ing under the excitement of the induction discharge in vacuo, I have found very great differences in the duration of the residual glow. Some earths continue to phosphoresce for an hour or more after the current is turned off, while others cease to give out the light the moment the current stops. Having succeeded in splitting up yttria into several simpler forms of matter differing in basic power (Roy. Soc. Proc. vol. xl. pp. 502-509, June 10, 1886), and always seeking for further evidence of the separate identity of these bodies, I noticed occasionally that the residual glow was of a somewhat different colour to that it exhibited while the current was passing, and also that the spectrum of this residual glow seemed to show, as far as the faint light enabled me to make out, that some of the lines were missing. This pointed to another difference between the yttrium components, and with a view to examine the question more closely I devised an instrument similar to Becquerel's phosphoroscope, but acting electrically instead of by means of direct light.

The instrument, shown in Fig. I, A and B, consists of an opaque disk, a bc, 20 inches in diameter, and pierced with twelve openings near the edge as shown. By means of a multiplying wheel, d, and band, ef, the disk can be set in rapid rotation. At each revolution a stationary object behind one of the A apertures is alternately exposed and hidden twelve times. commutator, g (shown enlarged at Fig. 1, B), forms part of the axis of the disk. The commutator is formed of a hollow cylinder of brass round a solid wooden cylinder. The brass is cut into two halves by a saw cut running diagonally to and fro round it, so as to form on each half of the cylinder twelve deeply cut teeth interlocking, and insulated from those on the opposing half cylinder by an air space about 2 mm. across. Only one half, hhh, of the cylinder is used, the other, i ii, being idle; it might have been cut away altogether were it not for some little use that it is in saving the rubbing-spring, j, from too great friction when passing rapidly over the serrated edge. To a block beneath the commutator are attached two springs, one, k, rubbing permanently against the continuous base of the serrated hemicylinder hh, and the other, j, rubbing over the points of the teeth of h h. By co necting these springs with the wires from a battery it will be seen that rotation of the commutator produces alternate makes and breaks in the current. The spring, j, rubbing against the teeth is made with a little adjustment sideways, so that it can be said to touch the points of the teeth only, when the breaks will be much longer than the makes, or it can be set to rub near the base of the teeth, when the current will remain on for a much longer time and the intervals of no current will be very short. By means of a screw, I, attached to the spring, any desired ratio between the makes and the breaks can be obtained. The intermittent primary current is then carried to an induction coil, m, the secondary current from which passes through the vacuum tube, n, containWhen the commutator, the ing the earth under examination. coil-break, and the position of the vacuum tube are in proper adjustment, no light is seen when looked at from the front if the wheel is turned slowly (supposing a substance like yttria is being examined), as the current does not begin till the tube is obscured by an intercepting segment, and it ends before the earth comes into view. When, however, the wheel is turned more quickly, the residual phosphorescence lasts long enough to bridge over the brief interval of time elapsing between the cessation of the spark and the entry of the earth into the field of view, and the yttria is seen to glow with a faint light, which becomes brighter as the speed of the wheel increases.

To count the revolutions, a projecting stud, o, is fastened to the rotating axis, and a piece of quill, p, is attached to the fixed With support, so that at every revolution a click is produced. a chronograph watch it is easy in this way to tell the time, to the tenth of a second, occupied in ten revolutions of the wheel. Under ordinary circumstances it is almost impossible to detect any phosphorescence in an earth until the vacuum is so high that the line spectrum of the residual gas begins to get faint; otherwise the feeble glow of the phosphorescence is drowned by the greater brightness of the glowing gas. In this phosphoroscope, however, the light of glowing gas does not last an appreciable time, whilst that from the phosphorescent earth endures long enough for it to be caught in the instrument. By this means, therefore, I have been able to see the phosphorescence of yttria, for example, when the barometer gauge was 5 or 6 mm. below the barometer.

When the earth under examination in the phosphoroscope is yttria free from samaria, and the residual emitted light is ex

At o'000875 sec. interval the highest speed the instrument

could be revolved with accuracy, the whole of the lines usually seen in the yttria spectrum could be seen of nearly their usual brightness.

I have already recorded (Phil. Trans., 1883, Part III. pp. 914-16), that phosphate of yttria, when phosphoresced in vacuo, gives the green lines very strongly whilst the citron band is hazy and faint. The same tube of yttric phosphate was now examined in the phosphoroscope. The green lines of GB soon showed themselves on setting the wheel into rapid rotation, but I was unable to detect the citron band of G8 even at a very high speed.

427

The action of barium on yttrium was now tried. The following mixtures (as ignited sulphates) were made :

Yttrium Barium

About I per

50

40

60

It

The effect of calcium on the phosphorescence of yttria and samaria has been frequently referred to in my previous papers. It may save time if I summarise the results here cent. of lime added to a badly phosphorescing body containing yttrium or samarium always causes it to phosphoresce well. diminishes the sharpness of the citron line of Gd but increases in brightness. It also renders the deep blue line of Ga extremely bright. The green lines of GB are diminished in brightness. Lime also brings out the phosphorescence of samarium, although by itself, or in the presence of a small quantity of yttrium, samarium scarcely phosphoresces at all.

In the phosphoroscope the action of lime on yttrium is seen to entirely alter the order of visibility of the constituents of yttrium. In a mixture of equal parts yttrium and calcium, the citron Gồ line is the first to be seen, then comes the Ga blue line, then the Go green line, and finally the Gn red line. This may, I think, be explained somewhat as follows:-Calcium sulphate has a long residual phosphorescence, whilst yttrium sulphate has a comparatively short residual phosphorescence. Now with yttrium, although the green phosphorescence of GB lasts longest, it does not last nearly so long as that of calcium sulphate. The long residual vibrations of the calcium compound induce, in a mixture of calcium and yttrium, phosphorescence in those yttric molecules (G) whose vibrations it can assist, in advance of those (GB) to which it is antagonistic; the line of Gd therefore appears earlier in the phosphoroscope than that of GB, although were calcium not present the line of GB would appear first.

Experiments were now tried with different mixtures of yttria and lime as ignited sulphates, to see where the special influence of lime on Gỗ ceased.

In the phosphoroscope the GB line appears earliest, but the blue Ga line is the next to be seen, whilst the red line of Gn is the latest in appearing. As the percentage of yttrium increases the blue line more and more overtakes the red and increases in brightness.

Spectrum similar to the above. As the percentage of yttrium increases the spectrum grows brighter. In the phosphoroscope the earliest line to appear is the GB green, then the Gn red, and next closely following it the Ga blue. In the radiant-matter tube all these mixtures give similar spectra. The GB green is a little brighter and the Go citron is a little fainter than in the corresponding mixtures of yttrium and calcium, but the whole of the yttrium lines are seen. In the phosphoroscope the GB green is the first to appear, then the Gn red. The Go citron

is not visible at any speed. Red line of Gn is much brighter; G is very faint, and the green of GB is stronger. In the phosphoroscope the order of appearance is, first the line of GB, then the red line of Gn.

Phosphoresces with difficulty, of a light blue colour, but turns brick-red in the focus of the pole. Spectrum very faint. Order of appearance to phosphoroscope, -GB first, the others too faint to be seen..

The next experiments were tried with strontium, to see what modification the addition of this body to yttrium would produce. The following mixtures of ignited sulphates were experimented with :

Yttrium Strontium

Order of appearance in the phosphoroscope. -GB, Ga, Gd, and Gŋ. The citron line of Go is only to be seen at a high speed, and is then very faint. Order of appearance in the phosphoroscope. -Ga, GB, and Go (citron and blue) together, and lastly Gn (red). At a very high speed the green lines of GB beconie far more luminous than any other line. Order of appearance.-Go and Ga together, then GB, and lastly Gn.

Order of appearance.-Go and Ga simultaneously, then GB, and lastly Gŋ. The residual phosphorescence lasts for 30 seconds after the current stops. The light of this residual glow is entirely that of Gd. The line of GB comes into view at an interval of 0.0045 second. At 0'00175 second the line of Gŋ is just visible.

90

Order of appearance.-Gô, Ga, GB.

95

99

0'5

Order of appearance. -G8, Ga. The green lines of GB could not be seen in the phosphoroscope; they would probably be obliterated by the stronger green of the continuous spectrum given by the calcium.

99'5

In the

A very good yttrium spectrum. phosphoroscope the order of appearance is, first the green of GB, then the Ga blue, lastly the Gn red. No Go citron line could be seen.

In the phosphoroscope the green of GB is very prominent at a low speed, standing out sharply against a black background. With a higher velocity the Ga and Gn lines come into view.

The ordinary spectrum of this and the neighbouring mixtures is very rich in the citron line of G8, but I entirely fail to see a trace of this line in the phosphoroscope at any speed. The line of GB is the first to come, then the blue line of Ga. At about this point a change comes over the appearance in the phosphoroscope. The blue line of Ga is now the earliest to appear, and it is followed by the Gŋ red and GB green. No Go line is seen. These mixtures are very similar to each other in the phosphoroscope. The line of Ga comes first, next the Gn line, then GB line. No G8 citron line has been seen in any of these mixtures.

In a paper read before the Royal Society, June 18, 1885 (Phil. Trans., 1885, Part II., p. 716), I described the phos

OF EARTHQUAKES

IN the course of a lecture delivered recently before the Rigaku Kyôkai, or Science Society of Tokio, on the causes of earthquakes, F:cf. Milne classified the theories as to the cause of these phenomena into three kinds-unscientific, quasi-scientific, and scientific. In the former class he included the explanations of the Negro preachers at Charleston after the late earthquakes there, that they occurred in consequence of the wickedness of the population. The Mussulmans in Java recently prayed to the volcanoes there to cease their shakings, at the same time promising reformation of life. That earthquakes are the direct result of man's wickedness is an idea that has always been About 1750 earthquakes were felt in many parts of Europe, which were widely attributed to this cause, and innumerable sermons were preached inculcating the lesson that if mankind would live better lives there would be no more earthquakes. In 1786, after a shock at Palermo, the people are recorded to have gone about scourging themselves, and looking extremely humble and penitent. An English poem called "The Earthquake," published in 1750, alleged, in somewhat halting

common.

verse, that the disturbances were not due to an unknown force, nor to the groanings of the imprisoned vapours, nor yet to the shaking of the shores with fabled Tridents:

"Ah no! the tread of impious feet

The conscious earth impatient bears
And shuddering with the guilty weight,

One common grave for her bad race prepares." From this theory, which can scarcely have satisfied the poet himself, Prof. Milne passed on to the myths which attribute earthquakes to a creature living underground. In Japan it is an "earthquake-insect" covered with scales, and having eight legs, or a great fish having a certain rock on his head which helped to keep him quiet. In Mongolia the animal was said to be a frog, in

ing hog, in North America a tortoise. In Siberia there was a myth, connected with the great bones found there, that these were the remains of animals that lived underground, the trampling of which made the ground shake. In Kamchatka the legend was connected with a god, Tuil, who went out hunting with his dogs. When these latter stopped to scratch themselves, their movements produced earthquakes. In Scandinavian mythology, Loki, having killed his brother Baldwin, was bound to a rock face upwards, so that the poison of a serpent should drop on his face. Loki's wife, however, intercepted the poison in a vessel, and it was only when she had to go away to empty the dish that a few drops reached him and caused him to writhe and shake the earth. The lecturer had no means of collecting the fables of the southern hemisphere; but they would obviously be worth knowing for purposes of comparison. As to quasiscientific theories, these endeavoured to account for earthquakes as parts of the ordinary operations of Nature. was supposed, for instance, that they were produced by the action of wind confined inside the earth. The Chinese philosophers said that Yang, the male element, entered the earth and caused it to expand, and to shake the ground in its efforts to

It

escape. Its effects would be more violent beneath the mountains than in the plains, and therefore earthquakes in the north of China, which was mountainous, were said to be more violent than those in the south. It was supposed that when the wind was blowing strongly on the surface of the earth, there was calm beneath, and vice versa. Aristotle and many other classical writers attri buted earthquakes to wind in the earth. Shakespeare, in "Henry IV.," speaks of the teeming earth being pinched and vexed with a kind of colic by the imprisoning of unruly wind within her womb. Then came the theory of electrical discharges, which was advocated in 1760 by Dr. Stakely, as well as by Percival and Priestley. They are strongly held in California at the present day, where it was believed that the network of rails

protected the State against any dangerous accumulation of electricity. But Prof. Milne showed that the laying down of rails in Japan had no such effect. He thought the electric phenomena which sometimes attended earthquakes were their consequences, not their causes. He had himself experimented with dynamite placed in a hole; an earth-plate was fixed about thirty yards away from the dynamite, and from it a wire was carried some distance to another earth-plate. When the dynamite charge was exploded there was certainly a current produced, as was indicated by a strong deflection of a galvanometer-needle at the end of the wire. He attributed this to chemical action. When the ground was shaken there was always a greater or less action by increase or decrease of pressure in connection with the earthplate. Earth currents unquestionably accompany earthquakes, but, as has been said, they appear to be the consequences, not the causes, of the latter. Next came the chemical theories, which were very strong in Europe up to the beginning of the present century. It was imagined that underground there were various substances, such as sulphur, nitre, vitriol, which, by their action on each other, resulted in violent changes, giving rise to vapour, the sudden production of which, in certain cases, would shake the ground. It was only in 1760 that Dr. Mitchell, who wrote a good deal on the subject, first threw out the theory that earthquakes were connected in some way with volcanoes, because they were most frequent in volcanic countries. observed that large quantities of steam were given off from volcanoes, and came to the conclusion that an earthquake was produced at the time that an attempt was made to form a volcano, that steam got in between certain strata, and, as it ran between them, caused pulsations. Prof. Rogers, about the same time, in North America, endeavoured to show that it was not steam, but really lava, that ran along underneath the ground, causing it to rise and fall, thus producing an earthquake. Prof. Milne having thus dealt with unscientific and quasi-scientific theories, passed on to those of modern science. It is unnecessary here to follow him into this portion of his subject, although it occupied the main part of the lecture.

He

ON THE EFFECT OF CERTAIN STIMULI ON VEGETABLE TISSUES1

THE object of our paper is to describe the behaviour of

turgescent pith when placed in water and treated with certain reagents. If from a growing shoot the external tissues be removed, a well-known result is seen: the pith suddenly lengthens, becoming longer than the specimen was at first. This experiment shows that turgescent pith is normally in a compressed condition-it is always trying to get longer-and when it is freed from the coercion of the unyielding external tissues, it at once does become longer. This tendency to become longer is further manifested by allowing turgescent pith to remain in damp air, or in water, for some time, when a great increase in length takes place. In such a piece of pith we have the essential, active factor in growth, freed from interference, and at liberty to perform its function rapidly and freely. The tendency in turgescent pith to get longer is the very power which calls forth that increase in length which we call growth; so that in studying turgescent pith we are studying the active agent in the production of growth. We do not suppose that our results are necessarily directly applicable to normal growth, but we think that they have a bearing on normal growth sufficiently close to give interest to our experiments.

3

The pith, after being cut into pieces about 6 inches in length and inch in thickness, was ready for use. The lower end of the pith was fixed to a hook at the bottom of a narrow jar, the upper end was attached by a silk thread to the short arm of an auxanometer lever. The jar was then filled with water, and as the pith elongates the short arm of the lever ascends and the long arm rapidly descends. Its movement, read off on a millimetre scale, gives an index of the rate of "growth" of the pith. The lengthening of the pith is, in fact, observed like the normal growth of a plant, the only difference being that the "growth of the pith is so rapid that the descent of the long arm is clearly visible to the naked eye and is correspondingly easy to measure. It is most striking to see the index travelling down thus quickly Abstract of a Paper by Anna Bateson (Newnham College) and Francis Darwin (Cambridge), read before the Linnean Society, January 20, 1887. 2 For the sake of convenience we shall nevertheless use the word "growth" to mean the elongation of the pith under observation. 3 Sunflower and Jerusalem Artichoke.

and traversing (it may be) 10 mm. (3 inch) in a minute. We used a stop-watch to determine the time in which the point of the long arm of the lever travelled over a certain distance, and we could thus estimate the changes in the rate of growth from minute to minute.

The first thing needful to know is the ordinary course of growth of the pith in water. It was found that an interesting phenomenon-an apparent grand period-takes place. That is to say, the growth is at first slow, then more rapid, and ultimately becomes slow again, the whole period taking perhaps twenty minutes to complete. This is precisely the series of changes which a growing organ exhibits in the course of days instead of minutes. We do not suppose that our grand period is necessarily of a kindred nature to the grand period of normal growth. we are aware that purely mechanical processes, such as the moistening of a hygroscopic awn, exhibit the same thingthe awn at first untwists slowly, then more quickly, and then again slowly. But the knowledge of the fact is of great importance to us, since unless we know the normal course of growth we cannot study the effect of reagents.

For

Warmth.-Before going on to consider the action of reagents, we will say a few words as to the stimulation caused by an increase in the surrounding temperature. If the water in the jar is gradually warmed, the growth of the pith increases in speed in the most striking manner. The increase is fairly steady from, say, 17° C. to about 35°, the rate at this latter temperature being perhaps four times as high as it was at first. It then usually becomes irregular, with some diminution; and, just before a temperature is reached which kills the tissues, a sudden and rapid fall in the rate of growth sets in. This we found usually to occur at about 55° C. This is, no doubt, an unusually high temperature, but not higher than plants are known to be able to survive.

The chief interest in these temperature experiments is this: they show that the phenomena we are considering is a truly vital one. We have always been on our guard in this matter, and have wished to make certain that the observed phenomena are not in some mysterious way mechanical, instead of, as we believe, the response of living tissues as living tissues. Therefore, when we find that heat has a normal effect on our material, we are encouraged to believe that our other results-to which we now pass on-are also vital phenomena.

Alcohol.-The pith was attached to the auxanometer, and the jar filled with water. As soon as the rate of growth was found to be steadily diminishing, a small quantity of alcohol was added. The result was an immediate and striking increase in the rate of growth. For instance, when 2 per cent. of spirit was added, the growth was accelerated within two minutes by 50 per cent. The result is temporary, so that in the course of another two minutes the rate of growth sinks to what it was before stimulation. Similar results were obtained with ether, and here the pith was allowed to grow in damp air, and was subjected to ether in the form of vapour. When the vapour was present in the proportion of o 27 per cent., the acceleration was 56 per cent. ; with 04 per cent., the acceleration was 100 per cent. Here, as

in the case of alcohol, the result was temporary, the rate falling in a few minutes to what it was before stimulation.

When the ether amounts to 3 per cent. of the atmosphere, the pith is killed, and shows no increase, but, on the contrary, a decrease in length. Elfving has shown that ether has a stimulating effect on respiration, and on the sensitiveness of swarm-spores to light. He also tested its effect on the growth of phycomyces. His results differ from ours, inasmuch as he found no stimulating effect: the ether produced either no effect whatever, or else it retarded, or even stopped, growth.

66

Ammonia.-We employed the Liquor Ammonia fortior of the British Pharmacopoeia" for the preparation of our solutions, and we found that various strengths ranging between 0.5 and 24 per cent. produced acceleration of growth. Here again, as with ether and alcohol, the acceleration was very temporary.

Acids. As a rule, acids produced no acceleration, but caused either retardation, or flaccidity and death. Thus, for instance, acetic acid (0'5 and 1 per cent.) produced retardation; 5'4 per cent. produced death.

Hydrocyanic Acid did not cause flaccidity such have described in the case of acetic acid. The action of this reagent is comparable rather to that of alcohol, but is not

1 This contraction is simply a symptom of flaccidity, and usually of death.