Page images
PDF
EPUB

minute in a normal solution of ester, and the corresponding dissociation 0-12 x 10" of a gramme-molecule per litre.

Fig. 6.

3

50 100 150 200 250 minutes

The great importance of this result is seen on comparison with the following numbers. Shortly before, Ostwald found in quite a different way that the electrolytic dissociation of water lay between 0.74 × 10-6 and 0.27 × 10-6 gram.-mol. per litre; whilst from Kohlrausch's conductivity determination he calculated that the maximum value did not exceed 0.6 × 10−6. Arrhenius obtained 0.11 x 10-6 from Shields' experiments, and Bredig estimated it at about 0.6 × 10-6 gram.-mol. per litre.

Is the theory of electrolytic dissociation hereby proved? By no means. In the domain of saponification it has nevertheless accomplished far more than could have been expected.

XLVIII. Note on the Elasticity of Spider Lines.

By JAMES H. GRAY, M.A. (Glas.), B.Sc. (Lond.)*.

THE

HESE tests were made last summer in the Physical Laboratory of Glasgow University, in the course of some work in which a very sensitive mirror-galvanometer was required.

When the mirror was suspended with a fine silk fibre, it was found that, when deflected, it took an inconveniently long time to return to zero. This defect, common to silk and most metal suspensions, and which has been called creep," was, at the suggestion of Dr. J. T. Bottomley, F.R.S., obviated by using a spider's thread for supporting the mirror.

66

*Communicated by the Author.

In the Proceedings of the Royal Society for 1889, page 291*, an account is given of some very interesting tests made of the value of the torsional rigidity of spider's thread by Professor A. Tanakadate, of Japan. He found that the torsional rigidity is rather less than one sixth of that of silk fibre of the same thickness.

For the tests about to be described, several garden spiders, species Epeira Diadema, about one eighth of an inch in length of body, were caught and placed each in a separate box. Fresh threads, free from dust, could thus be obtained. If more than one spider was placed in the same box the number soon became reduced to one, the others having been killed and eaten by the survivor. It is necessary to employ the garden spider, as the house species, which makes a closely woven web quite different from the beautiful spiral of the Diadema, cannot be made to spin a thread. The spinner was taken out of its box and placed on a card or piece of wood held about three feet or so above the ground. A slight tap was sufficient to cause the spider to drop from the support, spinning a thread as it fell. Sometimes it dropped to the floor, but in most cases it stopped when it had fallen six or eight inches, and, after hanging apparently motionless for a second or two, rapidly threw out an exceedingly light thread which floated outwards and upwards in the direction of the slightest draught. In the space of ten seconds sometimes as many as ten feet of thread were thrown out. If the outer end of this thread happened to attach itself to one of the adjacent supports, the spider immediately endeavoured to make its escape. The threads thus spun were exceedingly thin, and could only be seen when a strong light was thrown upon them and with a black background.

While it was still floating, two marks were made about 50 centimetres apart, by carefully fixing with gum two pieces of the thread, each 3 millimetres long, at right angles to the length. Small wire weights from to 30 milligrammes were made, and half a milligramme was fixed a little beyond the furthest thread-mark. This was sufficient to bring the thread down to the vertical without appreciably stretching it. On a vertical stand were fixed two paper scales graduated in halfmillimetre divisions, the one scale about 50 centimetres above the other. Directly in front of these scales the thread was hung, and by means of a telescope placed six feet away the readings corresponding to the two marks on the thread were obtained. Precautions were taken to prevent draughts, as

"Note on the Thermoelectric Position of Platinoid," by J. T. Bottomley, M.A., F.R.S., and A. Tanakadate.

the slightest current was sufficient to drive the thread against the scales, from which it could only with difficulty be freed.

The small weights were gradually fixed to the lower end of the thread, and the corresponding readings of the marks taken. The data were thus obtained for a stress-strain diagram, and are shown on the annexed curve. The abscissæ represent percentage stretchings, and the ordinates weights applied. The actual stretchings obtained are given below:

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][ocr errors][merged small][merged small][merged small][ocr errors][ocr errors][merged small][merged small][merged small][ocr errors][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][ocr errors]

Original length between the marks, 54-15 centimetres. One milligramme corresponds to 127.3 kilogrammes per square centimetre.

The breaking-weight of this thread was found to be 17 milligrammes, that is, 2:16 × 106 grammes per square centimetre. This number does not differ greatly from those for the breaking-weights of copper (annealed), bronze, drawn gold, palladium, and silver. It is rather greater than that for cast iron, but considerably less than steel, and indeed than silk fibre.

The value for the breaking-weight per unit of cross section of the spider's thread is given on the assumption that the thread is circular. This is not quite correct, however, as it is really composed of four or six strands parallel to each other. Each of these strands, again, is made up of about one thousand exceedingly fine threads. The cross section of the complete thread being, therefore, virtually four or six small circles instead of one large circle, the value given above for the breaking-weight is probably from five to ten per cent. too small.

The diameter of the thread was very carefully measured for me by Dr. William Snodgrass, of the Physiological Laboratory, by means of a very finely-divided micrometerscale and a powerful microscope. As measured by Dr. Snodgrass, the diameter was found to be almost exactly 1000 of a millimetre.

One interesting point is at once apparent on looking at the form of the curve. Whereas in all metal threads, whenever

the elastic limit is exceeded, the extension increases in a greater proportion than the tension, in spider's thread, on the contrary, the extension at first increases more slowly than the tension, and afterwards goes on at exactly the same rate up to the breaking-point. It will be seen that the latter part of the curve is practically a straight line.

[merged small][graphic][subsumed][subsumed][subsumed][subsumed][subsumed][subsumed]

Stretching-weight, in milligrammes. 1 milligramme corresponds to a stress of 127.3 kilogrammes per square centimetre.

[blocks in formation]

The behaviour of spiders' thread under stress is very similar to that of muscle and all animal tissue. They show a kind of tetanus tendency when stretched. It also resembles indiarubber, as shown in a stress-strain diagram obtained by Professor Archibald Barr in a testing-machine designed by him.

Subject to the error caused by want of circularity of the thread, the Young's modulus was found to be 7.769 x 106 grammes per square centimetre. This is much lower than the value for any metal, the lowest being that of lead, which

6

is 51 x 106, and even than any of the woods as given in volume iii. of Lord Kelvin's Collected Papers.' On account of the thread not being circular, the calculation of the Rigidity Modulus would of course be valueless.

A rough trial was made of the Torsional Rigidity; but as the thread was very much finer than that used by Professor Tanakadate, being about one third the diameter, it was difficult to obtain a light enough twisting-weight. The result obtained, however, agreed with that given by Prof. Tanakadate, in so far as it showed that the torsional rigidity of spiders' thread is considerably less than that of silk fibre.

XLIX. The Method of Fractional Distillation illustrated by the Investigation of Kerosene. By J. ALFRED WANKLYN and W. J. COOPER *.

A LIQUID known as kerosene, or Russian kerosene, is imported into this country in immense quantities. The liquid is produced by the distillation of crude natural Russian petroleum in Baku, and is conveyed in tank-ships to London. It arrives in this country in the condition of an almost colourless oil, with very little smell, and of sp. gr. 0.825 at 15°5 C. Apparently it is very constant in quality and composition.

The knowledge of it which we owe to former investigators is that it is a mixture of hydrocarbons of the general formula CAH2n; that these hydrocarbons are not olefines, but are isomers of the olefines, being not readily attacked by reagents. Very few of the members of the series have been described by former investigators. We have made a study of kerosene as follows:

When kerosene is rapidly distilled in a glass retort, 70 per cent. comes over below 250° C., about 20 per cent. between 250° and 300°, and the residual 10 per cent. may be almost completely volatilized; the last 5 per cent. requiring, however, a temperature much above the boiling-point of mercury.

By careful quantitative experiment, a carefully measured litre of kerosene being distilled in two operations, half a litre at a time, it was ascertained that there is neither expansion nor contraction during distillation. The observation was also made that not until the temperature had risen to 170° C. did the first drop of liquid distil over.

* Communicated by the Authors.

« PreviousContinue »