms is anxious to correct an oversight which influences some of the numbers in the third table given in that essay (vol. vi. p. 328). This oversight will be found in the head or title of the third column in each table, and consists in the statement of the atom of hydrogen being composed of two volumes instead of one[32], upon which latter supposition the tables are actually constructed, except in the instances corrected in the third table as follows, and in a sentence in the first paragraph on p. 330, beginning "This table also exhibits," &c. which is to be expunged.

Name.Sp. gr. hydro. being 1Wt. of atom, hydrogen being 1.Wt. of atom, oxygen being 10.Wt. of atom, oxygen being 10, from experiment.Sp. gr. atmospheric air being 1.Sp. gr. atmospheric air being 1, from experiment.Wt. of 100 cub. inch. Bar. 30. Ther. 60.Wt. of 100 cub. inch. from exper.Elements by volume.No. of vol. after condensation.Elements by weight.Observations.
Carburetted hydrogen8455.09.5555.5555(1)16.99916.9991 hyd + .5 car.51 hyd + .5 car(1)Dr. Thomson.
Olefiant gas1478.758.86.9722.9740(1)29.65229.721 hyd + 1 car.51 hyd + 1 car
Sulphureted hydrogen171721.2521.321.18051.17736.00635.891 hyd + 1 sul11 hyd + 1 sul
Muriatic acid18.53746.2545.421.2841.27839.18338.9791 hyd + 1 chl21 hyd + 1 chl
Hydriodic acid62.5125156.25157.534.34024.3463(2)132.375--1 hyd + 1 iod21 hyd + 1 iod(2)Gay-Lussac.
Ammonia8.51721.2521.5(3).5902(4).590018.00318.0003 hyd + 1 az23 hyd + 1 az(3)Dr. Wollaston. (4)Sir H. Davy.
Cyanogen262632.532.521.80551.8064(5)55.068--2 car + 1 az12 car + 1 az(5)Gay-Lussac. Ann. de Chim. Aug. 1815.
Hydro-cyanic acid13.52733.7533.846.9374.9360(5)28.593--1 cya + 1 hy21 cya + 1 az
Chloro-cyanic acid316277.5--2.15272.1111(5)65.659--1 cya + 1 chl21 cya + 1 chl

In this table it will be also observed that the new determinations of Gay-Lussac respecting the prussic acid, &c. are inserted, to show that they correspond with, and further corroborate, the views which have been brought forward in the essay above referred to.

There is an advantage in considering the volume of hydrogen equal to the atom, as in this case the specific gravities of most, or perhaps all, elementary substances (hydrogen being 1) will either exactly coincide with, or be some multiple of, the weights of their atoms[33]; whereas if we make the volume of oxygen unity, the weights of the atoms of most elementary substances, except oxygen, will be double that of the specific gravities of bodies in their gaseous state (either with respect to hydrogen or atmospheric air), by means of Dr. Wollaston's logometric scale.[34]

If the view we have ventured to advance be correct, we may almost consider the πρωτη υλη[35] of the ancients to be realised in hydrogen; an opinion, by the by, not altogether new.[36] If we actually consider this to be the case, and further consider the specific gravities of bodies in their gaseous state to represent the number of volumes condensed into one; or, in other words, the number of the absolute weight of a single volume of the first matter (πρωτη υλη) which they contain, which is extremely probable, multiples in weight must always indicate multiples in volume, and vice versa; and the specific gravities, or absolute weights of all bodies in a gaseous state, must be multiples of the specific gravity or absolute weight of the first matter (πρωτη υλη), because all bodies in a gaseous state which unite with one another unite with reference to their volume.


[1]Many scientists and historians of science refer to "Prout's hypothesis" (singular); however, William Brock takes great pains to distinguish these two hypotheses [Brock 1985]. I will follow his usage and refer to the "multiples" hypothesis and the "protyle" hypothesis. (Protyle comes from the Greek πρωτη υλη, prime matter.)

[2]For example, Jean Servais Stas carried out an extensive set of chemical analyses in the hopes of vindicating the multiples hypothesis; however, he observed, and in 1860 reported, just the opposite. [Stas 1860] In an immediate response to Stas' report, Charles Marignac urged that the protyle hypothesis was still viable, speculating (wildly for the time) that conservation of mass might not be strictly observed at the microscopic level among the components of prime matter. [Marignac 1860] Alembic Club Reprint #20 contains Prout's and Marignac's papers as well as extracts from Stas'.

[3]Prout published this paper and its companion, reproduced below, anonymously. The identity of the author was made public in 1816.

[4]Prout knows that his paper is speculative, and warns his readers at the outset that he is not certain of its results. He offers the essay for what it is worth, as a hypothesis to be supported or contradicted by facts. As an inspiration to future investigations, the article proved wildly successful.

[5]A modern chemist will notice several problems already. First, Prout incorrectly treats air as a chemical compound. Second, he appears to think that this belief about atmospheric air was an original idea. In fact, it was not as rare as Prout seems to think. (An article by Dalton on the constitution of the atmosphere clearly referred to "those who consider the atmosphere as a chemical compound" [Dalton 1805]. The constancy of composition of air is an observation which could lead one to believe that air is a chemical compound, but it is not a proof.) Third, Prout assumes the composition of air to be two nitrogen (azote) atoms for every oxygen atom, even though he correctly states the ratio by volume as close to four to one. The calculations in this paper appear to be based on the unstated assumption that equal volumes of gas contain equal numbers of molecules--except for oxygen: a volume of oxygen atoms contains twice as many molecules as an equal volume of other gases. Finally, the atomic weights in this paper appear unfamiliar not only because of the exceptional treatment of oxygen, but because the scale is based on oxygen = 10. (The absolute masses of atoms were not known or even reliably guessed until much later in the 19th century, so all atomic weight scales were relative. The most common basis was hydrogen = 1; however, oxygen = 10 was not all that rare.)

[6]Let x = sp. gr. of oxygen. 22.22 = a
y = sp. gr. of azote. 77.77 = b
Then (x+4y)/5 = 1.
And x:4y::a:b.
Hence 5-4y = 4ay/b
And y = 5b/(4a+4b) = .9722. And x = 5-4y = 1.11111. [Prout's original note--CJG]

[7]This idea of Prout's makes sense. It was difficult to determine the density of hydrogen because it was difficult to accurately weigh so light a substance. Consider that two grams of hydrogen gas takes up roughly 25 liters. A container that holds that much hydrogen would weigh at least that much, so the weight of the hydrogen would have to be determined as the relatively small difference difference between a filled and empty container. Finally, and not least, an experimenter must contend with the buoyancy of such a hydrogen sample: to the extent that the hydrogen-filled container is less dense than air, it would rest on the air and not push down a balance pan with the full force of its weight. One could measure the density of a heavier gas which contained hydrogen, and compute the density of hydrogen as a fraction of that heavy gas' density. Stas credited Prout for the accurate determination of the density of hydrogen. [Stas 1860]

[8]Let x = sp. gr. of hydrogen.
Then (3x+.9722)/2 = .5902.
Hence x = (1.1804-.9722)/3 = .0694. [Prout's original note--CJG]

[9]1.11111 / .0694 = 16. And .9722 / .0694 = 14. [Prout's original note--CJG]

[10]These are the first two observations that support the multiples hypothesis to be suggested below.

[11]This adjustment of experimental data is done on insufficient grounds. Prout alleges a similarity in method between experiments on muriatic acid and oxygen and nitrogen above. Such a vague similarity of measurement would be too casual a reason to warrant changing experimental data, particularly two sets of data by independent investigators. (The fact that Prout failed to include in the paper the oxygen and nitrogen correction to which he alludes makes his adjustment of the muriatic acid data even less justified!)

It is not all that rare for scientists to adjust their data; however, the adjustment of experimental data ought to be judged with a critical eye. For example, investigators who have long experience with a particular method may know that it typically overestimates the quantity it purports to measure by, say, 10%. In reporting a measurement made by this method, these investigators would be obliged to note the typical overestimate, and to regard a downward adjustment by 10% as their best estimate of the desired quantity.

[12]As 1.104:1.11111::1.278:1.286.
And as .969:.9722::1.278:1.283. The mean of these is 1.2845. [Prout's original note--CJG]

[13]Let x = sp. gr. of chlorine.
Then (x+.0694)/2 = 1.2845.
And x = 2.569-.0694 = 2.5 very nearly. [Prout's original note--CJG]

[14]Dr. Thomson is Thomas Thomson (1773-1852; see portrait at the Edgar Fahs Smith collection, University of Pennsylvania), editor of Annals of Philosophy (the journal that published this paper), friend of Prout, champion of Dalton's atomic hypothesis, and later champion of Prout's multiples hypothesis. Thomson was one of the chemists who took up the program, suggested by Prout, of measuring atomic weights to test the multiples hypothesis. His results supported that hypothesis [Thomson 1825]. Unfortunately, those results were severely flawed and harshly criticized. Jons Jacob Berzelius, the foremost analytical chemist of the time, opined, "This work belongs to those few productions from which science will derive no advantage whatever. ... The greatest civility which his contemporaries can show the author, is to forget that it was ever published." [Berzelius 1828]

[15]Annals of Philosophy, vol. iv. p. 13. [Prout's original note--CJG]

[16]Ditto, vol. vi. p. 126. [Prout's original note--CJG]

[17]That is, Prout carried out his own measurement here, the first of several reported in this section. Prout was a careful chemical analyst, and the combining ratio he reports here agrees with the accepted value to within 0.2%.

This paper, however, is not primarily an experimental paper presenting original observations. It is mainly a review of other work and an attempt at organizing that work and finding regularities in it.

[18]As 25.8:100::40:155. According to experiment 8th, stated below, the weight of an atom of zinc is 40. Dr. Thomson makes it 40.9, which differs very little. See Annals of Philosophy, vol. iv. p. 94. [Prout's original note--CJG]

[19]One volume of hydrogen combines with only half a volume of oxygen, but with a whole volume of gaseous iodine, according to M. Gay-Lussac. The ratio in volume, therefore, between oxygen and iodine is as 1/2 to 1, and the ratio in weight is as 1 to 15.5. Now .5555, the density of half a volume of oxygen, multiplied by 15.5, gives 8.61111, and 8.61111 / .06944 = 124. Or generally, to find the sp. gr. of any substance in a state of gas, we have only to multiply half the sp. gr. of oxygen by the weight of the atom of the substances with respect to oxygen. See Annals of Philosophy, vol. v. p. 105. [Prout's original note--CJG]

[20]Here is another example of a flaw in Prout's paper. He assumes the weight of carbon. He does not say he accepts it on the basis of his own or others' experiments or claim that the result is well established.

[21]Some of my experiments approached nearer to 20 phosphorus and 40 phosphoric acid. [Prout's original note--CJG]

[22]I quote on the authority of Dr. Thomson, Annals of Philosophy, vol. iii, p. 376. Dr. Wollaston makes it somewhat different, or that carbonate of lime consists of 43.7 acid and 56.3 lime. Phil. Trans. vol. civ. p. 8. [Prout's original note--CJG]

[23]In phosphorus and calcium one can see the scientific weakness of the multiples hypothesis, which Prout hints at near the end of this paper. The basis for that hypothesis is ostensibly the set of atomic weights proposed in the early portions of the paper. Yet it appears that these atomic weights are the result of selective citations of experimental values and rounding to the nearest multiple of hydrogen. The experimental weight of phosphorus cited in Table I is 17.4--not a multiple of hydrogen's, which is 1.25 on this scale. He rounds this to 17.5, and says in his own note that some of his experiments give a result closer to 20. Thus the uncertainty in the atomic weight of phosphorus is roughly equal to twice the atomic weight of hydrogen! Obviously, one cannot claim with any degree of probability, let alone certainty, that its atomic weight is an integral multiple of that of hydrogen.

In his calculation for calcium, Prout uses Marcet's analysis, and even then rounds the weight from 25.1 to 25 (a multiple of hydrogen); he notes, but does not use Wollaston's analysis, which would lead to a weight of 25.4 (not a multiple).

[24]The experiments Prout carried out and those he cites were accurate by the standards of his time. That accuracy, however, was not sufficient to support or to refute the multiples hypothesis if his data were submitted to a modern statistical treatment. There was sufficient leeway within the experimental error of his methods to select data which supported the multiples hypothesis. Take the case of potassium, for example. Suppose one of the measurements Prout cites was mistaken by only 1% so that 20 grams of super-carbonate of potash were equivalent to 10.0 grams carbonate of lime rather than the reported 9.9; this small error would change the computed weight of potassium from 50 to 48.75. So if Prout's analyses were accurate to ±1%, then the weight could be anywhere between 48.75 and 51.25 or anywhere from 39 to 41 times that of hydrogen. Clearly one cannot be sure that potassium is an integral multiple of hydrogen.

[25]In this table, columns 1-3 are all based on Prout's measurements or calculations. Columns 1 and 2 are identical for every element except oxygen. This reflects the unstated assumption remarked upon earlier (footnote 5) that a volume of oxygen contains twice as many molecules as a volume of other gases. Columns 2 and 3 are simply related to each other, for they are the same atomic weights but based on a different standard (hydrogen = 1 and oxygen = 10 respectively). Columns 3 and 4 represent atomic weights according to Prout and to other experimenters respectively; similarly columns 5 and 6 compare Prout's densities to those of other experimenters, as do (in different units) columns 7 and 8.

Superscript numbers in parentheses(#) in the table refer to Prout's table notes (Observations) at right.

[26]The original says that two volumes of hydrogen consitutes the unit. Prout corrected this heading in a subsequent paper.

[27]Table II presents a great deal of data on oxides. The first eight columns of data (i.e. all the numerical columns in my part a and the first two in part b) represent the same quantities as in Table I. The next two columns refer to the composition of the compound by volume. For example, half a volume of oxygen and one of hydrogen form one volume of water. The next column refers to the composition by atom, on the mistaken impression that oxygen has twice as many atoms per volume as other elements. Thus water is taken to consist of one atom of oxygen and one of hydrogen. Note that there are entries for several unnamed substances. For example, at the top of the second section (in my part b), there is an entry for an unnamed and then unknown entity made up of one volume of oxygen and one of hydrogen. These hypothetical compounds for which no data are presented do little more than display Prout's belief that such compounds existed or at least were possible. As in Table I, superscript numbers refer to Prout's table notes.

[28]Some papers of this time use the word atom as synonymous to equivalent or in a sense similar to that of the modern mole; that is, they use atom to refer to a macroscopic quantity and not to a molecule or ultimate particle. This reference to 1.5 atoms of hydrogen appears to be such a usage.

[29]See Gay-Lussac's memoir on iodine, Annals of Philosophy, vi. 189. [Prout's original note--CJG]

[30]These observations seem to refer to Table IV. In the original, however, they are labeled Table III, and there is no "observation" for Table IV.

[31]Prout does not say so explicitly, but all of his atomic weights are multiples of hydrogen. He does note that most are even multiples of 2 or 4 times hydrogen. He speculates that all elements might therefore be compounded of hydrogen and perhaps something else. He does not explicitly say so, but the notion of hydrogen as a primary material (the protyle hypothesis) flows from the multiples hypothesis: atomic weights would be multiples of hydrogen if the elements were composed of hydrogen units in an attraction so strong that chemical analysis cannot break the units apart. But why oxygen might be considered as another possible primary material is apparently not because of the atomic weights presented here. (That is, there is no multiples hypothesis apparent for oxygen from which a protyle hypothesis would spring naturally.)

[32]I have made this correction in labeling in the tables of the previous paper. (See footnote 26 above.)

[33]Here is an explicit statement of the multiples hypothesis.

[34]William Hyde Wollaston (1766-1828; see portrait at Carl Zeiss, Germany) was a well-respected chemical analyst whose work is cited often in the above tables. His "logometric scale" was an early chemical calculator, a table of equivalent weights attached to a device like a slide rule. [Wollaston 1814]

[35]πρωτη υλη is transliterated as prote hyle, from which the term protyle was coined.

[36]The idea of a primary matter is ancient. It was a part of Aristotle's conception of matter, for example (although Aristotle did not believe it could exist without embodying certain qualities which made up the elements). A belief in building blocks of matter simpler than the atoms of chemical elements persisted in the 19th century, even as the number of known elements increased. A basic belief that nature was simple led some scientists to believe that the elements were not the most fundamental pieces of matter: there seemed to be too many elements. Among those scientists was England's most prominent chemist, Humphry Davy. William Brock makes a case that Davy influenced Prout's thoughts on this subject. [Brock 1985] Dalton's belief in indestructible atoms [Dalton 1808], different kinds for different elements, was fundamentally opposed to this notion.


Back to the top of the table of contents of Elements and Atoms.
Back to the top of Classic Chemistry.