In this "primitive atmosphere" of nitrogen, containing some carbonic acid, and vapour of water, vegetation is found to be tolerably prosperous, in spite of the confined condition of the air: the carbonic acid is absorbed and replaced by free oxygen, so that after a certain lapse of time the gas in the graduated tube approaches the composition of ordinary atmospheric air, and can even be made richer in oxygen than the latter.

First 75 cubic centimetres of pure nitrogen (reduced to 0° C. and 30 inches barometer) are introduced, and the little plant being put in, makes the whole 102 centimetres.

Then a small amount of carbonic acid is let in, and the volume of gas oscillates during the experiment from 102 to 130 c.c., according to the temperature and barometric pressure at the time of observation.

The little plant of Convolvulus arvensis was introduced on the 25th July 1893, its height being then 30 divisions of the graduated tube.

On the 26th July it had grown to 37 divisions.

On the 31st July it had formed a new leaf and was curving; it now occupied 52 divisions of the tube.

On the 1st August it had curved considerably, as all plants of the genus Convolvulus do, and measured only 50 divisions in height; but on the 2nd August it had shot up again to 64 divisions.

It appeared very healthy.

On the 3rd August it attained to 68 divisions.

On the 5th August there were new leaves formed, and the plant measured now 70 divisions in height.

During the 6th, 7th, 8th and 9th of August, the plant was healthy and two more new leaves had formed.

The water being well supplied with carbonic acid, and a little of this gas again introduced into the graduated tube, I left the experiment (in order to make a journey) until the 18th September.

On the 18th September my Convolvulus had grown to 90 divisions; and by the 30th of that month to 94; nearly to the top of the tube.

On the 2nd October, it began to turn yellow, as did also the two plants growing in water outside the apparatus as "witnesses"; they all three put on their autumnal tints at the same time, and had ceased to vegetate by 30th October.

The gas in the graduated tube (reduced to 0° C. and 30 inches barometer) then measured 95 cubic centimetres. It was analysed on the 30th October, and gave exactly the following result:—

Thus in the course of three months and seven days, or ninety-eight days, the little plant had grown from 30 to 94 divisions, not counting the curve, natural to the Convolvulus, and had converted all the carbonic acid into oxygen, leaving the nitrogen exactly as it was at the commencement of the experiment.

At the end of these fourteen weeks the atmosphere of the graduated tube was richer in oxygen than ordinary atmospheric air.

II.-ADDITIONAL NOTES.

1. RUMBLING THUNDER.-There is a curious passage in Galvani's paper, De Viribus Electricitatis, etc., 1791 (Bologna), on the use of a frog's leg, properly prepared, as an electrometer, or electroscope.

Dr. Galvani wished to try whether the electricity of the atmosphere, or the clouds, would produce the same effect on the prepared limbs of a frog as is produced by the artificial electricity of ordinary machines. He had already noticed, independently of the convulsions caused by the direct contact of the limbs with the conductor of an electric machine, that whenever a spark was taken from a large conductor situated at some distance from the prepared limbs, the legs moved with a kind of spasmodic contraction, sometimes strong enough to cause them to jump a little.

In order to ascertain whether atmospheric electricity would act in the same manner, he placed a conductor at the top of his house in Bologna, and put this conductor in connection with the prepared frog placed upon a table in the open air, or enclosed under a glass bell, and he soon observed that lightning produced the same effect as the sparks of an electric machine at a little distance. The same contractions of the limbs occurred, and they were stronger, or weaker, according to the distance and quantity of the lightning.

Thus far the effects might, naturally, have been expected;