ume of the tubes between the three-way stopcock e, and the stopcocks of the globes, a a. But here the volume of the air drawn from the burette to fill the vacuous tubes had to be corrected for difference of temperature as well as for degree of exhaustion; the determination was made with these tubes covered with ice, just as they were to be covered with ice in the measurement of the pressure of hydrogen in the globes. Afterwards, the three-way stopcock was connected in the same way to the tube leading to the barometer and its capacity determined. The level of the mercury was noted, and determinations of volumes made at different levels. Taking account of the degree of exhaustion and of the differences of temperature at the burette and at the barometer, if any, the volumes contained to different levels in the barometer were computed, and a table made showing them. The tube leading to the barometer, which had been separated, was now fused in place, and a few determinations made of the united volumes.

As an example of the process is given the first determination made.
Blank space, 0,3 cc.

Connecting tubes which are kept covered with ice; vacuum, 6 mm., burette, 19.0°. Air drawn in on opening stopcock, 48.1 cc., 48.4 cc., 48.5 cc., 48.5 cc.; mean, 48.4 cc. ; reduced to 0° and corrected for 6 mm., 45.6 cc., the capacity of the ice-covered tubes. Fused three-way stopcock to tube leading to barometer; blank space same as before. Air left in vacuum, negligible.

Tubes kept in ice.

The details of the computation are unimportant. The tubes were renewed several times; the table following gives their capacities at different dates :

CAPACITY OF CONNECTING TUBES AT DIFFERENT DATES.

Barometer and tubes leading to it.
Mark.

Nov.-April

May-Oct.

January 18

66

240

41.1

56.9

53.8" 50.8"

270

46

47.7

41.6"
38.6

66

29.4
26.4"

15.-HYDROGEN BY NEW METHOD. EXHAUSTION OF GLOBES.

When the capacity of the connecting tubes had been determined, the stopcocks of the globes were opened, the pump was set in action, and a sufficient exhaustion obtained and measured with the McLeod gauge which was mentioned on page 18. The measurement was not made till about half an hour after the pump had been stopped. The automatic pump by acting during one night would produce a vacuum of ten, twenty, or forty millionths, according to circumstances; when acting through one day and two nights, it would leave from two millionths to three ten-millionths of an atmosphere. But such a degree of exhaustion was unnecessary in experiments in which simply the increase in pressure due to the admission of a known weight of hydrogen was to be determined.

16. HYDROGEN BY NEW METHOD.

MANIPULATION OF TUBE CONTAINING PALLADIUM.

It was thought to be of prime importance to use on the tube containing palladium no stopcock requiring lubrication or admitting the possibility of leakage. It is necessary to determine the volume of this tube before and after each experi ment. If the lubrication of the stopcock is exposed to water during the hydrostatic weighings, a source of uncertain error is introduced. It would be possible to wait till the effect of this exposure should be thought to have passed off, or to immerse the palladium tube only to a certain mark in the hydrostatic weighing. But the manipulation without any stopcock is easy and more accurate, and was always used in the experiments of this series, and in all the following experiments by this method.

A strong argument in favor of dispensing with the stopcock is the difficulty of knowing that, in a given case, the stopcock did not leak. A stopcock is often found, after an experiment, not to have leaked. I have repeatedly exhausted a globe, weighed it, connected it again to the pump, exhausted the connecting tubes to the same degree as before, and then have opened the globe and again measured the vacuum. When the lubrication was freshly applied, not infrequently the amount of leakage in one, two, or even four days has been found to be negligible. But and even purpose, these stopcocks were of very special dimensions ordered for the these often leaked, causing great annoyance and loss of labor. If some such measurement, or other proof that leakage has not taken place, cannot be applied, it would be necessary to use some method which should entirely eliminate leakage.

The tube holding the palladium was therefore arranged as shown in Fig. 25. At d is the tube by which hydrogen is admitted; at e is a drying tube, filled with phosphorus pentoxide; at g is a plug of fusible metal, at h is a small bulb filled

with asbestos, to prevent loss of fusible metal by projection when melted. i is a ground joint by which the tube can be connected with other apparatus. The tube between 9 and the point k is exhausted and then closed at k by fusion, and a notch made between i and k.

g

the

FIG. 25.-Palladium tube with no stopcock,

When the palladium is to be charged with hydrogen, is fused to the source of gas, and the palladium is heated. When the temperature is such that no absorption would take place, hydrogen is passed through the tube from d to a, the tube having been broken off at this point. The heat is raised, and the current continued for half an hour, when a is closed by fusion, and the heat withdrawn. Hydrogen is admitted from an apparatus in which it is produced by electrolysis of sulphuric acid in voltameters, and then freed from oxygen, and compounds containing oxygen, by passing over heated copper, potassium hydroxide, and phosphorus pentoxide. When the pressure of the gas in the palladium tube becomes equal to that of the atmosphere, the point a is again broken off, and the current continued for three or four hours; during part of this time, the gas is allowed to escape at f. By this long-continued passage of the gas escaping at a and f, it was hoped to remove very nearly all the nitrogen which, if once present in the gas, would not be removed by the purifying train. Since the tension of the gas in the palladium, especially when it is saturated, is considerable, the removal of the nitrogen with the air-pump cannot be effected.

When the charging with hydrogen is completed, the ends a and ƒ are closed by fusion, and the connection with the source of hydrogen at d is closed in the same way, when the palladium is ready for weighing.

17.-HYDROGEN BY NEW METHOD. HYDROSTATIC WEIGHING OF TUBE CONTAINING

PALLADIUM.

When the tube was charged with hydrogen, and closed by fusion, its volume was determined by hydrostatic weighing in a way which calls for no remark. The tube exposed nothing to the water but a continuous surface of glass. It was therefore not likely to suffer much change of weight by immersion. The tube was of the common German glass; it would now be possible to construct it of one of the new sorts of glass which resist the action of water in a much greater degree. But

the exposure to water lasted less than two minutes, and solubility of the glass was thought to be negligible.

18.-HYDROGEN BY NEW METHOD. WEIGHING THE TUBE CONTAINING PALLADIUM.

The tube containing palladium was weighed against a counterpoise of nearly equal volume and weight. The volumes on the two pans of the balance were then made equal within the twentieth of a cubic centimetre by means of the small equating flasks of various volumes but of weights made equal if weighed in a vacuum, which were mentioned on page 48. The tube of palladium was hung on one of the auxiliary pans of the reversing mechanism shown in Figs. 16 to 19, on pages 38 to 41, and its counterpoise on the opposite pan, the equalizing flasks were added, and the weights were laid on the pan which carried the palladium. After the closet had been shut for an hour, weighings were made by reversals just as in the second and third series of determinations of the density of oxygen. oxygen. Weights smaller than ten milligrammes were not put on the auxiliary pans, but were put on the pans of the balance itself; the fractions of a milligramme were determined by observation of the value of a scale division and computation from the position of equilibrium in the two weighings by reversal. Four weighings at suitable intervals were thought sufficient; but sometimes time was allowed for more.

19.-HYDROGEN BY NEW METHOD.

INTRODUCTION OF HYDROGEN INTO THE GLOBES.

The palladium tube was put in an iron trough containing magnesia, and laid on the furnace where it was to be heated. The point , Figs. 25 and 26, was broken off at the notch previously made, and scrutinized to see if some minute fragment were detached. Air was in this way admitted as far as the plug . To the ground joint i, was then cemented the corresponding joint 7, and this was then fused to the tube m n o p, Fig. 27, by which hydrogen was to be admit

to

ted to the globes; o in Fig. 27 is the same as FIGS. 26 and 27.—Palladium tube, ready to connect to apparatus. Palladium tube, connected d in Fig. 23. The air-pump was then con- apparatus, connection exhausted. nected to n, Fig. 27, the space betwen g and o was exhausted, and the connection with the pump was closed by fusion. After some few minutes, the fusible metal valve at o, Fig 27, d, Fig. 23, was opened by fusion; if the joint at 7, Fig. 27, and the fusible valve at y were tight, the fused metal would not be forcibly projected. Before this time the vacuum in the globes had been measured, and the

connection with the pump had been closed. The fusible metal plug g was now removed, and the palladium was heated.

As

When the delivery of gas ceased, d, Fig. 23, was closed by fusion. the palladium cooled off, most of the hydrogen remaining between g and o, Fig. 27, was absorbed. If the vacuum obtained originally in this space was equal to the vacuum produced by this re-absorption, this blank space had absolutely no effect on the experiment. When the palladium was cold, the tube was closed by fusion at g, Fig. 25, i was separated from the joint 7, Fig. 26, and cleaned, and the whole was made ready for weighing.

20.-HYDROGEN BY NEW METHOD.

MEASUREMENT OF PRESSURE.

When d, Fig. 23, had been closed by fusion, the pressure of the gas in the globes was determined by observations of the syphon barometer, just as in series second. Ice was heaped around the connecting tubes as shown above e, and the ice around the globes was renewed. Then readings were taken, as has already been sufficiently described.

The readings were continued at intervals for one hour; sometimes, in test cases, for a day, but there never seems to have been reason for suspecting the accuracy of a reading taken after fifteen minutes from the end of the filling, when the ice had been renewed an hour previously.

21.-HYDROGEN BY NEW METHOD.

SECOND WEIGHING OF THE PALLADIUM.

The tube containing the palladium now consisted of three parts: the large part, closed again by fusion, the short tube containing the asbestos, and the point which had been broken off at k, Fig. 26. The small parts were commonly united by fusion. The large part again presented a continuous surface of glass; its volume was determined by hydrostatic weighing. The volume of the smaller parts was computed from their weight and specific gravity. In some earlier experiments for a different purpose, the volume of the palladium tube was changed a little by the heat which expelled the gas; but this did not occur in any experiment of this series. If any change had occurred, the second hydrostatic weighing would have shown its amount.

The tube was now weighed against the same counterpoise as before, with the exception that the flasks for equalizing volumes were not the same as before. A larger volume of weights was now required, and the palladium tube was smaller by the capacity of the part between 9 and 7, Fig. 25, which was closed when first weighed, but was now open to the air. These effects nearly counterbalanced

each other.