FRANKLAND ON THE COXPOSITION OF V.-On the Compositionof Air from Moat Blanc By DR. E. FRANKLAND, F.R.S. THEgreatly increased perfection of gasometric analysis and the numerous minute investigations made by Regnaul t Bunsen and others during the past ten or fifteen years have rendered our lcnowledgc of the composition of atmospheric air exceedingly complete as far at least as its two chief constituents are concerned. The earliest analyses of air led chemists to believe that the rela- tive proportions of these constituents were liable to very consider- able fluctuations. As the processes employed however became more accurate these discrepancies gradually disappeared ;until at length carefully conducted experiments showed an apparent uni-formity in the composition of samples of air taken from the most widely different localities; but it remained for subsequent and far more delicate methods of research to demonstrate that notwith- standing these apparently accordaxit results the composition of the atmosphere does in reality exhibit certain fluctuations confined however within very narrow limits The first series of experiments belonging to the latter category and which may be said to have first established the variability of the percentage of atmospheric oxygen are those of Bunsen made AIR PROM MONT BLANC.upon samples of sir collected at Marburg on ten different days and exhibiting when freed from carbonic acid a percentage of oxygen varying from 20.Y73 to 20.840 whilst no two analyses of the same sample differed more thm -031 per cent from each other.Then followed a most elaborate series of determinations by Regnault begun in December 1847 upon the air of Paris and continued in January 1848 The variations in the percentage of oxygen here observed were in December 1847 from 20.90 to 2l*SO,and in January 1848 from 20.89 to 20.99. Lewy collected air near the surface of the sea on the 18th December 1847 at 3 P.M. in lat. 21" 9' N. and long. 42' 52' W. of Paris temperature 24"C. and found it to contain :-Nitrogen . 78*886 Oxygen . . 21.060 Carbonic Acid . *054 Another specimen collected December 4th 1847 at 3 A.M. in la&47" N. and long. 13' W. temperature 13"C. contained :-Nitrogen . 79.006 Oxygen .20.961 Carbonic Acid . . *033 r__-100-000 A Iarge nnmber of analyses by the same ehemist*demonstrate that the air near the surface of the sea contains about2 the same proportion of carbonic acid as that resting upon the land and that the sea air is richer both in oxygen and carbonic acid by day than by night; a fact which he explains by assuming that dissolved air is liberated during the day from the heated surface-layer of the ocean such dissolved air being as is well-known much richer than atmospheric air in the two gases just named. Lem'y also made a very extended series of analyses of air collected at New Granada and Bogota during thc dry and rainy seasons. The mean of eleven analyses of air from New Granada gave :- FRANKLAND ON THE COMPOSITION OF Nitrogen .78.946 Oxygen . . 21.014 Carbonic Acid . . oo40 1oo'ooo Analyses of air from Bogota collected during the dry and rainy seasons gave the following mean numbers :-Dry Season. Rainy Season. Nitrogen . . 78.932 78.966 Oxygen. . . 21.022 20.996 Carbonic acid . -046 *038 100~000 100~000 On some occasions the air of New Grenada was found to con-tain as much as 049 per cent. of carbonic acid on which occasions the percentage of oxygen fell as low as 20.331. This abnormal composition is ascribed to volcanic eruptions and extensive confla- grations. In nearly all cases Lemy seems to have found that an increased amount of ca~bonic acid in the air was accompanied by an increase in the percentage of oxygen.Messrs. €€.and A. Schlagintweit determined by weighing the amount of carbonic acid in the air at great elevations in the Eastern Alps by absorption with p6tash; but as the increase in weight of the potash apparatms was only from 3 to 6 milligrammes in each experiment and as the two weighings were made at intervals of nearly two months these detcrrninations can only be regarded as approximative. Messrs Schlagintweit found the volume of carbonic acid to vary from #032 to *058per cent Up to an altitude of 11,043 feet they found a gradual increase of carbonic acid but at tliis elevation they conceived that a constant maximum was arrived at. In some later experiments on the air of Monte Rosa at heights varying from 13,374 to 13,858 feet the same experimenters found a mean percentage of carbonic acid equal to *079and a maximum of *095.In the year 1852 M. Itcgnault published an extensive series of determilzatioris of the percentage of oxygen in samples of air AIR FROM MONT BLANC. 25 deprived of carbonic acid from different localities. The following is a condensed summary of his results Percentage of Oxygen. 100 specimens from Paris and neighbourhood a (1848. . 20.913 20.999 9 ?> , Lyons Montpeliier & St. Martin-aux-Arbres . 20.918 20.966 30 , , Berlin (1848 & 1849) . 20.908 20.998 >, , Madrid (1848) 20.97 6 20,982 23 2 , Geneva Mont Sa%ve and Mont Buet . 20,909 20.993 15 1 , Toulon the Mediterra- nean and Algiers . 20.912 20.982 5 , taken during a voyage from Liverpool to Vera-Cruz .310.918 20-965 1 , from Guallalamba South Ame- rica . 20.960 2 9 , the summit of Pichincha (15,924 feet) -. 20,949 20.988 In this investigation a few remarkable deviations from these normal amounts of oxygen were observed vix. :-Percentage of Oxygen. Air collected in the I-farbour of Algiers June 5th 1851 . . 20.42 20.395 Air from Bay of Bengal February lst 1849 . 20.46 20-45 Air from Ganges March 8th 1849. Tempera-ture 35"C. foggy weather much decompos- ing organic matter in the water. Corn-mcncement of an outbreak of cholera . 20.390 20.387 It would be interesting to know how far these results were really abnormal as regards the relative proportions of oxygen and nitro- gen since it is not improbable that the apparently small per- centage of oxygen indicated mas in reality due to the presence of gaseous organic matters in larger quantity than usual; the ignition of these with excess of hydrogen in the mode adopted in these analyses would have the eEect of converting the carbon of such organic matters into carbonic oxide thus diminishing the contraction of volume on explosion and consequently the apparent FRANKLAND ON THE COMPOSPTION OF percentage of oxygen.Should such abnormal specimens of air be again encountered it mould be desirable subsequently to the analysis made in the usual manner to igiiitc other partions of them with an equal volume of mixed electrolytic gases so as to convert the carbon of any organic matter that might be present into carbonic acid which could then be estimated in the usual manner by absorption with potash.Dr. Miller examined air collected during a balloon ascent in August 1852 at a height of 18,000 feet and also a sample collected near the surfme of the earth at the same time with the following results Air from altitude Air near the of 18,000 feet. earth. Percentage of Oxygen . . 20.88 20-9.2 After these numerous and minute analyses establishing as they do with few exceptions the slight but still undoubted variations in the relative proportions of the tmo chief atmospheric gases any further contributions to this particular branch of our knowledge can only be of comparatively small value. Nevertheless as an apportunity was afforded me for collecting specimens of air whilst accompanying Dr.Ty n dall during the past summer in his ascent cf Mont Blanc for the establishment of thermometric stations I did not regard a fern further experiments upon air from great altitudes as entirely SLI~CX*~~LIOUS ;since the discovery of the causes determining the variations in the composition of the atmosphere will probably only be arrived at by the accumulation of vast numbers of obscrvations made at VW~QUSpoints at and above the earth’s surface. It ~ill also bc perceived on reference to the various analytical results above given that with the exception of XI. Lemy no experimenter lzas made contemporaneous determi- nations of the three chief gaseous constituents of the air; and I was thcrcfore anxious if possible to render these samples avail- able for the determination of carbonic acid as well as of the two other chief gases.The very minute changes of volume which the instrument I use for gaseous analysis is capable of registering led me to hope that it might not be impossible to make direct determinations of carbonic acid in the few cubic inches of air which are usually sealed up for analysis ;and a number of estimations of the carbonic acid in air collected. at St. Bartholomcw’s Hospital proved that All% FROM HONT BLANC. this gas can be thus estimated by absorption with caustic potash. Such an amount of air was taken for each determination as was capable of supporting a column of mercury from 600 to 800 millimeters high which rendered any diminution of volume to the extent of about T&T part distinctly appreciable.The following results were obtained those on "the 26th of January being seveii successive determinations extending over about three hours Percentage of Carbonic Acid. . -042 . 964 . ,077 . *098 . 0087 ._. . . *098 .i . *085 . *098 * -110 * -101 Although it cannot be doubted that there are other methods by which carbonic acid can be more minutely and accurately determined whenever they can be carried out with the usual conveniences of a laboratory at hand yet it is very questionable whether any of these processes em rival this purely eudiometrical one in cases where the operations have to be performed in the midst of all the inconveniences attending an experimenter at great altitudes.Any such estimations involving the weighing of potash tubes at intervals of several days or even weeks are obviously not worthy of implicit confidence. ln the above mode of determination the actual change of volume self-corrected for temperature aqueous vapour &c. being actually observed by the operator aD error exceeding -01 per cent. is probably rarely or never committed. Iu the following analyses of air from &font Blanc the carbonic acid was absorbed by a single drop of concentrated solution of caustic potash and the oxygen was then determined by exploding the residual gas with excess of electrolytically prepared hydrogen. Specimen collected at the Grands Mulets (altitude 11,000 feet) August 20th 1859 at 6.45 P.M Wind north hail falling but a moderately clear sky FRANKLAND ON THE COMPOSITION OF Estimation of Carbonic Acid.I Obs. vol. Temperature C. Air used . . . 539*5 5.0' After absorption of carbonic acid . 538.9 5.0" Estimations of Oxygen. 11. Obs. vol Temperature C. Air used . 0 . 290.8 5.2" After admission of hydrogen 4865 5*2O After explosion 8 . . 304.9 5*2* 111. Obs. vol. Temperature C' 0 Air used . . . 248.7 5.2' After admission of hydrogen . 443% 5-2' After explosion . 288.7 5.2" Percentage Composition I. 11. 111. Mean. Nitrogen . .. 79096 79.124 79.110 Oxygen . .. 20.793 20.765 20.779 D.Carbonic acid 0.111 .. *111 100*000 Specimen taken at the summit (altitude 15,732 feet) August 2lst at 8.45 A.M. Wind north ;weather bright and sunny; air filled with particles of snow whirled up by the wind. Estimation of Carbolzic Acid. Obs. vol. Temperature C. Air used * 4 . 326.1 5.2' After absorption of carbonic acid . 325-9 5*2O A133 FROM MONT BLANCo Estimations of Oxygen. Ob,s. vol. Temperature C. . . . 166.8 5.3O Air used . After admission with hydrogen . 273.0 5*3O After explosion . . '168.1 5*3O 111 Obs. vol. Temperature C. . 159.9 5*3O Air used . . 0 After admission of hydrogen. . 302.7 5*3O After explosion . . . 202.7 5*3O Percentage Composition. I. Is. III. Mean.Nitrogen .. 78.989 78.988 78.989 Oxygen . .. 20.950 20.951 20.950 Carbonic acid @061 .. .. -061 100*000 Specimen collected at Charnonix (altitude 3000 feet) August 23rd at 2 P.M. Wind north; sky clear. &stirnation of Carbonic Acid. L. Obs. vol. Temperature C. Air used . . . * . 474.1 5*4O After absorption of carbonic acid . 473.8 5*4O Estimations of Oxygen. 11. Obs. vol. Temperatare C 8 8 0 Air used . 255.6 5*Oo After admission of hydrogen * . 456.8 5*0° After explosion . . 295.5 5*Oo 111. Obs. Val. Temperature C. Air used . . 217.4 5.0" After admission of hydrogen . . 377.3 5*0° After explosion . . 241.1 5 *Oo Percentage of Copposition. I. 11. 111. Mean. Nitrogen *. 79.0-15 79.007 79.056 Oxygen ... 20-892 20.870 20.881 Carbonic acid 0063 .. .. 0063 So far as the nitrogen and oxygen are concerned the composi- tion of these samples of air falls within the limits of variation noticed by former experimenters ;but although the comparatively high percentage of carbonic acid at the Grands Mulets confirms the observations of the Messrs. Schlagintweit as to the presence of a larger amount of this gas at great elevations yet the diminu- tion of the quantity to abaut the normal amount upon the summit shows either that this gas attains a maximum at a height of abotd 11,000 feet and again diminishes above this altitude or as is much more probable the percentage of' carbonic acid is generally but not invariably greater in the higher regions of the atmosphere.These results also exhibit a correlation between atmospheric oxygen and carbonic acid for when the one increases the other diminishes,-a fact tvhich will be better seen from the following comparison :-Percentage of Mean percentage of oxygen, carbonic acid. in air free from carbonic acid Grands Mulets . . -111 20.802 Summit . . -061 20,963 Chamounix . . -063 20.894 This result if it be confirmed cannot be regarded as altogether unexpected when we consider the effects of vegetation combus- tion and respiration upon the coiistituents of the atmosphere ; but both this and the comparative amoqnt of carbonic acid at great altitudes are problems the solution of which must be left to future and more extended inquiries.