96 



CHEMISTEY. 



a quantity for an accurate assay, and, if so, 

 there is no reason why very much larger 

 weights should not be used ; indeed, the process 

 is well adapted for working on a large scale, 

 and, if there ever should be any great demand 

 for the alkaloid, would probably be employed. 

 New Unit of Light for Photometry. In a 



Saper read before the chemical section of the 

 ritish Association for the Advancement of 

 Science, at its last meeting, Prof. A. Vernon 

 Harcourt proposes a new unit of light for 

 photometric purposes ; the unit or standard in 

 present use being, according to his view, neither 

 scientifically accurate, nor a sufficiently close 

 approximation to accuracy for practical re- 

 quirements. This unit, as fixed by act of Par- 

 liament " a sperm candle of six to the pound 

 consuming 120 grains of sperm per hour " is 

 objectionable, because of its great liability to 

 variation in illuminating power, both from the 

 varying composition of the sperm and the 

 structure of the wick ; and it is only by a free 

 use of averages that it can be made to yield 

 approximately trustworthy results. Accord- 

 ing to the author : 



Three conditions need to be fulfilled for the pro- 

 duction of a standard flame: 1. The combustible 

 must be of known and definite composition. 2. 

 The conditions of burning must be of a simple and 

 definable character. 3. The nature of the combusti- 

 ble and of the conditions of burning must be such 

 that atmospheric changes may produce a minimum 

 of effect upon the light. A fourth condition might 

 be added as highly desirable namely, that the 

 operator should be able to verify for himself the 

 composition of the combustible he employs. No 

 chemist, at least, would willingly spend time upon a 

 quantitative operation for the accuracy aud signifi- 

 cance of whicn he was dependent upon the care 

 exercised by the workmen of the best maker of 

 candles or refiner of oil. 



For reasons of another kind, it is desirable that a 

 new unit of light should be made to correspond to 

 the average value of the existing unit, the light given 

 by a sperm candle consuming 120 grains of sperm 

 per hour. No change in the nomenclature of pho- 

 tometry would thus be required ; the unit of light 

 might still be called a candle, and gas which is de- 

 scribed as 14-candle gas or 16-candle gas would re- 

 tain its appellation, yielding, when burned under 

 standard conditions, the light of 14 or 16 of the new 

 units. 



After trying many different plans, he at last chose 

 for the standard combustible a mixture of air with 

 that portion of American petroleum which, after 

 repeated rectifications, distills at a temperature not 

 exceeding 50 Cent. This liquid consists almost en- 

 tirely of pentane, the fifth member of the series of 

 paraffins. 



To prepare the gas (writes the author), I draw into 

 the gas-holder the required volume of air, chosen ac- 

 cording to the capacity of the holder, and corrected 

 for pressure, temperature, and tension of aqueous 

 vapor. The volume may be measured either by 

 meaus of a meter, or by a scale upon the gas-holder. 

 Then the corresponding proportion of pentane is 

 poured into the gas-holder from a measuring-flask, 

 connected by means of glass and caoutchouc tubing 

 with a-tapjn the upper plate. If the liquid pentane 

 comes in contact with the plug of the tap. it acts on 

 the grease which is used to lubricate the plug, and is 

 liable thus to cause leakage. Contact is easily pre- 

 vented by plncing in the mouth of the tap a piece of 

 caoutchouc holding a glass tube, which can slide, 



air-tight, up or down. The upper end of this glass 

 tube is connected with the flask charged with 

 pentane ; the tap is opened, and the glass tube 

 pushed down through the opening. When the con- 

 tents of the flask have been poured through, and a 

 minute or two allowed for drainage, the glass tube ia 

 drawn up until the tap can be closed, and then the 

 flask and connecting-tubes are removed. The pro- 

 portion which I propose to maintain between the air 

 and the pentane is 600 volumes of air to 1 volume of 

 pentane, measuring the liquid at or near 60 Fahr. ; 

 or, measuring both as gases, 20 of air to 7 of pentane, 

 the vapor-density of pentane being such that it occu- 

 pies as gas, under the normal conditions of 60 Fahr. 

 and 30 inches barometric pressure less the tension of 

 aqueous vapor at 60, 210 times the volume -which it 

 occupies as a liquid. 



When the pentane is poured upon the water in the 

 gas-holder with thrice its vapor-volume of air above, 

 it volatilizes rapidly and completely. A few minutes 

 are sufficient for the volatilization of the liquid, and 

 a few hours suffice for perfect diffusion. It is clearly 

 essential to the uniformity of the air-gas thus made 

 that the liquid should be free from any admixture 

 with non-volatile hydrocarbons, which would ac- 

 cumulate on the surface of the water, and dissolve 

 or give up portions of the gaseous hydrocarbons ; but 

 the fulfillment of this condition is insured by the 

 repeated rectifications which are necessary to separate 

 pentane from the hydrocarbons of higher boiling- 

 point. It is also essential that this vapor of gas 

 should be so slightly soluble in water that the pro- 

 portion in the air-gas standing over the large volume 

 of water in the tank of the gas-holder may not be 

 appreciably affected by changes in the temperature 

 of the water. Fortunately the gaseous paraffins are 

 most sparingly soluble in water. I have inclosed the 

 vapor of pentane at the tension of 261 millimetres 

 over boiled-out water in an eudiometer for 24 

 hours, and 100 volumes of water dissolved only 

 0.92 volume of the gas. No doubt the first sample of 

 gas made in the gas-holder filled with fresh water 

 would suffer and I find that it does suffer some . 

 diminution in the proportion of pentane ; but the 

 error, if standard gas is Kept in store for photometry, 

 fresh portions being made from time to time over the 

 same water, would only be such as could arise from 

 the difference between the solubility of gaseous 

 pentane, under a pressure of a quarter of an atmos- 

 phere, at one temperature and at another; and this 

 error, though real, is likely to be infinitesimal. Many, 

 and, as far as I know, all other substances, which 

 otherwise might be used as the luminous ingredi- 

 ents of a standard gas, such as olefiant gas, or ether, 

 or benzine, are excluded by the necessity of storing 

 and measuring gas over water, and the comparative 

 solubility of these substances in water. 



The product is a perfect gas not suffering con- 

 densation except under conditions that would not 

 occur in practice. 



Another problem remained, namely, to devise a 

 fitting burner. Prof. Harcourt exhitiitd a burner 

 consuming standard gas at the rate of half a cubic 

 foot an hour, and producing a flame whose height 

 from the surface of the burner to the luminous 

 tip, which in still air is almost as fixed and definite 

 as the ivory point marking the level of the mercury 

 in the reservoir of a barometer, is 2iV inches. Not 

 only is the height of the flame serviceable as a checi 

 on the preparation of the gas, but the regulation 

 of the height of the flame is more important for 

 the production of a fixed amount of light than the 

 regulation of the rate of consumption. 



The unit which I propose, and which has heen 

 adjusted to correspond to the light of "a sperm , 

 candle consuming 120 grains of sperm per hour," is : 

 " The light given by a mixture of 7 volumes of pen- 

 tane gas with 20 volumes of air, burning from a i-mch 

 orifice at the rate of half a cubit foot per hour, under 

 the standard conditions of 60 Fahr. and 30 inches 



