Page images
PDF
EPUB

Referring to my former paper above mentioned for details regarding the construction and use of the apparatus employed, I need here describe only such modifications as have been found expedient to adapt the process to this special purpose.

1.

The combustion-tube being packed with pure asbestus betweeen the points a and b, fig. 1, and the spaceabout two inches in length -between b and c left vacant, a plug of pure asbestus is placed at c, and

a

d

the space between c and d, about three or four inches in length, then filled with a mixture of pure asbestus and peroxyd of lead, and finally a plug of asbestus is placed at d. As the sulphuric acid formed is to be absorbed by, and finally determined from, the peroxyd of lead,-in order to obviate the necessity of treating the whole of the asbestus in the tube to obtain the sulphuric acid, which would be troublesome, and at the same time preserve the asbestus packing in the posterior part of the tube in a fit condition for future use,-it is important that the asbestus plug at c should be packed closely enough to prevent any particles of the peroxyd of lead from passing back of this plug.

As already stated, the object of mixing asbestus with the peroxyd of lead is to prevent the formation of a channel along the top. In this manner but a short column of the mixture of asbestus and peroxyd of lead will suffice to secure complete conversion of the sulphurous acid. The combustion is conducted precisely as for the determination of carbon and hydrogen alone, except that the portion of the tube which contains the peroxyd of lead is maintained at a gentle heat, sufficient to prevent condensation of water in that part of the tube and at the cork, but avoiding a temperature which would decompose the peroxyd of lead. As usual, the water formed is absorbed in a chlorid of itself, so heavy a powder. Through such a channel sulphurous acid might pass, in so small proportion, without coming in contact with the peroxyd of lead. It will be seen that the liability to the formation of a channel is obviated in my process by mixing the peroxyd of lead with a large proportion of asbestus. The asbestus serves also to increase the porosity of the mass, and in this manner also to lessen the chances of escape of sulphurous acid without coming in contact with the peroxyd. I may here add that, in making the combustion with oxygen in presence of asbestus, the quantity of sulphurous acid which reaches the peroxyd of lead is by no means very large. In a preliminary experiment, in which carbonate of soda was employed instead of peroxyd of lead, (the substance burnt being bisulphid of carbon), the carbonate of soda was found to contain within about 9 per cent of the equivalent of sulphur; and a portion of the deficiency it is not unlikely may have been taken up by the impure asbestus that was employed in this instance.

Concerning the other source of error in the determination of carbon which Carius mentions, it will suffice to remark that, in my process, the peroxyd of lead is kept at so high a temperature that the absorption of carbonic acid appears to be prevented.

AM. JOUR. SCI.-SECOND SERIES, VOL. XLI, No. 121.-JAN., 1866.

calcium tube, and the carbonic acid in Liebig's potash bulbs with a mulder tube attached.

2.

After the close of the combustion, when the tube shall have sufficiently cooled it is carefully removed from the furnace, the mixture of peroxyd of lead and asbestus cautiously drawn out into a beaker glass, by means of a bent iron wire, and the tube then inverted within another tube, e e, closed at one end, as shown in fig. 2. The mixture of peroxyd of lead and asbestus contained in the beaker glass is now treated with a strong solution of bi-carbonate of soda, and left to stand for about twenty-four hours, with frequent shaking."

Solution of bi-carbonate of soda is also poured into the tube e e until the level of the liquid shall have reached a point, f, on the combustion-tube, a little above that which was occupied by the plug c, and this is also left to stand as the other. After the lapse of sufficient time for the reaction to be completed, the solution is filtered from the asbestus mixture, including also the solution in the tube ee, and not omitting to carefully rinse out the anterior portion of the combustion-tube. The asbestus mixture upon the

filter is then thoroughly washed, the filtrate concentrated by evaporation, and the sulphuric acid precipitated with chlorid of

[graphic]

barium.

The following results of analyses of bi-sulphid of carbon indicate the degree of accuracy afforded by this process.

The preparation employed was commercial bi-sulphid of carbon, which was first subjected to re-distillation.

Analysis 1. 01414 gram of bi-sulphid of carbon gave 0.0806 of carbonic acid, and 0-8592 of sulphate of baryta.

[blocks in formation]

Analysis 2. 0.274 gram of the same substance gave 0.158 of carbonic acid, and 1.6768 of sulphate of baryta.

[blocks in formation]

'H. Rose, Chimie Analytique, new French edition, p. 662.

Analysis 3. In this analysis, in which I was prevented from determining the carbon, 0·1537 of bi-sulphid of carbon gave 0-9461 of sulphate of baryta, corresponding to 84.5 per cent of sulphur.

The mixture of asbestus and peroxyd of lead employed was of that which had already been used in the preceding analyses, and may possibly have contained a trace of undecomposed sulphate of lead, as the per-cent of sulphur found in this case is 0.3 per-cent above, while in the preceding analyses it was a fraction below the theoretical quantity. Trusting, however, that the results already obtained will be deemed sufficient to show the method to be a good one, I have not thought it advisable at this time to further repeat the analysis of this substance. I may here state that I have already applied the process in the analysis of bodies containing hydrogen, and have obtained satisfactory results which will soon be published.

The important advantage thus gained of being able to deter mine the different elements from the same portion of substance, eonsidering also the simplicity of the process, can hardly fail, I think, to secure for this preference over the older methods.

ART. IX.-Description of an Automatic Registering and Printing Barometer; by G. W. HOUGH, A.M., Director of the Dudley Observatory.

THE science of meteorology is as yet in its infancy. Universally interesting as its phenomena have ever been, and powerfully affecting the most important relations of society, it is but recently that the subject has engaged the systematic and com. bined effort requisite for its development, since its laws are still regarded as the most recondite problem in Physics. The first thing to be done is of course the collection of facts, and much is now being done in England and on the continent in this direction. The chief obstacle, hitherto, has been in the imperfection of the methods of observation. The results, in order to be of value as data from which to construct a science, should present a continuous record of the phenomena during a considerable period of time, and taken at as many different stations as possible. By the ordinary method of personal observation, this is well nigh impracticable. It would demand at every station the services of several observers, at great expense, and their results could only at best be more or less of an approach to what is desired. To obtain this, the only alternative is to substitute some mechanical means for the labor of personal observation; in short, to make the instrument record its own changes. If this can be done in a single instance, it can be done continuously.

The only method by which this has been hitherto attempted with success has been by the application of photography. This, though a very considerable advance, and probably all that could be desired in respect of continuity and accuracy of the record, is liable perhaps to the objection that it is too complicated a process for general use. If we consider the skill requisite in the preparation of the paper, the delicacy of manipulation involved by the apparatus, and the labor of interpreting the results, as compared with the average capacity and means of the great number of observers desired and likely to volunteer or be employed for such a purpose, it would seem that a simpler process is both desirable and necessary. This it has been my intention to furnish, and with what success remains for time and experience to determine. The importance of the subject will justify me perhaps in presenting some account of the new method.

The problem to be solved, was to cause any meteorological instrument, by means of suitable mechanism, simply and effectually to record its own changes. The instrument selected for experiment was the barometer. When any delicate instrument is made to record its own changes by mechanical means, the chief difficulty is that of getting sufficient power for the mechanism attached to make a distinct and continuous record, without taking a perceptible amount of force from the instrument itself, and thereby vitiating the results. The use of electricity naturally suggested itself as the best means of overcoming this obstacle. This agency has not as yet been made economical or certain as a motor, but is chiefly valuable in controlling power obtained through some other means. By it, as may be seen in its application to clock work, and in the telegraph, the movements of one machine may be reproduced in an other with no greater expenditure of force than is requisite for electrical contact. In the cases cited, however, the motion to be reproduced is sensibly uniform and in the same direction. For the solution of our problem, a mechanism is demanded that shall repeat the changes of the original in every form, whether the motion be uniform or variable, forward or reverse.

The feasibility of this plan was discussed with my friend Mr. Thomas Simons as early as the year 1862, and some steps were then taken to apply it to the thermometer. I may here express my acknowledgments to Mr. Simons for valuable suggestions in the construction of the present machine. Various plans were considered for effecting the electrical contact with the fluctuating medium which is the basis of this method. It was at first proposed to do this at the surface of the mercury in a siphon barometer, by means of a platinum wire which should be carried continually toward the mercury surface by suitable mechanism, and on touching the surface, a galvanic current would be formed

which should operate by an electro-magnet on the mechanisin so as to reverse the motion of the wire and break the circuit. This would be immediately restored by the normal movement of the mechanism, and thus the point of contact would be kept oscillating at the surface continually. The consumption of battery power by this plan would have been considerable, and it was thought the oxydization of the mercury by the electric circuit would in time be appreciable. It was therefore concluded to make the connection outside of the barometer tube, by means of a float resting upon the mercury column. By this plan there is no demand of action from the battery until some change takes place in the barometer, and a considerable saving of battery elements is effected.

Attention was then given to determining the degree of delicacy with which changes of the mercury surface could be represented by this process. It was found by experiment that a motion of less than 0005 of an inch was readily shown, a quantity within the limits of reading of a first class standard barometer.

The next step was to devise the proper mechanism for repeating the motion thus transferred, and recording it in some legible form. A finely cut screw was considered as best adapted to measure such minute intervals of space. To this screw a forward or reverse motion was given by a double system of clock work, each operated by an electro-magnet in connection with the float, and raising or lowering the screw by intervals corresponding with the changes indicated in the mercury column.

In respect to the permanent record of results, it was decided not only to attempt the production of a linear diagram or curve of atmospheric pressure, as an interesting method of presenting the recorded changes to the eye, but to avoid the tedium and uncertainty of measuring up such results, by producing at the same time a printed record of such variation, to any extent deemed advisable.

Having thus endeavored to give some conception of the design. and principal features of this method, I will proceed to explain more fully the details of its execution as at present arranged.

In order to make any self-recording machine of this kind practicable, we need to attend to two points. First, to reduce the consumption of electricity to the smallest possible amount consistent with certainty in the results; and secondly, to secure the greatest amount of useful work with the minimum of labor. We at once decided to adopt the "make" circuit; for so long as there is no motion, there will be no consumption of battery elements. The battery which we have adopted for recording transits is essentially that of Daniell; sulphate of copper being the exciting agent. A battery of this kind will maintain sufficient power for chronographic records for two or three months,

« PreviousContinue »