extinguished by slight error in adjustment of the stopcocks for admitting the gases, the pressure of the hydrogen would, before it was possible to make readjustment, so increase as to blow off through the mercury, as once happened even when the safe pressure was two atmospheres, and the two atmospheres, and the working pressure only half an atmosphere.

8. SYNTHESIS OF WATER. THE COMBUSTION.

The combination of forty-two litres of hydrogen with twenty-one litres of oxygen took about an hour and a half. When the When the pressure of the oxygen became very small, extreme care was required to continue the combustion. I have once carried it till the pressure of oxygen and water vapor together was only eight centimetres of mercury. The attempt to do the same again was unsuccessful.

When the combustion was ended, the stopcocks m and n were closed, the palladium tube was closed by fusion at the point 1, Fig. 34, and the globes containing oxygen were closed at their stopcocks. In experiments numbered 10, 11, and 12, the globes were also closed in the same way as the palladium tubes.

The lower part of the combustion chamber was now immersed in a freezing mixture. When the water produced had been frozen and cooled as much as possible, the stopcocks of the apparatus were opened, the mercury was withdrawn from the tubes 4 and 5, and the Toepler pump was used to transfer the gas remaining in the apparatus and its connecting tubes to an eudiometer. The tension of the vapor of the water at -18° C. being only 1.1 mm., the exhaustion could be made sufficiently complete. The phosphorus pentoxide tubes showed no signs of harmful deliquescence during the half hour they had to serve.

When the exhaustion was complete, the combustion apparatus was closed by fasion at hand h, Fig. 36. The parts fused off from the different parts of the apparatus were now cleaned from adhering wax or paraffin, and the points which had been broken off were put with those to which they belonged. The four parts of the apparatus were then weighed.

9. SYNTHESIS OF WATER. WEIGHING THE OXYGEN AFTER A COMBUSTION.

It is obvious that the volumes of the globes when full and when partly empty would differ by the amount of the corresponding elastic compression. In the first and second experiments, the oxygen was withdrawn from the globe as completely as possible; the globe after the experiment was therefore weighed against the counterpoise proper for the empty globe, and a correction computed for the difference of volume. In the experiments from the third to the seventh, the

oxygen was nearly half withdrawn; the globes were therefore weighed, one against the counterpoise for the empty globe and the other against the counterpoise for the full globe, and a correction computed for the difference of volumes; the two corrections nearly balanced each other. In the eighth and ninth experiments, the globes, both full and half empty, were weighed against the counterpoise of the empty globe, and corrections computed; the counterpoises for the filled globes had suffered injury. In the last three experiments, the globes were closed by fusion; they were weighed when half empty against the counterpoises of the full globes, and corrections computed for the difference in volumes.

10.-SYNTHESIS OF WATER. WEIGHING THE PALLADIUM AFTER A COMBUSTION.

The volume of the tube containing palladium was determined by hydrostatic weighings. In the cases where the tube was provided with a stopcock, it was immersed only to a certain mark at this second hydrostatic weighing. When no stopcock was used, the tube was submerged. In all cases, the volume of the part which had been fused off was determined from its weight and specific gravity.

From the change in the volume of the tube a correction to the observed weight was computed. These experiments were made before the construction of the system of flasks for making equal the volumes of the masses placed on the opposite pans of the balance.

11. SYNTHESIS OF WATER.

EUDIOMETRIC ANALYSIS OF THE RESIDUE LEFT
UNCOMBINED.

The gas extracted from the apparatus after the combustion was ended, was measured and analyzed. Only one point needs mention.

The amount was not
If it came from car-

In some experiments some carbon dioxide was found. large, but the question as to its origin was an important one. bon weighed with the hydrogen, the experiment was valueless. The probability was that it came from the asbestos, where the combustion was carried on for a few seconds. This asbestos had to be handled in the construction of the tubes, and it was not easy to free it from dust by ignition, as, after ignition, my sample was not well suited for its use. To settle the question, some two or three hundred grammes of water were formed in a combustion apparatus of the same kind, but without weighing the gases introduced. The water produced was allowed to flow out, as formed, by a tube fused to the bottom of the apparatus; it was kept from contact with the atmosphere. In this case there was no asbestos, but the hydrogen was the same as that used in all the experiments.

To the water produced in this experiment, barium hydroxide was added, and the whole showed no trace of turbidity. It was then evaporated to the bulk of five cubic centimetres; of course carbon dioxide was not admitted. Even when reduced to this bulk, there was no turbidity. It is obvious, therefore, that the hydrogen used in these experiments did not contain carbon compounds. It is precisely the same in quality, obtained from the same apparatus, with the same solution, as that used in all my experiments so far; the only change ever made has been the introduction of fresh distilled water to take the place of that removed by electrolysis.

If now this carbon dioxide was formed by the oxidation of organic matter adhering to the asbestos, it was supposed that each cubic centimetre of carbon dioxide formed represented a cubic centimetre of oxygen withdrawn from the globes but not used in forming water.

In some experiments, nitrogen was also found. The quantity was small. When the globes for oxygen contained air, and were exhausted and filled with oxygen, a certain amount of nitrogen remained in the globes. If the globes were exhausted again when partly filled with oxygen, the nitrogen was mostly eliminated. As the amount of nitrogen found agreed fairly in amount with what might be expected from the previous contents of the globes and the degree of the last exhaustion, no hesitation was felt in counting this nitrogen as having come from the globes of oxygen.

12. SYNTHESIS OF WATER. SOURCES OF ERROR.

In the weighing of the hydrogen, no source of appreciable constant error has yet been thought of. The magnitude of the accidental errors must be learned from the concordance of the experiments.

In the weighing of the oxygen, leakage might act in two opposite directions. If leakage occurred before the combustion began, the amount so lost would be extracted by the Toepler pump in the exhaustion preliminary to the combustion, and so would seem to have been used in the combustion. If the leakage occurred in the contrary direction, when the half-exhausted globe was hung on the balance for the final weighing, the amount of oxygen taken from the globe would appear less than the truth. Now the leakage during exhaustion, if noticeable in amount, would be readily detected by the behavior of the pump during the preliminary exhaustion. A very slight leak can be detected when the vacuum obtained is something like a ten-thousandth of an atmosphere. Such a leak would be the more certainly detected, since the pump was always left at rest for fifteen minutes

after exhaustion and before the measurement of the vacuum, so as to make sure of an equilibrium of pressure. Once such a leak was detected; when the apparatus was dismounted, and the globes weighed again. A leakage during the second weighing ought to be detected by the slow change of weight which it produces. It was not detected during any of the experiments of the present series. It may be said that the stopcocks of five of my globes were of large size so as to make the distance which air must travel around the key, from the entrance to the exit, as much as four centimetres. These were often found to be tight for some days when freshly lubricated and turned but a few times.

It was to avoid both these errors that, in the last three experiments, the globes holding oxygen were closed in the same way as the tube containing the palladium

Two sources of error in the determination of the weight of water produced were known to exist. The platinum wires for the passage of the spark might become loosened by the heat, so as to admit some water or air. On the other hand, the exhaustion by means of a Toepler pump, unless regulated by opening the stopcock but partly, may sometimes draw the gas through the drying tubes faster than it can be dried. Probably this happened, to a slight extent, in some of the experi ments, but the amount was not very large. The manipulation which might have easily and completely prevented this was not thought of till the second (unfinished) series of experiments was begun.

13.--SYNTHESIS OF WATER.

COMPLETENESS OF DRYING OF GASES WITH PHOSPHORUS

PENTOXIDE.

In a paper describing some of the earliest of my experiments relating to oxygen and hydrogen, it is shown that the amount of moisture left by phosphorus pentoxide in even a thousand litres of a gas is negligible. My experiments proved that a current at the rate of three litres an hour was made absolutely dry by a rightly filled tube whose capacity was twenty-five cubic centimetres; it was therefore thought safe to pass a current of twelve litres an hour through a tube containing one hundred cubic centimetres.

14. SYNTHESIS OF WATER. IS HYDROGEN GIVEN OFF BY PALLADIUM FREE FROM WATER?

A source of error would exist if, as has been thought, water should accompany the hydrogen delivered from palladium. This would imply that oxygen in some way found access to the palladium, and that the water produced was not removed with any nitrogen which might have accompanied the hydrogen. Keiser considered his experiments as proving that water is given off from palladium together with

hydrogen. I made five experiments, which show that a tube containing palladium aud filled with hydrogen as in all my use of it, gives off no water. The tube containing six hundred grammes of palladium was charged in the usual way, and then connected with a U tube containing phosphorus pentoxide, which had been carefully weighed against a tube similar in shape and equal in volume. The hydrogen was expelled through the drying tube at the rate of not more than six litres an hour, it being collected over water to control the rate. This was repeated five times. The drying tube showed the usual slight variations of weight from day to day, but the weight did not increase. When two hundred litres had been passed through it, the weight was the same as at first, and the algebraic sum of all the slight variations during the course of the experiment was zero. It is obvious therefore that if palladium is charged with hydrogen as in my experiments, it will give off no water.

15. SYNTHESIS OF WATER. IS OXYGEN ABSORBED BY THE PHOSPHORUS

PENTOXIDE USED?

A source of error in the present series of experiments would exist, if the drying tubes used were capable of absorbing oxygen in their ordinary use. If this were the case, the drying tubes b, b, Fig. 36, would absorb oxygen, which would seem to have been combined with hydrogen. But all the pentoxide used in these experiments had been examined to determine whether it were suitable for the purpose.

d

b

FIG. 39.-Apparatus for detecting absorption of oxygen by phosphorus pentoxide; filling with oxygen.

A drying tube filled with the pentoxide, a b, Fig. 39, was fused to the short arm of a syphon barometer lying on its side as seen. Pure dry oxygen was passed into c, escaping at d, and these two tubes were closed by fusion while the current was passing. The barometer was then set upright, and securely fixed to a wall, so as to remain undisturbed for months. A tank of water was put SO as to cover the drying tube, giving the means of knowing the temperature of the gas. A thermometer divided into fiftieths of a degree was placed in the water, and two other thermometers gave the temperature of the column of mercury. The apparatus was set up in a basement room of tolerably constant temperature. As soon as it was set up, readings were taken till equilibrium was attained, when a split tube was fixed with its lower edge tangent to the upper meniscus of the mercury, at e, Fig. 40. The reading having been verified several times, by warming or cooling the water till the mercury was again tangent to the mark, the pressure of the enclosed oxygen was known.