478 



NATURE 



[March 14, 1901 



various planetary systems with the 26-inch refractor of the 

 United States Naval Observatory at Washington. In the 

 Astronomische Nachrichten (Bd. 154, Nos. 3686-7) he gives 

 the details and results of the investigations of Saturn, including 

 measures of the planet, rings, and the satellite Titan, He 

 attributes much of the consistency of the values determined to 

 the use of various colour screens placed between the eye and 

 the telescope, which reduces or eliminates entirely the secondary 

 spectrum produced by the objective, thereby enabling a much 

 more sharply defined disc to be obtained. 



In addition to his own recent measures, the author also brings 

 together previous work on the subject from 1659. The following 



is a summary of the new determinations : — 



„ • km. 



External diameter of the outer ring ... =40-304 278,768 



Internal diameter of the outer ring, or\ _ .»g_ 240 610 



External diameter of Cassini's division ] ^ 

 DiameterofthecentreofEncke's division =37777 261,290 



Width of Encke's division = 0-107 740 



Total width of the outer ring ... ... = 2-758 19.076 



Width of the outer part of the outer ring = i "237 8,556 



Width of the inner part of the outer ring = 1*414 9>78o 



Width of Cassini's division ... ... = 0-418 2,891 



External diameter of central ring ... =33-951 234,827 



Internal diameter of central ring, or\ _2c.qi-2 170 coi 



External diameter of dusky ring ... ) ~ 5 ^J '^'-' 



Width of central ring = 4-000 27,667 



Internal diameter of dusky ring ... =20-582 142,359 



Black space between Saturn's globe and \ _ ,.^(^- jq g^g 



dusky ring ... ... ... ... / ~ ■' ' ' 



Equatorial diameter of Saturn =17-448 120,682 



AssumedoblatenessofSaturn(H. Struve) = o'loi3 



Polar diameter with this oblateness ... =15681 108,457 



Assumed mass of Saturn (Bessel) ... = 1:3501-6 



Resulting mean density of planet ... = o"i234 = o679 



that of water. 

 Diameter of the satellite Titan ... =0-487 3,368 



HYDROGEN IN AIR. 



TN a recent number of the Anttales de Chiinie ei de Physique 

 "*■ (January, ,1901), M. Armand Gautier, professor of che- 

 mistry at the Ecole de Medicine, Paris, gives a connected 

 account of his researches on the combustible gases of the atmo- 

 sphere. These researches have occupied some years, they have 

 been carried out with extraordinary care and completeness, and 

 they have yielded results of very great interest both in regard to 

 their main object and also in relation to incidental scientific 

 questions. 



The most striking fact elicited by M. Gautier is that pure air 

 contains free hydrogen as a normal constituent to the extent of 

 about two volumes in 10,000. This conclusion, taken in con- 

 junction with the recent experiments of Profs. Liveing and 

 Dewar (Nature, December 20, 1900, p. 189), in which they 

 record the isolation of a fraction of air containing 43 per cent, 

 of hydrogen, which they actually exploded, seems to admit of 

 no doubt. 



Analytical Methods. — M. Gautier set himself to determine the 

 character and amount of combustible gases in the atmosphere by 

 aspirating a large volume of air through a train of absorbents. 

 Nothing could be more obvious and simple in principle than 

 such a method ; the difficulty of making it available for deter- 

 mining with any degree of certainty the character and quantity 

 of very small amounts of combustible gases will, however, be 

 thoroughly appreciated by chemists. 



The first part of M. Gautier's memoir is devoted to a descrip- 

 tion of the preliminary work which was necessary for the 

 selection and proper use of the absorbents. Beginning with 

 carbon dioxide, he confirmed the previous observation of 

 Boussingault and Eliot and Storer that carbon dioxide is very 

 difficult to absorb. from a large admixture of other gases. After 

 circulating 90 litres of ordinary air during forty-eight hours 

 through a tube 80 centimetres long and containing glass 

 beads moistened with caustic potash solution of density 1-3, it 

 was found that 10-7 c.c. per million of CO2 remained unabsorbed. 

 A satisfactory absorbent in respect both to rapidity and complete- 

 ness was found in barium hydrate, either dissolved or simply 

 moistened with water. This substance would, of course, also 

 absorb other acid gases, such as HoS, SOo and NOo. 



The desiccation of air by sulphuric acid was also shown to be 

 incomplete ; a satisfactory agent was found in phosphoric oxide 

 previously heated with oxygen at 260° C. to get rid of lower 

 oxides. 



For the absorption of minute quantities of carbon monoxide 

 the ordinary reagents are ineffective and a new one was found in 

 iodine pentoxide. Air containing i/ioo,oooth of the gas loses it 

 completely and at once when passed over the pentoxide heated to 

 70° C. Carbon dioxide, oxygen, nitrogen, hydrogen, marsh gas 

 have no action on the oxide at that temperature, and other more 

 strongly reducing gases, such as alcohol vapour and benzene 

 when much diluted are also without action. When carbon 

 monoxide acts upon iodine pentoxide, iodine and carbon dioxide 

 are produced ; the iodine is absorbed by a tube containing finely 

 divided copper and the carbon dioxide by barium hydrate. The 

 estimation of hydrogen and hydrocarbons is next dealt with. 

 When a mixture of 50 c.c. of hydrogen with 235 litres of pure air 

 was passed over heated oxide of copper the hydrogen was 

 entirely burnt provided that a tube of 70 centimetres was 

 employed. With a tube of 30 centimetres, only 70 per 

 cent, of the hydrogen was burnt. When using shorter tubes in 

 subsequent experiments the weight of water obtained had to be 

 multiplied by a factor in order to give the effect of an " infinite " 

 tube of copper oxide. 



When diluted marsh gas is passed over heated copper oxide 

 there is neither complete combustion nor equivalent combustion 

 of the carbon and hydrogen, and here also the use of factors is 

 necessary. The hydrogen burns in greater proportion than the 

 carbon. With a diluted mixture of marsh gas and hydrogen it 

 was found that the presence of the hydrogen facilitated the com- 

 bustion of the hydrogen of the marsh gas but retarded that of the 

 carbon. Admixture of the copper oxide with spongy platinum or 

 with other metallic oxides did not improve the efficiency. It was 

 found that the copper oxide, after continued heating to redness, 

 gradually lost its power of oxidising, and after 1500 hours it was 

 without effect upon hydrocarbons, and it only partially oxidised 

 free hydrogen. 



For a detailed description and drawing of the apparatus the 

 original memoir must be consulted. The air was filtered from 

 suspended impurities by filtration through glass wool, its carbon 

 dioxide was then absorbed — and in this connection the author 

 devised a special form of absorption tube which he strongly 

 recommends — water was then absorbed and the air entered the 

 combustion tube. The combustion tube was heated in a furnace 

 of special construction, in which great uniformity of temperature 

 could be maintained from end to end. Water and carbon 

 dioxide formed by combustion were then absorbed, and the 

 apparatus terminated in an aspirator, a "decanteur" and a 

 tneter. There were in all twenty-eight pieces in the train ; 

 they were connected together by clamped india-rubber joints 

 made from purified tubing, which experiment showed to be 

 proof against diffusion. 



The Air of Paris. — Beginning first with towns, M. Gautier 

 examined the air in the region of the Ecole de Medicine. 

 The average ratio of carbon and hydrogen found corresponded 

 pretty nearly to CII4, but there was evidence at times of some 

 more highly carburetted hydrocarbon being present, at others 

 of free hydrogen. 



There was no evidence of hydrocarbons of the ethylene or 

 acetylene series being present. 



The quantity of carbon monoxide found was extremely small, 

 it averaged 2-1 1 volumes per million, but this included one very 

 abnormally high instance. Neglecting this one instance the 

 average of -56 volume per million was obtained. The quantity 

 of carbon monoxide varied in different places ; it increased in 

 densely populated places. In a small room at the laboratory 

 heated by an old-fashioned faience stove and illuminated for 

 several hours by three gas jets 12*3 volumes were found. On 

 the whole the' quantity of CO and unsaturated hydrocarbons 

 in town air may be said to be insignificant. 



The Air of Forests.— 1\i& next experiments related to the air 

 of forests, and the station fixed upon was a clearance in the 

 middle of a wood at Lainville, 70 kilometres from Paris. 



Here the proportion of carbon to hydrogen pointed distinctly 

 to the presence of free hydrogen along with marsh gas. It did 

 not seem probable that the hydrogen came directly from living 

 vegetation, but it was possible that it might arise from decom- 

 position going on in the soil, and it was therefore decided to 

 make analyses in localities where this possible source would be 

 removed to a large extent. 



NO. 1637, VOL. 63] 



