ART. XII.-Contributions from the Sheffield Laboratory of Yale College. No. XIX.

1. On the estimation of Carbonic Acid ;* by S. W. JOHNSON.

IN the processes usually employed, carbonic acid is determined by the loss of weight of an apparatus which contains no carbonic acid gas at the beginning and which must be completely emptied of this gas at the conclusion of the analysis. It is a matter of experience, however, that accurate results are not attainable with certainty, in this way. Nothing short of actual boiling for some time will expel all carbonic acid gas from the dilute acid liquid. This cannot be done conveniently without loss of aqueous vapor. The fact that good results are often obtained is due to the compensation of opposite errors, as the analyst may convince himself by repeatedly heating and sucking through air. If the suction go on to just the right extent, the loss of the apparatus will exactly correspond to the carbonic acid that was contained in the substance, but further exhaustion of the air will diminish the weight of the apparatus, not by complete removal of the carbonic acid, but by loss of aqueous vapor which easily escapes the desiccating material. By continued working on a carbonate of known composition one may soon learn how long to exhaust in order to bring out the proper loss, but where the analyst is out of practice, an error of 1 to 2 per cent is not unlikely to happen, and the process itself furnishes no means of judging when it will give a correct result.

The writer employs a simple modification of this method which, under proper conditions, gives very accurate results and furnishes to a great extent its own control. The process is novel in this particular, viz: the charged apparatus is in the first place filled with carbonic acid gas, the substance is then decomposed, and as soon as disengagement of gas ceases, the apparatus, still filled with carbonic acid gas, is weighed again. In this manner all aspiration is done away with, and the desiccating material has simply to dry as much gas as is yielded by the substance under analysis.

It is, however, essential that the substance under examination dissolve freely and completely in cold acid; it is likewise necessary that the analysis and weighings be conducted in an apartment not liable to change of temperature.

The apparatus may consist of a light flask or bottle with

The substance of this paper was verbally communicated to the National Academy of Sciences in August, 1868.

wide mouth which is closed by a soft rubber stopper, through which there passes, on the one hand, a chlorid of calcium tube, the lower bulb of which contains cotton, and, on the other, the neck of a vessel which contains the dilute acid. This acid reservoir is so constructed that on suitably inclining it, its contents will flow freely into the flask. For this purpose the tube connecting with the latter has an internal diameter of seven millimeters, and its extremity is cut off obliquely; at its other end, the acid reservoir terminates in an upturned narrow tube, b. This and the upper termination of the CaCl tube are chosen of such diameter that they fit quite snugly into short, narrow and thick-walled rubber connecters which are again provided with glass-rod stoppers; all these joints must be gastight. In the figure the apparatus is represented in one-third its proper dimensions.

The weighed substance, in case of carbonate of lime, e. g., is placed at the bottom of the flask, most conveniently in the form of small fragments. The acid vessel is nearly filled with hydrochloric acid of sp. gr. 1.1. It and the CaCl tube are tightly adjusted to the neck of the flask, and the glass-rod stoppers being removed, the apparatus is connected at c with a selfregulating generator of washed carbonic acid, and a rather rapid stream of the gas is transmitted through the apparatus for 15 minutes, or until the liquid in b is saturated and the air is thoroughly displaced. Then the opening at d is stopped and afterward the apparatus is disconnected with the carbonic acid generator and stopped at c. During these as well as the subsequent operations, the apparatus must be so handled that its temperature shall not change. It is immediately weighed. When removed from the balance, loosen the stopper at d, and, holding the flask by a wooden clamp, incline it so that the acid may flow over upon the carbonate. The decomposition should proceed slowly, so that the escaping gas may be thoroughly dried. As soon as solution of the carbonate is complete, replace the stopper at d and weigh again. Should there be any leak in the apparatus the fact is made evident by a slow but steady loss of weight, when it is brought upon the balance. If all the joints are sufficiently tight, the weight remains the same for at least fifteen minutes.

When properly executed the process gives extremely accurate results; a slight change of temperature or of atmospheric pressure between the two weighings of course greatly impairs the results or renders them worthless. Since the apparatus usually rises a little in temperature during the solution of the carbonate, it is better, as soon as the substance is decomposed, to stopper the CaCl tube and let the whole stand fifteen minutes, then to connect as before with the CO generator and pass dried CO, for a minute, and finally to stopper again and bring upon the balance. In seven analyses of pure calcite in quantities ranging from 0.5 to 0.9 grm., the writer obtained the following percentages of carbonic acid, viz: 44.07, 44·07, 43·98, 44·01, 4404, 44·11, 44·16; calculation requires 44.00.

In case of alkali-carbonates which absorb carbonic acid gas, it is necessary to modify the apparatus. Instead of the light flask a we may employ a small bottle of thick glass and wider mouth, and a thrice-perforated rubber stopper. Through the third orifice pass a narrow tube 3 to 4 inches long enlarged below to a small bulb to contain the carbonate. This bulb must be so thin that on pushing down the tube within the bottle, it shall be easily crushed to pieces against the bottom of the latter. The carbonate is weighed into the bulb-tube, the latter is wiped clean, down to the bulb, corked and fixed in the stopper. The apparatus is filled as before with CO, and weighed. Then the bulb is broken and the process finished as before described. In three estimations on carbonate of soda the writer found 41.54, 41.64 and 41:58 per cent of CO,. Calculation requires 41-51 per cent.

2. On the construction of Bunsen's Air-Pump.

In supplying the Sheffield Laboratory with the means of quick filtration, the following plan has been adopted. The laboratory being on the ground floor, with no well accessible, and having hydrant water under an adequate head, a lead supplypipe of inch bore was connected with the service pipe and carried up an empty flue to the garret. There it returned in a short curve, and at a height of about 35 feet, was soldered into the end of a vertical piece of lead pipe 8 inches long and of 1 inch bore, within which it penetrated two inches. Another leaden air-pipe of inch bore was also soldered into the top of the inch tube and ran down parallel with the one just mentioned, terminating at the laboratory desk below. To the bottom of the inch pipe in the garret flue was attached a third leaden fall-pipe of 4 inch bore which passed down to the cellar bottom with a vertical height of more than 40 feet, and there entered into a drain pipe. The supply-pipe and air-pipe were AM. JOUR. SCI.-SECOND SERIES, VOL. XLVIII, No. 142.-JULY, 1869.

brought out at a convenient point above the assistant's desk The former was provided with two cocks, one to regulate the flow of water and another for emptying it in cold weather.

At 3 feet above a shelf of the desk, a branch was fitted to the air-pipe which was connected by rubber to a vertical glass tube whose lower end stands in a small vessel of mercury and serves as a manometer. This arrangement was put up by a plumber in two days' time and gives the full Torricellian vacuum less the tension of water-vapor.

For the working students, four similarly constructed pumps were made, each being in the center of a double table that accommodates four operators. In case of these pumps it was not practicable to place the pipes within flues; so they are carried up to the ceiling of the laboratory, a height of 13 feet, and the fall-pipes pass below into a drain in the cellar bottom. The pipes are supported on the faces of a narrow vertical plank secured above to the ceiling, and below to the reagent shelf. The glass manometer-tube is brought into a slot in this plank so as to stand in full view of all four operators. The water supply is regulated from either side of the desk by a cock with a double lever-handle eight inches long, which works against a stop when open to the point of maximum exhaustion. The air-pipes terminate in a horizontal brass tube screwed on a shelf and open at either end for connecting with the filtering apparatus. These pumps when in action support a column of mercury of about 19 to 20 inches, which is sufficient for most ordinary purposes, and their use is in great favor with the students. It is needful to intervene between the filter flask and the metal pipes a small bottle to collect the liquids which condense in the latter, as they contain lead in solution or suspension.

ART. XIII-On the Aurora seen in New York, April 15, 1869; by W. S. GILMAN, Jr. (In a letter to Prof. ELIAS LOOMIS, Yale College.)

I SEND herewith a brief account of the grand aurora of last evening, as seen by a party of three from the roof of the Observatory of Mr. Jacob Campbell, in Brooklyn.

On coming out of doors to go to the Observatory at past 7 in the evening the writer called the attention of one of the party to thin streaky clouds pointing upward from the horizon, and remarked that it was the beginning of an auroral display. Hardly had we paused when by their changing, wavy light we perceived we were not in error, and in a few minutes afterward saw the formation of the corona as in the sketch below.

N

The nucleus or apex of the phenomenon was a cloud, shaped like the internal area described by a bow with the string drawn aside. The southwest was clear, as it was generally during the evening. From this point the wind had been blowing all day.

s'

M

S

To the south the auroral beams E did not extend nearer the horizon than Sirius, and indeed it was noticeable that at no time did the display exhibit great strength on the horizon.

Aurora of April 15, 1869, 7:40 P. M., as seen from space. M, Mars; O, Orion; P, Polaris; S, Sirius; Z, Zenith.

At 7.40 rosy tints appeared at different points of the compass. Remarkably deep tints were noticed both to the east and west of the zenith, and generally at an altitude of 45° or more. At 8 P. M. the phenomenon brightened after a dull period of some minutes. The rose tints were now very beautiful. In the west a very deep carmine was observed. Sulphurous-yellow and greenish hues were also noted.

At 10 P. M. the aurora was faint, but still quite striking in its arrangement.

W

At 10.30 it brightened again, the corona reappearing with

M

N

E

Aurora of April 15, 1869, at 10 30 P. M., as seen from space. M, Mars; P,

Polaris; S, Sirius; Z, Zenith.

N.N.E. to N.N.W.

redoubled brilliancy. The nucleus of light overhead was much larger than earlier in the evening, and in shape resembled the denser part of the Great Nebula of Orion. Indeed the entire phenomenon wore the appearance of that nebula, as if it had been suddenly brought near to our system, so as to cover a quarter of the sky.

At 2 A. M. we ceased observations. At this hour there were still faint beams in the north proceeding from the dark segment along the horizon, from the

I rank this aurora as inferior in brilliancy to the grand aurora I witnessed at Mount Desert, Maine, on the 15th of September last, but possessing more interest than the latter in the greater grotesqueness of the clouds forming the corona, and