*-"' I PROCEE&INGS OF THE ROYAL SOCIETY OF LONDON. From Jtcac 11, fo December 13, 1900.' VOL. LXV1I. LONDON: HARBISON AND SONS, ST. MARTIN'S LANE, |}rmtrrs in ©rbinartj to ptr late ^tajrslg. FKHEUART, 1901. M\ LONDON HARBISON" AND SON'S, PRIVTRBS IX ORDINARY TO HER LATE MAJESTV, ST. MARTIN'S LANE. CONTENTS. VOL. LXVII. No. 435 Page Meeting of June 21, 1900, and List of Papers read 1 Investigations on Platinum Thermometry at Kew Observatory. By C. Chree, Sc.D., LL.D., F.E.S., Superintendent. Communicated by the Kew Observatory Committee 3 A Comparative Crystallographical Study of the Double Selenates of the Series E2M(SeO4)o,6H2O.— -Salts in which M is Zinc. By A. E. Tutton, B.Sc., F.E.S 58 No. 436. Certain Laws of Variation. I. The Reaction of "Developing Organisms to Environment. By H. M. Vernon, M.A., M.D., Fellow of Mag- dalen College, Oxford. Communicated by Professor E. Eay Lan- kester, F.E.S 85 On the Diffusion of Gold in Solid Lead at the Ordinary Temperature. By Sir W. Eoberts- Austen, K.C.B., F.E.S., Professor of Metallurgy, Eoyal College of Science 101 On Certain Properties of the Alloys of the Gold- Copper Series. By Professor Sir W. Eoberts-Austen, K.C.B., F.E.S., and T. Kirke Eose, D.Sc. (Plate 1) 105 The Crystalline Structure of Metals. Second Paper. By J. A. Ewing, F.E.S., Professor of Mechanism and Applied Mechanics in the University of Cambridge, and Walter Bosenhain, B.A., St. John's College, Cambridge, 1851 Exhibition Eesearch Scholar, Melbourne University 112 On the Estimation of the Luminosity of Coloured Surfaces used for Colour Discs. By Sir William de W. Abney, K.C.B., F.E.S 118 The Diffusion of Ions produced in Air by the Action of a Eadio-active Substance, Ultra-violet Light, and Point Discharges. By John S. Townsend, M.A., Clerk Maxwell Student, Cavenish Laboratory, Fellow of Trinity College, Cambridge. Communicated by Professor J. J. Thomson, F.E.S 122 Static Diffusion of Gases and Liquids in Eelation to the Assimilation of Carbon and Translocation in Plants. By Horace T. Brown, F.E.S., LL.D., and F. Escombe, B.Sc., F.L.S 124 The Electrical Effect* of Light upon Green l*-a\.-s. (Pi«-limiiiar\ < '1.111- munication.) By Augustus D. Waller, M.D., F.R.S " 120 On the Viscosity of Gases as affected by Temperature. By lx>rd Rav- leigh, F.R.S :... 137 Report of Magnet ical Observations at Falniouth Observatory for the Year 1897 " 13!) Rejwrt of Magnetical Obsei \ati..n> at Faluu.uth OWrvatory fi.r the Year 1898 ' 144 Report of Magnetical Observations at Fahuouth Observatory for the Year 1899 lf,3 No. 437 Data for the Problem of Evolution in Man. V. On the Correlation between Duration of Life and the Number of Offspring. By Mis> M . Beeton, G. U. Yule, and Karl Pearson, F.R.S., University College, London *.... 1">9 On the Effects of Changes of Temperature on the Elasticities and Internal Viscosity of Metal Wires. By Andrew Gray, LL.D., F.R.S., Professor of Natural Philosophy in the University of Glas- gow, and Vincent J. Blyth, M.A., and James S. Dunlop, M.A., B.Sc., Houldsworth Research Students in the University of Glasgow 180 On the Connection between the Electrical Properties and the Chemical Composition of Different Kinds of Glass. Part II. Bv Pn-t Andrew Gray, LL.D., F.R.S., and Professor James J. Dobl>i«-. M. A.. D.Sc .". 1!»7 On the Change of Resistance in Iron produced by Magnetisation. By Andrew Gray, LL.D., F.R.S., Professor of Natural Philosophy in the University of Glasgow, and Edward Taylor Jones, I >.S< ., Pro- fessor of Physics in the University College of North Wales „ 208 The Exact Histological Localisation of the Visual Area of the Human Cerebral Cortex. By Joseph Shaw Bolton, B.Sc., M.D., RS. (Loud.). Communicated by Dr. Mott, F.R.S 21G Underground Temperature at Oxford in the Year 1899, as determined by Five Platinum-resistance Thermometers. By Arthur A. Ram- l.unt, M.A., D.Sc., Radclitfe Observer. Communicated by E. H. (Griffiths, F.R.S 218 On the Kinetic Accumulation of Stress, illustrated by the Theory of Impulsive Torsion. By Karl Pearson, F.R.S., Professor of Applied Mechanics, Univeisity College, London 222 The Nature and Origin of the Poison of Lotus Arabian. Preliminary Notice. By Wyndham R. Dunstan, M.A., F.R.S., Sec. C.S., Director of the Scientific LYpaitment of the Imperial Institute, and T. A. Henry, B.Sc.. Lond., Walters' Company's Research Fellow 2i'4 No. 438. Page On the Spectroscopic Examination of Colour produced by Simultaneous Contrast. By George J. Burch, M.A., Reading College, Reading. Communicated by Francis Gotch, F.R.S., Professor of Physiology, University of Oxford „ 226 An Experimental Investigation into the Flow of Marble. By Frank D. Adams, M.Sc., Ph.D., Professor of Geology in McGill University, Montreal, and John T. Nicolson, D.Sc., M.Inst.C.E., Head of the Engineering Department, Municipal Technical School, Manchester. Communicated by Professor H. L. Callendar, F.R.S 228 Lines of Induction in a Magnetic Field. By H. S. Hele-Shaw, F.R.S. , and A. Hay, B.Sc '. 234 The Distribution of Molecular Energy. By J. H. Jeans, Scholar of Trinity College, and Isaac Newton Student in the University of Cambridge. Communicated by Professor J. J. Thomson, F.RS 236 On the Capacity for Heat of Water between the Freezing and Boiling Points, together with a Determination of the Mechanical Equivalent of Heat in Terms of the International Electrical Units. Experi- ments by the Continuous-flow Method of Calorimetry performed in the Macdonald Physical Laboratory of McGill University, Montreal. By Howard Turner Barnes, M.A.Sc., Joule Student. Communicated by Professor H. L. Callendar, F.RS 238 Energy of ROntgen and Becquerel Rays, and the Energy required to produce an Ion in Gases. By E. Rutherford, M.A., B.Sc., Mac- donald Professor of Physics, and R. K. McClung, B.A., Demonstra- tor in Physics, McGill University, Montreal. Communicated by • Professor J. J. Thomson, F.R.S. ..." 245 On Expressed Yeast-cell Plasma (Buclmer's " Zymase''). By Allan Macfadyen, M.D., G. Harris Mori-is, Ph.D., and Sydney Rowland, M.A. Communicated by Sir Henry E. Roscoe, F.R.S 250 On the Thermodynamical Properties of Gases and Vapours as deduced from a Modified Form of the Joule-Thomson Equation, with Special Reference to the Properties of Steam. By H. L. Callendar, M.A., LL.D., F.R.S., Quain Professor of Experimental Physics, University College, London » 266 Note on Inquiries as to the Escape of Gases from Atmospheres. By G. Johnstone Stouey, M.A., Hon. D.Sc., F.RS 286 South African Horse-sickness : its Pathology and Methods of Protec- tive Inoculation. By Alexander Edington, M.B., C.M., F.R.S.E., Director of the Colonial Bacteriological Institute, Cape Colony. Communicated by Sir David Gill, F.R.S 292 Note on the Occurrence of a Seed-like Fructification in certain Palaeo- zoic Lycopods. By D. H. Scott, M.A., Ph.D., F.RS., Honorary Keeper of the Jodrell Laboratory, Royal Gardens, Kew 306 No. 439. The Demarcation Current of Mammalian Nerve. (Preliminary Com- munication.) I. The Demarcation Current of Mammalian Nervt-. vi !'«(.• By J. S. Mardonald, I',. A.. 1. .!,'.( . I'. F... I ni\.'i>itv ('..llt^o, Li\ity College, Liverpool, Research Scholar of the British Medical As.->i>rinti<>n. Communicated by Profe.^i.r Sherrington, r.li.S , :!!:• The Demarcation Current of Manimaliau Nerve. (Preliminary Com- munication.) III. The Demarcation Source and " the Concentration Law." By J. S. Macdonald, B.A., L.R.C.P.E., University College, Liverpool, Research Scholar of the British Medical Association. Communicated by Professor Sherrington, F.R.S 325 Meeting of November 15, 1900, and Proceedings 328 List of Papers received during the Recess and List of Papers read 329 Argon and its Companions. By William Ramsay, F.R.S., and Moms W. Travers, D.Sc 329 Data for the Problem of Evolution in Man. VI.— A First Study of the Correlation of the Human Skull. By Alice Lee, D.Sc., with some assistance from Karl Pearson, F.R.S., University College, London 333 Total Eclipse of the Sun, May 28, 1900. Preliminary Account of the Observations made by the Solar Physics Observatory Eclipse Expedi- tion and the Officers and Men of fl.M.S. " Theseus," at Santa Pola. By Sir Norman Lockyer, K.C.B., F.R.S 337 Total Solar Eclipse of 1900 (May 28). Preliminary Report on the Observations made at Bouzareah (in the Grounds of the Algiers Observatory). By Professor H. H. Turner, M.A., F.R.S., and H. F. Newall, M.A., Sec. R.A.S 346 Solar Eclipse of May 28, 1900. Preliminary Report of the Expedition to the South Limit of Totality to obtain 'Photographs of the Flash Spectrum in High Solar Latitudes. By J. Evershed 370 Preliminary Note on Observations of the Total Solar Eclipse of 1900, May 28, made at Santa Pola (Casa del Pleito), Spain. By Ralph Cope-land, Ph.D., F.R.A.S., F.R.S.E 385 Total Eclipse of the Sun, 1900, May 28. Preliminary Account of the Observations made at Ovar, Portugal. By W. H. M. Christie, C.B., M.A., F.R.S., Astronomer Royal, and F. W. Dyson, M.A., Sec. R. A.S. (Plates 2-5) f. 392 No. 440. Meeting of November 22, 1900, and List of Papers read 402 Further Note on the Spectrum of Silicium. By Sir Norman Lockyer, K.C.B., F.R.S 403 On Solar Changes of Temperature and Variations in Rainfall in the Region surrounding the Indian Ocean. By Sir Norman Lockyer, K.C.B., F.RS., and W. J. S. Lockyer, M.A. (Carnb.), Ph.D. (Gott.).... 409 VI] Pa.sre On the Restoration of Co-ordinated Movements after Nerve Crossing, with Interchange of Function of the Cerebral Cortical Centres. By Robert Kennedy, M.A., D.Sc., M.D., Assistant Surgeon to the Western Infirmary, Glasgow. Communicated by Professor McKendrick, F.E.S \ 431 Anniversary Meeting and Meeting of December 6, 1900 436 The Histology of the Cell Wall, with Special Reference to the Mode of Connection of Cells. By Walter Gardiner, M.A., F.R.S., Fellow and Bursar of Clare College, Cambridge, and Arthur W. Hill, B.A., Scholar of King's College, Cambridge. Part I. — The Distribution and Character of " Connecting Threads " in the Tissues of Pinua Sylvestris and other Allied Species. By Arthur W. Hill, B.A., Scholar of King's College, Cambridge 437 On the "Blaze Currents" of the Frog's Eyeball. By A. D. Waller, F.R.S _ .' 439 On a Bacterial Disease of the Turnip (Brossica napu-s). By M. C. Potter, M.A., F.L.S., Professor of Botany in the University of Durham College of Science, Newcastle-upon-Tyne. Communicated by Sir M. Foster, Sec. R.S „ 442 The Micro-organism of Distemper in the Dog, and the Production of a Distemper Vaccine. By S. Monckton Copeman, M.A., M.D., F.R.C.P. Communicated by Sir M. Foster, Sec. R.S 459 On the Tempering of Iron hardened by Overstrain. By James Muir, B.Sc., B.A., Trinity College, Cambridge, 1851 Exhibition Research Scholar, Glasgow University. Communicated by Professor Ewing, F.R.S 461 No. 441. Meeting of December 13, 1900, and Proceedings 466 List of Papers read 467 On the Spectrum of the more Volatile Gases of Atmospheric Air, which are not Condensed at the Temperature of Liquid Hydrogen. Preliminary Notice. By S. D. Liyeing, M.A., D.Sc., F.R.S., Pro- fessor of Chemistry, University of Cambridge, and James Dewar, M.A., LL.D., F.R.S., Fullerian Professor of Chemistry, Royal Institution, London 467 Additional Notes on Boulders and other Rock Specimens from the Newlands Diamond Mines, Griqualand West. By T. G. Bonney, D.Sc., LL.D., F.R.S., Professor of Geology, University College, London __ 475 The Distribution of Vertebrate Animals in India, Ceylon, and Burma By W. T. Blanford, LL.D., F.R.S \ 484 On the Intimate Structure of Crystals. IV. Cubic Crystals with Octahedral Cleavage. By W. J. Sollas, D.Se., LL.D., F.R.S., Pro- fessor of Geology in the University of Oxford 493 Index 497 PROCEEDINGS OF THE ROYAL SOCIETY June 21, 1900. The LORD LISTEE, F.R.C.S., D.C.L., President, in the Chair. Mr. George James Burch, Dr. Leonard Hill, Mr. Joseph Jackson Lister, Professor Arthur A. Rambaut, Mr. William James Sell, Mr. Philip Watts, and Mr. Charles T. R. Wilson were admitted into the Society. The Right Hon. Sir Ford North, a member of Her Majesty's Most Honourable Privy Council, was balloted for and elected a Fellow of the Society. A List of the Presents received was laid on the table, and thanks ordered for them. The following Papers were read : — I. "On the Effect of Changes of Temperature on the Elasticities and Internal Viscosity of Metal Wires." By Professor A. GRAY, F.R.S., V. J. BLYTH, and J. S. DUNLOP. II. " On the Connection between the Electrical Properties and the Chemical Composition of Different Kinds of Glass. Part II." By Professor A. GRAY, F.R.S., and Professor J. J. DOBBIE. III. " On the Change of Resistance in Iron produced by Magnetisation." By Professor A. GRAY, F.R.S., and Professor E. T. JONES. IV. " Underground Temperature at Oxford in the Year 1899, as determined by Five Platinum Resistance Thermometers." By Dr. A. A. RAMBAUT. Communicated by E. H. GRIFFITHS, F.R.S. V. " On the Kinetic Accumulation of Stress, illustrated by the Theory of Impulsive Torsion." By Professor K. PEARSON, F.R.S. vor>. LXVII. B •J /. • •'/',• VI. '• Line- of Induction in a Magnetic Field." By Professor III I.I SHANV, F.U.S., and A. BAY. VII. ''On the Spectroscopic Examination of Colour produced by Simultaneous Contrast." By G. J. BUKCH, M.A. Commu- nicated by Professor GOTCH, F.R.S. VIII. "An Experimental Investigation into the Flow of Marble." By Dr. F. D. ADAMS and Dr. J. T. NICOLSOX. Communicated by Professor CAT.I.F.NDAI;, F.R.S. IX. "A Criticism of the Young-Helmholtz Theory of Colour Percep- tion." By Dr. F. \V. EDRIDGE-GREEX. Communicated by Sir LAUDER BRVXTOX, F.R.S. X . " On the Viscosity of Gases as affected by Temperature." By LORD RAYLKK;H, F.R.S. XI. " On the Thermodynamical Properties of Gases and Vapours as deduced from a Modified Form of the Joule-Thomson Equa- tion, with Special Reference to the Properties of Steam." By Professor H. L. CALLEXDAR, F.R.S. XII. " Note on Inquiries as to the Escape of Gases from Atmo- spheres." By Dr. G. JOHNSTOM: STOXEY, F.R.S. XIII. "The Distribution of Molecular Energy." By J. H. .1: B.A. Communicated by Professor J. J. THOMSON*, F.RS. XIV. " Energy of Rontgen and Becquerel Rays, and the Energy required to produce an Ion in Gases." By Professor E. RUTHERFORD and R. K. M\, F.R.S. XV. " On the Capacity for Heat of Water between the Freezing and Boiling Points, together with a Determination of the Me- chanical Equivalent of Heat in Terms of the International Electrical Units." By Dr. H. T. BARNES. Communicated by Professor CALLEXDAR, F.R.S. XVI. "On Expressed Yeast-cell Plasma (Buchner's Zijm 40-45. Constancy of platinum thermometers 42 46-49. Differences between the thermometers 4V> 50-52. Accuracy of the observations 53 53. Improvements suggested in apparatus 56 54. Further experiments wanted 57 55. Acknowledgment of assistance 57 Preliminary. § 1. In 1895 the Kew Observatory Committee decided, in the words of their annual Report, " to instal platinum thermometers at Kew, and to institute an independent series of experiments into their behaviour." The Report adds, "Attention will, in the first instance, be directed more especially to the question of the fixity of the zero and of the funda- mental interval." In pursuance of this policy the Committee built a special room, furnished with a fume closet, and purchased from the Cambridge Scientific Instrument Company six platinum thermometers and a Callendar-Griffiths resistance bridge, which was regarded at the time as embodying all the latest improvements. In the choice of apparatus arid the construction of the room, the Committee hud the advantage of the advice of Mr. E. H. Griffiths ; while Mr. C. T. Heycock, and Mr. F. H. Neville, as well as Mr. Griffiths, B 2 Dr. C. Chree. Im-ixt /.//>//.*, /.s- ,,/t kindly vMted the Observatory, and illustrated the methods of 'In- apparatus. An account of the original installation was given by Mr. (iriffiths in ' Nature,' Nov. 14, 1895, pp. 39-46. As this paper and several more recent papers by Professor Callendar and others have discussed the fundamental facts of platinum thermometry very fully, I have judged it unnecessary to go into such details here. Thermometers. § 2. The six original thermometers, distinguished as KI to K«, were all made from one sample of platinum wire. In 1896 a seventh thermometer, K-, was obtained from the Instrument Company. It is believed to be of the same sample of platinum wire as the others, but this is not absolutely certain. Particulars as to the type of the thermometers are given in the following table : — Table I. Approximate Ther- mometer. Material of tube. Length from end to edge of wooden collar. Length from end to terminals. Diameter of tube. resistance in ohms, answering to fundamental interval. cm. cm. mm. KI porcelain 33 0 37 '5 11-5 1-0 K., » 355 40-0 11-5 1-0 K, ,, 26-5 31-0 13-5 1-0 K. u 26-5 31-0 13-5 1-0 K, glass 35-5 41-5 10-5 10 K, M 35-0 41-0 J4-0 2-5 K; » 34-5 41-0 8-0 1-0 The resistance at 0° C. of K« is about 6-5 ohms, while the resistance of each of the other thermometers at 0° C. is about 2'6 ohms. Since it came to Kew, K0 has been exposed to no temperature above the steam point; whilst K3, K4, K5, and K7 have not been taken above the sulphur point (444°'53 C. according to Callendar and Griffiths, under normal pressure). In their early days, KI and Iv were heated on several occasions to the temperature of melting silver (approx. 960° C.). After a few months' use, Kj began to behave unsatisfactorily. The Instrument Company reported after inspection that the tube was slightly short, and a new tul»e was fitted in March, 1896. The career of K-J has l>een chequered. The first tul>e broke in melting silver in Platinum Thcrmonutry at Kew Observatory. 5 March, 1896 ; the second tube was found to be broken in August, 1896, after a silver point experiment. After repair the thermometer behaved badly, and had to be sent to Cambridge. A third tube cracked in molten silver in December, 1896. The fourth tube lasted until the end of 1898, when it was broken at Sevres. On more than one occasion the thermometer had practically to be remade, so that the observations taken with Ko at different stages of its existence are not comparable. The spiral at present in K? is believed, however, to be of the same sample of platinum as the original one. The other five thermometers are not known to have had any mis- adventures. The thermometers K2 and K5 were taken by Dr. Harker to Sevres in July, 1897, and did not return to Kew until the end of 1898. No use was made of K5 at Sevres, but some observations were made with Ko up to a temperature of about 600° C. § 3. In an ordinary platinum thermometer, it is possible for the air inside the tube to become unduly moist, with consequent deterioration of the insulation. In a glass tube the presence of moisture may be detected by the cloudy appearance when the thermometer is cooled in ice ; in the case of a porcelain tube the only guide is the behaviour of the galvanometer. If there is- no sensible creep in the galvanometer, it is probably best to leave the thermometer alone, even if slight cloudiness is visible. There is some risk of altering the apparent zero of the thermometer in removing it from the tube, and replacing it after the tube has been dried out. On the rare occasions when a tube has been dried out, check observations have been taken before and after the process. The room in which the platinum thermometers are kept is heated night and day when necessary by a gas stove, whose combustion products escape by a flue opening outside. As the room is naturally a dry one, the risk of moisture has thus been small. Resistance Box. § 4. The box as it originally existed in 1895 had a plug system similar to that of the ordinary Post Office pattern. The plug holders were of brass. Towards the end of 1896 it was found that pulling out a plug influenced the tightness of its neighbours. .Early in 1897 the Committee arranged with the authorities of the International Bureau of Weights and Measures that Dr. J. A. Harker should proceed to Sevres, and take part in a comparison of platinum and gas thermometers. At first it was proposed to take the existing Kew box to Sevres ; but, on hearing of the difficulties experienced with it, the Committee decided that Dr. Harker should examine into their reality before a decision was come to. Finding that with an ordinary standard of plug tightness sufficiently consistent results were not 6 IM. ( '. Chree. ///'•< *l«jations on attainable. Dr. Hanker tried the effect of greater tightness This, however, made matters rapidly worse, and it shortly became obvious that the box in its existing condition was of no fuither use; it was accordingly sent for repair to Cambridge. Having high hopes of a new system of plug holders — fusible metal inside Doulton ware — the Instrument Company introduced this, at their own expense. The plug holders in the restored box are supported tely, and the pulling out of a plug has never shown any ten- dency to influence others. This is undoubtedly a great improvement ; but I am somewhat doubtful of the expecb'ency of the other changes made when restoring the box. After some time it was found that taking out and replacing a plug sometimes exerted a very sensible influence on the reading, and from the appearance of the plugs it was suspected that this arose partly from the state of the plug holes. The fusible metal seems disposed to develop a coating of light-coloured powder, and presumably this affects the plug resistances. On being applied to, the Instrument Company supplied a simple arrangement for cleaning out the plug holes without undue friction, and it has certainly improved matters. The accompanying sketch, fig. 1, shows diagrammatically a vertical section of the original box, perpendicular to its longest dimension. C represents the coil chamber, AAA a copper tank containing water. This tank could be heated from below by a gas burner, the flow of gas being determined by a gas regulator, whose mercury bulb was sur- rounded by the water inside A. The coils hung in air, and their temperature was deduced by means of a mercury thermometer whose bulb was inside C. When altering the plug system, the Instrument Company altered the shape of the water tank and the form of the coil chamber. Fig. 2 gives a section of the existing form taken in the same way as fig. 1, A C A A FIG. 1. FIG. 2. the letters having the same significance. The changes were presumably dictated by the altered nature of the plug arrangements. I am dis- posed to think that the coil chamber in the existing form is not so well protected from external influences as in the earlier form, and I believe that in accepting the readings of the mercury thermometer, as giving the temperature of the coils, there is more risk of error now than formerly. Platinum Thennometry at Kew Observatory. 7 Resistance Coils. § 5. The coil resistances are on the binary scale, H = G/2 = F/4 = E/8 = D/16 = C/32 = B/64 = A/128 = Aa/256, with two extra coils, " Cal " and FI — or I, as it will be called here — the former used merely in calibrating the bridge wire. The nominal value of H is 5 box units. The unit is very approximately O01 of an ohm, so that the fundamental intervals of all the thermometers except KG are nearly 100 box units. This has the advantage that when the read- ing R0 in ice is subtracted from the observed reading, one has approxi- mately the temperature in degrees Centigrade on the platinum scale. The coil I is approximately 100 box units, and practically all the fundamental interval determinations, except with KG, have been referred to it. The coil Aa was added along with the coil Cal in 1897, when the box was altered. It has not been used in any of the work now to be described, and the mean box unit has been based throughout on the eight coils H to A. The coils are of platinum silver. The bridge wire is fully 30 cm. long, readings to the right of the centre representing a temperature above, and readings to the left a temperature below that answering to the sum of the coils whose plugs are out. The bridge wire is also of platinum silver, and possesses the same temperature coefficient as the coils. The bridge- wire scale is divided to 1 mm., and the vernier reads directly to 0'02 mm., and allows O'Ol mm. to be estimated. A differ- ence of 1 cm. on the bridge-wire reading answers very nearly to 1 box unit, or to a difference of 1° in temperature with all the thermometers except KG. Thus, in referring to differences of bridge-wire readings I shall usually, for brevity, speak of them as differences of temperature. Nature of the Observations. § 6. The expediency of using the same coil combinations for all the thermometers except KG was soon recognised. Also a point has been made of using two combinations for each temperature, according to the following scheme : — Kh By using the same coil combinations for different thermometers we may at least hope to detect any sensible relative changes, whilst Thermometers. --'? KS, K4, KS, KV, In ice. fCDF ICDGH In steam. fCDFI I CDGHI In sulphur vapour. fAE LAFGH KG TAG I AH fACDF LACDGH 8 Dr. ('. rhnv. 7/,/v.s/ /,/,//;„/,* an by using two coil combinations we get a better idea of the order of :uTur;icy of the observations ;nnl h;ive some check mi the constancy of the smaller coil resistances. As a rule, each fixed point determination depended on at least iiliservations Thus, in ice we might have a balance a^ain-' ('!>!•'. next against CDGH, and finally against CDF again. By com- paring the readings in the first and third observations one can make reasonably sure that the ^thermometer has reached the tme tempera- ture. This is a very necessary precaution, especially in sulphur point observations. When only three observations were taken, only half weight was allowed to the first and last, so as to give equal weight to the two combinations of coils. 1 1 was originally intended to take a complete set of " fixed point " observations, ice, steam, and sulphur points, with each thermometer once a month. The regular observations have been in reality less frequent than was intended, and the intervals between them more irregular. In addition to the regular oljservations there have l)een a variety of special occasions on which ice and steam point determina- tions have been made, more especially with K;. This thermometer has been used in comparing occasional high-range mercury thermo- meters. Sources of CJuinge or Error. $ 7. Before considering the main question, I propose discussing a number of possible sources of change in the readings obtained with platinum thermometers. An error of invariable amount from a source external to the thermometer is perhaps not immediately germane to this inquiry, but the invariability of error is a very difficult thing to ensure. Excluding observational error or defects, such as moisture in the tube, which an experienced observer is likely to detect, I would chronicle the following sources of trouble : — 1 . Change in the resistance of the platinum spiral ; 2. Change in the thick platinum wires connecting the spiral to the thermometer terminals, or in the compensating loop inside the tube; 3. Change in the relative resistance of the leads, or in the relative resistance of the proportional arms of the box ; 4. Change in the box coils or in the bridge wire ; 5. Faulty action of the contact piece ; 6. Shift of the bridge centre ; 7. Thermo-electric currents ; 8. Heating due to the battery current ; Platinum Thermometry at Kew Observatory. & 9. Error in the temperature coefficient of the coils, or differences between the temperature coefficients of different coils ; 10. Change of zero or other error in the mercury thermometer inside the coil chamber ; 11. Failure of the box mercury thermometer to give the true tem- perature of the coils and bridge wire ; 12. Insufficient immersion of the thermometer; 13. Slowness of platinum thermometers in acquiring the true tem- perature. There are other possible sources of error affecting only certain " fixed point " observations, and mostly not peculiar to platinum thermometry, for instance : — 14. Impurity in the ice used in getting E0, or variability in the method of treating it ; 15. Error in the barometer, whether constant or varying, or error in the reduction of its readings ; 16. Impurity in the sulphur used, and other uncertainties in the determinations of the boiling point of sulphur ; 17. Error in the formula assumed for the variation of the boiling point of sulphur with pressure. I have seen no adequate consideration of these sources of error in any treatment of platinum thermometry, and in the original equipment obtained for Kew Observatory no provision was made for the detection of most of them ; thus our experience may be useful to others. Without at least a 'general idea of their nature, the reader would be unable to judge correctly of the degree of probability attend- ing the conclusions reached, or even of the nature of the evidence on which these conclusions are based. § 8. I propose deferring until the end the discussion of the evidence bearing on the first two items ; only calling attention in the meantime to the fact that without taking a platinum thermometer to pieces it would be very difficult, if not impossible, to distinguish between changes in the resistance of the spiral and changes in the resistance of the thick platinum wires connecting the spiral to the terminals. The (inalterability of a platinum thermometer cannot be proved by experi- ments on an isolated platinum spiral. The connecting platinum wires inside the tube are by no means of negligible resistance. They ought to be very approximately equal in resistance to the compensator loops intended to neutralise their variation with temperature. The resist- ance of the compensator loops has been measured by Dr. Harker with the following approximate results. The data answer to a temperature of about 15° C. 10 l)i i ' ' i . . 'ions on TaMe II. — Resistance of Compensator Loops. Thermometer K, K i\ K, K K K7 istance (in ohms) 0'20 0'22 0-15
., about 7 per cent, of the fundamental interval of one of the ordinal v thermometers. Thus a change of one-thousandth part in tin- resistance of one of the leads would answer to an alteration of aliout 0-014 in the quantity tabulated in Table III, or to an apparent shift of 0'007 in the zero of an ordinary thermometer. There are various suggestive features in the table. Between ()< tol>er 18 and November 11, 1898, the difference between two leads, immaculate to the eye, altered steadily in one direction by an amount equivalent to a change of 0°'05 in the zero of a platinum thermometer. Then, without apparent cause, there occurred a sudden alteration in the opposite direction, equivalent to a change of 0°"12 in the zero. After this very erratic behaviour the leads seem to have remained practically constant for ten days. Next we come to absolutely new leads of the best construction. No one, I think, can question the reality of the differences between the results of December, 1898, July, 1899, and September, 1899. But with the occurrence of differences of the size shown during 1899, we must anticipate errors of the order of at least 0°"01 or 0°'02 in the absolute readings, unless adequate provision is made for eliminating this source of uncertainty. Changes in the leads have unquestionably been the principal cause of the apparent changes in the zeros of the platinum thermometers during the investigation. These changes have added considerably to the difficulty of working up the results. The difference between the proportional arms was recently found by Dr. Marker to be less than 1/6000 of either resistance. Doubtless the resistances were originally made as nearly equal as possible, so that there is a strong presumption that if any differential change has occurred it has been small. There is, however, unfortunately no direct evidence bearing on this point. § 10. It is only proper to remark that from some points of view slow, regular changes in the relative resistances either of the leads or of the proportional arms are not of primary importance. Such changes, at least when small, are equivalent, the one to the addition of a con- stant quantity to the observed resistances, the other to the multiplica- tion by a constant factor, so long as the measurements dealt with cover only a short time during which no sensible variation occurs. Supposing sufficiently numerous observations made of the resistances R« and EI in ice and steam, it is clear that the platinum temperature Pt, given by pt = 100 (R - Ro)/(Ri - Bo), would remain practically unaffected. But, on the other hand, the necessity of frequent zero point obser- vations is one of the drawbacks most frequently dwelt upon in the case Platinum Therinometry at Kew Observatory. 13 of mercury thermometers, and it is only fair to recognise that they do not in this respect necessarily suffer by comparison with platinum thermometers. Coil Changes, § 11. The box was originally calibrated at Cambridge by Mr. E. H. Griffiths, who referred everything to a mean box unit based on the eight coils H to A. Since the box came to Kew Observatory it has been thrice calibrated, with the aid of a convenient apparatus designed "by Dr. Harker. The observations were made on the first of these occasions by Dr. Harker ; on the other two occasions mainly by Mr. W. Hugo, Senior Assistant at the Observatory. I took a certain number of check observations during the two last calibrations, and made all the calculations necessary to construct the correction tables based on these and on Dr. Barker's calibration. In doing so, I followed Mr. Griffiths' procedure, except in some minor details. The results of the four calibrations were as follows : — Table IV.— Coil Values in Terms of Mean Box Unit. Coil. Nominal value. Calibration I (Sept. 1895). II (May, 1897). III (July, 1897). IV (Mar.-Apr., 1899). A. 640 639 -941 639 -973 639 -980 639-975 B 320 320-169 320 -093 ! 320 '090 320 -090 C 160 160 -062 160 -023 160 -025 160 -028 1) 80 80-010 80-001 | 79-991 79-991 E 40 39 '968 40 -037 40 -028 40 -023 F 20 19 -963 20 -111 20 -115 20-115 G 10 9-945 9-760 9-766 9-774 H 5 4-942 5-002 5-005 5-004 I 100 99 -957 99-989 99-975 99 -426 Mean 1 cm. bridge wire 0-995 0-993 0-994 0-995 The considerable differences between the results of Calibrations I and II are probably mainly due to changes attending the reorganisation of the box in the spring of 1897. Coil I broke at one end, and had to be resoldered just before the last calibration. It is improbable that the algebraic sum of the resistances of the eight coils H to A was unaffected by the box changes in 1897, and thus the mean box units in Calibrations I and II were almost certainly slightly different. The data obtained with the thermometers, as will be seen later, suggest a slight increase at this time in the mean box unit ; and this is, at least, consistent with the apparent change in the bridge- wire resistance. 14 I >r. ('. ' 'Inn-. / an \Vr shi-iiM rather expect the resistance of the bridge wire to im especially near its centre, through constant rubbing by the coi; pici-f, .-'inl the figures obtained in Calibrations III and IV somewhat favour thi> view. Too much significance ought not, however, to !><• •KM! to the small differences apparent. Some of the differences in the coil values in Calibrations II, III, and IV may be experimental errors ; but the changes shown in the case of, at least, E and G are, I think, too large to be accounted for in this way. 8 12. Perhaps the clearest evidence of the reality of coil variations is that afforded by an examination of the thermometric results obtained with the different coil combinations. It would occupy too much space to go into details, so I merely record in Table V the mean differences between the values of RO, RI, and Rx (resistance in sulphur vapour) obtained with the two coil combinations used during different specified epochs. The unit in the table is 0*01 mm. of bridge wire, answering approximately to CT'OOl C. with the ordinary thermometers. In the final means equal weight is allowed to each observation, so that some thermometers exert more influence than others. During some of the epochs — especially 1895 and the first part of 1899 — the data with any one thermometer were very scanty. Again, it must be remembered that the part of the bridge wire at which readings are taken differs according to the thermometer used, and also varies for any one thermometer according to the temperature of the coils, the position of the bridge centre, and the difference between the leads. It is also different in the ice, steam, and sulphur point observations, fluctuating considerably in the latter two wises with the barometric pressure. Thus the fluctuations in the table amongst the results for a common epoch are due to many causes. $ 13. It is, I think, most instructive to start with Calibration III nvide in July, 1897. During the rest of that year the results from the two coil combinations show almost perfect agreement; in 1898 the results drift apart, and the drift is accentuated in 1899 prior to Cali- bration IV. Again, for some months after that calibration there is an excellent agreement, though a tendency to drift soon manifests itself. ' Calibration II seems less successful, but it was made by an observer different from the one who took the readings on which the results in Table V are based. A different standard of plug tightness does not influence all plugs alike, and in my experience the personal equation in this matter requires to be reckoned with. The data from March 12 to 19, 1897, were so outstanding that they are given separately. The exposed parts of the box had, I believe, Ixjen cleaned shortly before, and conceivably one of the coil supports ni iv have got a knock. There was, however, no suspicion of this at the time. This fact emphasises the necessity of a constant outlook for possible changes. Platinum TJicrmoinctri/ at Kew Observatory. 15 1 §d a-" CO 00 ! iH — l . IM rfi (N 1 + • III M O CO >* i : : : : , rf •^ 00 TP iH CO O i-l NO i-H MS II + + 111 + i M 00 (M O CO O ^H CD (N i-l . .l-H CO MS + 1 + 1 1 1 1 _ M i-4 IM l> O •* i-H . •* . l-l I ' + + 1 + rf oqj>. T*I o cocoiMt* '>, •i 7 . ^ ^ " '"'* rH W - *T O 'O 1--' *«• f- 3 '^ 0^ C* • • 1 Oi C5 C-- o&; C5 Ha ^ "^ ^ C3>. US 00 QOS 002 00^" 00 00 00 OOeS rH rH rHl^ rH^ rH l-H rH -- rHI_( f£ *^ 1 1; Dr. < '. < 'hive. / "u \ nmplete calibration is a laboi ions process, and its frequent repeti- tion would add seriously to the work entailed by platinum tlu-:iiio- metry. If, however, the results in Table V are normal, it would appear that frequent calibrations can hardly be avoided in the case of physical k of the highest accuracy. Faulty Action of the Contact Piece. §14. Parallel to the real bridge wire in the Kew resistam ••• box runs an exactly similar wire connected to the galvanometer, and the contact piece works by pushing down a short cross wire so as to span the interval between the wires. In January, 1898, it was noticed that mere moving and resetting the contact piece might alter the reading, and I found that the pressure to which the cross wire was exposed when contact was made — always acting on the same part of its surface — had cut a groove in it. The Instrument Company put this to rights, but the phenomenon had repeated itself by January, 1899, though to a smaller extent. On that occasion the Instalment Company made an alteration which, it is hoped, will prevent the recurrence of this trouble. It is difficult to keep the two parallel wires equally tight and exactly at the same horizontal level ; thus the cross wire may bear unduly heavily on one wire before it makes good contact with the other. This defect would be of little consequence in an open scale bridge wire ; but in the Kew box it produced, when at its worst, uncertainties of the order 0 -04, or even 0°'05, in individual readings. Of course this merely tended to introduce irregularity in the readings, and supposing a number of observations taken, could hardly simulate a change in a thermometer. Shift of the Bridge Centre. S 15. By the bridge centre I mean the vernier reading when a balance is made with all the plugs in their holes, the platinum thermometer and compensator resistances being cut out by short-circuiting straps. The original departure 0*006 in this centre from zero on the scale was given as a fixed correction in the first calibration table, and as no provision existed for determining the centre, I did not for some time properly appreciate the situation. My attention was first roused by a sudden apparent discontinuity in the values of RQ in the spring of 1897, for which the only apparent cause was a cleaning up of the box. The Instrument Company then supplied two short-circuiting straps, to be put across the CC and PP box terminals, and since then determinations of the centre have been made at the beginning of each observation day, and usually at intervals throughout it. Soon it became apparent that during a day's observations the bridge centre is apt to drift towards the mimis side of the scale ; the Platinum Thermometry'at Kcw Observatory. 17 same side as it goes to when the plugs are dirty and have an increased resistance. The amount of the drift is very variable. One day it may be negligible, while the next day it may be as much as 0°'04 ; it is occasionally more than this. The phenomenon is probably partly due to the fact that frequent pulling out and putting in of the plugs seems inclined to produce the grey powder already referred to. It may also arise from the observer's standard of plug tightness falling off as he becomes tired, or from asymmetric heating of the box through the proximity of the observer's person to the minus side of the scale. There are probably various influences at work, some of which may be peculiar to the particular box. One cannot be perpetually taking bridge centres, so that even the more recent observations are exposed to some uncertainty from this source. The practice of regularly taking bridge centres was not introduced until 1897, and it is of course possible that there was little or no occasion for it in the original box with brass plug holes. Thermo-electric Currents. § 16. The principal seat of thermo-electric currents seems to be the junc- tions at the head of the thermometer ; and it is desirable to shield the head as much as possible from heated air or vapour. When a Griffiths key is used these currents are not necessarily a source of error, but they . tend to increase the difficulty of taking readings. Originally it was sup- posed that the use of a Griffiths key rendered further care unnecessary, and no commutator was used until after the restoration of the box in 1897. As the Instrument Company then transferred certain terminals to the key-board, it is possible that the original box suffered but little if at all from the source of trouble now to be described. When the commutator was introduced the following phenomena were observed. When the day's observations began the readings with the current d and r (direct and reversed) were usually about the same. After a few minutes' observations the difference between the d and r readings began to increase, and nearly always in one direction. After a little the difference usually assumed a fairly constant value, but con- siderable fluctuations might occur, especially if the temperature of the room was rendered unsteady. On the advice of Mr. W. N. Shaw the thermo-electric key was enclosed in a padded box. This has decidedly diminished the evil, but it still appears expedient to take readings with the current both ways. Table VI shows the state of matters typical before and after the introduction of the protecting box. When the thermo-electric effect is not eliminated the error in the reading is (d cx< r)/2. Supposing d-r to remain constant through the ice, steam, and VOL. LXVII. C 18 < hree.' Investigations on Table VL Without protecting box. With protecting box. Thermometer. Hour. Difference r-d. Thermometer. Hour. Difference r-d. K, h. in. 11 45 11 64 12 12 12 39 2 56 3 30 3 41 4 20 o-ooo 0-064 0-094 0-106 0-086 0-105 0-107 0-121 K,.. h. in. 11 2 11 35 12 20 12 50 2 31 2 50 3 33 3 50 3 59 4 26 o-ooo 0-003 0-017 0-016 0-020 0-014 0-018 0-027 0-022 0-019 K K,. . K,. . K4. . sulphur point observations with a particular thermometer, the KO, KI, and R* would be wrong to the same amount, so that the fundamental interval and pt, (the value of pt answering to the boiling point, of sulphur) would be unaffected. In practice, however, we cannot antici- pate so favourable a contingency. After some experiments, I found that I could at pleasure change the amount, and even alter the sign of the r-d difference by heating with the finger or cooling by an air-blast one of the terminals of the Griffiths key. Heating due to the Battery Current. § 17. In the Kew apparatus there is an option of two resistances, viz., 100 and 20 ohms, in the battery circuit ; and our original instruc- tions were to use the 100 ohms when observing with an ordinary thermometer in ice and steam, and the 20 ohms when observing at the sulphur point. The object, doubtless^was to ensure sufficient galvano- meter sensitiveness at high temperatures. In an early sulphur experiment I was surprised to find that the reading was about 00>07 higher with the 20 ohms in use than with the 100. There being only a single dry cell in use, I had not anticipated a sensible difference. Supposing the cell-to [remain constant, the heating effect would of course remain the same for a given thermometer at any one fixed point, and so would not influence the results immediately in view. It seemed, however, inexpedient to trust to this, and we have accordingly employed the 100 ohm resistance in all the regular observa- Platinum Thermometry at Kew Observatory. 19 tions, merely using the 20 ohms occasionally to furnish data from which to calculate the heating effect under the normal conditions. To under- stand how this is done we must glance briefly at the theory. Let E be the E.M.F., R' the internal resistance of the battery, K the rest of the resistance in the battery circuit. Let r\ be the resistance in either of the proportional arms, r the resistance in the bridge arm containing the platinum thermometer, whose spiral has a resistance p. Then if i denote the current in the spiral, H the heat given to it in unit time, we have H = *> = E2r12PH-[(R + K')(r + r1) + 2rr1p ......... (1). The heating of course is gradual, and theoretically it might be possible, by rapid, skilful manipulation of the tapping key, to obtain a balance before there is a sensible effect. In practice, however, this is hardly possible in work of the highest accuracy. Only by the remotest chance does one hit the balance at the first attempt, and, as a rule, the key must be put down a good many times. Also, unless the key is held down a sensible time, a small absence of balance may be overlooked. The weaker however the current, the longer the time before sensible heating exists, and with the 100 ohms in the circuit it seems possible to get a fair balance before the heating effect is appreciable. Thus in comparing platinum and mercury thermometers at high temperatures, where accuracy of the order 0°*01 C. is usually much above what is necessary, the bridge has been regarded as balanced when no movement appears in the galvanometer on depressing the key. In the fixed point observations, on the other hand, the resistances have been adjusted until no movement appears on releasing the key. Under these circumstances the current has exercised its full heating effect. The platinum spiral is presumably heated to such small extent above the surrounding temperature as is required for its gain of heat from the current to balance the loss by radiation, conduction, &c. We may pretty safely assume that this excess of temperature is proportional to the heat given to the wire, but it must also depend on the specific heat of the platinum, on its radiating and conducting power, and also conceivably on the shape, dimensions, and material of the enclosing tube. I understand that several authorities'* have proposed to keep i-p constant in the spiral by suitably altering the battery resistance. This, however, as the above reasoning shows, cannot secure constancy in the temperature excess of the spiral at all temperatures. § 18. The application of the formula (1) presents difficulties. We know — at least, very approximately — the resistance r\ of the propor- tional arms, and the resistance p of the spiral, at any "fixed point" temperature, but it is not customary to measure directly the resist- * For instance, Waidner and Mallory, 'Phil. Mag.,' July, 1899, p. 14. c 2 20 l>r. ('. Chree. In >/is on ance r. It exceeds p by an amount equal to the resistance of the leads, including the thick platinum wires inside the thermometer tul>r. The mean temperature of these latter wires depends partly on the temperature of the spiral, partly on the length of tube immersed, and partly on the temperature of the external air. Of course we can measure r on any given occasion, but its fluctxiations would be consider- able under normal conditions. My object being rather to get a fairly exact general idea of what to expect under the varied conditions of normal use, than to measure the phenomena with extreme exactness under an arbitrary set of conditions, I have accepted f or r - p a rough approximation, good enough for my special purpose. The internal resistance of a dry cell varies very much according to its freshness. I have thus made two calculations, in which this resist- ance, R', is taken as 0 and as 5 ohms respectively. The latter figure actually applies to the later experiments on the relative heating in ice, steam, and sulphur vapour ; while the former answers sufficiently to the case of a new cell. The E.M.F. of a dry cell falls off in time, though not to a very large extent, and for our present purpose such a change does not concern us. We do not need to know the value of H, as given by (1), in absolute measure, but only the relative values of H with varying p and r, but with constant E and r\. In the following calculations I have supposed r\ = 5 — the propor- tional arms having each a resistance of 5 ohms — and for an ordinary thermometer p = 2*6 in ice, 3*6 in steam, and 6*8 in sulphur, while r - p = 0'4 ; for K« we have p = 6'45 in ice and 8'95 in steam. The calcu- lated values of (H/EVr) x 107 are given in Table VII. Table VII.— Values of (H/E2^2) x 107. Fixed point. Ordinary thermometers. K«. Difference Difference E'. R-100 60 40 30 20 20 and 100 R = 100 20 20 and 100 Ice.... JO 15 38 100 212 357 720 34 — — — 491 682 457 37 484 447 Steam. . {I 41 107 225 375 744 37 92 182 286 513 703 476 35 437 402 Sulphur {S 41 — — — 681 37 — — — 478 640 441 — — — The most striking features in the table are the comparative smallness of the differences between the supplies of heat at the three fixed Platinum Thermometry at Kew Observatory. 21 points, with a given value of K', and the relatively large effect of an addition of 5 ohms to the battery circuit when the rest of the resistance is only 20 ohms. No very serious error would arise from treating the heat supplied as independent of the temperature, at least up to 445° C., so long as there is a large resistance of the order of 100 ohms in the battery circuit ; and under these conditions a moderate change in the resistance of the battery itself is not of much consequence. § 19. At any one fixed point, as already stated, we may reasonably expect the rise of temperature of the spiral to be proportional to the supply of heat. . It is desirable, however, to check this conclusion by experiment. I thus compare in Table VIII the above theoretical results with some experiments made with thermometers Ko and KS at the steam point in June, 1897. In these the additional resistance in the battery circuit — beyond that of the battery itself — was given several values intermediate between 20 and 100 ohms. Assuming Heating effect/heat supplied = C, a constant, I determined C by equating the observed and calculated values of the difference of the heating effects with the two extreme additional resist- ances 20 and 100 ohms. This quantity is denoted by H2o - HIOO, and a similar notation is employed for the other similar difference effects. The differences in the table are really lengths of bridge wire, but for practical purposes they may be treated as temperatures. Table VIII.— Heating Effects of Battery Current at Steam Point. B'. H.-H™. H-p 30 -f- 100- U IT Jdjo — ^MOO* HTT 60—±1100- Hlw. H«. TT "•W H30- H.JO. .- /Calculated! K-2-) I 0 5 0-165 0-078 0-086 0-043 0-050 0-015 0-019 o-oio 0-013 0-025 0-032 0-053 0-063 0-088 0-099 0-175 0-178 1 Observed ... >. o-ioo 0-047 0-019 £ f Calculated -j 0 5 0-185 0-088 0-097 0-048 0-057 0-017 0-022 0-011 0-014 0-028 0-036 0-059 0-071 0-099 0-111 0-196 0-199 '•Observed ... 0-097 0-061 0-026 The resistance of the battery itself was not determined on either occasion ; but from the large size of the heating effects compared to those of later experiments with an older cell, I am inclined to suppose that the resistance was nearer 0 than 5 ohms. The values calculated with R' = 5, however, unquestionably present the closest accordance with experiment. In fact, taking into account the difficulty of the experiments, the agreement in this case is altogether too good, and must be largely a matter of luck. However this may be, there is, I think, every reason to suppose that the values deduced for the heating L'L* Dr. < '. l.'lnvr. / I on effects HIOO, &C., with the additional battery resistances 100, &c., are a very fair approximation to the truth. There seems little doubt that with a fresh dry cell the reading at the steam point is raised fully the hundredth of a degree, even with 100 ohms in the Iwittery circuit. If 20 ohms only were inserted, the heating effect would amount to nearly two-tenths of a degree. In this latter event, an increase of 5 ohms in the battery resistance would lower the observed value of Rj by about 0°-06 C. $ 20. We have next to consider the relative values of the heating effects at the three fixed points. These have been determined by a series of special determinations of the values of HSO r- HJOO- During the majority of the experiments the battery resistance was approxi- mately constant, and not far from 5 ohms. The following table gives the observed values of the ratios (H-20 - HI oo at steam point)/(Hjg - HIOO at ice point) and (H->o - HIOO at sulphur point) (H2o - HIOO at ice point) for the several thermometers, as well as the values calculated from Table VII for the ratios of the heat supplies. Table IX. — Differential Heating Effects at Ice, Steam, and Sulphur Points (found experimentally). 1 K;. K... K,j. K.J. Ks. KT. Mean K^OK;. K6. Steam ice 0-77 0-77 0*74 0-75 Ratio of heat supplies. . . . 0-67 0-75 >R' = 0 0-74 1-03 0-54 Sulphur/ice. . . . 0-33 0-26 0-17 0-40 Ratio of heat supplies. . . \ R' = 5 0 '42 0 -29 r R' = o 1-04 0-31 0-94 0-90 .. " " \ R' = 5 0-97 The table is based on eighteen experiments comparing the (steam/ice) ratios, and eight experiments comparing the (sulphur/ice) ratios. In most cases the several steam/ice observations with the same thermometer showed a remarkably good agreement, but the results with K5 and K- were less harmonious than the others. The sulphur/ice observations were so few that little weight attaches to the results for the individual thermometers ; but the final mean should not be much in error. The highest value of the sulphur ice ratio observed was 0*42 in the single experiment made with K5. A considerable number of heating experi- ments have been made in sulphur, but I have employed in Table IX — as in the case of steam — only those which were made on the same day Platinum Thernwmetry at Kew Observatory. 23 as a corresponding experiment with the same thermometer in ice. This is necessary on account of possible changes in the resistance of the dry cell. At the sulphur point the heating effect is only about one-third of what it would be if it depended only on the value of i-p in the spiral, and even at the steam point the heating is less than three-quarters of what it would be on this erroneous hypothesis. § 21. If the heating effect due to a given i2p were the same at all temperatures, Tables VII and VIII would justify the conclusion that with 100 ohms in the battery circuit all that is to be feared is a trifling fall, of the order 00-003, in the apparent zero E0 as the cell deteriorates. But, according to Tables VIII and IX, the true conclusion is that even with 100 ohms in circuit the heating effect is likely to diminish the fundamental interval by several thousandths, and the ice-sulphur interval by not much less than a hundredth of a degree. The use of 20 instead of 100 ohms in sulphur point observations would largely increase the uncertainty. The changes in Rg - EQ and RI - RO, due to varying heating effect, are in the same direction, so that the effect on pts is relatively small, and, in view of the numerous uncertainties existent, I have not thought it worth while to attempt a " reduction to an infinitely small current." It is clear, however, that unless the current can be materially reduced by the employment of a more sensitive galvanometer, the necessity of providing for this reduction must be kept in view in all work of the highest accuracy. Before quitting the subject, I would remark that the diminution of the heating effect of a given i2p as the temperature rises is only what we should expect from the known increase with temperature of both specific heat and radiating power. It is possible that heating effect experiments with thermometers of other metals than platinum, or with spirals of varying gauge of wire, might prove a useful method for investigating radiating power. Error in the Temperature Coefficient of the Coils or Differences between the Temperature Coefficients of Different Coils. § 22. The standard temperature selected by Mr. Griffiths was 20° C., and he" found for the relation between the resistance rT of a sample of the coil wire at temperature r, and its resistance r20 at the standard temperature, TV = r20 [1+0-00026 (T- 20)] (2). All temperature corrections have been deduced by (2) since the box came to Kew Observatory. Suppose after the calibration corrections are applied, that the observed box reading makes p the resistance of a certain platinum •_M Dr. C. Chrec. ////•< .•>•//>/. Now suppose R« the true resistance at t° of a platinum thermometer whose true resistance at 0° is RO, and whose fundamental interval is I. Then the platinum temperature is pt = 100(Rt-Ro)/I (3), and corresponding to any correction (or error) AR,, arising from a cause other than errors in RO or I, we have for the correction (or error) in pt the formula Ap/ = AR,(100/I) (4). Supposing (2) correct, and the -box at temperature T, we must have in order to balance the bridge R, = p [l + 0-00026 (T- 20)] (5). If then Ap/ is the correction to pt arising from our application of a temperature correction to the box readings, we have Ap/ = (100/I)p x 0-00026 (T - 20) (6), where p, like I, is measured in mean box units. For one of the ordinary Kew thermometers the following are suffi- ciently near values for purposes of illustration : — Ice point. Steam point. Sulphur point. P = 258 358 678 Corresponding to these representative values, with I = 100, we have Ice point. Steam point. Sulphur point. Apt = 0-067 (T - 20) 0-093 (T- 20) 0-176 (T- 20) Using Calendar's formula, t-pt = S[(*/100)2-*/100] (7), we have for the small increment A/ on the air scale corresponding to a small increment A/tf on the platinum scale A* = Apt -=- [1-10-4S(2<-100)] (8). Assuming for the sulphur point on the air scale Callendar and Griffiths' Platinum Thermometry at Kew Observatory. 25 value, tg = 444° '53, then in the Kew thermometers, as will be seen later, 8=1'5 nearly, and we have approximately At the ice point A£ = bpt x 0'985 = 0'066 (T - 20), *| „ steam „ A/ = Apt x 1-015 = 0'094 (T - 20), I ... (9). „ sulphur „ & = Ap^x 1-134 = 0-200 (T- 20) J § 23. Supposing that we aim at an accuracy of 0°-001 C., and are able to keep our coils within 1° C. of the standard temperature,, then it will suffice to know our temperature coefficient correctly to 1 part in 66, to 1 part in 94, or to 1 part in 200, according as the " fixed point " concerned is the ice, steam, or sulphur point. Accuracy to 0°'001 C. at the boiling point is not a very extravagant aim, and it thus appears desirable to know the temperature coefficient correctly within 1 per cent, unless we can keep nearer than 1° to the standard temperature. From correspondence with Mr. Griffiths, I found that he did not anticipate so high a degree of accuracy as this in his determination of the temperature coefficient, and as a matter of fact two fairly complete experiments by Dr. Marker have given values from 0'00025 to 0'000245 for combinations of several of the actual box coils. The allowance of a departure of 1° C. from the standard temperature is a narrow one, except under specially favourable conditions. A good thermo-regulator, no doubt, can control the gas supply with consider- able accuracy, even under unfavourable conditions; but what the thermo-regulator controls is its own temperature. That temperature is probably in general close to the mean temperature of the immediately surrounding water. But supposing, as may happen in the platinum- thermometer room at Kew, that the air temperature is only 14° one day and 19° the next, then the temperature of the coil chamber will be much higher on the second day than on the first, supposing the thermo- regulator to have been left untouched. One cannot tell in advance what is going to happen during the night, and if one finds in the morn- ing the box thermometer indicating 17° and attempts to bring it up to 20", then either a long time is wasted in waiting during a very gradual rise, or else one is very uncertain as to the uniformity of the tempera- ture in the coil chamber. Again, in summer the room may be at 22° or 23°, or more, and the only simple way of keeping the coil chamber at 20° — viz., constant supplies of cold water in the bath — would probably introduce uncer- tainties greater than those it removes. On the whole, experience at the Observatory suggests that it is best to keep the box nearly at the temperature of the room, so that seasonal* variations of T between 17°'5 and 22° '5 have been common. Supposing * In the course of a single daj's experiments the temperature changes were usually only a few tenths of a degree. 26 Dr. ('. < 'hive. ////-.*//»/"/ r-20 = 2°-5, an error of 5 per cent, in the temperature coefficient would introduce the following errors in /: — Near 0° C. Near 100° C. Near 445° C. 0-008° 0-012 0-OJ.-> The above remarks suppose that we are concerned, as in the Kew experiments, with the u/i*ili>ff value of the platinum thermometer resistances, and have to do with observations taken on different days. In ordinary thermometric work uncertainties in the temperature coefficient would be much less serious, supposing one to determine the fundamental interval daily, and to combine together only such observa- tions as are taken on the same day. Under such circumstances it would not matter how far removed from 20" the temperature was, or how much in error the temperature coefficient was, so long as the tem- perature of the coils was strictly constant. Unless, however, the room containing the platinum thermometers is kept at a constant tempera- ture artificially, or is exceptionally situated, its diurnal range of tem- perature will seldom be inappreciable. Another source of uncertainty is the possibility of differences between the temperature coefficients of the different box coils, or between those of the two proportional arms. Such differences have not been observed at Kew, but there has been hardly any direct experiment, and the com- parison of the results from two coil combinations, such as CDF and CDGH, in which only the small coils are different, does not afford a delicate test. As a difference between the temperature coefficients of the propor- tional arms was suggested by the Cambridge Instrument Company as a possible explanation of some phenomena (due, however, to some other cause), it is presumably a contingency which should not be dis- regarded. Error in the temperature coefficients, it should be noticed, would only introduce irregularity into the results of the Kew experiments. It might conceivably introduce an apparent seasonal variation in R) or in the fundamental interval, but could hardly simulate a true secular change. Change of Zero or other Error in the Mercury Thermometer inside the Coil Chamber. § 24. An error of 0°'l C. in the estimated temperature of the coils affects the calculated temperature in the case of one of the ordinary Kew platinum thermometers by 0"'007 at 0° C., 0°-009 at 100° C., and 0°-020 at 445° C. Thus for accuracy of the order 0°'001 at 100° C., or 0°'002 at 445 C.. the temperature of the coils must be known to 0°'01 C. An ordinary Platinum Thermometry at Kevi Observatory. 27 tin calibrated mercury thermometer, even Avith an open scale, cannot be relied on to give temperature differences to this degree of accuracy under the most favourable conditions. The thermometer in use in the Kew box is a good calibrated one, and errors in its graduation are unlikely to exceed 0°'01. It is, how- ever, subdivided only to fifths of a degree, and individual readings cannot claim an accuracy of 0°'01. Fortunately, errors in reading merely introduce irregularities in the results, and so are not of fundamental importance in the present inquiry. A. more subtle source of trouble is the secular change of zero, normal to mercury thermometers. A rise of 0°'l in the zero — which is not very much in excess of what actually occurred during the four years occupied by the experiments — would, in fact, exactly simulate a rise of 0°-007 in the value of K0. In the ordinary use of platinum thermometers, at least when there are moderately frequent observations of RO, the circumstances are different. The slow secular change of zero in the mercury thermo- meter is then of little moment, while great importance attaches to the accuracy of the individual readings. Failure of the Box Mercury Thermometer to give the True Temperature of the Coils and Bridge Wire. . § 25. If the room temperature is below 20° C., and artificial heating is employed, we may have a practically stationary reading on the box thermometer, and yet find it differing by several tenths .of a degree from a second thermometer whose bulb is at a different level in the coil chamber. Under such circumstances the temperature inside the coil chamber also varies with the distance from the side or end walls. The coils are of various shapes and diameters, some coming much nearer the top and bottom of the coil chamber than others, and they are necessarily at different distances from the ends. Thus, under the conditions specified above, different coils may possess different mean temperatures, and the temperature given by the mercury thermometer may differ sensibly from that of any one of the coils. Even with no artificial heating of the box, the temperatures at different parts of the coil chamber may differ by several tenths of a degree, when the room temperature is changing moderately fast. In such a case the presence of water in the protecting tank seems a positive drawback, as it makes the temperature of the coil chamber lag behind that of the rest of the room. When the temperature of the coil chamber is altering, whether through artificial heating or otherwise, the various coils doubtless alter 28 Dr. C. Chree. / tions on their temperature at different rates, and unless the mercury thermo- meter is specially chosen it may have a greater or a smaller lag than any one of the coils. The mercury thermometer used in the Kew box, like most open-range thermometers, contains a considerable quantity of mercury, and when the temperature changes fast it lags very sensibly behind the coils. For instance, in some special steam point observations with KI, when the recorded box temperature rose 2°'5 in twenty minutes, the corrected bridge-wire reading fell about 0°'06, instead of remaining practically constant as it ought to have done. The figures really show that the temperature of the box thermometer had lagged about 0°*7 C. behind that of the coils in use. This of course was an extreme case, and under ordinary experimental condi- tions the difference between the temperatures of the box thermometer and the coils is hardly likely to exceed 0°'3, and is probably seldom half this. Necessarily, however, a good deal depends on the depth of the bulb of the thermometer inside the coil chamber, and any alteration of that depth should be carefully avoided. § 26. The temperature of the bridge wire is open to much greater uncertainty owing to its relatively exposed position. The whole bridge wire resistance in the Kew box is fully 30 box units, and if this were all on the platinum thermometer side of the Wheatstone bridge the error arising from the assumption that it possesses the temperature of the box thermometer would be approximately O'OOSr, where T is the error, in degrees Centigrade, in the temperature assigned. As T may easily amount to several degrees, an error of several hundredths of a degree might easily creep in from this cause if one used a large fraction of the bridge wire. Theoretically it is possible to keep the bridge-wire reading within 5 units of the bridge centre,* thus reducing the uncertainty to less than one-sixth of that existing in the extreme case supposed. This, however, has not been compatible with our practice of employing the same coil combinations for all the thermometers of the same pattern. Even with the drawback of temperature uncertainty it is often advantageous to make a liberal use of the bridge wire. It facilitates finding the balance, and in work such as the comparison of mercury thermometers at high temperatures, where rapid reading is essential, it is a great convenience. It is thus highly desirable that the bridge wire should be better protected than in the Kew box, and that its temperature should be measured directly. Insufficient Immersion. § 27. Supposing a correct mercury thermometer to be immersed in ice up to the reading - 10° C. on the stem, there is a column * This was the policy recommended by Mr. Griffiths originally. Platinum Thermometry at Kew Observatory. 29 of mercury of about 10-degree divisions exposed to the influence of the surrounding atmosphere, whose temperature is, say, 20° C. This emergent column possesses a mean temperature somewhere between 0° and 20°, and so occupies a larger volume than if all were at 0° ; on this account alone the thermometer must read too high. The emergent column has a second influence tending in the same direction ; it serves, conjointly with the glass tube, as a path for the conduction of heat to the bulb. In high-temperature measurements the error due to a long emergent column may amount to several degrees, the thermo- meter in this case reading too low, The conduction along the stem is usually much the less important influence of the two. A platinum thermometer is not wholly free from immersion diffi- culties. It is not sufficient to have the spiral inside the bath or chamber whose temperature is desired, so that the defect in a platinum thermo- meter is most analogous to the second source of error described in the mercury thermometer. If, for instance, a platinum thermometer is buried in ice to only a little over the top of the spiral, the tube emerging in a room at 20° C., the heat communicated to the spiral through the surrounding air or through the connecting wires raises the reading appreciably. Similar insufficient immersion in the hypsometer sensibly lowers the reading. It was difficult to settle on a suitable basis- for comparing the effects of insufficient immersion in the different platinum thermometers. The spiral is wound on a mica frame, and usually one of the small mica discs which hold the wires apart is situated immediately above the top of the frame. When this occurred I have taken the " bulb " of the thermo- meter as extending to the first mica disc ; but in one case, where the interval between the first disc and the frame was considerable, I took instead the top of the frame itself. As a rule, the spiral stops short of the top of the mica frame by about f cm. ; so the definition is at least a lenient one to platinum thermometers. The tubes of K5, K6, and K7 being of transparent glass, the length of the " bulb " was visible from outside ; but in KI and K4 the tubes had to be taken off to allow of the requisite measurements being made. In the immersion experiments in steam the thermometers were carried in the usual way by a cork, about 3'25 cm. long, fitting the neck of the hypsometer, and the immersion was counted up to the lower face of the cork. Thus fully 3 cm. of the tube immediately above the portion exposed to the steam was protected from direct cooling by the air. § 28. Table X gives particulars as to the results obtained. It includes the errors observed with the various immersions, measured on the hypothesis that an immersion of " bulb " + 10 cm. was in all cases sufficient. It also gives the lengths of the " bulbs," and the mean tem- perature in the room during each series of experiments with each 30 Dr. C. Chiv.-. Investigation U •*• i>. + §— 1 O t^ U? o S o -H 5 o o o «o o o o o b 1— 1 9 o A e s i 0 us + §O N OS **• S 8 1 o °O 0 O O O b ? o I a • J s o HOI 64 + -1 « MS •<* ri Oi O o rt O >— • •— i O ^^ oo o o o o 0 •-I o ^ 0 1 e g 6 •*» 1-5 CO 00 O 1-1 do M »? oo us O •* N •-( *) § IM 5 1 + °b o o o o o 0 02 d jf. .0 O ^? *>• O 00 B 9 ^P B 4 IM O lO OO •* t^ $ iH » °O O O 0 0 o E HM t- °o b o o o o 9 b a _o 'S O us •* d «O «i M + CO 1O O -H C5 9 ? ? 9 9 "o o o o b o 9 o CO ff o •c & H • § h-l 1 -*» i— i •»• §CO Tji US t^ N N O '-3 *1 N rH iH °O 0 O O O t^ us I— 1 o s o O + JZ en us to Tf ^ 00 US OO US 00 t-« US CO i-t N « eo iH •z P5 °0 O O 0 O o l~i 3°; a a d «*• -» _c. go Tti •* «O •* us ^> a 0 £ |5 :- - Q00 O 00 US ID o 1 Jp !•§ _ x- 00 a- a KM c o tj «2 c a. ^H VO «O « *? •* M « N 6 • s = II — a EH M W W «" M" 1 s -Si -S 9 i* - Platinum Tliermometry at Kew Observatory. 31 thermometer. This last information is of most consequence in the case of the experiments in ice, because the errors must then increase as the temperature of the room rises, and during the immersion experiments the room temperature was much below its ordinary summer value. Some of the figures are based on two or three complete independent experiments, but others depend on only one experiment. In most cases there were in addition several check experiments, with only two or three different immersions, but I have left these out of consideration in constructing the table. The results cannot claim any very high degree of precision, because it is difficult to secure uniformity in the conditions, more especially when the immersion is in ice. The difficulty is greater the less the immersion. When the " bulb " only is immersed in ice, the reading after attaining a minimum usually rises somewhat. The phenomenon seems mainly due to melting of the surface ice, with consequent deterioration of contact between it and the thermometer ; but possibly a difference in the circumstances under which the thick platinum leads and the compensator wires cool may possess some influence. In taking the minimum readings, as I have normally done, I have of course, if anything, favoured the thermometers. This has also been done to some small extent in assuming that "bulb " + 10 cm. is in all cases ample immersion. As a matter of fact, an unmistakable difference was observed in the reading of several of the thermometers when the immersion was increased from " bulb " + 10 cm. to " bulb " + 12 '5 cm. This difference was not, however, always in one direction, .the phenomena in one case distinctly suggesting that cooling an additional 2-5 cm. of the compensator wires had more effect than cool- ing an additional 2 '5 cm. of the platinum leads. The experiments are not sufficiently varied to justify any too con- fident explanation of the causes of the large differences between the different thermometers. Thus the pre-eminence of the errors in K4 might at first sight be attributed to the fact that the tube is of porcelain and of exceptionally large diameter ; but KI is also of porcelain, while KG has a much larger diameter than K;. Perhaps the most probable explanation is that K4 is short for its tube, so that the mica frame — which is exceptionally short — only conies to within about 1 cm. of the end of the tube. Ceteris paribus, we should expect that with a given immersion, the error would vary approximately as the difference between the tempera- ture of the immersed portion of the tube and that of the air of the room, but to confirm or refute this conclusion would require a consider- able number of careful experiments in which either the temperature of the bath or that of the room alone is varied, all the other conditions remaining unchanged. We cannot make use of the experiments in ice and steam for this purpose, because the conditions under which the transfer of heat occurs are radically different. We can meantime only :',L' I >i. ( '. < 'In. •!-. / i on note as an isolated fact that the immersion errors were only about 50 per cent, greater in steam than in i«-c, \vhm-as the (litlcivnce between the temperatures of the room and the immersed portion of the ther- mometer was about six times as large in the former case as in the latter. It may 1)6 only a coincidence — but if so it is a very curious one — that in the mean results, both in the ice and steam experiments, the errors diminish in geometrical progression as the immersions increase in arithmetical progression. An addition of 1 '25 cm. (i.e., one-quarter the length of the " bulb ") lowers the error almost exactly one-half. § 29. Returning to the general question, we see from Table X that the immersion should be at least 10 cm., in addition to the length of the "bulb"; in all, in the Kew thermometers, at least 15 cm. (6 inches). If the whole length immersed is less than 6 cm., the error in an ice or steam point experiment is unlikely to be much under 0°-1 C. ; and until the total immersion exceeds 10 cm., the error is likely to exceed 0°-01 C. In " fixed point " experiments it is usually easy to have an immersion exceeding 15 cm., and the main results dealt with in this paper are, I believe, free from appreciable uncertainty on this ground. There are, howeverj other circumstances under which immersion difficulties arise, as for instance in the comparison of mercury and platinum thermometers at high temperatures. Unless both the material of the bath and the contained liquid are transparent, we must have the divisions of the mercury thermometer which we have to read emergent. But with ordinary stirring it is also essential, except in a specially protected bath, that the bulb of the mercury thermometer and the spiral of the platinum thermometer should be about the same level. These conditions are often inconsis- tent with sufficient immersion of the platinum thermometer when we are comparing the lower part of the scale of the mercury thermometer. In fact, my attention was first directed to the question of immersion of platinum thermometers through the unexpected result that raising equally the height of K; and a certain mercury thermometer in a bath of molten metal lowered the reading of K7 most. Another case where insufficient immersion is to be feared is in the determination of melting points of metals such as silver. The crucible supplied for use with melting silver at Kew would, if full to the lip, allow an immersion of only 10 cm. in all. As the silver is stirred during part of the experiment, and is too precious a material to splash all over the furnace, the immersion cannot well exceed 9 cm., and may be considerably less. It is by no means impossible that insufficient immersion may have Platinum Thermometry at Kew Observatory. 33 been accountable for the fact that the mean values of the melting point of silver from experiments by Mr. Hugo and myself (six in all) were only 958°'2 C. with Kx and 957"'5 with K2; while Mr. Heycock and Mr. Neville, at Cambridge, found 961°'l C. with the same sample of silver but a different thermometer. This is 2 individual observation at Kew Observatory. above the highest Slowness of Platinum Thermometers in Acquiring Hie True Temperature. § 30. When a mercury thermometer is taken from a room at one constant temperature into another room at a different constant tempera- ture, some time elapses before the reading becomes steady ; while, if the temperature of the surrounding medium alters, the thermometer shows a lag which varies with the thickness of the glass walls and the amount of mercury in the bulb. As we have seen, the naked coils in the Kew resistance box have a smaller lag than an ordinary mercury thermometer, but it is otherwise with platinum thermometers. Here the wire does not come into direct contact with the external medium, and the attainment of the steady state is somewhat slow. Experiments have been made with some of the Kew thermometers which have been suddenly transferred from a bath about 15° C. into ice, or into a hypsometer in which the water is freely boiling. The interval, in seconds, was observed which elapsed before the bridge-wire reading came to 1°, 0°'5, or other fixed distance of the final stationary position ; this position had been found in a preliminary experiment. In all cases the immersion was similar to that of the ordinary fixed point experiments. Two mercury thermometers, Nos. 686 and 750, were experimented on in an exactly similar way for the sake of comparison. They are both Kew standards of the following dimensions : — Length of fundamental interval. Distance of zero mark from lower end of bulb. Length of bulb. Diameter of bulb. K.S. 686 K.8.760.... cm. 35 32 cm. 21 8 cm. 2-5 2 2 cm. 0-6 0-5 I have not thought it necessary to go into full details of the observa- tions. The following table gives the number of seconds required to get to 1° and 0°-5 of the stationary temperature in ice, and to 1°, 0°'5, and 0°'25 of the stationary temperature in steam. It was found that VOL. LXVII. D 34 Dr. C. Chree. Investigations on tin- rapidity of attaining the steady state in ice depended greatly on the tightness of the packing and the moistness. Thus, particulars are given of experiments made with the ice variously treated — Table XI. Thermometer... Ice. Steam. As usual, but not packed tight. As usual, packed tight. Freshly moigtened. Mixed with a good deal of water. 1°. 0°'5. 1°. 0°-5. 1°. 0°-5. I8. 0°-5. 1°. o"-6. loo-as. K, . 51 103 132 38 50 41 31 9 11 50 58 01 II 13 20 33 44 44 53 M 35 :<7 M «5 43 9 6-5 II 58 «1 — 77 89 — 64 K K... icj.ll.!!. ! ... K!S. «»« K S. 750 Most of the data are based on at least two experiments. The different experiments with any one thermometer in steam were in good agreement. The agreement in the ice experiments was pretty fair when the ice was moistened or was tightly packed ; when, however the ice was not freshly moistened and was only loosely packed the results were very variable. The intervals 1°, 0°'5 were generally only approximate, and to elucidate the exact law of cooling or heating would require a more complete investigation. The data are, however, exact enough for general conclusions. The most striking fact is the extreme slowness of all the platinum thermometers as compared to the two mercury thermometers ; and it should be noted that these two thermometers are not of a type intended to be rapid. To get within 1° of the steady reading, the temperatures of the thermometers had to alter about six times as much in steam as in ice. The times required for these two changes are pretty much alike in all the platinum thermometers, supposing the ice tightly packed, as it usuilly is in the normal experiments. Thus the approach to the steady state is decidedly slower in ice than in steam, and a longer time must be allowed in the former case than in the latter. § 31. Some rough calculations, based on the observations with KI, summarised in Table XI, point to the conclusion that 3*5 minutes' im- mersion in ice and 2 '5 minutes' immersion in steam would suffice to bring the reading within 0°'0005 of its stationary value in the case of this thermometer. K» and K6 are decidedly less rapid, and I should con- clude that ten minutes' immersion in well-packed ice and five minutes' exposure to steam are by no means very excessive allowances in the Platinum Thcrmonutry at Kew Observatory. 35 case of an average platinum thermometer. It has been customary at Kew to allow a fully longer immersion than ten minutes in ice before taking the first reading, and I think that in the case neither of the ice nor the steam point observations has there been any sensible error through excessive hurry. At the sulphur point the attainment of the steady state is a tedious process. Readings taken with the short-stem thermometers KS and K4 within thirty or forty minutes of their first exposure to sulphur vapour had nearly always to be rejected, being conspicuously low. One had in fact to allow about an hour with these thermometers. Even with the long-stem thermometers, whether glass or porcelain, at least forty minutes' exposure to sulphur vapour was found desirable. A good deal depends on the gas supply and on the quality of the Bunsen burner used. In some of the sulphur point experiments at Kew it is open to doubt whether the stationary temperature had been absolutely attained, but I do not think any serious uncertainty was introduced in this way. Impurity in the Ice used in yetting KO, or Variability in the Method of Treating it. § 32. The ice employed at Kew Observatory is supplied in large blocks, which are washed prior to use. As the testing of thermometers in ice is a frequent occurrence, the assistants have no lack of experience, and since a planing machine was introduced some years ago the ice as prepared has been very uniform and finely divided. The ice employed in the platinum thermometry has been taken from the supply used in the ordinary test work, and as the purity of this is frequently checked by observations with Kew standard thermometers no serious impurity need be feared. The check would, I allow, be hardly adequate to settle a question of two or three thousandths of a degree, and no doubt for work of the highest accuracy a more stringent test would be desirable. Strong confirmatory evidence of the uniform purity of the ice is afforded, as will be seen presently, by the small variability in the observed values of the fundamental intervals of the thermometers. The method of preparing and moistening the ice seems to exert a small but sensible influence on the depressed zero readings of mercury thermometers, possibly because the amount of the depression must be influenced by the rate of cooling. There seems, however, to be very little if any effect on the zero reading of a platinum thermometer, so long as the immersion is ample and the temperature of the room is moderate. Error in the Barometer. § 33. Ordinary errors of reading, lag, or improper temperature correction would merely introduce irregularities into the calculated D 2 I >r. ( '. < 'lnvf. / boiling points; l»ut an undetected change of zero in the barometer, or SIM alti-ration in the method of reduction, would simulate a change in K| and K,, the same for all the thermometers. After the first few observations the Royal Society's old double tul>e iiHMvury barometer — repaired some years previously by Messrs. ;ti and Zambra — was set up in the platinum thermometer room, and it has since then l>een used in all the steam and sulphur point o]»ervations. A recent comparison between it and the Observatory standard barometers shows that no certain change of zero has occurred since it was set up. The surface of the mercury in the cistern of the barometer is about 6 inches lower than the surface of the water in the hypsometer, and about 11 inches lower than the surface of the boiling sulphur. No allowance has been made for these differences of level, as they would introduce extremely small and practically constant errors in the two boiling points. The same reduction tables have been in use throughout, and the con- stant factor 1 '0006 has been applied in the reduction to gravity at latitude 45°. Until standard gravity is more exactly defined, and relative values of gravity more carefully determined, the accuracy of the reduction is open to question, hut for the purposes of the present inquiry the uncertainty is quite immaterial. As to the accuracy of the individual barometer observations, I need only say that readings by Mr. Hugo and myself seldom differed by more than O001 inch. The instrument has a tube of 0'6 inch in diameter and responds rapidly to alterations of pressure, so that no serious error need be apprehended from lag. fif in tin1 S"Jj>hnrt £c. § 34. The sulphur made use of in the experiments has all been obtained from 'Messrs. Baird & Tatlock; it has been produced by Chance's process. Though several supplies have been obtained, at different times, no discontinuity has l>een observed in the sulphur points; thus, unless our experience has been unduly favourable, sulphur has at least one great merit as a medium for supplying a fixed high temperature. There are, however, certain disadvantages which have introduced some variability into the conditions of experiment. After an experiment the sulphur solidifies in the tube containing it, and the remelting it on a subsequent occasion not infrequently ends in cracking the tube. On one or two occasions a crack went right round the tube, at a considerable height above the bulb, after the sulphur had been melted and experiments were progressing, and on other occasions the tube was found to have cracked a day or two after the sulphur had cooled. Of one batch of tul>es, of thicker glass than usual, several Hew into pieces after the sulphur had liquefied, and the observer was lucky Platinum Thermo metry at Kc,w Obseri:ati>r;/. 37 in escaping with no damage except to his clothes. In consequence of this defect, we have had to employ about a score of tubes, and these have not been strictly uniform in length, thickness, or diameter. We have also had to employ a considerable number of small asbestos cones, of the pattern recommended by Mr. Heycock and Mr. Neville, and the tightness with which these have fitted the tubes has varied sensibly. Thus the sulphur point observations have been taken under more vari- able conditions than those at the steam point. We have not indeed observed any certain consequences to follow any of the minor changes described above, but more careful experiments would be necessary to justify the conclusion that the variations are immaterial. I am, in fact, inclined to think that variations in the length of the asbestos cones, or in the tightness with which they fit the tube, are sufficiently probable sources of uncertainty to deserve investi- gation. Error in the Formula assumed far the Variation of the Boiling Point of Sulphur with Pressure. § 35. In another direction there is, I think, little doubt that appre- ciable error exists, viz., in the reduction of sulphur point observations to the standard barometric pressure. Professor Callendar and Mr. Griffiths accept a simple linear relation Af = (p - 760) x 0-082 (10), where A£ is the increment in the temperature of sulphur vapour, on the air scale, when the barometer is p mm. instead of 760. This is based on some experiments by Regnault, who examined a wide range of pres- sure, but had so few experimental points that interpolation is very uncertain. A priori, having regard to the corresponding phenomenon in steam, one would hardly anticipate great accuracy from a linear relation, unless restricted to a very limited range ; while in ordinary every-day use we must be prepared to cover at least a range of 30 mm. The Kew experiments, having been taken without any special regard to the atmospheric pressure, naturally cover a fairly wide range, and thus afford a favourable opportunity of testing the accuracy of (10). To this end I have carefully reduced all the sulphur point experiments in a uniform way. Assuming Callendar's formula t-pt = 8[(r. < '. ' where /„ and j>/f are the temperatures, on the gas and platinum - of the sulphur boiling point at standard pressure. II ML-C if (10) be correct, we have, pi answering to the temperature of sulphur vapour when the pressure is p mm., pi, = pi - 0-082 (;> - 760) [1 - 100--S(2/, - 100) - 100-28 x 0-082(^-760)] (12). Supposing, for instance, 8 = 1-5, and tg = 444°'53, then 2>ft = pi - 0-082 (/) - 760) [0-88164 - 0-000012 (p - 760)] . For any ordinary barometric range the term containing p - 760 inside the square bracket is negligible, and so we may take pt, = pt- 0-0723 Q»-760) (13), where the last significant figure is, of course, uncertain. In actual application pt is the observed value of 100(R - Ro)/(Ri - RO) in a sulphur point experiment taken when the reduced barometer read- ing is p, while ptg is the value calculated from this experiment for the platinum temperature of the boiling point of sulphur at standard pres- sure, on the assumption that (10) is correct. § 36. In calculating pt# I adopted two methods : First I took for RI - RO in each group of experiments with a given thermometer (a >jrnnp including all the experiments made during one year, or during the time one particular calibration was in vogue) the mean of the observed values of the fundamental interval during the epoch in question. Next I used for each sulphur experiment the individual value of RI - RO found from the corresponding ice and steam point observations. These observations were usually taken on the same day as the sulphur point observation, or on the previous day. I thus obtained two series of values of plg. In the first, one principal source of uncertainty was largely eliminated ; while, in the second, another principal source of uncertainty was largely reduced. The following table shows the results for the longest epoch con- sidered, viz., from September, 1897, to January, 1899, inclusive, cover- ing the time when Calibration III was in vogue. During a set of sulphur experiments with the different thermometers the barometric height of course varied somewhat, thus the pressures given in the first column of the table are only approximate. The break in the table separates the cases where the reduced barometric height was al>ove 760 mm. from those in which it was below. The results in the last two lines are based on the exact barometric readings observed during the experiments on the individual thermometers : — Platinum Tliermomctry at Kew Observatory. 39 •d «2 4- CO CO CO 2 bi'" rHufuSUS^^US 1 CO Iw 1^ CO O 11 g rM (M • • "3 T r* r-« ,_, a _J 3 g •^TflrHOi-l-*OiCO US^>rHO05 CO CM o o rs "j? •r °t»l>USCDCOCOUS US-f^CDuj CO US rH C ^1-- M (N •" "S &c ~ US c c 'N "*"' -* ^jt>-COcbuSCOCD'* USTflUSUSUS 05 CO rH 10 CO US CO US 00 0 C 0 M •££ (M ' O *3 00 A, 1> "*? co co 3 i o OONUSIMCOCO i^coeorHco {§55 rH O "3 1 S - -2 g " is a <0 S w" o+SSSSS? 1 ? ? IS Si us •* co oo oo o rH O 00 O • • CO US « 9 C § 2 g so w o+og§.goo ^«sg Ol O CO » ^ 00 ID "# O Q C5 CO M K.T rHCOCO'^CQCO''*-!? US^CO-^I^ US Tf! rH O CO US 95 43 'o 88 fi A v 0 h •^ p co co O5 1> co us co oo cc co rH CD t> t^ — CC CO CO CO IO US 1O US ^1 ressure » 1 « 0 ^8 a m P- -° j rS S O O 0 t> *M ^ C c o o ® E c to ni ** &j * £H 5w OH a^ n ^H ^2 * fi« 40 Dr. C. Chree. Investigations on § 37. A similar treatment of the entire number of observations with these four thermometers — viz., twenty-four with K3, twelve with K7, ami twenty-one with both KI and KI — making use of the individual fundamental intervals only, gave the following results : — Table XIII. Pressure. Mean pressures 700 + Mean values of pt,, 421 + KT7 ). A ,. K4. K7. K,. K3. K4. K7. Above 760 Below 67-57 66 54-08 54 13-49 12 n ptt per 1 •66 66 -19 68 '77 06 54-63 53-91 60 11 56 14-86 mm. of pressure •508 •468 •040 0-0030 •612 •485 •127 0-0101 •625 •526 •099 0-0086 • •533 •402 •131 0-0088 Differences .... Thus variation i As we shall see later, the thermometer KI has undergone a change. Its pts has risen, and the apparent smallness in the variation in its case in Table XIII seems largely due to the fact that amongst the latest experiments there was a preponderance of cases where the barometer was below 760 mm. § 38. Tables XII and XIII prove, I think, beyond a doubt that the Callendar-Griffiths reduction formula (10) assigns an appreciably too small value to the variation of the temperature of sulphur vapour with pressure near 760 mm. This is the conclusion to which the results from every single thermometer point, and the phenomena are too con- spicuous to be attributed to accident or to experimental errors. The agreement between the results in Tables XII and XIII is, I admit, not all that might be desired, and any numerical conclusions based on them are perhaps hardly likely to be final. Still I have thought it worth while to see what these conclusions are, proceeding as follows: — Suppose that the Callendar-Griffiths formula (10) is defective in omitting a term in (p - 760)2 and in assigning an incorrect value to the constant multiplier of the term in /> - 760. Then, supposing as before 8 = 1-5, we should replace (13) by pt =^+(0-0723 + x) (p- 760) + »/(/> -760)2 (14), where x and y must be determined from the experiments. As all the observations had been already worked up applying (10), it seemed simplest to take for data the values calculated in this way for ptg, determining x and // by least squares from the observed differences between the calculated values and their mean. Platinum Thermomctry at Kew Observatory. 41 I first took the data in Table XII and added to the individual values for pts in KI, K3, and KT the mean of the differences between the values of pts found with these respective thermometers and the corresponding values in K4. As the corresponding values of pts in these four thermo- meters answered to nearly the same barometric pressure, one got in this way a mean ptg answering very fairly to the mean barometric pressure. Applying formula (14) to the figures so obtained, I obtained by least squares x =+0-0063, y= +0-00018 (15). Next I took all the observations made with K3 and K4 — excluding one when the asbestos cone fell off — and combining the observations made on the same days got twenty -four values of pts as given by the formula (10). Combining these two and two in the order of the corre- sponding barometric pressures, I found from the resulting twelve data by least squares x= +0-0087, y = +0-00021 (16). For ready comparison I show side by side the formulae resulting from (15) and (16), as well as those based on the Callendar -Griffiths value. In these formulae, tg and ptg are, as before, the temperatures on the air and platinum scales of sulphur vapour boiling under the standard pressure 760 mm., while t and pt are the corresponding temperatures when the pressure is not 760, but p mm. It is supposed that S = 1-5 in the formulas for pt. Callendar and Griffiths, {pt = pts + 0'072 (p - 760), from Regnault. l* = ts + Q-082(p - 760). From (15), i.e., by ex-^ periments on KI, K3, | pt = pts + 0-078 (p - 760) + 0-00018 (p - 760)2, K4, and K?, between }» Sept. 1897 and Jan. t = t, + 0'Q88(p -760) + 0-00020 (^-760)2. 1899. » K4 between Oct. 1895 [ f = , +o-092(p -760) + 0-00024 (p-760)2. and Sept. 1899. In restricting the second calculation to K3 and K4 I was guided by the obvious change in KI and by the comparative fewness of the experi- ments at the sulphur point with K>, K5, and K^. According, however, to Table XIII the variation of pts, per 1 mm. pressure, deduced from the experiments on K3, is exceptionally high. Further, the agreement between observed and calculated values in the second calculation was not so good as in the first. I am thus disposed to attach somewhat 4'J I>r. ( '. Clnvr. / on more weight to the result of the first calculation. In any case I have little doubt that the formula * = /, + 0-090 (^-760) + 0-0002 (^-760)2 ......... (17) will prove considerably more exact than the formula (10) now in use. Further direct experiments on the point are, in my opinion, desirable. § 39. In our previous calculations we have employed Callendar and Griffiths' value, 444° -53 C., for the boiling point of sulphur. This is supposed to be measured on the scale of the constant pressure air thermometer. It is about 0°'7 C. lower than the value recently found by Chappuis and Harker, employing the constant volume nitrogen thermometer. We do not as yet know enough about gas thermometry to judge of the consistency of the two results, or of the degree of accuracy either can claim. It is thus desirable to know how the previous results would be influenced by a slight change in the accepted value of /„. Such a change would not affect any formula like (10) con- necting / and p, and it would modify formulae like (13) containing^ only indirectly through change in 8. The change AS in 8, answering to a small change A/g in tf, is given by AS = A/. [104 - 8(2*, - 100)] -=- [/,(*, - 100)]. Using Callendar and Griffiths' value of tg we get approximately, when 8 = 1-5, AS = 0-0575A/,. For instance, if A/x = 0'7° C., then AS = 0'04, approximately. Constancy oi" Variability of Plot i/nim Thermometer*. § 40. Supposing the RQ of a thermometer, determined in the usual way, to show a distinctly lowered value as time proceeds, the cause might be a change in the leads, or in the proportional arms, or in the box coils, and not a real change in the thermometer itself. There are no certain data as to the original equality of the proportional arms ; but as Dr. Harker recently found them to differ by only about three parts in 20,000, and as they were, doubtless, made as nearly as possible equal at first, there is reason to suppose that any uncertainty on this head is small. On the other hand, as shown by Table III, changes in the leads have been large and fluctuating. As no data exist for eliminating this uncertainty in a direct way in the earlier observations, I have based the inquiry into the constancy of the thermometers mainly on an inter- comparison of themselves. As the same coil combinations were used for every one except K and Kj arise from the absence of these thermometers At Sevres and from the changes in the former thermometer during 1896. § 42. Table XV shows that if any relative change has occurred between K3 and KI since 1895 it must be very small. The figures are not, however, absolutely inconsistent with a slight reduction in the difference between the two resistances. There is the clearest evidence of a gradual fall in the RO and RI of KI relative to the corresponding quantities in K3 and K4, the change being greatest in the case of RI. On the other hand, the R* of KI has apparently risen during the last two years relative to the mean R, of K3 and Kj. As the resistance of KI at 0° C. is intermediate between the resistances of K3 and K4, it is impossible to attribute the phenomena to changes in the proportional arms, or in the coils, or to any other external cause. I thus see no alternative to the conclusion that in KI there has been a diminution in the resistance in ice and in the funda- mental interval, accompanied by an increase in the ice-sulphur and steam-sulphur intervals. In its early days KI was several times exposed to the temperature of molten silver, but the last exposure preceded 1897. Table XV shows no other certain example of relative change except in the case of KS when quite new. The first experiment made with this thermometer was on February 12, 1896, by which time the observers had become fairly Platinum Thcrmometry at Keiv Observatory. 45 M 1 ~ 1 w ~~ P 1 1 i 1 01 CO CO « ^^"^^ c^j -i* ^i *.o o ?i c~- ifl iO «O ^ iO *O •** CO CO — o CO CN coco COO i ^^^ 0> Ol 1 1 O oico 1 1 co C5C5 CO f -* 00 JO JO rH : c-> a o o o CO CO CO C SM C-i| CO - 9 CO - !£ CO s rl 6 £ lO O Ol ",0 O rH X S *** !Z *® *'^ — 1 Ol rf r CT. O* G ;IIIIi -r iT^ - Ol JO s O bboo b b 0 b be 50OOOO 0 o o •: ; oo o « Si 1 M •' i i iS i i b 1 1 if 1 1 1 1 1 1 1 1 1 1 S 1 lO 1 1 to 1 tS b b 1 ill! V o | S 0 < JO 1— t bbc 1 '" * ^P > b f -** 1 Ol N -t J* JO O 1 0 0 O o bbc 5 b b (KS+K4). CM JO OS •-* rH - OO OS rH OS ITS I pj ;» co co 01 co 1 Ol Ol Ol Ol CM ^?O1 rf — OO ~ to oo o -r cc 01 £5 rH O O O O O 00 IO p 1 1 1 1 1 1 8 , pill 1 1 1 1 1 1 1 1 II 1 1 1 Mill 1 1 i1 8 s 00 T •• ^ Ol cj OS CO ff o o e I 1 1 Ol j n< 5 3 CO 1 M* S'lO 00 CO Ol OV a OO CO »C ^f CO Ol P? t-jo jo jo joto T « 1 M 1 1 1 | | 1 1 1 1 1 1 SP ° • S S i ?S5! D tO to 1 0,0 oooo O b ' o o c 5 OO o 3 \ £ Mill « 1 I 1 M r! M M 1 1 1 1 II 1 1 1 1 MM 1 1 II 1 1 i 1 1 1 1 1 1 1 1 " Mill 1 1 M 1 1 1 1 1 1 1 1 5 « oo a i CO 0 1 » V 00 T IO i CO 0 H lO ' 5 3 >- i! i|| ?• 1 CO p CO M £> T w bob 00 CO JO t- i i -H O JO O p* 1 (M 0 1 -H -H OOOO O JO CO O J= I -5!^ £2 b to 00 b 0 ?i Cl M »O ** I |S§ 2 rH bb b b bbbb b o 00 9 b 1 b oi oi i bbc OSOl t- -* lO If ??: o oc JOlO i- r o >— co r- «o S Ol Tl< cc CO 00 >bbbbb - -H CM — 1 IO rH 5 'C O Ol Ol rH 5 p p O O p >b o bb b 1 S JC -V JO ii IO > OOOO O t s 8 <=i 0 91 £ O M t- CO 04 C b b i 1 1 1 bbc 1 rH JO C •Ol t— it 0 0 C 5 S > b > b 1 ; to oo 5 O O b to ? o 1 to 10 0 b b b o 0 OOOO 0 O o b b c 50 OOOO 0 o o « >00 0 JO O -t" lO p CO i I b bb b b ?_ lOrH rH O , C£ -^ 0 J; Ol 5 b b b b b j? r b c: co a IO JO if tp JO o bbc cr. os a 8 « ? bbc : S 5 tO > b 3 >O O joss sob s o b i b 1 Mean for 1896 — . Si iJiiti 1 Mean for 1897 1898. Jan nary — Pebt uai March i«_n 11 00 1 1 | 1= >, 2 rt 3 v5 : H» S r- b > cc Mean for 18i»9 t«; I >r. ( '. ( 'hive. / expert. On thi< day observations were taken at the ice, steam, and sulphur points, and on the same day observations were also taken with K3 and K(. Further, both Mr. Hugo and myself observed, each with two sets of coils, at the ice and steam points, and our observations were in excellent agreement. Thus I think the evidence is almost conclusive as to the reality of a fall in the R0 and RI of K5, in consequence pre-um- ably of its first exposure to the sulphur point temperature. Thi> fall was apparently about 0°'l in the case of RO. This is not the only case in which this phenomenon occurred. Thus etween April 24 and July 2, 1896, the R\ of K2 showed a total apparent fall of 0°'735, whilst the corresponding mean apparent fall in K3, K4, and K^, as obtained by interpolation, is only about 0°*015. The mean fall in these last three thermometers is doubtless a nearly correct measure of contemporary change in the leads ; and we thus conclude that two exposures to the sulphur point temperature and two exposures to molten silver had produced a fall of approximately 0°-7 in the zero of K, The thermometer KI had been in molten silver in September, 1 895, before it came to Kew. On its first exposure to molten silver at the Observatory — the exposure being as usual preceded and followed by steam point observations — the fall of Rj was only 0°-08; and on two subsequent occasions the effect, if an)r, was very small. For instance, on the last occasion the apparent fall was only 0°'009. * It was said to bare been heated in September, 1895, at Cambridge, to fully as high a temperature. Platinum Thcrmometry at Kew Observatory. 47 § 44. We have next to consider the constancy of the fundamental intervals. Corresponding ice and steam point observations were almost invariably taken in immediate succession to one another, so that ordinary changes in the leads could hardly affect the results. I have divided the data into groups according to the date, and give only the mean results. In the first group there' were only two observations and in the last group only three, except in the case of KI and K5, where five observations had been made. In the three intermediate groups the means are nearly all based on six observations. A few observations made in the early part of 1897 have been wholly omitted, as the state of the leads was certainly not then satisfactory. Table XVI. — Mean Values of Fundamental Intervals at different Epochs. Calibra- Epoch. tion r,. JLgi K3. K4. x* KG- K7. vised. 99 + 100 + 99 + 99 + 100 + 250 + 99 + 1895 I . . . . •956 •711 . , •246 1896 I •812 . .. •981 •709 •247 •301 •295 Sept., 1897, to 1 Apr., 1898. J III •776 •• •942 •672 • • •157 •265 Apr., 1898, to "1 Jan., 1899. / III •752 •• •945 •671 .. •163 •257 May to Sept., 1899 IV -712 •857 •925 •656 •188 •171 •256 According to the table there is a fall in the fundamental interval of every thermometer between 1896 and 1899. It is probable, however, that this is mainly if not entirely fictitious except in the case of KI. When the resistance box was altered in the summer of 1897, the coils, as Table IV shows, were slightly disturbed ; and in all probability the mean box unit was slightly altered. An increase in this unit of about 35 parts in 100,000 would account pretty satisfactorily for the apparent differences in the results of the second and third epochs of the table. During the third and fourth epochs nothing happened to the box, and Kj is the only case where there is any noteworthy difference "between the corresponding means in the two epochs. Between the fourth and fifth epochs the coil I broke, and its resist- ance was diminished by about 0'66 of a box unit by the resoldering. This fact, and the comparatively small number of observations on which the means are based during the fifth epoch, predispose me to query the apparent falls in the fundamental intervals of K3 and K4. In the case of Kx, however, the fall is too large to be explained away. § 45." Instead of attempting to trace changes in the ice-sulphur 48 Dr. C. Chree. Investigations on + M" 5!??? M" rH rH rH rH . « • § § * « • 1 1 . »H rH P rH P "^ 00 rH O CO -d gj See 8 8 , ffl 4J — '--;.-..- -S iT *9 ^r tO tO 1 i 9 1 i _ _ _ _ •5 £ • ^ CO 4> ^3 O O O ^ CO I 5 M rl tO »O tft N rH t O tO tO tO b 1 M rH rH rH rH '— - £3 0 f3 5 M* + M . 55 rH W 1 •* 3* **N t^ 4 • II* rH rH e? c8 1 + 10 •* 05 00 £ ( 2 oS S» f-f r4" rH CO T* Tf S 1 M rH rH l-l tH 'N-x • rH ^ t 1 I— I ^ "*" i *y«3; o . o -• ao IO O O iil M M • • « 3 uf • • • rH rH rH ^«» _c » OS 1— 1 08 „ «*H •,J • + t» . N «0 3 ? O 0 | W 55 f3 1 1 *? •8 I w" ^ 1 1 '-° o I ' w 5 rH r~l •I £ § OD i 0 4 •M r. -r y -H y 10 IO to 0 i 1 M «P 00 O 30 O5 C5 O O rH rH rH rH s i c TJ 1 s 1 0 OS rH O -H rH t—l "3 M* t^ •* CO •* -H to IO IO to (N • • • • 1 HH JS "d 4 O5 Q O O ^ i25 to to ^ O r-( O rH rH rH rH X 0 HH • 9 ^ M HA 1 M 'J< •f- s9 1 1 ? ' •<* 1 1 \A rH rH Ot-«l- W _^ O tffl rH Q »* lO tO tO -4" rH rH r- rH i m *? tjOO 05 tO C5 O» 05 -. y -f. 'f 00 rH rH rH lH I 00 rH rH rH rH Platinum Thennomctry at Kcw Observatory. 49 interval R* — R0 directly, I pass to the consideration of ptg and £, as quantities of more immediate interest. The constancy of one of these quantities is involved in that of the other ; but each has an interest of its own. I have thought it desirable to give the results obtained from formula (17) as well as those obtained by accepting the Callendar- Griffiths formula (10). The thermometer called K2 in 1896 is wholly different from that so denominated in 1899. With the exception of the 1899 results in K2, and the 1897 results in Kt, every mean in the table depends on at least four observations ; in some instances eight were available. The existence of a gradual rise in the ptg of KI, with corresponding fall in its 8, is placed I think beyond doubt. In the case of the other thermometers there is no fluctuation which might not reasonably be attributed to errors of observation or uncertainty in the pressure reduc- tion formula. Differences between tJie TJiermometers, § 46. A difference between the values of RO or I in two thermometers does not necessarily imply any difference between them as measurers of temperature. The temperature pt is a ratio, and so should depend solely on the law of increase of resistance with temperature. If the drawing of platinum wire affects its physical properties, we may perhaps expect differences between the temperature relations of .wires of different diameters. But there it no obvious reason a priori why differences should exist between specimens of wire of the same thickness from the same sample of platinum. It is clearly desirable to ascertain whether such differences exist, and if they do whether they follow any recognisable law. This is the more important at present owing to the recent proposal made by Prof. Callendar to the British Association to set up standard platinum thermometers made of some one sample of platinum wire. Referring to Tables XVII and XVIII, we see evidence of differences between certain of the thermometers. The consistency of the mean values found in different years for ptg in K3 and K4 renders it im- possible to ascribe to experimental error the very considerable differ- ences between the values of pts in these two thermometers, and in the original KI and K2. § 47. The evidence afforded by Tables XVII and XVIII as to differ- ences between K3, K4, K5, and K; is less conclusive. I have thus had recourse to two more sensitive methods of detecting differences. The first of these consists in taking the ratios borne to one another by the resistances at the ice, steam, and sulphur points. The taking note of the value of Ri/Ro was in fact recommended originally by Mr. Griffiths as a delicate means of checking the accuracy of individual VOL. LXVII. E 50 . / • 'I'lf'ums on observations, and the \r U continued for sometime after the experiments commenced. The value of these original calculations is, however, prejudiced 1>y the fact that the results are affected by any want of equality between the leads, unless this can be eliminated. I have thus included in the following table only those observations in which the difference between the leads was known and could lie allowed for. The thermometers are arranged in descending order of Rj/Ro during the last epoch. Table XIX. Thermo- meter. Mean values of Ri/R0. Mean values of R,,'R». Sept. 21, 1897. (Cal. III). Oct., 1898, to Jan., 1899. (Cal. III). May — Sept., '1899. (Cal. IV). Mav — Sept., 1899. (Cal. IV). K. !• A3 K< K, K- K2 Mt 1 -38771 1 -38751 1 -38707 1 -38709 1 38670 •ana, excluding K 1 -38792 1 -38777 1 -38759 1 -38714 1-38706 1 -88664 1-38787 1 -38774- 1 -38751 1 -38711 1 -38691 1-38664 1-38610 2-63527 2 -63331 2 63165 2-63112 2 -62937 2 -62709 1 -38702 2-63090 The last figure is retained mainly to give an idea of the concordance between the several results for the same thermometer. Without realising this, one is not in a position to judge of the weight to be attached to the apparent differences between the different ther- mometers. It will be observed that KI alone has altered its relative position since September, 1897. Again, it will be noticed that the order of R,/Ro is the same as that of Ri/Ro- But for this fact we should have found larger differences between the values of ptt and 8 in the several thermometers. § 48. The above table will, I think, be considered pretty conclusive as to the reality of the differences between the different thermometers. I shall, however, also give the results of the second method of attacking the problem, as they possess an independent interest. If two platinum thermometers differ only in the absolute resistance of their spirals, the difference may be ascribed to a definite portion of the wire forming the spiral of higher resistance. Thus, in such a case, the difference between the resistances should increase M'ith the temperature in precisely the same way as the total resistance of Platinum TJiermomctry at Kew Observatory. 51 either thermometer. Consequently, if the platinum wire in the several thermometers were identical, the differences between the resistances of any two of them at the steam point and at the sulphur point — or the differences between any one and the mean of any two others — should bear to the difference of the resistances in ice constant ratios. And according to Table XIX the values of these two constant ratios should be approximately 1*387 : 1 and 2*631 : 1. Table XX shows what the ratios actually are in the cases examined. In treating KG I utilised only those observations in which the differ- ence between the leads could be allowed for. The meaning of the results will be rendered clear by an example. The mean of the differ- ences between corresponding observed values of RI in K3 and K4 during 1896 was 0'665, while the mean of the differences between the observed values of RO was O405 : the ratio of these two quantities is 1 '64, and it is given under the heading RI/RQ. This is here used as an abbreviation for (R! in K3 less RI in K4) -7- (Ro in K3 less R0 in K4). Table XX. KS-K*. K! - HK3 + K4) . K2 - i(K3 -f K4) . K5-i(K, + K4). Epoch RI/RQ. E«/BO. R)/RO. RS/RO. RJ/RO. RS/RO- RI/RO. RS/RO- 1896 1'64 3-78 0-89 0-53 1-61 3-65 1897 1-62 3-64 0-68 0-67 . . . . 1-61 • • 1898 1-70 3-97 0-35 2-57 . . .. . . , . 1899 1-68 3-79 -0-29 3-93' 1-30 2-24 1-62 3-79 Epoch. |K.-i(K, + K4). KK3 + K4)-K7 K2-K3. K2— K;. RI/RO. RI/RO- RS/RO- RI/RO- RS/RO . R!/RO. R«/RO- 1898 1-57 • • 1897 1-63 1-58 3-48 . , . . . . , . 1898 1-64 1-59 3-51 t t f f t ^ t e 1899 1-67 1-57 3-50 1-23 1-90 1-36 2-51 In only one case, K2 - K;, are the ratios even approximately equal to the mean values given by Table XIX, viz., 1'39 : 1 and 2 '63 : 1. Even in that case the differences are too big to ascribe to experimental error, because the R0 of K2 exceeds that of K; by fully 4 '5 box units. E 2 52 I >r. ( '. < 'hive. / i on In the case of K3-K4, ami <>f lv{-t-K4), the differences of resistance at the ice point are only about 0*4 of a box unit ; but the ratios are based on a considerable number of harmonious observations, and the smallness in their fluctuations from one year to another appears incompatible with any serious error. The only case in Table XX in which there is unmistakable change in the values of Ri/Ro and R»/Ro with the time is KI - '(K:< + K4). The change in this case is conspicuous, and supports, of course, our previous conclusion that KI has undergone some alteration. § 49. If we omit combinations containing KI or Kj, we notice a curious resemblance between the values of the ratios in the different cases, and this same similarity appears in other combinations of the five thermometers K3, KI, KS, K;, and K«j (when multiplied by 0'4) which I have tried. We always get something pretty close to : — Ice point difference : steam point difference : sulphur point differ- ence :: 1 : 1-6 : 3'6. The phenomena presented by these five thermometers are thus pretty much the same as if their spirals consisted in the main of given equal lengths of identical platinum, with, in addition, variable lengths of a material whose temperature coefficient is higher than that of platinum. In this imaginary material the resistances at the ice, steam, and sulphur points would be approximately as 1 : 1 '6 : 3*6. Another way of stating the fact would be that the changes in the resistance of the imaginary material due to temperature are about 1'56 times as great as in platinum. In addition we can closely reproduce the figures resulting from the comparison of K-2 with the other thermometers by supposing that Kj has a smaller quantity of this imaginary foreign material than the others, but a larger quantity of pure platinum. The figures to which my calculations lead for the resistances at 0°, due to the foreign material and the pure platinum, are as follows, in terms of the mean box unit : — Resistance, Resistance, Thermometer. foreign material. platinum wire. K2 Z Z + 5'150 K7 2 + 0-637 x K4 -s + 1-391 x K3 2+1-783 x K5 2 + 2-223 * Taking, for instance, the cases Ko - K^ and K2 - Ky in 1899, we have, in mean box units — Platinum Thermomctry at Kcw Observatory. 53 Observed differences. Calculated differences. Ice point. Steam point. Sulphur point. Ice point. Steam point. Sulphur point. K,-K5.. Ko - K7 . . 2-927 4-513 3-604 6-124 5 -564 | 11 -306 j (2 -927) (4 -513) 3-586 (6 -124) 5-547 11-257 The numbers enclosed by brackets were used in my original calcula- tion, so their coincidence with the observed values is without signifi- cance. The coincidences could have been improved by supposing that the amounts of pure platinum in KS, K4, KS, and K^ are only nearly and not absolutely the same, and by slightly altering the values 16/10 and 36/10 assumed for the ratios borne by the resistances of the foreign material at the steam and sulphur points to the resistance at the ice point. That the above coincidences are purely accidental seems improbable, but what the real nature of the physical fact underlying the figures is, I do not know. Accuracy of the Observations. § 50. It is difficult to say what is an equitable basis on which to dis- cuss the question. Presumably if the room containing all the appa- ratus were kept at a constant temperature, and a uniform source of current were employed ; and if arrangements were made for taking the bridge centre, for reversing the battery, and for interchanging the leads and proportional arms at every observation, a variety of sources of uncertainty affecting the Kew results would be much reduced. To provide, however, for all possible interchanges must add to the cost of platinum thermometry, and tend to make observations more tedious. I am thus disposed to think that, except in the case of the most refined physical work, where apparatus is abundant and time and expense of relatively small account, the conditions are unlikely to be more favour- able, and will in general be less favourable, than those prevailing during the experiments at Kew Observatory. Under any circumstances a great deal may depend on whether it is, or is not, deemed essential to make a fixed point observation — either in ice or steam — immediately before or after each series of observa- tions. If a fixed point observation be not taken, we may presumably expect uncertainties in the calculated temperatures of the same order as the fluctuations observed at Kew in E0 or RI ; while if a fixed point observation is taken, we may presumably expect, when mea- suring a steady temperature between 0° and 100° C., a similar degree of accuracy to that met with in the determinations of fundamental intervals in the present inquiry. 54 1'r. (,\ Chree. ////'/"//••. Aii idea of the degree of accuracy in the two cases is ail'or1'. ( '. Cliri-r. ////•». •;////'///„/,< OH other thermometers, has to l>e multiplied by 0'4 to render it comparable as an interval of temperature. § 52. Particulars of the degree of accuracy attainable in determining pt, may also be of interest. In this case, of course, the uncertainty existing as to the true law of variation of the boiling point of sulphur with pressure should be lx>rne in mind. Taking all the observations made with K-_>, Ka, K4, Kj, and K7, excluding one with K4 and the first three with K7as probably faulty, and employing the Callendar-Gritiiths formula (10), I find for the average departure of individual values of ptt from their arithmetic mean the following results : — Table XXIV. — Average Departure of Values of ptg from their Arithmetic Mean. 1 K2. K,. K4. K.. K7. Number of observations. . . 6 24 21 10 12 Ayerage departure ±0 '04 ±0-09 dkO'07 ±0*08 ± 0-07 The substitution of formula (17) for the Callendar-Griffiths formula reduces the average departures in the case of K3, K4, and K7 to ± 0'07, ± 0*05, and ± 0'05 respectively, but increases the mean departures in K > and K5. Roughly speaking, the average departure from the mean in the case of ptt is nearly ten times as large as in the case of the fundamental interval. This, however, may partly arise from the fact that sulphur point observations were frequently taken on the day subsequent to that on which was determined the value of RO used in calculating ptt. Improvements suggested in Apparatus. § 53. The experiments have suggested various desiderata, which it may be well to summarise. 1. It is desirable that some simple and more perfect method should be devised for eliminating the effects of changes in the relative resistance of the leads and the proportional arms. 2. The temperature of the resistance coils and bridge wire should be exposed to much less uncertainty than in the Kew box, or else these resistances should be made of some material with a much smaller temperature coefficient than platinum silver. 3. The bridge-wire scale is too much contracted for very exact work with thermometers of such low resistance as the ordinary Kew ones; it is desirable, if accuracy to 0°'001 C. is aimed at, that 1 cm. of bridge wire should not answer to more than 0°*1 C. Platinum Tkcrmometry at Kew Observatory. 57 4. If possible, the causes of wanderings in the bridge centre and of thermo-electric currents — not eliminated by the Griffiths key — should be removed. 5. Unless the plug resistance uncertainties can be reduced, it would seem desirable to increase the resistance inside platinum thermometers intended for very exact measurement of very slowly altering temperatures, such as occur in "fixed point" observations. 6. Unless the heating effect of the current can be exactly ascertained and allowed for, the sensitiveness of the galvanometer should be largely increased, so as to render unnecessary currents whose heating effects will sensibly influence the readings. 7. It is certainly desirable that the entire room containing the appa- ratus should be kept approximately at the temperature accepted as the standard temperature for the coils. At the same time it is not desirable that the observer should be continually exposed to so high a temperature as 20° C. A temperature such as 62° F. is very much healthier, at least in winter, and is more bracing. There is no obvious reason why there might not be one standard temperature for winter and another for summer, so long as the temperature relations of the resistance coils are accurately known. Further Experiments Wanted. § 54. I should like to see experiments made on the following points : — 1. The true law of the variation of the boiling point of sulphur with pressure ; 2. The behaviour of new platinum thermometers under a variety of conditions ; 3. The effect of long or frequent exposures to temperatures extending from that of sulphur vapour to the highest temperature for which it is claimed that platinum thermometers are suitable ; 4. The effect of exposure to low temperatures, such as are now attain- able by means of liquid air ; 5. The degree of accordance in the results deduced by Callendar's formula from observations taken with different platinum thermometers — varying in their 8 — at two or three fixed temperatures differing considerably from that of sulphur vapour, these temperatures being obtainable with a degree of accuracy not inferior to that attained in sulphur point observations ; 6. The cause of the curious phenomenon described in §§ 48 and 49. § 55. The great majority of the observations on which this paper Mr. A. K. Tutt»n. A >:/<,jr»j>/< /'<•<>/ depends ha\e been made by Mr. W. Hugo, senior assistant at I\c\\ Observatory, to whose care and exactitude the research owes a very great deal. The reduction of the observations is also mainly due to Mr. Hugo. In several of the special experiments Dr. J. A Harker haa given valuable assistance, and various operations have been much facilitated by means of apparatus devised by him. The main work of checking the reductions and the various calculations requisite in discussing the observations have been done by myself. I have also taken a considerable number of observations, and in discussing the various sources of trouble I speak from personal experience, supple- mented of course by frequent exchange of ideas with Dr. Harker and Mr. Hugo. At various stages of the investigation I have had valuable advice and assistance from Professor Carey Foster and Mr. W. N. Shaw, to whom the initiation of the research is largely due. The original cost of the apparatus was partly defrayed by a grant of £100 from the Government Grant Fund, and a contribution of £120 was received last year from the Gunning Fund towards this and other cognate work undertaken by the Kew Observatory Committee. " A Comparative Crystallographical Study of the Double Selenates of the Series E2M(Se04)2,6H20.— Salts in which M is Zinc." By A. E. TUTTON, B.Sc^F.RS. Received March 5,— Eead March 15, 1900. (Extended Abstract.) In two communications to the Chemical Society,* the author presented the results of a detailed study of twenty-two salts of the series of monoclinic double sulphates R2M(S04)o,6H.,0, in which R was represented by potassium, rubidium, and caesium, and M by magnesium, zinc, iron, manganese, nickel, cobalt, copper, and cadmium. The first of these memoirs dealt with the external morphology of the crystals, and the second with their internal physical properties. The present investigation refers to the less-known analogous double selenates, in which R is again represented by the alkali metals potassium (atomic weight 39), rubidium (atomic weight 85 -2), and caesium (atomic weight 132'7). The work on the group containing zinc has been com- pleted, and the results are now communicated. Topsoe and Christian- sent included in their well-known investigation the potassium salt of the group. * ' Journ. Chem. Soc. Trans.,' 1893, vol. 63, p. 337, and 1896, vol. 69, p. 344. t ' Ann. de Chim. et de Phys.,' 1874, p. 5. FIG. 1. Study of the Double Seknates of the Series E2M(SeO4)2,6H;,0. 59 The rubidium and csesium salts have never been investigated, and the author has been unable to find any evidence of their preparation. The section-plates and prisms employed in the optical work were prepared by means of the author's cutting and grinding goniometer, in the manner already described in previous memoirs. The new special adjusting apparatus described to the Royal Society in the memoir in the ' Phil. Trans.,' A, for 1899, p. 461, has proved of the greatest service in the preparation of the prisms. All the prisms were obtained with its aid by one adjustment, instead of two (one for each of the two required surfaces). A very useful addition to it has been made in the shape of a pair of special grip-holders, besides those referred to on page 460 of the memoir just quoted. One of these new holders is represented in fig. 1. The upper portion, the stem by which it is attached to the lower end of the crystal-adjusting apparatus, carries two vertical grooves instead of only one, so that the plane within the crystal which is desired to bisect the angle (60°) of the prism can always be preliminarily roughly arranged at right angles to the 120° adjusting segment, whatever be the situation of the most conve- nient direction of gripping. The stem passes down into a thicker solid cylindrical portion, surrounded by a closely fitting hollow cylinder capable of movement for somewhat over 90°, and fixation at any position, by means of slits at different levels and two clamping screws passing through them into the solid cylinder. This enables the necessary final azimuth adjust- ment of the crystal to be effected, so that the particular bisecting plane referred to can be set exactly to the required orientation, with the aid of the goniometer and its graduated adjusting movements. The gripper is attached rigidly to the underside of the movable •cylinder. Its two prongs are arranged to be drawn together by means •of a screw manipulated with a milled-headed key, and the gripping lower part is thickened and padded with chamois leather. The use of one of these holders, which only differ in the size of the gripper so as to accommodate different sized crystals, avoids the use of warm wax, which may cause efflorence, or cracking, of the crystal. Every prism employed was prepared by the aid of this instrument so that the two surfaces were symmetrical to one of the three principal planes of the optical ellipsoid (indicatrix), and its edge parallel to one of the two rectangular axes of the indicatrix lying in this plane. It therefore yielded two of the three refractive indices directly, namely, those corresponding to vibrations parallel to the two rectangular axes •of the optical indicatrix lying in the bisecting plane. The salts were prepared in the following manner : — A quantity of Mr. A. K. Tiittnn. A C»mparat< • Vr«//wy ,//«:«/ lk;ilino selenate crystals, prepared as described in the author's memoir on the normal alkaline selenates,* adequate to yield after addition to the calculated amount of the M-selenate sufficient of the double salt for all the purposes of the investigation, was dissolved in the minimum of distilled water. An equal molecular proportion of pure zinc sulphate crystals was also weighed out, dissolved in dis- tilled water, and precipitated by a solution of sodium carbonate. The precipitate of basic zinc carbonate was isolated by decantation and prolonged washing with hot water, and subsequently dissolved in a solution of pure selenic acid, whose strength had previously been accurately determined by titration with a solution of pure anhydrous sodium carbonate of known strength. Sufficient of the acid was run in upon the zinc carbonate to provide for one drop of excess, as this prevents the possibility of a basic salt forming after addition of the selenate of the R-metal. The solutions of the two selenates were mixed and allowed to crystallise spontaneously, taking all the pre- cautions to avoid disturbance enumerated in the previous memoirs. The spherical projection given on p. 343 of the double sulphate memoir applies equally to the double selenates. POTASSIUM ZINC SELENATE, K.,Zn(SeO4)1.,6H.jO. A determination of the zinc in a quantity of the crystals employed gave the following result: — 0-9152 gramme yielded 0*1396 gramme ZnO, corresponding to 12*24 per cent, of zinc. Calculated per cent. Zn = 12-11. Goniometnj. Ten excellent crystals were employed, derived from several different crops. Habit : Short prismatic to tabular. Axial angle— ft = 75° 48'. Ratio of axes — a : b : c = 0-7458 : 1 : 0-5073. Forms observed — a = {lOOJooPco; b = {OlOjco^oo; c = {001}0P; p = {110}ooP; p' = {120}oo^2; y = {OllJ^oo; <>' = (Ill} + P;r' = {201} + 2Poo. The results of the angular measurements are given in the accom- panying table. The habits observed resembled those exhibited in the following figures given in the double-sulphate memoir : — Fig. 7 without b faces, fig. 8 without the smaller faces, fig. 20, fig. 21, fig. 24, and fig. 25. * ' Journ. Cheui. Soc.,' 1897, p. 846. Study of the Double Schnates of the Series K.,M(Se04)i,,6H30. 61 Morphological Angles of Potassium Zinc Selenate. Angle measured. K'o. of measure- ments. Limits. Mean observed. Calculated. Diff. foe = 100 : 001 = ft 5 0 / ° . ' 75 43— 75 55 75 45 0 / 75 48 3 as = 100 : 101 — — — 46 19 — so = 101 : 001 — — — 29 29 — J cr' = 001 : 201 16 63 4— 63 25 63 15 63 12 3 ' cs' = 001 : 101 — — 38 22 s'r = 101 : 201 — — — 24 50 — r'a = 201 : 100 5 40 50— 41 0 40 56 41 0 4 [_r'c' = 201 : 001 16 116 35—116 56 116 45 116 48 3 Cap = 100 : 110 8 35 33— 36 8 35 52 * j>p' = 110 : 120 2 19 1— 19 45 19 23 19 28 5 I p'b = 120 : 010 1 — 34 39 34 40 1 1 pi = 110 : 010 10 53 51— 54 24 54 7 54 8 1 j pp = 110 : 110 20 71 34— 71 54 71 43 71 44 1 \^pp = 110 : 110 20 108 6—108 32 108 16 108 16 0 [ cq = 001 : Oil 36 26 0— 26 23 26 8 # 1 \ qb = Oil : 010 12 ' 63 35— 64 8 63 52 63 52 0 [ qq = Oil : Oil 13 127 38—127 53 127 44 127 44 0 f ao = 100 : 111 ^ _. _ 49 35 | .->q = 111 : Oil — — — 27 42 — -j aq = 100 : Oil — — — 77 17 — qo' = Oil : 111 — — — 34 31 — (jt'a = 111 : 100 — — — 68 12 — f co = 001 : 111 — — — 35 11 op = 111 : 110 — — — 43 21 — j cp = 001 : 110 42 78 22— 78 41 78 32 * — | po' = 110 : 111 3 56 42— 56 53 56 48 56 50 2 o'c = 111 : 001 3 44 37— 44 42 44 39 44 38 1 \^pc = 110 : 001 42 101 21—101 38 101 29 101 28 1 fbo = 010 : 111 _ _ 69 51 tt/i'i'}>}iical The " ;ni(l l> faces were usually both well defined; the hcmi-pyramid 0' was occasionally \\cli drvdnprd, l»ut more usually small, while small p' faces were only observed on two crystals. The general type exhi- bited p,r,r and q as the main faces, r always large, and ?•' occasionally large but generally rather smellier than the q faces. Topsoe and Christiansen (loc. cit., p. 77) give for the ratio of the and the axial angle, a : b : c = 0'7441 : 1 : 0*5075, and ft = 7~> 46'. The measurements were made by Topsoe in the year 1870. There is an excellent cleavage parallel to the faces of r'{201}, as stated by Topsoe and Christiansen. Volume. Relative Density. — The following four independent determinations were made : — Weight of salt Sp. gr. at employed. 2074°. 5-5358 2-5535 6-1058 . 2-5544 5-1597 2-5524 5-0657 2-5546 Mean 2-5537 Molecular Volume.—^- = -fff^ = 210'13. a 2'5537 Distance Ratios. — Combination of the axial ratios and axial angle previously given, with the molecular volume, affords the following distance ratios : — X : ^ : (o = 6-1941 : 8'3054 : 4'2133. Optics. Orientation of Axes of Optical Ellipsoid. — The plane of the optic axes (optic binomials) is the plane of symmetry. The sign of the double refraction is positive. Two section-plates ground parallel to the symmetry plane afforded the following extinction angles, relative to the normal to the basal plane : — Section 1 5° 30' 2... 4 36 Mean 5° 3' The direction is behind the normal, nearer to the vertical axis. This axis of the optical ellipsoid (indicatrix) is the second median Study of the Double Sclcnates of the Series EoM^SeOJo.BILO. 63 line. The first median line is consequently situated in the obtuse angle of the morphological axes ac, and is inclined 5° 3' to the axis a. The second median line lies also in the obtuse angle af, and is inclined 9° 9' to the vertical axis c. Refractive Indices. — The results of the determinations with six prisms, ground on six different crystals, are given in the accompanying table. The values obtained by Topsoe and Christiansen are appended in the last column. The (3 values, which were alone determined directly by them, are observed to agree well with the authors' values. The intermediate refractive index of potassium zinc selenate, cor- rected to a vacuum, is accurately expressed as far as the neighbourhood of F, for any wave-length A, by the following formula : — B = 1 -5010 + 694 10° - 3 °°5 70° 00° 00° + X2 A4 As the dispersion increases with the numerical value of the index, the a and y indices are not precisely reproduced by diminishing and increasing the constant 1-5010 by fixed amounts ; but they are on the average less and greater respectively than the (3 values by 0-0060 and 0-0154. Alteration of Refraction lij Rise of Temperature. — Determinations carried out at 60° indicated that the indices are diminished by about 0'0020 for 45° rise of temperature. Axes of the Optical Ellipsoid. — The calculated values of the axial ' ratios of the two optical ellipsoids are as follows : — Axes of optical indicatrix : a: /3:y = 0'9960 : 1 : 1-0101. Axes of optical velocity ellipsoid : a : fa : t = 1-0040 : 1 : 0'9900. Mr. A. K. Tiitton. A Co// CV //*/'//•' I ^ 3 -1 I *«* 4 3 — l~ iO O I ri rt t^ -N r. — -C -H — IM *I t^ l~ *> O C5 C> -H — — .7 f — •'. z r. i --. - — i - '• TI :i r; ?i :c ?7 ~ — -* IO IO 10 IO MB Mi o >o o m t^ *« 17 c ~ it / • ^> •* t» C -*• O o _ r: "C c — ^^^5jo^w M m 09 00 9 ? o o 1.0 o o irs us ua ift »a >j5 us >-7 ^f ** ^* t^» I*? * ?1 l^» l>» * •>• • x w r. T r: -^ c — C — . M I I I I I I 0^«CO" >_-: >7 ip o o I I I I I I •- O^»*l^-~- rrtjc"— rtxn^ — _ i_t 1.7 IS it •-. x o^-oo M — L-5i^— I1' ^Hi— i-lM IS O O 7 ' : — ~ >O C. « t^- t^. » 1 ».O Cl L7 l.t ^ — . 71 CO — O— i-I^M — i-l — riMCO CS >O lO «.O O it "tit it it it Study of the Double Seienates of the Series R3M(SeOt)2,6H30. 65 Molecular Optical Constants. — Following are the values of these con- stants : — Axis of optical indicatrix. a. 0. 7- 2 i r Q Specific refraction, ., - ( — = n ] n 0-1170 0-1199 62-76 0 -1181 0 -1211 63-37 0 -1210 0-1242 64-93 64-33 0-0029 64-98 0-0030 66-65 0-0032 1-57 1-61 1-72 Molecular refraction M • • C 107-00 108-24 111 -41 d Optic Axial Angle. — Three excellent pairs of section-plates were ground, perpendicular to the first and second median lines respectively. Determination of Apparent Angle in Air of Potassium Zinc Selenate. Light. Section 1. Section 2. Section 3. Mean 2E. Li o / 112 12 0 / 111 34 Ill 56 Ill 54 C 112 14 111 40 112 0 111 58 Na 112 24 112 8 112 19 112 17 Tl 112 40 112 35 112 41 112 39 P 112 58 113 1 113 10 113 3 Topsoe and Christiansen give for the angle in air 111° 50', and for the true angle 66° 8', both being measured in sodium light. The dispersion is observed to be extremely small, and the most accurate measurement is required to determine it, employing sections which afford very small rings and sharp brushes. A valuable confirma- tion of the nature of the dispersion is afforded by immersion in cedar oil, whose refraction is almost exactly the same as that of the crystals. A section perpendicular to the first median line shows, in cedar oil, brushes separated at their true angle and frinjed with colour accord- ing to the following scheme : — red | blue 1st M.L. no colour | no colour. The obtuse morphological axial angle ac is assumed to be to the left. The optic axial angle is therefore greater for blue than for red, and measurements in Li and F light gave an angle larger in the latter case by about 5', thus confirming the accuracy of the values given in the table. VOL. LXVII. F 66 Mi. A. !•!. Ttitton. A O/////>"/§"///v CryttaUographieal Determination of True Optic Axial Angle of Potassium Zinc Selenate. No. of No. of Light. section JKTpfll'li- culur 1st median Observed values of 2Ho. section perpendi- cular 2nd median Observed values of 2Ho. Calculated values of 2Va. Mean value of 2Va. line. line. 1 o / 60 7 i. 0 / 100 12 o / 66 16 1 2 59 54 2a 100 8 66 8 i. 66 1 L' 3 60 3 3a 100 14 66 13 \ r 1 60 5 la 100 8 66 16 1 2 59 52 2a 100 4 66 8 U6 13 1 3 60 0 3a 100 6 66 14 i Na 4 1 2 3 59 54 59 43 59 49 la 2a 3a 99 42 99 39 99 39 66 18 66 10 66 16 1 66 15 : r 1 59 39 la 99 5 66 20 I Tl 4 2 59 31 2a 99 3 fifi 14 Us 17 ! I 3 59 36 3a 99 5 66 18 J -{ 1 2 3 59 17 59 11 59 19 la 2a 3a 93 17 98 15 98 25 66 22 66 18 66 20 [•66 20 Dispersion of the Median Lines. — In cedar oil no appreciable move- ment of the right brush, that corresponding to the optic axis situated in the acute morphological angle ac, was detected on altering the wave-length of the light, the total movement occurring at the left brush. Hence the first median line lies nearer to the morphological axis a for red than for blue. The amount does not exceed 5'. Effect of Rise of Temperature on the Optic Axial Angle. — Measurements at 65° indicated that 2E increases 2£° for 50° rise of temperature. RUBIDIUM ZINC SELENATE, Rb2Zn(Se04)2,6H20. An estimation of zinc in a specimen of the crystals employed afforded the following guarantee of purity: 1-1130 grammes yielded 0'1426 gramme ZnO, which corresponds to 10'28 percent, of zinc. Cal- culated Zn = 10-33. Goniometry. Ten suitable small crystals were employed, belonging to four differ- ent crops. Habit : thick tabular, sometimes prismatic. Study of the Double Selenates of the Series K.,M(Se04)2.6H30. 67 Morphological Angles of Rubidium Zinc Selenate. Angle measured. No. of measure menti. Limits. Mean observed. Calculatec Diff. (~ac <= 100 : 001 = /S 0 / 0 / 0 / 74 44 i as = ]00: 101 . — — — 45 47 — sc = 101 : 001 — — — 28 57 — . cr' = 001 : 201 9 63 45— 64 2 63 53 63 42 11 | cs' = 001 : 101 — — — 38 24 s'r' = 101 : 201 — — — 25 18 — r'a = 201 : 100 — — — 41 34 — \,r'c = 201 :001 11 115 57-116 15 116 4 116 18 14 Cap •= 100 : 110 — — — 35 38 [ pp' = 110 : 120 — — — 19 28 — \ ,p'b = 120 : 010 — — — 34 54 — ^ pb = 110 : 010 2 54 15— 54 27 54 21 54 22 1 pp = 110 : 110 20 71 6— 71 39 71 15 * — \jpp = 110 : 110 20 108 21—109 1 108 45 108 45 0 [cq = 001 :011 21 25 35— 26 7 25 50 * \\qb = Oil : 010 2 64 10— 64 16 64 13 64 10 3 1 Iqq = Oil :01I 17 128 3—128 26 128 25 12S 20 5 ! fao = 100 : 111 — — 48 59 ! | oq == 111 : Oil — — — 27 19 — \\aq = 100 : Oil — — — 76 18 — qo' = Oil : 111 — — — 34 37 \j)'a = 111 : 100 — — — 69 5 — C co = 001 : 111 . — 34 34 I op = 111 : 110 — — — 43 5 \\cp = 001 : 110 35 77 28— 77 57 77 39 * 1 po' = 110 : 111 3 57 33— 57 37 57 35 57 43 8 o'c = 111 : 001 3 44 40— 44 47 44 45 44 38 7 [_pc = 110 : 001 34 102 10—102 36 102 21 102 21 0 fbo = 010 : 111 — — 70 13 _ 1 os = 111 : 101 — — — 19 47 — \ bo' = 010 : 111 3 65 9— 65 17 65 14 65 14 0 •{ oV = 111 : 101 — — 24 46 [o'of = 111 : III 1 — 49 34 49 32 2 f *y = 101 : Oil 38 2 I qa' = Oil : 121 I — 35 46 35 33 13 ! n'p = 121 : 110 1 — 50 44 50 57 13 j qp = Oil : 110 26 86 14— 86 49 86 37 86 30 7 pS = 110 : 101 — — _^_ 55 28 \jpq = 110 : Oil 26 93 6— 93 36 93 23 93 30 7 {s'q = 101 : Oil _ 45 8 qp = Oil : 110 28 63 26— 63 55 63 38 63 30 8 ps' = 110 : 101 — — — 71 22 — pq = 110 : Oil 28 116 6—116 38 116 23 116 3D 7 f>V = 201 : 111 3 34 51— 34 53 34 52 34 49 3 I o'p = 111 : 110 3 92 44 — 92 47 92 45 92 39 6 ] pr' = 110 : 201 29 52 12— 52 32 52 21 52 32 11 (.r'p = 201 : 110 28 127 23—127 50 127 39 127 28 11 Total number of measurements, 354. 68 Mr. A. I-!. Tutt'in. A Comparative Orystattograph Axial angle : /3 = 74° 44'. Katio of axes : n : I : < • = 0'7431 : 1 : 0-5019. Forms observed: a = {100}ooPco; b = {010}co£oo; e = {001 p = {110}ooP; faces as a rule. Only a trace of the orthopinacoid a \VH«; discovered ; the clinopinacoid b was invariably small when present, and the hemi-pyramid n' was only present on one of the crystals measured. The d faces were fairly well developed, and afforded good images of the signal. Frequently the only faces present were c, p, /, and q. The c faces varied in relative importance from the breadth usually exhibited in the potassium salts to the narrow strip charac- teristic of the caesium salt. There is an excellent cleavage parallel to the faces of the orthodome r'{201}. Density. — The following four determinations were made with independent quantities of the finely-powdered crystals : — Weight of Sp. gr. salt employed. at Mr/49. 7-8234 2-8596 6-0506 2-8611 7-8824 2-8601 6-7490 2-8608 Mean ...... 2*8604 Molecular Volume.— = = 219-90. Ratios. — On combining the axial angle and the ratio of the axes already given, with the molecular volume, the following ratios are obtained : — X : ^ : <•> =6-3062 : 8-4863 : 4'2593. Optics. /ifulion of Axes of Optical A7////.W'/. — The optic axes (optic bi- normals) lie in the plane of symmetry. The sign of the double refrac- tion is positive. The following extinction angles were exhibited by two section-plates Study of the Double Scknates of the Series E2M(Se0.1)i,,6H.,0. 69 ground parallel to the symmetry plane, with reference to the normal to the basal plane : — Section 1 2° 0' Section 2 .. 27 Mean 2 3 The direction is behind the normal, towards the vertical morpho- logical axis. This axis of the optical indicatrix is again the second median line. The first median line thus lies in the obtuse morphological axial angle ac, and is inclined 2° 3' to the axis a. The second median line lies also in the obtuse angle ac, and is inclined 13° 13' to the vertical axis c. Refractive Indices. — Six prisms, ground on six different crystals belonging to various crops, were employed in the determinations. The results are given in the accompanying table : — Kefractive Indices of Eubidium Zinc Selenate. Index. Light. Prism 1. Prism 2. Prism 3. Prism 4. Prism 5. Prism 6. Mean. fLi 1 -5123 _ 1 -5132 1 -5135 1 -5126 1 -5129 8 I C 1-5128 — — 1 -5136 1 -5141 1 -5132 1 -5134 Vibr. 1 Na 1 -5157 — — 1-5165 1 -5167 1 -5159 1 -5162 par. { Tl 1 -5188 — — 1 -5197 1 -5201 1-5190 1 -5194 2M.L. F 1 -5228 — — 1 -5234 1 -5240 1 -5229 1 -5233 LO 1 -5284 — — 1 -5288 1 -5295 1 '5284 1 -5288 a fLi 1 -5187 1 -5175 1 -5198 1-5190 — 1 -5188 P TT4V,» 1 c 1 -5192 1 -5181 1 -5201 — 1-5197 — 1-5193 vior. j Na 1 -5220 1 -5210 1 -5230 1 -5227 1 -5222 par. 1 Tl 1-5251 1 -5241 1 -5261 1 -5258 1 -5253 symn. F 1 -5291 1 -5280 1 -5302 1 -5297 1 -5293 axis. u 1-5350 1-5338 1 -5360 — 1 -5354 — 1 -5351 fLi 1 -5286 1-5299 1 -5298 1 -5294 1 -5294 7 0 — 1 -5290 1 -5304 1 -5303 — 1-5298 1 -5299 Vibr. j Na — 1 -5323 1-5336 1 -5335 — 1 -5330 1 -5331 par. 1 Tl — 1 -5355 1 -5369 1 -5370 — 1 -5367 1-5365 1M.L. 1 F — 1 -5396 1-5408 1 -5411 — 1 -5405 1 -5405 u — 1-5457 1 -5469 1 -5472 — 1 -5466 1-5466 The intermediate refractive index of rubidium zinc selenate, cor- rected to a vacuum, is expressed as far as F, for any wave-length X, by the formula : — B = 1-5067 + 592 314-1 397 60° 00° 000 + The a indices are very closely reproduced by the formula if the con- 70 Mr. A. E. Tutton. A Com j>" rut ire Ory&afloffraphical stant 1-5067 is diminished by 0*0060; owing to apprecial.lv increased dispersion in the y direction, the y indices are not so accurately repro- duced, but are approximately given if the constant is iiuTcasud by 0-0109. .Ill, ft> /inn <>f l;74 55 The dispersion is thus seen to be small, but it is distinctly greater than in the case of the potassium salt. In cedar oil, whose refraction is almost exactly the same as that of this salt, a section perpendicular to the first median line exhibits the brushes fringed with colour as follows : — blue | red 1st M.L. red | blue The optic axial angle is therefore greater for red than for blue, thus confirming the results of the measurements. Dispersion of the Median Lines. — In the foregoing representation of the coloured brush-fringes as seen in cedar oil, the obtuse angle of the morphological axes ac is situated to the left. The hyperbolic brush on this side was more faintly tinted at the edges than the right-hand brush, indicating feebler dispersion of the optic axis to the left. Measurements in cedar oil confirmed this, and showed that the differ- ence between the position of this optic axis for C-light and for F-light was 8 — 10', whilst for the right-hand axis it was 20'. Hence the median lines are dispersed so that the first median line lies nearer to the morphological axis a by 4 — 5' for red light than it does for blue. Effect of Rise of Temperature on the Optic Axial Angle. — Very little change is introduced in the optic axial angle of this salt by variation 72 Mr. A. E. Tutt«»n. A Comparative OryftaUographical of temperature. Measurements in succession at 10'5° and 60'5° indi- cated that 2E increases 30' for 50° of rise of temperature. C/ESIUM ZINC SELENATE, The analysis of a specimen of the crystals of this salt afforded the following numbers : 0*7027 gramme gave 0'0790 gramme of ZnO, which corresponds to 9 -02 per cent, of zinc. The calculated percentage is 8-98. Goniometry. Eleven of the most suitable small crystals, derived from five dif- ferent crops, were employed in the goniometrical measurements. Habit : flattened prismatic. Axial angle : (3 = 73° 49'. Katio of axes : a : b : c = 0'7314 : 1 : 0'4971. Forms observed: b = {010}oo^oo; c = {001 }oP; p = {110}ooP; q = {011}?oo; 7?i = {021}2?oo; o' = {Ill}+P; r' = The results of the measurements are presented in the accompanying table of angles. The habits of the crystals of the various crops are typified by the following figures of the double sulphate memoir (loc. cit.) : figs. 9 and 10 in the description of caesium zinc sulphate, fig. 16 in the descrip- tion of caesium ferrous sulphate, fig. 21, but with much larger q faces, and fig. 34 in the description of caesium cadmium sulphate as far as regards the relation of the c, r and q faces. They are characterised by large q faces relatively to the faces of the basal plane, which latter is usually only represented by a mere strip. The faces of the hemi- pyramid o' are often considerably developed. The clinopinacoid b is frequently present, but the orthopinacoid a was never observed. The clinodome m was found developed on one of the crystals measured. The cleavage parallel to /{201}, common to the series, was well developed. Study of the Double Selenates of the Series E2M(Se04)2,6HoO. 73 Morphological Angles of Caesium Zinc Selenate. Angle measured. No. of measure- ments. Limits. Mean observed. Calculated Diff. (ac = 100 : 001 = 0 0 / 0 / o / 73 49 _: as = 100 : 101 — — — 45 3 — sc = 101 : 001 — — — 28 46 — j cr' = 001 : 201 22 64 36— 64 53 64 45 64 36 9 1 cs' = 001 : 101 — — — 38 52 — s'r' = 101 : 201 — — — 25 44 — r'a = 201 : 100 — — — 41 35 — (^0 = 201 : 001 19 115 7—115 32 115 16 115 24 8 ( ap = 100 : 110 — 35 4 — pp1 = 110 : 120 — — — 19 28 — j p'b = 120 : 010 — — — 35 28 — | pb = 110 : 010 6 54 46— 55 4 54 54 54 56 2 pp = 110 : 110 19 69 55— 70 21 70 7 * (.pp = 110 : 110 20 109 25—110 16 109 52 109 53 1 fcq = 001 :011 30 25 22— 25 42 25 31 * — \qb = Oil : 010 4 64 18— 64 43 64 33 64 29 4 [qy = Oil : Oil 14 128 40—129 12 129 0 128 58 2 (ao = 100 : 111 48 12 — | oq = 111 : Oil — — — . 27 14 •{ aq = 100 : Oil — — 75 26 — qo' = Oil : 111 — — — 35 6 (jo'a = 111 : 100 — — — 69 28 — fco = 001 : 111 34 13 op = 111 : 110 — — — 42 36 — j cp = 001 : 110 37 76 34— 76 57 76 49 • — I po' = 110 : 111 16 57 52— 58 34 58 10 58 14 4 o'c = 111 : 001 17 44 35— 45 21 45 0 44 57 3 (_pc = 110 : 001 36 102 59—103 39 103 11 103 11 0 f bo = 010 : 111 — _ 70 37 — 1 os' = 111 : 101 — — — 19 23 — f bo' = 010 : 111 65 22 _^_. 1 o's' = 111 : 101 — — — 24 38 — C*q = 101 : Oil — _ 37 43 J qp = Oil : 110 36 87 24— 88 3 87 39 87 37 2 ] ps = 110 : 101 — — 54 40 — Ipq = TlO : Oil 36 91 52— 92 36 92 21 92 23 2 fs'q = 101 : Oil — 45 21 J qp =-• Oil : 110 35 62 50— 63 25 63 7 63 5 2 ] ps' = 110 : 101 — — 71 34 — Ipq = 110 : Oil 35 116 35—117 11 116 53 116 55 2 Cr'o' = 201 : 111 11 34 42— 35 15 34 58 35 2 4 | o'm = 111 : 021 1 36 56 36 54 2 j mp = 021 : 110 1 — 55 47 55 49 2 1 o'p = 111 : 110 12 92 32— 92 58 92 48 92 43 5 | pr' = 110 : 201 29 52 2— 52 28 52 12 52 15 3 (r'p = 201 : 110 28 127 25—128 2 127 48 127 45 3 Total number of measurements, 464. 74 M:. A. E. Tutton. A (.'• CryttdOograpMeal Volume. Relative Density. — Four determinations with independent material afforded the following results : — Weight of Sp. gr. saltempltm-.l. at 20y/4°. 5-2958 3-1148 5-5387 3-1175 5-1525 3-1126 5-5317 3-1164 Mean 3-1153 Molecular Volume.— ^ = 72*'° = 232-40. Distance Ratios. — Combination of this molecular volume with the axial angle and axial ratios already given, affords the following dis- tance ratios : — X : ^ : w = 6-3860 : 8'7311 : 4-3402. Optics. Orientation of Axes of Optical Ellipsoid. — The plane of the optic axes (bi-normals) is again the plane of symmetry. The sign of the double refraction is also still positive, as for the potassium and rubidium salts. Stauroscopic observations carried out with two section-plates ground parallel to the symmetry plane gave the following extinction angles with reference to the normal to the basal plane : — Section 1 6° 7' Section 2 5 26 Mean 5 46 The direction is in front of the normal, nearer to the inclined morpho- logical axis a. This axis of the optical indicatrix is the second median line. The first median line is accordingly situated in the acute angle of the morphological axes at, and is inclined 5° 46' to the axis a. The second median line lies in the obtuse angle ac} and is inclined 21° 5?' to the vertical axis c. 11 iff active Indices. — The results of the refractive index determina- tions with six prisms, ground on crystals from different crops, are presented in the accompanying table. The intermediate refractive index of caesium zinc selenate, corrected to a vacuum, is expressed to near F, for any wave-length A, by the following formula : — i.*i AT _,_ 704 232 2 877 60° ft - 1 ol87 + -_ Study of the Double Selcnates of the Series R3M(Se04)3,6H30. 75 The a indices are reproduced with precisely equal accuracy if the constant 1-5187 is diminished by 0'0036; owing to slightly increased dispersion exhibited by the y indices, these values are reproduced with slightly less accuracy when the constant is increased by the average amount of 0-0050. Alteration of Refraction by Rise of Temperature. — Measurements of the prism angle and minimum deviation at 60°, indicated that the refractive indices are reduced by 0'0015 for 45° rise of temperature (15° to 60°). Axes of the Optical Ellipsoid. — These values are as follows : — Axes of optical indicatrix : a : /3 : y = 0*9977 : 1 : 1'0033. Axes of optical velocity ellipsoid : a : ft : t = 1'0024 : 1 : 0-9968. Kefractive Indices of Caesium Zinc Selenate. Index. Light. Prisin 1. Prism 2. Prism 3. Prism 4. Prism 5. Prism 6. Mean. a fLi 10 — 1 -5285 1 -5290 1 -5287 1 -5292 1 -5296 1 -5302 1 -5293 1 -5298 — 1 -5290 1 -5295 Vibr. j Na — . 1 -5321 1 -5323 1 -5332 1 -5327 — 1 -5326 par. 2M.L. 1 Tl IF — 1 -5353 1 -5395 1 -5356 1 '5396 1-5362 1-5404 1 -5360 1 -5402 — 1 -5358 1 -5399 l<* — 1-5454 1 -5456 1 -5464 1 -5461 — 1 -5459 ft Vibr. par.- symni. axis. fLi .0 J Na 1 Tl 1 F l» 1-5323 1 -5327 1 -5359 1 -5391 1 -5432 1 -5491 — 1 -5326 1 -5332 1 -5361 1 -5396 1-5434 1-5497 — 1-5327 1 -5332 1 -5364 1 -5395 1 -5437 1 -5496 1 -5328 1 '5334 1 -5365 1 -5395 1 -5437 1 -5498 1 -5326 1 -5331 1 -5362 1 -5394 1 -5435 1 -5495 rLi 1 -5371 1 -5371 — 1 -5381 1 -5376 1 -5375 7 Vibr. lc J Na 1-5377 1-5409 1 -5377 1-5408 ~ 1 -5385 1-5418 z 1 -5381 1 "5411 1-5380 1-5412 par. 1M.L. 1 Tl IF 1 -5443 1 -5484 1-5444 1 -5486 — 1-5451 1 -5493 — 1-5445 1-5488 1 -5446 1-5488 IG 1-5544 1-5547 — 1 -5554 •""* 1 -5549 1 -5549 Molecular Optical Constants. — These are as under : — Axis of optical indicatrix. a 0 7 0 -0991 0 -0996 0 -1004 ' (n" + 2)d l<* „ 2 i ]y£ r Q 0 -1016 71-73 0-1022 72 -14 0 -1030 72-69 ' ri- + 2 d l« 73-58 0 -0025 73-98 0-0026 74-58 0 "0026 Molecular dispersion, HTQ. — m 1-85 1 -84 1-89 Molecular refraction M . ...... (J 123 -06 123 -89 125 -03 d Mr. A. K. Tuttoii. A C<>in/i»r"/irc (',•//*/»//<»/, (>j'ti>- A n«l .Imili'. — The optic axial angle in air 2E is so large as to l>c only measurable with some difficulty, even in the cases of the largest and most perfectly transparent sections. Owing to the relatively small double refraction of all the salts of this series, sections require to have a thickness of at least a millimetre in order to afford small ring* md sharp liru.shes. Hence, if the section is not of considerable, and not always attainable, relative breadth the brushes become obscured. Two of the sections prepared enabled trustworthy measurements to be dlit. lined, but the others, although excellent for 2H«, did not exhibit adequately clear brushes in air. Apparent Angle in Air of Caesium Zinc Selenate. Light. Section 1. Section 2. Mean 2£. O / O / 01 Li 162 50 165 20 164 5 C 163 12 165 33 164 22 Na 165 43 166 30 166 6 Tl 168 21 167 23 167 :.i- No measurements were obtainable beyond the green, the angle becoming too large. From the results given in the accompanying table for the true angle, it will be observed that the dispersion of the optic axes is considerably greater than in the cases of the potassium and rubidium salts. To confirm its nature, and determine the dispersion of the median lines, a liquid was sought for whose refraction was the same as that of the crystals, and which was without action on them. Pure methyl salicylate fulfils these conditions, and the interference figures afforded by sections perpendicular to the first median line were observed, while immersed in a cell of this liquid. Measurements in C and F light afforded angles almost identical with those calculated from 2Ha and 2Ho as measured in monobromonaphthalene, and showing the same amount and order of dispersion. Four pairs of section plates were employed in the determinations of the true angle. Study of the Double Selenatcs of the Series RoM(Se04);>,6H30. 77 Determination of True Optic Axial Angle of Caesium Zinc Selenate. Xo. of No. of Light. section perpendi- cular 1st median Observed values of 2Ha. section pei"pendi- cular 2nd median ! Observed values of 2H0. Calculated values of 2Va. Mean value of 2V«. line. line. r 1 0 / 76 27 la o / 87 33 o / 83 37 o / 1 Li 2 3 76 12 76 15 2a 3a 87 28 87 30 83 30 83 31 1 83 33 I 4 76 20 4a 87 30 83 34. J r 1 76 23 la 87 32 83 34 J 2 76 6 2a 87 27 83 27 CQ Qft i 3 76 9 3a 87 29 83 28 OO Ot-' I 4 76 14 4a 87 28 83 32 r 1 75 43 la 87 26 83 12 1 Na J 2 3 75 26 75 25 2a 3a 87 23 87 27 83 4 83 1 jss . L 4 75 29 4a 87 24 83 5 J r 1 75 1 la 87 18 82 50 "1 Tl J 2 3 74 47 74 49 2a 3a 87 18 87 24 82 41 82 39 1 82 43 I 4 1 74 49 4a 87 20 82 41 j . I 1 2 3 74 17 73 53 73 55 la 2a 3a 87 9 87 11 87 21 82 26 82 10 ! 82 6 }82 14 I 4 74 0 4a 87 14 82 12 Dispersion of tlie Median Lines. — When the true angle is observed in methyl salicylate in white light, the brushes are seen to be highly coloured in accordance with the following scheme : — blue | red 1st M.L. red | blue. The angle is thus indubitably the larger for red. If the section is arranged so that the obtuse angle of the morphological axes ac is situated to the left, the optic axis to the right is found to be less dis- persed between C and Tl light by about 10' than the left-hand axis. Hence the median lines are dispersed by about 5' between the same wave-lengths, and so that the first median line lies nearer to the mor- phological axis a for red than for green. Effect of Pdse of Temperature on the Optic Axial Angle. — Repeated measurements of 2E with the best of the sections at 60°, indicated that the angle in air decreases about 3° for 50° rise of temperature. 78 .Mr. A. K. Tut ton. A tollographical ••{ th< Tlin> A concise statement of the conclusions to be «ij Comparison of the Optical Ellipsoids. Optical Indicatrix. a £ y Double refraction. KXn scluiJitc 0-9960 : 1 : 1-0101 141 Kl./n „ 0-9961 : 1 : 1-0072 111 CsZn 0-9977 : 1 : 1-0033 :>»; KZn selunate RbZn „ . I'sXli Optical Velocity Ellipsoid. a b c 1-0040 : 1 : 0-9900 1-0039 : 1 : 0-9929 1-0024 : 1 : 0'9968 140 110 56 Comparison of the Optical Indicatrices when ^K7.n= 1. a 0 y 0-9960 : 1 : 1-0101 1 -0099 KXn 28 27 IJhZn CsZn 0-9988 108 1-0096 1-0027 1-0119 92 The last series of ratios shows the total change of the ellipsoid ou passing from one salt to another, and the numbers are obtained by considering the initial length of the /3 axis, that is, its length in the potassium salt, as unity. They are graphically expressed by the dotted curves in fig. 3. FIG. 3. c« 3 6 Tl 66 17 75 2 82 43 F 66 20 74 55 82 14 The symmetry plane is the common plane of the optic axes. The double refraction is positive for all three salts, and the disposition of the median lines is also identical, subject to the rotation of the whole optical ellipsoid. Comparison of the Molecular Optical Constants. o -i Specific Refraction, — — — = Jt. For ray C (Ua). For ray lly near G. KZu sel. 0-1170 0-1181 0-1210 0-1199 0-1211 0-1242 119 119 130 121 122 134 EbZnsel. 0-1051 0-1062 0-1080 0-1078 0-1089 0-1108 60 66 76 62 67 78 CsZnsel. 0-0991 0-0996 0-1004 0-1016 0-1022 0-1030 84 [ftheScr I; M - 3,0 I M Molecular refraction, — — - • — = m. For ray C (Ha). //- + -2 -/ For ray Hy near O. a 0 7 a 0 7 KZn sel. 62'7G 63 :?7 64'93 64-33 64'98 6ti 3-37 3-40 2-99 3*47 3-49 3*05 Kl.Znsel. 66-13 66'77 67-92 67-80 68-47 69-70 5-60 5-37 4-77 578 .V.ll 4-88 CsZnsel. 71 -7:'. 72-14 72-69 73'58 73-98 74-58 Specific Dispersion, n(i - nc. a & 7 KZn sel 0-0029 0-0030 0-0032 RbZnsel 0-0027 0-0027 0-0028 CsZn sel 0-0025 0-0026 0-0026 Molecular Dispersion, m^ - mt- a /3 7 KZn sel 1-57 % 1-61 172 RbZnsel 1-67 170 178 CsZnsel 1-85 1-84 1-89 // - 1 Molecular Refraction (Gladstone), M. Q a 0 7 KZn sel 107-00 108-24 111-41 5-90 5-96 5-12 RbZnsel 112-90 114-20 116-53 10-16 9-69 8-50 CsZnsel 123'06 123-89 125-03 The rules stated in the Abstract regarding specific and molecular refraction are independent of the wave-length, and whether they are calculated by the formulas of Lorenz or Gladstone and Dale. They are also independent of temperature, for it has been shown with regard to each salt that the refraction is diminished by rise of tem- perature, and the density, the other factor in the calculation, is naturally affected in the same direction by increase of temperature. The replacement of sulphur by selenium in these zinc double salts is accompanied by an increase of molecular refraction of 7-0 — 7-4 Lorenz units or 13'0 — 13'9 Gladstone units for the ray C, according to the direction compared. The increase due to each atom is thus 3-5 — 3-7 or 6-5 — 6 -9 units. The values derived from a comparison of the simple sulphates and selenates were 3*4 — 3*8 or 6*2 — 7 -2. The mean values derived from the two series are thus identical. In the next communication the magnesium group of salts Anil be described. Certain Lavs of Variation. 85 *' Certain Laws of Variation. T. The Reaction of Developing Organisms to Environment." By H. M. VERNON, M.A., M.D., Fellow of Magdalen College, Oxford. Communicated by Professor E. HAY LANKESTER, F.E.S. Eeccived March 7, — Head March 29, 1900. In a former paper* it was shown that the ova of the Echinoid •Strong i/locentrotus limdus were extraordinarily sensitive to their environ- mental conditions at the time of impregnation. For instance, by keep- ing the mixed ova and spermatozoa in water at about 26° or 8° C. for an hour, the plutei obtained after eight days' development were some 5 per cent, smaller than those from ova kept at about 20° at the time of impregnation. It was even found that the effect produced was nearly as great if the time of subjection to the abnormal temperature were reduced to one or three minutes, though if reduced to ten seconds it was not so great. This latter result was probably due to the time being insufficient for all the ova to become impregnated at the abnormal temperature. These observations have now been repeated and confirmed, and in .addition others have been made upon the reaction of the ova to environment in the later stages of their development. It has thereby been found that the degree of this reaction diminishes in more or less regular proportion from the time of impregnation onwards. The method of experiment is fully described in the above-mentioned paper, so it will be sufficient to state here that it consists in shaking pieces of the ovaries and of the testes of several specimens of the Echinoid in small beakers of water, and then bringing portions of the contents to the required abnormal temperature. These portions are then mixed, and after an hour the temperature is gradually brought to the normal by floating the beakers in large vessels of water. The now impregnated ova are then poured into covered jars holding 2| to 4 litres of water, and after eight days the plutei into which they develop are killed and preserved, and measured under the microscope with a micrometer eye-piece in groups of fifty. In addition to these plutei, •others are obtained in each case from ova impregnated at a normal temperature, but allowed in all other respects to develop under similar •conditions. These constitute the normal or standard larvae, from which the variations in the mean size of the other larvae are calculated. The particular dimension measured was the length of the calcareous skeleton of the " body " of the larva, that of the arms as a rule not being deter- mined in the present research. The results obtained, both in the old and the present series of experi- * ' Phil. Trans./ B, 1895, p. 577. VOL. LXVII. U 86 1 >1. II. M. YrllKUI. incuts, a iv -iv< in the accompanying table, those of the old series 1 icing the i : .several observations : — Nuinlii-r of Time of exposure to Temperature of impregnation of Percentage diminution i observuti abnormal temperature. Normal larvae. Abnormal larvae. in Bize produced. f8. . . 1 hour 0 19-9 o 8-7 4-2 Old I5" »» 19 -9 25 -5 5-2 1 «{ 4 ... 1 or 3 niins. 19 -6 8 -0 3-3 series j „ 19 -8 25 -5 5-1 L4.. 10 sees. 19-4 7 -7° or 25 -7° If {I... 1 hour 14-2 1-0 3-6 I... ,, >» 8-3 4-2 1. .. » » 25-5 9-4 1... H » 27-7 6-0 Here it will be seen that, on an average, a more unfavourable effect was produced by exposure of the ova to a temperature of about 25'5" for an hour or a minute at the time of impregnation, than to one of about 8". Thus 5'7 per cent, diminution was produced in the 10 ob- servations at 25-5°, and only 3'9 per cent, in the 14 at about 8°. All these observations were made on the pluteus of Strongylocentmti/* lividus. A further series was also made with the pluteus of Sphcerechiiiux » 22-1 27-1 2'5 5 Bunotei _ „ 4-3 1 hour 20-9 26-7 0-9 5 minute >» » 0-9 • For figures of these larvae, vide 'Phil. Trans.,' B, 1898, p. 468. Certain Laws of Variation. 87 Larvae impregnated for an hour at about 11° were, on an average, 4'0 per cent, smaller than the normal, or practically the same as in the case of Stroii yylocentrot us larvae. An abnormally high temperature does not seem so effective, however, judging from the few results available. Thus one hour's expostire to about 27° caused only 1*7 per cent, diminu- tion, and five minutes' exposure 2 -6 per cent. It should be mentioned that both in this case and that of Strongylocentrotus, the conditions of the short-time exposure experiments differed in one respect from the others, as the beakers of abnormally cooled or warmed ova were poured directly into jars of water at normal temperature, and were not first gradually warmed or cooled. That the shock of this sudden change of temperature cannot be held accountable for much of the effect produced, is proved by the fact that in those experiments in which the time of exposure was reduced to ten seconds, only 1*7 per cent, diminution in the size of the larvae was produced altogether. Experiments were now made to determine the effect of exposure to abnormal temperatures during later stages of development. In each case all the ova were kept for the first hour during impregnation at the same temperature, and were then divided up into two portions, which were poured into jars of water at different temperatures. In the first experiment, made in March, the temperature of the Aquarium tank water was on an average 12 '9°, or distinctly low. Some of the ova, after an hour's impregnation at 13'2°, were accordingly poured into a jar of water kept at 22°. At various later periods the contents of this jar were stirred up, and portions of it poured into smaller jars, which were then transferred to the tank of running water at 12 '9°. The temperature of 22°, which previous experiments had shown to be about the most favourable for the development of the larvae, was maintained practically constant by keeping the jar in a larger vessel of water, which in its turn rested on the top of a water-bath warmed very slightly by means of a gas flame provided with a regulator. The results obtained in this experiment are given in the following table, the body length of the normal larvae, or those kept at 12 '9° during the whole of their development, being taken as 100 : — Conditions. Size. Percentage increase per hour. During hours Normal larvae (12'9°) 100 '00 1 — 11 hours at 22° 99 "96 nil 1— 11 1—28 110 -86 0-40 1 — 28 1—71 „ 116 '25 0 125 28—71 For some unaccountable reason, the larvae developing from ova kept H 2 I)i. H. M. Vi only ten hours at 22° were apparently not aH'e<.-tfeen very slight. It should be pointed out that there is probably in almost every case a possible experimental error of some 2 per cent, in the determination of the growth of these larvae, and occasionally, as we saw in the preced- ing experiment, this error may for some unknown reason Ije consider- ably greater. In the next experiment only approximate results can be calculated. Thus larvae were grown at respectively 13-3° and 20'3° during the whole period of development, but some of them were also preserved and measured after only 3i days' growth. The following values were obtained : — 3J days. 8 days. Keptatl3-3° 88-28 100:00 20-3° . 109-63 111-98 Here we see that the larva? grown only 3£ days at 20-3° are 9 -6 3 per • cut. larger than those grown 8 days at 13'3°. If they had been kept Certain Laws of Variation. 89 an additional 4£ days at 13*3° they would doubtless have grown some- what more. The increase for the 1st to 84th hours is therefore some- what more than this 9 '63 per cent., though less than 11 '98 per cent. Let us take it as 10'80 per cent., or 0'130 per cent, per hour. Again we see that the larvae kept 8 days at 20'3° are only 2*35 per cent, larger than those kept 3i days at this temperature. Hence the maximum effect capable of being produced by the more favourable temperature during this 84th to 192nd hour must be somewhat less than 0*022 per- cent, per hour, or not a fifth of that produced in the earlier period. The next experiment was made in July. After an hour's impregna- tion at 22 '7°, some of the ova were poured into a jar of water which stood in another jar which was surrounded by water and ice. By this means the developing ova were kept at about 12°. Every few hours the water was stirred up and portions of it poiired into jars, which were then transferred to the tank water. This had a mean temperature of 22-5°. The following results were obtained : — Conditions. Size. Percentage diminution per hour. During hours Normal larvse (22 '5°) 100 '00 ] —6 hours at 12° 93 '61 1-28 1 — 6 1—10 , 92-37 0 -31 6—10 1—21 90*09 0'21 10 — 21 Here we see that the effect produced during the 1st to 6th hours was four times as great as that during the 6th to 10th hours. In another experiment some ova, which had been impregnated at 1° C., and had thereby given rise to larvae 3*6 per cent, smaller than the normal, were kept for the next eight hours at 6°. The larvae resulting therefrom were still 9*34 per cent, smaller, or, on an average, were diminished 1*17 per cent, for each hour of exposure to the abnormal temperature. As it was found somewhat troublesome to keep a considerable volume of water some twelve or fifteen degrees below that of the atmosphere for many hours, the rest of the observations were made on the effects of keeping the developing ova at a higher temperature than the normal. As the temperature of the air in the summer months at Naples, where these experiments were made, varies but little from day to day, it was easy to keep the water in a tank holding about 30 litres at a practically constant temperature throughout the experiment. In fact, it did not vary more than 0'3° or 0-5° at the most. In such a tank, if left un- covered, the temperature of the water was found to fall by evaporation to about 25°, or about 2° lower than that of the atmospheric tempera- ture. By covering it up, this could be diminished if wished, and small I'll I>r. 11. M. V.-nion. • (uantities of hot or cold water couM In- added t«> l>rin- tin- temperature . tly what was required. The room in whieh this tank was kept was shut up closely at night so as to prevent cooling. In the first experiment some of the ova, after an hour's impregnation at -'2 •- , were kept for varying periods in this tank of water at 26'0J, and portions of them transferred in smaller jars to Aquarium tank water at a mean temperature of 2.'i-.~> . The following results were obtained : — Conditions. Size. Percentage \arintion per hour. During hours Normal larvae 1 — 4 hours at . 1-8 „ , 1-12 „ „ 1-22 „ ,. 1-144 „ „ (23-5°) .. 100-00 88-33 88-90 -.a -56 99-31 98-43 -3-89 + 0-14 + 1-42 + 0-47 -0-007 1—4 t— 8 8—12 12—22 •2-2—144 1 26° Here we see that three hours' exposure of the developing ova to a temperature of 26° produced a diminution of 11 '7 per cent, in the size of the larvae. Further exposure, on the other hand, not only failed to produce a further diminution of size, but gave an actual increase, which gradually became more and more marked. It is obvious, therefore, that a temperature of 26°, though harmful to the ova in their earlier stages of development, becomes advantageous in the later stages. The reason of this will be made evident further on. It might be thought at first sight that this and other similar experiments in which the environment produces a varying effect, could be of no use in deciding the question under discussion. By judicious selection of certain of the values, how- ever, useful results are obtainable. Thus, in the present instance, we see that by the end of the 8th hour the favourable action of the high temperature has already established itself, and it remains established from that time onwards. All results obtained after this period are, therefore, of value, and the figures which show that the effect produced between the 8th and 12th hours is three times as great as that between the llth and 22nd hours, are genuine ones. The apparent slight diminution of size occurring between the 22nd and 144th hours is doubt- less due to experimental error. In all the observations made during the summer months the larvae were killed and preserved after only six days' growth, instead of eight. This was because they practically reach their maximum size in this period, the rate of growth being so much greater than at the lower temperatures experienced in the spring. In the next experiment, the adverse effect produced during the first few hours' exposure was extraordinarily urcat. so that the favourable Laics of Variation. '.)! influence of the later hours was only very partially able to counteract it. Thus, during the 1st to 4th hours, the diminution effected was no less than 6 -45 per cent, per horn-. There was even a slight additional diminution during the next 3-1- hours, but after that the increase in size noticed in the above experiment set in. It is probable that an un- favourable effect persisted even to the first portion of the 7£ to 11 hour Conditions. Size. Percentage variation per hour. During hours Normal larvae (24 '2°) .... 1 — 4 hours at 28 '0° 100-00 80'64 — 6'45 1—4 1 — 7i 79 '24 -0-40 4— 7i 1—11 „ „ 1 — 22 80-21 84-25 + 0-28 + 0-20 7i-ll 11 — 22 1—144 „ , 87-27 + 0 -025 22—144 period, as the percentage increase per hour is considerably less than one would expect. The only values which are unequivocally genuine are, therefore, the last two. From these we see that the effect produced between the 22nd and 44th hours is only an eighth of that between the llth and 22nd hours. It was thought that perhaps the very marked diminution produced in the size of the larvae might be in part due to the rather sudden changes of temperature to which the developing ova were subjected. These changes were not, as a matter of fact, by any means remarkably sudden, as the water in which the ova were placed after impregnation took about fifteen minutes to attain its temperature of 26', whilst the reverse change from 26° to 24-2°, the temperature of the tank water in this experiment, took about ten minutes. Still, to test this supposition, some of the ova used in this experiment were subjected to several changes of temper.ature. Thus, one portion, directly after the first hour's impregnation, was kept three hours at 26', then seven hours at 24'2°, then fourteen hours at 26°, and the remainder of the time at 24-2°. The size of the larvae obtained therefrom was 85-89, or, if anything, somewhat larger than one would have expected. Another portion of the ova was kept for the 1st to 7|th hours at 24'2°, the 7|th to 22nd hours at 26'0C>, and the remainder of the time at 24'2°. The size of these larvae was 94-93, or 5 per cent, less than the normal. One would have expected them to be if anything slightly larger than the normal, as they were kept at the lower and favourable temperature during the first 6| hours. Still, these two experiments, taken together, show that the effect of even several changes of temperature can only be slight. Dr. H. M. Y Still. a^ain, it was thought that the vigorous .^tin-ing of the water wliich w.i- necessary in order to distribute the organisms evenly through it previous to withdrawal of a portion, might perhaps exert a retarding influence on development. However, this was evidently not so, as, in two experiments, in which the wat-r \va- absolutely unstirred throughout, the resulting larvae varied by respectively + 2*0 and - 2-& per cent, from those derived from frequently-stirred water. In the last experiment to lx> described the developing ova were kept at 25° and not at 26°. Consequently, the diminution produced in the size is not so great. In this case, also, two parallel series of observa- tions were made, one with Strongyloccntrt>tn.< ova as usual, and another with the ova of Sphwrechinus granulari*. In the former case, the- unfavourable effect of the high temperature persisted till the end of the Conditions. Strongylocentrotns larva. During hours Sphcerechiiiut larva-. Size. Percentage variation per hour. Size. Percentage variation per hour. Normal larvae (23'3°) 1—4 hours at 25° .. 1-9 „ 1-21 „ 1-144 „ 100 00 93-95 92-64 97-68 96-88 -2'-02 -0-26 + 0-42 -0-007 1-U 4—9 9—21 21—144 100-00 97-56 94-20 93-96 95-17 -0-81 -0-67 -0-02 + 0-009 9th hour, and, in the latter, until at least the 21st hour, and possibly even later. The Sphwcchmus ova did not react so much to the environ- ment as the Stronyylocenirotu-s, just as was found to be the case in the experiments on the effect of temperature at the time of impregnation. Still these observations on Spha-rerhiwis, as far as they go, more or less support the conclusion drawn from the ArVMgjfiMmJrisfM experiments, viz., that there is a diminishing reaction to environment as the stages of development progress. We see, then, that in all of these four sets of ol»servations the originally unfavourable influence of the high temperature is later on converted into a favourable one. What is the cause of this 1 No- absolute explanation was arrived at, but some observations made on the maximum or death temperatures of the developing ova gave a very satisfactory partial explanation. In these observations portions of the water containing the ova in various stages of development were placed in a beaker, and this was placed in a larger l>eaker of water which was gradually warmed. The beaker containing the ova was continuoiiftly stirred with a thermometer, and when the required tern- Certain Laws of Variation. 93 peratuve had been reached it was removed from the warm water and quickly cooled down by a stream of cold water. After keeping for twenty-four hours, corrosive sublimate was added to kill off any of the embryos still surviving, and they were all collected in a small glass cell and examined under the microscope. From the different stages of development attained by the developing ova killed at the time of heat- ing, and those only killed twenty-four hours later by the sublimate, one could easily determine the effect of the various degrees of high tem- perature. For ova at the time of impregnation the fatal heat temperature is probably about 28'5°. Thus only 31 per cent, of some ova heated to 27 '7° at the time of impregnation were found to have developed to normal blastulse twenty-four hours later, whereas some of the same ova impregnated at a normal temperature (14'2°) were all found without exception to have reached the blastula stage. On the other hand, in another case not a single ovum out of a number heated to 30° at the time of impregnation showed any sign of normal development twenty- four hours later. As regards subsequent stages, portions of some developing ova, four hours after impregnation, were heated to respectively 29°, 32°, 35°, and 38°. Next day all the embryos heated to 29° and 32° had nearly or quite arrived at the pluteus stage, whilst none of those heated to 35° or 38° had got further than the half-formed blastula stage. The fatal temperature must therefore have been between 32° and 35°, or say 32 '5°. Other portions of the same stock of developing ova were heated in a similar manner twelve hours after impregnation. Next day 'all those heated to 29°, 32°, and 35° had arrived at the full or semi- pluteus stage, whilst all of those heated to 38° were either normal blastulae or blastulse just beginning to invaginate. The death tempera- ture in this case must therefore have been about 36 '5°. Still other portions of the same stock of embryos were heated to various tempera- tures twenty-eight hours after impregnation. They had now arrived at the free-swimming pluteus stage, and hence it was quite easy to determine by naked-eye observation what effect had been produced. Of the plutei heated to 37°, none were affected, but all of those heated to 39° sank to the bottom of the beaker in a few minutes. However, about a third of them had recovered an hour after, and all of them had recovered several hours after. None of these plutei were heated above 39°, so the actual death temperature was not determined ; but other results showed that the death temperature is only slightly above the heat paralysis temperature, so one may conclude that it was in this case about 39'5°. On heating some of the six days plutei obtained from the same stock of ova, it was found that a quarter of an hour after heating, three- fourths of those heated to 39° had sunk to the bottom of the beaker, l>r. II. M. and all <>f those heated to 40 J and 41 . After an hour, all of those heated to 'M , ami half ot those heated to 10 . were free-swimming. After four hours four-fifths of those heated to JU were free—wimming, luit none of those heated to 41° had recovered. The death temperature w.-is therefore about 40'3\ These death temperature observations perhaps become more striking if put in tabular form. Thus: — Stage of development. Time after impregnation. Death temperature. 28-5° 1 hours :w -.-> Bl»i*tiilfe and semi-gastrulse u 36 '5 Plutei and semi-plutei 28 39'5 Plutei 6 dav« 40*3 It should l>e remarked that the embryos used in these ol>servation8 had been kept at 26° during development. Six days' plutei obtained from the same stock of ova, but allowed to develop at 23-5° instead of 26°, were found to have a death temperature of 39 -.3 \ Thus the higher temperature of development had produced a certain amount of acclimatisation. The bearing of these resxilts on the curious double effect of exposure of the developing ova to high temperature is obvious. Thus, if a tem- perature of 29 J is fatal to the vitality of ova at the time of impregna- tion, the temperatures a few degrees below this are doubtless unfavour- able to development. Still lower temperatures, on the other hand, are known to exert a favourable influence. Now as in the course of development the death temperature gradually rises, one is quite justified in concluding that the lower limit of the unfavourable tem- perature rises too, and very probably to a more or less similar extent. In the above experiments it was found that up to the end of four hours a temperature of 26 J was distinctly unfavourable to growth. During the next four hours it was more or less neutral, but after this time it was most distinctly favourable. Now the present observations show that between the 4th and 12th hours the death temperature rises about 3°, so what was an 'unfavourable temperature to the earlier stage of development may have become converted into a favourable one to the later stage. Let us now return to the results on the effects of temporary subjection to abnormal temperatures. These were obtained under such a variety of conditions that one is scarcely warranted in grouping them all together, but the majority of them can be split up into three more or less homogeneous groups. In one the so-called normal larvae were Certain Laws of Variation. 95 kept at about 20', and the abnormal ones were kept for varying num- bers of hours at about 8°, whereby a negative effect on growth was produced. In another, the normal larvae were kept at 13°, the abnormal at 22', a positive effect being produced, and in the third the normal larvae were kept at about 24-° and the abnormal ones for varying periods at about 26 \ whereby a negative effect, followed by a positive one, was produced. Normal larvae at 20°, Normal larvae at 13°. Normal larvae at 24°, abnormal at 8°. abnormal at 22°. abnormal at 20". Time of exposure in hours. Mean time in hours. Per cent, variation in size per hour. Time of exposure in hours. Mean time in hours. Per cent, variation in size per hour. Time of exposure in hours. Mean time in hours. Per cent, variation in size per hour. 0—1 0-5 -4-14 _ .. 0—1 0-5 -5-92 1—6 3-5 -1-23 — — — 1—4 2-5 -6-45 1—9 5-0 -117 1— S 4-5 + 1-08 1—4 2-5 -3-89 6—10 8-0 -0-31 1—11 6-0 0-0 1—4 2-5 -2-02 10—21 15-5 -0-21 8—19 13-5 +0-37 8—12 10-0 + 1-42 — — 1—28 14-3 + 0-40 9—21 15-0 ; +0-42 — — — 19—43 31-0 0-0 11—22 16 "5 +0-20 — — — 1—84 42-5 i -t- 0-130 i 12—22 17-0 + 0-47 — — — 28—71 49-5 +0-125 21—144 82-5 o-o — — — i9— 192 105-5 o-o 22—144 83'0 + 0-025 — — — 84—192 138-0 + 0-022 22—144 83-0 o-o '! i The results are arranged in this table according to the times duiing which the larvae were exposed to abnormal conditions. The means of these times are also given, as comparisons are thereby rendered easier. In the first line of the left portion of the table is given the average effect produced in the ten experiments already quoted, in which the ova were kept at about 8° at the time of impregnation. (The experiments in which the time of exposure was one to three minutes have been omitted, as the effect produced in this case was probably something special, directly connected with the act of impregnation.) The results in this group of observations show a fairly regular and very rapid diminution in the effect produced on the size of the larvae with pro- gress in development, but unfortunately they extend only to the 21st jhour. The results in the middle portion of the table extend to the 192nd hour, but they are very irregular. Nevertheless they also, on "the whole, show a rapidly diminishing effect. The results in the right portion of the table bear out this result more fully. In the first line is given the mean of the seven observations in which the ova were kept at 26° at the time of impregnation. In the next three observations in the table, in which the mean time of exposure was 2 '5 hours, the mean effect produced was 4*1 per cent. In the 10th hour it was 1*4 per •cent., in the 15th to 17th hours on an average 0'36 per cent., and in the 83rd hour on an average only O008 per cent. All the observations made, therefore, whether taken in the small groups in which they were OU l'i H. M. \Yrni.n. originally obtained, <>: taken colirrtively, agin; in .showing that the effect of temperature- on the growth of an organism diminishr- rapidly from the time of impregnation onwards. It is to l>e noticed that the effect produced was, as far as could be ascertained, a perma- nent one. At least it persisted to the full larval growth of the organisms, for the larvae were found to practically cease growing after six to eight days' development. How much would have persisted through the metamorphosis to the adult Echinoid stage is, of course, another matter. It seems highly probable that what is tnie for temperature is true for other environmental conditions, and that future research will justify one in assuming the existence of a definite Law of Variation. This might be worded as follows: " Tlte permanent effect of environment on tlie growth of a iiiff organism diminishes regularly and rapidly from the time of im/nrfpiation onwareen completed, and when it was too late for me to put it to u proper experimental test. This criticism depends on the obvious fact that all organisms must be confined within comparatively narrow limits in their powers of growth, so that, for instance, supposing a Strongylocentrotvs pluteus under average conditions attains a size of 100, then probably under no conditions whatsoever could it be made to attain a larger size than 120 to 125, or a smaller one than 80 to 75. Thus in the most extreme variation noticed in any of the numerous observations made on these larvae, the range ran from 19 '2 per cent, above the normal to 18-0 per cent, below it. Now supposing that during the first hours of development an embryo is placed under especially favourable conditions, then it may happen that thereby it is stimulated to undergo all, or nearly all, the increased growth of which it is capable. In subsequent hours, therefore, little if any more favour- able effect may be produced, simply because the organisms from their very nature are tumble to show it. If this principle be examined in relation to the present experiments, I think it can be shown, however, that though no doubt the relation 1>etween the reaction of the organism during the earlier hours to that in the later hours has thereby been exaggerated, yet that there still remains plenty of evidence behind to prove that the diminishing re- action to environment exists in addition. Let us first consider the three series of experiments in which the developing ova were kept at 26° or 25°. Here a diminution of size amounting to from 7 '36 to 20'76 per cent, is produced by the first few hours' exposure to high temperature, so that after this, when the environment begins to exert a favourable influence, we know that it has at least this range of growth capacity at its disposal, plus what- Certain Laws of Variation. 97 •ever amount of increased growth one might have been able to effect in the " normal " larvae, by exposing them to the most favourable condi- tions of growth possible. Now we see that in no case did the favour- able environment succeed in forcing on the growth of the larvae to that of the original normal larvae, so there was always plenty of growth capacity at its disposal. In the experiment in which the ova were kept at 12° instead of 22'5°, there is no doubt that the larvae could have been diminished at least 10 to 15 per cent, more if the conditions had only been suffi- ciently unfavourable and sufficiently long continued. Thus in the .above-mentioned paper it is shown* that larvae kept during the whole period of development at 10°, instead of about 20°, are diminished in size by no less than 24 per cent. We see, therefore, that in two of the different methods adopted for acting on the larvae there was always a considerable amount of growth capacity still present. This may have been true also for the third method, though in this case one cannot prove it. Now we have seen in the above tables that the reaction in the latest periods of development was not a hundredth or even a five hundredth part of that in the first hour, and hence, even admitting the growth capacity was diminished, there can be no doubt whatever that the sensitiveness of the organism to the environment undergoes an enormous gradual diminution. In order to determine exactly the sensitiveness of the developing ova to environment during the various stages of growth, one should keep various portions of them at the normal temperature for the first three, six, &c., hours, and then expose them to the abnormal tempera- ture for a few hours. Then they should be transferred, for the remainder of their developmental period, to the normal temperature. In this way there would always be the same amount of growth capacity for the environment to work upon, and so the effects obtained for the various periods would accurately express the true capacity for reaction. The Effect of other Environmental Conditions. It is obvious that in order to demonstrate the principle under dis- cussion, almost any sufficiently powerful condition of environment might have been chosen. Temperature was hit upon first as being the most convenient one, but further series of experiments were made with another condition also, that of salinity of the water. It has been shownf that growth of the larvae in water of a certain dilution may increase the size by as much as 15 '6 per cent., whilst growth in pure * ' Phil. Trans.,' B, 1898, p. 481. f ' Phil. Trans.,' B, 1895, p. 587. 98 l»r. II. M. Yeiix.n. sea water, instead <>i A<|ii;iriuiii tank water, may increase it by as niiu-li as 19'2 per cent.* As ;i rule, however, the effect produced is nut so great as this. The developing ova, after impregnation for one hour under normal conditions in ordinary Aquarium tank water, were kept for various periods in diluted sea water or pure sea water, and were then trans- ferred to ordinary tank water again. Once the ova have reached the free-swimming blastula stage, or within about five hours in the middle of the summer, it is practically impossible to separate them from tin- water in which they are swimming. In all the experiments, therefore, one part of the diluted or pure sea water, after vigorous stirring to distribute the embryos evenly through it, was poured into ten parts of the normal water. The subsequent growth of the embryos was there- fore continued in tank water containing an eleventh part of the foreign water, but, as will soon be seen, this could have made very little differ- ence to their size. In the first experiment, made in the beginning of April, the develop- ing ova were placed in diluted water made by adding 100 c.c. of fresh water to 1900 c.c. of Aquarium tank water. The specific gravity of this water was found to be 1 '02736 at 15-56' C., whilst that of the unadulterated tank water was 1 '02869. The following were the results obtained : — Normal larvae (13-8°) lOO'OO 1 — 6 hours in diluted water ... 95'44 1—12 „ „ ... 93-55 1—25 „ „ ... 89-02 1—192 „ „ ... 102-08 Here we see that larvae kept first in diluted water, and then trans- ferred to normal water, are considerably diminished in size, those transferred after twenty-four hours' development being diminished by no less than 1 1 per cent. The next experiment was made in July, when the temperature of the water during development was 21-5°, or nearly 8° higher than in the April experiment. The following values were obtained : — Normal larvae (21-5°) 100-00 1 — 6 hours in diluted water . . . 96'90 1—11 „ „ ... 96-63 1-25 „ „ ... 103-28 1—144 „ „ ... 103-66 In this case the larvae reached their minimum size after ten hours' growth, and then so rapidly increased that fourteen hours later they were 3 per cent, larger than the normal. * ' Mittheilungen a. d. Zool. Stat. zu Neapel,' vol. 13, p. 376. Certain Laws of Variation. 9f> The next and last experiment was made in August, when the tem- perature of the water was on an average 24'5°. In this case the larvae were first kept in pure sea water collected several kilometres from the shore. The specific gravity of this water was 1 '02868 at 15-56°, that of the normal tank water being 1-02901. Normal larvas (24-5°) lOO'OO 1 — 4 hours in pure sea water ... 90*38 1—8 „ „ „ ... 93-61 1-12 „ „ „ ... 96-98 1-22 „ „ „ ... 97-04 1_H4 „ „ „ ... 103-61 Here we see that the larvae were reduced to their minimum size \yy only three hours' development in pure sea water, and that longer treatment produced a more and more favourable effect, though only those larvae kept for the whole period of development in the pure water were larger than the normal. We see, therefore, that in each of these series of experiments, though the ultimate effect of the diluted or pure sea water was a favourable one, yet a temporary immersion in it was always unfavour- able. The times of production of the maximum diminution of size were respectively twenty-four, ten, and three hours, or apparently very variable. It is to be noticed, however, that these times more or less correspond with the period at which the developing ova reach the free-swimming blastula stage. Thus at a temperature of 13-8° this was found to be some twenty to twenty-four hours, whilst at a temperature of 24° it was about five hours. At 21-5° it is probably about eight hours, though no exact observations were made to deter- mine it. To what is this unfavourable effect upon the larval growth due ? It is impossible that the pure sea water can of itself be a less favourable medium for the early stages of development than the impixre tank water, and probably the same is true as regards the diluted tank water. In all probability the harmful effect is to be attributed to the shock attendant on the transference of the embryos from water of a lesser degree of salinity to that of a greater. Owing to the differences of osmotic pressure thereby set up, the tissues would immediately undergo a certain amount of shrinkage, and it is a ready assumption that their growth is thereby for a time delayed* The sensitiveness of the embryos to a change of salinity would seem to be less and less the more advanced the state of development, so that after a day or two's growth the harmful influence becomes entirely in abeyance. The reverse process of transference of the developing embryos from more saline to less saline water does not, on the contrary, appear to be 100 /";'- I." I''* • "f }'n i-iiit inn. attrmlfd with any unfavourable result. Thus some of tin- -aine stock of impregnated ova used in the second of the above experiment^ \\eic kept for respectively five and twenty-four hours in on Unary tank water, and were then transferred to diluted water, and kept there for the remainder of their development. The larvae so obtained were respectively 2 '6 and 2 '8 per cent, larger than the normal, or but slightly smaller than the larvae kept for the whole period of develop- ment in diluted water. In another instance, also, embryos kept for respectively twenty-three hours and two days in normal water, and for the rest of development in diluted water, were 3'2 per cent, and Tl per cent, larger than the normal. These experiments therefore prove that the condition of salinity is not a favourable one for the determination of the reaction to environ- ment. Still they serve to emphasise the extraordinary sensitive m-^ of the embryos to their environmental conditions, and also show that this sensitiveness is much greater in the earlier stages of development than in the later ones. Only one observation was made on the effect of keeping the ova in diluted water for an hour at the time of impreg- nation. In this case a diminution of 2 '2 per cent, was produced in the size of the larvae. In the paper already mentioned, however, five experiments of this nature are recorded,* the water being of various degrees of salinity. The effects produced were respectively - 4*3, + 4'1, -1'8, -2'9, and - 2'4 per cent., or on an average -1'5 per cent. Summary. By keeping the impregnated ova of the Echinoid Stfongylocentntiu lividus for various periods during development at an abnormal tem- perature, and comparing the size of the larvae into which they developed with that of larvae allowed to grow throughout under normal con- ditions, it was proved that the permanent effect of temperature on flie gfoidh ifi/iiiiiiislied rapidly and regularly from the time of impregnation onward*. For instance, it was found that exposure of the ova to a temperature of about 8a for an hour at the time of impregnation produced an average diminution of 4'1 per cent, in the size of the larvae measured after eight days' growth ; during the 4th hour after impregnation the diminution produced for each hour's exposure was about 1'2 per cent., and during the 15th hour about 0-2 per cent. In another series, exposure to a temperature of 22° produced an increase in size, this amounting to about I'l per cent, for each hour's exposure in the 4th hour; to 0*4 per cent, in the 14th hour; 0'13 per cent, in the 46th hour, and O'Ol per cent, in the 120th hour. Exposure to a temperature of 26° during the first few hours of development produced a diminution of from 20'8 to 7 '4 per cent., * ' Phil. Trans.,' B, 1895, p. 588. On the Diffusion of Gold in Solid Lead. 101 but in the later hours it produced an increase of from 4'3 to 11*0 per cent. The reaction of the organism to a constant environmental con- dition was thus a variable one. This is probably explicable by the fact that the temperatures necessary to kill the organisms, and presumably also those which cause an unfavourable effect on growth, rise steadily during development. Thus the death temperature is about 28'5° for unsegmented ova, 34° for blastulse, and 40° for plutei. The impregnated ova Avere also found to be much more sensitive to changes in the salinity of the water during the early stages of development than during the later ones. " On the Diffusion of Gold in Solid Lead at the Ordinary Tempe- rature." By Sir W. ROBERTS-AUSTEN, K.C.B., F.K.S., Professor of Metallurgy, Royal College of Science. Eeceived April 5, — Eead May 10, 1900. In the Bakerian Lecture, " On the Diffusion of Metals,"* delivered in 1896, evidence was given to show that gold placed at the base of a column of fluid lead 16 cm. high, maintained at a mean temperature of 492°, or 166° above the melting point of lead, diffuses to the top of the column in an appreciable amount in a single day, the diffusivity ex- pressed in centimetre-day units being 3'0. If the lead be heated, say to 251°, or 75° below the melting point of the metal, diffusion takes place at a much slower rate ; it may still be readily measured, though the diffusivity is only 0'023 in centimetre-day units. In the experi- ments on diffusion in solid lead, the latter metal was prepared with great care, and possessed a high degree of purity. The method of preparation consisted in the reduction of carefully purified carbonate of lead by cyanide of potassium, the reduced metal being cast in carbon moulds. It became evident that at the ordinary temperature the rate of diffusion of solid gold in solid lead must be very slow, and I stated in the Bakerian Lecture that cylinders of lead had been set aside with discs of gold affixed to their bases, in order that, after a sufficient lapse of time, the diffusion occurring at the ordinary temperature might be measured. By the month of March in the present year, four years had elapsed since the experiment began, and the time appeared to be sufficiently long to justify the attempt to ascertain how far the gold had diffused. In starting the experiments the bases of the lead cylinders were carefully brought to a smooth surface, and the discs of pure gold were specially cleaned, the discs of gold being held against * Delivered February 20, 1896. ' Phil. Trans.,' A, vol. 187 (1896) pp. 383—415. VOL. LXVII. I 102 sir W. Roberts-Austen. tin- l>ases of the cylinders by means of clamps. The lal>oratory in which the cylinders were placed consists of a vaulted chamber situated in the basement of the Mint, and its temperature varied but little from a mean of 18° C. The diameters of the cylinders were in all cases 0*88 c.m., their lengths varied somewhat, the longest being 25 cm. At the end of the four years the discs of gold were found to be adherent to the lead. The cylinders were divided into thin slices at right angles to the axes of the cylinders, the first slice was approximately 0'75 mm. thick, but the succeeding layers were about 2'3 mm. thick. By the ordinary methods adopted by assayers, which were conducted with extraordinary precautions, gold wsis found in each of the four lower slices, while only the minutest traces of gold could be found in any slice beyond the fourth from the base. The amount of gold that had diffused in the different cylinders of lead was, however, not uniform. The variation is probably due to difference in contact Gold Spangles. 4th layer. Gold Globule. 3rd layer. Gold Globule. 2nd layer. Gold Globules. 1st layer. • ft between the cylinders of lead and the discs of gold. The results in all four experiments were, however, of the same order, and it will be sufficient to give the actual amounts of gold found in a single cylinder. The richest layer was, of course, the one in direct contact with the gold, and from it a globule of gold was extracted which weighed 0*00005 gramme. There is in the Mint a balance that will readily weigh such globules. The gold extracted from the 2nd and 3rd layers was too small to be weighed, but the amounts could be approxi- mately determined by measurement under a microscope. Actual On the Diffusion of Gold in Solid Lead. 103 photographs of the gold extracted from the successive layers of a cylinder are, moreover, appended ; the magnification being in all cases the same (56 diameters). It may be thought that the amounts are but small, but from the point of view of the assayer, who is accustomed to determine minute quantities of precious metal in large masses of material, the results assume very substantial proportions. Thus the amount of gold found in the richest layer of lead represents no less than 1 oz. 6 dwts. of gold per ton, which could be profitably extracted, while the amount in even the poorest layer is 1^ dwt. per ton. The significance of these results may perhaps be made clearer if it is stated that the amount of gold which would diffuse in solid lead at the ordinary temperature in 1000 years is almost the same as that which would diffuse in molten lead in a single day, provided no more gold is supplied in either case than can be held in solution. This will serve to show how important temperature is in relation to diffusion. As an example of the relative effects of temperature on this purely physical change and on a chemical change, it may be interesting to refer to the case of the dissociation of auric chloride. At the ordinary temperature, the tri-chloride of gold is very stable though it decomposed rapidly at 180°, and my colleague, Dr. Rose,* has shown that though the decomposition of auric chloride may be perceptible at a temperature of 70°, it would nevertheless require, at that tempera- ture, about twenty-five years for its nearly complete change into mono- chloride. I believe, with Eobert Boyle, that though solid gold may have its " little atmosphere," " no man has yet tried whether gold may not in time lose its weight," but the rate at which gold can possibly evaporate into the air at the ordinary temperature must be far less than that at which it diffuses into lead. This shows that the action of a solvent for the gold is necessary, and this solvent is provided by bringing gold into contact with solid metallic lead. I would express my warm acknowledgment to Dr. A. Stansfield, who aids me in conducting the Metallurgical Laboratory at the Royal College of Science, for the care he has devoted to the tedious manipu- lation involved in these experiments. His help has given me great confidence in the accuracy of the results. It may be well to add that I propose to prepare suitable cylinders of lead and gold on the lines indicated in this paper, and to offer them to the National Physical Laboratory with a view to their being examined after such a lapse of time as may be deemed fully adequate. [Note, May 28. — In the Graham Lecture, delivered at Glasgow on the 18th of April last, after speaking of the diffusion of gold in solid lead, * ' Journ. Chem. Soc.,' rol. 67 (1895), p. 904. I 2 104 On (In- />///"/'- Gold in ,W?V/ Lead. I stated that I was "trying to ascertain whether diffusion in the solid metal is, or is not, accelerated by the simultaneous passage of a strong electric current." I again referred to the subject in answer to Lord Kelvin during the discussion which followed the reading of the present paper, and stated that the experiments were incomplete. Su<-h experi- ments take a long time, and it may be well to add that the arrange- ment was just as is described above, except that the lead ordinarily used for assaying was employed. Two cylinders, each 0-88 cm. in diameter, with gold clamped to their respective bases, were maintained at a temperature of 150° for 544 hours, beginning on the 31st of January of the present year. A current of 1-5 amperes was passed through one of the cylinders only during the whole time, the current passing from the gold to the lead. The amount of gold which had diffused into each of the lead cylinders was then ascertained by the method which has already been described. Gold was detected at a height of 7'5 mm. in the case of the cylinder through which the current had passed, while in the other case with no current it had reached a height of 10 mm., the amoxint of gold in each section being also greater. Subsequent experiments showed that a part at least of this difference was due to imperfection in the contact between the lead and the gold. Other experiments are now in progress in which far greater current densities are employed. If these experiments confirm the previous one, they will show that a solution of gold in lead does act, to a small extent, as an electrolyte. The following method was adopted for ensuring contact between the gold and the lead : — My assistant, Mr. W. H. Merrett, succeeded in joining by fusion discs of gold between two cylinders of lead, as is shown in the accom- panying figure. Contact between the metals is, therefore, above re- proach, but it will be many weeks before the results can be recorded. Thirteen years ago I was unsuccessful in the attempt to electrolyse a solution of gold in metallic lead by the passage of a current of 300 amperes through the molten mass.* The failure may have been due to the fact that at the high temperature produced diffusion must have been very rapid. If, therefore, separation of gold from the lead did take place, uniformity of the solution may have been restored by diffu- » British Association Report, 1887, p. 341. On Certain Properties of the Alloys of the Gold-Copper Series. 105 sion. I succeeded in 1895 in obtaining some evidence as to the separa- tion of gold from its solution in metallic lead by electrolysis through a glass septum.* This is, however, only indirectly connected with the electrolysis of alloys.] " On Certain Properties of the Alloys of the Gold-Copper Series." By Professor Sir W. KOBERTS-AUSTEN, K.C.B., F.E.S., and T. KIRKE PiOSE, D.Sc. Eeceived and read May 10, 1900. [PLATE 1.] Notwithstanding the extraordinary importance from a technical point of view of the members of this series, which constitute the gold coinages of the world, singularly little is known respecting either their molecular constitution or even their physical constants. Both the authors of this paper possess unusual facilities for studying them, and they felt that time should not be lost in beginning a systematic exami- nation of the series. The other alloys used for coinage have, on the other hand, not been so neglected. Many years ago one of us,t in submitting his first paper to this Society, gave a curve representing the freezing points of the members of the silver-copper series. This curve, corrected in accordance with more recent work and interpreted in a modern way, proved to be one with two branches meeting at a point where the eutectic alloy of the two metals occurs. The presence of the eutectic has also been since readily detected in standard silver and in several other members of. the series, and possesses a melting point of 778°. As is well known, different portions of a mass of any of the solidified alloys of the silver-copper series, except the eutectic alloy, exhibit divergences in composition which usually amount to about two or three parts in a thousand. The gold-copper series, on the other hand, has long enjoyed a reputation for homogeneity, and it was supposed that the variations in the composition either of the alloy which contains 916-66 parts of gold in 1000, and is used for the coinage of the Empire, or of the alloy which contains 900 parts of gold in 1000, and is one adopted by the Latin Union and in the United States of America, need not exhibit greater divergences than O'l part in 1000. It was, moreover, believed that such a divergence was not the result of any systematic molecular grouping. This view was shaken by one of usj in 1895, when evidence was obtained by chemical analysis that in the case of a gold- * Third Keport to the Alloys Research Committee, ' Proc.Inst. Mech. Engineers, 1895, p. 240. t Roberts- Austen, ' Roy. Soc. Proc.,' vol. 23 (1874), p. 481. £ Rose, ' Chem. Soc. Journ.,' vol. 67, 1895, p. 552. 106 sir W. Etoberte-Auaten and Dr. T. Kirk.- Rose. copper alloy containing 0'2 per cent, of impurity a certain amount of the gold was driven to the inside of the mass by solidification. Cor- roborative evidence was subsequently obtained by the aid of the cool- ing curves afforded by the recording pyrometer, a description of which has already been submitted to this Society. To decide the point finally it was desirable to show to what group of alloys the gold-copper series belongs, and in particular to determine whether the freezing points of the various alloys would lie on a single continuous curve connecting the freezing point of gold with that of copper. Freezing-point curves were accordingly taken by the recording pyrometer of a comprehensive series of alloys. In each case 100 grammes of the alloy were employed, and the thermo-couple, protected by a very thin clay tube, was inserted in the molten mass, which had been previously thoroughly stirred. The rate of cooling was pro- longed as much as possible by allowing the crucible and its contents to remain in position in the gas furnace in which the melting had been effected. The freezing points of this series have, so far as we are aware, never been published, except a few at the copper end by Heycock and Neville.* MM. Charpy and Riche have, however, recently stated that the curve of fusibility of the alloys of gold and copper consists of two branches meeting at a point corresponding to the eutectic alloy which, according to these experimenters, contains 55 per cent, of gold, alloyed with 45 per cent, of copper, and fuses at 940°. t This conclusion is not confirmed by the results of our experi- ments, which are given in the accompanying table and are plotted in the curve, fig. 1. » ' Phil. Trans.,' A, vol. 189 (1897), p. 25. f ' Administration des Monnaies et J Medaillcs. — Rapport au Mini? tre des Finances,' 1899, p. xixviii. On Certain Properties oj the Alloys of the Gold-Copper Series. 107 FIG. l. i,io8\ 1,000 8. 900 800 Copper IO 20 JO 4O 5O 60 70 80 90 100 Percentage of Copper in Atoms. Freezing points of Alloys of Gold and Copper. Percentage weight of gold. Number of atoms of gold per 100 of alloy. Freezing point. 100-00 100-00 1063° C. 95-00 86-00 979 91-60 77-80 951 90-07 74-45 946 88-06 70-32 926 82-05 59-49 905 79-97 56-19 907 73-83 47-54 916 66-26 38-69 928 (32 -20 34-58 941 52-03 25-84 957 51-87 25-72 963 40-45 17-88 994 27-70 10-96 1022 11-15 3-87 1059 0 0 1083 The initial freezing points of the gold-copper alloys are easy to deter- mine, but the subsidiary or eutectic points are very difficult to detect even if a sensitive autographic recorder is employed. We have great confidence in our conclusion that the alloy containing about 82 per cent, of gold and 18 of copper, and not the one which contains 55 per 108 Sir \V. l:.il,»ms-Au>t«-n ami Dr. T. Kirkc Hose. cent, of gold, is really the eutectic. The reason, apart from micro- graphic evidence, is not only that the freezing point of the 82 per cent, gold alloy is lower than that of any other member of the series, but the autographic record reproduced in fig. 2 shows that the angles at A Fio. 2. and B, where the solidification begins and ends, are quite sharp, while the portion between A and B, which represents the actual solidification of the alloy, is horizontal. Neither of these conditions are met with in the autographic records of the other alloys of the series. Moreover, the fracture of the 82 per cent, alloy is conchoidal, as in the case of a great number of other eutectics, owing to the extremely fine state of division of the constituents, which makes these alloys appear to be homo- geneous. The exact composition of the eutectic is, however, difficult to determine. A comparison of the freezing-point curve of the gold-copper with that of silver-copper alloys shows that there are striking similarities when the number of atoms in the alloys are taken as abscissae. It was shown by Levol* as long ago as 1852 that the only homogeneous alloy of silver and copper corresponded in composition with the formula Ag3Cuo, and Heycock and Nevillet confirmed the anticipation of one of us,} which was not verified at the time, that it would prove to be the eutectic of the series. In the gold-copper series the alloy contain- ing 59-49 atoms of gold and 40'41 atoms of copper has a lower freezing point than any other alloy examined, although it is hardly to be distingu shed from the alloys containing a little more copper. The curve of fusibility of the series is much more rounded near this point than that of most binary alloys, and bears a superficial resemblance to that of two substances forming a continuous series of mixed crystals, but micrographic study of the series conclusively shows that it possesses a eutectic. • Levol, ' Annales de Chim. et de Phys.,' rol. 36 (1852), p. 193 ; vol. 39 (1853), p. 163. t ' Phil. Trans.,' A, vol. 189 (1897), p. 25. J Roberts-Austen, loc. cit. On Certain Properties of the Alloys of the Gold-Copper Series. 109 It has been shown by Osmond,* moreover, that silver and copper are each capable of holding a small percentage of the other in solid solution, but that if both metals are present in considerable amounts, the two solidified solutions exist side by side. It is evident therefore that they form an interrupted series of " mixed crystals," and that the substance first solidified in cooling a solution of one metal in the other is not a pure metal, but an isomorphous mixture of the two metals containing only a small percentage of one of them. This conclusion agrees well with the general shape of the curve of fusibility of the silver-copper series, and the still greater concavity of the curve of fusibility of the gold-copper series suggested that a similar condition of things is here met with, but that the gap in the series of mixed crystals is much smaller, and that the mutual solubility of these two metals is greater. Microscopic examination of the alloys of gold and copper affords evidence that this is really the case, but appears to point to the con- clusion that more copper can be dissolved in gold than gold in copper. Alloys containing only a small percentage of copper consist of large crystals similar in shape to those seen in pure gold, and showing no signs of cement between them. They differ from those of pure gold in their colour, which is reddish or reddish-brown, after treatment with nitrohydrochloric acid. When magnified 1580 diameters these crystals show a minutely granular structure which resembles that of pure gold, and affords no evidence of separation into two constituents. Even in standard gold containing only 9T6 per cent, of gold the structure is nearly the same, and is not unlike that of the ground mass of standard silver containing 92 -5 per cent, and 7*5 per cent, of copper prepared in a similar way. On the other hand, the alloys con- taining less gold than the eutectic show crystals of copper set in a matrix which consists apparently of the eutectic. The following examples of photomicrographs of the series are shown in Plate 1 :— Fig. 1, Plate 1, represents the characteristic surface of a small ingot of standard gold. The structure was not developed by etching, and the magnification is only 4'5 diameters. Fig. 2 is a polished section of standard gold etched by immersion for about 15 seconds in a boiling mixture of equal parts of nitric and hydrochloric acid. The magnification is, as in the case of No. 1, 4*5 diameters, and the structure consists of sections cut in various direc- tions by a plane passing through the crystals, of which the mass is composed. Fig. 3 is the eutectic of the gold-copper series ; it contains 80 per cent, of gold and 20 per cent, of copper etched as in the case of the alloy shown in fig. 2 ; the magnification is, however, 1580 diame- * ' Bull, de la Soc. d'Encouragement,' 5th Series, vol. 2 (1897), p. 837. 110 Sir \V. !!..l..Tt<-AnsicM an.l Dr. T. Kirke Rose. ten, whirh reveals the banded structure characteristic of a eutectic alloy. Fig. 4 is standard gold etched as before and magnified diameters. Fig. 5 is a section, etched as before, of an alloy containing 27 per i cut. of gold, and 73 per cent, of copper. In it the presence of two distinct constituents can be seen. The darker portion, which has been readily attacked by the acid, is copper, and the lighter is mainly the eutectic. This fact is proved by fig. 6, which is a very high nia.u'ni- tiratinn (6300 diameters) of the lighter portion of fig. 5, and this on close examination reveals the presence of the laminated or banded eutectic. Another resemblance between the series of gold-copper and silver- copper alloys is to be found in their relative tensile strengths. In both cases the eutectic alloys are extremely brittle, and have a lower tenacity than the other members of the series. In the case of gold, one of us* has shown that the tenacity of an unworkedcast bar of pure gold 7*5 mm. wide and 5'2 mm. thick is 7 tons per square inch, the metal elongating 30'8 per cent, before rupture. Both tenacity and extensibility are greatly increased by the first additions of copper, the tenacity rising in the case of standard gold which contains 8'3 per cent, of copper to more than twice that of pure gold. Under similar conditions, however, we have found that the eutectic alloy of gold and copper has a tensile strength of only 7 '87 tons per square inch, with an elongation of only 3*3 per cent. It is in fact about as brittle as pure gold alloyed with 0'24 per cent, of lead, which has an elongation of 4*9 per cent. We also determined the extensibility of the eutectic alloy of silver and copper to be only 2 '2 per cent., and its tensile strength 29'1 tons per square inch. These are the first cases observed in which eutectic alloys appear to show less tenacity and extensibility than the other members of the series to which they belong. The eutectics of lead and tin, of copper and tin, and of iron and carbon are in each case the strongest alloys of the series, and are not at all brittle. The eutectic of the copper-zinc series is more extensible than any other member of its series, while its tenacity is considerable. The gold-copper and silver-copper alloys differ therefore from other alloys, which appear to be brittle and of low tensile strength only if they have passed through a pasty stage in solidifying, and possess two freezing points, the lower of which is that of the eutectic. It is clear, from the results given above, that gold and copper cannot be expected to form a series of alloys of uniform composition, but will show evidence of liquation similar to that exhibited by silver and copper, though in a less degree. Much evidence on this point was obtained in the course of the preparation of the standard gold trial • Koberts- A usten, ' Phil. Iran*.,' A, vol. 179 <1888), p. 339. Roberts -dutsten 8f Rose.. Roy. Soc. Proc., Vol. 07, PI, I Fig. I. Surface of an Ingot of Standard Gold, (unetched) -\4. 51). Fig. 3. Eutectic, (80 per cent Gold, 20 per cent Copper) x 15800. Fig. 2. Standard Gold. Fig. 4. Standard Gold, x 15800. 5- 27 per cent Gold, 73 per cent Copper* X4-5D Fig. 6. Eutectic portion of 27 per cent Gold, 73 per cent Copper, x 63200 On Certain Properties of the Alloys of the Gold-Copper Series. Ill plate, which contains 91 '6 per cent, of gold and 8'3 per cent, of copper. This alloy was cast into flat bars of various dimensions, and assays were made on pieces cut from all parts of the plates into which the bars were rolled, care being taken to adopt all the precautions described by one of us with a view to ensuring accuracy in the determinations.* This enabled the limit of error to be reduced to 0*02 per 1000 on a mean of three assays. In all, nine plates were prepared before one of the necessary accuracy was obtained, and over 900 determinations of the proportion of gold present in the assay pieces were made. It was found that in general there was a tendency for the outside of the ingots to be richer in gold than the interior, but that this distribution was hardly so regular, and was not so pronounced as that observed in a contrary sense, in standard silver, in which case silver accumulates in the centre of the mass. It may be added that the differences in composition of different parts of the gold bars, though small, are many times larger than the possible errors of assay. The plates were of various dimensions, and were prepared from pure gold and electrodeposited copper, well stirred to ensure uniformity while in the molten condition, cast in iron moulds coated with carbon,, and rolled out to a width of about 17 '5 cm., a thickness of 1 mm., and a length of from 1 to 1'25 metres, the weight being from 3'7 to .4*6 kilos. Series of discs were then cut out in parallel lines, one down the centre of the plate, and two others distant 1 cm. from the .edges. In the case of plate No. 1, intermediate series of discs were cut half way between the centre line and the edges. The means of all assays of each series taken from three typical plates were as follows,, each result giving the mean of from 21 to 27 assays : — No. 1. Ko. 2. No. 3. Left side 916-59 916 '65 917 -03 Left intermediate .... 0'65 Centre line 0-49 916 '55 916 '71 Right intermediate . 0-66 Right side 0'61 916 '58 916'95 Mean of all assays 916 -589 916'598 916 -895 Greatest differences from mean (per 1000 parts) f +0-35 \ —0-36 + 0-14 — 0 '13 + 0-60 — 0-40 Difference between richest and poorest parts of the plate (per 1000 parts) 0'71 0-27 1 -00 In the case of standard silver plates of the same size prepared in a. similar way, the difference between the amounts of silver in the richest * Eose, ' Chem. Soc. Journ.,' vol. 63 (1893), p. 704. Ill' I'n.f. .!. A. Kwiug. and poorest parts of the plate is usually from I'D to 3*0 parts per thousand, or three or four times as great as that in the case of standard gold. The poorest part in the gold plate is, however, always in the centre and the richest part at the outside. The assays made on the " get " of the gold plates, the place on the top of the casting where shrinkage of the mass on solidifying is marked externally by a depression, showed that this part was usually richer in gold than any other part of the plate. These assays are not included in the means given above. Conclusion. It will be evident from the results given above, that when a small proportion of copper is added to gold, the alloy sets as a whole, and forms a solid solution. If small amounts of copper are successively added, the limit of solubility of that metal in gold is at length reached, and a eutectic separates, which forms the whole mass when about 82 per cent, of gold and 18 per cent, of copper are present. Comparatively small additions of gold to copper saturate the latter, and the eutectic makes its appearance before the proportion of gold reaches 27 per cent. The composition of the eutectic corresponds approximately to 60 atoms of gold and 40 of copper, while the silver- copper eutectic also contains nearly 60 atoms of silver and 40 of copper. In other respects also, in the brittleness of the eutectic, in the limited mutual solubility of the two metals, and in the liquation which attends solidification, the gold-copper and silver-copper series resemble each other closely. The main difference is that copper appears to be more soluble in gold than in silver, so that the charac- teristics of the gold-copper alloys are less marked, and consequently have been less easy to detect. " The Crystalline Structure of Metals." Second Paper. By J. A. EWIXG, F.R.S., Professor of Mechanism and Applied Mechanics in the University of Cambridge, and WALTER ROSENHAIN, B.A., St. John's College, Cambridge, 1851 Exhi- bition Research Scholar, Melbourne University. Received May 17,— Read May 31, 1900. (Abstract.) The investigations described in this paper deal principally with the phenomena of annealing. The first section of the paper describes experiments made in the hope of observing under the microscope the process of recrystallisation in strained iron. It is well known that The Crystalline Structure of Metals. 113 rearrangement of the crystalline structure of iron occurs when the metal is heated to redness, and it is .believed that such changes are associated with the evolutions of heat which are indicated by " arrest •points" during the cooling of iron. We hoped that by keeping a polished and etched surface of the strained metal under microscopic observation while the specimen was gradually heated, we should see a more or less sudden change in the crystalline pattern at a temperature corresponding to one of the " arrest points." This attempt, however, to watch the process of recrystallisation failed, although the experi- mental difficulties of keeping a specimen under microscopic observa- tion while it was being heated were successfully overcome. The specimen was electrically heated in a vessel with a thin glass or mica window, and the microscope-objective was kept cool by directing a strong blast of cold air on it and on the surface of the window. In one set of experiments the specimen was kept in an atmosphere of pure hydrogen during the heating, but it was found to become so much tarnished as to obliterate the crystalline pattern. At a red heat, how- ever, the uniform luminous surface of the specimen was seen to develop a number of dark patches which, on slightly raising the temperature, spread over the entire field. No corresponding change was visible during cooling, but the phenomenon would recur every time the speci- men was heated, provided it had been cooled below redness after the previous heating. This phenomenon was absent when the specimen was heated in a vacuum, and we believe that it indicates a chemical action between hydrogen and iron, possibly corresponding to the hydrogen arrest point discovered by Sir W. Eoberts- Austen.* In the next series of experiments the specimen was heated in a vacuum. On prolonged heating the specimen still became tarnished, but at first the crystalline pattern remained visible up to a bright red heat. No change in the pattern was observed, but subsequent polishing and etching of the same surface showed that a real change of crystal- line structure had occurred. The original etched pattern on the sur- face had persisted after heating, simply because the differences of level and surface texture on which it depended had in no way been disturbed by the recrystallisation. Any crystalline pattern seen under the microscope, whether it be produced by etching or relief polishing, consists either of coloured surface deposits, or of steps, or pits, or other differences of level in the surface, and these differences of superficial texture, like mechanical scratches, are not affected by rearrangement of the crystalline elements. Coloured surface deposits would also remain unaffected. All attempts to observe the actual process of recrystallisation must therefore be unsuccessful. The next section of the paper deals with the changes of crystalline structure which go on in lead and other metals at comparatively low * Alloys Research Committee Report, ' Proc. Inst. Mech. Eng.,' 1899. 114 1W. -I. A. Kwi: temperatures. Our attention was directed to this l»y noticing th.-u u piece of plumber's sheet laid, when etched with dilute nitric acid, exhibits a strikingly crystalline structure, with large crystals. The character of this appearance led us to the view that a slow process of* anm-aling or recrystallisation was at work in such lead at ordinary atmospheric temperatures, and we have satisfied ourselves that this is the case. The method of investigation consisted in taking a series of micro-photographs, at low magnifications, of certain marked areas in the surface of a specimen, in order to watch the change which went on through lapse of time, or after application of some thermal treatment. It was necessary, for the reasons given above, to re-etch the surface l)efore each photograph was taken. We have observed that when a piece of cast lead is severely strained by compression, the originally large crystals, after being considerably flattened, are driven into and through one another, so that the etched surface of a strained specimen presents a fine grain, whose crystalline nature only becomes apparent under considerable magnification (80 to 100 diameters). A piece of lead severely strained in this way, and kept for nearly six months in an ordinary room without any special thermal treatment, was found to be undergoing continuous change during that time. A series of photographs of this specimen, taken at intervals during the six months, show that a great number of the small crystals have grown larger at the expense of their neighbours. In similar specimens which have been kept at 200° C., the growth has been much more rapid and more pronounced. The rate of growth is a function of time and temperature, but some specimens show much more rapid changes than others under similar conditions of tempera- ture ; in some cases five minutes' exposure to a temperature of 200° C. is sufficient to alter the crystalline pattern completely. Experiment* have also been made at 100° C. and 150D C., leading to the general result that crystalline growth will occur at any temperature from that of an ordinary room, i,e., 15° C. or 20° C. up to the melting point of lead, and that in general the higher the temperature the more rapid is the initial rate of change. No numerical data can be given, as the crystals are quite irregular, both in size and shape. So far as our observations go, they lead to the result that when such crystalline growth has con- tinued for some time at a given temperature, the structure becomes more or less stable, so far as that temperature is concerned, but expo- sure to a higher temperature may cause further growth to occur. A comparison of micro-photographs of the same specimen at various stages reveals the fact that the growth of an individual crystal occurs, not in uniform layers all round it, but by the formation of arms and branches that invade the neighbouring crystals, the intervening portions sometimes changing at a later stage. This action is analogoi to the formation of skeleton crystals in a metal during solidificatic The Crystalline Structure of Metals. 115 from the liquid state, the space between the branches filling in as solidification proceeds. A marked feature observed in several specimens was the large and rapid growth of one or two individual crystals ; in several instances such individuals grew until they were some hundreds of times larger than their neighbours. We have not been able to discover the deter- mining cause of such growth nor, in general, why one crystal should grow at the expense of its neighbour. Generally the most aggressive crystals were found near the edges of the specimen. It is noticeable that at times a crystal which has already grown considerably is swallowed up by a more powerful neighbour. Some light is thrown on the nature of these actions by the fact that this growth only occurs in crystals that have been subjected to severe plastic strain. By casting the metal in a chill mould, specimens of lead can be obtained having a crystalline structure quite as minute as that found in a severely strained specimen, but this structure remains unchanged at temperatures which produce rapid change in a strained specimen. The investigation of the effects of such comparatively moderate temperatures was extended to other metals, viz., tin, zinc, and cadmium. In tin, the various phenomena of crystallisation from the fluid state are strikingly illustrated on a large scale by the thin layer of that metal which constitutes the surface of commercial tin-plate. The effects of rapid and slow solidification in producing small or large crystals respectively are well marked, and an examination of the etched surface of tin-plate under the microscope reveals beautiful geometrical markings or pits, whose oriented facets produce the well- known selective effect of oblique illumination. The study of the crystalline structure affords an explanation of the nature and method of production of patterns in " moiree metallique," a process which has long been in use for the decoration of articles manufactured of tin- plate. In tin, also, we find that the smallest structure obtainable by quenching the melted metal in water remains unchanged at all tem- peratures up to the melting point ; on the other hand, specimens whose crystalline structure has been modified by great plastic strain exhibit phenomena of recrystallisation at lower temperatures similar to those observed in lead. In a piece of strained tin, an hour's expo- sure to 150° C. produced complete recrystallisation. Exposure to lower temperatures for this short time produced no visible change, but we have not investigated gradual time effects in this metal. The behaviour of strained zinc and cadmium is analogous to that of tin and lead. Exposure to 200° C. is sufficient to produce rapid recrystal- lisation in both zinc and cadmium. This is particularly marked in the case of ordinary sheet zinc. On etching commercial sheet zinc,' no llti Prof. J. A. Ewing. large crystals are visible. In this state the metal is strong and tough, bending quite noiselessly. Aftrr exposure to 200° C. for half an hour, it shows on etching a large brilliant structure, and the metal is then wrak and brittle, and when bent, breaks along well marked cleavage planes and' emits a " cry " like that of tin. In cadmium the recrystallisation is comparatively slow at 200° C., and a time effect has been observed ; the action is rather different from that observed in lead. In cadmium the size to which the crystals grow appears to be much more uniform ; no arms or branches are thrown out and no twin-lamellae are found. The final section of the paper deals with an hypothesis, which is advanced as an attempt to explain the mechanism of the growth of crystals in apparently solid metal.* According to this hypothesis, the metallic impurities which are present in a metal, play an important part in the action. When a metal solidifies from the fluid state, the metallic impurities ultimately crystallise as a film of eutectic alloy in the inter-crystalline junctions ; when fairly large quantities of such eutectics are present, the microscope reveals their presence as an inter- crystalline cement, such as that formed by " pearlite " in slowly cooled mild steel ; very minute quantities of eutectic, however, will be invisible and yet capable of forming a thin film of fusible cement. We conceive that the changes of crystalline structure which go on while the piece is in the solid state are accomplished by the agency of eutectic films between the crystals, in dissolving metal from the sur- faces of some crystals and depositing it on others. When a metal is severely strained, these films of eutectic will be also strained and in many places broken, thus allowing the actual crystals to come into contact with one another. The difference in the rate of etching of adjacent crystals and the phenomena of the electrolytic transfer, in an acid solution, of lead from one crystal to another in the same mass of metal, support the supposition that there is a difference of electric potential between the crystal faces which are brought into contact by severe strain. If it be assumed that a film of eutectic alloy when fluid, or even when in the pasty condition that precedes fusion, can act as an electrolyte, we may regard any two crystals thus in contact, with a film of eutectic interposed in places, as a very low-resistance circuit, and the growth of the positive crystal at the expense of the negative would result. Moreover, such growth would be more rapid at higher temperatures, and its rate at a given temperature would vary in different specimens according to the nature and quantity of the impurities present. That an alloy can act as an electrolyte has not been established experimentally, but the assumption is supported by the close general analogy between alloys and salt solutions. This analogy extends to the very question of the growth of crystals, as * It is proper to say that this hypothesis is due to Mr. Rosenhain. — J. A. E. The Crystalline Structure of Metals. 117 Joly has shown that when crystals of a salt are immersed in their mother-liquor, growth of one at the expense of others will take place. It should be added that solution of one crystal into the intervening film of eutectic, along with deposit on the neighbouring crystal from the eutectic, may occur as a consequence of differences of orientation, producing differences of " solution pressure " apart from actual electrolysis, but the fact that growth has not been observed to occur except in strained crystals favours the view that the action is electrolytic. Some further results which have been deduced from the above hypothesis have been verified by experiment. It follows from the hypothesis that an inter-crystalline boundary containing no eutectic would be an impassable barrier to crystalline growth, but if the eutectic could in any way be supplied, growth across the boundary might take place. In an absolutely pure specimen of lead, there would be no eutectic at the inter-crystalline junctions, but as extremely minute traces of impurity would suffice to set up the action, it is almost hope- less to verify the hypothesis in this way. Some experiments on the cold welding of lead have, however, borne out our conclusions. Two clean, freshly scraped lead surfaces will unite under great pressure in the cold state, and if a piece so welded be annealed, we find that the crystalline growth due to the annealing, with very rare exceptions, never crosses the inter-crystalline boundary formed by the welding surface. To test whether the presence of some eutectic would allow growth to take place, we have scattered small quantities of a more fusible metal over the freshly scraped surfaces of lead before squeezing them together. Then, after- a cold weld had been made by pressure, .on annealing by exposure to 200° C. it was found that crystal growths frequently crossed the line of the weld, as the above theory led us to expect. This experiment has been repeated many times with the uniform result that whenever a small quantity of eutectic, or of an impurity capable of forming a eutectic with the lead, was scattered over the clean surfaces before welding, a distinct growth of crystals across the boundary took place as a resxilt of annealing. On the other hand, a large number of welds were made without introducing any/ impurity, and with very rare exceptions they showed no growth across the boundary, even after the annealing process was continued for some weeks. In rare exceptions a minute amount of growth across the boundary was observed, but these may fairly be accounted for by the almost unavoidable presence of traces of impurity. The result as a whole goes far to confirm this solution theory of crystalline growth in annealing. VOL. LXVII. 118 Sir W. de W. Almey. On the Estimation of the " On the Estimation of the Luminosity of Coloured Surfaces used for Colour Discs." By Sir WILLIAM DE W. ABNEY, K.C.B., F.RS. Received May 5,— Read May 31, 1900. When a source of light is small, such as the points of an arc light, a candle, or lamp, it is comparatively easy to find the luminosity of any coloured surface which is illuminated by it, using the method which has been described in " Colour Photometry, Part II " ;* but when the source of light is a large surface, such as the sky, the method therein described is much more difficult to apply. Quite recently, when examining the question of providing suitable screens for producing the negatives required for three-colour photographic prints, it became necessary to devise a plan by which rings of different colours could be made of equal luminosity in ordinary daylight by rotating them with the proper proportions of black. The rings were concentric and rotated as a disc, Fia. 1. S is the nut of the spindle. Vis a yiolet disc (methyl violet). S is a portion of a blue ring (French ultramarine). R „ „ red ring (Yenniliou). G „ „ green ring (emerald green). 1* „ „ yellow ring (chrome yellow). W „ „ white ring. see fig. 1 , and the difficulty encountered was to ascertain what amount of black ought to form part of each ring. In " Colour Photometry, Part III,"t it was shown that only one ray * Abney and Testing, « Phil. Trans.,' A, 1888. t Abney and Festing, ' Phil. Trans.,' A, 1892. Luminosity of Coloured Surfaces used for Colour Discs. 119 of the spectrum, a greenish-yellow, progressed in luminosity at the same rate as white light. Thus, if part of a white screen were illuminated by this colour and another part by white light, and the luminosities were equal (say) to one candle, then if the two beams were equally diminished they would still match in luminosity until the light was so feeble that it ceased to stimulate the retina. Other rays lying not far from this ray, both on the red and green side of it, gave practically the same results. When, however, the red was compared with the white, each being made equal (say) to one candle, equal diminution of the beams did not show the luminosities as the same, the red becoming rapidly less luminous than the white. With the blue-green, the blue, and the violet the reverse was the case, the white becoming darker than the colour as the beams were equally diminished. A more extended research which is nearly complete shows that the observations recorded in Part III of " Colour Photometry " are correct and can be applied to the problem which I wished to solve. Further, it was shown in the same paper that colour disappeared from all rays of the spectrum long before (except in the case of the pure red) their light was extinguished, this last owing to the feeble stimulation of the retina. Naturally, as the colour began to disappear, the matching of the luminosity of the ray under consideration with that of white became easier to carry out. These facts made it possible to devise a ready method to ascertain the luminosity of any colour. If we take two yellow discs, one (say) 8 inches in diameter and the other 4 inches, and between them sand- wich a pair of interlaced black and white discs of 6 inches diameter, and rotate the four discs on a rotating machine at a speed which will make the black and white into a grey without scintillation, this grey can be made, by altering the proportion of black to white, to match the luminosity of the yellow. A very exact match can be obtained by observing the discs through a black transparent medium, such as the black obtained on a photographic plate after development with methol or amidol developers. The deposit may be so dense that the yellow colour may practically disappear, and the two dull greys may then be readily matched. The luminosity of the yellow in terms of the white is given by the angle which the white subtends when the small proportion of white reflected from the black annulus is added to it. The same procedure may be adopted for a green colour and its luminosity be obtained. It may be stated that four or five observa- tions for each colour should be made if great exactness is required. When the luminosities of these two colours have been determined, 4-inch discs of them may be interlaced with a blue, and a grey formed, which can be matched with a grey formed of black and white as before. K 2 120 Sir "NV. . Wall.-r. Other plants. — Leaves of Tropseolum and of Mathiola, as far as I have- yet seen, give a response to light, that is in the main the contrary of the ordinary Iris response, viz., " positive " during illumination, and subsequently " negative."* Li- 0-01 I hour us FiO. 4.— Series of Normal Responses of Mathiola annua ("positive" during illu- mination, subsequently " negative.") (1793.) Leaves of Nicotiana reacted like Iris. Leaves of Begonia have given a variety of responses strongly sug- gestive of the simultaneous action of two opposed forces effecting a resultant deflection in a + or - direction. As regards Mathiola and Tropaeolum, leaves empty of starch have acted better than leaves laden with starch. Leaves of Ulva gave no distinct response (only one series of trials). Leaves of ordinary garden shrubs and trees, &c. (e.g., Lilac, Pear,. Almond, Mulberry, Vine, Ivy), and petals of flowers, gave no distinct response. Anesthetics. — I was able to make only three satisfactory experi- ments with Iris leaves, before the supply of available material had come to an end.t The first was made upon a vigorous young leaf on May 15th, the test (five minutes' illumination) being made at intervals of rather more than half an hour, with the following result : — * " Negative " as tbe term is employed in physiological literature, /.«., negative pole of positive element (" rincative "). t Note added July 16th. — I have made further trial of anesthetics during the past month upon leaves of Begonia. The effect vas perfectly clear, but slow — viz.r temporary abolition of response by ether vapour, permanent abolition by chloro- form vapour, augmentation by " little " COS, temporary suppression by " much " CO). I think it possible that the refractory behaviour of the Iris leaf mentioned in the teit may have been due to a primary effect of the anaesthetic upon stomatal guard-cells. ( ride ' Proc. Physiol. Soc.,' June 30.) The Electrical Effects of Light upon Green Leaves. 133 Response before COo = 0-008 volt. during and after COo = nil. subsequently = 0'013 ,, during and after COo = nil. subsequently = O'OIO ,, The second experiment was made upon a rather "old" leaf on May 21st, the test being applied at intervals of 40 minutes, and the leaf chamber being at 25°. 1. Normal response = 0'004 to 0*005 2. After chloroform = 0-003,0-002,0-005 3. After more chloroform = 0'005, 0*008 4. After carbon dioxide ... = 0-002, nil, O'OOl, 0 012, 0-005 Upon other leaves (Mathiola, Tropaeolum) I have witnessed — 1. Augmentation of response in consequence of an air-supply con- taining 1 to 3 per 100 of C0,>. 2. Prompt abolition of response when a full stream of CO^ is run through the leaf-chamber. 3. Gradual abolition of response when the air-supply to the leaf- chamber has been kept clear of CO^ ; followed by gradual recovery on the readmission of a small amount of COo. The action of ether upon a leaf of Nicotiana Tabaeum was as follows : — Time 0 Normal = -0-0016 15 mins. ^Etherisation^ -0-0016 30 „ J L - 0-0004 45 „ -0-0008 60 „ -0-0008 90 „ -0-0016 150 „ -0-0020 each period of illumination lasting for 2 minutes. Nature of the Normal Response (Iris Leaves). Direction. — The accidental or " normal " leaf currents observed when the electrodes are first applied to a leaf are of no significance, as regards the response to light. Such " normal " current may be due to accidental injury or to physiological inequality or to unequal imbibi- tion of contacts, and necessarily includes the small amount of current that may arise from the unpolarisable electrodes. It may be positive negative, or non-existent. The regular and normal response to light is independent of such VOL. LXVII. L i:*4 Dr. A. D. \Vullrr. arc Mental currents, provided they !•»• not due to excessive physio- logical differences. The immediate effect of light is to arouse current in the half-shaded leaf, directed from the illuminated to the shaded half (is., in the galvanometer from shaded to illuminated; i.e., from resting to active tissue, as in muscle and nerve). AYith illumination of moderate duration, i.e., not exceeding a few minutes, this first effect lasts as long as its cause, rising towards a maximum. With longer illumination, a maximum is reached from which the effect begins to decline. The current drops to or beyond zero, giving place to the reversed current, which is the regular after- effect of illumination. At the end of an illumination of moderate duration, the current rapidly subsides and gives place to a reversed current directed in the leaf towards the previously illuminated half. This effect and after-effect of illumination are similar in appearance to the effect and after-effect in nerve produced by tetanisation, extend- ing, however, over longer periods of time (figs. 2« and 2l>). MH f/iii fmle. — The electromotive force of the response has a value that usually ranges from 0'005 to 0'020 volt. The leaf resistance (interpolar distance = 5 cm., and breadth = about 1 cm.) is generally between 500,000 and 1,000,000 ohms. The current deflection with these values is between 5 and 40 cm. of scale, with a possible accidental effect of ± 1 cm.* L. R. FIG. 5. — Interval of Time between Illumination L and Response R, of a rigorous Leaf of Iris. (1733.) * The sensitiveness at which this galvanometer was used was such that 10~9A » 1-eni. scale. With the recording galvanometer, 1 cm. of ordinate = 3'10~8 A. The Electrical Effects of Light upon Green Leaves. 135 Resistance. (Ohms.) i ii i§ |o"o'o''o'"ooly| »» »oo QO"OO"OOOOO ooo'oooo 'rM^l^-HOCi CQ CDIO O t* 1> t— f CO 1O *O O C5 »O 10 >O *O O O 10 r-T ^^__, co" c^ «" •*" .2° 0 3 o ^~^ f After-effect. lot.0 us 33 ^ ^H rH O jy o o o 1 I o ^3^i 0 O O O O ^* ^" COtfDOOO !>OOxO W5OO o o o o o o o ^^ o o o ++++ +++ +++ M rHin-^ocoooio IH w us t» -^ox •* ooxoot-coeococooiooo OMNi-ii— IQQ^^OQQO OMO Q OOi-(i-li-l'M •— ( O _i i— I i— og.-.-.^o c fcS «!!«!» 'a -siiiiiiiiiikk!!^ " y " r- 1 i— |o a 0 -» » O : : : : : « ' ~ .s 1 1 : : : : : : g a^ ent experimt $ response is ..... J-8 « ..... s ® .a" ^ :::::::«£ aB B^efl c. „ (all subsequ water tank) Showing that th< perature 6 iH 09 OQ ^ US » .c- i> ~i>. -i> t- 1- 1> i> t» t» t>- o ft _.o t— i 91 M ^m cot- rg-H rHr-li-H- t-( i-l i-lrH *H >% L 2 136 The EI ,,/,/ ,'/„,,, Qretn Leaves. Summary of Observations on Iris — eimlin>i<.I. Resistance. (Ohms.) § §§§§§§§§ 9.9. , to FH" ef |S 1 .y « I | |f | | H| |ff | MMII + + 4- + + d 1 ||| lUiliiiJii^ ooo 00000000^000^^ 1 + 1 1 1 1 1 1 1 1 1 III Li JP •N jj CO i-H r-l r-( ^JIM-NINIMNMJMIN 71 o ' I I I I I I " " :::::£ ::::::::: : ooo •0 •« »J 7j n ri :::::§ :::.::::: 3 0 : : : : : g ::::::::: a, ... p n and arc (no response) .eg . c . 3 S -2 o 3 9 . . ! & 43 ^ . 2 ° X 3 1 l^>fc :g::S:o-'S 1 . - l== :£ : •£ '•* : t- CM ** ^ .— ^ ® ^J p cj o £ o « 13 : iil :J'* J <£ o K 00 O O •-! 9) 1-1 1-1 l-( co 1—1 gglll'g ^gggggggg g ^ i-H P » — jr • 03 minutes' duration) progressively diminish if repeated at " short " in- tervals (10 minutes). At intervals of about 1 hour, successive illumina- tions of 5 minutes produce approximately equal effects. With the leaf of Mathiola, I have used periods of illumination of 2 minutes at intervals of 15 minutes without provoking any obvious sign of fatigue. Conclusions. — The leaves of certain plants under favourable conditions of life exhibit electromotive effects and after-effects, amounting to ± 0'02 volt in response to illumination. As in the case of animal tissue, it is possible that the negative (zincative) effect may be significant of dissimilation, and the opposite effect or after-effect significant of assimilation. The absence of distinct response in petals indicates that chloroplasts are essential to the reaction. The absence of distinct response in the green leaves of trees and shrubs is possibly due to a lower aArerage metabolism in such leaves, as compared with the activity of leaves of small young plants, in which leaf-functions are presumably concentrated within a smaller area. " On the Viscosity of Gases as affected by Temperature." By LORD EAYLEIGH, F.RS. Received June 20, — Read June 21, 1900. A former paper* describes the apparatus by which I examined the influence of temperature upon the viscosity of argon and other gases. I* have recently had the opportunity of testing, in the same way, an interesting sample of gas prepared by Professor Dewar, being the residue, uncondensed by liquid hydrogen, from a large quantity collected at the Bath springs. As was to be expected,! it consists mainly of helium, as is evidenced by its spectrum when rendered luminous in a vacuum tube. A line, not visible from another helium tube, approxi- mately in the position of D5 (Neon) is also apparent. J The result of the comparison of viscosities at about 100° C. and at * 'Roy. Soc. Proc.,' vol. 66 (1900), p. 68. t ' Roy. Soc. Proc.,' vol. 59 (1896), p. 207; vol. 60 (1896), p. 56. J I gpeak doubtfully, because to my eye the interval from T)l to D3 (helium) appeared about equal to that between D3 and the line in question, whereas, accord- ing to the measurements of Ramsay and Travers ('Roy. Soc. Proc.,' vol. 63 (1898), p. 438), the wave-lengths are - of G ns the temperature of the room was to show that the temperature was the same as for ///" //'•/•// Observations at (18* + We*t.) Table I. — Hourly Means of Declination at the Fahnouth on Five selected quiet Day- in lours Mid. 1 2 3 4 5 6 7 8 9 10 11 Winter. .897. t / / / / / / / / / / / an. .. 43-5 437 43-9 44-2 44-0 44-0 43-9 43-8 43 -3 43 -3 43-9 45-4 eb. .. 43-0 43-1 43-2 43-5 43-4 43-5 43-0 427 \-l 1 42-2 4-2 •>• 43-8 'arch . 41-8 42-0 42-2 42-3 42-1 42-0 41-8 41-4 40-0 39-2 40-6 42-9 ct. .. 40 '9 10*8 40-9 41-3 40-9 1 40-8 40-4 39-9 39-1 38-9 40-2 42-6 or. .. 38-9 39-3 39-7 40-1 40-1 39-7 39-6 39-6 39-5 39-1 39-7 41-0 ec. .. 38-5 38-7 39-2 39-4 39-3 39-3 39-1 39-2 38-9 38-7 39-0 39-5 Cleans 41-1 41-3 41-5 41-8 41-6 41-6 41-3 41-1 40-5 40-2 41-0 42-5 Summer. Lpril . . 43-1 42-9 43-2 42-7 42-5 42-7 42-1 40-8 39-0 38-8 39-9 42-3 lay .. 43-4 43-3 43-0 42-9 42-8 42-0 40-8 39-6 38-9 39-7 41-6 44-3 une .. 42-4 42-4 42-0 41-7 41-6 40-2 39-1 39-0 38 -5 39 •« 41 -6 43-6 uly .. 42-3 42-0 41-8 41-6 40-8 39-8 38-7 38-9 38-5 38-7 40-7 427 LUg. .. 40-8 41-0 40-7 40-6 40-3 39-7 ! 38-7 38-2 38-0 38-5 40-6 433 ept. . . 39-6 39-7 39-7 39-6 39-1 39-8 39-5 39-1 38-4 38-8 40-0 42-1 Means 41-9 41-9 41-7 41-5 41-2 40-7 39-8 39-3 38-6 39-0 40-7 43-1 Table II. — Diurnal Inequality of the Falniouth [ours Mid. 1 2 3 4 5 6 7 8 9 10 11 Summer mean. -0-5 -0-5 -0-7 -0-9 -1-2 -1-7 -2-6 -3-1 -3-8 / -3-4 -1-7 +0-7 Winter mean. / -0-9 -07 -0-5 / -0-2 / -0-4 -0-4 -0-7 -0-9 / -1-5 / -1-8 / / -10 + 0-2 Annual mean. / -0-7 / -0-6 -0-6 / -0-6 -0-8 / -1-1 / / -1-7 -2-0 / -27 / -2-6 / / -1-4 +0-5 Falmouth Observatory for the Year 1897. 141 Observatory, determined from the Magnetograph Curves each Month during 1897. Noon 1 2 3 4 5 6 7 8 9 10 11 Mid Winter. 46-6 47-2 46-5 45-6 45-1 45'2 44-6 44-3 43-8 43-6 43-4 43-6 43 •{ 45-7 47-0 46-9 46-1 45-3 44-7 44-5 44-2 43-8 43-5 43-1 42-9 42 •{ 45-8 48-1 48-8 47-9 46-1 44-3 43-3 42-7 42-7 42-6 42-4 41-9 42 '( 44-6 45-2 44-7 44-1 42-6 42-2 41-8 41-1 41-3 40-9 40-8 40-9 40-5 42-5 42-9 42-5 41-6 41-2 41-0 40-2 40-1 39-8 38-9 38-3 38-3 38'! 40-3 40-7 40-5 39-7 39-6 39-2 38-9 38-6 38-5 38-1 38-1 38-2 38 -J 44-3 45-2 45-0 44-2 43-3 42-8 42-2 41-8 41-7 41-3 41-0 41-0 41-] Summer. 45-5 48-3 49-0 477 46-5 45-0 43-9 43-2 43-8 43-7 437 43-4 43'] 47-0 47-8 48-3 46-5 45-0 44-3 43-9 43-6 43-6 43-6 43-5 43-3 43-' 45-5 46-4 46-8 46-2 45-2 44-1 43-9 43-3 43-0 42-9 42-9 42-1 42 •( 45-4 46-3 46-6 46-0 44-8 43-2 42-3 42-2 42-2 42-1 42-4 42-0 42 •( 46-0 47-5 47-5 45-4 44-6 42-8 41-6 41-2 41-2 41-3 41-2 41-0 41 •( 44-0 45-6 45-4 44-7 43-7 42-8 42-7 42-2 41-3 41-0 40-8 40-8 40^ 467 47-0 47-3 46-1 45-0 43-7 43-1 42-6 42-5 42-4 42-4 40-1 42 •( Declination as deduced from Table I. Noon 1 2 3 4 5 6 7 8 9 10 11 Mid Summer mean. + 4-3 + 4-6 + 4-9 + 3-7 + 2-6 + 1-3 + 0-7 + 0-2 + 0-1 o-o o-o -0-3 -0- 1 Winter mean. / / / i ' , t / / / , / / / + 2-3 + 3'2 + 3-0 + 2-2 + 1-3 + 0-8 + 0-2 -0-2 -0-3 -0-7 -1-0 -1-0 -0-< Annual mean. , 1 , / , i / , , , , , , + 3-3 + 3-9 + 4-0 + 3-0 + 2 0 + 1-1 + 0-5 o-o -o-i -0-4 -0-5 -0-7 -0-1 142 Report of Magnetical Observations at 0-18000 + (C.G.S. units). Table III. — Hourly Means of the Horizontal Force at Falmouth on Five selected quiet Days in [Jours Mid. 1 2 3 4 5 6 7 8 9 10 11 Winter. 1897. Ian. . . 578 578 578 580 582 583 584 584 581 576 569 567 Feb. .. 580 579 578 578 579 580 581 581 580 576 571 569 Uaivh . 584 ; 585 585 585 585 586 588 589 584 574 566 565 Oct. .. 607 605 603 603 605 605 606 603 598 589 583 .-,si Nov. .. 598 598 599 600 602 604 605 603 603 597 591 590 Dec. .. 587 588 586 589 592 594 596 595 595 594 588 .-,S5 Means 589 589 588 589 591 592 593 593 590 584 578 577 Summer. April . . 596 596 594 593 593 595 599 597 591 581 571 563 tfay .. 597 596 596 596 596 596 592 587 580 573 571 574 June . . 604 603 601 602 602 599 596 592 587 582 581 586 July . . 610 611 609 608 608 606 603 599 593 588 581 579 A.ug. .. 613 613 612 612 611 609 606 603 595 589 689 593 Sept. . . 604 606 604 604 604 601 604 601 595 587 582 586 Means 604 604 603 603 602 602 600 597 590 583 579 580 Table IV. — Diurnal Inequality of the Falmouth ours Mid. 1 2 * 4 6 6 7 8 9 10 11 SumiiuT UH-in. + -00004 + -00004 + -00003 1 1 + -00003 1 + -00002 + -00002 -00000 -•00003 1 -•00010,- -00017 - -00021 -•00020 Winter •00000 •ooooo 1 1 - -00001 -00000 + -00002 + -00003 + -00004 + -00004 + -00001 - -00006 - -00011 -•o Annual m6&n. + •00002 + •00002 + -00001 + -00002 + -00002 + -OOOOJ + •00002 + •00001 -•00006- -00011 - -00016 - -00016 Falmouth Observatory for the Year 18i)7. 143 Observatory, determined from the Magnetograph Curves each Month during the year 1897. Noon 1 2 3 4 5 6 7 8 9 10 11 Mid. Winter. 568 573 579 578 579 581 583 585 584 583 581 580 579 571 576 576 578 577 579 579 580 583 581 583 583 582 566 572 576 578 582 583 585 590 591 592 590 590 588 590 596 600 602 602 603 607 608 609 609 610 610 609 590 591 594 i 592 597 600 602 602 603 601 601 602 601 588 591 594 593 592 592 593 596 596 595 593 593 592 579 583 587 587 588 590 592 594 594 594 593 593 592 Summer. 566 572 575 587 595 599 599 603 604 604 604 605 605 579 588 595 598 600 605 609 608 609 604 603 605 603 591 593 593 597 600 605 608 612 613 612 612 610 607 584 591 596 605 609 614 617 618 620 618 616 614 612 595 598 600 605 612 617 618 621 621 620 618 617 616 589 595 599 602 606 604 607 611 613 611 611 611 611 584 ! 590 593 599 604 607 610 612 613 612 611 610 609 Horizontal Force as deduced from Table III. Noon 1 2 6 7 8 9 10 11 Mid. Summer mean. - -00016 - -00010 - -00007 - -00001 ; + -00004 + -00007 + -00010 + '00012 + -00013;+ -00012 + -000111+ -00010 + -0000 Winter mean. - -00010 - -00006 - -00002 - -00002 - -00001 + -00001 + -00003 + -00005 + •00005 + -00005 + -00004 + -00004 + -OOW Annual mean. •00008 - -00005 -•00002 + •00002 + -00004 + -00007 144 hYlM.it of Magnetical Observations at Falmouth Observatory for the Year 1898. Latitude 50° 9' 0".N., Longitude 5° 4' 35" W. ; heiglit, 167 feet above mean sea-level. The Declination and the Horizontal and Vertical Forces are deduced from hourly readings of the photographic curves, and so are corrected for the diurnal variation. The results in the following tables, Nos. I, II, III, IV, are deduced from the magnetograph curves which have been standardised by observations of deflection and vibration. These were made with the Collimator Magnet, marked 66A, and the Declinometer Magnet, marked 66c, in the Unifilar Magnetometer No. 66, by Elliott Brothers, of London. The temperature correction (which is probably very small) has not been applied. In Tables V and VI the Vertical Force values, also deduced from the Photographic Curves, have been standardised by observations of Dip and of Horizontal Force, and are published for the first time. The January results are based on four days' means, and the June and Octoter results on the means of three days only. No temperature correction has been applied, and this probably has modified to some extent the apparent law of variation of the Vertical Force throughout the twenty-four hours. As is not unusual with a new instrument, some discontinuities occurred in the course of the year. In Table VII, H is the mean of the absolute values observed during the month (generally three in number), uncorrected for diurnal varia- tions and for any disturbance. V is the mean of the products of the tangent of Dip and H. In Table VIII the Inclination is the mean of the absolute observations, the mean time of which is 3 P.M. The Inclination was observed with the Inclinometer No. 86, by Dover, of Charlton, Kent, and needles 1 and 2, which are 3£ inches in length. The Declination and the Horizontal and Vertical Force values given in Tables I to VI are prepared in accordance with the suggestions made in the Fifth Report of the Committee of the British Association on comparing and reducing magnetic observations^ and the time given is Green wich Mean Time, which is 20 minutes 18 seconds earlier than local time. The following is a list of the days during the year 1898 which were selected by the Astronomer Royal as suitable for the determination of the magnetic diurnal variations, and which have been employed in the preparation of the magnetic tables : — Report of Mayndical Observations at Falmouth Observatory. 145 January 3, 4, 7, 9, 23. February 1, 3, 7, 26, 27. March 1, 3, 4,24,31. April 1, 9, 21, 22, 29. May 7, 19, 21, 23, 25. June 5, 13, 17, 20, 21. July 2, 10, 15, 16, 18. August 1, 8, 10, 15, 25. September 6, 7, 12, 21, 26. October 4, 8, 12, 16, 18. November 5, 10, 14, 29, 30. December 11, 12, 17, 23, 26. EDWARD KITTO, Magnetic Observer. 1 1'i (18° + W«t.) ]!• l»'rt of Maynetical Obscrvafi» Table I. — Hourly Means of Declination at the Falmouth on Five selected quiet I >a\ •> in Hours Mid. 1 2 3 4 5 6 7 8 9 10 11 Winter. 1898. Jan. . . / 37-4 , 37-6 , 38-9 37-5 37-9 38-3 38-4 38-3 38-0 37-8 37-7 , 37-4 Feb. .. 37-8 38-0 38-0 38'0 38-1 38-0 37-7 37-6 37-5 37-1 37-3 387 March . 37-9 38-0 37-9 38-0 377 38-4 37-9 377 37-1 36-2 36 -6 38-7 Oct. .. 34-9 35-2 35-3 35-4 35-0 35-3 35-1 34-5 33-3 33-2 34-7 :<7 -L' Nov. .. 35-7 35-8 36-4 36-5 36-5 36-4 36-2 36 1 36-1 36-0 36-8 38-1 Dec. .. 34-9 35-2 35-7 36-1 35-8 35-8 35-6 35-6 35-3 35-5 36-3 36-6 Means 36-5 36-7 36-9 37-1 36-9 37-0 36-7 36-5 MM 35-9 36-6 38-0 Summer. / / i • / / / / i , , April . . May .. June . . July .. 38-6 38-7 36-9 37-1 37-6 37-9 37-8 37-3 38-4 36-9 37-5 37'0 38-4 38-0 36-8 36-2 37 "3 36 -6 36-7 36-2 37-9 35-1 35-3 35.4 38-1 33-5 33-8 34-1 37-5 32-2 33-2 33-8 36'4 31-7 33-2 33-7 36-0 32-5 33-1 34-3 36 9 35-1 35-0 36-2 38-9 38-6 38-1 38-4 Aug. .. Sept. . . 36-8 36-9 35-5 35-6 36-3 35-0 35-8 35-2 35-7 34-5 35-4 34-3 34-7 33-8 34-3 33-9 33-9 33-6 34-8 33 9 36-8 35-6 39-3 37-9 Means 37'2 37-3 36-9 36-7 36-2 35-6 34-7 34-2 33-8 34-1 35 9 38-5 1 Table II. — Diurnal Inequality of the Falmouth Hours Mid. 1 2 3 4 5 6 7 8 9 10 11 Summer mean. -0-4 -0-3 -0-7 -0-9 -1-4 -2-0 -2-9 -3-4 -3-8 -3-5 -1-7 + 0-9 Winter mean. -09 -07 -0-5 -0-3 -0-5 -0-4 -0-7 -0-9 -,', -1-5 -0-8 + 0-6 Annual mean. -0-7 -0-5 -0-6 -0-6 -1-0 -1-2 -1-8 -2-2 -2-6 -2-5 -13 + 0-8 Falmoutk Observatory for the Year 1898. Observatory, determined from the Magnetograph Curves each Month during 1898. . 141 Noon 1 2 3 4 P 6 7 8 9 10 11 Mid. Winter. 40-1 40-2 39-9 39-0 38-9 38-7 38-2 37-7 37-5 37'2 37-2 37 -4 37 '8 40-2 41-4 41-3 40-8 39 -8 39 -1 38 7 38-1 38-1 37-7 37-5 37-5 37-3 41-6 43-1 43-4 42-4 41-2 j 40-1 39-6 39-3 38-9 38-8 38-5 38 -4 38 -4 39-7 40-4 40-1 39-2 37-2 36-5 36 '6 36-2 35-6 35-3 35-2 35-3 35-1 39-4 39-9 39-4 39-0 38-1 37-7 36-4 36-3 36-2 35-8 35-6 35-5 36-0 37-6 37-5 37-4 36-8 36-2 35-7 35 '3 35-1 34-8 34-5 34-4 34-6 ; 34-8 39-8 40-4 40-3 39-5 38 -6 38 -0 37 '5 37-1 36-9 36-6 36-4 36 -5 36 -6 i 1 Slimmer. 41-5 43-8 44-9 43-6 42-2 41-2 40-3 39-2 39-2 39-2 33-7 38-6 38-3 41 -4 ' 43 -0 42-7 41-1 39-4 37 8 36-6 36-6 36-8 36-9 36-8 36-8 36-1 40-7 42-1 41-8 41-0 39-8 39-0 38-0 37 -7 37 -1 37-4 37-5 37-8 37-8 41-2 42-6 42-3 41-5 40-1 38-8 38-2 38-0 37-9 37-9 37-7 37-5 37-5 41-7 i 43-4 43-2 42-6 40-9 39-4 38-2 37-5 37-6 37-3 37-3 37-2 37-1 40-8 41-9 41-3 40-0 37-8 36-3 35-7 35 -8 35 -8 35-2 35 -3 35 -6 35-5 41-2 42-8 42-7 41-6 40-0 38-8 37-8 37 '5 37 '4 37-3 37-2 37-3 37-1 Declination as deduced from Table I. Noon 1 2 3 4 5 6 7 8 9 10 11 Mid. Summer mean. + 3-6 + 5-2 + 5-1 + 4-0 + 2 -4 + 1-2 + 0-2 -o-i -0-2 -0-3 -0-4 -0-3 -0-5 Winter mean. + 2-4 + 3-0 + 2-9 + 2-1 + 1-2 + 0-6 + 0-1 -=0-3 -0-5 -0-8 -1-0 -0-9 -0-8 Annual mean. + 3-0 + 4-1 + 4-0 + 3-1 1 + 1-8 + 0-9 + 0-2 -0-2 -0-4 -0-6 -0-7 -0-6 -0-7 [48 T;il tie III. — Hourly Means of the Horizontal Force at Falmouth 0-18000 + (C.O.S. unit.). Kve ^lected quiet Vny* in lours Mid. 1 2 3 4 5 6 7 8 9 10 11 Winter. I 1 MIS. an. .. 8M 604 604 605 607 610 612 612 610 607 £99 (98 eb. .. 623 CL'L' 621 at 623 625 627 8M 626 62.-1 621 615 iarch . 623 620 621 622 621 622 624 626 625 619 614 610 ct. .. 638 639 639 636 636 637 636 636 632 623 616 618 OT.. . 635 634 632 634 635 639 641 642 63S 631 <\H «;ji ec. .. 635 635 636 638 637 638 639 639 638 636 635 633 Moans 626 626 G26 626 627 629 630 630 628 624 618 615 1 Summer. pril.. 622 619 620 619 618 617 617 617 613 608 600 :/.»<; [ay .. 636 634 632 630 631 630 624 615 609 602 598 599 une . . 639 637 635 635 635 634 630 626 618 614 610 612 uly.. 630 629 629 628 628 628 624 616 610 605 604 611 tug. . . 648 646 643 641 642 639 636 632 625 619 617 620 ept. .. Ciiii 624 tt-2-2 621 619 618 616 613 609 603 597 596 Means 633 632 630 629 629 62H 625 620 614 609 604 606 Table IV. — Diurnal Inequality of the Falmouth Summer mean. + •00006 + •00006 + •00003 + -00002 + •00002 + •00001 -•00002 - -00007 - -00013 - -00018 - -00023 -•00021 Winter rne&n. •ooooo •ooooo •ooooo •ooooo + -ooooi + -00003 + -00004 + -00004 + -00002 -•00002 - -00008- -00011 Annual nu-:in. + -00003 + -00001 •*- -00002 + '00001 + -00002 + -00002 + -00001 - -00002 - -00006 .- -00010 - -00016 - '00016 Falmouth Observatory for the, Year 1898. 149 Observatory, determined from the Magnetograph Curves on each Month during the Year 1898. Noon 1 2 3 4 5 6 7 8 9 10 11 Mid. Winter. 602 606 608 606 606 608 609 612 612 611 611 611 609 615 619 621 621 621 621 623 624 624 626 625 625 626 609 613 616 617 620 621 622 626 627 627 626 627 627 619 625 633 636 636 639 640 642 643 642 641 641 640 626 632 632 633 636 639 640 643 643 642 638 637 636 635 636 637 637 639 639 640 639 640 639 639 639 637 618 622 625 625 626 628 629 631 632 631 630 630 629 Summer. 600 606 611 614 618 623 628 628 627 626 623 626 625 601 609 617 624 629 635 639 641 644 643 641 640 636 620 625 631 634 637 640 642 644 646 644 641 641 638 617 620 624 630 632 635 637 639 638 639 638 635 634 628 631 630 634 639 645 649 653 656 655 653 654 651 604 613 615 618 620 623 623 629 630 631 630 628 627 612 617 621 626 629 634 636 639 640 640 638 637 635 Horizontal Force as deduced from Table III. Noon 1. 2 3 4 5 6 7 8 9 10 11 Mid. Summer mean. -•00015 - -00010 - -00006 - -00001 + -00002 + -00007 1 + -00009 + -00012 + -00013 + -00013 + -00011 + -00010 + -00008 Winter mean. - -00008 - -00004 - -00001 -•00001 •ooooo + -00002 + -00003 + -00005 + -00006 - -00006 + -00004 + -00004 + -00003 Annual mean. - -00012 -•00007 -•00004 -•00001 + -00001 + -00005 + -00006 + -00009 + -00010 + -00009 + -00008 + -00007 + -00006 (TOL. LXVII. 150 Report of Magndical Observations at Table V. — Hourly Means of the Vertical Force at Fal mouth Five selected quiet Days in 0 -43000 + (C.GK8. units). Hours. Mid. 1 2 3 4 5 6 7 8 9 10 11 Winter. 1898. Jan. .. 610 611 613 613 613 612 611 609 608 607 606 603 Feb. .. 616 616 615 616 616 615 615 614 614 614 612 606 March . 614 616 618 618 619 619 620 621 622 621 615 608 Oct. .. 561 564 565 566 566 567 566 566 565 563 556 550 NOT. .. 555 555 556 557 557 557 556 556 555 554 552 553 Dec. .. 531 531 531 530 530 530 530 529 529 528 528 526 Means 581 582 583 583 r>s4 583 583 583 582 581 578 574 Summer. April . . 557 559 561 562 563 563 563 563 562 559 554 542 May .. 602 603 603 605 607 610 610 609 604 596 585 577 June .. 596 599 601 603 605 608 609 606 601 596 587 570 July .. 523 525 526 527 528 529 530 530 527 521 517 505 Aug. .. 554 555 557 560 563 567 569 570 567 560 554 548 Sept. .. 557 557 557 558 559 560 562 565 563 557 559 539 Means 565 566 568 569 571 573 574 574 571 565 559 547 Table VI. — Diurnal Inequality of the Falmouth tourt. Mid. 1 2 8 4 5 6 7 8 9 10 11 Summer mean. + •00004 + •00005 + •00007 + •00008 + •00010 + •00012 + -0001J + •00013 + •00010 + •00004 -•00002 -•00014 Winter mean. •00000 + •00001 + •00002 + •00002 + •00008 + •00002 + •00002 + •00002 + •00001 •ooooo -•08008 -•00007 Annual mean. + •00002 + •00008 + •00006 + •00005 + •00006 + -00007 + -00008 + 00008 + •00006 + •00002 -•00002 -•00010 Falmouth Observatory for the Year 1898. 151 Observatory, determined from the Magnetograph Curves on each Month during 1898. Noon 1 2 3 4 5 6 7 8 9 10 11 Mid. Winter. 604 607 609 611 611 612 610 609 609 607 608 609 609 605 606 609 612 615 614 613 611 611 609 609 608 606 603 606 611 617 621 623 624 624 625 625 627 629 629 550 551 553 562 566 564 564 563 564 565 567 568 569 553 555 558 559 560 558 557 556 556 555 556 555 555 526 527 532 533 532 531 530 529 528 528 527 527 526 574 575 579 582 584 584 583 582 582 582 582 583 582 Summer. 537 538 545 553 557 560 563 562 560 559 559 559 559 575 580 586 596 602 605 606 606 601 597 598 598 598 563 570 575 582 585 590 590 587 582 578 578 580 581 498 499 504 511 514 517 517 515 515 512 513 515 518 543 542 544 550 556 559 559 558 556 556 556 558 558 530 527 531 539 544 545 545 545 545 546 543 544 544 541 543 548 555 560 563 563 562 560 558 558 559 560 Vertical Force as deduced from Table V. Noon. 123456789 10 11 Mid, Summer mean. - -00020 -•00018 - -00013 - -00006 - -00001 + 00002 + -00002 + -00001 + -00001 - -00003 - -00003 - -00002 Winter mean. -•00007 - -00006 - -00002 + 'OC001 + -00003 + -00003 + -00002 + -00001 + -00001 + -00001 + -00001 Annual mean. - -00018 - -00012 -•00007 -•00003 + '00001 + -00002 + -00002 + •00001 •ooooo -•00001 -•00001 •ooooo •000 152 Report of Magnetical Observations at Fat mouth Observatory. Table VII. — Magnetic Intensity. Absolute Observations. Falmouth Observatory, 1898. 1898. C.GKS. measure. Hor Horizontal force. Vor Vertical force. 0 -18603 0-18600 0-18585 0 -18593 0-18607 0 -18611 0 -18611 0-18628 0-18602 0-18609 0-18624 0 -18636 0-43611 0*43583 0-43562 0-43528 0 -43536 0-43545 0 -43510 0-43564 0-43559 0-43548 0 -43541 0-43544 May July 0-18609 0-43553 Table VIII. — Magnetic Inclination. Absolute Observations. Falmouth Observatory, 1898. Month. Mean. Month. Mean. o / 66 54-5 July 9 o / 66 50-5 19 66 54*3 21 66 50-5 28 66 52-8 29 66 50-5 66 53-9 66 52-5 August 10. ......... 66 50-5 66 51-2 21 66 55-0 15 66 50-6 26 66 52-4 28 66 51-0 March 10 66 53 -3 66 53-0 September 4 66 50-9 66 51-6 19 66 65'9 25 66 51-9 29 66 52-2 30 66 53-9 April 7 66 53-7 66 52 -5 October 14 66 52-5 66 50-7 18 66 52*4 21 66 51 '5 19 66 52-2 29 66 52-8 20 66 52 -3 28 66 51-6 66 51-7 66 50-8 66 52 -2 19 66 50-1 May 6 66 51-8 29 66 50-5 17 66 62-2 27 66 50-6 66 50-5 66 49-9 66 51 -5 21 66 49-4 June 10 66 51-5 31 66 50-1 21 66 51-3 29 66 51-7 66 49*8 66 51*5 153 Report of Magnetical Observations at Falmouth Observatory for the Year 1899. Latitude 50° 9' 0" K, Longitude 5° 4' 35" W. ; height, 167 feet above mean sea-level. The Declination and the Horizontal Force are deduced from hourly readings of the photographic curves, and so are corrected for the diurnal variation. The results in the following tables, Nos. I, II, III, IV, are deduced from the magnetograph curves which have been standardised by observations of deflection and vibration. These were made with the Collimator Magnet, marked 6 6 A, and the Declinometer Magnet, marked 66c, in the Unifilar Magnetometer No. 66, by Elliott Brothers, of London. The temperature correction (which is probably very small) has not been applied. In Table V, H is the mean of the absolute values observed during the month (generally three in number), uncorrected for diurnal varia- tions and for any disturbance. V is the mean of the products of the tangent of Dip and H. In Table VI the Inclination is the mean of the absolute observations, the mean time of which is 3 P.M. The Inclination was observed with the Inclinometer No. 86, by Dover, of Charlton, Kent, and needles 1 and 2, which are 3| inches in length. The Declination and the Horizontal Force values given in Tables I to IV are prepared in accordance with the suggestions made in the Fifth Report of the Committee of the British Association on comparing and reducing magnetic observations, and the time given is Greenwich Mean Time, which is 20 minutes 18 seconds earlier than local time. The following is a list of the days during the year 1899 which were selected by the Astronomer Koyal as suitable for the determination of the magnetic diurnal variations, and which have been employed in the preparation of the magnetic tables : — January ... March May July September November 1, 7, 10, 13, 27. 4, 5, 26, 27, 30. 13, 14, 24, 25, 29. 15, 17, 22, 28, 29. 5, 6, 7, 14, 20. 2, 10, 16, 20, 27. February April... June... August October , December 4, 5, 7, 8, 18. 13, 15, 16, 21, 22. 6, 7, 17, 25, 26. 12, 16, 18, 19, 23. 2, 3, 10, 20, 29. 6, 11, 14, 15, 24. EDWARD KITTO, Observe): M 2 164 (18° + \\ - . Table I. — Hourly Means of Declination at the Falmouth on Five selected quiet Days in 'ours Mid. 1 2 3 4 5 6 7 8 9 10 11 Winter 899. n. . . 35-0 35-5 35-3 35-6 35-4 35-2 35-1 35-0 34-9 35-2 35-8 36-6 •b. .. 33-9 36-2 36-3 36-4 36-3 36-4 36-3 36-3 3f. 5 3 -s 32-7 32-6 32 5 :\-2 •:{ 32-0 31-9 32 7 34-1 S •uiniiirr >ril . . 33-4 33-7 33 •« 33-6 33 4 33-0 32 6 31-6 30-2 30-2 31-6 34-2 »y .. 31-6 31-7 31-3 31-5 31-1 30-1 28-8 27-6 27 3 28-3 30 5 33-3 Ill- .. 32-7 32-6 32-5 32-4 32-1 30-8 29-3 28-8 28-5 L'S .-, 30-6 33-0 ly •• 32-2 31-7 31-7 31 6 31-2 30-5 29-6 29-6 28-9 29-4 31-1 33-0 igust. :n-8 31-8 31-9 31-8 31-6 30-9 30-3 29-3 28-6 29-7 32-2 35-0 pt. .. 28-7 28-7 29-2 28-8 28-5 28-3 27-6 26-5 25-9 20-3 28-6 31-3 if cans 31-7 31-7 31-7 31-6 31 3 30-6 29-7 28-9 28-2 28-7 30-8 33-3 1 Table II.— Diurnal Inequality of the Falmouth oure Mill. 1 2 3 4 5 6 7 8 9 10 11 summer mean. -0-7 -0-7 i -0-7 -0-8 -1-1 -1-8 -2-7 -3-5 + 4 2 -3-7 -1-6 + 0-9 Winter mean. -0-7 -0-4 -0-4 -0-3 -0-4 -0-6 -0-6 -0-8 -1-1 -1-2 -0-4 + 1-0 Annual mean. -0-7 -0-6 / -0-6 -0-6 -0-8 -12 -1-7 -2-2 -2-7 -2-5 -1-0 + 1-0 1 Falmouth Observatory for the Year 1899. 155 Observatory, determined from the Magnetograph Curves each Month during 1899. Noon i 2 3 4 5 6 7 8 9 10 11 Mid. Winter. 37'2 37-6 36-4 36-0 36-0 35-7 35-6 35-2 35-2 34-7 34-8 34-8 34-9 39-5 39-7 39-0 33-1 37-1 37-1 37-0 367 36-4 36-1 36-1 35-8 35-2 37-0 38-5 38-6 36-7 35-3 34-1 33-3 33-0 32'6 32-7 32-6 32-4 32-7 33-5 34-7 35-0 34-2 32-9 32-1 31-8 31-6 31-2 31-1 31-0 30-8 30-9 33-0 33-3 32-5 31-4 30-7 30-6 30-3 30-2 30-1 29-9 30-0 30-0 30-1 31-8 32-3 31-8 31-2 307 30-3 30-1 29-8 29-6 29-7 29-8 29-8 30-0 35-3 36-0 35-6 34-6 33-8 33-3 33-0 32-8 32-5 32-4 32-4 32-3 32-3 Summer. 37-2 39-5 40-2 39-0 37-1 35-7 34-4 33-3 33-4 33-7 33-6 33-4 33-6 35-8 36-9 36-7 35-7 34-0 33-0 32-4 32-2 32-1 31-9 31-8 31-8 31-8 36-2 37-9 38-4 37-8 36-6 35-1 34-4 33-8 33-3 33-1 33-2 33-2 33-0 35-4 36-8 37-4 36-7 35-2 34-0 33-4 33-1 32-9 32-5 32-5 32-2 32-2 36-6 37-7 37-3 36-3 34-6 33-3 32-5 32-4 32-5 32-5 32-3 32-0 31-8 34-3 35-4 35-2 33-4 31-5 29-9 29-3 28-9 29-1 29-0 28-6 28-4 28-7 35-9 37-4 37-5 36-5 34-8 33-5 32-7 32-3 32-2 32-1 32-0 31-8 31-9 Declination as deduced from Table I. Noon 1 2 3 4 5 6 7 8 9 10 11 Mid. Summer mean. / + 3-5 / + 5-0 / + 5-1 i + 4-1 / + 2'4 / + 1-1 / + 0-3 / -o-i / -0-2 / -0-3 / -0-4 / -0-6 / -0-5 Winter mean. + 2-2 + 2-9 / + 2-5 / + 1-5 + 0-7 + 0-2 / -0-1 -0-3 -0-6 / -0-7 / -0-7 / -0-8 / -0-8 Annual mean. + 2-9 +;, + 3-8 + 2-8 + 1-6 + 0-7 + 0-1 -0-2 -0-4 -0-5 -0-6 , -0-7 -07 L56 Report of Magnetical Observations at Table III. — Hourly Means of the Horizontal Force at Falmouth Five selected quiet Days in 0-1800 + (C.GKS. unite). Hours Mid. 1 2 3 4 5 6 7 8 9 10 11 Winter. 1899. Jan. . MB 649 649 650 652 653 654 655 654 647 645 646 Feb. . 651 618 648 648 649 650 653 653 652 649 646 645 March 656 654 652 653 654 657 657 660 655 647 640 <::<: Oct. . 667 669 667 667 668 669 670 668 663 657 649 648 NOT. . 671 670 671 673 675 676 677 675 671 665 662 658 Dec. . 679 677 679 680 681 682 683 682 681 679 677 675 Mt a:.- 662 661 661 662 663 665 666 666 663 657 653 652 Summer. April . . 661 662 661 662 660 661 660 658 654 645 638 631 Mav .. 669 668 666 665 665 663 659 652 643 636 634 il:i'> June . . 667 665 665 663 662 664 660 656 651 646 643 C.W July .. 667 666 664 665 665 663 660 657 654 651 647 645 Aug. .. 672 670 670 669 668 666 663 658 652 647 644 646 Sept, . . 681 i 681 680 680 679 678 676 670 662 653 646 649 Means 670 669 668 667 667 666 663 659 653 646 643 642 Table IV. — Diurnal Inequality of the Falmouth Hours Mid. I 2 3 4 5 6 7 8 9 10 11 Summer mean. + •00006 + •00006 + •00004 + -00003 + •00003 + -00002 - -00001 - -00006 -•00011 -•00018 -•OOOZz'-'OOOB Winter mean. •ooooo - -ooooi -•ooooi •ooooo + •00001 + -00003 + -C0004 + -00004 + -00001 -•00006 - -00009 -•00010 Annual mean + •00008 + -00002 + -00002 + -00002 + -00002 + •00003 + TOOOT - -00001 -•00006 -•00012 - -00016 - -OOOW Falmouth Observatory for the Tear 1899. 157 Observatory, determined from the Magnetograph Curves on each Month during the year 1899. Noon 1 2 3 4 5 6 7 8 9 10 11 Mid. Winter. 649 653 652 648 645 647 650 650 650 649 649 648 647 646 649 652 651 649 649 651 654 656 656 656 655 657 641 649 654 655 655 656 656 659 658 656 656 657 658 648 653 657 662 664 666 669 671 671 671 672 669 669 661 665 669 672 675 676 676 676 675 674 673 672 672 677 678 678 678 679 681 683 682 682 682 685 681 682 654 658 660 661 661 663 664 665 665 665 665 664 664 Summer. 632 639 650 658 660 664 668 663 666 666 664 664 665 648 657 663 666 665 666 671 674 676 674 672 670 670 646 649 654 662 666 669 673 677 677 676 673 671 670 650 653 661 665 665 668 669 674 676 674 674 671 670 649 657 664 666 669 670 673 680 680 679 677 677 676 660 670 676 681 682 684 684 687 685 683 683 684 682 648 654 661 666 668 670 673 676 677 675 674 673 672 Horizontal Force as deduced from Table III. Noon 1 2 4667 9 10 11 Mid. Summer mean. 016 -•00010 -•00003 + •00002 + •00004 + '00006' + -00009 + -00012 + '00013 + -00011 + -00010 + •00009 + -000( Winter mc:in. - -00008 - -00004 - -00002 -•00001 -•00001 + -00001 + -00002 + -00003 + -00003 + -00003 + -00003 + •00002 + •0001 Annual mean. - -00012,- -00007 - '00001 + -00001 + -00001 + '00005 + -00006 + -00008 + -00008 + '00007 + -00007 + '00006 + -000 158 Report of Magnetical Observations at Falmouth Observatory. Table V. — Magnetic Intensity. Absolute Observations. Falmouth Observatory, 1899. C.G.8. i neasure. 1899. Hor Horizontal force. Vor Vertical force. January 0-18636 0 -43558 0-18646 0-43548 0-18631 0 43543 0*18642 0 -43509 May 0*18640 0*43539 0-18633 0-43474 July 0 -18655 0*43515 August 0-18662 0 -43525 (September 0-18654 0 -43572 October 0-18655 0*43498 November 0*18660 0*43548 December 0 -18656 0*43545 Means ..,......, 0 -18647 0 -43531 Table VI. — Magnetic Inclination. Absolute Observations. Falmouth Observatory, 1899. Month. Mean. Month. Mean. January 15 66 49' -5 July 8 66 48 4 23 66 51-0 21 66 46-9 81 66 50-0 29 66 47*8 February 10 66 50-2 66 48*6 August 10 66 47-7 66 47-3 18 66 49*3 22 66 49 -1 27 66 50-1 29 66 46-1 March 10 66 493 66 50-7 September 9 66 47-5 66 48*8 24 66 51 -3 23 66 48-6 30 66 48-9 30 66 50-8 April 7 66 60-1 66 48-9 October 11 66 49-4 66 47-0 14 66 47-9 21 66 46-9 31 66 47-6 May 2 66 48-4 66 49-8 fifi 47-2 66 48-9 8.. . 66 49-1 20 66 48-1 17 66 60-3 29 66 48-0 27 66 48-3 June 8 66 49-4 66 48-4 fi6 48 -3 66 48-5 19 66 47 -4 19 66 48-4 29 66 48-2 29 66 48*6 66 48-0 66 48-5 Data for the Problem of Evolution in Man. 159 " Data for the Problem of Evolution in Man. V. On the Cor- relation between Duration of Life and the Number of Offspring." By Miss M. BEETON, G. U. YULE, and KARL PEARSON, F.R.S., University College, London. Eeceived April 19,— Read June 14, 1900. 1. According to the Darwinian theory of evolution the members of a community less fitted to their environment are removed by death. But this process of natural selection would not permanently modify a race, if the members thus removed were able before death to pro- pagate their species in average numbers. It then becomes an important question to ascertain how far duration of life is related to fertility. In the case of many insects death can interfere only with their single chance of offspring ; they live or not for their one breeding season only.* A similar statement holds good with regard to annual and biennial plants. In such cases there might still be a correlation between duration of life and fertility, but it would be of the indirect character, which we actually find in the case of men and women living beyond sixty years of age — a long life means better physique, and better physique increased fertility. On the other hand, there is a direct correlation of fertility and duration of life in the case of those animals which generally survive a number of breeding seasons, and it is this correlation which we had at first in view when investigating the influence of duration of life on fertility in man. The discovery of the indirect factor in the correlation referred to above was therefore a point of much interest. For it seems to show that the physique fittest to survive is really the physique which is in itself (and independently of the duration of life) most fecund. In continuing our study of the inheritance of longevity,! it occurred to us that it would be possible at the same time as extracting data for duration of life to extract data bearing on the size of the family. Accordingly Miss M. Beeton, in working upon family histories, made records of this additional character. Meanwhile Mr. G. U. Yule, who had been independently at work on this very point, drew my attention again to the matter in connection with a passage in the ' Grammar of Science.'^ We agreed to unite our material, and the result is the following joint paper.§ * Of course longer life may denote greater chance of male or female meeting female or male, but in this case we have not a graduated fertility, the individual is or is not once fertile. t ' Roy. Soc. Proc.,' vol. 65, p. 290. £ Second edition, p. 445. § We have also to very heartily thank Mr. L. N. Filon, M.A., and Mr. K. Tressler for aid in the calculations and in the preparation of diagrams. VOL. LXVII. N 160 Miss M. Beetim, Mi. (i. I . Vulr, an. I IY<»f. K. Pearson. 2. The data dealt with in this paper consist of four series, the first three collected and reduced by Miss M. Beeton, and the fourth series 1>\ .Mr. G. U. Yule. The sources from which they were extracted are the following : — Mothers. Length of Life and Size of Family. Series I. — Taken from the ' Whitney Family, of Connecticut,' a well- known history of an American Quaker family. In order to complete a thousand and more entries some very few additions were made from the ' Backhouse Family,' the history of an English north-country Quaker family. This series may be taken to substantially represent American women more or less closely connected with one strain of blood, either by inheritance or by marriage. As soon as these results were tabled it was noticed that the average age at death of mothers was immensely below the average age at death of Englishwomen. Further, the maximum frequency of deaths which occurs at 35 to 40 was actually greater than the maximum which occurs between 70 to 75 ! Either then American women of this class die very early, or the women of the Whitney family suffer under some hereditary taint, e.g., phthisis. Series II. — Taken from purely English Quaker records. The data for this series were drawn from a great variety of histories and records most kindly placed at our disposal by Mr. Isaac Sharp, Secretary of the Society of Friends, arid by the Secretary of the well-known insurance office, the Friends' Provident Association, both of whom we desire to cordially thank for their aid. The object here was to avoid the selec- tion which may unconsciously be made when the data are drawn from the records of a single family.* In these two series, as in the third series, we selected the records of the Society of Friends because — ('(.) They appear to be the most trustworthy and complete of the family histories available. (b.) The ages at death of the women are given ; these are rarely recorded in other genealogical works. (c.) The artificial limitation of fertility seems to be less probable in a strongly religious community like the Friends than in other classes of the population. In this series the mean age at death, the modal age, and other con- stants are quite fairly in accord with what we know of the population at large. * Of course a '; family " history like tliut of the Whitney family, professing to deal with all the descendants of a single pair, really contains an immense addition through marriage of other strains. Data for the Problem of Evolution in Man. 161 Fathers. Length of Life and Size of Family. Series III. — The great bulk of the data was extracted from the American Whitney Family. Here the features noted for the women were again observed in the men, but to a much less marked degree. There was a rather high maximum frequency of death at 45,* but not so high as the maximum at 75, and the average age at death was some- what lower than we find for the general English population. On the whole the series is a very good one. Series IV. — Extracted from Burke's ' Landed Gentry.' It has been stated elsewhere! that this is a good class for such data. It possesses a higher average fertility than the Peerage, and is a class in which there is probably comparatively little artificial restriction. Unfortu- nately it offers no material for the age at death of women. 3. The following are the chief results obtained from the reduction of these series : — • I. — Table of General Results. S.D. Correla- Regression. Series. Parent. Mean age at Mean size of fertility Li/e 50 Life 50 death. family. Age at death. Size of family. duration of life Whole table. years and years and • under. over. I Mother 53-292 5-269 4-091 3-409 0-5003 0-4174 0-8085 0-2237 II Mother 61-183 5-811 3-769 3-479 0-2374 0-2191 0-7029 0-0941 III Father 58-086 5-469 3-213 3-453 0-4926 0-5282 0-8414 0-2186 IV Father 63-577 5-336 3-037 3-387 0-2010 0-2240 0-5940 0-0720 In this table the unit for the standard deviation of the age at death is 5 years, the unit of the grouping in the accompanying tables. Thus age at death of mothers 35 gives the frequency of all the group of mothers dying between 32 '5 and 37 '5. Of course the age at death of certain parents would lie exactly on the boundary of a group, but such exact information is very rarely forthcoming, and when it is in a few cases forthcoming, i.e., the day of both birth and death is given, it is very improbable that the age of death exactly bisects the year. Thus no fractionising was found necessary in the first three tables. In the ' Landed Gentry,' owing to the nature of the record, Mr. Yule found a small amount of fractionising necessary, and this appears in the table for Series IV. In the regression coefficients above tabulated 5 years is again the unit, and the coefficient of regression is the constant by which * The existence of a modal Talue about 45 has been already noted io the resolution of the mortality curve; it is the mode of the middle age mortality component. See ' Phil. Trans.,' A, vol. 186, p. 403, and Plate 16. t ' Phil. Trans.,' A, vol. 192, p. 257. x 2 Miss M. I1,.- •;«:!. Mi (>. I". Vuli-, au.l 1W. K. the deviation in the age at death from the mean age at death, UK -a-u ••••! in 5-vear units, must he multiplied in order to obtain the probable deviation of the family from the mean family. II.— Table of Regression Formulae or Curves. y = Size of Family, = Duration of Life. Serifs I. American Mothers. (a) For all lives. Straight line : y = 0-821 1+0-083,47-' (/») For lives of 50 years and under. Straight line : y = -l-9881+0-163,233z (c) For lives of 50 years and over. Straight line : »/ - 3-5531 +0-044,748* ('/) For all lives. Cubical parabola. Origin of x at 55 years and1 unit = 5 years : // - 6-0208 + 0-328,474./- 0-035,056*2 + 0-003,000.^ Origin of .<• at birth and unit of .f one year. Series II. English (a) For all lives. Straight line : y = 3-1781 +0-043,819z Origin of x (V) For lives of 50 years and under. Straight line : at birth and // = - 0-6222 + 0'140,584.r f unit of r one (c) For lives of 50 years and over. Straight line : year. // = 4-9341 +0-018.810./- (d) For all lives. Cubical parabola. Origin of -x at 57 -5 years and unit = 5 years : y = -079,120x-0-052,719.r2 + 0- Series III. Fathers. («) For all lives. Straight line : y - - 0-6819 + 0-105,G44./ (/') For lives of 50 years and under. Straight line : y = - 2-6766 + 0-168.: (r) For lives of 50 years and over. Straight line : y = 3-3976 + 0-043,726* (rf) For all lives. Cubical parabola. Origin of x at 55 years and unit = 5 years : y = 5-8 187 +0-363, 122^-0-047,438^ + 0-003,0. Origin <>f i at birth and f unit of x one year. Data for the Prol)lc,,t <>f Ecoli'tion 'm M.«n. 163 Series IV. EnyUxh Fatlwrs. (a) For all lives. Straight line : •y = 2-4877 + 0-044,800* Origin of s, (b) For lives of 50 years and under. Straight line : j at birth and »/ = -1-006 1+0-118,800.?; f unit of v one (c) For lives of 50 years and over. Straight line : year. y = 4-6717 + 0-014,400.^ (j - mean tj = coefficient of regression ( or r~ } x (x - mean .*•).* The cubical parabolas have been fitted by the method of moments. t The whole of this system of formulae has been plotted, and is exhibited graphically in the accompanying diagrams (pp. 176 — 179). These diagrams suffice to give the entire graphical solution of this prob- lem to an exactness sufficient for most practical purposes. A careful ex- amination of these diagrams will enable the reader to follow our general conclusions even more clearly than inspection of the algebraic formulae. 4. General Conclusions. — (i.) The regression straight line for all lives, ua, does not give a satisfactory picture of the relation between age at death of a parent and the average number of offspring. We see at once that it is too steep at the beginning and not steep enough at the end of life. Accordingly, starting from 50 years as the sensible limit to woman's child-bearing period, the mothers were broken up into two groups, and the regression lines calculated separately for lives of 50 years and under, and for lives of 50 years and over. In this way quite a reasonable fit was obtained to the observations. For con- venience, the age of 50 was also taken as a dividing age for fathers. In all four cases the regression line cc for parents living beyond 50 years shows a quite sensible deviation from the perpendicular, or fertility is correlated with longevity even after the fecund period is pasted. If we take American mothers there is no doubt of this increasing * See Yule, ' Roy. Soc. Proc.,' vol. CO, p. 477. t I hare shown in a memoir not yet published (a) how to fit all types of curves, but particularly parabolas of any order, by the method of moments; and (b) that sucli method gives results practically of the same order of exactness as those given by the method of least squares. — K. P. 164 Miss M. Beeton, Mr. <:. U. \ 'ulc. ami Prof. K. Pearson. fertility even up to 90 years of age. With English mothers it is less marked, hut appears to be quite true up to 75 years. Beyond 75 there appears to be a slight decrease. Turning to the two series for fathers we see that we might possibly have better taken 60 than 50 as a dividing age, for the general trend of the observations is much the same up to 60 years. After this there is still a sensible trend in the American results, so that aged fathers are again the most fertile. With the English fathers this relation is, as in the case of English mothers, far less marked, although it is sensible if we take fathers above 50 years. Thus I think we might sum up : That the peculiar physique in both men and women which leads to longevity is also associated with greater fecundity. Of two women who both live beyond 50 years, the longer lived is likely to have had l>efore 50 the larger family. The association is, however, much greater for American than English parents, although the American parents dealt with are, in the great majority of cases, of Anglo-Saxon race. Climate, mode of life, generally selection and environment, seem to be differentiating in this respect the English and the Anglo-American. The English Friends, we should suppose, would be a class very comparable with the American Friends, yet their average life is longer, their fertility greater, and there is less association between longevity and fecun- dity. In lx)th cases our algebraical formulae show that American men and women are more alike, and English men and women are more alike than the women to the women or the men to the men of the two races. This is the more remarkable, as the English Friends as a class are by no means identical with the Landed Gentry. (ii.) In order to represent the continuous change in the regression, which cannot be done by two straight lines, which only enable us to distinguish the fecund and non-fecund periods of life, the statistics were fitted with cubical parabolas. The regression line at any age in life may then be looked upon as the tangent to the cubical parabola at that age. An inspection of Diagrams 3, 4, 7, 8 shows what an excel- lent expression such parabolas are for these statistics. For American mothers and fathers we see dijjdx consistently positive throughout life, and we have a most excellent graphical demonstration of the physical characters which tend to longevity being also associated with fecundity. In the English fathers the same feature appears in a much less marked degree ; there is a point of inflexion in the curve, although dy/dx remains positive. Up to about 75, however, the number of offspring continues to increase with duration of life, and when we break off at 95, the curve has got a renewed outward trend. With English mothers, however, the curve has a small but sen-iUe trend inwards in old age. For fifteen years after the climacteric increased life connotes larger family, i.e., shows fecundity associated Data for the Problem of Evolution in Man. 165 with the physique peculiar to longevity, but beyond 65, as judged by the parabola, longevity is slightly unfavourable to fecundity.* The following are the values of the regression coefficients obtained by differentiating the cubical parabola and referring to birth as origin and a year as unit : — Table III. — Regression Coefficients showing their Change with Duration of Life. Series. Old method, line aa. Cubical parabola. I II III IV 0 -0835 0-0438 0 -1056 0-0448 0 -437,741-0 -010,72402 + 0 -000,0720*2 0 -711,949-0 -019,99552 + 0 -000,137222 0 -501,693-0 -011,80742 + 0 -000,07282- 0 -546,143 - 0 -013,8269z + 0 -000.087322 By simply substituting the number of years of life £, we can find the value of the regression at any age. 5. Illustrations of these Results. — (i.) What is the probable family of an English mother dying at 40 ? (a) gives 4'93, (b) 5'00, and ((/) 5-24, all of which might equally well have been read off on the diagrams. The actually observed number is considerably in excess of all these, i.e., 6'23. In fact, if an English mother lives to 40 years, she will, on the average, have very nearly completed her family. For an American woman (a) gives 4' 16, (b) 4*54, and (d) 4*64. But if she lives another ten or twenty years she will probably have a family of 5 or even 6. (ii.) Compare the strength of the relationship between duration of life and size of family for American fathers dying at 40 and 70 respectively. We find the slope of the cubical parabola at the points corresponding to 40 and 70 years to be 0-1459 and 0-0249 respectively. The mean regression for the whole of life is 0-1056 ; for the first fifty years 0-1683, and for the last fifty 0-0437 (see Table I, and reduce to year as unit). It thus appears that the influence of mere number of years as * It has been suggested that this is due to the nature of the record, there being a tendency to enter only the children who survive their parents. Thus the longer the latter live the fewer would be the offspring entered. In other words, we should be under-estimating the correlation between fertility and longevity. But the Quaker birth -records include all children, and their system is uniform. There does not appear any reason on this ground for English and American returns differing so sensibly. 1G<; M;^ M. Beeton, Mr. (I. I'. Yule, ami I'r-.f. K. 1'. arson. (•(•injured with the physique which tends to longevity has an effect on fertility of about 5 or 6 to 1. (iii.) Weismann has suggested that it may be an advantage to a species that its duration of life should l>e shortened. This is not, & priori, con- firmed for the case of man in the American series : the longer the parents live the greater the number of their offspring. But if we can lay any stress on the bend-in for the English mothers, and on the similar but less marked tendency for the English fathers, we might argue that reproductive selection was possibly in England working against extreme longevity, although favouring parents living till 65 or 70. Indeed those who rush rapidly to brilliant but not over-stable conclusions might emphasise Weismann's views by showing how in an old community, with much greater pressure on the material resources, there is a tendency to reduce the fertility of the long-lived parents ; while in a new community, with plenty of food and occupation for all, the longest-lived parents are the most fertile ! However, all that we can safely say is that there is a marked difference between English and American parents, and that this distinguishing characteristic is almost equally visible if we take opposite sexes of such diverse classes as English Friends and English country gentlemen. We would leave to further investigations its true interpretation. 6. Admitting a substantial correlation between length of life and fertility, it is of great interest to investigate what effect, other things being equal,* reproductive selection would have in modifying the duration of life. The following table gives the mean length of life of parents taken singly and of parents weighted with their offspring : — Table IV. — Mean Duration of Life of Parents in Years. Series. Unweighted parent?. Weighted parents. Progression. I 53-292 59-920 6-628 II 61-183 63-839 2-656 III 58-086 63 -082 4-996 IV 63 -577 05 '510 1-930 Now these are substantial differences even in the case of the English parents (II and IV), but they are very large differences in the case of the American parents (I and III). If we suppose no assortative mating on the basis of characters tending towards longevity, then it is easy to * Omitting, for example, the effect of natural selection as evidenced possibly in a greater death-rate in large families, &c. Data for the Problem of Evolution in Man. 167 obtain a rough approximation to the effect of reproductive selection in modifying the duration of life. It has been shown* that if there be no assortative mating the average deviation, /^, of an array of offspring from the mean of the general population of offspring due to parents deviating h-2 and hs from the means of the general populations of parents is given by : hi = fa— Ai-r-ris— *8j °"2 0-3 where r^ and r13 are coefficients of parental inheritance, and a-v o-2, o-3, the standard deviations in offspring and parents for the character in ques- tion. When that character is longevity our data are not yet complete, but two of us have shown that the value of ri20"i/for the Americans = 0-1 682 (2-656 + 1-900) 1 , ,, ,-, r , = 0.77U 'j> for the English Thus the increased duration of life would be about 2 years per generation from the American data, and about 9 to 9'5 months per generation from the English data. The result for the American series shows us how an especially low expectation of life, due possibly in this case to some family character,! will be rapidly Raised by reproductive selection, if there be no opposing * ' Phil. Trans.,' A, vol. 187, p. 288. t ' Roy. Soc. Proc.,' vol. 65, p. 297. The Landed Gentry would appear to be closer than the Peerage to our present material. J It is by no means certain that this is the true view of the case. We have seen that the American women have their maximum mortality in early middle- life, and only a secondary maximum at 70. The maximum mortality of the table prepared by J. P., F.R.S., for the years 1728-57 (' A Collection of the Yearly Bills of Mortality from 1657 to 1758 inclusive,' London, 1759) occurs about 41 years, and there is no evidence of a maximum at 70 at all. Thus the American data appear to resemble London data of two centuries back. 168 M: .M Beeton, Mr. C. 1". Yule and I'n.f. K. I', Mi-son. factor of evolution. The English results on the other hand show us a small but sensible tendency in reproductive selection to prolong the duration of life. Allowing three generations to a century, we might expect the duration of life to be raised about 2 years in a century by this factor of evolution. In making this statement we are supposing that parents are not a short-lived selection out of the general adult population. There seems no reason why they should be, and we have some statistics to show they are not. Thus for the ' Peerage ' and ' Landed Gentry ' we have shown that for fathers and sons living 20 or 25 years and upwards, the age at death of the father is substantially greater than that of the son.* Further, from data for the Society of Friends, Miss Beeton has found the average age at death of women in general to be 59*831, and the average age of mothers at death to be 59*793, sensibly the same. In the table for 1871 to 1880 given by the Registrar-General, the expecta- tion of life of women in general at 20 years of age is given as 41*66 years, or the average duration of life is 61 '66 years. This is only very slightly greater than our average! for English mothers above, i.e., 61*183, and substantially less than our average for mothers weighted with their offspring, i.e., 63*082 years. Again, the general population of males of 20 had (187 1-80 returns) an average life of 64*48 years, which is not com- parable with our ' Landed Gentry's' sons surviving 20 with an average life of 60*915 years, but with that of their fathers, i.e., 65*96 years. We do not think, therefore, that parentage, in particular maternity, corre- sponds to any shortening in the expectation of life. Thus reproductive selection appears to indicate a real increase in the expectation of life. Such an increased expectation of life is usually considered to have come into existence during our century owing to better sanitary conditions, greater care of the sick and invalided, &c., &c. Its exact estimation is a matter of some difficulty. \Ve find F. G. P. Neison,| working on the Registrar-General's returns before 1841, gives (Table D, p. 8) expecta- tions of life from 10 years onwards. For males of 20 and 25, his mean durations of life are 60*69 and 62*35 • for females of 20 and 25, 61 *60 and 63*36 respectively. These are not substantially less than the Registrar- General's returns for 1881 to 1890, which gives males 60*27 and 61*28, females 62*42 and 63*50 respectively. In fact, the males show reduc- tion. If we stick to the Registrar-General's returns as given for three different periods, and presumably more comparable with each other than with Neison's work, we have the following results : — * ' Roy. Soc. Proc.,' vol. 65, p. 297. t The average age at death of mothers must in our case closely give the expecta- tion of life of •women of 20, for there are few marriages below 20, and we have in our tablet included all cases of sterile unions. J ' Contributions to Vital Statistics,' London, 1846. Data for the Problem of Evolution in Man. Expectation of Life. Ann 1838-54. 1871-80. 1881-90. Se- nisi)»le part of the increased duration of life of this century may l»e due to the inheritance of longevity and the correlation of longevity with fertility. Further determination of the inheritance of duration of life in the case of minors may help to throw additional light on the matter. 7. The following method of illustrating the influence of longevity on fertility may serve to impress the matter on the reader: — In Series I the longer-lived moiety of the mothers produce 64'0 per cent, of the children, and the shorter-lived moiety 36'0 per cent. In Series III the longer-lived moiety of the fathers produce 61 -1 per cent, of the children, and the shorter-lived moiety 38'9 per cent. In Series II the longer-lived moiety of the mothers produce 55'2 per cent, of the children, and the shorter-lived moiety 44'8 per cent. In Series IV the longer-lived moiety of the fathers produce 53-5 per cent, of the children, and the shorter-lived moiety 46'5 per cent. Thus, while the results are all very sensible, those for the American parents are markedly so. In both American and English statistics the influence of longevity on the fertility of the mother is greater than its influence on the father. 8. (\nii-lnJimj l!<-)iwrkf. — A somewhat widespread view of evolution stops at the survival of the fitter without discussing the mode whereby the less fit leave no, or fewer, offspring than the fit. Of course, if the unfit are exterminated before adult life, there is no chance of their reproducing themselves. It has been shown in the second paper of this series that a selective death-rate does exist for adults, so that the whole work of selection does not take place before the reproductive stage is reached. But Miss Beeton's data for the correlation of duration of life in the case of brethren dying as minors seem to show that the selective death-rate for children is rather less than greater than its value for adults."* Hence, for the reduction or extermination of stock unsuited to its environment, we should have to look largely to selection in the adult state. In the present paper we have made what we believe to be the first quantitative determination of how a selective mortality reduces the numbers of the offspring of the less fit relatively to the fitter. In the case of life under wild conditions, the correlation l>etween fertility and power of surviving would probably be far greater. But for such life it is almost impossible to get statistics of this nature ; we are thrown * The matter is still under investigation, so that this conclusion is stated subject to modification. Of course, the selective death-rate among children may largely remove those not weak from inherited constitution, but by physical or physiolo- gical accident. These our method of investigation would throw into the non- selective death-rate. Data for the Problem of Ecolution in M«n. 171 hack upon measuring the effect in man, and thus obtaining what may well be considered as a minimum value of the influence under dis- cussion. In the course of our investigations we have seen that the relationship between fertility and duration of life does not cease with the fecund period. We thus reach the important result that characters which build up a constitution fittest to survive are also characters which encourage its fertility. This result is of great value from the standpoint of the differentiation of type, where it is absolutely necessary that the fittest to survive should also be the most fertile.* On the other hand, we note that duration of life is a character capable of modification by reproductive selection, and we suggest that a considerable part of the increased expectation of life observed in recent years may be due to this cause. In the case of the American statistics, we see at once how it can replace a remarkably short-lived stock by a longer-lived stock, the bulk of the offspring coming from the longer-lived members. * ' The Grammar of Science,' second edition, pp. 448— 171* Mi^s M. Beeton, Mr. ('-. V. Yule, ami I'mf. K. 1'oarson. T I 3 " i s 8 S •s o 2 > r. ~ •- -i :: z ~ z ~ z co OOOO s 8 s C0t»0t»00 <-* -H 05 OS «0 •-< — tal M i O o fo Dcita for the Problem of Evolution in Man. 173 O OS O -* rH O r-i •fruudisjjb fo . 174 Mi>s M. IS.i-i.,11. Mi. C. I'. Yiih-, iiinl IV.f. K. IVui i § £ o 1 eac. M — ^H 8 * 8 8 8 s g I ^ U) «* . C r-l •Guudsffo fo VOL. LXVII. 176 Mi- Mr. s M. l'..-.-t..n, Mr. < :. I". Vul.-, an.) I'mf. K. Age I:K\V GRAY, LL.D., F.R.S., Professor of Natural Philosophy in tin1 Uni- versity of Glasgow, and VINCENT J. BLYTII, M.A.. and JAMES S. DUNLOP, M.A., B.Sc., Houlds worth Research Students in the University of Glasgow. Received May 24, — Read June 21, 1900. At the outset the object of this investigation was to determine the variation, produced by rise of temperature, in the rigidity-modulus and in the Young's modulus of wires of different metals, but the obser- vations made for this purpose yielded what seemed to us interesting information as to the viscosity of the substances examined, and led to an extension of the research. Heretofore but little attention seems to have been paid to alteration of rigidity with temperature, though several results, of apparently very different degrees of value, are available for Young's modulus. The earliest of these are probably those of Wertheim's experiments,* but on account of the smallness of the quantities observed it is not possible to regard them as even nearly correct. Experiments were made about 1870 by F. Kohlrausch and F. E. Loomis.f After referring to the difficulties attending the method adopted by Wertheim, they remark : " All these difficulties disappear, however, and at the same time the most accurate method of observa- tion is obtained, by employing for investigation the torsion elasticity, whose choice is further to be recommended from the fact that torsion is so generally employed in measurements. If a wire is loaded with a weight and set in vibration about its vertical axis, the reciprocal value of the square of the time of vibration affords a direct measure for the coefficient of the torsion of the wire. Since observations of the period of vibration are among the most accurate known in physics, the varia- tions of elasticity may be thus determined with all the rigour de- sirable." The authors seem here to indicate that the investigation of torsional elasticity would yield information to be compared with that obtained by Wertheim for Young's modulus, and this impression is confirmed by the remark which occurs later, that the results obtained " show no trace at all of the remarkable phenomenon of a maximum, alluded to at the beginning of this article, which would seem to be indicated for iron by the investigations of Wertheim." The Young's modulus is an essentially composite one, involving both the rigidity-modulus and the * ' Annales de Chimie ct de Physique,' tome 12, 1844. t ' Pogg. Ann.,' bd. 141, 1870, or ' Amer. Jour. Sci.,' rol. 50, 1870. On the Effects of Changes of Temperature on Metal Wires. 181 bulk-modulus, and it is impossible, except on strict experimental evi- dence, which has not so far been forthcoming, to suppose that there is any invariable numerical relation between the two latter moduli. Wertheim, it may be remarked, imagined that he had found evidence of an increase of the Yoiing's modulus from 0° C. to 100° C. and a diminution from 100° C. to 200° C. This result is negatived, as Messrs. Kohlrausch and Loomis notice, by the fact that if "two tuning forks are in vibration, and one of them is heated, the number of vibrations changes in the manner demanded by the assumption of a decrease of elasticity for increasing temperatures." It was found by Messrs. Macleod and Clarke,* in an experiment on the change of frequency of a tuning fork produced by alteration of temperature, that the period was increased by the fraction 11 x 10~5 for one degree of rise of temperature. The amount of this change due to expansion was very small in comparison with that due to diminu- tion of the Young's modulus. For the linear expansion of steel is about 1*2 x 10~5 per degree, and as the period of a fork is altered in the direct ratio of the square roots of its corresponding linear dimen- sions before and after the expansion, if there is no change of modulus, the period was augmented by the expansion, by the fraction 0-6 x 10~5 for each degree rise of temperature. Hence the increase of period due to diminution of the Young's modulus was about 10'4 x 10~5 for each degree rise of temperature. And as the period is inversely as the square root of the Young's modulus, the fractional diminution of the Young's modulus must have been twice this amount or 20'8 x 10~5 for each degree rise of temperature. It will be seen below that the change of Young's modulus for mild steel is, according to our experi- ments, a diminution of about the same amount, though no doubt the change may be very different for different specimens of material. In the 'Philosophical Magazine' for June, 1899, Mr. G. A. Shake- spear has described an application of an interference method to the investigation of Young's modulus for wires, and has given values of the temperature-changes found by measurements made in this way. Thus it was found that the final value of the Young's modulus was for copper, iron, steel, and hard brass lower at the higher temperature than that at the lower temperature by the respective percentages 3 '6, 1*6, 3'2, and 3. Also it was noticed that repeated heating and cooling of the specimen seemed to produce an augmentation of the tempera- ture-change of the modulus, the material apparently settling down to a steady state. The method adopted by Mr. Shakespear was to elongate the speci- men of wire (which was in each case about 28 cm. long and 0'75 mm. in diameter) by applying a weight of about 2 kilos. The elongation changed the difference of path of two rays from the same source, and * 'Phil. Trans.,' vol. 171, Part I, 1880. Fie. i. Source Pi-.f. A. Gl«y, Mr. V. .). lilyth, an.l Mr. -I. S. I Mini,,],. so produced a rhan^e in the number of interference bands, which measured the retardation of one ray relatively to the other. The elongation was thus measured by the number of interference bands which passed across the field of view of the observing apparatus. The weight being kept hanging on the wire, the temperature was alter- - nately raised and lowered a num- ber of times until the number of bands measuring the change be- came approximately constant. It was noticed that the first heating produced a diminution in the number of bands measuring the elongation, and the succeeding heatings an increase which finally settled down to a constant value. In our experiments the method adopted was entirely different. The specimens were wires of radius varying from about 0-036 cm. to 0'06 cm., and each had a length of about 5 metres. The exact dimen- sions are given in the tables of results below. The experimental arrangement followed in the case of each wire was practically the same. The wire was hung vertically in the laboratory from a plate at the top screwed firmly to the lower side of a wooden beam in the ceiling. The upper end was secured by being passed through a hole in the plate and firmly soldered at the back. The wire passed along the axis of a double-walled tube or jacket (fig. 1), consisting of two coaxial cylinders of tin-plate of diameters 3'5 cm. and 5 cm. re- spectively. The upper and lower ends of the jacket were closed by corks, through which the wire loosely passed. A scale-pan weighing about 4 Ibs. was hung on the lower end of the wire to receive weights for the stretching experi- ments, and served when no other weight was applied to keep the wire taut. Sink On the Effects of < '//"////>:* of Tcnifxr^ttirc on Metal Wires. 18.'> Steam was generated under atmospheric pressure in an ordinary boiler, and was led into the upper end of the space by what is called in the diagram the " source-pipe," while what remained uncondensed escaped by the " sink-pipe " below. It was found that the space in which the wire hung was thus heated to a fairly uniform temperature. This temperature was determined by means of four thermo-electric couples of copper and German silver, the arrangement of which is shown in fig. 2. A German-silver wire passes along the axial space close to the FIG. 2. wire under test, and four copper wires are soldered to it at intervals as shown. These copper wires are led out along the axial space in which the wire hangs. Junction (1) is 25 cm. from the top of the jacket, junction (2) 150 cm. lower down, junction (3) 150 cm. still lower, and 1S4 I'1-..f. A. ' ir.iv. Mr. V. .1. r.lytl,. aii.l Mi. .1. S. Dunl-.j,. junction ( 1) J5 cm. al>ove the lower end <.f the jacket. An arrange- ment of mercury cups enables any of the thermal junctions to lie brought into the circuit of the galvanometer and the cold junction. 1 is of course taken to ensure that the wires leading from the junc- tions to the galvanometer are never in contact with the inner wall of the heating jackef or with the wire. Each thermo-electric couple before use, and from time to time after- wards, was carefully calibrated in the usual way by comparison with thermometers plared with the junctions in water, the temperature of which could be conveniently altered through the range over which the experiments extended. The results were expressed in curves from which the temperatures of the junctions were obtained in the experi- ments. The sensitiveness of the galvanometer was such as to give alxnit eight divisions deflection per degree of difference of temperature. This arrangement worked satisfactorily ; but in future experiments it will probably l»e replaced by a platinum thermometer. nf YHIHI v'x Mul ill n.<. Experiments were made for each wire, first as to the variation of Young's modulus with temperature, next as to the variation of the rigidity modulus with temperature, and, in connection with this latter determination in each case, of the rate of dying out at the different temperatures of torsional oscillations. The results of the last-men- tioned observations are, we think, very interesting, but we give here in the first place the information obtained with regard to Young's modulus. A scale-pan, weighing about 4 Ibs., was attached to the wire and loaded to such an extent as not to produce gradual increase of length of the wire. A small reading-microscope, with a convenient scale in its eye-piece, was rigidly mounted on a heavily loaded table resting on the solid stone floor of the laboratory, and was focussed on the point of a very fine needle soldered to the wire just below the lower end of the heater. The needle was set so as to be as nearly as possible at the proper end of the scale to give a displacement along the scale when the weight was removed from the wire. A similar arrangement at the upper end served to determine the amount by which the support yielded with different loads. This microscope was supported on a horizontal board firmly attached to two massive roof-girders at a distance from the beam to which the wire was suspended. Enough of the load was then removed to cause the needle to traverse nearly the whole length of the scale, and the readings of the top and l»ottom microscopes were observed. A series of readings for weights on and off were first taken at the ordinary temperature of the room, and again after steam had l>een blown for some time through the heat- On the Effects of Changes of Temperature on Metal Wires. 185 ing jacket. The temperature of the internal space was observed before and after each series of readings, and in most cases also while the series of operations was in progress. The distance between the needles was measured at the lower tem- perature ; the distance at the higher temperature was obtained with sufficient accuracy from the difference of temperature and the result of a determination specially made of the coefficient of linear expansion of the wire. This coefficient was determined during the heating of the wire, and, as the change of length was beyond the range of the micro- scope scale, a scale was fixed behind the lower needle, and the increase of length read off on it by means of a telescope. The radius of the wire was determined after the wire was finally taken down by means of a micrometer from twenty readings taken at different points by each of two observers. In two cases (wires II and III) the radius was determined by careful weighing of the wire in air and in water, and the results so obtained, properly corrected, agreed with those got by the micrometer method to within Taff per cent. Experiments were made for the following wires : — I, German silver ; II, mild steel ; III, brass ; IV, copper (commercial) ; V, copper (hard drawn electrolytic) ; VI, soft iron. At first everything did not work quite smoothly, the arrangement of the microscopes required a little adjustment, and the determinations of moduli for I were somewhat doubtful in point of accuracy, and for II were certainly in error. The apparatus was, however, got into thorough order before the experi- ments on III were begun,' and those on II were thereafter repeated. The following is the table of results for Young's modulus, in which is included the value obtained for German silver, notwithstanding the doubt as to its correctness, and the later value for mild steel, since there were good grounds for rejecting that obtained at first. The radii given in column VI were calculated for the higher temperatures from the expansions observed. The results of three separate determinations are given for the wire of soft iron (VI). Each elongation was obtained from the mean of about ten separate observations, agreeing to less than 0*5 of a microscope division. The elongations were obtained in micro- scope divisions, the value of which was 0-005948 cm. The difference between the second and third values of the coefficient of diminution for soft iron and the first of the three values, depends upon O2 of a microscope division. The degree of accuracy to which the deflections could be read was to 0-1 of a division. The lower value, 0-000136, is probably more nearly correct than the higher, 0-000197. Comparing the results with those obtained by Mr. Shakespear, re- duced for comparison, we have for steel 0'000247 as against his 0-00038, for brass 0-000373 as against 0-000352, for copper (electro- ! \ Mr. V. -I. Hlytli, and Mi. .!. S. l>unl.>i.. On the Effects of Chanr/cs of T<'nt/>rir'ture on Metal JVires. 187 Coefficients of Diminution of Young's Modulus with Temperature. Coefficient of Difference diminution Wire. Modulus. Difference. of of modulus temperature. per unit pei- degree 0. I 1-3046 0'03i5 66'6 0-000397 II 2 -1279 0*0340 64-8 0 -000247 Ill 1 '0257 0 -0287 70-8 0 -000373 IV 1-1132 0 -0129 74-9 0 -000155 V 1 -2954 0 -0416 73-7 0 -000436 ("(a) ... VI. 1 (6) ... 1-545 1 -5578 0 -0238 0 -0155 78-3 73-0 0 -000197 0 -000136 LOO... 1 5536 0 -0198 77-8' 0 -000136 lytic) 0-000436 as against 0-000414 for "a specimen of copper," and for soft iron 0*000166 as against 0-000157, so that there is fair agree ment. The difference between the coefficients for commercial copper (which was ordinary copper wire of the kind used by bell-hangers) and electro- lytic copper is very great, but it is not more striking than the difference between the electric conductivities of the two materials. It was thought desirable to make a series of determinations for soft iron with the view of finding whether any progressive change due to alternate and repeated heating and cooling disclosed itself as in Mr. Shakespear's experiments. Nothing of the nature of change of sign (see Mr. Shakespear's paper) was observed, though there was some in- dication of a diminution of the coefficient. Experiments on the Rigidity Modulus, and Observations on the Internal Viscosity of the Wires. Experiments were made on each wire, as has been stated above, to determine the change of rigidity produced by the alteration of tem- perature. These experiments were made by the torsional oscillation method. Cylindrical vibrators were attached to the wires such that the moment of inertia of the vibrating system was practically that of the cylinder only ; and observations were made for each wire with two separate vibrators of different weights, except in the cases of German silver and steel, for which one vibrator only was used, and electrolytic copper, for which four vibrators were used. The length of wire used was, with the vibrator attached, about 505 cm. in each case. The lengths and radii at the different tempera- tures were calculated from the expansions as in the Young's modulus experiments. The results of the experiments are given in the following table :— 1SS IW. A. Gray, .Mr. V. .1. lilytli. ;,iM Mr. J. S. \Vin-. W.-ight of vibrator. Period. Temp. Modulus in 10" (hue-* per sq. cm. 12-7 •J.1 -10 20-7 :\ -1256 U7 -i;:, ss :i 840M II Mild steel 18-7 35-71 17-0 7-'.' 3612 80-1 7-7701 f 12"-7 -'.» I.'. -I 3 -55 19 Ill Brass j 50-71 S7 7 3-4913 L 3"- 4 26 -58 1C, -1 :( •; 26-9 ^7 " 3 -3214 r 12-7 20-81 ]9'3 4 -4S22 J 21-07 4 '3549 L 18"'7 i :*440 25-98 '.'J-u 4-2230 C 12-7 Ifi-.'n; 23-0 1 2.">96 V. Copper, hard drawn, elec- troljtic •{ IS -7 1 „ 17-20 , '90-7 20-25 •!-•:> •2\--M 90-1 3 -9402 4-1908 3 •7»;i»2 L 24-7 23 -23 19 -8 4-1994 24-6 89-8 3 7305 r 12-0 1233 13-6 8-2855 „ 12-52 93-8 8 -0125 VI. Soft iron 4 18-0 15-13 15-G 8 '094 T 15-44 93-8 : 7601 ^ 24?-0 17 23 16-0 8'23OO " 17 -4G 95 -1 7-9903 The following table shows the coefficient of diminution of the rigidity modulus as deduced from the results given in the last table : — Wire. Vibrator. DilFerem-e of temperature. ( 'i I'lHcient of diminu- tion of rigidit> jier unit per degree C. I Ibs. 12-7 67'6 0 -000528 II 18-7 72-2 0 -000338 Ill 3-4 70-9 0 -000352 IV 12-7 12 -7 72-3 73 -1 0 -00< u 1 7 0 000392 V 18-7 12-7 72-9 67-7 0-000413 O'OOlll VI 18-7 ^4-7 12-7 67-6 70 0 MI •;? 0 -00149 0 -00160 0 '00041 18-7 2f7 7:>'l O'OO 0-00037 With respect to this table, it is to l>e observed that the results for soft iron and brass agree fairly with those obtained by Kohlrausch and Loomis for those metals, which were coefficients of diminution On the Effects of Changes of Temperature on Metal Wires. IS!) G'000447 and 0*000428 respectively. The result given above for com- mercial copper agrees also moderately well with that obtained by these experimenters, viz., 0'000520. But here, as indeed in all the substances examined, the differences in results are not surprising, in view of the probable difference of composition of the specimens. The great excess of the coefficient in the case of hard-drawn pure copper above the value for commercial copper is very noteworthy. It will be interesting to examine, as we propose to do immediately, the behaviour of copper wire soft drawn as well as hard, and to trace the effect, if any, of repeated heating and cooling. It was noticed that the rate of dying out of the torsional oscillations was very different at the different temperatures, being, except in one case, that of German silver, very much more rapid at the higher tem- perature than at the lower. This is undoubtedly the effect of a con- siderable increase of the internal viscosity of the wire with rise of temperature, for no practical difference existed in the immediate sur- roundings of the vibrator, which of course was the part mainly affected by the air in the vibrations. The top of the cylindrical vibrator in each case was about 8 cm. below the lower end of the heater, and the length of the cylinders, which were of brass, varied from 7 to 13 cm. There was no appreciable change of the air-temperature at the vibrator produced by the heater. This difference in the rates of subsidence of the torsional oscillations seems to us very remarkable, and, so far as we know, has not been observed before. It is shown in the diagrams numbered I — VI. \Xote added June 20, 1900. — The change in rate of subsidence pro- duced by alteration of temperature had, we have since found, been observed by Streintz and by Pisati (' Pogg. Ann.,' 153, 1874, and ' Gazzetta Chimica Italiana,' 1876, 1877, also ' Sitzungsb. cl. Wien. Akad.,' Ixxx, Abth. 2, 1879). In the first and last papers here cited, Streintz gives an account of his own work and of that of Pisati, principally as bearing on the effect which he called " Accommodation." In certain wires, e.g., steel, copper, silver, brass, and platinum, examined by Streintz and Pisati, a marked diminution of rate of subsidence was produced by keeping the wire for a considerable time in continual torsional oscillation. This is contrary to results obtained by Lord Kelvin at an earlier date, 1864-5 (Art. " Elasticity," < Collected Papers,' vol. III). Since learning of the investigations here referred to, we have begun to extend our experiments to the question of " fatigue of elasticity," and have found that the rate of subsidence appears to be, in some cases at least, a function of the temperature and of the amplitude of vibration. It seems possible thus to reconcile the discordance of re- sults thus referred to, and we hope to make a communication on the subject at an early date.] IW. .\. \lr. V. .1. lilvlli. ami Mr. -I. S. I>unl<.]>. In each riirvr tin- abscissae are amplitudes <>f vibration in degrees, an vibrator respectively, and in this order. Here it will be seen that the rate of subsidence is greater with each vibrator in the order of its weight ; that is, the larger weight in each case corresponds to a greater rate of subsidence, whether the temperature be the higher or lower of those used in the experiments. Also the rate of sub- VOL. LXVII. P 192 Prof. A. Gray, Mr. V. J. Myth, and Mr. J. S. 1 1 to to so SO An DIAGRAM III. ,, Brass Amplitudes 9 /0° £0° 30° 4Cf> &0° 60° TO8 8t X / ^ 7 /^ / 7 / // 7 " 2 t. ~urve I J. ; Temp* » A5 /°^ '.. 3 ^6. 1 'tbmCor * w * 4 j » 87-0°- /A 4°. 877°' 3 * 4? - * DIAGRAM IV(a). Copper (commercial). 10° £0° 30° 40° SO0 60° 70° 60° On the Effects of Changes of Temperature on Metal Wires. 193 sidence is in each case greater at the higher temperature than at the lower. It was observed by Lord Kelvin that increase of period as well as increase of stretching force affected the rate of subsidence, these effects being opposite. The effect of increase of stretching force increased the rate of subsidence. The effects of increased mass of vibrator given here are therefore mainly those due to increase of pull in the wire. We are, however, arranging to alter the period without altering the pull, and to alter the pull without changing the period, by using suitable vibrators. There is one characteristic of these curves for commercial copper which was shown also in the case of soft iron, but which does not appear in the other experimental results. The falling off of amplitude goes on very quickly in curves (3) and (4), and (5) and (6) in such a way that the curves are almost straight lines, until the amplitude has come down to only some 3 or 4 degrees, and then the subsidence becomes comparatively slow. The curves for the 3 Ib. vibrator show at the low temperature a logarithmic decrement increasing in the ratio 27 to 60 as the amplitude falls from 80° to 24°. At the higher temperature the logarithmic decrement is practically constant. DIAGRAM IV(i). Copper (commercitAL). O 60 JOO ISO 200 Diagram IV(&) shows subsidence at four temperatures which, taken in the order in which the experiments were made, were 10'2°, 51'1°, 53'3°, 78-4° C. In this diagram 150 divisions of abscissa represent 34°. Diagram V(«) gives the rates of subsidence for pure electrolytic copper at temperatures 23-2° and 92'1° with the 3 Ib. vibrator. Here there is no sign of the change of character of the curves of subsidence at low amplitudes which has just been noticed. The curve of subsi- dence at the higher temperature is indeed nearly an exponential curve at all amplitudes less than 20°. The curve for the lower tem- perature shows a logarithmic decrement at first somewhat quickly, p 2 194 Prof. A. Gray, Mr. V. J. Blyth, and Mr. J. S. Dunlop. diminishing then more slowly. The initial and final values are in the ratio of 73 to 44. Diagrams V(6), V(c), V(d) are the curves of subsidence at two tem- peratures— for the 12*7 Ib. vibrator, the 18'7 Ib. vibrator, and the 24'7 Ib. vibrator respectively. It will be observed that the curves of sub- sidence are nearly alike in the two sets V(6) and V(c), only a very slight difference in the direction of faster subsidence with the heavier vibrator being visible. The curves of subsidence at the higher tem- perature with the 24 '7 Ib. vibrator show distinctly more rapid sub- sidence at the lower temperature than in any of the other three cases. Diagram VI shows the curves of subsidence for soft iron. Curves (1) and (2) were obtained with the 12*7 Ib. vibrator; the former curve gives the subsidence at the lower temperature, the latter at the higher temperature stated in the diagram. It will be observed that the rate of subsidence at the higher temperature is very much greater than at the lower. These two sets of experiments were made in direct suc- cession to one another on May 15. Next, on May 16, two sets of experiments at temperatures 150-6 and 930>8 C. were made with the 18-7 Ib. vibrator. The results of these are shown in curves (3) and (4). Between these curves there is a much smaller difference than between (1) and (2), though the tempera- tures were very nearly the same. The high-temperature experiments were repeated on both days with reproduction of practically the same curve. After twenty-four hours nearly had elapsed the experiments at the lower temperature were repeated with the same vibrator, when the rate of subsidence was found to be much slower at 13° '5 C. than it had been on the previous day at 15°P6 C. On May 18 two curves (5) and (6) were obtained with the 24*7 Ib. vibrator at the respective temperatures 16° C. and 95°'l C. Curve (6) does not differ very widely from the curve (4) obtained on May 16 with the 18-7 Ib. vibrator; but the curve at the lower temperature lies much to the right of that (3) obtained on May 16 at the lower temperature, the companion curve of (4). All the results in curves (3) — (6) show rates of subsidence lying between those shown in (1) and (2). Curve (1) is approximately exponential. It will be observed also that curves (2) and (4) show the eomewhat rapid change of direction when the amplitudes have become small to which attention is directed above in connection with commercial copper. Curve (6), however, obtained at 95'1° C. with the 24'7 Ib. vibrator, shows no such change of direction ; but a considerable part of its curve, after the amplitude has been reduced to about 10° C., is approximately exponential in character. All the curves except (1) and (4'), however, show very marked diminution of rate of subsidence at the lower amplitudes. It seems certain that On the Effects of Changes of Temperature on Metal Wires. 195 DIAGRAM V(«). O /Q° so0 3O9 4O° 5O° 6O° 7O° 60 DlAGBAM V(6). Copper . 5 * BO SO 60 60 Cun-e \,zUb. 10 20 0 fO° £0° J0° 40° 30° 60° TOP 00° 60 Amplitudes. 60 Curve ao-4°C.i IB /ait. vibrator. DIAGRAM Copper (pure>. 30° 4O° 30° 60° 70° Curve I; temp 2+U). £4 ' 00° vibrator. On the Electrical Properties of Different Kinds of Glass. 197 DIAGRAM VI. — Soft iron. 70° BCf 90° AmpLitudes. O 10° 20° SO 40° 50° 60C the successive operations have had a considerable effect on the behaviour of the wire ; and experiments are being made to elucidate this effect as far as possible. Some other points of interest which were observed in these experi- ments are reserved for further investigation in a continuation of the work now in progress. " On the Connection between the Electrical Properties and the Chemical Composition of Different Kinds of Glass. Part II." By Professor ANDREW GRAY, LL.D., F.B.S., and Professor JAMES J. DOBBIE, M.A., D.Sc. Eeceived May 25, — Head June 21, 1900. In a former paper* we described experiments on the electrical qualities of specimens of glass of which the chemical composition was determined by analysis. Results were given for a lead-potash glass made by Messrs. Powell and Sons, of London, a lead-potash glass made by Messrs. Schott and Co., of Jena, a barium glass, and a zinc-soda glass (" Jena glass "), both made by Messrs. Schott and Co. These * ' Koy. Soc. Proc.,' No. 390, April 20, 1898. 198 Profs. A. Gray and J. J. Dobbie. specimens are referred to in the paper by the numbers XXI — XXIV. We have now to communicate the results of some further experi- ments, the object of which was to throw light on various points which had arisen in connection with the previous experiments, and to afford information as to whether the resistance or capacity of the glass was affected by the process of annealing, or varied with time. The speci- mens here referred to are numbered XXV — XXXII, and were all made, as nearly as possible, according to a previously prescribed com- position, and in the form of flasks, with long thick-walled necks, adapted for experiments by the direct-deflection method formerly employed. Full particulars of the different specimens are given in Table I below. The method of experimenting and the method employed were the same as those described in our former paper. The chief sources of error which had to be guarded against were, as before, surface conduc- tion, due to moisture on the surface of the glass, and leakage at other parts of the apparatus, due to the want of perfect insulation. The most careful watch was kept throughout the experiments against the possibility of inaccuracy from these causes, and tests were made in connection with each determination to make sure that everything was working correctly. The Resistance Experiments. Only a few days before the meeting of the Royal Society at which the paper referred to above was read, a rough test was made of the resistance of a flask (XXVII below) made for us by Messrs. Powell and Sons, which had approximately the same composition as Specimen XXI, the potash, however, being replaced by soda. It will be seen by a reference to our former table of results, that XXI was a lead-potash glass of very high specific resistance, certainly above 18000 x 1010 at 1 30° C. It was anticipated from our experiments that the substitution of soda for the potash in this glass would very greatly diminish the specific resistance, and it was stated when the paper was read that this conclusion had been verified. More accurate determinations made since that time have confirmed this result, as will be seen by a com- parison of Table I with the table given in the former paper. While XXI had the resistance at about 130° C. quoted above, the specific resistance of XXVII at about the same temperature was only 136 x 1010; so that the substitution of soda for potash in the composition of the glass diminishes the resistance of the glass to y^ of its former amount. The influence of the substitution of soda for potash is still more clearly brought out by a comparison of XXIX with XXXI, and XXX with XXXII. Specimens XXIX and XXX are lead-potash glasses, On the Electrical Properties of Different Kinds of Glass. 199 XXX and XXXII lead-soda glasses, in which the amount of soda is nearly equivalent chemically to the potash in XXIX and XXX respec- tively. Reference to Table I will show that in both cases the potash glasses have very much higher resistances than the soda glasses. Both XXI and XXVII were lead glasses ; but in the previous paper a glass was discussed which was made by Messrs. Schott and Co., and was composed mainly of barium oxide and alumina in combination with silica and boron trioxide. This, which we shall call the Jena barium glass, had a very high resistance, one quite unmeasurable, indeed, within the range of temperatures covered by the experiments. Moreover, it was found that the inductive capacity in a plate of this glass was exceedingly low, and that the glass showed little or no effects of dielectric polarisation. It was thought that it might be of interest to find whether the high resistance of this glass was associated with the presence of the large percentage of barium oxide. Accordingly, further experiments have been made on flasks of barium-potash glass manufactured by Messrs. Powell and Sons. This glass is numbered XXVIII of Table I. Its resistance is very low in comparison with that of either the lead-potash glass or the Jena barium glass, which, however, it must be remembered contained no potash. It was found to be subject to a somewhat rapid disintegration of its surface, and it probably differed in physical con. stitution from the Jena glass. The Jena glass, moreover, contains a considerable quantity of boron trioxide and alumina, which are absent from the glass made by Messrs. Powell. The presence in glass, how- ever, of a considerable percentage of potash with lead is, as shown by XXIX and XXX, consistent with a high specific resistance. It is interesting also to compare XXVIII and XXVI, which contain approximately the same percentage of barium and lead oxides respec- tively, and are otherwise very similar. The lead glass, XXVI, contains soda, from which the barium glass, XXVIII, is free ; but in spite of this, the resistance of XXVI is about three times that of XXVIII at the same temperature. Comparison with Ordinary Glass. — Experiments were next made to test how the glasses of specially prescribed chemical composition, already experimented on, compared with the ordinary kinds of glass used in the construction of apparatus. A common lime glass, XXV, and common lead glass, XXVI, were examined. The resistances of both of these were low ; that of XXV was low in comparison with the resistance of any other glass in the table.* As has been pointed out, * It may be noticed here that the first experiments on the electrical properties of specimens of glass, which were afterwards subjected to chemical analysis, seem to be those described in Mr. T. Gray's paper, ' Roy. Soc. Proc.,' vol. 34, p. 199. The analyses of the specimens (which were of glasses used for different purposes in the arts) were made with great care by Professor Divers, of Tokio. 200 Profs. A. Gray and J. J. I)oM»i«-. however, the resistance of XXVI considerably exceeded that of the barium glass, XXVIII. Effect of Annealing. — The last four specimens mentioned in Table I (XXIX — XXXII) were experimented on to find the effect of an- nealing on the resistance of glass. It is well known that in most conductors, especially pure metals, the effect of annealing is to diminish the specific resistance. XXIX and XXX were specimens of Messrs. Powell's lead-potash glass. Of these, XXIX and XXXI were carefully annealed in the usual way by Messrs. Powell : XXX and XXXII were left in the unannealed state. The results are given side by side in Table I, and show that the effect of annealing glass is very greatly to increase its specific resist- ance. In the case of XXX and XXXII, the lead-soda glasses, the specific resistance has been raised to three times its former value. Annealed glass is therefore a much better insulator than unannealed Variation of Resistance with Time. — The question of variation of resistance with time has been investigated by testing flasks, which had been set apart for the purpose, at intervals of about six months. The results are shown in Table II. XXI, XXVII, and XXVIII have had their resistances determined three times, and so far no change has dis- closed itself. As noticed above, the surface of the glass XXVIII seems to become disintegrated in course of time ; for the surface, which was originally cleaned till quite clear, has gradually acquired a milky appearance, and is now quite opaque. TJte Capacity Experiment*. The specific inductive capacity of the glass of these flasks was determined by the method described in our former paper, and the results are shown in Table I. It will be seen that the specific inductive capacity of Powell's lead-soda glass, XXVII, is rather low in com- parison with that of the corresponding lead-potash glass, XXI, which was about 8. It is noteworthy, however, that the lead-soda glass was free from the dielectric polarisation effects which were so troublesome in the case of XXI. Of the glasses discussed in the present paper, XXVI, one of the ordinary glasses experimented on, was the only one that showed dielectric polarisation conspicuously. No trouble from dielectric polarisation was experienced with Nos. XXIX — XXXII, so that annealing does not cause any marked differ- ences in the " electric absorption " of the glass. On the Electrical Properties of Different Kinds of Glass. 201 Measurement of Residual Tmst of Glass Fibres, The object of these experiments was to compare fibres of the glasses, on which the electrical experiments already described had been carried out, as regards imperfection of their torsional elasticity. The figure shows a sketch of the apparatus used. A long cylindrical glass tube (a), about 23 inches in length and 2 inches in diameter, was fixed permanently in a vertical position in a wooden frame, as indi- cated in the diagram. A tightly-fitting cork closed the upper end of the tube, and through a narrow slit in this passed a rectangular piece of copper foil, from the centre of the lower edge of which projected a short tag of the foil. This tag was quite rigid, and carried attached to it with shellac cement the glass fibre, care being taken that the fibre was so placed that it hung vertically along the axis of the tube, when the cork was fixed in position. The lower end of the fibre was attached in a similar manner to a cross-piece, c, weighing about 1 gramme, and having a length of about If inches. The fibre was attached to the upper end of the cross, and to the lower end was fixed a small mirror. The lower end of the glass tube fitted into a groove cut in a wooden sole-plate, capable of being rotated about a vertical axis, which coincided as nearly as possible with the axis of the tube, that is, with the position of the fibre. Two brass pins (&) were fixed vertically at two points in the sole-plate within the tube, and, projecting upwards, stood one on one side the other on the other side of the horizontal arm of the cross-piece. By turning the sole-plate round, any required twist could be given to the fibre, since the tube was held fast in its supporting frame. A lamp and scale set in front of the mirror enabled the position of the lower end of the fibre to be observed. 202 Profs. A. Gray and J. .1. D<.bbie. After fixing the fibre and allowing the arrangement to stand undis- turbed for a day or two to allow the cement to harden, the scale wjis placed in position with its ends at equal distances from the mirror. The zero position of the spot of light on the scale was observed, and the sole-plate turned round to give a total turning of the lower end of the fibre relatively to the upper of 360°. The couple was kept applied for thirty seconds, and then taken off gradually without any jerking of the fibre. This prevented torsional oscillation of the fibre after the removal of the couple. In fifteen seconds after the removal of the couple a reading of the position of the spot of light was taken. From this the angular deflection of the lower end of the fibre was obtained, and the angle so measured, divided by the length of the fibre, gave the residual twist. The following table gives the results in radians per centimetre of the length of the fibre x 104. The diameter of the fibre was found by weighing a known length of it, and calculating from the known density of the glass. No. of specimen. Diameter in centimetres. Residual twist. XXI 0-0109 2-09 XXII 0-016 1-84 XXIII 0-010 0-93 XXIV 0-0105 3-83 XXV 0-0115 7-19 XXVI 0-015 5-41 XXVII 0-0187 2-44 XXVIII 0-0105 5-52 XXIX 0-0142 0-83 XXX 0-0183 0-8 XXXI 0-0148 1-9 XXXII 0-0191 2-8 The experiments on Specimens XXI — XXVII inclusive were made in June and July, 1898, and it was then thought that there ap- peared to be some connection between the resistances and the residual twist of the fibres. But since XXI to XXIII were glasses whose resistances were too high for measurement, it was of course impossible to draw any numerical conclusion on the point without examining more specimens. A very little reflection, however, showed that no exact comparison was possible from this point of view, as the residual torsion was no doubt influenced in a very marked degree by the immediate previous history of the fibre. But it is noteworthy that the residual twist is very low in the case of the Jena barium glass and in Messrs. Powell's lead-potash glasses, viz., in XXIII, XXIX, XXX, and is abnormally high for the two ordinary glasses, XXV, XXVI. On the Electrical Properties of Different Kinds of Glass. 203 It was found that for the same fibre the same twisting couple pro- duced the same residual twist, provided the two determinations were not made in immediate succession. The rate at which the residual twist came out was very great immediately after the twisting couple had been taken off; and this rate diminished rapidly as the spot of light approached its zero position on the scale. Chemical Composition of the Specimens of Glass. The results of the chemical analyses made for the different speci- mens are stated briefly in Table I, which affords a conspectus of all the results of the experiments now described. The following notes regarding the different specimens, containing approximate empirical formulae for their composition, are, however, set down here. XXV. This is an ordinary lime glass. The alkalies were not esti- mated separately, and no formula can therefore be given for it. XXVI. This is an ordinary lead-alkali glass, containing consider- able quantities both of potash and soda. After deducting ferric oxide, alumina, and manganese, its composition may be represented by the empirical formula 43Si02, 5PbO, 5Na20, 3K20. Found. Calculated. Si02 59-60 60-18 PbO 26-44 26-00 Na20 7-44 7-23 KoO 6-50 6-57 99-98 99-98 XXVII. This glass is composed of silica, lead oxide, and sodium oxide, and is free from potassium. Allowing for the small amount of ferric oxide and alumina which it contains, its composition may be expressed by the empirical formula 10Si02> 3PbO, 3Na20. Found. Calculated. Si02 40-87 41-24 PbO 45-33 45-98 Na20 13-80 12-78 100 100 XXVIII. This is a barium-potash glass, free from lead, and con- taining only a very small amount of soda. After allowing for the 204 Profs. A. Gray and J. J. Dobbie. Table I. Resistance. Dimensions. Number of speci- men. Description of glass. Density. d ** thickness of bulb. « = effective surface. r = external radius. Tempe- rature, centi- Specific resistance in ohms x 10W. grade. XXV 2-487 d » 0-128 cm. 149° 0-202 * = 174 -614 sq. cm. 116 1-874 r = 3 -829 cm. 93 11-901 72 89-15 55 531-05 XXVI Lead glass, with pot- 2-99 d - 0 -0663 cm. 150° 8-535 ash and soda * - 188 -19 sq. cm. 140 18-64 r - 3-971 cm. 130 33-59 101 4-42 -70 88 195H -50 66 18034-0 XXVII Lead - soda glass 3-552 d = 0 -092 cm. 142° 136-5 made by Messrs. * = 211 -54 sq. cm. 116 797-3 Powell and Sons, r = 4 -226 cm. 90 5249 -0« London XXVIII Barium-potash glass 3-11 d - 0-127 cm. 138° 8*41 made by Messrs. * = 249 -1 sq. cm. 125 16-01 Powell and Sons r = 4 -566 cm. 95 178-05 77 1115-50 XXIX Lead - potash glass 3-41 d = 0 -086 cm. Resistance too big] made by Messrs. * = 295 -411 sq. cm. to measure. COT Powell and Sons r = 4 '422 cm. tainly above 2900< (annealed) at 140° XXX Lead - potash glass 3-34 d =• 0 '996 cm. 142° 1328-6 made by Messrs. * = 211 -286 sq. cm. Too high to me* Powell and Sons r - 4 -2014 cm. sure at lower tern'1 (unannealed) peratures XXXI Lead - soda glass 3-408 d = 0 -059 cm. 141° 4-874 made by Messrs. * = 226 -705 sq. cm. 122 20-497 Powell and Sons r -4 -3199 cm. 102 102 -820 (annealed) 84 515-94 XXXII Lead - soda glass 3-36 d = 0-0986 cm. 140° 1-691 made by Messrs. * - 213 -246 sq. cm. 120 4 -927 Powell and Sons r « 4 -1921 cm. 104 20 -821 (unannealed) 83 144-640 73 215-130 * This result at 90° is not accurate, b« On the Electrical Properties of Different Kinds of Glass. 205 Table I. Capacity. Chemical Composition. Tempe- rature, centi- grade. Specific inductive capacity. J 02 *l s g ^ £ o Fg i-2 "HI Potassium oxide. X 0 • l-t •!-( ra W o o 02 Ferric oxide and alumina. o fl i *-^ H32 11° 129 6-26 6-79 69-04 very small trace •• 7-59 19- 23 2-73 1-31 10° 130 7-06 7-90 58-82 26-10 • • trace 6-42 7-35 0-43 0-81 8° 130 5-42 5-69 40-75 45-19 •• •• •• 13-76 0-29 22° 147 6-93 7-18 52-28 •• 26 -09 •• 19-74 0-64 0-25 18° 7-22 140 7 -42 48-25 40-80 •• •• 10-65 •• 0-29 18D , 140 6-76 7-05 50-11 39-74 •• •• 10-03 0-12 20° 140 8-013 8 302 50-42 40-24 •• •• trace 8-77 0 38 19° [ 130 7-376 8-44 51-46 38-94 •• •• •• 9-13 0-25 serves to show the order of the quantity. 206 Profs. A. (Jray and J. J. Dobbie. Table II. Number of tpecimen. Dimensions. d = thickness of bulb. * — effective surface. Date of test. Tempera, turo, centi- grade. Resistance. Specific resistance in ohmB x 10 w. XXI April 27, 1898 Up to 140° Resistance too high to mea* sure. Dec. 15 „ „ July 3, 1899 XXVII d = 0 -085 cm. {138° 106 -17 g = 215 '15 sq. cm. Mar. 31, 1898 107 90 1609-3 Resistance too high to mea- •ure. f!42 73-3 Mar. 19, 1898 < 119 505 -35 I 90 Resistance too high to mea- sure. fl44 72-28 July 4, 1899 < 115 I 90 535-9 Resistance too high to mea. : sure. XXVIII d = 0 '158 cm. {145° 2-37 * = 230 -686 sq. cm. June 7, 1898 114 93 23-32 152 -79 72 1039 -25 fl30 9-35 Jan. 13, 1899 { 106 6177 I 89 296-97 (142 5 -045 July 5, 1899 105 91 -56 85 461-64 XXIX d = 0-125 cm. July C, 1899 Up to 140° Resistance too t = 204 -306 sq. cm. high to mea- sure. XXX d = 0-1135 cm. July 7, 1899 140° 1777*5 t = 210 -14 sq. cm. Could not be measured at lower tem- perature. XXXI d = 0-0483 cm. July 11, 1899 136° 7 '533 » = 219'5 sq. cm. 100 102 -04 89 316-40 XXXII d = 0-1438 cm. July 10, 1899 141° 0 -8502 * = 208 '8 sq. cm. 110 8-175 i 69 298 -95 Oil the Electrical Properties of Different Kinds of Glass. 207 small quantities of ferric oxide, alumina, and sodium oxide, its compo- sition may be represented by the formula 20SiOL>, 4BaO, 5K20. Found. Calculated. Si02 53-29 52-54 Bad 26-59 26-87 K,0 20-12 20-58 100 99-99 XXIX and XXX. Both of these glasses are from the same pot, XXIX being annealed, and XXX unannealed. Although they yielded slightly different numbers on analysis, they are essentially the same, and may be represented by one formula, viz., 23SiOo, 5PbO, 3K-.0. Found. XXIX. XXX. Calculated. SiO, 48-39 50-17 49-69 PbO 40-92 39-78 40-15 K20 10-68 10-04 10-15 99-99 99-99 99-99 XXXI and XXXII. These two glasses correspond, closely to XXIX and XXX, soda, however, being substituted for potash. It was hoped, in preparing them, to obtain a glass closely corresponding to XXIX and XXX, soda being substituted for potash in exactly equivalent quantity ; but Messrs. Powell found that in order to make the glass Avorkable, it was necessary to add a slight excess of soda. XXXI was annealed ; XXXII remained unannealed. The composition of both may be expressed by the formula 24Si0.2, 5PbO, 4Na20. Found. SiO-> 'xxxi. 50-70 XXXII. 51-7 Calculated. 51-37 PbO 40-47 39-12 39-77 Xa.,0 ... ... 882 9-17 8-84 99-99 99-99 99-98 We desire, in conclusion, to express our thanks to Messrs. Powel. and Sons for their kindness in preparing glasses for us, and to Messrs. 0. W. Griffith and Robert Abell for their help in the electrical and chemical experiments respectively. VOL. LXVII. 208 i .'•. • .-iiiil K. 'I'. .' " On the Change of Resistance in Iron produced by Magnetisation." I'y ANDREW GRAY, LL.D., F.R.S., Professor of Natural Philo- sophy in the University of Glasgow, and EDWARD TAYI.OI; JONES, D.Sc., Professor of Physics in the University College of North Wales. Received May 30,— Read June 21, 1900. The experiments described below were made with the object of determining simultaneous values in a specimen of soft iron wire of the magnetising force, the magnetisation, and the change of resistance due to magnetisation. In all the measurements hitherto made of magnetic- changes of resistance no attempt seems to have been made to deter- mine at the same time the magnetisation ; in fact, the results which have been obtained for bismuth might appear to indicate that there is but little, if any, connection between magnetisation and change of resistance. The results herein described, on the other hand, indicate that there is a somewhat close connection between the two phenomena, and make it clear that further measurements on similar lines would have some value. Mo'., q (fig. 1), and two nearly equal coils A, B in a, b. A and B were both kept in one bath of oil. All connecting wires were of thick copper. The galvanometer was a Kelvin low-resistance astatic instrument. The magnetising current was measured by a Kelvin graded galvanometer, standardised by electrolysis and by comparison with a Kelvin deci- ampere balance. In making observations, the iron coils P, Q were first demagnetised by reversals, and the position of the sliding contact key S found, which gave no galvanometer deflection. Then a weak magnetising current was applied, reversed several times, and the position of S again found. The change of position of S indicated that the resistance of the longitudinally magnetised wire became greater than that of the transversely magnetised wire. This was repeated for a large number of field-strengths, each greater than the one preceding it. The resistances of the auxiliary coils A, B were so chosen that the change of position of S with the greatest mag- netising field used amounted to about 16 cms. The difference of the /AP AQ fractional increments of resistance of P and Q ( -^- - -^- = A<£ j was \ * Vs / calculated from the observed displacement of S from its zero position, and the results given below represent values of 8 for different field- strengths. 210 Tints. A. day und K. T. J-mes. Measurement <>f The magnetisation of the iron was determined in separate experi- ments. For this purpose a narrow glass tube, 60 cm. long, was filled with a number of lengths of iron wire, cut from the same specimen, from which the insulation had been removed. A number of turns of fine insulated copper wire was wound on the glass tube near its middle and connected to a ballistic galvanometer, standardised by solenoid and secondary coil. The tube, with the lengths of iron wire, was placed within the magnetising coil, and the BH curve of ascending reversals was determined for the iron in the usual way, and the mag- netisation I calculated from the equation I = B - H/4jr. No account was taken of the demagnetising factor of the iron wires ; the factor for a cylinder of the same length and of cross-section equal to the sum of the section of the thirteen pieces of wire in the glass tube is but of the order 0-0008. Results. The following are particulars as to the coils, &c., used : — Diameter of soft iron wire used in. resistance and magnetisation experiments ........................ 0'0745 cm. Resistance of iron coil P (longitudinally wound) ... 1 -045 ohms. Resistance of iron coil Q (spirally wound) ......... 1 • 1 1 8 „ Do., do., auxiliary coil A .............................. .1 -382 „ Do., do., auxiliary coil B .............................. 1 -476 „ The coils A, B were of German silver. Resistance per centimetre of bridge wire (a-) •= 0*0000536 ohm. The difference of the fractional increments of resistance of P and Q was calculated from the approxi- mate equation AP AQ A + B **-T— T •AB-"*' where 8x is the displacement of the contact S (fig. 1) from its zero position, i.e., from its position when the wires were not magnetised. The mean temperature of the iron wire during both the change of resistance and the magnetisation measurements, determined by fre- quently observing the temperature of the water entering and leaving the magnetising coil, was 5'5° C. The results of all the measurements are shown in the accompanying curves. Fig. 2 is the magnetisation curve (I, H, c.g.s.) of ascending reversals. Fig. 3 the curve B, H for the longitudinal coil. Fig. 4 shows A as a function of the magnetising field H. The general resem- blance between this curve and the magnetisation curve suggests that the change of resistance depends on the magnetisation rather than on magnetising force. Change of Resistance in Iron produced by Magnetisation. 211 Sg O •o I'll' 15 Profs. A. Gray and E. T. .1 Fio. 4. 10 so too ISO 200 JBO ZOO 550 Figs. 5, 6, 7 show A<£ as a function of B4, I4, Ili respectively. Of these, fig. 6 approximates closely to a straight line except in the neighbourhood of the origin. It may thus be stated as an empirical result for the specimen of soft iron and for the range of magnetisation employed in these experiments that the change of resistance is ap- proximately proportional to the fourth power of the magnetisation. Note added, June 18. — The results are complicated to some extent by hysteresis. It was found that when the iron wires were first thoroughly demagnetised, and after having been left long enough to take the temperature of the surroundings, were subjected to a rather strong magnetising current, kept in for less than a second, so that no appreciable heating could arise, about a third of 8 remained after the magnetic force was removed. The above experiments were carried out in the Physical Laboratory of the University College of North Wales, and we wish to take this opportunity of expressing our high appreciation of the value of the assistance of two students of the College, Mr. Guy Barlow and Mr. Godfrey Rotter, who, by making many measurements and calculations, enabled us to complete the work. Cka/if/e of Resistance in Iron produced ~by Magnetisation. 213 75 FIG. 5. JO B** 5 IO 214 A- Q j !•- • FIG. 6. 15 10 I***' Change of Resistance in Iron produced ly Magnetisation. 215 FIG. 7. 10 T6 16 I X/0 . 6 6 •Jin Dr. J. S. r.olton. 2 t Hbtologicdl Localisation " The Exact Histological Localisation of the Visual Area of the Human Cerebral Cortex." By JOSEPH SHAW BOLTON, !'• ^ M.D., B.S. (Loncl). Coinnmnicated by Dr. MOTT, F. ! i:« « .-iveil May 11, — Head June 14, 1900. (Abstract.) Previous Research. The previous research concerning the human visual area has l>een carried out in three directions. (1) The study of lesions causing blindness. (2) The study of the myelination of the corona radiata. (3) The histological examination of " occipital " or " calcarine " cortex as regards — (a) Cell form. (b) Subdivision of this variety of cortex into layers. (c) The modifications caused in (a) and (b) by long-standing blind- ness. > Examination of the literature on the first two subdivisions demon- strates the extreme diversity of opinion which exists regarding the situation of the primary visual area of the cortex. The object of the present research has been to indicate the exact region of the cortex to which the visuo-sensory function is limited. For this purpose it has been unnecessary to pay attention to the special neuronic structure of this portion of the cerebrum, but the general histology of the cortex referred to in (3), (b), and (c) has been considered minutely in the third section of this paper. The Exact Distribution of the " Occipital " Lamination. (1) The " occipital " lamination in the region of the calcarine fissure has been histologically mapped out, in six normal and pathological brains, as a well defined cortical area. (2) The general distribution of this area is as follows. It occupies — (a) The body of the calcarine fissure, including the anterior and posterior annectants, and extending upwards to the parallel cuneal sulcus and downwards to the collateral fissure. (b) The posterior part of the calcarine fissure extending to the polar sulci surrounding its extremities. ('•) The inferior lip of the stem of the calcarine fissure (including the superficial surface and lower lip of the cuneal annectant) nearly to its anterior extremity, just posterior to which the area tails off to a sharp point. of the Visual Area of the Human Cerebral Corte>. 217 (3) The approximate outline of this area is consequently pear shaped with the apex anteriorly and the thick end at the pole of the hemisphere. (4) The area is much decreased in extent, but not in distribution, in cases of old-standing optic atrophy. (5) In anophthalmos the area is much contracted as regards both extent and distribution. It occupies the usual position in the stem of the calcarine fissure, but only extends backwards as far as the posterior cuneo-lingual annectant, and it is confined to a portion of the inferior lip of the fissure and to the cortex between this and the collateral sulcus. The General Histology of the Cortex Cerebri in the Region of the Calcarine Fissure. (1) The following classification of layers has been adopted for the purposes of micrometer measurements : — (rt) The cortex of the area of special lamination which has just been described. I. The superficial layer of nerve fibres. II. The layer of small pyramidal cells. Ilia. The outer granule layer. Illb. The middle layer of nerve fibres, or line of Gennari. Illr. The inner granule layer. IV. The inner layer of nerve fibres. V. The layer of polymorphic cells. (&) The cortex surrounding the area of special lamination. I. The superficial layer of nerve fibres. II. The layer of small and large pyramids. III. The layer of granules. IV. The inner layer of nerve fibres. V. The layer of polymorphic cells. At the junction of these two varieties of lamination an abrupt change takes place, the line of Gennari suddenly ceasing, and the outer granule laj^er joining the inner one, the conjoined layer being approxi- mately of the thickness of the former outer layer. (2) The average of very numerous micrometer measurements of the cortex of the area of special lamination and of the neighbouring con- volutions gives the following results : — (a) In the area referred to, in cases of old-standing optic atrophy, the line of Gennari is decreased nearly 50 per cent, in thickness, and the outer granule layer more than 10 per cent. 218 Dr. A. A. Rambaut. £//>• . rature at (b) On the other hand, in the cortex surrounding the area nf. to, old-standing optic atrophy causes no modification of the lamination. (r) In anophthalmos the conjoined outer granule layer and line of Gennari (for the granules in the former layer are not suffi- ciently obvious to admit of easy micrometer measurement alone) are narrowed down to two-thirds of the normal thick- ness, the other layers of the cortex being approximately un- changed. This amount of narrowing is the same as that found in cases of old-standing optic atrophy. ((/) The majority of the layers of the cortex either inside or outside the area of special lamination do not vary appreciably in thick- ness as a result of age or chronic insanity, but there is an almost exact correspondence between the thickness of the con- joined first and second layers cf the cortex and the degree of amentia or dementia existing in the patient. Summary of Ctiii<-lii*i<>n* tJ mien from the present Reseur<-}t. (1) The area located and described in this paper is the primary visual region of the cortex cerebri. (2) The part of this area to which afferent visual impressions primarily pass is the region of the line of Gennari. (3) A marked contraction of the area in both extent and distribu- tion, without absence of the line of Gennari, occurs in anophthalmos. (4) This area can probably be described as the cortical projection of the corresponding halves of both retinze. In this projection the part above the calcarine fissure represents the upper corresponding quadrants and the part below the lower corresponding quadrants of both retinse. " Underground Temperature at Oxford in the Year 1899, as determined by Five Platinum Resistance Thermometers." By AKTHUR A. RAMBAUT, M.A., D.Sc., Radcliffe Observer. Com- municated by E. H. GRIFFITHS, F.R.S. Received May 17, — Read June 21, 1900. (Abstract.) I. Description of the Appuratus. The instruments with which the earth-temperatures given in this paper were observed were five platinum resistance thermometers of the Callendar and Griffiths pattern. The thermometers were inserted in undisturl>ed gravel, the first four as determined ~by Five Platinum Resistance Thermometers. 219 lying one under the other in a vertical plane beneath the grass of the south laAvn of the Radcliffe Observatory, and within a few feet of the Stevenson's screen in which the dry bulb and the wet bulb, the maxi- mum and minimum thermometers, are suspended. A fifth thermometer was subsequently placed at a depth of about 10 feet in a separate pit. The actual depths of the various thermometers as measured in October, 1898, were as follows : — Thermometer. I. II. III. IV. V. Depth 6| iii. 1 ft. 6 in. 3ft. 6£ in. 5 ft. 8| in. 9 ft. 11^ in. The resistance box is in its general design similar to that described by Mr. Griffiths,* but simplified to suit the particular class of work for which it was intended. It is provided with three principal coils, A, B, and C, whose nominal values are, 20, 40, and 80 box units respectively, a box unit being about 0*01 ohm. The apparatus is provided with a slow motion contact maker, of Mr. Horace Darwin's pattern,! and Mr. Griffiths's thermo-electric key 4 In the standardisation of the apparatus the method described by Mr. Griffiths, in his article in ' Nature,' referred to above, was in the main followed. The temperature coefficient was determined by Mr. Griffiths, in his own laboratory at Cambridge. Two separate series of observations led to the following results : — Range of Temperature Date. temperature. coefficient. July 27 9-18° 0-000242 Augusts 12-51 0-000240 The value actually used in the reductions was 0-00024. From observations made when the instrument was mounted in situ at Oxford, the values of the coils were found to be C = 80-1581 B = 39-979 S-mean box units, A - 19-863J and one scale division of the bridge wire is equal to 1-0134 mean box units. One of the most important considerations in connection with this subject is the degree of permanence in the fundamental points, as determined at considerable intervals of time; but the process of standardisation is not one that can be very frequently applied. * ' Nature,' rol. 53, November 14, 1895. t ' Nature,' vol. 53, November 14, 1895. t ' Phil. Trans.,' A, pp. 397-8, vol. 184 (1893). 220 I >r. A. A. li'.mil'iuit. du All the instruments wore very carefully standardised by means of observations extending over three days, in October, 1898, and on October 6, 1899, taking advantage of a visit from Mr. Griffiths, I had the 6-in. thermometer dug up, and we examined its zero point after exactly a year's continuous observations. The readings agreed to within 0-004° C., being In 1898 0-306 In 1899 0-302 For reasons given in the paper, it was not thought necessary to re-examine the boiling point. For another thermometer (A), kept in the observing room, the fundamental interval was found to have remained practically unchanged, being In 1898 101-067 And in 1899.. 101-059 II. Discussion of the Observations. The first step in the discussion of the observations is to group them into monthly means, and thence to deduce the harmonic expressions which will represent the readings of each thermometer throughout the year.* These monthly means expressed in degrees Fahrenheit are given in the following table : — Mean Monthly Temperature of the Ground at the Radcliffe Observatory, Oxford, 1899. Thermometer 1 2 3 4 5 Depth 6.\ in. 1 ft. 6 in. 3 ft. 6} in. 5 ft. 8} in. 9ft. Hi in. 40-47 42°07 44?68 46°80 49°97 40-09 41-34 43-25 45-08 48-33 41-34 41-91 43-21 44-74 47-42 48-77 47-66 46-61 46-40 47-37 May 54-86 52-95 50-96 49-54 48-53 66-73 62-89 58-29 54-73 50-88 JUly 69-43 65-93 62-15 58-74 53-85 69-23 67-79 64-88 61-66 56-39 59-34 61-12 62-03 61-19 57-80 48-99 51-14 54-29 56-18 56-71 46-73 48-43 50-99 52-69 54-48 38-14 41-08 45-33 48-58 52-33 * Professor W. Thomson, " On the Reduction of Observations of Underground Temperature," ' Trans. Roy. Soc. Edin.,' vol. 22, p. 409. as 52-013 -12-932 -3-128 + 2-130 +0-976 13-305 256 24-3 2-343 65 23 -3 1 52-010 -15-337 -1-507 + 3-017 + 0-511 15-411 264 23 -2 3-060 80 23-2 Air 50 -396 -12-776 -1-621 +2 -847 + 1-410 12-878 262 46-1 3-177 63 39-2 From each wave as observed at any pair of thermometers we obtain two determinations of the diffusivity (&) of the gravel, one from the diminution of amplitude and the other from the retardation of phase. In computing the value of the expression ^/(ir/k), the Paris foot and the Fahrenheit degree have been used. I have omitted the results for the thermometer No. 1 (6 inches), which are too much affected by the diurnal changes and other causes. From six comparisons of the amplitude and retardation of the annual wave at the remaining four thermometers we obtain twelve determina- tions of the value of v/(;r/&), the mean of which is O'l 189. For the half-yearly wave the mean value obtained in a similar way is O'l 187. This close agreement of the mean values of x/(7rA") derived from the annual and half-yearly waves is very remarkable, and seems to indicate a high degree of precision in the results. The paper deals with the observations of a single year, and the results accordingly exhibit some discrepancies between theory and observations which, although they are less than might have been expected, are greater than one would like to see. These discrepancies are due partly to the fact that the temperature variations are not strictly of a periodic character, as the theory supposes, and as such they might be expected to be diminished in the mean of a number L'L'L' I'i'ii'. K. Pearson. OH ik* Kinetic Accumulation qf 8tr of years, and partly to irregularities, physical and formal, i?i the surface of the ground. Another source of irregularity affecting previous observations of this sort, namely, thermometer errors arising from the uncertainty as to the temperature of the liquid in the long stems of the mercury or alcohol thermometers, does not in this case apply ; and if other errors peculiar to the platinum thermometers exist, they seem to be confined within much smaller limits. " On the Kinetic Accumulation of Stress, illustrated by the of Impulsive Torsion." By KARL PEARSON, F.R.S., Profe- of Applied Mechanics, University College, London. Received May 29,— Read June 21, 1900. (Abstract.) 1. It is usual in engineering practice to double the value of the stresses, calculated statically, when a live load comes onto a girder ; and further various empirical laws, such as those due to AVohler, are adopted in the case of repeated loading to measure the effective resistance of a structure. While these methods, practically adopted, show very clearly that there is a just appreciation that loading varying with the time differs in its nature very considerably from purely permanent loading, they yet fall considerably short of the defmiteness required from the theoretical standpoint. Occasionally it must be confessed that they would fail even from the practical standpoint were it not for the large factor of safety usually adopted. So soon as a live load comes onto a girder, even without impulse, vibrational terms arise in the strains, and the same thing occurs also in the parts of machinery subjected to external forces changing with the time. The discussion of the strains in a girder due to a rolling load was first undertaken by Sir George Stokes in 1849,* and his results have been considerably extended in later papers by Phillips, Renaudot, Bresse, and de Saint- Venant.t The latter has further dealt with a considerable number of problems of what I have elsewhere termed wm-impul*ir? /'>//////" ,+ as well as a variety of cases of impulsive resilience in the case of bars receiving longi- tudinal or transverse impacts.§ The numerical results of Saint- Venant's papers, as well as his graphical representations, hardly seem • See ' History of ElaMicity,' vol. 1, arts. 1276 and 1417. t IMC. cil., vol. 2, art?. 372 — 382. J Loc. rif., vol. 2, arts. 355—357. § Loc. cit., vol. 2, arts 401 — 414. (For the history of the subject see art. 341.) illustrated by the TJieory of Impulsive Torsion. 223 to have been sufficiently considered in the light of practice. They show that there can be a kinetic accumulation of stress at sections where a wave of vibrations is reflected, and that in many cases this kinetic stress can considerably exceed double the value of the statical stress at that section due to the same load permanently applied. In fact, every case of repeated loading or live-loading forms a kinetic problem which must be independently solved, and which, if it presents difficulties, still presents difficulties such as the mathematician is in duty bound to overcome. While the papers to which I have referred give a fairly complete solution of a number of problems in longitudinal and transverse loading of bars and girders, the problem of torsional loading seems to have been untouched up to the present. Yet the question of torsional vibrations is one that arises very frequently, as in the cases of shafting and axles. The present memoir endeavours to give from the mathemati- cal standpoint a fairly comprehensive solution of the problem of the kinetic accumulation of stress in bars, shafts, axles, &c., owing to repeated, impulsive, or changing systems of torsional loading. The attempt — following the traditions of de Saint- Venant — has been made, however, not to leave the results in the form of long series unintelligible to the average practical man. A very large amount of numerical reduction has been undertaken, and many of the results are shown graphically. The whole of the calculations, as well as the original large diagrams, are my own work ; but I owe to Mr. G. Baker and Mr. J. Longbottom, who have at one time or another been my assistants at University College, the reduced diagrams which accompany this paper, and I have to cordially thank them for the care and labour they have given to the reproduction of my drawings. 2. With regard to the general conclusions of the memoir, I should wish to draw special attention to the following points : — (rt.) When a load is repeated, or applied, reversed, and repeated, there will in general be kinetic accumulations of stress ; the amount of this accumulation varies with the times of incidence and of release of the load, but it may easily exceed the double of the stress due to the statical application of the load. (b.) Wohler's empirical laws must either be considered as allowing for this kinetic accumulation, or not. If they do not allow for it, but are based merely on the assumption of a statical load gradually applied and gradually removed, then they do not cover an immense range of repeated loading which occurs in practice. If they do allow for it, then the conception of the material being worn out by a maximum stress lower than the elastic limit is a false one, for the explanation of the destruction of the material lies in the kinetic accumulation of a much greater stress. VOL. LXVII. R 224 Messrs. W. K. Dunstan and T. A. Henry. (c.) The solution of the problem in both the cases of long shafts and short axles depends upon the discovery of a series of discon- tinuous functions. These functions would appear to have con- siderable interest for the mathematician. Like functions first appear, I think, in a paper by Boussinesq,* and they offer ;in alternative to the usual solution of problems in vibration by Fourier's series. The latter in such cases often give easy ana- lytical expressions, which, however, may be almost useless for the purposes of numerical calculation owing to the slowness of their convergence. In the course of the paper the numerical solution by the use of discontinuous functions, is compared for one case with the solution obtained by a Fourier's series. A verification of the work is thus obtained, and the advantages of the novel functions illustrated. (d.) There are many points in the memoir which suggest possibilities for physical research, and it seems to me that both from the purely scientific and the engineering sides a well-devised series of experiments on continuously and on abruptly varying torsional loads would lead to results of much interest and practical value. The memoir endeavours to complete as fully for torsional loading the theory of a changing load as the latter has been completed for longi- tudinal and transverse loading by de Saint- Venant, Boussinesq, and Flammant. The methods, analytical and graphical, are analogous, but the whole of the results, algebraic and mimerical, are, I believe, novel, and apply to a series of cases which, if possible, have even greater practical importance. " The Nature and Origin of the Poison of Lotus Arabiais. Pre- liminary Notice." By WYNDHAM R. DUNSTAN, M.A., F.R.S., Sec.C.S., Director of the Scientific Department of the Im- perial Institute, and T. A. HENRY, B.Sc. Lond., Salters' Company's Kesearch Fellow. Received June 7 — Head June 14, 1900. IjOtus Arabicus is a small leguminous plant resembling a vetch, with pink flowers, indigenous to Egypt and Northern Africa. It grows aliundantly in Nubia and is especially noticeable in the bed of the Nile from Luxor to Wady Haifa. It is known to the natives ;is " Kh.mb.er," and old plants with ripe seed are used as fodder. The dried plant is unusually green, and possesses the aroma of new-mown hay. At • See ' History of Elasticity ' vol. 2, arts. 401, 402. The Nature and Origin of the Poison of Lotus Arabians. 225 certain stages of its growth it is highly poisonous to horses, sheep, and goats, the poisonous property being most marked in the young plant up to the period of seeding. Owing to the trouble which this plant has given to the military and civil authorities in Egypt, the assistance of the Director of Kew was sought in order that the precise nature of the poison might be ascertained, and, if possible, a remedy found. The matter having been referred to the Scientific Department of the Imperial Institute, Mr. E. A. Floyer, Director of Egyptian Telegraphs, collected some of the material for investigation. It was found that when moistened with water and crushed, the leaves of the plant evolved prussic acid in considerable quantity, the amount being greatest in the plant just before and least just after the flowering period. Further investigation has shown that the prussic acid originates with a yellow crystalline glucoside (Co^HioNOio), which it is proposed to name lotusin. Under the influence of an enzyme, also contained in the plant, lotusin is rapidly hydrolysed, forming prussic acid, sugar, and lotoflavin, a new yellow colouring matter. The hydrolysis may be effected by dilute acids, but is only very slowly brought about by emulsin and not at all by diastase. The peculiar enzyme, which it is proposed to call lotase, appears to be distinct from the enzymes already known. Its activity is rapidly abolished by contact with alcohol, and it has only a feeble action on amygdalin. Old plants are found to contain lotase but no lotusin. The mgar has been proved to be identical with ordinary dextrose. Lotoflavin, the yellow colouring matter, has the composition expressed by the formula CiaHioOc. It belongs to the class of phenylated pheno-y-py rones, and is a dihydroxychrysin, isomericwith luteolin, the yellow colouring matter of Reseda luteola, and with fisetin, the yellow colouring matter of Rhus cotinus. The decomposition which ensues on bringing lotase in contact with lotusin, as happens when the plant is crushed with water, is therefore probably expressed by the following equation : — Co,H19N010 + 2H20 == C15H10Oo + HCN + CoH^Oo. Lotusin. Lotoflavin. Prussic acid. Dextrose. Hydrocyanic (prussic) acid occurs in small quantity in many plants, and according to Treub and Greshof is often present in the free state. The only glucoside at present definitely known which furnishes this acid is the well-known amygdalin of bitter almonds, which under the influence of the enzyme emulsin, also contained in the almond, breaks up into dextrose, benzaldehyde, and prussic acid. Owing to the scientific interest which attaches to this new gluco- side, its properties and those of its decomposition products have been very fully studied, and the characteristics of the new enzyme have also been investigated. 226 The Nature and Origin of the Poison of Lotus Arabicus. \\ V are much indebted to Mr. Floyer for the great pains he has taken to collect, in Nubia, the necessary material for this investigation, and also to Sir W. T. Thiselton-Dyer for having grown the plant at Kew, from seed obtained from Egypt. The Nature and Origin of the Poison of Lotus Arabians. 225 certain stages of its growth it is highly poisonous to horses, sheep, and goats, the poisonous property being most marked in the young plant up to the period of seeding. Owing to the trouble which this plant has given to the military and civil authorities in Egypt, the assistance of the Director of Kew was sought in order that the precise nature of the poison might be ascertained, and, if possible, a remedy found. The matter having been referred to the Scientific Department of the Imperial Institute, Mr. E. A. Floyer, Director of Egyptian Telegraphs, collected some of the material for investigation. It was found that when moistened with water and crushed, the leaves of the plant evolved prussic acid in considerable quantity, the amount being greatest in the plant just before and least just after the flowering period. Further investigation has shown that the prussic acid originates with a yellow crystalline glucoside (C22Hi9NQio)> which it is proposed to name lotusin. Under the influence of an enzyme, also contained in the plant, lotusin is rapidly hydrolysed, forming prussic acid, sugar, and lotoflavin, a new yellow colouring matter. The hydrolysis may be effected by dilute acids, but is only very slowly brought about by emulsiri and not at all by diastase. The peculiar enzyme, which it is proposed to call lotase, appears to be distinct from the enzymes already known. Its activity is rapidly abolished by contact with alcohol, and it has only a feeble action on amygdalin. Old plants are found to contain lotase but no lotusin. The sugar has been proved to be identical with ordinary dextrose. Lotoflavin, the yellow colouring matter, has the composition expressed by the formula CjsHioOo. It belongs to the class of phenylated pheno-y-pyrones, and is a dihydroxychrysin, isomeric with luteolin, the yellow colouring matter of Reseda luteola, and with fisetin, the yellow colouring matter of Ehus cotinus. The decomposition which ensues on bringing lotase in contact with lotusin, as happens when the plant is crushed with water, is therefore probably expressed by the following equation : — C22H19N010 + 2H20 == C^HioOo + HCN + C6H120«. Lotusin. Lotoflavin. Prussic acid. Dextrose. Hydrocyanic (prussic) acid occurs in small quantity in many plants r and according to Treub and Greshof is often present in the free state. The only glucoside at present definitely known which furnishes this acid is the well-known amygdalin of bitter almonds, which under the influence of the enzyme emulsin, also contained in the almond, breaks up into dextrose, benzaldehyde, and prussic acid. Owing to the scientific interest which attaches to this new gluco- side, its properties and those of its decomposition products have been very fully studied, and the characteristics of the new enzyme have also been investigated. VOL. LXVII. S L'L'ti Mr. (1. .1. r.urc-h. (hi t1i> S We are much indebted to Mr. Floyer for the great pains he has taken to collect, in Xuliia, the necessary matt-rial for this investigation, and also to Sir W. T. Thiselton-Dyer for having grown the plant at Ke\v, from seed obtained from Egypt. •'On the Spectroscopic Examination of Colour produced by Simultaneous Contrast." By GEORGE J. BURCH, M.A., Reading College, Reading. Communicated by FRANCIS GOTCH, F.H.S., Professor of Physiology, University of Oxford. Received June 12,— Read June 21, 1 900. In a previous communication I have described some methods of using the spectroscope to analyse sensations of successive contrast. In those experiments the eye, after having been fatigued by monochro- matic— preferably spectral — iight, is exposed to a second stimulus, consisting also of spectral light, exciting one or more colour-sensations which may or may not include that fatigued by the primary sensation. The question naturally arises, whether the spectroscopic method might not be applied to problems of simultaneous contrast. With this view I made a number of experiments with the Marl- borough spectroscope during the summer of 1897, of which the follow- ing may be mentioned. A piece of thin cover-glass was fixed in front of the eye-piece at an angle of 45° with the optic axis, so as to reflect into the field of view a small complete spectrum furnished by a 3|- inch direct-vision spectroscope. In order that this might be visible against the bright field of the larger spectroscope, a glass disc, with an opaque spot of the required size painted on it, was inserted in the eye-piece close to the diaphragm. With this arrangement it was easy to see the effect of contrast upon the smaller spectrum, but the lack of a comparison spectrum made the experiment far less striking than I had anticipated. Recently a device has occurred to me by which this difficulty may be got over, namely, the production of simultaneous contrast by dif- ferent colours in the two eyes. This method is employed in the well-known experiment by Hering, to show that the apparent alteration of colours by contrast is not due to an error of judgment, but to some real effect produced in the eye itself. An ordinary stereoscope is very convenient for this purpose, a square of red glass being inserted on one side of the central partition and a square of blue glass on the other. A small black wafer is then fixed at the centre of each glass, with a white wafer close to the left side of the one on the right-hand glass, and another on the right side of that o/ Coloi'i- i>n>o when cold. The tube was then expanded by heat- o as to allow the marble to pass completely into it and leave sil unit 1'25 inch of the tube free at either end. On allowing the tube to cool, a perfect contact between the iron and the marble was obtained. In some experiments the tube was subsequently turned down, so as to be somewhiit thinner immediately around the marble. Into either end of the tube, containing the column, an accurately fitting steel plug or piston was then inserted, and by means of these the pressure wa> applied. The high pressure required was obtained by means of a powerful press, especially constructed for the purpose, consisting of a double hydraulic " intenbifier," the water pressure being in the first instance obtained from the city mains. By means of this machine, pressures up to 13,000 atmospheres could be exerted on the columns having a diameter of 0*8 inch, and the pressures could be readily regulated and maintained at a constant value for months at a time, if required. It having been ascertained that the columns of the marble 1 inch in diameter and 1£ inch in height crushed at a pressure of from 11,430 to 12,026 Ibs. to the square inch, the column enclosed in its wrought-iron tube, in the manner above described, was placed in the machine and the pressure applied gradually, the exterior diameter of the tube being accurately measured at frequent intervals. No effect was noticeable until a pressure upon the marble, varying of course with the thickness of the enclosing tube, but generally about 18,000 Ibs. to the square inch, was reached ; when the tube was found to slowly bulge, the bulge being symmetrical and confined to that portion of the tube surrounding the marble. The distension was allowed to increase until the tube showed signs of rupture, when the pressure was removed and the experiment concluded. The conditions under which the marble was submitted to pressure were four in number : — 1. At the ordinary temperature in the absence of moisture. (Cold dry crush.) 2. At 300° C. in the absence of moisture. (Hot dry crush.) 3. At 400° C. in the absence of moisture. (Hot dry crush.) 4. At 300° C. in the presence of moisture. (Hot wet crush.) Eight experiments were made on marble columns at the ordinary temperature, in the absence of moisture, the rate at which the pressure was applied differing in different cases, and the consequent deformation being in some cases very slow and in others more rapid, the time occupied by the experiment being from ten minutes to sixty-four days. The amount of deformation was not in all cases equal, as some of the tubes showed signs of rupture sooner than others. On the comple- tion of the experiment the tube was slit through longitudinally by means of a narrow cutter in a milling machine, along two lines An Experimental Investigation into the Flow of Marble. 231 opposite one another. The marble within was found to be still firm and compact, and to hold the respective sides of the tube, now com- pletely severed from one another, so firmly together that it was impossible without mechanical aids to tear them apart. By means of a steel wedge driven in between them, however, they could be separated, but only at the cost of splitting the marble through longitudinally. The half columns of the marble now deformed generally adhere so firmly to the tube that it is necessary to spread the latter in a vice in order to set them free. The deformed marble, while firm and compact, differs in appearance from the original rock in possessing a dead white colour, somewhat like chalk, the glistening cleavage surfaces of the calcite being no longer visible. The difference is well brought out in certain cases owing to the fact that a certain portion of the original marble often remains unaltered and unaffected by the pressure. This when present has the form of two blunt cones of obtuse angle whose bases are the original ends of the columns resting against the faces of the steel plugs, while the apices extend into the mass of the deformed marble and point toward one another. These cones, or rather parabolas of rotation, are developed, as is well known, in all cases when cubes of rock, Portland cement, or cast iron are crushed in a testing machine in the ordinary manner. In the present expe- riments they seldom form any large portion of the whole mass. In order to test the strength of the deformed rock, three of the half columns from different experiments, obtained as above described, were selected and tested in compression. The first of these, which had been deformed very slowly, the experiment extending over sixty-four days, crushed under a load of 5350 Ibs. per square inch; the second, which had been deformed in 1| hours, crushed under a load of 4000 Ibs. per square inch ; while the third, which had been quickly deformed, the experiment occupying only 10 minutes, crushed under a load of 2776 Ibs. per square inch. As mentioned above, the original marble, in columns of the dimensions possessed by these before deformation, was found to have a crushing weight of between 11,430 and 12,026 Ibs. per square inch. These figures show that, making all due allowance for the difference in shape of the specimens tested, the marble after deforma- tion, while in some cases still possessing considerable strength, is much weaker than the original rock. They also tend to show that when the deformation is carried on slowly the resulting rock is stronger than when the deformation is rapid. Thin sections of the deformed marble, passing vertically through the unaltered cone and the deformed portion of the rock, were readily made, and when examined under the microscope clearly showed the nature of the movement which had taken place. The deformed portion of the rock can be at once distinguished by its turbid appearance, differing in a marked manner from the clear transparent mosaic of the 232 S'n.i1. F. I*. A.lams ami Dr. .1. T. tfi unaltc Thi- turbid apj>earance is most marked along a series of n-tirulating lines running through the sections, which when highly magnified are seen to consist of lines or bands of minute calcite granules. They are lines along which shearing has taken place. The cal.-ite individuals along these lines have broken down, and the frag- ments so produced have moved over and past one another, and remain as a compact mass after the movement ceased. In this granulated material are enclosed great numbers of irregular fragments and shreds of calcite crystals, bent and twisted, which have been carried along in the moving mass of granulated calcite as the shearing progressed. This structure is therefore cataclastic, and is identical with that seen in the felspars of many gneisses. Between these lines of granulated material the marble shows move- ments of another sort. Most of the calcite individuals in these posi- tions can be seen to have been squeezed against one another and in many cases a distinct flattening of the grains has resulted, with marked strain shadows, indicating that they have been bent or twisted. They show, moreover, a finely fibrous structure in most cases, which, when highly magnified, is seen to be due to an extremely minute polysynthetic twinning. The chalky aspect of the deformed rock is in fact due chiefly to the destruction by this repeated twinning of the continuity of the cleavage surfaces of the calcite individuals, thus making the reflecting surfaces smaller. By this twinning, the calcite individuals are enabled under the pressure to alter their shape somewhat, while the flattening of the grains is evidently due to move- ments along the gliding planes of the crystals. In these parts, there- fore, the rock presents a continuous mosaic of somewhat flattened grains. From a study of the thin sections it seems probable that very rapid deformation tends to increase the relative abundance of the granulated material, and in this way to make the rock weaker than when the deformation is slow. When the marble is heated to 300° C. in a suitably-constructed apparatus and is then subjected to deformation under conditions which otherwise are the same as before, the cataclastic structure is found to l)e absent and the strength of the deformed marble rises to 10,651' ll>s. to the square inch, that is to say, it is nearly as strong as the original rock. The calcite grains, which in the original rock are practically equidimensional, are now distinctly flattened, some of them being three or even four times as long as they are wide. Some grains can be seen to have been bent around others adjacent to them, the twin lamella- curving with the twisted grain. In others again of these twisted lamellae, the twinning only extends to a certain distance from the margin, leaving a clear untwinned portion in the centre. The rock consists of a uniform mosaic of deformed calcite individuals. An EoLperliiK'iital Investigation into the Flow of Marble. 233 When the deformation is carried out at 400° C., no trace of cata- clastic structure is seen. An experiment was then made in which the marble was deformed at 300° C., but in the presence of moisture, water being forced through the rock under a pressure of 460 Ibs. per square inch during the deformation, which extended over a period of fifty-four days, or nearly two months. Under these conditions the marble yielded in the same manner as when deformed at 300° C., in the absence of moisture, that is, by movements on gliding planes and by twinning, but without cataclastic action. The deformed marble, however, when tested in compression, was found actually to be slightly stronger than a piece of the original marble of the same shape. The structure developed was identical with that of the marble deformed at 300° C. in the absence of water. The presence of water, therefore, did not influence the character of the deformation. It is quite possible, however, that there may have been a deposition, of infinitesimal amount, of calcium carbonate along very minute cracks or fissures, which thus helped to maintain the strength of the rock. No signs of such deposition, how- ever, were visible. By studying the marble deformed at a temperature of 300° C., or better at 400° C., it will be seen that structures induced in it by the movements, and the nature of the motion, are precisely the same as those observed in metals when they are deformed by impact or by com- pression. In a recent paper by Messrs. Ewing and Rosenhain, " Experiments in Micro-metallurgy : Effects of Strain," which ap- peared in these Proceedings, three photographs of the same surface of soft iron, showing the results of progressive deformation under pressure, are shown, which photographs could not be distinguished from those of thin sections of the marble described in the present paper, at corre- sponding stages of deformation. In both cases the movements are caused by the constituent crystalline individuals sliding upon their gliding planes or by polysynthetic twinning. In both cases the motion is facilitated by the application of heat. The agreement between the two is so close that the term "flow" is just as correctly applied to the movement of the marble in compression under the conditions described, as it is to the movement which takes place in gold when a button of that metal is squeezed flat in a vice, or in iron when a billet is passed between rolls. In order to ascertain whether the structures exhibited by the deformed marble were those possessed by the limestones and marbles of contorted districts of the earth's crust, a series of forty-two speci- mens of limestones and marbles from such districts in various parts of the world were selected and carefully studied. Of these, sixteen were found to exhibit the structures seen in the artificially-deformed marble. In these cases the movements had been identical with those developed 234 Messrs. II. S. Il.-Ii-Sliaw and A. Hay. in the C'.'irrara marMe. In six other cases the structures bore n analogies to those in the deformed rock but were of doubtful origin, while in the remaining twenty the structure was different. The following is a summary of the results arrived at : — 1. By submitting limestone or marble to differential pressures ex- ivcding the elastic limit of the rock and under the conditions described in this paper, permanent deformation can be produced. i'. This deformation, when carried out at ordinary temperatures, is due iii part to a cataclastic structure and in part to twinning and gliding movements in the individual crystals comprising the rock. 3. Both of these structures are seen in contorted limestones and marbles in nature. 4. When the deformation is carried out at 300° C., or better at 400° C., the cataclastic structure is not developed, and the whole move- ment is due to changes in the shape of the component calcite crystals by twinning and gliding. 5. This latter movement is identical with that produced in metals by squeezing or hammering, a movement which in metals, as a general rule, as in marble, is facilitated by increase of temperature. 6. There is therefore a flow of marble just as there is a flow of metals, under suitable conditions of pressure. 7. The movement is also identical with that seen in glacial ice, although in the latter case the movement may not be entirely of this character. 8. In these experiments the presence of water was not observed to exert any influence. 9. It is believed, from the results of other experiments now being carried out but not yet completed, that similar movements can, to a certain extent at least, be induced in granite and other harder crystal- line rocks. " Lines of Induction in a Magnetic Field." By H. S. HELE-SHAW, F.If.S., and A. HAY, B.Sc. Received June 13,— Read -Turn- 21, 1900. (Abstract.) When a viscous liquid flows in a thin layer between close parallel walls, the motion takes place along stream-lines identical with those of a perfect liquid. The course of the stream-lines may be rendered evident by injecting into the clear liquid thin bands of coloured liquid. If the thickness of the liquid layer be varied, then there will lie .1 decrease of resistance to the flow wherever there is an increase of Lines of Induction in a Mcnjiidic Field. 235 thickness. As a consequence, there will be a convergence of the stream-lines on the area of greater thickness. When experiments with liquid layers of variable thickness were first tried, a general resemblance was noticed between the stream-lines so obtained and the lines of induction due to the presence of a permeable substance in a uniform magnetic field. The main object of the present paper was to investigate accurately whether complete correspondence between the two cases really existed, and, should correspondence be established, to apply the method to the solution of a number of two-dimensional magnetic problems. The investigation thus involved — (1) A mathematical treatment of the subject, by means of which plotted diagrams could be obtained for comparison with experimental results. (2) The construction of apparatus capable of giving exact results which could be photographed. (3) The investigation of the laws connecting the rate of flow with the thickness of film of the liquid used. The theoretical case selected as a test case was that of an elongated elliptic cylinder placed with its major axis along the field, the permeability being assumed to be 100. The lines of induction for this case are shown in the accompanying diagram, and were calculated and plotted by the method explained in the paper. DIAGRAM. If, for the moment, we assume that the liquid stream-lines are identical with lines of magnetic induction, then the following corre- spondence between the two cases holds : — 236 Mr. -I H. .loans. Li'jni'1 Flow. ('() Pressure gratlient. (/<) Kate of flow per unit width of liquid layer. ('•) Ratio of (b) to (a). • •fir Iii'/'irtion. (a) Magnetic intensity or f < . (/?) Magnetic induction, (y) Permeability = ratio of (y8) to (a). From this it is evident that the permeability corresponding to a given ratio of thicknesses of the liquid layer is given by the ratio of the rates of flow, per unit width of layer, for the two thicknesses, iissuming the same pressure gradient for both. The connection l>etween the rate of flow and the thickness for a given gradient of pressure was carefully investigated in a series of preliminary experi- ments, and it was found that the rate of flow varied as the ail*- of the thickness — a result which was afterwards confirmed by a theoretical investigation. The permeability in the magnetic problem is thus given by the ratio of the culms of the two thicknesses. A stream-line diagram corresponding to the theoretical diagram given above was next obtained, and on superposing the two it was found that their lines were practically coincident. The soundness of the method as applied to two-dimensional problems in magnetic induction having been thus established, the authors pro- ceeded to apply it to a number of special cases, many of which could not be successfully attacked by any other method. The paper is accompanied by a large number of photographs, showing the results obtained. Some of these are of importance from an electrical-engi- neering standpoint. The method described is the only one hitherto known which enables us to determine the lines of induction in the substance of a solid magnetic body. It is equally applicable to two-dimensional problems in magnetic induction, electrical flow, and heat conduction. " The Distribution of Molecular Energy." By J. H. JEANS, K.A.. Scholar of Trinity College, and Isaac Xewton Student in the 1'niversity of Cambridge. Communicated by Professor J. J. THOMSON, F.R.S. Pteceived June 14,— Read June 21, 1900. (Abstract.) This paper attempts to examine the well-known difficulties in con- nection with the partition of energy in the molecules of a gas. A definite dynamical system is first considered, an ideal gas in which the molecules are loaded spheres, that is, spheres of radius «, of which the centre of mass is at a small distance, r, from the geometrical centre. It The Distribution of Molecular Energy. 237 is shown by direct methods that the energy will, after an infinite time, distrib\ite itself equally between the five degrees of freedom, but when a wave of sound is passed through the gas, the energy will never have sufficient time to attain to its equilibrium distribution. It is shown that sounds of different period will be propagated with appre- ciably different velocities, except in the extreme case in which the ratio of r to a is almost, but not necessarily quite, zero. In this case, the ratio of the two specific heats, as determined from indirect experiments on the velocity of sound, would be If, while direct experiments might give any value from 1§ to If, the value varying with the duration of the experiment. It is suggested that an escape from this dilemma is made possible by regarding the molecules as forming an incomplete dynamical system, of which the ether is the remaining part. For purposes of illustration, it is imagined that the interaction between the two parts of this com- plete system consists of a frictional force which retards the rotation of the molecules. A steady state is now impossible, but it is shown that when the energy (i.e., temperature) of the gas is sufficiently low, the gas tends to assume an approximately steady state, in which the energy of rotation vanishes in comparison with that of translation. It is then shown that these conclusions may be generalised, so as to apply to a more complex system of molecules, these molecules possess- ing an indefinite number of degrees of freedom, and internal potential energy as well as kinetic. The molecules exert forces on one another O«/ at any. distance, and the radiation is of a more general type than before. In Part III some of the "physical consequences of the view here put forward are examined. The final conclusions are briefly as follows : — The degrees of freedom must be weighted, not counted. The weight of a degree of freedom may be anything between unity and zero, and may vary with the temperature. A degree of freedom which does not radiate energy will always be of weight unity ; for a non-luminous gas, one which does radiate energy when the gas is heated is of weight zero. As the gas is heated, the radiation and internal energies will in- crease much more rapidly than the temperature, until finally, at infinite temperature, the energy is distributed equally between all degrees of freedom. Finally, it is pointed out that this view is in accordance with ordin- ary thermodynamics for a non-luminous gas, but that the ordinary thermodynamics must be supposed to break down above the tempera- ture of incandescence, a view which has already been put forward, in a modified form, by Wiedemann. Dr. II. T. I', • //••// "On the Capacity for Heat of Water between the Free/in^ and I'xiilin^ Points, together with a Determination of the Mechanical Equivalent of Heat in Terms of the International Electrical Units.— Experiments by the Continuous-flow Method of Calorimetry performed in the Macdonald Physical Laboratory of McGill University, Montreal." By HOWAIID TUKNEK BAHXES, M.A.Sc., D.Sc., Joule Student. Communi- cated by Professor H. L. CALI.KNDAK, F.RS. Received June 15,— Read June 21, 1900. (Abstract.) At the Toronto meeting of the British Association in 1897, a new method of calorimetry was proposed by Professor Callendar and the author for the determination of the specific heat of a, liquid in terms of the international electrical units. At the Dover meeting in September, 1899, some of the general results obtained with the method for water over a part of the range l>etween 0° and 100° were communicated, with a general discussion of the l>earing of the experiments to the work of other observers. In the present paper the author gives a summary of the complete work, in the case of water, to determine the thermal capacity at different temperatures between the freezing and boiling points. Theory of the HetJiod. If a continuous flow of liquid in a tube be made to carry off a con- tinuously supplied quantity of heat EC, in electrical units, then after nil temperature conditions have become steady JsQ (0i - 00) t + (0! - 0«) ht = ECt where J = mechanical equivalent of heat, Q = flow of liquid per second, s = the specific heat of the liquid, 00 = the temperature of the liquid flowing into the tube, 01 =• the temperature of the liquid flowing out of the tube, h = the heat loss per degree rise of temperature from the liquid flowing through, t = the time of flow. In the case of water, E represents the E.M.F. across an electrical heating conductor in the tube, and C the current flowing. In this case, which is treated of entirely in the present paper, Jx is replaced Water betiueen the Freezing and Boiling Points. &c. 239 by 4*2 (1 ± 8) where S is a small quantity to be determined, and varies with the thermal capacity of the water, which is not exactly equal to 4'2 joules at all points of the range. Substituting in the general equation, rearranging terms, and dividing through by t, the equation is given in the following form : — 4-2Q(0i - 00)5 + (0i - 00)h = EC - 4-2Q(^ - 00), which is termed the general difference equation of the method. The two terms 8 and h may be determined by using two values of Q, giving two equations of the form 00)h = EA - 4-2Q1(01 - 00) 4-2Q2(02 - 60)8.2 + (0-2 - 00)h = E2C2 - 4'2Q2(02 - 00). For the same value of 0o, if the electrical supply for the two flows is regulated so that 6\ = 02, then 8, = 82 = 8, and by eliminating h, 8 = - 00)) - (E2C2 - 4-2Q2(0! - fl0)) 4-2(Q1-Q2)(01-00) which corresponds to the mean temperature where (0X - 00) is not too great. In the present method the flow tube is of glass, about 2 mm. in diameter, connected to two larger tubes forming an inflow and an outflow tube, in which the temperature of the water is read, by a differential pair of platinum thermometers, before and after being heated by the electric current. A glass vacuum jacket surrounds the fine flow tube and a part of the inflow and outflow tubes, to reduce the heat loss as much as possible. A copper water jacket encloses the inflow tubes and vacuum jacket, in order to maintain the glass surface of the vacuum jacket always at a constant temperature equal to the inflowing water. The heat loss from the water is then the loss due to radiation from the flow tube through the vacuum jacket, and conduc- tion from the ends of the flow tubes. In testing the accuracy of the method, the dependence of the heat loss on the rise of temperature was found, and the dependence of the heat loss on the flow. Measurement of Fundamental Constants. The electric heating current supplied to the wire conductor in the fine-flow tube was taken from four large 200-ampere hour accumulators. It was passed through a standardised resistance in series with the wire I>r. II. T. n.-imr.. Un tl. ' /.' ami in addition a specially constructed rheostat, by which sma.ll adjustments to the circuit could IKJ made for regulating the heat supply. The ipeasiirenient of the different constants entering into the general difference equation of the method is treated of under two heads, Electrical and Thermal. In the first, the Clark cell and resist- ance form the principal measurements, and in the second the measure- ment of temperature, time, and weight have to be considered. An exceedingly accurate potentiometer was employed to determine the difference of potential across the resistance and calorimeter in terms of the E.M.F. of the Clark cell. nful Proof of the Theory of the Method. In this section the author shows that the dependence of the heat loss per degree rise in the calorimeter varies in a linear relation to the flow in proportion to 4'2 QS beyond certain limits of flow, and that this is essential for the fulfilment of the theory of the method. For very small flows the conduction effect at the outflow end, due to the rise of temperature in the water, appears and causes the line repre- senting the relation of heat loss to flow to approach an infinitely large value of the heat loss for a zero flow. The limits of flow chosen in the present measurements are safely included within the linear relation. The relation of the heat loss to the rise of temperature shows that for rises of from 2° to 8° and beyond, the heat loss is directly pro- portional to the rise. The thermal capacity of the calorimeter is cal- culated, and it is shown that for the small changes in the temperature of the calorimeter during an experiment this is negligible. Effect of Stream-line Motion. Some of the earlier results are given in this section, showing the effect of stream-line motion on the distribution of heat throughout the water column for a calorimeter with a 3-mm.-bore flow tul>e for different flows. The temperature of the heating wire used for these experiments is also calculated, and found to vary considerably when moved from the centre to the sides of the tube. It was found neces- sary to thoroughly stir the water in its passage through the flow tube, in order to ensure a perfectly uniform temperature throughout the water column. Preliminary Measurement*. The preliminary measurements of the value of J, which were made in the summer of 1898, were affected by the presence of stream-line motion in the tube, as at that time no device was introduced to Water betiveen the Freezing and Boiling Points, &c. 241 obviate it. Owing to the calorimeter which was used then, however, having only a 2 mm.-bore flow tube, the effect was not so large as for the tube with a 3 mm. bore. The value which corresponds to a tem- perature of the water of 30° C. is 4- 1805 joules, which agrees to 1 part in 2000 with the later and more accurate measurements which were obtained for all the calorimeters with the various devices for elimi- nating the stream-lines. Experiments between 0° and 100°. In this section the complete list of fifty-five tables is included, giving upwards of forty-five complete experiments at different parts of the range. The experiments have extended over just a year, and divide themselves naturally into eight separate series. The results with different calorimeters and with different rises of temperature are included. Summarising the results and plotting the values of 8 for all the experiments, the following values of 8 and the corresponding values of J are obtained from the smoothed curve : — Summary of the Specific Heat of Water from Smoothed Curve. Temperature. S. J. °C. 5 +0-00250 4-2105 10 -0-00050 4-1979 15 -0-00250 4-1895 20 -0-00385 4-1838 25 . -0-00474 4-1801 30 - 0-00523 4-1780 35 -0-00545 4-1773 40 -0-00545 4-1773 45 -0-00520 4-1782 50 -0-00480 4-1798 55 -0-00430 4-1819 60 - 0-00370 4-1845 65 -0-00310 4-1870 70 -0-00245 4-1898 75 -0-00180 4-1925 80 -0-00114 4-1954 85 -0-00043 4-1982 90 +0-00025 4-2010 95 +0-00090 4-2038 Mean value 4-18876 The values of 8 represent the specific heat of water in terms of a thermal unit equal to 4-2000 joules, which occurs at 9° C. It is more VOL. LXVII. T l>i. II. T. Kirnes. On the Capacity for Heat of suit;ible to select a thermal unit at a more convenient part of the scale. The mean value of the mechanical equivalent of heat from these measurements over the whole range is 4-18876 joules, which is very nearly equal to the value at 16° C., which is 4*1883 joules. It seems desirable to select a unit at a temperature which, if at the same time at a convenient part of the scale, may be equal to the mean value over the whole scale. The author has in consequence adopted a unit at 16° C., and has expressed the specific heat of water in the following table in terms of this unit : — Variation of the Specific Heat of Water in Terms of a Thermal Unit at 16°C. Temperature. Observed values. Calculated values. 5 1-00530 10 1-00230 15 1-00030 20 0-99895 25 0-99806 30 0-99759 35 0-99735 40 0-99735 45 0-99760 50 0-99800 55 0-99850 60 0-99910 65 0-99970 70 1-00035 75 1-00100 80 1-00166 85 1-00237 90 1-00305 95 1-00370 Mean... 1-00012 1-00446 1-00206 1-00024 0-99894 0-99807 0-99757 0-99735 0-99735 0-99757 0-99807 0-99894 0-99910 0-99972 1-00036 1-00100 1-00166 1-00233 1-00301 1-00370 In expressing the results in a formula it is impossible to fit any one simple expression over the whole scale. It is seen that the curve falls rapidly from 0°, passes through a minimum point at 37-5°, and in- creases again less rapidly towards 100°. Two formulae can be fitted very accurately over the scale. Between 5° and 37-5° C. the following expression in terms of a thermal unit at 16° is found to read, S = 0-99733 + 0-0000035 (37-5 -Q2 + 0-00000010 (37 -5 -0s. The same formula holds between 37-5° and 55° by simply consider- Water between the Freezing and Boiling Points, &c. 243 ing all values of the cubical term positive. Above 55° the simple formula S = 0-99850 + 0-000120 (£-55°) + 0-00000025 (t- 55)2 holds with great accuracy. Both these formulae are given for comparison in the second table. They fit very closely except below 5°, where the specific heat curve increases more rapidly. These loAver values are within 1 part in 1000, however. A summary of the values obtained for the radiation loss shows that the absolute value for any one calorimeter cannot be relied on to an •order of accuracy greater than 1 part in 1000 over extended periods. This is particularly true when the temperature of the calorimeter is widely changed. It was found most essential to always eliminate the heat loss from at least two different flows in order to be completely independent of its absolute value. The complete independence of the results from the value of the heat loss, provided this remained con- stant throughout the time of an experiment, was shown by employing calorimeters with different degrees of vacuum involving widely differ- ing values of the heat loss. The temperature coefficient of the radia- tion loss was found to be almost exactly linear over the range of these -experiments. Relation to the Work of other Observers. It is at once apparent that the value of the mean mechanical equiva- lent of heat obtained from these meastirements, which is 4-18876 joules, is somewhat larger than the exceedingly accurate and trustworthy measurements of Eeynolds and Moorby. Their value, which is 4-18320 joules, is lower by 0-132 per cent., or a little over 1 part in 1000. It is evident that this error may be attributed to the neglecting of some correction factor in the present series of experiments at the •extremities of the range, which would cause the variation curve to increase more rapidly than it truly does ; but from the order of accu- racy with which the theory of the present experiments holds at the •extremities of the range, it is far more likely that the variation curve is correct, and that the difference in the two results is to be attributed to an error in one of the constants. The thermal constants employed in the two different experiments are referred to the same values, but the introduction of the value of the electrical units into the present series of experiments, which do not enter into the calcula- tion of Eeynolds's and Moorby's result, renders it highly probable that the error is to be looked for here. In view of the immense amount of labour expended in establishing the value of the inter- T 2 244 On tlic Capacity for Heat of Water, &c. national ohm, it is probable that the error is not there. The recent work on the absolute value of the Clark cell, which is demanding so much attention just now, and which has so far given so many incon- sistent results, makes it very probable that the value of the Clark cell adopted in the present work is in error. If this is so, then, as is pointed out, the value of the Clark cell must be taken as 1*43325 int. volts at 15° C., in order to bring the present series of experiments, involving both the international volt and ohm, into absolute accord with the result by the direct mechanical method of Reynolds and Moorby. Having considered the above relationship, the mean value of the mechanical equivalent given by Rowland's experiments between 6° and 36° C. is compared with the same mean value from the present series of experiments over the same range. By expressing this latter value in terms of the value of the Clark cell 1 -43325 volts, or as may be said in terms of Reynolds's and Moorby's determination, instead of the original value 1 '43420 volts used in calculation, it comes equal to 4'1817 joules. The value obtained from Rowland's corrected curve is 4'1834 joules, which agrees with the present series of experiments to 1 part in 2000. This is a discrepancy so small as to be, if not within the limits of error of these several determinations," at least negligibly small in comparison to the great range covered by the present series of experiments. By far the most difficult part of the present series of experiments is the comparison of the absolute value of the mechanical equivalent of heat obtained from these experiments with the values obtained by the electrical method used by Griffiths, and by Schuster and Gannon, even when our several results are expressed in terms of the same values of the units used. There is every reason to believe that the values of the resistance standards used in the present work were the same as those used by both these investigators. It is also highly probable that the values of the Clark cells in the present series of experiments were in correct agreement with all the best results that have been obtained in setting up this electro-chemical combination. It is probable that the difference in the values obtained by Griffiths, and by Schuster and Gannon, from the value obtained in the present series of experi- ments must be attributed not to these, but to the radical difference in the methods of calorimetrv. Energy of Rontgen and Bccquerel Rays, Sfc. 245 " Energy of Eontgen and Becquerel Kays and the Energy required to produce an Ion in Gases." By E. RUTHERFORD, M.A., B.Sc., Macdonald Professor of Physics, and R K McCLUNG, B.A., Demonstrator in Physics, McGill University, Montreal. Communicated by Professor J. J. THOMSON, E.R.S. Received June 15,— Read June 21, 1900. (Abstract.) The primary object of the investigations described in the paper was the determination of the energy required to produce a gaseous ion when X rays pass through a gas, and to deduce from the result the amount of energy radiated out into the gas by uranium, thorium, and the other radio-active substances. In order to determine this " ionic energy " it has been necessary to accurately measure the heating effect of X rays and the absorption of Rontgen radiation in passing through a gas. The coefficient of transformation of a fluorescent screen excited by X rays as a source of light has also been investigated, and a simple practical method of expressing the intensity of Rontgen radiation in absolute measure has been explained. The method adopted to determine the ionic energy was briefly as follows : — The maximum current between two electrodes produced by the ionization of a known volume of the gas by the rays was determined. In order to ionize the gas energy has to be absorbed, and the intensity of the radiation falls off more rapidly than the law of inverse squares. Assuming that the energy of the radiation absorbed in the gas is expended in the production of ions, then, knowing the coefficient of absorption of the rays in the gas, the total current produced by the complete absorption of the whole radiation given out by the bulb into the gas can be deduced. Let i = maximum current produced by the total ionization of the gas by the rays, n = number of ions produced, e = charge on an ion. Then i = ne. Let H = heating effect due to the rays when absorbed in a metal, E = total energy of the rays in ergs, Then E = JH, where J = Joule's equivalent. If W = average energy required to produce an ion, then nW = E = J.H, w JH JH* .'. W = = : — • n i 246 I E. Kuthertunl ami Mr. II. K. .MrClun- The values of H and i are experimentally determined, and, assuming the value of «, namely, 6*5 x 10~10 electrostatic unit, determined by J. J. Thomson, the value of W is found in absolute measure. In the course of the investigation the following subjects have been considered : — (1) Measurement of the heating effect of X rays. (2) Efficiency of a fluorescent screen excited by X rays as a source of light. (3) Absorption of X rays in gases at different pressures. (4) Determination of the energy required to produce an ion in air and other gases, including deductions on — (") Distance between the charges of ions in a molecule. (b) Minimum potential difference required to produce a spark. (5) Energy of Becquerel rays and emission of energy by radio-active substances. Heating Effect of X Hays. An automatic focus tube was employed, excited by a large induction coil with a special form of Wehneldt interrupter giving fifty-seven breaks per second. The bulb gave out intense rays of a very pene- trating character. The heating effect was measured by determining the variation of resistance of a special platinum bolometer when the rays fell upon it. A platinum strip, about 3 metres in length, 0'5 cm. wide, and 0'003 cm. thick, was wound on a light mica frame 10 cm. square. Two such " grids," as similar as possible, were constructed, and formed the two arms of a Wheatstone bridge. A balance was obtained for a momentary pressing of the battery key, using a sensitive galvanometer. The raya were then turned on for 30 or 45 seconds, and the deflection from zero determined immediately after the rays were stopped. In order to measure the heating effect, a current was sent for the same time as the rays acted through the grid, and its value adjusted until the deflection due to the heating of the grid was the same as for the rays. When this is the case the heat supplied per second to the grid by the rays is equal to the heat supplied per second by the current. Thus, heating effect of rays per second = 0'24i- R calorie, where t = current through the grid of resistance R. The grids were enclosed in a lead vessel with an aluminium window to let in the rays. The whole was surrounded by a felt covering, and several aluminium plates intervened between the bulb and the grid, so that any heating effect, except that due to the rays, was completely eliminated. Energy of Itontgen and Becqiierel Rays, &c. 247 About 0'55 of the energy of the incident rays was absorbed in the grid. Some of the energy of the rays was used up in exciting secondary radiation at the surface of the platinum grid, but the amount was not large, and was neglected in comparison with the total energy of the rays. The rate of supply of heat to the grid area 92 '2 sq. cm. at a distance of 26 cm. from the source of rays was 0'00014 gramme-calorie per second. The total energy of the rays given out from the front surface of the platinum antikathode (omitting absorption of rays in the glass of bulb, in air and screens) was O011 gramme-calorie per second, or 0-046 watt. The number of discharges per second was 57, and assuming 10~5 second* as the average duration of the rays during each dis- charge, the maximum rate of emission of energy from the bulb = 19*5 calories per second. The heating effect of the sun's rays falling normally on 1 sq. cm. of surface = 0'035 calorie per second. The maximum rate of emission of energy from an X-ray bulb is thus 560 times greater than the energy of the sun's rays at the surface of the earth. Some experiments were made on the heating effect of the rays, using a thermopile, but it was found to be a very unsuitable instrument for such a determination. Efficiency of a Fluorescent Screen. Photometric comparisons were made of the light from a fluorescent screen with that of the standard Hefner- Alteneck amyl lamp, using a Lummer-Brodhun prism. With a screen of platinocyanide of barium Intensity of light from screen _ Intensity of light from amyl lamp Tumlirzf has shown that the energy of the visible light from an amyl lamp falling normally on 1 sq. cm. surface at unit distance = 0-00361 gramme-calorie per second. For X rays of- the same intensity as were used in the photometric measurements, the energy under the same conditions = 0-0023 calorie, * Trouton, ' Brit. Assoc. Keport,' 1896. t ' Wied. Annal.,' vol. 38, p. 640. Prof. E. Rutherford and Mr. I,1. K. M« clung. or the rate of emission of energy per second as visible light from the Hefner lamp is nearly twice the rate of emission of energy from the X-ray tube. 0'73 of the energy of the rays was absorbed in the screcMi. The efficiency of the transformation of X rays into visible light by the screen (compared with the Hefner lamp) = 0*044 or 4-4 per cent. Assuming this transformation factor for a fluorescent screen, two simple photometric measurements are required to express the energy of any bulb in absolute measure. The light from a fluorescent screen is first compared with the standard Hefner lamp. The absorption of the rays in the screen is determined by placing a piece of the screen in the path of the rays. Let p = ratio of intensities of light from bulb and lamp, pi = ratio of transmitted to incident radiation on the screen. Then it is shown that the intensity in absolute measure 0-082p , . = - E gramme-calorie per second. 1 -pi The absorption in the cardboard of the screen is supposed to be negligible, but if necessary can be readily allowed for. Absorption of X Rays in Gases. A null method was employed, as the absorption of the rajs in air at atmospheric pressure was small. The rays passed through two long brass tubes with aluminium ends, and the current produced by the rays, after passing through one tube, was balanced against the current due to the other. On exhausting one tube the electrometer balance was disturbed. From measurements of the deflection per second from the balance and the deflection per second due to the rays after passing through one tube, the absorption can be calculated. The mean value of the coefficient of absorption of the rays in air at atmospheric pressure was found to be 0-000279, or the rays would pass through 24'7 metres before absorption reduced the intensity of the radiation to one-half. The absorption was found to be proportional to the pressure from a hulf atmosphere to three atmospheres. The coefficient of absorption in carbonic acid gas was found to be 1'59 times the absorption in air. Energy of Rontgen and Becguerel Rays, &c. 249 Energy required to produce an Ion. The current produced when a given volume of the gas was ionized by X rays was determined by means of an electrometer. In order to get rid of the secondary radiations set up when X rays strike on a con- ductor, the rays passed between two charged parallel plates without striking them. A guard-ring method was employed to ensure uniformity of the electric field. The value of the ionic energy was deduced from the determination of the current, heating effect, and absorption of the rays. The mean value of the energy required to produce an ion in air at atmospheric pressure and temperature was found to be l-90xlO-10erg. This value is much greater than the energy required to produce hydrogen and oxygen ions in the decomposition of water. The ionic energy of air was found to be approximately the same from pressures of one-half to three atmospheres. The method of determining the ionic energy for other gases is described, and the evidence that the " ionic energy " is the same for all gases is discussed. Distance between the CJuirges of the Ions in a Molecule. On the assumption that the energy absorbed in producing an ion is due to the work done in separating the ions against the forces of their electrical attraction, it can be shown that the mean distance between the charges of the ions in the molecule is 1-1 xlO-°cm. This is only •£$ of the probable diameter of the atom. This result is in accordance with the view recently advanced by J. J. Thomson, that ionization is produced by the removal of a negative ion from the mole- cule, and that the negative ion is only a small fraction of the mass of an atom. Minimum Potential required to produce a Spark. If the production of ions is necessary before a spark can pass, it can readily be deduced from the value of ionic energy that a spark cannot pass for a potential difference less than 175 volts. Experiments have shown that the minimum value is over 300 volts. The theoretical value is of the same order, but from the complexity of the phenomena a very close agreement could not be expected. 250 Dr. A. Macfadyen, J)r. (>. II. Morris, and Mr. ,S. ll»>\vl;uif Age of yeast After 24 hours. After 48 hours. IB from yeast. collection. Alone. With 40 per cent, sugar. Alone. With 40 per cent, sugar. C.C. c.c. c.c. c.c. A* Fresh 520 280 600 472 A* 1 day 240 808 — — A* 2 days 133 164 — — A* Fresh 308 186 308 324 A* Fresh 400 228 400 340 A* 1 day 285 270 285 320 A* 1 day 162 84 162 150 Af 2 days 990 234 990 290 At Fresh 90 170 — — At 3 days 760 560 788 592 AJt Fresh 900 2165 — — A* 5 days 280 220 — — C* 1 day 550 180 — — ct 1 day 540 100 — — C* 3 days 160 200 — — A* 7 days 120 200 — — A* 1 day 65 90 — — * Toluol used as antiseptic. t Sodium arsenite used as antiseptic. £ In this experiment 10 per cent, cane-sugar was employed. 2.')6 Dr. A. Macfadyen, Dr. G. H. M-.rrK ami Mr. S. Rowland •t >H •s «i L I o s 1 1 5 1 Z - ^ Z •* CO CO X •* •* eo 1-1 t !/; S 1 l§ S S 1 oo as 88888 ill I o 60 iO ^" LO O •-• •* eo fe "c 8 ' S.3 ^ *1 O tO 1 S o o o lO 1O CJ5 •* 1 CO 00 CO CM O 00 CO CO Ci OS OJ CM CM CM CO 1 40 bo <•* o 1 SU3 O «5 U5 _ -- *1 O QO g ?D r-t CM CM «O O O O O 0 O Q C^l Ct ^ •^ CM as 8 rH 3 E,00^ rH i-l i-H CM O M^M 1 1 "o *i ii 181 ICO 30 1 CO to x 1 CM g|S S »0 E. ° ec eo co eo eo eo 3 -H O . a m *oo S I I § g 0 O O O §O | 1,0 c i-H R OO ac t* ^ 1 eo eo co •r. A. Id .-n, Dr. G. H. Morris. an-1 Mr. S. i:,i\vl:in«l. experiments to determine the most favourable concentration of sugar ; the results show that the smaller amounts — 5 to 10 per cent. — give the most favourable results, whilst the larger quantities sensibly retard the action, i.e., less gas is obtained from the juice plus sugar than from the juice alone. This probably explains to some extent the results we Fro. 2 7-0 Showing the influence of ag« of after 48 hours. expressed juice 6-0 3-0 3-0 /•O o \° \ &4C- uncorrected for&uCo-fermwitaCion. Age of yeAaC in days. obtained in our earlier experiments in which 40 per cent, of sugar was employed. Some of the results are shown diagrammatically in fig. 3. Influence of different Sugars. — In order to determine if the nature of the sugar employed had any influence on the amount of gas evolved, we carried out a series of experiments with cane-sugar, dextrose, maltose, and levulose at different concentrations, using the same sample of juice On Expressed Yeast-cell Plasma (Buchner's " Zymase "). 259 for each set of experiments. The results, as a whole, show that more carbon dioxide is given off from cane-sugar than from either of the other sugars. FIG. 3. — Showing the influence of different concentrations of sugar on gas evolved. :I 3-0 'hO a. a' 10 15 25 20 tf~sug&r. a, After 46 hours (uncorrected) ; a', corrected for &u Co- ferm. 6, « 24 " " ; b , " " " Influence of Temperature. — We made several experiments to ascertain the most favourable temperature for the action of the juice. As an example of the results obtained, we may quote the following : — The juice was mixed with 10 per cent, of cane-sugar in the usual way, and there were obtained — At 0° C.... 0'41 gramme of carbon dioxide in 48 hours. „ 10 0-83 5> 2o 1 '05 ,, ,, ,, 37 1-17 55 °' 1 II ,, „ „ The higher temperatures therefore appear to increase the activity of the juice. Influence of Filtration. — In order to ascertain what influence, if any, nitration through Chamberland and Berkefeld niters had on the activity of the juice, we carried out a series of experiments with different juices, carefully testing their gas-producing activity before and after filtration. Thymol was used as an antiseptic in each case The results are given in Table III, and it will be seen that filtration decreases to a considerable extent, but without entirely destroying, both the auto-fermentation and the action of the juice on sugar. This decrease in gas-evolving power is accompanied by a very considerable U 2 260 Dr. A. Macfadyen, Dr. G. H. Morris, and Mr. S. Rowland. fall in gravity of the juice. These experiments agree with those of Buchner on the same point. Table III. — Influence of Filtration on the Activity of the Juice. Gas evolved from 100 c.c. of juice. Source Gravitv Ag? before Gravity Before filtration, after After filtration. of of «H« nil — , \c;i-'. Jeast- tion. tion. With 10 With 10 Alone. per cent. Alone. per cent. C.8. O.S. days. gramme. gramme. •MBUD6. gramme. B« 3 — — 0 -67 0 -31 B 2 -^ 0-43 0-65 0-00 0-10 B 2 0-18 1-56 0-25 0-84 Bt 2 — 1-23 0-83 O'll , 0-43 B 2 1055 1018 0 -57 0-65 0-24 0-23 B 3 1045 1030 0 -25 1-55 0-24 1-17 * In this experiment the filtration was through a Chamberland filter ; in the remaining experiments a Berkefeld filter was used. f After 72 hours in this case ; all the others after 48 hours. Influence of Dilution. — In considering the nature of the action of the juice and of the agent to which the evolution of gas was due, it appeared important to ascertain the effect of dilution on the action of the juice. All experiments were conducted by adding the weighed quantity of sugar to the juice itself, so that no water at all was intro- duced. If the action were a purely enzymic one, dilution to a limited extent should not appreciably affect the result ; whereas if the action were due to other causes, it might be influenced to a greater or less extent. We accordingly carried out a series of determinations on dilution with water alone, with physiological salt solution (0'75 per cent, sodium chloride), and with water in the presence of cane-sugar. The experiments with sugar were made in two ways : in the one, the sugar was added to the juice in the usual way (10 per cent.), and water was then added to bring about the desired dilution, the ratio of the sugar to juice being therefore kept constant ; in the other, the dilution was made with a 10 per cent, solution of sugar, so that the ratio of sugar to the total volume was maintained throughout. The results obtained are set out in TaUlc IV. An examination of the results will at once show that the auto-fermentation of the juice is greatly influenced by dilution both with water and with salt solution. The addition of an equal volume of water sensibly retards the action, On Expressed Yeast-cell Plasma (Buehncr's " Zymase "). 261 M it"1 | 1 1 || 1 ft i— I g lo oo 01 E a 1 39^9 | g o o o o b b ' 1 0 rz3 bD ,fis q * ~ _g r-t o * O5 « I S rH O rH 1 9 ^ § h 3. i-* 1 d o 1 8 b a ' S ^ bC ; 03 S 1 I 'o > o CO rH I Q s ! || '! o • 1 "a o O 09 o a reserice 0 0 rH rH b o •s o • o n te a 8 . « £* ~ r*i bfi o rrj t ? .- ~ rj- "8 rH Xft U5 CD — .-..? »» o °0 . j « 1o g | O ^ 02 " •/'//<.-• th<- evolution '/'/">•. With salt solution, the action is still more marked. In the presence of sugar the retarding action is still distinctly apparent, especially when the concentration of the sugar decreases with dilution. In this case the effect of dilution is fully as marked as in the case of water alone or of salt solution. When the strength of the sugar solution is maintained constant, the retardation is still con- siderable, but not so great as in the other cases. This paralysing effect of dilution on the activity of the .juice is so contrary to the behaviour of enzymes in general under similar condi- tions, that in our opinion it forms a grave objection to the acceptance of Buchner's enzyme theory. Since the above experiments were made, we find that Wroblewski* has conducted dilution experiments with like results. In connection with the question of the influence of dilution on enzyme action, it may be mentioned that when a sample of six-day juice was diluted to 1 in 1000 with cane-sugar solution, 50*5 per cent, of the cane-sugar was found to be inverted, whilst with another juice, three days old, a dilution of 1 in 100 showed an inversion of 79'5 per cent, of the cane-sugar present. This offers a great contrast to the effect of dilution of the juice on the production of carbon dioxide. Ratio of Carbon Dioxide to Alcohol. — In connection with the question whether we had to do with a true alcoholic fermentation, it became important to determine if carbon dioxide and alcohol were produced in the proportions ordinarily found, and if the amount of sugar which disappeared during the experiment bore any relation to the alcohol and carbon dioxide. We carried out a large number of experiments with a view to elucidate these points, and the results of some of the experiments are shown in Table V. In experiments 1 to 5, the alcohol and carbonic acid estimations were made on the same fermenta- tion, but in experiments 6 to 15 we carried out duplicate fermenta- tions, under identical conditions with the same juice, for the two determinations. We did this in order to ensure greater accuracy in the alcohol estimation, since the escaping gas could l»e washed by passage through a little water, which was subsequently added to the distillation flask. When we were estimating both products from the same experiment this was not possible. It will be noticed from the table that the juice as it comes from the press always contains a considerable amount of alcohol, and we found on examination that this agrees fairly closely with the amount of alcohol contained in the yeast, oven after the thorough washing and pressing to which it had been subjected in the preliminary treatment. When corrections are made for the amounts of alcohol and of carbon dioxide formed during the auto-fermentation of the juice, the ratio » 'Centralbl. f. Physiol.,' 1899, p. 284. On Expressed Yeast-cell Plasma (Buclmers " Zymasc"}. 263 05 Q .S ^ O o ^3 cc 53 H 05 O o o .2 H -4-2 I M o |« s;iraoioioifSifl>oSI 4n ^ rt^J>OOCOlffllO •^lO^aqcorHiot- ^•H bC O o I5* Q* | S CcoJoN^lO ^JooSS cSS P §0 Cr-iOOOO OOOOO SrHCO ^ Si •d 03 0 S -s g 00 5<1 -H O WS i— 1 P CO G^ O | (M Ol JO tn2SS^ S^co 3 a B 00 tN i-H 1 J>- lO i — t i— 1 Ti< O iQ CO OOCO 02 w -^i>coi-ii-i "irax £ SUD cS'1 0 g • §SSggS£§ •M O • 01 § 'o V q-i rJC 1 °J M * ElOO OlO OO"5 S3 r-l •* C iH COO oppoiraco ooo 'S C :I1 ^ ° I>r. A. Macia.lym, I)r. (>. H. Morris, ;iml Mr. S. l!.i\vlaml. between the residual alcohol and the carbon dioxide is very variable, an „ „ Oil Expressed Yeast-cell Plasma (Bnchner's " Zymase"). 265 plasm (anabolism) ; and (2) a breaking clown of this complex material into simpler products, of which carbon dioxide and alcohol are the constant and principal constituents (katabolism). May it not be that after the expression of the cell-juice from the cell the same series of actions continues to take place, at least for so long a time as the rapidly changing and unstable cell-juice remains in a condition ap- proximately identical with that in which it existed in the living cell ? If this hypothesis be admitted, then the varying activities of the juice are at least partly explicable, for if we designate by x the hypothetical protoplasmic constituent of the cell with which the sugar combines, then we may imagine the processes which take place in the expressed cell-juice (in which we assume x to continue to exist) to be somewhat as follows : — . (a) In the case of auto-fermentation the x-sugar combination, built up during the life of the cell, continues to decompose, after the ex- pression of the juice, yielding carbon dioxide and alcohol. (b) In the case of the disappearing sugar, the formation of the X-sugar combination continues to a certain point, depending on the activity of the juice, but the decomposition of this combination comes to an end before the whole of the sugar has been liberated in the form of carbon dioxide and alcohol. In the case of a very active juice we may imagine this process to continue until practically the whole of the combination has been decomposed. In the case of a weak juice, the building-up process takes place more rapidly than the breaking-down process, and, consequently, when the activity of x ceases, there remains an excess of sugar in the form of the x-sugar combination. We are continuing our investigations with the yeast-cell plasma, and shall hope to communicate our further results to the Society in due course. In the meantime it may be convenient to briefly summarise the results we have already obtained, which so far appear to be leading us in the direction not of an enzyme explanation of the process, but rather of a theory which refers the phenomenon to the vital activity of the yeast-cell protoplasm. (1.) The top-yeast of English breweries yields, by suitable treat- ment, a cell- juice which possesses the transient power of decomposing sugar into alcohol and carbonic acid. (2.) The amount of gas formed by an actives juice is as great as, or even greater than, that found by E. Buchner. (3.) The cell-juice as prepared by us undergoes a very considerable auto-fermentation, in some instances exceeding that given by a mixture of the same juice and cane-sugar. (4.) A moderate dilution (1 : 2) with water or physiological salt solution practically stops all fermentative activity. (5.) Only with a very active cell-juice does the ratio between the 266 I' 'i. IT. L. Callendai. 0 col alcohol and carbon dioxide formed approximate to that found in ordinary alcoholic fermentation. (6.) When the cell-juice is allowed to act on sugar — either cane- sugar or dextrose — the quantity of sugar which disappears is considerably in excess of that which can be accounted for by the production of carbon dioxide and alcohol. " On the Therm ody n amical Properties of Gases and Vapours as deduced from a Modified Form of the Joule-Thomson Equa- tion, with Special Reference to the Properties of Steam." I'.y H. L CALLENDAK, M.A., LL.D., F.R.S., Quain Professor of Experimental Physics, University College, London. Received and Read June 21, 1900. At the present time, the relations between the specific heats and other thermodynamical properties of gases and vapours, and the devia- tions from the behaviour of the ideal gaseous substance in isothermal and adiabatic expansion, remain extremely obscure. The variation of the latent heat of a vapour, and of its saturation pressure, are generally expressed by purely empirical formulae, without theoretical foundation. Various equations, such as those of Van der Waals, and Clausius, have been proposed and have been very generally adopted to represent some of the simplest of these relations, but owing to their complexity, and to the number of empirical constants involved, their utility is seriously limited, and the results to which they lead are in some cases undoubtedly erroneous. The object of the present paper, which is founded mainly on experi- ments on steam, is to develop the application of a modified form of the Joule-Thomson equation, which is sufficiently simple to be of great value in the discussion of the thermodynamical relations of gases and vapours, and which leads directly to accurate formulae for many pro- perties which have hitherto been represented empirically. To take the case of steam as an example, all tables of the properties of steam are at present founded on Regriault's formula for the total heat H of saturated steam at / C. reckoned from 0° C., namely : H = 606-5 + 0-305/ ........................ (1), and on his empirical formula for the pressure of saturated steam, namely : = a + bB' + cC* ........................ (2). The latter formula contains five empirical constants, but it is usual Properties of Gases and Vapours, &c. 267 to employ two different formulae to cover the range 0" — 220° C. of his experiments. The specific volume of saturated steam, owing to the effects of surface condensation,* cannot be determined by direct experiment, and is generally deduced from the above empirical formulae, by the application of the well-known thermodynamical relation, (3), where L is the latent heat, v the specific volume of the saturated vapour, and b that of the liquid, 0 the absolute temperature, and dp/dB the rate of increase of the saturation-pressure with temperature. Eegnault also determined the specific heat of superheated steam at atmospheric pressure by condensing highly superheated steam in a calorimeter ; but owing to the small proportion which the superheat of the steam bears to the latent heat, and to the difficulty of calori- metric work at high temperatures, the measurements were not very certain, and many recent experimentalists and writers (e.g., Ewing, Perry, Grindley) have preferred to adopt widely different values deduced by other methods from the formula for the total heat. It was proved by Rankinet that the rate of change of the total heat of steam at low temperatures, at which it very nearly follows the laws of an ideal gas on account of its low pressure and large specific volume, must be very nearly equal to the specific heat of the vapour at con- stant pressure. Therefore either the specific heat of steam at low temperatures must be O305, increasing considerably with the tempera- ture so as to reach the value 0'48 between 100° and 200° C., or else the observations of Regnault must be wrong. In any case it is clear that the variation of the total heat should not be linear, unless we abandon the experimental evidence in favour of the constancy of the specific heat of an ideal gas. It is most likely that the source of the discre- pancy is to be found in the difficult calorimetric measurements of the rate of change of the total heat at low temperatures. The determina- tions of the latent heat by Griffiths, 572'6 calories at 40'2° C., and by Dieterici, 596'7 calories at 0° C.,J are from 6 to 10 calories smaller than Regnault's, and imply a rate of change of total heat about 30 per cent, larger, and more nearly equal to the theoretical value. At temperatures above 100° C., the determinations of Regnault are more consistent, but it is very likely, from the method which he employed, that they may be considerably in error. His observations show a sudden increase of six calories above 175° C., which is explained by the discovery and rectification of a leakage of steam through the * Eamsay and Young, 'Phil. Trans.,' A, 1892. t ' Koy. Soc. Edin. Proc.,' 1850. I Griffiths, ' ROT. Soc. Proc.,' December, 1894. 268 I'M-:'. II. L. Callfiular. On the TV/-/-/ >>-al «li>tril>uting tap intx) the idle calorimeter. It is clear that he regards the observations ;il»ove this point with greater confidence. In any case it is unlikely that the order of accuracy attained in his experiments was greater than one-half of 1 per cent, at any point, because his thermometers were not sxifficiently perfect, and because it is practically certain from the recent determinations of Reynolds and Moorby of the mean specific heat of water between 0° and 100° C., and from the work of Callendar and Barnes on the variation of the specific heat over the whole range 0° to 100° C., that his value for the specific heat of water at 100° C. is at least 1 per cent, too large. He was also ignorant of the considerable changes which occur in the specific heat of water at low temperatures, and it is evident that his work, though far in advance of his time, requires revision when considered in the light of the great advances which have been made in the last fifty years. It is obvious from the nature of the problem that the most appropri- ate method of determining, either the variation of the total heat of steam, or the specific heat of steam, is by the application of some differential method, which shall be independent of the determination of the latent heat. In the papers which follow, I have described the application of two such methods to the case of steam. By means of the " Differential Throttling Calorimeter " it is possible, following the method of Joule and Thomson, to determine accurately the variation of the total heat of steam, and the deviations of the specific volume from the ideal gaseous state, in terms of the specific heat at constant pressure. By the "Electrical Method of Measuring the Specific Heat," which is exactly similar to the method already applied* in the case of water, it is possible to determine the specific heat without reference to the latent heat. The details of these experiments are reserved for subsequent communications, the object of the present paper is to explain the thermodyriamical relations involved, and to exhibit the calculation of the variations of the specific volume, the specific heats, the total heat, the latent heat, and the pressure of saturated steam in terms of the quantities which are directly observed. The theory of the method is applicable, and has been already applied, to some problems connected with gases, but in dealing with vapours some additions are required, and it is clear that the original equation of Joule and Thomson requires some important modifications. Modification of tlie Jwle-Thotruon A'y""//,i ?//• Tl» /•/,/<»/////"////' tion, which is equal to 3/x ~1 per unit mass ;it any tempt- follows from this assumption that the limiting value of the SJH-. -iti<- heat of a gas in the ideal state (p — 0, v = oc ), cither at constant ]>• or at constant volume, must IKJ constant, if the molecule is stable, since it is directly proportional to />/•/#, which tends to a constant limit \\lien p = 0, even in the case of vapours at temperatures far below their boiling points. These constant limiting values of the two fundamental specific heats will l>e denoted by the symbols S° and a" respectively. As a further simplification we may a-sumc that the kinetic energy of a vapour is proportional top (v - f>) at all stages and not only in the limit. On this assumption it is also necessary to suppose that the index of 0 in the small term «/R#- in the Joule-Thomson equation is not 2, but n = *°/R, the ratio of the limiting value of the specific heat at constant volume to the limiting value of prjd. If we adopt the hypothesis of Clerk Maxwell with regard to the distribution of energy between the various degrees of freedom of a molecule, which, in the absence of certain knowledge with regard to the exact nature of a molecule, appears to be the only practical working hypothesis, the theoretical value of this limiting ratio should be 1'5 for a monatomic gas like argon, 2*5 for a diatomic gas like oxygen or hydrogen, 3*5 for a tri- atomic gas like steam or COo, and so on, increasing by unity for each additional atom in the molecule. The value 3*5 for the index is closely verified in the case of steam by the experiments to be described on the Joule-Thomson effect, and also by the experiments on the specific heat, by which this relation was first suggested. Adopting these two modifications, of which the second is the more important, the equation may be written in the form, v-b = R6Jp-c°(60/6)n = V-r (6), in which V is taken as a convenient abbreviation for the ideal volume ~RO/p, and the co-volume b is taken as constant and equal to the volume of the liquid. The small correction c, representing the state of co- aggregation of the molecules, is called the "co-aggregation volume," and is a function of the temperature only, varying inversely as the //th power of the absolute temperature, where the index n is used as an abbreviation for .<;0/R. It is a quantity of the same dimensions as a volume, and is measured in cubic centimetres. The numerical value of f in the case of steam at 100° C. is 26'5 c.c., as deduced from the experiments on the Joule-Thomson effect and the specific heat. The calculated value of c° at 6° = 273'0° is 79*0 c.c. It is obvious on the simplest considerations that the co-aggregation volume r cannot remain accurately constant at high pressures, since there is an obvious limit to the possible co-aggregation of the molecules. If, for instance, the molecules are simply paired, the pairing must cease when v-b = c. But it is certain from the differential experiments that the modified Properties of Gases and Vapours, (fee. 271 equation represents a very accurate approximation to the facts at moderate pressures, although n is not necessarily equal to s°/R in all cases. Variation of th/j Specific Heats. It would appear at first sight as though the modified equation were more complicated than the original of Joule and Thomson, but it leads as a matter of fact to far simpler relations between the thermody- namical properties, and makes it possible to attack problems which would be quite intractable with the more complicated forms of empiri- cal equations in vogue. If S and s are the specific heats at constant pressure and at constant volume respectively, and <£ is the entropy, we have the well-known relations, (dS/dph = 8d*/dO dp = - 0(^/d0% ............ (7), (dsjdv)9 = O&ydO dv = + 0(d*p/d0*)v ............ (8). Assuming the characteristic equation (6), it is easy to prove from these relations that the values of the specific heats at any temperature and pressure are given by the simple formulae, S - S°(l+»c/V) = S° + n(n + I)pc/6 ............ (9), s = s°(l+ncj\)(l-c/V) ........................... (10), where S° and s° are the constant limiting values of the specific heats when p = 0. The ratio of the specific heats g — S/s is given by the relation, where g" stands for the constant limiting value of the ratio, and is equal to (n+ \}fn. Isentropic Relations. The isentropic relations are greatly simplified by the assumption n — s°/R, since in this case the ratio of the co-aggregation volume c to the ideal volume V, or to the difference of specific volumes of the vapour and liquid v-b, is constant at constant entropy. From the characteristic equation (6), we have for the isothermal elasticity Ee written in the form, c/\ = (cY where e is the base of natural logarithms, and the symbols p°, c°, V°, 6°, °, refer to a standard state such as 0° C., and 760 nun. The Change of Entropy <£ - ° in expansion from p° to p at a con- stant temperature 6" is deduced from the relation, d = (d*t>/dp)edp = -(dv/d6)pdp = -(R/p + ne9/P)dp ...... (17), whence -° = Rlog_p°/p + (/>° -p^c'/B" ............... (18). The Heat Absorbed is given by the expression, Q - Q° = K,e°logp0/p + (p0-p)nc0 ............... (19). The Work Done, W = R 6" log p°lp, is the same as for an ideal gas between the same limits of pressure. The Change of Intrinsic Energy is, therefore, E-E° = nc°(p°-p) ........................ (20). Heating at Constant Pressure. The Heat Absorbed in a rise of temperature from 6" to 6 at constant pressure p° is given by the equation Q - Q° = JSrf0=S'(0-09) + (»+l)/(e°-r) ......... (21). Change of Energy, E-E° = s9(0-tf) + np'(c'-r) .................. (22). The Work Done, W = R(0-6°)+p°(ce-r) ..................... (23). Change of Entropy, = s° log ofr + np°(co/e' - c/e) ...... (24). In passing from any one state represented by the co-ordinates p°, 0°, to any other state represented by p, 0, both the change of energy and the change of entropy must be independent of the process by which Properties of Gases and Vapours, &c. 273 the change is effected, and equal to the values calculated by combining isothermal expansion with heating at constant pressure. We thus obtain the following general expressions for the change of energy and the change of entropy in any transformation from the state p°, 0°, to the state p, 0 : — E-E° = i(d-B°}-n(pc-p°c°} (25). $ - ° = S° log 0/0° - E logjp// - n(cp/6 - c°p°/e°) (26). These expressions are true for any value of n. Calculation of the Specific Volume of Saturated Steam. As an illustration of the numerical application of this method, I propose to take the case of steam, as the most important and interest- ing. But the methods and reasoning would be equally applicable to any other gases or vapours for which the requisite experimental data were available. The deviations of the specific volume from the ideal state are imme- diately given by the values of the co-aggregation- volume c, which are easily calculated. It is quite a mistake to suppose, as is frequently stated, that there is any sudden or rapid change in the co-aggregation as the saturation point is approached. This idea has arisen merely from experimental errors due to surface condensation. Under certain conditions it is well known that the vapour can exist in stable equi- librium at pressures greatly in excess of the saturation value, provided that there is no liquid present, or any nuclei, or other aids to con- densation. I have not, for obvious reasons, succeeded in investigating the properties of supersaturated steam by the method of throttling ; but there does not appear to be any reason to suppose that its behaviour could not be predicted with great probability by assuming that the co-aggregation volume remains constant at constant temperature, which is certainly a very close approximation to the behaviour of steam in the superheated condition down to the temperature of satura- tion. In the following table, which contains a few sample values of the co- aggregation c and the specific volume v, the ideal specific volume of steam at 100° C. and 760 mm. pressure is taken as 1698O c.c., which is calculated by assuming the density of oxygen, corrected for its probable co-aggregation, and taking the ratio of the molecular weight of steam to that of oxygen to be 18/32. The value of c for steam at 100° C. is taken as 26 '5 c.c., and the values at other temperatures are calculated by the formula (6) : — VOL. LXVII. 274 Prof. If. L Callendar. On 11 'dynamical Table I. — Specific Volume and Co-aggregation of Steam. Temp. Cent. Satura- tion- pressure. Co-aRgre- gation. c. Ideal volume. V. Specific volume. r. Co-aggre- gation. Ratio. r/V. Deviation. Ratio. (<•-*)/ V. e atrnos. c.c. c.c. c.c. 0 0-00613 79-0 202680-0 202602-0 0-000389 0-000385 20 0 -02323 61-7 57370 -0 57309-0 0 -001076 0-0010GO 40 0-0731 49-0 19490-0 19442-0 0-002515 0-002463 60 0-1267 39-4 7710-0 7671 -0 0-00512 0-00498 80 0-4670 32-14 3438-0 3407-0 0-00932 0-00903 100 I'OOOO 26-50 1698-0 1672 -5 0 -01560 0-01500 120 1-961 22-07 911-7 890-6 0-02425 0-02305 140 3-570 18-56 526-0 508-4 0-0354 0-0333 160 6-01 15-74 321-7 307-1 0-0490 0-0455 180 9-93 13-41 207-7 195-3 0-0648 0-0594 200 15-37 11-55 140-1 129-6 0 0825 0-0742 The values of the specific volume r are calciilated for the saturation- pressures, given in atmospheres in the second column. The values for any other pressures can he calculated with equal ease, since the co-aggregation-volume c, which is given in the third column, depends only on the temperature. It is7 necessary in each case to calculate the appropriate value of the ideal volume V = HO/p. The values of the specific volume v are found by subtracting (c - 1} from the ideal volume. The values of the co-aggregation ratio c/V, and the deviation- ratio (c - b)/V, are also given, as the first is useful in calculating the values of the specific heats by formulae (9), (10), and (11), and the second affords the most convenient means of representing graphically the variations of the specific volume, since it is quite impracticable to piot the specific volume itself on an adequate scale. Graphic Representation of the Variations of Specific Volume. The best method of representing these results graphically appears to be that adopted in fig. 1, of plotting the ratio of t?/V or (c-6)/V against p. We have — pv[R6 - t;/V = l-(c-J)/V = l-(c-b)p[RO (27). Since r is constant at constant temperature, the isothermals are Properties of Gases and Vapours, 275 straight lines inclined at different angles to the axis, but all intersecting at the same point, v/V — 1 when p = 0. These lines are drawn for each 20° of temperature in the figure, and are all represented as terminating in the saturation curve, although, as a matter of fact, it is 99 -98 .§ -97 « *t> •95 o -94 1 •53 V O .92 •91 •90, O-2 of Pressure for o°-/oo°C. O-6 OO I-O <3,tmO3. £OOTx. 4.6 5 10 IB M Sc&Le of Pressure in Atmospheres. /(X?-£OoC. Fig. 1. — Specific volume of steam. possible to suppose them produced beyond it, if condensation does not occur. The first part of the curve up to 100° C. is represented on a ten times larger scale of pressure in the upper part of the figure. It will be observed that the whole deviation from the ideal volume at 100° C. is only 1'5 per cent, at saturation-pressure. This method of x 2 27G l'n>I'. 11. L < 'ullrmhir. On the T/n-rtin»/iin»i/i;- plotting is analogous to that rendered familiar by Amagat and others in the case of the deviation of gases from Boyle's law. It is usual to plot the product pv against p, but it seems to me to l>e preferable to plot />v[R6 instead of pv, because the pv method confuses the diagram by introducing the effects of the variation of temperature, so that the different isothermals cannot be so well compared, and their relations til.-erved. In the diagrams of Amagat and others, who have adopted the direct method of measuring the whole specific volume instead of the differ- ential method of observing only the deviation from the ideal volume, the isothermals are not accurately straight, but always bend down- wards more steeply as saturation is approached, so that they are concave to the axis of pv. There seems reason to believe that this peculiarity may be partly due to the effect of surface condensation so well established by the observations of Ramsay and Young.* It is true that a similar though smaller increase in the slope results from the work of Xatanson on the Joule-Thomson effect for CO., at pressures up to 26 atmospheres, at a temperature of 20° C. But in that case also the effect may be explained by condensation in the pores of the porous plug, as is indicated in some of the work of Joule and Thomson. I was not able to find by the differential method, which would eliminate any error of this kind, any trace of this effect at moderate pressures. In fact, the cooling effect appeared to diminish very slightly with increase of pressure, as it should, on account of the small increase in the value of the specific heat with increase of pressure. If this is generally true for other vapours, it would appear possible that many of the complica- tions which have been introduced in current forms of empirical equa- tions of the fluid state, may serve only to represent errors inherent in the experimental methods on which they are founded. Values of the Specific Heats of Steam. The values of the specific heats at any temperature and pressure are easily calculated from their limiting values at zero pressure, by means of the formulae (9), (10), and (11) already given. The actual value of the specific heat at a pressure of one atmosphere was experimentally determined by the electrical method to be subsequently described. The value so found, though slightly larger than Regnault's, agreed so well with the theoretical value deduced from the characteristic equation (6) by means of Maxwell's assumption, that there can be little doubt that the method of deduction employed is valid. The value of the constant K for steam is readily found from the value of the ideal volume already assumed, we thus obtainf * « Phil. Trans.,' A, 1892. t Assuming that the pressure due to a column of mercury 760 mm. in height at 0°C. and sea-level in latitude 45° is equal to 1*0133 megadynes per sq. cm. Properties of Gases and Vapours, &c. R = 4-613 x 10° C.G.S. = 0-4613 joule/deg. C. = 0-11037 cal./deg. C. The unit of heat adopted in this paper is the thermal capacity of 1 gramme of water at 20° C., which is taken as being equivalent to 4-180 joules, from the mean of the results of Rowland and of Reynolds and Moorby, compared and reduced by the work of Callendar and Barnes on the variation of the specific heat of water over the whole range 0° to 100° C. The limiting values of the specific heats of steam, and of their ratio, in terms of this unit are as follows, S° = 0-4966 cal./deg. C. s° = 0-3862 cal./deg. C. g" = 9/7 = 1-2857. In the following table, the values of S, s, and g are given for saturated steam at the point of saturation, in order to illustrate the increase of specific heat with pressure. The values of the specific heat S' at a pressure of one atmosphere at various temperatures are also given, to show the diminution of the specific heat with rise of temperature. The values enclosed in brackets are of course imaginary, but are included to show more clearly the nature of the change. The value of the specific heat at constant pressure has been calculated by Zeuner from Regnault's observations to be 0'568, on the assumption that the specific volume of steam is a linear function of the temperature at constant pressure, in which case the specific heat at constant pressure is independent of the pressure. Another common assumption is that the pressure at constant volume is a linear function of the temperature (Van der Waals). Neither of these assumptions can be reconciled with the most accurate thermometric work at moderate pressures, or with the present experi- ments on steam by the method of the differential throttling calorimeter. The advantage gained by these assumptions is very slight and one-sided. The partial constancy of one specific heat is a small matter, if at the same time the other thermodynamical relations are rendered so com- plicated as to make the equations useless. The values of the specific heat of steam at constant pressure have been recently calculated by Grindley from his throttling experiments,* assuming the linear formula (1) of Regnault for the Total Heat of steam. These values are included in Table II for comparison. The numbers given in brackets are not given by Grindley, but are calculated by an extension of his method to show the effect of his hypothesis. The extraordinary differences between his values and mine, as shown in the adjacent columns (6) and (7) of Table II, are not due to any discrepancy in our experiments, but simply to his assumption of Regnault's formula for the total heat. The method of deducing the specific heat from the total heat, though it has often been applied, is unsound in principle, because the specific heat * ' Phil. Trans.,' A, vol. 194, 1900, pp. 1-36. 278 1'rof. H. L. Callendar. On the Tl "////'<•,// depends on the rate of variation of the total heat, so that all the errors in the formula for the total heat are enormously exaggerated in the calculation of the specific heat. The last column in the table contains the values of the so-called " Specific Heat of Saturated £>team," i.e., the value of the specific lu-at when dpjdO is determined by the condition that the steam is to remain saturated. The value in this case is given by the formula S(sat.) = dE/dO-L/0 (28), which is most readily obtained by differentiating the expression for the entropy of saturated steam, namely, g = <}>w 4- L/0, where <£,P is the entropy of water reckoned from 0° C. The values of S (sat.) also differ considerably from those which have been previously calculated on the usual assumption that (fH/dO is constant, and equal to O305, as in formula (1). It will be observed that the mean value of the specific heat of steam at a constant pressure of one atmosphere between the limits 120° and 200° C., the range of Regnault's experiments, is 0'512 according to the values given in Table II. This value is not so greatly in excess of the value 0'48 given by Regnault as to be beyond the limits of experi- mental error, especially if we consider that the method which he adopted must necessarily have given rise to constant errors in defect, and that the superheat was only one-sixteenth of the total heat to be measured. It is probable that the variation of the specific heats of all other gases and vapours, to which the Joule-Thomson equation can be applied, is of the same type as that exhibited above in the case of steam. I have succeeded in reconciling a good many of the appa- rently discordant experimental data on the subject by means of this hypothesis, but the experiments themselves are difficult, and the question of the variation of the specific heats of gases is obscured by unavoidable errors. Among the most remarkable and accurate of recent results are those of Joly on the specific heats of Air and CO* at constant volume, determined by means of his differential steam- calorimeter. The values which he obtained are much larger than those deduced from Regnault, and cannot be reconciled with them on the common assumption (Van der Waals) that the specific heat at constant volume is constant, but they agree remarkably well, considering the difficulty of the experiments, with the theory proposed in this paper. A fuller discussion of these and similar relations will be reserved for a future communication. Properties of Gases and Vapour*, &c. Table II. — Specific Heats of Steam. 279 Temp. Cent. t. At saturation point. Satura- tion- pressure. . P- At one atmosphere. S(sat.). Calories/deg. Eatio. Calories/deg. S. *. 9- S'. Grindley. Cal./deg. o 0 0 -4973 0-3866 1-2862 mm. 4-66 |(0 -6056) (0 -306) -1-680 20 0-4984 0 -3873 1 -2870 17-67 (0 -5758) (0 -308) -1-502 40 0-5008 0 -3885 1 -2888 55-55 (0 -5544) (0 -314) -1-351 60 0-5055 0 -3913 1 -2921 140-63 (0 -5416) (0 -326) -1-223 80 0 -5128 0 -3952 1-2977 355 -30 (0-5313) (0 -349) -1-116 100 0 -5236 0-4009 1-3060 760-00 0 -5236 0-387 -1-028 120 0 -5388 0 -4090 1 -3175 1491 -4 0 -5181 0-448 -0-955 140 0-5581 0-4188 1 -3327 2716 -5 0-5138 0-537 -0-895 160 0-5816 0 -4303 1 -3518 4657 -0 0-5105 0-665 -0-844 180 0-6086 0-4427 1 -3748 7546-0 0 -5079 (0 -827) -0-801 200 0 -6399 0-4568 1-4012 11684 -0 0 -5059 (1-043) -0-759 Variation of the Total Heat and Latent Heat. The Total Heat, H, of a vapour, defined in the usual manner, is related to the Latent Heat, L, by the simple equation H = L + A (29), if the Heat of the Liquid h is reckoned from the same zero as the total heat of the vapour. In the case of water both H and h are reckoned from the state of water at 0° C. The specific heat of water is so nearly constant that h may often be taken as equal to t. More generally we may write h = t + dh (30), where dh is the small difference of the heat h from the value t, which it would have if the specific heat were constant and equal to unity. According to the observations of Barnes, the variation of the specific heat is very nearly linear between 60° and 100° C. The value of h at temperatures above 60° C. may be taken as h = t + dh = * + 0-0001 10(*-60)2 (31). 1'SO Prof. H. L Callendar. On the Tkirmodjriumieal This simple formula agrees very nearly with Regnault's observations in the rate of variation above 60°, and also with the table of values of h given by Callendar and Barnes.* The accurate relation between the Total Heat and the Specific Heat of the Vapour is readily obtained by equating the intrinsic energy of steam evaporated at 0° C. at a pressure p", and then heated at constant pressure p° up to any temperature 0, to that of steam obtained by heat- ing the liquid up to the same temperature 0, evaporating it at 6 under the constant saturation pressure p, and expanding the vapour at constant temperature 0 down to the pressure p" of saturation at 0° C. The various changes of intrinsic energy involved in these processes are given by equations (20) and (22). After a few simple reductions we obtain the Equation of Total Heat, H-H° = S°(0-0°)-(n+l)(cp-cy) (32), which is simply the expression of the first law of thermodynamics as applied to the problem, and might also have been obtained in various other ways. If we omit the small terms depending on the co-aggrega- tion c, the equation is identical with that given by Kankine in 1850, on the assumption that saturated steam could be treated as an ideal gas. The small terms represent the effect of the deviations of steam from the ideal state, and become important at high pressures. The equation neglects the external work of expansion of the liquid, but this is less than one-thirtieth of a calorie at 200° C., although it may become important as the critical temperature is approached. Equation (32) gives only the variations of the total heat of the saturated vapour. In order to find the absolute values, it is neces- sary to know the actual value of the total heat at some particular temperature. The obvious value to select would be that given by Regnault at 100° C., namely 637 calories. His methods do not appear, however, to have been sufficiently exact, and I prefer to rely on a more recent determination by Joly, with his steam calorimeter (described by Griffiths), t Joly determined the mean specific heat of water between 12° and 100° C 'in terms of the latent heat of steam at 100 C. Now the mean specific heat of water between 12° and 100° C. is known in terms of the specific heat at 20° C. by the results quoted above. We can therefore reverse the calculation, and find the latent heat of steam at 100° C. The result of the calculation gives L at 100° = 540-2 calories at 20° C. This is considerably in excess of Regnault's value, but it is quite within the limits of probable error of his experiments, and it possibly still errs in the direction of being too low. Assuming this value as a starting point, I have calculated the following table of * 'Brit. Aasoc. Hep.,' 1899. t ' Phil. Trans.,' A, 1895, p. 322. Properties of Gases and Vapours, &c. 281 values of the total heat and latent heat, and of the rate of variation of the saturation pressure and the total heat with temperature, namely dp /dO (sat.), and dH/dO. It will be observed that the rate of increase of the total heat diminishes rapidly at high temperatures, while the rate of diminution of the latent heat increases. This must necessarily be the case, as the latent heat should vanish near the critical tem- perature, which occurs about 365° C., according to the observations of Cailletet and Colardeau, whereas the linear formula of Regnault would make the latent heat vanish at about 870° C. This is an additional indication of the impossibility of Eegnault's formula. It may be observed, however, that the average rate of increase of the total heat, according to Table III between 100° and 200° C., over the range of Regnault's experiments from which the linear formula (1) was calcu- lated, is only 0'330 calorie per degree, which differs so little from the coefficient O305 given by Regnault as to be well within the limit of accuracy of his experiments, considering the acknowledged leakage of the distributing tap, and that the whole difference is only one-half of 1 per cent, on the quantity of heat measured. The values of dp/dO, given in column 6 of the table, are calculated from those of L by means of the thermodynamic relation (3) already quoted, assuming the values of the specific volume from Table I. Table III. — Total Heat and Latent Heat of Steam in terms of the Thermal Capacity of Water at 20° C. Temp, cent. H-H° cals. H. cals. li.- cals. L. cals. dpjde. mrn./deg. dH/dO, cals./deg. Eegnault's formula. 0 O'OO 593-5 0 -00 593 -5 0 -3364 0 -4935 606-5 20 9-83 603-3 20 -06 583 -2 1-069 0-4887 612-6 40 19-53 613-0 40-02 573-0 2-952 0-4800 618-7 60 28-99 622-5 60-00 562-5 6-904 0 -4665 624-8 80 38-15 631-7 80-03 551-7 14-39 0-4467 630-9 100 46-83 640-3 100-14 540-2 27-164 0 -4202 637-0 120 54-90 648-4 120-3 528-1 47-35 0-3885 643-1 140 62-32 655-8 140-6 515-2 77-06 0-353 649-2 160 68-9 662-4 161-1 501-3 118 -62 0-314 655'3 180 74-9 668-4 181-6 486-8 173 -50 0-274 661-4 200 79-9 673 -4 202-3 471-1 243-0 0-237 667-5 I 'rut'. If. L. ('aileiuliir. On the T< <"l The values of e observed that the values of the total heat in the above table agree very closely with those of Regnault between 60° and 90% where his results (according to Griffiths)* are most reliable. The average of his observations over this range is in exact agreement with Table III. At lower temperatures, Regnault's observations are very discordant, but the values given in the table are well supported by those of Griffiths. He finds, for instance, the latent heat to be 572*7 calories at 40*2° C., where the table would give 573*1. The unit employed by Griffiths (calorie at 15° C.) is different, so that his value would require to be raised nearly 0*6 cal. to reduce to the same unit, which would make it 573*3 calories at 40*2°. There can be no doubt that his observations are entitled to much greater weight than those of Regnault, which are nearly 6 calories larger at this point. The value, 596*7 found by Dieterici with an ice calorimeter,! for the latent heat at 0° C., is 3*2 calories larger than that given in the table. But it must be remembered that observations of the latent heat at 0° C. are not at all easy, and that there is some uncertainty about the unit employed by Dieterici, as he finds by the same method the value of the mean specific heat of water between 0° and 100° C., about 1 per cent, larger than Reynolds and Moorby or Callendar and Barnes. It is possible that his result might agree with Table III if it could be reduced to the same units. Entropy of Wvder and Steam. The entropy of water, <£fr, is readily calculated from a table of the values of the specific heat when the variation of the specific heat is known. Since, however, the specific heat is nearly equal to unity, we may write, w = \og*6/8° + d (33), where d is the small difference of the value at any temperature from the value loge 0/0°, calculated on the assumption of constant specific heat. The values of d given in the following table have been calcu- lated from the table of values of the specific heat given by Callendar and Barnes. J It will be seen that the values of d are very small at temperatures below 100° C., but increase rapidly at higher tempera- tures. Above 200° C. the values of the specific heat of water are so uncertain that it is not possible to calculate the properties of steam satisfactorily by any method. The values of dh/6 = (h- /)/# are also * ' Boy. Soc. Proc.,' December, 1894. t 'Wied. Ann.,1 vol. 37, p. 504, 1899. J ' Brit. Assoc. Kep.,' 1899. Properties of Gases and Vapours, &c. 283 given in the table for comparison with those of d(f>. They are calcu- lated from the same table of the values of the specific heat, and the difference d - dh/0 is seen to be small. This difference occurs as a small correction in the equation for the saturation-pressure, to be pre- sently given. The values of the entropy of water, corrected for the variation of the specific heat are given in the column headed <£w. The values of the entropy of steam are obtained by adding the values of L/0 given in the next column, which are found from the values of the latent heat already given in Table III. Values of g, calculated from Regnault's formulae, are given in the last column for comparison. Table IV. — Entropy of Water and Steam. 1. dkje. cty. Log, 0/0°. ft* L/0. *«. Keg- nault. o 0 0 0 0 0 2-1740 2 -1740 2-214 20 0-00016 0 -00017 0-07070 0 -07087 1 -9908 2 -0617 2-089 40 0 -00008 0 -00009 0-13672 0-13681 1 -8308 1 -9676 1-982 60 o-ooooo 0-00001 0 -19867 0 -19868 1 -6892 1 -8880 1-890 80 0-00009 o-oooio 0 -25694 0-25704 1 -5628 1-8198 1-814 100 0 '00037 0-00039 0 -31208 0 -31246 1-4483 1 -7608 1-748 120 0 -00084 0 -00090 0-36428 0-36518 1-3438 1-7090 1-692 140 0-00152 0 -00163 0 -41404 0 -41567 1-2475 1 -6632 1-644 160 0 -00242 0 -00231 0 -46112 0 -46373 1 -1578 1 -6215 1-604 180 0 -00353 0 -00384 0 -50645 0-51029 1-0746 1 -5849 1-568 200 0 -00497 0 -00532 0 -54960 0 -55492 0-9960 1 -5509 1-536 The Entropy Equation. By the application of the second law of thermodynamics we may obtain a relation between the latent heat and the saturation-pressure. Expressing the fact that the entropy of steam evaporated at 0°, and heated at constant pressure p° up to any temperature 0, is the same as the entropy of steam obtained by heating water up to the same temperature, 0, evaporating it at saturation-pressure, p, and then expanding it at constant temperature, 6, down to the pressure, p°, of saturation at 0°, we immediately obtain the entropy equation (i - s°) ioge e/e° + L/e - L°/e° + K iogep/P° 0 = 0 (34). Prof. II. L. ('allrndar. On ih< Tin •/•?/*'«/////"/// /»• - dh/B) (35). Neglecting the terms depending on the co-aggregation, and on the variation of the specific heat of water, this equation is equivalent to one given by Dupr£ and Bertrand, and rediscovered in various ways by many other observers (e.g., Pictet and Hertz).* If the correct values of S°, L°, and R are inserted in the formula, the equation thus simplified gives very accurate values of the saturation-pressure at low pressures where the properties of steam satisfy approximately the fundamental assumptions made in deducing the formula. Bertrand, t although of course he was well aware that the formula thus obtained was not accu- rate at high pressures, has calculated numerical formulae of this type for a large number of vapours, choosing the constants empirically so as to obtain the best agreement over the whole range. The values of the constants so found do not, of course, agree with the correct values of L° or S°. The numerical values chosen by Bertrand in the case of water, for instance, give L° = 573 calories, S° = 0*575 cal. per deg., and the value of the steam pressure found is 763 mm. at 100° C. At low temperatures the first term in formula (35) is the most important, since log« djB" is very nearly equal to t\B when t is small. The formula then reduces to the simple type, \ogp = A + B/0, which has often been employed for approximate work, and is the basis of the useful relation of Ramsay and Young.J Adding a second term, C/02, to this formula * Pictet, ' Comptes Rendus,' vol. 90, p. 1070, 1880 (proof invalid) ; Hertz, ' Wied. Ann.,' vol. 17, p. 177, 1832. t ' Thermodjnamique,' p. 93. j ' Phil. Mag.,' vol. 21, p. 33. Properties of Gases and Vapours, &c. 285 to take account of the small terms in equation (35), we obtain the well- known empirical formula of Kankine (1849), which is very convenient and accurate. A nearly equivalent formula is that of Unwin,* log p = A 4- B06, in which the same effect is empirically secured by an arbitrary exponent. These formulae are purely empirical, but it is interesting to observe how they are related to the correct thermodynamical expres- sion (35). Saturation-Pressures of Steam. In employing equation (35) to calculate the numerical values of the saturation-pressure in the case of steam, we have only one empirical constant, namely, p°, which is determined by the condition that the saturation-pressure at 100° C. is 760 mm. The values of the other constants which occur in this equation have been already given, namely, S° - 04966 cal./deg. L° = 593-5 cals. R = 0-11037 cal./deg. The value of 6°, which is also one of the fundamental data, is taken as being 273°, but is uncertain to the extent of 0'1°. Dividing the equa- tion by R, and reducing to common logarithms by the modulus loge 10 = 2-3026 = m, we obtain the numerical formula loglopfp° = - (d<}> - dh/0)/mR ...... (36). in which F(#) stands for the function logio 0/B° - t/mO. I have used this form for calculation, and have -given the values of the separate terms in Table V so as to show their relative importance. It is also possible to write the formula in the shape, log p = A + ~B/6 + C log 6 + small terms, but this does not show so clearly the relative effect and im- portance of L° and S°. The values of the saturation-pressure in the column headed p are calculated by the complete thermodynamic formula (36). The values given in the column headed Regnault are those of Regnault, recalcu- lated by Peabody and reduced to latitude 45°. The difference expressed in degrees of temperature is given in the last column, and is probably within the limits of error of Regnault's observations and of the empirical formulae employed to represent them. If we refer to the actual obser- vations of Regnault, we find that the discrepancies of individual observations at any point, expressed in degrees of temperature, exceed the values of the differences shown in the last column. We also find that in most cases the actual observations agree better with the single formula (36) than they do with the two empirical formulae, each with five arbitrary constants, from which the values in the column headed * ' Phil. Mag.,' Tol. 21, p. 300. 286 I >r. (r. Johnston i > Stont-y. Table V.— Saturation -Pressures of Steam. (. -4-661F(0). 5B dt-,11, 8 ~«B " f- B«*- nault. Diff. o 0 0 0 0 0 4-66 t-,;,, 0-°18 •2 > 0*MM -0-0019 + 0-0003 0 17-68 17-40 11 i>n 40 1-0932 -0-0177 + 0-0003 0 :,:, •:>:> 54-91 0*10 60 1-5413 -0-0367 + 0-0020 0 149-63 148-80 0-12 80 1-9387 -0 0603 + 0 -0039 0 355-30 351-03 0-04 100 2 -2932 -0-0872 + 0-0066 -o-oooi 760-00 760-00 0 00 120 2 -6120 -0 -1169 + 0-0104 -0-0002 1491 -4 1490-5 0-02 140 2-8998 -0-1488 + 0 -0152 -0 0005 2716-5 2717 '9 0-02 160 3-1612 -0-1817 + 0-0211 -0 0008 4657 4652 0 05 180 3-3990 -0-2163 + 0-0279 -0-0012 7546 7537 0-05 200 3-6170 -02514 + 0-0357 -0-0018 11684 11664 0-08 Kegnault were calculated. Taken in conjunction with the differential throttling experiments, and with the direct measurement of the specific heat by the electrical method, this is very strong evidence that Regnault's formula for the total heat is incorrect, and that the values of the total heat and latent heat given in Table III, and supported by the experiments of Griffiths and Joly, should be accepted in its place. "Note on Inquiries as to the Escape of Gases from Atmospheres." By G. JOHNSTOXE STOXEV, M.A., Hon. D.Sc., F.R.S. Received and Head June 21, 1900. We have now three investigations which profess to supply informa- tion about the escape of gases from atmospheres. Two of them, those of Messrs. Cook and Bryan, reason forward by the help of the kinetic theory of gas from the supposed causes ; the third, which is that pre- ferred by the present writer, reasons backward by the help of the same theory from the observed effects. Mr. Cook's investigation, which will be found in the ' Astrophysical Journal ' for January, 1 900, seeks to compute the proportion of mole- cules which can attain the speed requisite for escape by means of the formula which Maxwell published in 1860, assigning the proportion of On Inquiries as to the Escape of Gases from Atmospheres. 287 particles whose speed lies between v and v + dv, in a system of colliding particles intended to represent an isotropic portion of gas. Professor Bryan's investigation* is based on the investigations made since 1866 into the way in which energy tends ultimately to be par- titioned among the various motions possible within a self-contained dynamical system of bodies. The system need not be isotropic, since the bodies may be moving in a constant field of force. An inquiry by the present writer into Mr. Cook's method of dealing with the problem is attempted in the May and June numbers of the 'Astrophysical Journal' for 1900, and in the present paper a similar attempt is made with reference to Professor Bryan's. Both Mr. Cook and Professor Bryan predict the proportion of mole- cules which can escape from an atmosphere by deducing the proportion from its supposed causes, and in this respect are in contrast with an investigation previously published, which sought to ascertain from the observed effects of escape where and on what scale it has in fact taken place. (See memoir by the present writer in the ' Scientific Transactions of the Royal Dublin Society,' vol. 6, Part 13, or in the 'Astrophysical Journal' for January, 1898. And for further evidence that helium is escaping from the earth, see 'Nature' of the 24th May, 1900, p. 78.) Where, as in the present instance, the a priori and a posteriori methods have led to inconsistent numerical results, there must be a mistake or mistakes somewhere, and it is incumbent upon us to search till these are detected. If they can be. found and corrected an • important advantage will be gained. Professor Bryan, at the end of his letter in 'Nature ' of the 7th June, 1900, indicated one place where a mistake may have been made, viz., in the assumed relation between temperature and the kinetic energy of the translational motions. Another mistake may perhaps have been made in assuming the legiti- macy of treating the partition of energy when molecules move in a field of force, as though the only partition to be considered is between these molecules, whereas no field of force can exist unless it has been produced by some physical agent, upon which every motion that goes on within the field must react. In consequence of these reactions no field of force in which any motion occurs can be accurately constant, and a partition of energy based upon the supposition of its constancy is a theorem in rational dynamics, but has no counterpart in nature. Thus, in the case of the earth's atmosphere, the anisotropic condition of its outer layers is due to the field of force which exists in the neigh- bourhood of the earth ; and when we are obliged to take into account this anisotropic condition, as we must when dealing with the escape of gases from atmospheres, this is to be done (when we are treating the problem as one of partition of energy) by including as molecules between which the partition has to take effect not only the gaseous molecules, * ' Roy. Soc. Proc.,' April, 5, 1900, p. 835. 288 I >r. < 1. .F.'lmston.' Stmicy. Init also all the other ait rat-tin-; molecules which provide the fieM of force. [So again with reference to the never-ceasing turmoil which goes on in the atmosphere, which near the surface of the earth exhibits itself in tempests, thunderstorms, and other phases of weather, and which in the tipper regions includes phenomena still more extensive and swift. It is manifest that these events increase the opportunities which gaseous molecules have of escaping from the earth, and that accordingly flm/ //m. H into '"•/•llint(1 either explicitly or implicitly, in every valid inquiry as to the rate of escape. To take them into account in an investigation based on the partition of energy, we have to extend that partition to whatever agency pro- duces the turmoil. Now the activity within the atmosphere (and in fact almost every molar activity upon the earth other than the little which is attributable to tidal action or to such minor agencies as earth- quakes and volcanoes) is caused by the shiftings about of energy which come in between the continuous advent of energy by radiation from the sun, and its continuous escape from the earth by radiation into space. Hence to render an investigation by the Boltzmann- Maxwell law valid it is necessary to extend the partition of energy beyond the atmosphere — first to the solid earth, so as thereby to take account of the anisotropic character of some of the atmospheric strata (which facilitates the escape of gas) ; and secondly to embrace at least the sun and the aether between the earth and sun, so as thereby to take into account the turmoil in the upper regions of the atmosphere (which further increases the rate of escape). It seems to be only when these extensions shall have been effected that a generalised law such as the Boltzmann-Maxwell law for the partition of energy between the various degrees of freedom can become competent to furnish any information with reference to the rate at which gaseous molecules actually do escape from the earth. — July 17, 1900.] Then as regards temperature. The temperature of a solid is in reality twofold ; it is either its radiation temperature or its conduction temperature. These are physically distinct, although in all but some exceptional cases they are so nearly proportional to one another that they may be given the same mathematical expression. So, again, when dealing with gases we do well to keep in mind the essential distinction between radiation temperature and what may be called convection temperature. The temperature of an isolated gaseous molecule moving by itself through space is of the first land only, and depends exclu- sively on the energy of the internal motions — those motions within the molecule which enable it to absorb or emit radiant heat — and it is in no degree affected by the kinetic energy of the translational motion of the molecule ; whereas if the same molecule form part of a gas, it meets with encounters with other molecules or with the walls of a containing On Inrjuirie* as to the Escape of Gases from Atmospheres. 289 vessel, and at each such encounter there is a partition of energy between the translational and the internal motions, and in consequence of this the kinetic energy of the translational motion becomes a part of what determines that average power of absorbing and emitting radiant heat which (when estimated over a time embracing a sufficient number of encounters) is the proper definition of the radiation temperature of the molecule. Accordingly the average kinetic energy of the transla- tional motions of the molecule enters into its mathematical expression. If the gas be dense, encounters are frequent, and A£, the time requi- site for the averages, may be brief. In this case the radiation tempera- ture of a molecule, while the gas is undergoing some change in its condition, is predominantly the oiitcome of its encounters, and depends mainly on the molecules that surround it ; whereas if the gas be very much attenuated, then the radiation temperature of the molecule during a period of transition will depend mainly on what influences then reach it from the surrounding aether, and will be but in a subordi- nate degree affected by the encounters to which the moleciile at aboiit that time happens to be subjected. This is a matter which needs to be very fully taken into account when we attempt to estimate the escape of molecules from the earth's atmosphere, inasmuch as a large part of the heat radiated by the sun to the earth is absorbed by the gaseous molecules which happen at the time to be moving about in those strata of the atmosphere from which alone there can be any effective escape. Accordingly it will .need to be carefully scrutinised whether this has been either explicitly or implicitly taken into account in the attempts which have been made to determine a priori the rate of escape. W4ien the molecules of a gas or of a mixture of gases move in a field of force such as that surrounding the earth, convection currents can exist, and the term temperature as applied to the gas becomes ambiguous. It may have either of two distinct meanings, one of which has reference to the transport of heat by convection and by the con- sequent sweeping of successive portions of gas against bodies immersed in it, and the other has reference to the exchanges of heat by radia- tion with those or with more distant bodies. These are different physical events, and the assumption that they stand in a fixed ratio to one another is convenient, but is often not true. It is probably legiti- mate to regard it as approximately holding good in a gas which has nearly reached a final, i.e., an unchanging condition, and where the problem with which we are dealing does not need our making any closer scrutiny than as to what on the average happens to a sufficiently large swarm of molecules within a sufficiently long duration ; but it is not true while gas is passing through transition stages, nor is it true of individual molecular events or of small swarms of events, even in gas which has reached its final state. VOL. LXVII. Y •J'.iii Dr. C. Johnston-- Si..n,.y. Now, none of the gases of the atmosphere have even approached any such state. Changes incessantly go on in the open air at the bottom of the atmosphere, and the extent and abruptness of the changes that as incessantly go on in its upper regions are probably greater. Again, the consequences of cumulative effects arising in the illimit- able trains and combinations of encounters that are taking place, and of associated events in the aither, will also need to lie either explicitly or implicitly taken into account in any valid investigation of the escape of gases from atmospheres by the deductive method. All the circumstances that have been referred to would have to appear among the data of an ordinary dynamical investigation of the escape of an individual molecule from an atmosphere, if such an investigation were possible ; and the claim of a generalised theorem like that of the partition of energy to render it unnecessary to go into these details, ought to be carefully scrutinised. In one case at least the claim does not appear to stand this test, viz., in reference to the supposed legitimacy of the assumption that the field of force surround- ing the earth is constant. Though its variations are minute they are none the less real, and are due to interactions between each gaseous molecule and all the molecules of the solid earth, as real as are the interactions between gaseous molecules when they encounter, and as much entitled to be taken into account, when we seek to carry on the investigation in the region of generalised propositions. It should be kept in mind that in reference to what happens within this region, the plea of l>eing so minute as to be of negligible amount is not admissible. Whether a very small factor may or may not be neglected must be determined independently in each individual case ; and in the above instance the decision is that it may not be neglected. Other corrections might be suggested along with the principal ones noticed above — that relating to the two kinds of temperature, that relating to the field of force, and that relating to turmoil in the atmo- sphere ; but what seems most to be wanted is that we should recognise that any law for a distribution of energy within the atmosphere by itself, can only come approximately into practical effect after the lapse of a sufficient duration, and throughout a column of the atmosphere from which accidents are excluded ; and that this law will not be the Bolt/mann-Maxwell law, which may not be so restricted. Thus, let us imagine a cylinder like a great Tower of Babel, reaching to the top of the atmosphere, with walls competent to intercept dynami- cal, electrical and all other extraneous influences other than gravitation. The air within this tube would consist of molecules, moving in a field of force caused mainly by the earth's attraction and rotation, and this column of air might perhaps after some such period as a month, a year, a century, or a thousand years nearly attain such a distribution of On Inquiries as to the Escape of Gases from Atmospheres. 291 energy as that indicated by some law. But if while this process is matiiring a wind overthrows the tower, sweeping away the air it con- tained and substituting other air under new conditions, and subject to all the chances of uprushes, downrushes, thunderstorms, auroras, cyclones, cloud, sunshine, rain, &c. ; then after all or any of these or of the like accidents, the tower would have to be rebuilt before any portion of the atmosphere extending from the bottom to the top could find itself in a position even to commence the first steps of an advance towards at some future time complying with the law. The supposition then that the Boltzmann-Maxwell law can be restricted within our existing atmosphere would appear to be a mistake ; and if so the inferences from that law are not part of a real interpre- tation of nature. It need not therefore be matter of surprise that, in the case of helium, the facts of nature seem to negative those inferences. The weather which will prevail over the earth this day month will be the outcome of the present molecular state of the earth, and of the molecular events which will happen in the meantime ; but our power of stating in mathematical form the existing state of the earth, and our knowledge of molecular physics, are not such as would enable us to predict that future weather by the a pviwi or deductive method of proof. The difficulties in this case are easily seen; and they are instructive, since the escape of gas from the earth depends on pheno- mena which are probably as complex as those which determine the weather and as little amenable to treatment by the deductive method. Any such distribution of energy as that assigned by the Boltzmann- Maxwell law would, if it could be realised, be brought into existence by the gradual eftacement of excesses which had previously existed ; from which it would appear to follow that excesses prevail in our existing atmosphere greater and more numerous than could exist in an ideal atmosphere that obeyed that law. It is probable, therefore, that in our actual atmosphere there are more opportunities for the escape of molecules than there would be in the ideal atmosphere— a conclusion which accords well with the fact that the actual rate of escape exceeds those computed by Professor Bryan and Mr. Cook, (See 'Nature' of May 24, 1900, p. 78, second column.) Y 2 292 Mr. A. Kdington. South A f rim n //,//••>, -*/,/ "Smith African Horse-sickness: its Pathology and Mrtho.N »f Protective Inoculation." By ALKXAM-II: KI>IN<;T<>N, M.l',.. < ..M.. K.li'.S.K., Director of the Colonial Bacteriological Insti- tute, Cape Colony. < 'uniiiiunicated by Sir DAVID (In.i., K.I; S. Received August '20, 1 900. This disease, so far as is known, is peculiar to South Africa. While affecting the Transvaal and Rhodesia every year — beginning about the end of October and continuing until the following May, or even later — it only affects the Cape Colony and Natal in an epizootic form in certain years, although sporadic cases occur annually in certain localities. .Iniiimh .tffi'rii'il. — It affects horses, asses, mules, and it has been asserted — although I have never seen such cases — that quaggas have also been killed by it. A disease which occurs to a limited extent among cattle, called by the natives Imapunga, and one which exists as a widespread plague among high-bred sheep and goats in the Eastern Province of Cape Colony, are each closely related in their pathology with this malady. .In «< //'/<> W. — The most dangerous areas are those which are relatively low-lying — independent of the absolute altitude of the district. 1'iriod of Infection. — It has been commonly observed that where animals during a season of sickness are not permitted to graze after sunset and before the sun has dried up the dew from the herbage, they do not so commonly become affected as where such a routine is not carried out. Horses which are stabled during the night are, as a rule, safe, but during last year 60 per cent, of the stabled horses in Eshowe, ZuJuland, died of this sickness. Veterinary Lieutenant Coley, A.V.D., who kindly made the observations for me, stated that these horses were mainly fed on Guinea or Ub«uiba grass mixed with forage or Indian corn. This grass was usually cut in the evenings and made into bundles till next day. Those who took particular care to have the grass thoroughly dried in the sun before using it did not lose their horses, while those neglecting this precaution lost heavily. The disease is only directly contagious ; for while inoculated horses have died in my stables among clean animals, I have never found, •luring observations extending over seven years, a single case of infec- tion from such a source. The annual mortality in Rhodesia and the low-lying parts of the Transvaal amounts to over 90 per cent. Animals which have recovered from the sickness are termed " salted," and are from six to ten times increased in market value. •its Pathology and Methods of Protective Inoculation. 293 Secondary Fever. Animals which are " salted " are liable to subsequent attacks of fever which have no necessary relation to fresh infection. I have observed numerous cases of this description among the " salted " animals under my observation and during periods when the sickness was unknown. Symptoms of the Disease. It occurs under two forms — the Dikkopziekte* and the Dunpaarde- ziekte.f In the former the head and neck swells up enormously, thus affording trustworthy indications of illness during life. In the latter form, as a rule, no symptoms appear until close to the period of death, when the animal becomes subject to very rapid breathing with heaving at the flanks. At the moment of death, in both forms, it is common to find a huge cloud of white foam ejected from the mouth and nose. This foam is produced from a free exudation of blood plasma into the air passages. Owing to the fact that the animals suffering from the Dunpaarde- ziekte show no symptoms until toward the end of the period of illness, it had come to be believed that the whole period of the disease was limited to a few hours' duration. Post-mortem Phenomena. The pericardium is almost invariably filled with a yellow fluid which, while usually clear, is sometimes blood-stained. Solidified gelatinous exudate is frequently found in relation to the beginning of the aorta. The pleural cavity is frequently occupied by yellow fluid, and the interlobular and sub-pleural tissues are also frequently dis- tended by this material. The interlobular tissue is frequently so dis- tended by exudation that the lung tissue proper is dissected up in all directions. The subcutaneous tissue, especially about the great vessels in the neck, is commonly found to be invaded by this exudation, while in the Dikkopziekte the swelling of the head and neck is due to this effusion. The yellow fluid of the pericardium and the pleurae is spontaneously coagulable in the presence of minute traces of blood. These represent the more characteristic pathological conditions obtaining in this disease, among which one characteristic is most noticeable by its absence, e.g., inflammatory phenomena. Pathological * Dikkopziekte, a Dutch word signifying "thick-head sickness," is applied to the form in which the swollen head is the most obvious symptom. f Dunpaardeziekte, " thin horse-sickness," applied to the form in which the head is little or not at all swollen. '294 Mr. A. l-Miii-toM. Xouth African Hur* -xicL-nes» : phenomena are, therefore, for the most part to be ascribed to the marvellous exudation of blood plasma, which, while seen more or less throughout the serous and subcutaneous tissues, is best marked within the thoracic cavity. In my annual reports as Director of the Colonial Bacteriological Institute I have referred to the morbid anatomy in greater detail. Inoculation Experiment*. For the purpose of conveying to healthy animals the infection from those already sick three materials have been made use of, viz. — 1. The yellow fluid from the trachea of infected animals. 2. The yellow fluid from the pericardium of infected animals. 3. The blood of infected animals. The use of the first two fluids has not always been successful in setting up the disease, but fresh virulent blood has invariably proved successful. of Use of the Material* mentioned. ( /'/ ' The yellow fluid from the pericardium of an animal which had //• / establishment of protection, whereas the higher degree of insuscepti- bility of the other animals resisted infection, and in this way evaded the onset of protection. This phenomenon forms the greatest barrier to protective inoculation, and has contributed to the enormous trouble I have experienced in devising a practical method of protective inoculation. The fresh infected blood of cattle, sheep, and goats is still more variable in its results than that obtained from the donkey. Numerous other experiments of the same nature have been made, all of which result in showing — (a.) That donkeys, oxen, goats, and sheep possess a very irregular susceptibility to the disease. (/>.) That the blood of donkeys which do not react may produce no effect when inoculated into the horse. (>'.) That the blood of donkeys which have evinced moderate reaction may produce intense reaction in some horses and practically none in others. (il.) That a mild reaction in the donkey furnishes no definite assurance as regards the reaction which its blood may set up in horses. Owing to the variable quality of the infection possessed by infected donkey's blood in the fresh state, I experimented with blood taken from donkeys and oxen which, after having l)een received, was pre- served in the manner already described. A large number of experiments carried out by this means furnished the following results : — 1. Protection was only obtained where a definite amount of fever had been produced on several occasions, but unless the reaction was severe, the animal did not resist the inoculation of 1 c.c. of preserved virulent blood at a later period. 2. The susceptibility of horses to such a weakened or attenuated virus varies enormously. Of two animals inoculated with the same dose of the same virus injected directly into the jugular vein, one had good reaction, the other very feeble. Neither were found to be protected when subsequently inoculated with virulent blood. One inoculated with the same amount of the same preserved material two months later died from the primary inoculation, thus showing that even the attenuated virus can be satisfactorily preserved for a considerable period of time. In the case of another animal which was inoculated intravenously with this virus no result followed. Fifteen days later the same inocu- its Patholofji/ and Method* of Protective Inoculation. 299 lation was repeated. The temperature began to be elevated on the fourteenth day, and it died of horse-sickness seven days later. The primary inoculation in this case, while being ineffectual to induce the disease, had evidently lowered the susceptibility, so that a fresh stimulus, by the same virus, was sufficient to overcome the resist- ance entirely. Having recognised that the blood of animals which lived beyond the ordinary period at which horses usually die from horse-sickness was lowered in virulence, I determined to attempt to produce this change in vitro. Having, 'therefore, prepared bottles containing citrate solution, and having thoroughly sterilised them, selected animals were bled under the most rigid aseptic management, and the blood received in the bottles which, after being replugged, were incubated at a temperature of 102 F. during ten days. In one such experiment, out of a total of fourteen bottles, thirteen remained perfectly free from extraneous organisms. Such blood after incubation was then preserved and tested. I found, in this manner, that it was possible to produce an atten- uated virus equally suitable for inoculation as that obtained from the donkey or the ox. While, howeA^er, these experiments demonstrated that it was possible to protect horses by repeated inoculations of an attenuated virus, they equally demonstrated the irregularity of action, owing to the varying susceptibility to the disease in its attenuated form which obtains among horses. Several important facts, however, which were elucidated, are deserv- ing of careful consideration, viz. : — Death in cases of horse-sickness cannot directly be ascribed to hyperpyrexia, inasmuch as several horses have recovered after having experienced temperatures of over 107 F. ; while others, which have died, and, in which characteristic lesions have been found, have not had a temperature exceeding 105 F. Protection can be arrived at without the production of very great reaction, provided that a number of inoculations are made into the animal, and that these have been so arranged as to proceed very gradually to the highest degree of virulence. 3. It is exceedingly difficult to determine the exact degree of attenuation in any particular sample of an attenuated virus. I have usually attempted this by the inoculation of the virus into one or, at most, two horses; but if the susceptibility of such animals happens to be of a low grade, then the reaction produced may not obtain in other horses for which it may subsequently be used. In other words, to determine .'>00 Mr. A. Ellington. Smith African Horm--sickn(-- rxactly the strength of an attenuated vims, it would IKJ necessary always to make the test on at least five animals. 4. The indication for future experimentation was thus to call for the discovery of some method by which a virus «>f standard virulence might l>e, at will, reduced to any required degree of attenuation. Experiments were also made to determine whether the blood of an animal suffering from " secondary " fever had any infective property. To this end animals under "secondary" fever, with temperature as high as 106 F., were bled and the blood used to inoculate clean animals, but in no case was any reaction produced thereby. I there- fore am convinced that the blood during " secondary " fever is non- infective. Experiments with Serum and Dejibriiuited Blood of ,-//// rise, but although the total amounts given exceeded 1000 c.c., no definite interference with the course of the disease was noticeable. Under Clause 2, " salted " animals were inoculated with progressively increasing doses of virulent blood. When these animals had been inoculated with doses of virulent blood equal to 1000 c.c. they were allowed to rest for eight to twelve days, after which they were bled. Of this serum, 500 c.c. was inoculated at one dose into a horse, which, during thirty-three subsequent days, manifested no signs of illness due to the inoculation. When this period was completed, it was inoculated with virulent virus and as a result died of characteristic horse-sickness. No evidence was shown that the serum had in any •zY-s Pathology mid Mrfhods of Protective Inoculation. .'501 way interfered with the action of the virus. Where, however, this serum was used to inoculate animals which were already infected, a very curious change in the character of the disease occurred. The animals became affected, usually in thirty-six hours, with hsemoglobinuria, which later passed into hsematuria and ended in- variably fatally, if the disease was virulent. In two cases, however, where the disease had been induced by an attenuated virus, the haematuria came to an end with the subsidence of the fever. In all, this curious condition was produced by serum in nineteen cases. Where animals are bled into citrate solution, the plasma is of a yellow colour, but in cases which eventually became the subjects of haematuria, I noticed, if they were bled about twenty-four hours pre- vious to the onset of this condition, that the plasma was red coloured. It is therefore evident that the condition has its origin in the blood. In several cases animals, which were partially protected, became subject in a slighter degree to this complication, if they were re- inoculated with virulent blood and were unable to resist it. This blackwater may have some relation to the blackwater fever in man. It is generally believed in Rhodesia that blackwater does not occur as a primary disorder, but only supervenes in persons who have previously been the victims of malarial fever. It seemed to me that this serum might in some way be associated with a residual infection. To determine this I inoculated a " salted " horse, which had also had repeated large injections of virulent blood, with 50 c.c. of fresh blood. I bled it eight days later, and with 5 c.c. of its blood inoculated a clean animal, which thereafter had a very slight rise of temperature on the eighth day. An animal similarly treated was finally inoculated with 300 c.c. injected intravenously and 20 c.c. subcutaneously thirty-nine days pre- vious to being bled. When bled, the blood was defibrinated, and 100 c.c. was injected into each one of six animals. No evidence was shown of any infectivity of the blood. I now determined to make use of the serum from animals which, under Clause 3, had been allowed to rest for periods over one month previous to the collection of their serum. This serum is that which is now being used for the purpose of pro- tective inoculation. I have determined with regard to it — 1. It possesses no curative action which in practice would be of any avail to restrain the course of the disease. 2. Since an injection of 100 c.c., into one animal has absolutely no effect in restraining the action of 1 c.c. of ordinary preserved virus inoculated subcutaneously on the other side, it does not possess any immunising power which would be of practical value in withstanding infection. 302 Mr. A. Kdington. South J//-//-"// II 3. Its germicidal activity is extremely weak, as is shown by the following experiments :— (e taken into account, and, in order to establish this conclusion satisfactorily, a con- siderable number of animals would require to have been simul- taneously dealt with. (c.) Equal volumes of serum and preserved blood were mixed and kept at ordinary room temperature for four days. Of this mixture, 2 '5 c.c. was injected subcutaneously into a clean animal. Fever set in after the usual period of incubation, pursued its characteristic course, and the animal died under circumstances and in the usual time which obtains after the use of pure virulent blood. Since 1 c.c. of virulent blood mixed with 100 c.c. of serum pro- duced a sharp febrile reaction in one animal but had practically no effect in some others, and since 1 c.c. of blood and 200 c.c. of serum produced a reaction in another animal, it was clear that under this method also I should have to meet differences of animal suscepti- bility. It was so far fortunate that preserved virulent blood acted equally well as fresh blood, so that a standard virus is easily prepared and maintained, and by mixing the serum of a considerable number of animals I am able to standardise a large volume of serum. I concluded, therefore, to determine the amount of serum which, when mixed with a definite amount of blood, would serve, acting in concert with the natural protective bodies in the system of the average horse, to ensure the production of the modified disease. After fourteen days should have elapsed subsequent to this inoculation, provided a severe reaction was not set up, I intended to re-inoculate with the same dose of virulent blood, but with a much reduced quantity of serum. Again, after fourteen days, the procedure should be repeated, the dose of virulent blood remaining a constant quantity, but the dose of serum l>eing still further reduced. Finally I intended to inoculate with virulent blood by itself. In the first three series of experiments sixteen horses were used. These were inoculated as follows : — its Pathology and Methods of Protective Inoculation. 303 1st Inoculation 1 c.c. virus and 100 c.c. serum (10 animals). 1 c.c. ,, 90 c.c. „ (4 animals). 0'5 c.c. ,, 50 c.c. „ (2 animals). r A slight variation 2nd Inoculation 0'5 c.c. virus in 30 c.c. ,, I of the quantities 3rd Inoculation 0'5 c.c. „ 15 c.c. „ | was made in *>- several cases. 4th Inoculation 0 • 5 c.c. pure preserved virulent blood. The following shows the results obtained, and where the remark "salted" is made, it is to be understood that the animal has, at later dates, withstood enormous doses of the most virulent blood. Animal. Eeaction. Result. 1st. No reaction at all Salted. 2nd. Reaction to 1st only „ 3rd. No reaction ,, 4th. Slight reaction after all four „ 5th. Eeaction to 4th Died. 6th. Slight reaction to 1st „ 7th. No reaction Salted. 8th. Reaction to 4th Died. 9th. Slight reaction to 3rd; after 5th... „ 10th. Slight reaction to 4th Salted. llth. Slight reaction to 4th ,, 12th. Reaction to 5th Died. 13th. No reaction Salted. 14th. Reaction to 5th „ 15th. Reaction to 4th „ 16th. Reaction to 4th ,, In the next experiment seven animals were used, which were inocu- lated as follows : — Inoculations. 1st. 1 c.c. virus and 100 c.c. serum. 2nd. 0-5 c.c. „ 25 c.c. „ 3rd. 0-5 c.c. „ 10 c.c. „ 4th. 0'5 c.c. „ 1'5 c.c. „ 2 had 0'5 c.c. pure virus. The results were as follows : — Animal. Reaction. Result. 1st. Slight reaction to 5th Salted. 2nd. ,, „ 3rd and 5th. „ 3rd. No „ 4th After a large dose of pure virus, died. ."i<>4 Mr. A. Kdiiiirti'ii. -s // Animal. 'ion. Result. 4th. Slight reaction to 2nd Salted. ~)th. No „ 4th After a large dose of pure virus, died. T.th. Slight .. -2nd Salted. 7th. „ .. 3rd —Where " 5th " is mentioned, it refers to a dose of pure virulent blood. Total animals inoculated 23 ,, died 9 „ „ salted 14 The tests thus applied have been of the most severe character, and despite the fact that these are only of the value of preliminary experi- ments, the results are extremely satisfactory. Obviously animals have been sacrificed which, under altered methods, might have been saved, for the outcome of these inoculations goes to show that, unless some reaction has been produced during the earlier reactions, there is no certainty that an animal is protected. Never- theless it is equally proved that some have become highly protected without having shown any reaction at all. The indication, therefore, has been to increase the dose of the virus used in the primary inoculations, even at some risk to the more susceptible animals. In a subsequent series of animals this has been carried out in the following manner : — 1st. Inoculation 2 c.c. virus and 50 c.c. serum. 2nd. „ 2 c.c. „ 20 c.c. „ Twelve animals have been simultaneously inoculated in this manner. The reactions produced have been as follows : — Animal. 1st Inoculation. 2nd Inoculation. 1 None. None. 2 Slight. Slight. » » 4 „ Severe. 5 Slight. None. 6 „ ,, 7 Severe. Slight. 8 Slight, 9 None. „ 10- Slight. 11 „ None. 12 Severe. its Pailiolorjy and Methods of Protective Inoculation. 305 Since one animal, after the first inoculation, had a severe reaction, it is evident that the limit of strength, consistent with safety, had been reached. The reactions, in the two cases, after the second, were extremely severe, and indicate that the limit of strength of virus for that inoculation had been slightly exceeded, if a widespread scheme of operation had been intended to be carried out among animals in the open. These results would seem to indicate that fortified serum, e.g., that obtained from animals which, after " salting," have been reinoculated with large doses of virus, exerts a peculiar and definite action on the virus. While, however, 100 c.c. of serum suffices to prevent 1 c.c. of virulent blood, when mixed with it, producing any great elevation of tem- perature, I have referred to a case in which a severe reaction was produced. Since, in another case, 200 c.c. of the same serum, with an equal amount of virulent blood, was followed by a reaction and a definite amount of protection, it is evident .that the difference in susceptibility between the latter animal and those which react slightly after 100 c.c. of serum and 1 c.c. of virulent blood is equal to 100 c.c. of fortified serum. Moreover, as already shown, when the virus is attenuated by its passage through less susceptible animals, such as the donkey or cow, its effect, when used in the same dose, either by sub- cutaneous or intravenous injection, varies very greatly in different animals ; in some producing no evident reaction, in others setting up some fever ; while, again in others, its use was f ollowed by the onset of the virulent disease resulting in death. If, therefore, the admixture of serum with virulent blood is followed, on inoculation, merely by a modified form of the disease, it must be concluded that the serum, of itself, cannot be credited with this result, but that a peculiar quality, existing in the animal body, and varying in amount from animal to animal, must play an important part. Whether this principle is a simple body, or is a combination of several, cannot at this moment be determined, but for convenience' sake I would suggest that the name " Antagones " should be applied to it. The term need not be taken to imply either an antitoxic or germicidal body, but merely to denote the " defensive " properties which are already existent to a greater or less degree in all animals, or are produced or increased under special stimulation. Since thoroughly " salted " animals and donkeys can be reinoculated and infection proved to exist in their blood for at least ten days sub- sequent, I am inclined to look upon the protection existing in " salted " animals as of the nature of a " tolerance," and to believe that true immunity, in horses, against this disease is never acquired. VOL. LXVII. I)i I >. II. Scott -V«/ // of a "Note on the Occurrence <>! a Seed-like Fructification in certain I'ahrozoic Lycopods." By D. H. SCOTT, MA., PhD., F.RS., Honorary Keeper of the Jodrell Laboratory, Royal Gardens, Kew. Received August 21, 1000. It has generally lu-rn assumed by palaeobotanists that the fossil seeds described by Williamson* under the name of Cardiocarpon, even if not necessarily co-generic with the Cardiocarputi of Brongniart, at least belonged to the same group of Gymnospermous plants. J Brongniart's specimens, often preserved with marvellous perfection, have proved to be the seeds of members of the extinct Order Cordaitew, or of allied plants. The same conclusion applies to certain of the British forms, notably the Cardiocarpon anomalum of Carruthers,§ which was certainly Cordaitean, and probably to some of Williamson's examples. The specimens to be shortly described in the present note show, however, that seed-like bodies, identical with those figured by Williamson under the name of Cardiocarpon anornalum,\\ were borne on Lepidodendroid cones, otherwise indistinguishable from Lepidostrobus. They thus prove that under the genus Cardiocarpon, and even under the " species " C. anomalum, totally different objects have been con- founded, namely, the seeds of Cordaiteae or Cycads on the one hand, and the integumented megasporangia of certain Palaeozoic Lycopods on the other. The latter organs present close analogies with true seeds, but are wholly distinct in detailed structure from the Gymnospermous seeds above mentioned. The discovery of the specimens of the new cone is due to Messrs. J. Lomax and G. Wild, who recognised it as a Cardiocarp on-bearing strobilus, resembling a Lepidottmbu&$ The original specimens, which are calcified and generally well pre- served, were derived from the Ganister beds of the Lower Coal- * " Organisation of the Fossil Plants of the Coal-measures," Part VIII, ' Phil. Trans.,' vol. 167, Part I, 1877, p. 254. t Founded in Brongniart's ' Prodrome d'une Histoire des Vegetaux Fossiles,' 1828. The forms Cardiocarpon and Cf,- un fhc O'ri'fi-citci of n The structure of this strobilus is sufficiently well preserved to show that the anatomy of the axis agrees with that of the less mature IIMIS, and, as the tissues are more completely formed, exhibits the Lepidostroboid characters even more clearly. M . Wild's specimen, then, demonstrates that the r,, /,//.„, ,,y,,,// iiii'tiiKiliiiii of Williamson was borne on a cone with all the characters of a Is-jtiil>i.-Ii>i*t and that it represents the matured condition of the -poran^ium and sporophyll. The detailed comparison of specimens in the young and the mature condition has shown the nature of the change, which converts the megasporangium, together with its sporophyll, into a seed-like organ. The nucellus of the latter retains almost unaltered the structure of the megasporangial wall, with its columnar layer. In the sporangial cavity the single large megaspore, accompanied by its abortive sister- cells, is present as before. A thick integument has, however, grown up from the sporophyll, completely overarching the megasporangium, except for a narrow crevice left open at the top. When seen in a section tangential to the strobilus as a whole, this crevice is cut across, and presents exactly the appearance of a micropyle; in reality it differs from a micropyle in being a narrow slit, extending almost the whole length of the sporangium, in the radial .direction, whereas the micropyle of an ordinanr seed is a more or less tubular passage. The integument springs from the upper surface of the sporophyll- pedicel ; it does not consist of the incurved margin of the pedicel, for, in the more distal region, the margin of the latter projects consider- ably beyond the insertion of the integument. From the frequency of detached specimens in the Cardiocarpan con- dition, it appears that in nature the sporophyll, bearing the integu- mented megasporangium, was shed as a whole, though parts of the sporophyll-lamina no doubt perished, only so much being persistent as was necessary to form a complete envelope to the " seed." In a strobilus associated with the seed-like specimens, and bearing microsporangia, it was found that the latter, like the megasporangia of the female cone, are provided with integuments. This specimen was figured by Mr. Maslen as a variety of Lepitlostrobu* OUfonwM,* though possibly deserving specific rank, a determination with which I agreed at the time. There is every reason, however, to suppose that this strobilus was a male fructification of the same species, the female of which bears the integumented seed-like megasporangia above described. The microsporangial integument is more widely open than that of the megasporangium. The Burntisland specimens, which from their horizon are presumably of a distinct species, are at present only known in the isolated Cunli'- condition. They are of interest for two reasons : in one speci- * MnsU-n, ' Structure of LepMost robu*,' p. 371, Plate 37, fig. 21. Seed-like Fructification in certain Palccozoic Lycopods. 309 men the ligule is clearly shown, enclosed by the integument, the only example of this organ, so far observed, in the mature, seed-like stage of the fructification. Another of the Burntisland specimens is the only one as yet observed in which the prothallus is present.* It fills a great part of the functional megaspore, which is almost co-extensive with the spo- rangial cavity, and consists of a large-celled tissue, resembling the pro- thallus of Isoetes or Selaginella, The peripheral prothallial cells are smaller than the rest, but no archegonia could be detected. The bodies described in this note resemble true seeds in the posses- sion of a testa or integument, and in the fact that one megaspore or embryo-sac alone came to perfection ; the seed-like organ was likewise shed entire, and appears to have been indehiscent. In many points of detail, however, the reproductive bodies in question differ from the seeds of any known Gymnosperms ; they afford no proof of the origin of the latter Class from the Lycopods. The newly discovered fructifi- cation nevertheless shows that certain Palaeozoic Lycopods, with strobili at first indistinguishable from Lepidostrobus, crossed the boundary line which we are accustomed to draw between Sporophyta and Sper- mophyta. As these fossils appear worthy of generic rank, I propose to found the genus Lepidocarpon for their reception; it may be briefly characterized as follows : — Lepidocarpon, gen. nav. — Strobilus, with the characters of Lepidostro- bus, but microsporangia and megasporangia each surrounded by an integument, growing up from the upper surface of the sporophyll. Megasporangium completely enclosed in the integument, except for a slit-like micropyle along the top. A single functional megaspore developed in each megasporangium. Sporophyll, together with the integumented megasporangium, detached entire from the strobilus, the whole forming a closed, seed-like, reproductive body. It is proposed to name the Coal-measure form Lepidocarpon Lanta.i-i, and that from Burntisland L. Wildianum. Both were included by Williamson under his Cardiocarpon anomalum, which, however, is quite different from the seed so named by Carruthers. A full, illustrated account of these fossils is in preparation, and will shortly be submitted to the Royal Society. * I have since examined a section, cut by Mr. Lomax from one of the Coal- measure specimens, in which the prollmllus is even better preserved. — Note, added October 9, 1000. Seed-like Fructification in certain Palaeozoic Lycopods. 309 men the ligule is clearly shown, enclosed by the integument, the only example of this organ, so far observed, in the mature, seed-like stage of the fructification. Another of the Burntisland specimens is the only one as yet observed in which the prothallus is present.* It fills a great part of the functional megaspore, which is almost co-extensive with the spo- rangial cavity, and consists of a large-celled tissue, resembling the pro- thallus of Isoetes or Selaginella. The peripheral prothallial cells are smaller than the rest, but no archegonia could be detected. The bodies described in this note resemble true seeds in the posses- sion of a testa or integument, and in the fact that one megaspore or embryo-sac alone came to perfection ; the seed-like organ was likewise shed entire, and appears to have been indehiscent. In many points of detail, however, the reproductive bodies in question differ from the seeds of any known Gymnosperms ; they afford no proof of the origin of the latter Class from the Lycopods. The newly discovered fructifi- cation nevertheless shows that certain Palaeozoic Lycopods, with strobili at first indistinguishable from Lepidostrobus, crossed the boundary line which we are accustomed to draw between Sporophyta and Sper- mophyta. As these fossils appear worthy of generic rank, I propose to found the genus Lepidocarpon for their reception; it may be briefly characterized as follows : — Lepidocarpon, gen. nov. — Strobilus, with the characters of Lepidostro- bus, but microsporangia and megasporangia each surrounded by an integument, growing up from the upper surface of the sporophyll. Megasporangium completely enclosed in the integument, except for a slit-like micropyle along the top. A single functional megaspore developed in each megasporangium. Sporophyll, together with the integumented megasporangium, detached entire from the strobilus, the whole forming a closed, seed-like, reproductive body. It is proposed to name the Coal-measure form Lepidocarpon Lomaxi, and that from Burntisland L. Wildianum. Both were included by Williamson under his Cardiocarpon anomalum, which, however, is quite different from the seed so named by Carruthers. A full, illustrated account of these fossils is in preparation, and will shortly be submitted to the Royal Society. * I have since examined a section, cut by Mr. Lomax from one of the Coal- measure specimens, in whicli the prothallus is eren better preserved. — Note, added October 9, 1900. VOL. LXVII. 310 Mr. .1. S. " The Demarcation Current of Mammalian Nerve. (Preliminary Communication.) I. The Demarcation Current of Mammalian Nerve." By J. S. MACDONALD, B.A., L.R.C.P.E., University College, Liverpool, Research Scholar of the British Medical Association. Communicated by Professor SHERRINHTOX, F.R.S. Received July 28, 1900. 1. A necessary preliminary to the study of the distribution of the demarcation current and source is an examination of the character of the resistance of the particular nerves in which phenomena are observed. The resistance per cm. of the nerves examined — vagus, phrenic, and sciatic nerves of dog and cat, &c., varies with the nerve and with the animal Vagus Horse, 2000 ohms per cm. Dog, 12,500 „ Cat, 31,000 .. Sciatic Dog, 3500 .. „ „ Cat, 4500 .. the variations depending upon the character of the nerve and presum- ably of the individual fibres, upon its sectional area, and probably upon intrinsic differences between the average salt content of the tissues of different animals. But taking any individual nerve the estimated value of the resistance per cm. varies with the length of the piece, from the determination of the resistance of which the estimation is made. Thus keeping one electrode at a fixed point of a nerve, the cross- section A, moving a second electrode from point to point, B, C, D, &c., and measuring the resistance included between A and these several other points, a series of determinations of the resistance of various lengths of the same nerve are made. From a knowledge of the length of the piece and from this determination, an estimation of the resistance per cm. of the nerve is obtained which is greater with each diminution of length. Experiment. — Vagus Nerve of Dog. Piece. Length. .Resistance. Resistance per cm. AB 15 mm. 29,500 ohms. 19,700 ohms. AC 30 .. 58,500 ,. 19,:XX) „ AD 50 „ 85,000 ., 17,000 „ AE 65 „ 108,000 „ 16,600 „ AF 83 „ 133,500 „ 16,080 „ The Demarcation Current of Mammalian Nerve. 311 The calculated resistance per cm. is also considerably greater when the resistance is measured from one point on the longitudinal surface to another than in the case in which the resistance of a similar length is taken from the cross section to a point on the longitudinal surface, and greater still than the resistance of a similar length bounded by two cross sections. When therefore a knowledge of the resistance is required as a basis for calculations, a direct determination of the resistance is only of value when the resistance is required to a current having the same path as that used for the measurement of resistance. In any other case, as when the fraction of the longitudinal re- sistance corresponding to a fraction of the length of the nerve is required, it is better to use a calculated value than the value of the directly determined resistance. This calculated value is best obtained from the resistance per cm. of the longest available stretch of the nerve. ***** 2. If a point on the longitudinal surface is connected to the cross section through a pair of non-polarisable electrodes and an outer wire path, and a current is found to traverse the circuit so formed, then the current can not only be found in the outer wire path by means of a galvanometer placed in it ; but can also be followed in a return direc- tion in the nerve, travelling in the opposite direction from the cross- .section to the longitudinal surface, by means of the new differences of potential, which the formation of the outer circuit immediately estab- lishes in every intervening point of the nerve. This is true when any arbitrarily selected point on the longitudinal surface is connected to the cross section. In each case the return current through the nerve is found established as a new phenomenon, due to the closure of the outer path, and is exactly the current due to the action of a source of E.M.F. of the value determined as the potential difference between the point on the longitudinal surface and the cross section in a circuit of the resistance found. Experiment. — Vagus Nerve of Cat. The nerve was excised and laid upon four non-polarisable electrodes, A, B, C, D. The cross section was at A, the nerve stretched a small distance beyond electrode D, and was suspended to the wall of the moist chamber by a silk thread. The potential difference (so called) between points A and D was determined as 0-00712 volt. The potential difference between the two intermediate points B and C was zero, the points being equipotential. Points A and D were now permanently connected through the 2 A 2 Mr. J. S. MacdoimM. resistance of a pair of non-polarisable electrodes and a wire joining them, forming in all a circuit of resistance 135,000 ohms. Nerve AD ............... = 128,000 ohms. Electrodes ............... = 7000 ohms, The length of nerve AD = 4-3 cm. „ „ piece BC =1*8 cm. The " calculated value " of resistance B, C = 54,000 ohms. The value directly determined = 63,000 ohms. After closure of circuit A, D, the intermediate point B, nearest to the cross section, became " + " to the more distant intermediate point C, which was " - ". The value of this difference was 0-0028 volt. If in this experiment it is assumed that, (1) the only source of E.M.F. is that found and measured as the potential difference between A ami D, 0-00712 volt : and (2) that the path of the current is the simple one of nerve electrodes and wire, i.e., the path through the nerve is simple, and not divided into two sets of resistances carrying a current in opposite directions (circuit completed in nerve itself) : then a difference of potential should be found between points B and C of this path T ' It E' = _ 0-00284 volt, and this is practically the value actually found. In many similar experiments which have been performed, this agreement of value found and value calculated has been found to hold good within a small limit of error, entirely owing to an alteration of the E.M.F. due to the cross section and to the lapse of time taken to perform the experiment. When there is a pre-existing difference of potential between the points B and C, this difference subtracts from the newly-acquired value due to the closure of the circuit A, D, and the value actually found is the algebraical sum of the pre-existing and the newly-acquired difference. Since a pre-existing difference between B and C is the source of the " longitudinal current," the last point in the above statement is con- sidered of importance, as tending to show that the source of the demarcation current and that of the longitudinal current can be so separable as to oppose one another in a conveniently arranged circuit. T/ie Demarcation Current of Mammalian Nerve. .">!.'> 3. Similar experiments show that the closure of a circuit for the observation of a longitudinal current also gives rise to similar changes of potential in the intervening stretch of nerve. Similar experiments show that the closure of a circuit for the obser- vation of an elect rotonic current affects the potential of every interven- ing point in the same way. So that if a path joining two points on a nerve is found to be carrying a current, whether it be demarcation, longitudinal, or electrotonic, this current can be traced in the longitudinal axis of the nerve, making use of its gross longitudinal resistance, and not interfered with by currents from any other of the possible sources of E.M.F., discovered by the determination of pre-existing differences of potential of intervening points. None of these sources are brought into action so as to affect a current in the longitudinal axis of the nerve until an additional outer circuit of non-polarisable electrodes and wire is formed for them; they then at once are brought into action and add to or subtract from the original. ***** 4. Long stretches of mammalian nerve — vagus, sciatic, and phrenic — of about 10 cm. long have been taken, and laid upon an ebonite scale. The E.M.F.s available between either cross section, and the points distant 1, 2, 3, 4, &c. cm. from one end, have been systematically measured, and from the measurements curves obtained with the nerve as abscissa and the ordinates the E.M.F. between the immediately •underlying point of longitudinal surface of the nerve and the cross section. Whatever be the difference between the E.M.F.s due to the two cross sections (in some cases a difference of height of maxima o 0-006 volt, the lower one, e.g., being 0*002 volt, the higher 0*008 volt), and whatever be the peculiarities of the curves, they are, notwith- standing the difference of level, parallel for the greater portion of their extent. The curve due to the available force between one cross section and the longitudinal surface repeats all the maxima and minima of the curve due to the other cross section at a different level, the relations between the maxima and minima being preserved unaltered. Such a condition of affairs is most readily explained by the assump- tion that the determination of E.M.F. between a point on the longi- tudinal surface and the cross section is always the determination of the algebraical sum of two opposing forces, one acting radially at the point on the longitudinal surface, and one at the cross section acting in the longitudinal axis of the nerve ; and that the radial force remains the same at a point of the longitudinal surface, whereas there may be and usually is a difference between the two longitudinal forces, one at each cross section. ***** 314 Mr. J. S. Macdonald. 5. Selecting one cross section and systematically measuring the E.M.F. available between this and points on the longitudinal surface, drawing a complete curve (force diagram) at regular intervals of 30' between each curve, the levels of the curves fall with diminishing rapidity, and a gradual change occurs in the form of each curve. After the nerve has been removed some hours from the animal, the maximum E.M.F., the highest point of the curve, may have sunk to one-tenth of its original value. If now the nerve is taken and placed for a short period (five minutes) in tap water, a maximum E.M.F. ia obtained considerably greater than the maximum obtained from any point of the same nerve when freshly removed from the recently-killed animal. If the value of the demarcation source is taken as an index of the condition of the nerve, then a rejuvenation has taken place with the immersion in tap water. This increased value remains for some time, the curve being of not much different form, and the rate of fall of level being similar to the rate of fall from the original maximum. If the nerve be left in tap water for twenty-four hours, a demarcation current, a difference of potential between each point on the longitudinal surface and the cross section is observed, giving a curve very similar to that obtained after the first immersion in tap water twenty-four hours previously. If one waits after the death of an animal until rigor mortis is com- pletely established, and a nerve be then removed, only a small trace of demarcation current is obtainable from it, and the curve of E.M.F. due to a cross section is at an extremely low level. If now this nerve is left for a short period (five minutes) in tap water, a maximum E.M.F. is obtainable from it higher than that obtained from the fresh nerve of the recently-killed animal, and as high as that obtained after the immersion of the fresh nerve in tap water, and there is no marked difference in the form of curve (force diagram). If a nerve is removed at once after the death of an animal, and the E.M.F. between a "maximal" point and the cross section is taken (1) immediately (2) after a short immersion in 0'9 per cent, saline, then it is seen that the 0*9 per cent, (normal) saline has diminished the E.M.F. If the nerve be now immersed in 0'45 per cent, saline the original value is returned, and is increased by a further immersion in 0'3 per cent., 0-2 per cent., O'l per cent, saline solution, each further dilution increasing the E.M.F. A return of the nerve through the series diminishes in each case the E.M.F., to be renewed by a sub- sequent return to the weaker solution. The maximum effect is obtained by the action of tap water, and a very considerable reduction of this maximum is obtained by a sub- sequent immersion in O'l per cent, saline, to be followed by a return to the maximum with a return to the tap water. There is no sign of any critical point marking the separation of two possible The Demarcation Current of Mammalian Nerve. 315 phenomena, one a function of the condition of life of the nerve and the other a physical phenomenon dominated by the salt content of the nerve, and capable of continuation long after its death. ***** 6. If a number of threads are twisted together to form a rope and the rope laid upon two non-polarisable electrodes of the usual type, no current is found between the electrodes if the thread rope is previously uniformly wetted with a saline solution or with tap water. If on such an uniformly wetted rope a drop of saline solution of a different concentration is placed at a point closer to one electrode than another a current is found in the circuit, and a source of E.M.F. quite com- parable in value to the maximal value of the demarcation source of a nerve. A drop of the same solution placed upon a corresponding point of the rope nearer to the other electrode may reduce, bring to zero, or reverse this difference of potential. This phenomenon is presumably due to the upsetting of the balance between the osmotic processes taking place in the two non-polarisable and " similar " electrodes. ***** 7. The close association of the value of the demarcation current with the salt content of the nerve suggests a similarity between the experimental phenomena observed in the thread and in the nerve, and the causation of the demarcation current of nerve as due to a balance between two unequal osmotic processes, one at the cross section and one at the longitudinal surface. ***** The expenses of this research have been partially defrayed by a grant from the British Association. " The Demarcation Current of Mammalian Nerve. (Preliminary Communication.) II. The Source of the Demarcation Cur- rent considered as a Concentration Cell." By J. S. MAC- DONALD, B.A., L.R.C.P.E., University College, Liverpool, Research Scholar of the British Medical Association. Com- municated by Professor SHERRIXGTON, F.R.S. Received September 25, 1900. The changes produced by the action of tap water upon the nerve have, in the interval, been more closely studied. Excluding alterations of E.M.F., they are as follows : — 316 Mr. J. S. Mu.-dniuM. •(n) weight. An increase of < <*> (c) rigidity. •(700 j 0-018 0-023 0-020 0-016 (>•' 13 0 Oil 1 0 207 0-213 0-221 O-l-'l 0 233 0 236 | 4-9 5'1 5-1 t>-l 51 5-1 An error is evidently introduced into the measurement of resistance, if we wish to consider the resistance of a cylinder of the nerve of con- stant length and cross section, by the increase in volume attending the increase of weight. Failing an actual determination of the area of the cross section, or one of the volume from which it might be directly obtained by dividing by the length, the value of the weight found in grammes has been treated as volume in cubic centimetres, and used in this way. The error introduced by a neglect of the specific gravity is not, in the case considered, appreciable. The " specific resistances " for unit length and cross section found in this way from the figures given are — A. 165 (1.) 184 (2.) (3.) (4.) (5.) 180 182 | 184 | 188 ohms. Experiment B. — Sciatic Nerve of Oat, The experiment was in every way similar to the last, with the single exception that the solution used was an NaCl solution of 0'75 gramme per cent. A. (1) (2.) (3.) (4-) (5.) Ohms 20700 206f>0 19,4 0 IS 900 19,200 Volts.... 0 018 0'018 0-018 0-017 0015 0-014 Grammes . . . 0 -210 0-218 0'222 0'22s 0-229 46 4-7 4'7 47 47 •17 " Specific resistances" 205 L'. :! •2 :< 200 196 199 T/ie Demarcation Current of Mammalian Nerve. :\ 1 0 Experiment C. — Sciatic Nerve of Cat. Similar experiment. Solution used 0-9 gramme per cent. NaCl. A. (1.) (2.) (3.) (4.) (5.) Oliins . . 20300 19 700 16 700 16900 17 100 16 I'lO Volts 0 :014 0-014 O'Oll 0-009 0 -009 0 001 Grammes 0 236 0-247 0-257 0-262 0'260 0 960 Centimetres 5 5 5 5 5 5 " Specific resistances " 191 184 171 177 177 167 Taking the three experiments and examining the variations in the specific resistance in each case, it is noticeably least affected in the 0-75 per cent, solution^ which therefore, from this point of view, most nearly approaches the " isotonic solution." Reducing the data from the three experiments to an assumed average value of 200 ohms for the inital resistance, we have Experiment with 0-6 per cent. ,, 0-75 „ ,, 0-9 Initial resistance. After successive immersions in the solutions. (1.) (2.) (3.) (4.) (5.) 200 200 200 222 198 192 218 198 178 220 194 184 222 190 184 228 194 174 or, taking averages from the five determinations made after immersion in the solution, Initial. Final specific resistance. Total increase. 0 "6 per cent 200 225 + 25 ohms 0-75 .. 200 195 -5 „ 0-9 200 185 -15 „ The alteration in weight in the three experiments is apparently anomalous, as there is an increase in each case, and judged from this criterion alone, none of the solutions are isotonic. Considerable care was taken to dry off the superficial moisture, and as far as possible the precautions were the same in each case. There is a difference in Mr. .1. S. the character of the increase which is confirmed by the results of other experiments, the increase in the hypotonic solution is progressive, in the other two there is an increase to, and a maintenance of, a steady maximal weight. It seems evident that in all three cases there is an imbibition of saline solution at first, apart from any question of the transference of water through a membrane. The appearances of rigidity were noticeable in the nerve immersed in the 0'6 per cent, solution, though the only index of their occurrence found in the figures given is provided by the increase in length. The occurrence of the appearances is, however, unmistakable in an actual experiment ; the difference between a nerve which has been immersed in 0-9 per cent, and one which has been in 0'6 per cent, is quite a marked one. ***** Markedly graduated as are the changes of weight, length, resis- tance, and rigidity following the immersion of nerves in solutions of graduated concentration, the changes in the E.M.F. available between cross-section and longitudinal surface are no less so. In so far is this true that it is possible, knowing the initial value of the E.M.F., to predict the value which will be found after immersion of the nerve for a given time in a solution of known concentration at a constant temperature. The accompanying data are taken from the records of 8 entirely separate experiments. In each, a sciatic nerve (cat), removed immediately after the death of the animal, was. cut to a definite length of 5 cm., and placed for 25 minutes in 500 c.c. of an XaCl solution, at a temperature of 17° C. The E.M.F. found after immersion is given expressed in terms of the initial E.M.F. used as unity, is called the KM.F. " recovered," and as will be seen is sometimes greater than the initial value. Solution used. Proportion of E.M.F. recovered. (1) Tap water 1-5 (2) 0-6 gramme XaCl percent 1*059 (3) 0-75 „ „ „ 0-921 (4) 0-9 „ „ „ 0-786 (5) 1 -8 grammes „ „ 0'388 (6) 3-0 „ „ „ 0-237 (7) 6-0 „ „ „ 0-107 (8) 9-0 „ „ „ 0-062 If in each case the value of the concentration of the NaCl solution is multiplied by the E.M.F. recovered (expressed in terms of the initial E.M.F.), we get a value which is almost constant for the whole series. The Demarcation Current of Mammalian Nerve. :\~1\ (2) 0-6 x 1-059 = 0-6354 (3) 0-75x0-921 = 0-69075-| (4) 0-9 X 0-786 = 0-7074 I (5) 1-8 X 0-388 = 0-6984 f (6) 3-0 X 0-237 = 0-7110 J (7) 6-0 X 0-107 = 0-6420 (8) 9-0 X 0-062 = 0-5580 Which relationship can be interpreted to mean that for a consider- able range of concentrations the E.M.F. "recovered" varies almost exactly inversely as the concentration, and outside this range the deviation from the Jaw is not great. The value of the constant, which is practically 0'70, is evidently the concentration of a saline solution in which the E.M.F. should be unaltered by the immersion. The preservation of a constant temperature throughout the series of experiments is of importance, the variation with temperature being considerable, and complicated. Data are given from a few experi- ments made to determine the interest of this point. The data are as before, each from a separate experiment upon a measured length (5 cms.) of the sciatic nerve of a cat ; the only difference being in fact that the temperatures of the solutions were varied instead of their concentrations. Temperature of solution. 9°C. 17 28 38 * 9°C. 17 27 34 2 Concentration of solution. •75 per cent. NaCl. 3-0 per cent, NaCl. E.M.F. " recovered " in terms of the original as unity. 1-01 0-92 0-62 0-62 * 0-25 0-24 0-11 0-04 0-15 It may seem an obvious and foregone conclusion that the isotonic solution, in which the nerve may lie with the minimal disturbance due to the transference of water and salts, should closely coincide with the solution in which the E.M.F. is retained constant, and also with the probable isotonicity of the solution which in the living body bathed the outer surface of the nerve. It may indeed be maintained that from the point of view of the established hypothesis the local short circuiting of the demarcation source would be affected by solutions varying in concentration from •"'L'l' Mr. J. S. Macdonald. the " normal saline " in just such a way as to cause exactly the variations described in the apparent value of the E.M.F., and that the variations would be connected by the simple law found. If so, an examination of the alterations taking place in the same " normal solution " when its temperature is varied presents an anomaly for explanation. The solution maintained at anything approaching the temperature of the body, in which the E..M.F. would remain constant, is not 0'70 per cent., but 0'45 per cent. NaCl solution. A more striking anomaly still is obtained when an appeal is made to solutions of electrolytes other than NaCl ; an extreme instance is given by the consideration of solutions of NaOH. The following data are taken from four separate experiments in which 5 cm. pieces of sciatic nerves (cat), removed immediately after death, were placed in each case in 500 c.c. of an NaOH solution at a temperature of 17° C., and left in it for 25 minutes. Solution. E.M.F. " recovered." (1) 0-025 gramme per cent. NaOH. 1-620 (2) 0-050 „ „ „ 0-922 (3) 0-050 ,. „ „ 0-800 (4) 0-100 „ ?, „ 0-422 Proceeding as before, and multiplying the concentration by the E.M.F., we have (1) 0-025x1-620 = 0-0405 (2) 0-050 x 0-922 = 0 0461 (3) 0-050 x 0-800 = 0-0400 (4) 0-100 x 0-422 = 0-0422. The concentration law is the same as for NaCl, but the " constant ' solution is 0-04 instead of 0'7. Dividing these figures by the mole- cular weights of NaOH and NaCl respectively, the proportion existing l>etween them is 1 to 12 ; and this, even allowing for the greater con- ductivity of NaOH solutions, is evidently a relation of a more complex kind than that found when passing from one concentration to another «.f a solution of the same electrolyte. ***** In certain types of experiments, such as those in which the effect of tap water was studied (vide supra) upon the nerves of animals in a state <>f li'jnr nif/rti*, the capacity of the established hypothesis to explain the facts is strained to an absurd degree. The following are brief descrip- tions of typical experiments concerning which the same statement may safely be made : — The Demarcation Current of Mammalian Nerve. 323 I. Experiment. Sciatic Nerve of Cat. Piece 5 cm. long. Difference of potential taken between the cross section and a point 1 cm. distant. Volt. At once OO15 After 25' in 9 per cent. NaCl solution at 17° C O'OOl Another 25' ., „ „ O'OOl Another 25' „ „ „ O'OOl Another 10' „ ., „ O'OOl After 15' in tap water at 17° C O011 After another 30' in tap water at 1 7° C 0-022 One hour and 40 minutes in the very " abnormal " saline solution of 9 per cent. NaCl and a subsequent 45 minutes in the very " abnormal " saline solution of tap water, and yet the vigour of the changes is unimpaired — they are displayed to the same abnormal degree by the effects of tap water upon the local circuit, and the contrasted states of activity and hyperactivity are shown in exactly the same situations. II. Experiments upon Degenerated Nerve. In these experiments the preliminary operations for section of the nerves were performed by Professor Sherrington, F.R.S. They are described in the briefest possible manner, but as the nerves were made the subject of systematic and detailed study the full description is withheld, as in case of the other experiments, for an opportunity for more detailed publication. Experiment a. Vagus of Dog. Preliminary Operation. — 1 cm. of nerve excised at upper limit, and 1 cm. of nerve excised at the lower limit of the nerve in the neck. Examination nine days afterwards. Degenerated Nerve. E.M.F. = 0-000 volt. Several cross-sections were tried. After an immersion of 25' in tap water — E.M.F. = 0-017 volt. Intact vagus of other side. E.M.F. = 0-006 volt. After an immersion of 25' in tap water — E.M.F. = 0-016 volt. •-•I-' I -Mr. -I. S. Espcrimcnt /3. Sciatic Nerve oj J'i liiniii'ifi/ Operation. — 1 cm. of nerve excised. Esnininntion twelve days aft Degenerated Sciatic. K.M.F. = 0-003 volt. After immersion in tap water. K.M.F. = 0-025 volt. Intact Sciatic. K.M.F. = 0-017 volt. After immersion in tap water. E.M.F. = 0-026 volt. It seems highly probable to the author, biased by the simplicity of the "concentration law," that the extreme case studied, namely, the nerve after immersion in tap water, is but an extreme variation of a pre-existing condition — in fact, that the internal structures of the nerve form what is to all intents and purposes a stronger aqueous solu- tion of electrolytes than is found in its superficial parts, just such an arrangement of solutions as the character of the resistance and internal pohvisation of nerves has always made probable. If this is true, all the arguments which can be adduced to explain the E.M.F. obtained from the extreme case can be transferred, when modified, to the normal condition. In this extreme case there is no need to invoke a difference in the distribution of the dissociation phenomena of life to explain the exist- ence of a source of E.M.F. The source is granted as soon as it is determined that solutions of different concentration, such as are present, are asymmetrically placed in the otherwise symmetrical arrangement of solutions connecting the metallic electrodes.* Failing an absolute knowledge of this asymmetry, there are many reasons which make it highly probable ; the anatomical conditions are obviously asymmetrical. The mathematical considerations determining the value of such a source have been so perfectly elaborated, and consequently simplified, that the data collected from the examination of a supposed instance, even of a complicated case, can be afforded a criticism of great exactness. With a view to such criticism the research is being continued, and for the present the conditions of a possible reversal are sought. Throughout the conduct of this research I have been most liberally assisted with information and advice by Professor C. S. Sherrington, F.R.S., and Professor Oliver Lodge, F.R.S., for which I take this opportunity for expressing my gratitude. I have also to thank Mr. B. Davies, Mr. A. Hay, and Mr. \V. H. Derriman for their frequently sought opinions, and Mr. \V. G. Lloyd for practical assistance in some of the experiments. * Concentration cells of Jfernst, 'Electrochemistry,' Le Blanc. The Demarcation Current of Mammalian Nerve. 325 " The Demarcation Current of Mammalian Nerve. (Preliminary Communication.) III. The Demarcation Source and ' the Concentration Law.' " By J. S. MACDONALD, B.A., L.E.C.P.E., University College, Liverpool, Eesearch Scholar of the British Medical Association. Communicated by Professor SHERKIXG- TON, F.K.S. Eeceived October 18, 1900. Since writing the previous statement the changes of E.M.F. occurring during the early part of the time spent in saline solutions have been studied more in detail, the nerves being removed every five minutes for examination. As a result of information so acquired, the following statement can be definitely made. Solutions of NaOH, HC1, NaCl, KC1 mainly affect the demonstrable value of the demarcation source according to their concentration, and differ intrinsically from one another in their effects upon this source only in minor characteristics. Each of these electrolytes produces an effect which is best regarded as a variation of the effect of water, and varies with the concentration according to a simple law. The concentration in each case determines whether the initial value of the demonstrable E.M.F. shall be increased, maintained, or diminished. A study of the comparative effects of various concentrations of the same electrolyte is of particular interest when the nerve is only immersed for a short period (five minutes), presumably because within this period processes of diffusion interfere least with the concentration of electrolytes in the internal parts of the nerve. The concentration law found to unite the effects of solutions of NaCl, KC1, HC1 is comparatively simple, the case of solutions of NaOH being apparently more complex. If " E " represents the initial value of the E.M.F., "En" the value after an immersion of 5' duration in a solution of concentration " n," " n " the concentration in gramme-molecules per litre, then in the special case of solutions of KC1 En = Elog- approximately. '/& Thus, taking the data from four experiments performed upon 5 cm. pieces of sciatic nerves (cat), determining the available E.M.F. between cross section and longitudinal surface (a) immediately upon removal from the recently killed animal, and (I) after an immersion of 5' in a solution of KCl at 17° C. :— VOL. Lxvn. 2 15 Mr. J. S. Macdonuld. Experiments with Solutions of KC1. Concentration E.M.F. recovered, Grain Gramme-mols. in terms of the per cent . per litn>. initial value E. 7-45 1 E x 0-10 = E log 1-2 3-72 £ ExO-34 = Elog2'2 1-86 I ExO-60 = Elog4 0-93 | ExO-90 = ElogS The law connecting the effects of solutions of Nad is not very different, as is shown by examination of the results of the following experiments made before the relation was discovered. The results are in this case none the less remarkable in so far as the concentrations of the solutions used bear no simple proportion to the normal solution of 1 gramme-molecule per litre (5 '85 per cent.). Experiments with solutions of NaCl. Concentration . E.M.F. recovered, Grammes Gramme-mols. in terms of E, per cent. per litre. the initial value. (1.) 0-45 — E x 1-15 - E log 14-1 13 (2.) 0-6 4? ExO-92 = Elog 8'3 y * i (3.) 0-9 Ex 0-77 = Elog 5-9 6 '5 (4.) 3-0 -1 ExO-34 = Elog 2-2 1 *9 (5.) 6-0 Ex 0-13 = Elog 1-3 0-97 Presuming temporarily the law connecting these results to be "En = Elog-," then the concentration multiplied by the number of n which the logarithm is the proportion between the final and initial value of the E.M.F., should equal unity. Thus (1.) X 13 1 1 •1 = 1O9 (2.) X 8 •3 = 0-85 (3.) X 5 •v = 0-90 (4.) 1 2 •2 = 1-15 (5.) 0^97X 1 •3 = 1-34 The Demarcation Current of Mammalian Nerve. 327 It is obvious that the law as stated approximately represents the truth in the case of NaCl and KC1 solutions. In how far the approxi- mation is different in the two cases must be left to the consideration of further and more exact experiments. That a similar law is true for solutions of HC1 is shown by the fol- lowing experiments. The stock laboratory solution of 0*4 per cent. HC1 was diluted twice, &c., to obtain the required variations of con- centration. Experiments with solutions of HC1. Concentration. Grammes per cent. (1 ) 0'2 -\ Gramme-mols. per litre. 1 in terms of E, the initial value. ... Ex 0-31 = Elog 2-0 (2 ) O'l 18-2 1 E x 0-60 - E log 4 (3 ) 0-0250 36-3 1 .. . Ex 1-21 - Elog 16-2 (4) 00125. 145-2 1 .... Ex 1-53 = Elog 33-9 '" 290-4 ' In this case presuming the statement of the law to be then in Case (1.) k = - x 2-0 = (HI lo'J (2.) t - ^ x 4-0 = 0-11 * - 1455 The law in this case is En = E log — It is obvious that in the three cases considered the concentration law is amply expressed as E = y^.E.log-2, n where k\ and k% are constants. The interest attached to this mode of expressing the law is seen, upon reference to the theory of concentration cells, to be considerable ; and the occurrence of the law cannot be otherwise regarded than as a 2 B 2 328 valuable confirmation of the reality of the assumption made as to the nature of the demarcation source of the nerve. Taking the simplest expression for the value of the E.M.F. of such a cell, in which there are two solutions of different concentrations N and n of the same electrolyte at the same temperature, E = Klog-. n If in such a cell the solution of concentration "N" is retained con- stant, whereas the solution of concentration "n" is given different values, 71 1, n.2, 7i3 ...... &c., and E», = Elog^x— *-. «i log N/n If X is constant, and also if N/n the original relation of the two solutions is treated as a constant, we have En, = E.fc.log*2, «1 and this is the relation found existing between the value of the demar- cation source of the nerve before (E) and after (EBl) the immersion of the nerve in a solution of concentration n\. November 15, 1900. The LORD LISTER, F.R.C.S., D.C.L., President, in the Chair. The Right Hon. Sir Ford North, Professor J. Bretland Farmer, Dr. Patrick Manson, and Professor James Walker were admitted into the Society. A List of the Presents received was laid on the table, and thanks ordered for them. In pursuance of the Statutes, notice of the ensuing Anniversary Meeting was given from the Chair. Professor J. D. Everett, Dr. J. H. Gladstone, and Dr. R. H. Scott were by ballot elected Auditors of the Treasurer's accounts on the part of the Society. The following Papers received during the Recess, and published or Argon and its Companions. 329 in course of publication, in accordance with the Standing Orders of Council, were read in title : — " South African Horse-sickness : its Pathology and Methods of Protec- tive Inoculation." By ALEXANDER EDINGTON, M.B., C.M., F.RS.E., Director of the Colonial Bacteriological Institute, Cape Colony. Communicated by Sir DAVID GILL, F.E.S. " Note on the Occurrence of a Seed-like Fructification in certain Palaeozoic Lycopods." By D. H. SCOTT, M.A., Ph.D., F.R.S., Honorary Keeper of the Jodrell Laboratory, Eoyal Gardens, Kew. " The Demarcation Current of Mammalian Nerve. (Preliminary Communication.) Parts I — III." By J. S. MACDONALD, B.A., L.RC.P.E. Communicated by Professor SHERRINGTON, F.R.S. The following Papers were read : — I. "Argon and its Companions." By Professor W. RAMSAY, F.R.S., and Dr. M. W. TRAVERS. II. " Data for the Problem of Evolution in Man. VI. — A First Study ' of the Correlation of the Human Skull." By Dr. ALICE LEE and Professor K. PEARSON, F.R.S. III. " Mathematical Contributions to the Theory of Evolution. IX. — On the Principle of Homotyposis and its Relation to Heredity, to the Variability of the Individual, and to that of the Race. Part I. — Homotyposis in the Vegetable Kingdom." By Pro- fessor K. PEARSON, F.R.S. IV. " A Chemical Study of the Phosphoric Acid and Potash Contents of the Wheat Soils of Broadbalk Field, Rothamsted." By Dr. BERNARD DYER. Communicated by Sir J. H. GILBERT, F.R.S. "Argon and its Companions." By WILLIAM RAMSAY, F.R.S., and MORRIS W. TRAVERS, D.Sc. Received November 13, — Read November 15, 1900. (Abstract.) The discovery of krypton and neon was announced to the Royal Society in the early summer of 1898 ; and subsequently atmospheric air was found to contain a heavier gas to which the name of xenon was applied. Mr. Baly, in the autumn of the same year, called attention to the presence of helium lines in the spectrum of neon, an observation 1'r.if. W. Ranisay and Dr. M. W. Travers. which confirms that made by Professor Kayser, of Bonn, and by Dr. Friedliinder, of Berlin. At the same time we imagined that we had obtained a gas with a spectrum differing from that of argon and yet of approximately the >.imc density ; to this gas we gave the name metargon. It has now been found that the presence of the so-called metargon is to be accounted for by the fact that in removing oxygen from the mixture of these gases, which was then in our hands, phosphorus containing carbon was employed ; this mixture when burned in oxygen yields a spectrum to some extent identical with that furnished by carbon monoxide, but ditt'ering from it in as much as lines of cyanogen are also present. We have no doubt that the so-called metargon, the spectrum of which is visible only at high pressure, and only when impure phosphorus has been employed to remove oxygen, must be attributed to some carbon compound. In spite of numerous experi- ments we have not yet succeeded in producing any gas in quantity which yields this composite spectrum. It is only to be obtained by a mixture of carbon monoxide with cyanogen. To obtain the heavier gases krypton and xenon, a large amount of air was allowed to evaporate quietly; the residue was freed from oxygen and nitrogen, and then consisted of a mixture of krypton, xenon, and argon, the last forming by far the largest portion of the gas; this mixture was liquefied by causing it to flow into a bulb immersed in liquid air, and the bulk of the argon was removed as soon as the temperature rose, the krypton and the xenon being left behind. By many repetitions of this process we were finally successful in separating these three gases from each other. While krypton has a considerable vapour-pressure at the temperature of boiling air, the vapour-pressure of xenon is hardly appreciable, and this afforded a means of finally separating these two gases from one another ; in the complete paper the operations necessary to separate them are fully described. For neon the process of preparation was different. The air liquefier furnished a supply of liquid air ; the gas escaping from the liquefier consisted largely of nitrogen; this mixture was liquefied in a bulb immersed in the liquid air which the machine was making. When the bulb had been filled with liquid nitrogen a current of air was blown through the liquid until some of the gas had evaporated. That gas was collected separately, and deprived of oxygen by passage over red- hot copper ; it contained the main portion of the neon and the helium present in the air. The remainder of the nitrogen was added to the liquid air used for cooling the bulb in which the nitrogen was con- densed. Having obtained a considerable quantity of this light nitrogen it was purified from that gas in the usual manner, and the argon containing helium and neon was liquefied. By fractional distillation Arc/on and its Companions. it was possible to remove the greater portion of the helium and neon from this mixture of gases, leaving the argon behind. Many attempts were made to separate the helium from the neon. Among these was fractional solution in oxygen, followed by a systematic diffusion of the two gases ; but it was not found possible to raise the density of the neon beyond the number 9-16, and its spectrum still showed helium lines. It was not until liquid hydrogen made by an apparatus designed and built by one of us (M. W. T.) had been produced in quantity, that the separation was effected ; the neon was liquefied or perhaps solidified at the temperature of boiling hydrogen, while the helium remained gaseous. A few fractionations serve to produce pure neon ; we did not attempt to separate the helium in a pure state from this mixture. That these are all monatomic gases was proved by determination of the ratio of their specific heats by Kundt's method ; the physical pro- perties which we have determined are the refractivities, the densities, the compressibilities at two temperatures, and of argon, krypton and .xenon the vapour-pressures and the volumes of the liquids at their boiling points. The results are as follows : — Helium. Neon. Argon. Krypton. Xenon. Refractivities (Air = 1) Densities of gases (O = 16) Boiling-points at 763 mm. 0-1238 1-98 '? 0 -2345 9-97 ? 0-968 19-96 86-9° abs. 1-449 40-88 121 -33° abs. 2-364 64 163 -9° abs. Critical temperatures Critical pressures ? p below 68° abs. 9 155 -6° abs. 40'2 210 -5° abs. 41 -24 287 -7° abs. 43 '5 metres metres metres Vapour-pressure ratio .... Weight of 1 c.c-. of liquid. . 9 ? P P 0 '0350 1-212 0-0467 2 155 0 -0675 3-52 ? ? grammes 32 '92 grammes 37 -84 grammes 36-40 The compressibilities of these gases also show interesting features. They were measured at two temperatures — 11 '2° and 237'3° ; the value of P.V. for an ideal and perfect gas at 11-2° is 17,710 metre-cubic- centimetres, and at 237'3° to 31,800. This is, of course, on the assumption that the product remains constant whatever be the varia- tion in pressure. Now with hydrogen at 1142° C. the product increases with the rise of pressure ; with nitrogen, according to Amagat, it first decreases slightly and then increases slightly. With helium the in- crease is more rapid than with hydrogen ; with argon there is first a considerable decrease followed at very high pressures by a gentle 332 Artjon and its increase, although the product does not reach the theoretical value at 100 atmospheres pressure ; with krypton the change with rise of pressure is a still more marked decrease, and with xenon the decrease is very sudden. At the higher temperature the results are more difficult to interpret; while nitrogen maintains its nearly constant value for P.V., helium decreases rapidly, then increases, and the same peculiarity is to be remarked with the other gases, although they do not give the product of P.V. coinciding with that calculable by assuming that the increase of P.V. is proportional to the rise of absolute temperature. These last experiments must be taken as merely preliminary ; but they show that further research in this direction would be productive of interesting results. The spectra of these gases have been accurately measured by Mr. E. C. C. Baly, with a Rowland's grating ; the results of his measure- ments will shortly be published. It may be remarked, however, that the colour of a neon-tube is extremely brilliant and of an orange-pink hue ; it resembles nothing so much as a flame ; and it is characterised by a multitude of intense orange and yellow lines ; that of krypton is pale violet ; and that of xenon is sky-blue. The paper contains plates showing the most brilliant lines of the visible spectrum. That the gases form a series in the periodic table, between that of fluorine and that of sodium is proved by three lines of argument : — (1) The ratio between their specific heats at constant pressure and constant volume is 1-66. (2) If the densities be regarded as identical with the atomic weights, as in the case with diatomic gases such as hydrogen, oxygen, and nitrogen, there is no place for these elements in the periodic table. The group of elements which includes them is : — Hydrogen. 1 Helium. 4 Lithium. 7 Beryllium. 9 Fluorine. 18 Neon. 20 Sodium. 23 Magnesium. 24 Chlorine. 35-5 Argon. 40 Potassium. 39 Calcium. 40 Bromine. 80 Krypton. 82 Rubidium. 85 Strontium. 87 Iodine. 127 Xenon. 128 Caesium. 133 Barium. 137 (For arguments in favour of placing hydrogen at the head of the fluorine group of elements, see Orme Masson, ' Chem. News,' vol. 73, 1896, p. 283.) Data for the Problem of Evolution in Man. 333 (3) These elements exhibit gradations in properties such as re- fractive index, atomic volume, melting-point, and boiling- point, which find a fitting place on diagrams showing such periodic relations. Some of these diagrams are reproduced in the original paper. Thus the refractive equivalents are found at the lower apices of the descending curves ; the atomic volumes, on the ascending branches, in appropriate positions ; and the melting- and boiling-points, like the refractivities, occupy positions at the lower apices. Although, however, such regularity is to be noticed, similar to that which is found with other elements, we had entertained hopes that the simple nature of the molecules of the inactive gases might have thrown light on the puzzling incongruities of the periodic table. That hope has been disappointed. We have not been able to predict accurately any one of the properties of one of these gases from a knowledge of those of the others ; an approximate guess is all that can be made. The conundrum of the periodic table has yet to be solved. 'Data for the Problem of Evolution in Man. VI. — A First Study of the Correlation of the Human Skull." By ALICE LEE, D.Sc., with some assistance from KARL PEARSON, F.RS., University College, London. Pteceived July 13, — Read November 15, 1900. (Abstract.) The substance of this paper was a thesis for the London D.Sc. degree ; it was shown to Professor Pearson, at whose suggestion con- siderable modifications were made, and a revision undertaken with a view to publication. In order to deal exactly with the problem of evolution in man it is necessary to obtain in the first place a quantitative appreciation of the size, variation, and correlation of the chief characters in man for a number of local races. Several studies of this kind have been already undertaken at University College. These fall into two classes, (i) those that deal with a variety of characters in one local race, and (ii) those which study the comparative value of the constants from a variety of races. Thus Dr. E. Warren has dealt with the long bones of the Naqada race,* Mr. Leslie Bramley-Moore has compared the regression equations for the long bones from a considerable number of races in a memoir not * ' Phil. Trans.,' B, vol. 189, p. 135. 334 Dr. A. Lee and Prof. K. Pearson. \ct published, Professor Pearson has dealt with the regression equations for stature and long bones as applied to a variety of races ;* Miss A. Whiteley has studied the correlation of certain joints of the hand,t and is investigating the correlation of the bones of the hand in a second local race ; Miss C. D. Fawcett has made a long series of measurements on the Xaqada skulls, and correlated their chief characters ; the present memoir, on the other hand, deals with only a few characters in the skull, comparing, however, the results obtained from a variety of local races. It is thus related to Miss Fawcett's work much as Mr. Bramley- Moore's to Dr. Warren's, i.e., it endeavours, by selecting a few characters and testing them, to ascertain how far results obtained for one local race are valid for a second. In Professor Pearson's memoir on the reconstruction of the stature of prehistoric races, results obtained from one local race were then extended to a great variety of other races. The degree of accuracy in this procedure can only be fully ascertained when the data now being collected in both English and German ana- tomical institutes are available for calculation. The skull, however, differs very widely from the stature and long bones ; for, while these have a very high degree of correlation in all local races, the chief characters of the skull are very loosely correlated, and such correlation as they possess varies in a remarkable manner with sex and race. This was first indicated by Professor Pearson ;J it has been amply illustrated in the measurements of Miss Fawcett, and is confirmed in a recently published memoir by Dr. Franz Boas. It may be said that this want of correlation in the parts of the skull is the origin of its great importance for the anthropologist; it is the source of its personal and racial individuality. But this anthropological advantage is, from the standpoint of organic evolution, a great dis- advantage. Cuvier introduced the conception of correlation with the idea of reconstructing from a single bone the whole skeleton and even the outward form of an extinct animal, but the great want of correlation between the parts of the skull, and between the skull and other parts of the human skeleton, renders quantitative reconstruction — and this is the really scientific reconstruction — of one character of the skull from a second, or of the skull and parts of the skeleton from each other ex- tremely difficult, if not impossible, for all but a very few characters. Among these characters one of the most feasible to deal with, and one of the most useful, is the capacity of the skull. This is correlated to a fairly high degree (although to nothing like the same extent as the long bones among themselves) with the maximum length and breadth, with the total and auricular heights, and with the horizontal and * ' Phil. Trans.,' A, vol. 192, p. 169. t ' Roy. Soc. Proc.,' vol. 65, p. 126. J ' Phil. Trans.,' A, vol.- 187, p. 279, and ' Roy. Soc. Proc.,' voL 60, p. 495. Data for the Problem of Evolution in Man. 335 vertical circumferences of the skull. The present memoir deals in the main with the problem of the reconstruction of the capacity from these characters. Three fundamental problems arise in the theory of reconstruction, i.e., the determination of the probable value of an unknown character from a known and measurable one, or from several such. Namely : — I. The reconstruction of the individual from data for his own race. II. The reconstruction of the average value of [a character in one local race from data determined for a second local race. III. The determination of the probable value in an individual of characters not measurable during life from characters which are measurable. These three problems are all dealt with for the special character capacity of the skull in the present paper. Their importance may be indicated by the following considerations : — (a.) Many, especially of the more ancient and accordingly more interesting skulls, are too fragile or too fragmentary to allow of their capacity being directly determined. (ft.) The methods for directly determining capacity are still not only very diverse, but divergent in result, and from the physical stand- point crude and inexact. In the concordat of the German craniologists — the Frankfurter Verstiindigung — the point was left for future con- sideration, and so it has remained for many years. The capacities of series of skulls determined during the past forty years in France, England, and Germany are, we are convinced, not comparable, at least if the argument from the comparison is to depend on a difference of 30 to 40 cm.3 While the same observer using different methods may be trained to get results within 4 to 6 cm.3 for the same skull, different observers, equally careful, using the same method, will easily get results for the same series diverging by 20 to 30 and even more cubic centimetres. Shortly, the personal equation — involved in the packing in the skull and in the measuring vessel — is very large. Accordingly a regression equation for the capacity as based on external measurements may, if deduced from a sufficiently large range of series measured by careful independent observers, give results fairly free from the error of personal equation and this sensibly as correct as, or more correct than, direct measurement when we require the mean capacity of a series. (c.) It is impossible to obtain a large series of skulls belonging to known individuals with a classified measure of intellectual ability. Actually we have only a few skulls of men of great intellectual power, sometimes preserved because they were large, and to compare with these the skulls of the unknown and often the ill-nourished, which reach the 336 /- •'/<'/•//" J'rofifi in <>f Ei-v'utivn in .1. anatomical institutes.* Accordingly it is an investigation of con- siderable interest to compare the pribabU capacity of the skulls of living persons with their roughly appreciable intellectual grade. It is only by such a comparison that we can hope to discover whether the size and shape of the skull is to any extent correlated with brain power. In the course of the memoir it is shown that the auricular height of the skull is a better measurement for determining skull capacity than the total height ; that the circumferences of the skull, while highly correlated with its capacity, give regression equations which vary widely from one to another closely-allied race ; that linear regression equations involving length, breadth, and auricular height, while giving fairly good results for individuals within the local race, have very divergent coefficients as we pass from local race to local race ; that the cephalic index has very little correlation with capacity at all (as a rule what there is may be summed up in the words : In a brachy- cephalic race the rounder the skull the greater the capacity, in a dolichocephalic race the narrower the skull the greater the capacity — the greater capacity following the emphasis of the racial character) ; finally, that the correlation of capacity with the triple product of length, breadth, and height gives a regression equation which is fairly constant from local race to local race, and is accordingly the best available. From this and other equations individual and racial reconstructions are made, and the deviations between the actual and predicted capacities in randomly chosen series of skulls are tabulated. The mean error made in the reconstruction of the individual capacity by the best formulae is 3 to 4 per cent., the maximum error, although of course infrequent, may even be 10 per cent. For the reconstruction of the mean capacity of a race, the mean error is about 1*2 per cent., with a maximum error of 2*5 per cent. If these errors appear large to the craniologist, we would remind him that his search for an absolutely correct formula giving cranial capacity from external measurements is the pursuit of a Will-o'-the-wisp. The theory of probability shows us exactly the sort of errors such formulae are liable to, and teaches us how to select the best. The whole basis of the theory of evolution, the variability of one character, even with fixed values for a number of others, would be upset if any such absolute formula were forthcoming. What we have to do is to select a few organs as highly correlated as possible, but, having done this, it has been shown elsewhere that we shall not sensibly decrease the error of our prediction by increasing the number of organs upon which the estimate is based, t Accordingly we do not believe that sensibly better reconstruction formulae than those * This argument applies also, in even an intensified degree, to the determinations of brain weight. t ' Phil. Trans.,' A., vol. UK), p. 466. Total Eclipse of the Sun, May 28, 1900. 337 found will ever be forthcoming, for, as we have already observed, we know from Miss Fawcett's wide series of skull correlations that we have practically chosen the organs of the highest correlation. Better data for determining the equations will undoubtedly be available as further craniological measurements are made, or as the great mass already made are quantitatively reduced. In the last place we turn to the third problem : the reconstruction of the capacity of the living head. The memoir contains tables of the skull capacity of some sixty men, and also of some thirty women, whose relative intellectual ability can be more or less roughly ap- preciated. It would be impossible to assert any marked degree of correlation between the skull capacities of these individuals and the current appreciation of their intellectual capacities. One of the most distinguished of Continental anthropologists has less skull capacity than 50 per cent, of the women students of Bedford College ; one of our leading English anatomists than 25 per cent, of the same students. There will, of course, be errors in our probable determinations, but different methods of appreciation lead to sensibly like results, and although we are dealing with skull capacity, and not brain weight, there is, we hold, in our data material enough to cause those to pause who associate relative brain weight either in the individual or the sex with relative intellectual power. The correlation, if it exists, can hardly be large, and the true source of intellectual ability will, we are convinced, have to be sought elsewhere, in the complexity of the convolutions, in the variety and efficiency of the commissures, rather than in mere size or weight. "Total Eclipse of the Sun, May 28, 1900. Preliminary Account of the Observations made by the Solar Physics Observatory Eclipse Expedition and the Officers and Men of H.M.S. 1 Theseus,' at Santa Pola." By Sir NORMAN LOCKYER, K.C.B., F.E.S. Eeceived June 22, 1900.— Eead at Joint Meeting of the Eoyal and Eoyal Astronomical Societies, June 28, 1900. The observing station selected for my party was determined upon from information supplied by the Hydrographer, Eear-Admiral Sir W. J. L. Wharton, E.N., K.C.B., F.E.S. Santa Pola appeared likely to meet the requirements of a man-of-war, and without such assistance as a man-of-war can render, the manipulation of long focus prismatic cameras in eclipse observations in a strange country is impracticable. Santa Pola lies very near the central line of the eclipse, and good anchorage was available, protected from some winds. 338 Sir Norman Lockyrr. Before leaving England, I communicated with Professor Francisco Iniguez e* Iniguez, Director of the Madrid Observatory, and Mr. Jasper Camming, II. M. Vice-Consul at Alicante. These gentlemen, together with Don Jose* Bonmati Mas, a large landed proprietor, and father of the Mayor of Santa Pola, very kindly made all the necessary prelimi- nary arrangements with the local authorities, who had also been instructed by the Spanish Government, after representations had been made by the Foreign Office, at the request of the Royal Society. As a result of the Royal Society's application to the Admiralty, II. M.S. "Theseus," commanded by Captain V. A. Tisdall, R.N., was told off to meet the expedition at Gibraltar, and convey the observers to Santa Pola. The expedition consisted at first of Dr. W. J. S. Lockyer, from the Solar Physics Observatory, Mr. A. Fowler, the demonstrator in Astronomical Physics, from the Royal College of Science, and Mr. Howard Payn, who joined as a volunteer ; I subsequently received orders to accompany and take charge of it. Mr. Payn went on in advance overland to make preliminary ar rangements and to lay out the camp on a plan which had been previously arranged, while the remaining observers left England on May 11, by the R.M.S. " Oruba," of the Orient line. On arriving at Gibraltar, the party at once went on board H.M.S. " Theseus," and left for Santa Pola, which was reached just before noon the following day, May 17. I was glad to find that great interest had been shown in the expedition before our arrival on board, and that lectures on the work to be undertaken had already been given by the Chaplain, the Rev. G. Brooke-Robinson, M.A. Assistants were at once forthcoming for working the prismatic cameras, and also for manipulating several cameras which I had brought out to be used by the ship's company in obtaining photo- graphs of the corona. Observing parties in charge of officers of the ship, to make observa- tions along several lines, were at the same time organised. On our arrival at Santa Pola, the following local officials came on board with Mr. Payn : — Sns. Francisco Bonmati Mas, Mayor of Santa Pola ; Antoine Bonmati Mas, Vice-Mayor of Santa Pola ; Jose* Bonmati Mas, Municipal Councillor ; Jose* Salinas Perez, Municipal Councillor ; Eladio Ponce de Leon, Secretary to the Mayor ; Michel Sempere, Justice of the Peace ; Jose" Hernandez, Captain of the Port ; Geronimo Agnati, Administrator of Customs ; Eduard Fernandez, 1st Lieut, of Coast Guards ; Tomas Bueno, Medical Officer. They informed us that permission had been given for land ing parties from the man-of-war, and special facilities granted for landing instru- ments and personal baggage without Custom's examination. The erection of the instruments, huts, and tents was commenced on Total Eclipse of the Sun, May 28, 1900. 339 the following morning, .May 18, and by the evening of May 21 the principal instruments were reported in approximate adjustment. Drills were begun on May 22, and were carried on several times a day up to the day of the eclipse. In this work the eclipse clock, which I have described in previous eclipse reports, was used. By permission of the Captain, three of the officers of the " Theseus," Lieuts. Andrews, Doughty, and Pattrick, R.N., occupied quarters on shore to superintend the work of the parties in the camp. On board, the Chaplain gave instructions in sketching coronas and recording stars, using for this purpose a lantern which had been placed at the disposal of the expedition by the Orient Steam Navigation Company. The weather was very favourable for the work of the expedition, but at times the landing and embarking of parties from the ship was rendered difficult by strong sea breezes and the consequent surf. Both day and night the instruments were carefully guarded by a detachment of " Guardias Civiles," told off for the purpose by the Spanish authorities. The groups of observers were as follows : — LIST OF ECLIPSE PARTIES. Timekeepers. Lieut. F. A. Andrews, K.N. J. Wale, 2nd Yeoman Signals. Mr. Boughey, Mid. W. Webb, P.O. 1. Mr. Lambert, Mid. Bugler Sneller, O.S. 6-inch Prismatic Camera. Dr. Lockyer. C. Willmott, O.S. S. Birley, E.R.A. A. Humphries, O.S. J. Green, A.B. G. Hyatt, O.S. C. Fishenden, O.S. 20-foot Prismatic Camera. Mr. Fowler. A. Maskell, A.B. W. F. Cox, Armr. E. Davies, O.S. A. Wbitbourne, A.B. H. Cristopher, O.S. F. Burt, A.B. W. Harrison, Sto. Mech. 4-inch Equatorial. Sir Norman Loc-kyer, K.C.B. C. C. Lambert, Mid. 3f -inch Equatorial. Lieut. H. M. Doughty, E.N. A. GK N. Lane, Mid. Long-focus Coronagraph. Mr. Payn. H. Eary, A.B. T. McGowan, A.B. W. Mann, O.S. E. Woodland, A.B. H. Brooks, O.S. 340 Sir Norman Lockycr. Graham Coronagraph. Mr. W. J. S. Perkins, Awt. Engr., R.N. J. Knowles, Chief Stoker. W. Walker, Lg. Stoker. De La Rue Coronagraph. Mr. H. W. Portch, Asst. Engr., R.N. H. Frost, Chief Stoker. W. Waterfield, E.R.A. Dallmeyer Coronagraph. Surgeon J. Martir, R.N. E. Quint, Chief Stoker. E. Buckingham, E.R.A. Discs. Mr. J. B. Bateman, Mid. E.N. T r Mr. J. A. Daniels, Torp. Gunuer, R.N. W. Fraser, Arm. Crew. I < Or. Fair, Armourer. R. S. Bradbrooke, A.B. J I E. Gordon, 8. Carp- H. W. Richardson, P.O. 2. i r W. Tucker, A.B. E. Vojle, Lg. Shipwt. I -I W. Brewer, A.B. T. Orange, Boy, 1 c. J I B. Salmon, Boy, 1 c. A. Mason, A.B. -| r A. May, A.B. A. Steven, A.B. I -I H. Bailey, A.B. C. Paul, Boy, 1 c. J I J. Entwistle, S. Std. Boy. Sketches of Corona without Discs (on shore). W. Butt, M.AJL. H. Meacher, Pte. R. M.L.I. G. Guilliame, A.B. H. Schmidtael, O.S. Sketches of Corona without Discs (on board). W. Baiter, A.B. J. Wheeler, Pte. R.M.L.I. W. Butts, Ttc. R.M.L.T. E. Willis, S.B. Attendant. C. Jacob, Pte. R.M.L.I. Observations on Stars (on shore). Mr. Bennett, Clerk. H. Angus, O.S. W. Eiches, L. Seaman. W. Kinrett, Pte. R.M.L.I. A. Pontifei, A.B. W. Oliver, Pte. R.M.L.I. W. Bos worth, A.B. Observations on Stars (on board). Rer. G. B. Robinson, M.A. E. Hammond, Sto. H. Croxon, S. Corpl. G. Andrews, Sto. A. Phillips, Leading Shipwt. G. Nightingale, Sto. E. Vigus, Corpl. R.M.L.I. S. Wilson, Sto. E. Price, Pte. R.M.L.I. E. Savage, Pte. R.1I.L I. Observations of Shadow Sands (on shore). Commander Hon. R. F. Boyle, R.N. Mr. J. G. Walsh, Mid. R.N. Mr. T. Slator, Xaval Instructor, R.N. Mr. F. C. Skinner, Mid. R.N. Meteorological Observations (on shore). Lieut. Pattrick, R.N. Mr. G. S. Hallowes, Mid. R.N. Total Eclipse of the Sun, May 28, 1900. 341 Meteorological Observations (on board). G. Donnelly, Yeom. Sig. W. Hearne, Sig. E. Gant, Lg. Sig. J. Beach, Sig. A. Enstidge, Sig. Landscape Colours (on shore). Capt. F. V. Whitmarsh, E.M.L.I. Lance-Corpl. Wade, E.M.L.I. Ship's Steward D. Green. W. Birkett, Writer. Landscape Colours (on board). Fleet Paymaster A. W. Askham, E.N. Lieut. W. J. Frazer, E.N. Shadow Phenomena (on shore). Mr. C. Prynn, Carpr. E.N. Shadow Phenomena (on board). Lieut. H. E. Shipster, E.N. Photographers. J. Knight, S.B. Steward. B. Bulbrook, A.B. Aide-de-Camp to Sir Norman Loekyer, K.C.B., F.R.S. Mr. C. C. Lambert, Mid. E.N. Time Arrangements. According to the Admiralty chart, the latitude and longitude of the place of observation are 38° 11' 20" N. arid 0° 33'66' W. respec- tively. For this point, the. times and position angles of contact derived from the formulae given in the ' Nautical Almanac Circular,' No. 17, were as follows : — Beginning of totality, May, 28 d. 4 h. 12 m., 51'7 s. End „ „ 4h. 14 m., 10'5 s. Duration of totality 1m., 18-8 s. Position angle of first contact, 87° 3-5' from N. towards W. last „ 93° 47-3' „ „ E. The experience of the Indian eclipse of 1898 suggested that the duration of totality was too long, and for the practical working during the eclipse the adopted time was 75 seconds, so that there would be no chance of spoiling the coronagraph plates by exposing them after totality. The face of the eclipse clock was graduated accordingly. The arrangements for securing signals at definite intervals before totality was identical with that employed in Lapland and India. An image of the sun projected by the finder of the 6-inch two-prism pris- matic camera was viewed on an adjustable screen, marked in such a way that it was easy to see when the cusps subtended angles of 90J VOL. LXVII. 2 C :\\'2 Sir Xiirniiin Lockyer. .mil ")5°, which occurred respectively at 16 sees, and 5 sees, before totality. The signals " Go " at the commencement of totality, and " Over " at the end, were given by myself, from observations made with the 4-inch Cooke telescope. Some of the photographs have not yet l>een developed, and the reports have not yet been received from the ship's parties, so that only a very brief reference to the work accomplished is possible. The discussion of the series of photographs taken with the prismatic cameras employed in the last three eclipses indicated that continued work with this form of spectroscope should be undertaken, (1) with the view of obtaining data strictly comparable with the previous photographs, and (2) that an effort should be made to extend the inquiry into comparative lengths of the various arcs. For the first purpose it seemed desirable to repeat the Indian work with the 6-inch camera having two prisms, while for the second an instrument of longer focus was necessary. Representations as to the importance of the latter instrument were made to the Royal Society, and ultimately the purchase of a Taylor triple lens of 6 inches aperture and 20 feet focal length was autho- rised. This was received so shortly before the expedition left England, that it was only possible to make a rough trial of the instrument before it was set up at Santa Pola. Both prismatic cameras were worked in conjunction with siderostats, calculations having shown that the position angles of contact were favourably situated after reflection. Dr. Lockyer took charge of the two-prism instrument, and Mr. Fowler of that having a long focus, and in each case the programme of exposures was successfully performed. The photographs which have been developed indicate the same suc- cession of phenomena recorded in the three previous eclipses, but the recent eclipse was specially advantageous, for the reason that the chromospheric arcs at the instant of contact were of greater length. A very complete record of the spectrum of the chromosphere at various depths has been secured with both instruments, and it seems probable that new information as to the distribution of the various vapours will be furnished by the photographs taken with the long- focus instrument. The spectrum of the corona shows the green ring at 5303*7, the blue ring at 4231, and the violet ring at 3987'0 : others may possibly appear on closer examination. All the rings are of totally different character from the chromospheric arcs, and have their greatest bright- Totnl Eclipse of the Sun, May 28, 1900. 34:5 ness in regions other than those where the chromospheric ares are brightest. As before, 3987 '0 is much more uniform in brightness1 throughout the extent of the ring than the others ; 5303-7 is especially •strong in one or two regions ; but on the whole is probably weaker than in 1898. The photographs show that the scale of the spectra is by no means too large for work with short exposures with a lens of 6 inches aper- ture. The spectra are 7*5 inches long from D3 to K, and the diameter •of the rings is 2'5 inches ; photographs taken with an exposure esti- mated at £ of a second are fully exposed. The Differences between tlie Coronas observed at the Periods of Sun-spot Maxima and Minium. My attention was called especially to these differences, because I •saw the minimum eclipse of 1878, while the phenomena of that of 1871 (maximum) were still quite fresh in my mind. My then pub- lished statements have been amply confirmed during the eclipses which have happened since 1878, but certainly the strongest confirmation has been obtained during the present one, which took place two more spot periods after 1878. 1. Form. "With regard to form, at the instant of totality I saw the 1878 corona over again, the wind vane appearance being as then most striking. 2. Tlie Spectrum. In connection with the eclipse of 1878 (minimum), I pointed out that, whereas in 1871 (maximum) the spectrum of the corona viewed by small dispersion was remarkable for the brightness of the lines ; in 1878 they were practically absent, and the continuous spectrum was remarkably brilliant. I determined therefore to make a similar observation in this year of maximum, and, as in 1878, used a grating first order spectrum placed near the eye. The result was identical with that recorded in 1878. I saw no obvious rings or arcs, but chiefly a bright continuous spectrum. 3. Tlie Minute Structure of the Inner Corona. Lieut. Doughty, E.N., and myself made observations on the minute structure of the corona, in order to see if any small details could be observed, and whether they were the same as those I saw so well and recorded during the eclipse of 1871, at a period of sun-spot maximum. This question was specially taken up this year, as exactly two sun- spot periods have elapsed since 1878. 2 c 2 344 Sir Xoruiun In 1871 I used a 6-inch object glass, and distinctly observed marked delicate thread-like filaments, reminding one of the structure of the prominences, with mottling and nebulous indications here and there ; some of these distinct markings were obvious enough to be seen till some minutes after totality.* This year, with a perfect 4-inch Taylor lens and a high power, not the slightest appearance of this structure was to be traced ; the corona some 2' or 3' above the chromosphere was absolutely without any detailed markings whatever. Lieut. Doughty duplicated and confirmed these observations with a 3f Cooke. Here, then, is established another well-marked difference between maximum and minimum coronas. Tlie Coronagraplis. Four coronagraphs were employed of various apertures and focal lengths. One, of 4 inches aperture and 16 feet focal length, was in charge of Mr. Howard Payn, while the others were controlled by officers of the ship. The results obtained are very satisfactory, those taken with the long-focus instrument being especially good. In this case the image is If inches in diameter, and the definition is perfect. The photograph taken with an exposure of 5 seconds shows a great wealth of detail in the inner corona and prominences ; the fine definition appears to be due to the fact that a Taylor photo-visual lens was employed, bringing the rays of various refrangibilities to the same focus. A long ex- posure photograph, with the same instrument, is remarkable for the perfect hardness of the moon's edge, notwithstanding the motion during totality. The three photographs secured by Asst. Engineer Portch, R.N., with the De la Rue lens of 4£ inches aperture, give also sharp images with much fine detail. Sandell triple-coated plates were used with this instrument. With the 6-inch Dallmeyer lens, two photographs on Sandell plates were obtained by Dr. Martin, R.N., one being exposed for about half a second, and another for 50 seconds. The longer exposure records the extensions to a greater distance from the dark moon than any of the other photographs obtained, with the exception of the one secured with the small-grating camera. This last-mentioned instrument consisted of a Zeiss anastigmatic lens of 9 inches focal length, with a small Thorp grating mounted in front of it. The exposure of the plate was 40 seconds during totality \ the longest streamer in the N.E. quadrant extends to a distance of 4i lunar diameters. * ' Solar Physics,' p. :<":>. Total Eclipse of the Sun, May 28, 1900. 345 Discs. Six discs for cutting out the bright light of the inner corona were erected, with the view of enabling the observers to detect the long extensions if there should be any. They were very carefully set up by Lieuts. Doughty and Andrews, E.N., and were provided with eye- pieces having all necessary adjustments. Mr. Daniels, torpedo gunner, then took charge of the party, and numerous rehearsals were given. In the trials remarkable skill in recording delicate details was dis- played. During the eclipse, the actual observer was blindfolded for five minutes before totality. No extensions of the nature observed by Prof essor Newcomb in 1878 were recorded. Observations on tlie Stars Visible during Totality. A large party for the observation of stars visible during totality was trained and organised by the Chaplain, Eev. G. Brooke-Eobinson, E.N., who was provided with a set of star charts for purposes of in- struction prior to the eclipse, and another set, prepared by Dr. Lockyer, for making records during the eclipse. Venus became visible at a very early stage of the eclipse, and during totality Mercury was a very conspicuous object near the extremity of one of the streamers, a Tauri, a and y Orionis were also recorded. No comet or unknown body was noted. Shadow Bands. The Naval Instructor on H.M.S. " Theseus," Mr. T. Slator, B.A., undertook this branch of the eclipse work, and during the eclipse worked in conjunction with the Commander, the Hon. E. F. Boyle. Very complete arrangements were made for securing the orientation of the bands (1) on a horizontal plane ; (2) on a plane in the meridian ; (3) on a plane in the prime vertical. The bands appear to have been very ill-defined, but the necessary observations were secured in planes 1 and 2. Meteorological Observations. A regular series of observations of temperature and pressure was established three days before the eclipse, and continued until two days after ; Lieut. Pattrick, E.N., taking charge of this branch of the work. During the eclipse the temperature fell 5° C., and the barometer also fell slightly. The thanks of the expedition are due especially to those named in 346 I'K.f. H. H. Turner and Mr. H. K. Nc-wall. the foregoing account, not only for assistance rendered, but also for their great kindness to us. I have already, in a letter, expressed to- the Royal Society my deep sense of the obligation they have laid u> under. As in the case of the " Volage " and " Melpomene," the officers and men of the "Theseus" not only assisted us with certain instruments, but organised crews for others, and many lines of work which it was. impossible for the observers sent out from England to attempt. Their skill, resourcefulness, and steadiness were alike truly admirable. Thanks are also due to the Managers of the Orient Steam Naviga- tion Company, who conveyed the instruments to and from Gibraltar freight free. I may add, the Civil Governor of the Province of Alicante, Senor don Hipoldo Caras y Gomez de Andino, visited the camp to a»ure himself that all the assistance the Spanish authorities could give had been rendered. " Total Solar Eclipse of 1900 (May 28). Preliminary Report on the Observations made at Bouzareah (in the Grounds of the Algiers Observatory)." By Professor H. H. TuKNERr M.A., F.R.S., and H. F. NEW ALL, M.A., Sec. R.A.S. Received June 28, — Read at Joint j Meeting of the Royal and Royal Astronomical Societies, June 28, 1900. The Report is presented in three parts. PART I. ORIGIN of THE EXPEDITION AHD GENERAL PREPARATIONS BY THE. Two OBSERVERS JOINTLY (§§ t — 10). PART II. SEPARATE REPORT BY PROFESSOR TURNER. §§ 11—12. The Cameras and Coalostat. § 13. The Polariscopes. §§ 14—16. Adjustments. §§ 17 — 19. Programme of Observations. §§ 20. The Standard Squares. § 21. Use of Green Screen. § 22. Integral Photometer. § 23. Development. PART III. SEPARATE REPORT BY MR. XEWAXL. § 24. The Four-prism Spectroscope with Slit. § 25. The Photographic Camera with large Objective Grating. § 26. The Polariscopic Camera (Savart Plates and Nicol Prism). § 27. Atmospheric Polarisation. § 28. General Observations. Total Solar Eclipse of 1900 (May 28). 347 PART I. 1. Origin of the Expedition. — This expedition was one of those organised by the Joint Permanent Eclipse Committee of the Royal Society and Royal Astronomical Society, funds being provided from a grant made by the Government Grant Committee. The expedition was most cordially and hospitably assisted by M. Trepied, the Director of the Algiers Observatory, and the observers are indebted to him in numberless ways for his kindness. He assigned good positions for the instruments in the Observatory grounds, and had brick piers built beforehand according to plans supplied to him by the observers. He made the arrangements for conveying the instruments to and from Algiers ; and put at the disposal of the observers a capacious dark room (which we believe he had specially arranged for the purpose) and the services of a carpenter. 2. Mr. Wesleijs Observations. — It may be here mentioned, although it does not come strictly within the scope of this report, that M. Trepied allowed Mr. W. H. Wesley, the Assistant Secretary of the Royal Astronomical Society, who has had great experience in drawing the corona from photographs, to use the equatorial coude of the Algiers Observatory during this eclipse ; and Mr. Wesley was thus enabled to make his first eye observations on the corona itself under most favour- able conditions. He joined the present expedition, but as he was the emissary of the Royal Astronomical Society and not of the Joint Committee, the report of his observations is not included here. That M. Trepied should have placed the finest instrument in the Observatory at the disposal of a foreigner is a striking instance of his scientific liberality ; and the observers call attention to it because it will indicate more clearly than any enumeration of details the kind of assistance for which they have to thank him. 3. Personnel. — The following persons took part in the expedition : — H. H. Turner, M.A., F.R.S., Savilian Professor of Astronomy at Oxford. H. F. Newall, M.A., Sec. R.A.S., Observatory, University of Cambridge. 4. Itinerary. — The observers left Charing Cross at 11 A.M. on Satur- day, May 12. They spent one day in Marseilles, and arrived at Algiers on Tuesday, May 15, proceeding in the evening of the same day to the little village of Bouzareah, which they made their headquarters, about a mile from the Algiers Observatory. The instruments had been sent round by sea (through the Papayanni Steamship Company), and should have arrived on May 10, but for some reason they did not arrive until May 17, and were delivered at the Observatory on the evening of May 18. Three whole Avorking days of the eleven which had been 3 I* Prof. H. H. Turner and Mr. H. V. Xewall. counted on were thus lost, and in order to carry out the programme .in undeniably great press of work was necessary. The day of the eclipse was fine, and many good photographs were obtained. The development of these and the packing up of the instruments fully occupied the observers till Friday, June 1. They left Algiers on Saturday, June 2, and arrived in London on Monday, June 4. But they would record the opinion that the time spent on the expedition was too short. The work was got through, but with practically no margin for contingencies, and would have been done l>etter with another week at least. 5. Position of Station. — The station was on the west side of the equatorial coude, and about 50 yards S.E. of the transit-circle, the position of which is Longitude 0"12m8»-7 E. of Greenwich. Latitude 36° 48' 0"-5 N. Height above mean sea level, 1123 feet. This spot was some distance from the central line, and 4 or 5 seconds of the 70 seconds of totality available were thus lost ; but the loss was more than counterbalanced by the many advantages of being at a fixed observatory. 6. Meteorological Conditioit*. — As regular meteorological observations were made at the Observatory, none were made by us. The day of the eclipse was the finest of our stay, and fine days preceded and followed it. On May 26, 27, and 28 the sun was seen to set in the sea, and the " green ray " was looked for and seen by several observers. The disc, when near the horizon on May 28, assumed remarkable shapes, of which the following four types were noticed by several observers : — There was at times considerable wind, as M. Trepied had warned us, but the day of the eclipse was calm. 7. In*ti-iini<'iitx, <(•'•.— (See separate reports of observers.) 8. Huts. — "Willesden canvas over wooden framework was used, and found very satisfactory, as before. Mr. Newall's hut was designed for his particular instruments, and the openings were obtained by leaving the canvas loose in the form of flaps, which were tied in the proper positions, either open or closed. Total Solar Eclipse of 1900 (May 28). 349 Professor Turner's hut was designed for general requirements, and has now been used, not only in this expedition, but as a transit hut in the determination of the longitude of Killorglin by the staff of the Koyal Observatory, Greenwich, in 1898. As it appears to satisfy the conditions, the following notes of its structure may be useful to others : — It is a skeleton wooden framework filled in by a series of panels, any one of which is removable without disturbing any other by simply taking out two screws. The panels forming the sides drop into a groove running round the base, and two screws are sufficient to hold them at the top. For the roof panels it is the upper edges which push into grooves along the central ridge, and the two fixing screws are near the eaves. The panels themselves are rectangular wooden frames with canvas stretched over them. For transport, the sides are unscrewed, and then the canvas is rolled round the ends like a window blind. The screws which fix the panels in position in the hut terminate in rings instead of the ordinary screw heads, so that they can be screwed up or unscrewed with the fingers instead of with a screw-driver, which may not be handy at the moment. It may be remarked that both the huts were securely fastened down on this particular occasion, as the wind sometimes blew a gale. 9. Assistance. — The observers were assisted in the exposures as follows : — • Mr. H. Wyles, of the Leeds Astronomical Society, counted seconds aloud from a metronome. Mr. J. Potter, of Leeds, carried from Mr. Newall's hut the informa- tion of the setting of the Savart prism (which Mr. Newall was to observe during totality) to Major K. 0. Foster, who set the correspond- ing instrument in Professor Turner's hut (see separate report of Mr. Newall). It was originally intended to shout this information, but as it was found in the rehearsals that there was occasionally difficulty in hearing, Mr. Potter undertook this conveyance as a safeguard. As the event proved, his assistance was all important, for at the actual eclipse there was so much noise from other observers in the neighbour- hood that the shout was not heard at all. Major K. 0. Foster, F.R.A.S., set Mr. Newall's savart between the second and third exposures, and at the same time changed the slit of Professor Turner's polariscope. He also uncovered the plates for long exposure soon after the beginning of totality and covered them before the end. Mr. F. L. Lucas, of Berkhamsted, made the exposures for Professor Turner at the objective. Master Eric Henn handed the plates. Mr. F. L. Crawford, of the Indian Civil Service (who had seen the 350 . H. H. Tomer and lie II. I-'. 1898 eclipse at Berar), received the plates, recorded the times, and also exposed for 10 seconds the integral photometer. Mr. Lovett Henn, of Algiers, made the exposures with the grating for Mr. Xewall. Mrs. Newall made observations of the atmospheric polarisation during totality. At 15 seconds before totality, as shown by the diminishing crescent of the sun, Professor Turner called " Stand by"; at totality, "Start": when Mr. Wyles counted from the metronome steadily up to 80. Totality lasted 64 or 65 seconds, and the extra 15 seconds was required by Mr. Newall for exposures at the second " flash." The signals were given with approximate correctness, though, by an oversight, no one timed the interval between the "Stand by " and the " Start." The operations were rehearsed several times on the day before the eclipse, and once or twice in dumb show on the actual day. It was not found possible to arrange for rehearsals earlier ; but, with the exception of the omission just noticed, everything went off at the time without a hitch. 10. Tlie Day of tlie Eclipse. — Perfectly clear all day — no anxiety. The contacts were not observed by us with special care as we had much else to do, and observations were being made by the staff of the Observatory. M. Sy kindly supplied the following predictions and observations : — Predictions. Observations. 3h I7m 18" 4 29 27 4 30 32 5 34 25 Local mean time (12ni 8*'7 in advance of G.M.T.). Lamps were not needed during totality. Owing to an accident (a signal being lost through noise made by others) the shadow was not observed. Major Kingsley Foster noticed the " shadow bands " on the white surface of the " double tube " near which he was stationed. PART II.— SEPARATE REPORT BY PROFESSOR TURNER. Instrumental Equipment. 11. The Cameras. — The double camera used at Fundium iu 1893 (by Sergeant Kearney), and at Sahdol in 1898, was modified on the present occasion. One of the 7 x 7-inch tubes contained, as before, the photo- heliograph objective No. 2 of 4 inch aperture and 5 feet focal length, with a Dallmeyer secondary magnifier of 7£ inches focus placed 1st contact.... 2nd „ .... 3rd . 3h 17'" 31" . 4 29 25 4 30 32 4th 5 34 31 Total Solar Eclipse of 1900 (May 28). 351 5 inches within the focus, giving an image of the sun 1£ inches in diameter; but the " Abney " lens was no longer used in the other tube. It had been decided by the Joint Permanent Committee to dis- continue the separate use of the two Abney lenses, and to recombine them into the original doublet, which Mr. Davidson was to use in the expedition organised by the Astronomer Royal. .Hence the other half of the double tube camera was set free, and it was utilised to good effect by arranging hco polariscopic cameras to give images on the same plate, a diagonal partition dividing the square tube into two. One instrument was arranged by Mr. Xewall, and is described by him. The other was similar to the apparatus used by me in India in 1898, but with improvements in detail as described below. The double camera is furnished with six plate holders, each taking two plates of 160 x 160 mm. (as in use for the Astrographic Chart), both plates being exposed by a quarter turn of one shutter. Alongside the double tube two other cameras were arranged for single exposures during the greater part of totality. One was a por- trait lens of 5| inches aperture and 30 inches focus, stopped down to / S ; the other was a small polariscopic camera, described below. 12. The Ccelostat. — All these cameras were pointed downwards at an angle of 18° with the horizon, in azimuth 42° west of south, to the 16-inch ccelostat used in India in 1898. The mirror of this instrument was made by Dr. Common. It was silvered and sent out to Algiers by the Improved Electric Glow Lamp Company, and had a very fine • surface. The mounting and clock of the instrument were made by Mr. J. Hammersley, from designs by Dr. Common. A steadier mount- ing is desirable on future occasions, though the present arrangement works well when there is not much wind to cause vibration. 13. The Polarucopes. — The arrangement used in India was as follows : — (A) Objective, 3i inches aperture, 18 inches focus. (B) Slit, of width 0'2 inch, in cardboard. (C) Collimator, 1| inches aperture, 6i inches focus. (D) Ehomb of spar, 1 inch aperture (clear). (E) Camera, 2 inches aperture, 9 inches focus. On the present occasion (E) was substituted for (A), which was of inconvenient width for the space at disposal. The primary image was thus reduced to half the size ; but this had the advantage that a larger part of the image fell on the slit, the width of which remained the same as before, being governed by the focal length of the collimator and the angular separation of the images by the rhomb. The colli- mator (C) and the rhomb (D) remained unchanged, but the camera lens (E) was now a photographic objective of 1^ inches aperture and 28 inches focus, made specially by Messrs. Cooke and Sons, of York. :!.-,ii Prof. II. II. Turner an.l Mi. II. K. NVwall. The plate holder wa< of course that of the double tube, us above explained. The slit (B) was arranged, as in 1898. in two portions, l.ut was on this occasion made in brass. The slit and rhomb were connected by a bar, and could l»e rotated sympathetically. They were set at such a position angle that the lines of the image parallel to the slit corresponded to vertical lines on the corona ; but this setting was found after the eclipse to be not quite accurate. The setting was not changed during totality, but the slit was moved in the direction of its length, so as to give a different part of the field between the second and third exposures. The small polariscope exposed separately resembles the objective prism spectroscope as opposed to the slit spectroscope. The reason for adopting the slit spectroscope form for the instrument alx>ve described is that the angular separation of images given by the large rhomb was not large, and if this rhomb had been simply placed in front of an objective, one image of the corona would have seriously overlapped the other. But a small rhomb (kindly lent me by Mr. Newall) gave a separation of 3£°, so that when the corona was viewed through this rhomb and an objective the two images polarised in perpendicular planes were clearly separated, though each was projected on the sky of the other. To cut out the sky backgrounds, a slit, of 1 inch aperture, was placed 15 '7 inches in front of the rhomb. Aeat sixty-three times to the minute. Further, the word " One " was called on an actual l>eat which came alxwt 0*5 second after the word " Start " (signifying the commencement of totality) had been called. This signal was given by me from a direct observation of the disappearance of the crescent, and Agreed well with the observations of others. The signal for 15 seconds before totality was given to Mr. Xewall by watching the length of the disappearing crescent on the focussing glass of the camera, and was Approximately correct, though by an oversight no one observed the interval ; but after giving this signal, as I found the direct light of the crescent did not hurt the eyes, I watched that in preference to the image on the glass. I saw the complete ring of the moon's disc quite 10 seconds l>efore totality, and from that moment the corona seemed to grow out from the limb in a most beautiful manner. 18. Programme for tlie Polariscopes. — The two polariscopes mounted in the double tube had of course exactly the same exposures as above. Between exposures of slides 2 and 3 the Xicol prism and Savart plate of Mr. Newall's polariscopic apparatus were rotated by Major K. O. Foster to the reading indicated by Mr. Xewall's eye observations ; and the slit of my apparatus was also moved by Major Foster to the second position, so that the second pair of photographs gave a different part of the corona from the first. The smaller polariscope was exposed from 5 to 60 seconds, counting from the beginning of totality. 19. Pngnunmefor the Portrait Isns. — A Sandell triple-coated plate was exposed in this instrument. The exposure was made by Major Foster from 5 to 60 seconds, counting from the taginning of totality. 20. The Mnii'liiril frjiHttr*. — On the six plates in slides, 1, 2, and 3, Sir "W. Abney's " standard squares " were impressed for photometric observations of the corona. The exposures were to a standard candle at 5 feet from the plate, which is approximately twice as bright as the full moon. Assuming the brightest part of the corona to be as bright AS the average surface of the full moon, the exposures to be given to the candle were calculated as follows : — An image of the moon in the Dallmeyer lens of 4 inches aperture -would be 1 i inches in diameter. The illumination of the object glass is thus concentrated — (4/l£)2 times = 7 times. Hence the brightest part of the corona will affect the plate about seven times as much as direct moonlight, or three and a half times as much as a candle at 5 feet. Total Solar Eclipse of 1900 (May 28). 355 Hence to compare with a 1 second exposure to the corona through the lens, we should expose the plate to the candle for 3~ seconds. Since the faintest of the standard squares obstructs some of the light, and since it is advisable to have an exposure on the plate from the standard light rather denser than the densest part of the image, the plates in slide No. 1 (to be exposed 1 second to the corona) were exposed for 6 seconds to the candle. Those in No. 2 (to be exposed 5 seconds to the corona) were exposed for 40 seconds to the candle, and those in No. 3 (20 seconds to corona) for 21 minutes. These exposures are even longer, relatively, than No. 1 ; but on previous occasions the squares had not been dense enough, and it was considered advisable to make sure of going beyond the point required. 21. Use of Green Screen. — In slide No. 4 a coloured glass screen, Ttindly provided by Mr. Shackleton, was placed in front of the plate, with a view of obtaining information on the distribution of coronium in the corona by comparing a photograph taken in green light with the others. But owing to the following circumstance this particular experiment was not a success. The plate should of course have been one sensitive to green light, and some Cadett " spectrum " plates were taken out to Algiers for the purpose. They were whole plates and required cutting down to fit the slide. In the stress of other work the experiment was forgotten until an hour before totality. There was still plenty of time to fill the slide, and I went to the dark room to do so. But the circumstances were scarcely favourable for manipulating the diamond in the dark, as these plates require. The first plate broke and cut my finger, not seriously, but enough to hamper me, so that I had no better success in cutting another plate. Indeed, after one or two attempts I had to give it up. It seemed just worth while putting in a " Rocket " plate behind the green screen, but there was very little on the plate when developed ; and the experiment was on this occasion of little or no value. I am sorry to have been unable to do justice to Mr. Shackleton's kindness in providing the screens, and hope that on another occasion I may make better use of them. 22. An Integral Photometer. — To obtain an estimate of the total light given by the corona, an Ilford " Empress " plate was exposed to its light for 10 seconds, in a small camera from which the lens had been removed, so that the corona shone directly on the plate, but side light was excluded. On the same plate standard squares were impressed by exposing it to the candle at 5 feet, as in § 8. The exposure given to the candle was, by an oversight, not recorded, but was either 10 or 20 seconds. The oversight was discovered before the eclipse, and another plate from the same batch was exposed to the candle and squares for 5, 10, and 20 seconds, and developed in the same dish. The effect of the corona was, however, considerably greater than that 356 Prof. H. H. Turner ami Mr. II. K. X.-\vall. of the 20 seconds' exposure through the thinnest of the square screens. 23. Development. — The plates in slide No. 1, and the corona picture of slide No. 2 (1 second), were developed with amidol : also the plate exposed to the integral photometer mentioned in § 12. All the others with metol. PART III. — SEPARATE RETORT BY H. F. NK \v.\u.. § 24. Tlw Four-prism Spectroscope with Slit. It was intended to attempt — (i) To secure photographs of the bright-line spectrum of the sun's limb at the beginning and end of totality, five photographs at the beginning, six at the end. (ii) To photograph the spectrum of the corona in two separate regions of the corona. It had l)een decided not to attempt to determine the velocity of rotation of the corona, for the duration of totality was not long enough to give satisfactory images of the lines in the spectrum of the corona at such distances from the limb as would ensure some measure of certainty that the observations would not deal with the local disturl>- ances known to exist near the chromosphere. The instrument arranged for the purposes above mentioned is a four- prism spectroscope with a single slit. It was used by the writer in India, at Pulgaon in 1898.* The only changes made in it were that (i) only one slit was used instead of two ; and (ii) one of the prisms which had been found to give imperfect definition on account of want of homogeneity in the glass had been replaced by another prism. The prism box and train of prisms had been used at the Cambridge Observatory for a star spectrograph, and were dismounted, for use in the eclipse, after the completion of certain observations of Capella. The train of prisms is of such dimensions and construction as to transmit a 2-inch beam of light, and to produce a minimum deviation of 180° for Hy. The collimator and camera are set parallel to one another. The whole spectroscope is mounted so as to turn about an axis parallel to the collimator. The axis is rotated (with a period of twenty-four hours) by clockwork, and is tilted so as to be parallel to the earth's axis. In this position the collimator points to the north pole, and the camera to the south pole. The tube of the collimator is prolonged beyond the plane of the slit, and is arranged to carry at its end a mirror of speculum metal and an * ' Roy. Soc. Proc.,' rol. 64, p. 55. Total Solar Eclipse of 1900 (May 28). 357 object glass, by means of which an image of the sun can be thrown upon the slit. The whole arrangement thus consists of a spectroscope combined with a polar heliostat, and in virtue of the fact that the spectroscope is rotated together with the mirror, the image of any celestial object thrown upon the slit does not rotate relatively to the slit. Further- more, the mirror is mounted in such a manner that the axis about which it can be tilted — namely, the declination axis — can be oriented relatively to the collimator tube, so that any diameter of the sun may be set parallel to the slit. A special plate holder was designed for use in Algiers in order to facilitate tlie rapid change of plates. It was charged with twelve plates, fixed film outwards on the outside of a cylinder (2 inches in diameter), whose axis was set parallel to the focal plane of the camera and in the plane of dispersion, free to turn inside a slightly larger covering cylindrical case. The arrangement was turned by hand, and worked admirably well. It is, however, only suitable for narrow spectrum plates, and might be used with very small alteration for a film on celluloid, such as is used in hand cameras of the Kodak type. The linear dispersion in the photographed spectrum is about 14 tenth- metres per millimetre at Hy. The width of the slit was adjusted to O03 mm. by a diffractional method. The scale of the photograph is such that one degree on the sky corresponds to about 9 mm. on the plate. The effective aperture of the combination regarded as an instrument for producing monochromatic images of a slit-shaped region of the corona is// 10. The adjustment of the axis of the instrument to parallelism with the earth's axis was accomplished in the same way as in India by means of a theodolite with declination circle and level, which was attached to a part of the frame of the spectroscope specially prepared for it. Programme of exposures, &c. : — I. Spectrum of the Sun's Limb at the Beginning of Totality. — Five exposures were made in 7 seconds, beginning 3 seconds before Professor Turner's signal " Start " was called, and ending as Mr. Wyles called the " fifth " beat of the metronome, Result. — The developed photographs show that the first plate was exposed at exactly the right moment to catch the spectrum of the " flash." It is filled with bright lines, and shows the part of the spectrum between Hf (3900) arid H^ (4861). The best part of the spectrum is that between wave-lengths 4100 and 4650. All the other four plates show bright lines, but the fall in the number of them is very abrupt between the first and the second plates. II. Spectrum of the Corona. — Six seconds after Professor Turner's signal " Start " a plate was exposed for the spectrum of the corona, and VOL. LXVII. 2 D 358 1W. H. II. Turner and Mr. H. F. the exposure was continued for 49 seconds, ending when Mr. \Yyles called " fifty-five." >It. — The developed photograph shows the spectrum of the corona in two regions situated at the ends of a chord whose length is approximately equal to the radius of the moon's image. The radial extension is even less than in Captain Hills's photograph taken atPulgaon in 1898. There is an abrupt fall in the intensity of the marked continuous spectrum at about 2£' from the limb, and at 3|' from the limb the spectrum is invisible. In the preliminary examination of the spectrum none of the ordinary Fraunhofer lines have been detected, a fact which is of remarkable import when considered in connection with the intensity of the polarisation of the light emitted from the corona. (See below, p. 364.) In one of the spectra the hydrogen lines are very strong, viz., Hp, Hy, Hj, Hc, and Hf. In the other they are barely visible, Hy appear- ing only very close to the limb, and not extending more than about a quarter of a minute of arc. The helium lines are strong in one and barely perceptible in the other. In one the calcium lines H and K are intensely strong and broadened, though the edges are defined and the lines very much shifted towards the red end of the spectrum ; in the other, the H and K lines are weak and well-defined narrow lines. It is important to note that the shift of the broad calcium lines is in the direction that one would anticipate if pressure were the cause of the broadening and of the shift. Whilst it seems clear that the presence of the hydrogen, helium, and calcium lines in one and not in the other of the two regions of the corona whose spectra have l>een photographed is probably due to a prominence, this explanation is difficult to recon- cile with the signs of pressure above referred to. There are several bright coronal lines discernible in both spectra ; and in the neighbourhood of one of the lines, viz., that of wave- length 4231, there seem to be two dark lines, apparently the only absorption lines visible in the spectrum. III. Sprctrum of thf Sttn's Limit fi. — Immediately after the end of the exposure of the plate for the spectrum of the corona, the image of the corona was readjusted on the slit, under unexpected difficulties however on account of the faintness of the light. Mr. Wyles called " sixty-four " as I reached the platform again to make the exposures, and the exposures were made as follows : — Plate No. 7 at 65 i „ 8 „ 66 9 „ 67 10 „ 68 H „ 69 „ 12 „ 70 and I gave the signal to Mr. Henn for his last exposure at 71. Tvtnl Solar Eclipse of 1900 (Mai/ 28). 359 It was found later that the faintness of the light was caused by a dark glass in front of the eye-piece, which was used for viewing the image on the slit. This was needed in the first exposures, but should have been removed by turning the hinged glass aside. It is evident from the photographs that the image was improperly adjusted in consequence of the faintness of the light ; there is no impression on the plates. The results obtained with the four-prism spectroscope may be summarised as follows : Five photographs of the spectrum of the vapours near the sun's limb at a fixed point, and a photograph of the spectrum of the corona at two points widely separated near the sun's, limb. • § 25. T/u; Photographic Camera with Large Objedive Grating. Visual observations of the green coronal ring made at Pulgaon, India, 1898, January 22,* convinced me that the ring could have been photographed with the objective grating and telescope then used. Accordingly preparation was made to attempt a photograph with a large grating at Algiers. For this purpose, use .was made of a plane grating by Rowland, 14,438 lines to the inch on a ruled surface- 5 x 3^ inches, fitted on an axis in front of a telescope of focal length 68 inches and aperture 4 inches. The grating is a very brilliant one,, and is ruled on an unusually fine-grained piece of speculum metal. The object glass is an excellent one by Cooke and Sons. Both of these belong to the splendid spectroscopic installation arranged by the late Professor Piazzi Smyth, with the aid of contributions from the Government Grant. The installation is now set up at the Cambridge Observatory, having been put at my disposal for spectroscopic investi- gations by the Royal Society. I am thereby put under a great obli- gation to the Society, and I venture to take this opportunity of making acknowledgment of it. In the recent eclipse the sun was about as far to the north of the celestial equator as it was to the south in the Indian eclipse of 1898 ; accordingly the grating and telescope could be mounted in almost the same relative positions in Algiers as in India ; it was only necessary to- reverse the positions along the polar axis, and arrange that the tele- scope pointed towards the south pole instead of the north. Accord- ingly the instruments were mounted so that the telescope was parallel to the earth's axis and pointed downwards towards the south pole. For the purposes of taking photographs this position was extremely convenient. A strong wooden bridge or yoke was fitted to the object glass end of the tube of the telescope, and projected in front of the object glass * ' P.oy. Soc. Proc-.,' vol. 61, p. 58 ; and ' Mon. Not., R.A.S.,' vol. 58, App., p. (58). 2 D 2 360 Prof. II. II. Turner and Mr. H. K X.-wnll. at such a distance from it that tin- grating could U> mounted free to turn on a spindle passing through the sides of the yoke at right angles to the collimation axn. A >Tnall brass cup or socket was attached to the middle point of the yoke so that it lay in the axis of collimation, and it was made the lower 1 tea ring, by which the whole instrument was supported on a pointed pivot, fixed, with a small amount of freedom for adjustment, on a low pillar of Im't kwork. The upper end of the tul»e of the telescope rested on antifriction rollers, supported on the west side of the higher pillar of 1 trick work, which also carried the four-prism spectroscope. Thus the polar axes of the two instruments, viz., the objective-grating camera and the mounting of the four-prism spectroscope, were side l>y side ; and it was not a difficult matter to link together the two mountings by means of a connecting rod, so that the same clockwork should drive both. Each mounting was connected by slow motions with the one clock-driven sector, and so each could be adjusted relatively to the sun without disturbing the other. The arrangement worked admirably. The light of the corona was incident on the grating at an angle of about 55°, and the diffracted beam utilised in the telescope left the grating at an angle of about 13° 40'. In this position of the grating the green of the second order was used and the magnifying power of the grating was a little greater than one-half, so that the coronal ring was distorted into an ellipse, in which the major axis was perpendicular to the length of the spectrum and parallel to the direction of daily motion. The axis of the instrument having been adjusted to parallelism with the earth's axis, it remained only (i) to set the grating so that the coronal ring should appeal1 in the middle of the field, and (ii) to focus the instrument. Neither of these operations could be done satisfactorily before the eclipse, that is, before the diminishing crescent of the sun made it possible to recognise the exact position of the spectrum in the field of view. Ten minutes before totality the dark lines were indistinctly visible in the spectrum, and a glance showed me that I had had an extraordinary stroke of good fortune in the rough setting of the grating, an operation which had been done by turning the grating till I thought the colour of the green was about right for the background of the magnesium lines. For the lines were only slightly displaced from the centre of the field, and the adjust- ment for the part of the spectrum required in the photograph was practically correct to a nicety. Accordingly no further adjustment was attempted. Two minutes before the beginning of totality the crescent was fine enough to show the dark lines in the spectrum very distinctly, a somewhat bewildering array of interlacing elliptical crescents, and the focussing was accomplished with ease. Mr. Henn then took charge of the instrument, and put a dark slide in position, Total Solar of 1900 (May 28). 361 and adjusted the exposing shutter. I am very much indebted to him for his admirable precision in carrying out the programme of ex- posures. The programme was carried out as follows : — Three plates were exposed. Plate X, 1. For the brightest chromospheric lines, at the beginning of totality — a short exposure, about 1^ seconds. This plate was to be exposed at the signal "Start," given by Professor Turner, and was to be closed between the time- keeper's calls " one " and " two." It was actually exposed at the signal " Start," and closed at the time-keeper's call " three." The time-keeper found that the first beat of the metronome after " Start " came so soon that he did not call "one," but called the next beat "two "without calling " one " at all. The exposure was thus probably 2£ seconds. Plate X, 2. For the green coronal ring — a long exposure, about 40 seconds. This plate was to be exposed as soon after Plate X, 1 as the change of plate holders would allow, and was to be closed at the call "fifty-five." It was actually exposed at " nine " and closed at " fifty-five," and thus had an exposure of about 46 beats. Plate X, 3. For the Fraunhofer lines immediately after the end of totality for comparison with any chromospheric lines that might appear on Plate X, .1. This plate was to be exposed when I gave the signal " Now," and was to be closed 1 second later. It was actually exposed at " seventy-one," and closed at " seventy-two." Results. — The Plates X, 1 and X, 2 show faint images, but have not been examined carefully yet ; a cursory examination shows that («) only a single chromospheric line appears on X, 1 ; and (6) continuous spectrum appears on X, 2, but no marked coronal ring is discernible. Plate X, 3 is a strong spectrum, showing the curved Fraunhofer lines between wave-lengths 5050 and 5460 ; the linear dispersion on the plate is, roughly speaking, 5 tenth-metres per millimetre. Remarks. — In an eclipse with longer duration of totality, the pro- cedure here described should give good results for the green coronal ring. The plates used were Edwards's Isochromatic Snapshot plates. It should be remembered that the effective aperture of the camera, viz., a little less than F/17, was rather dangerously small. .".OH IW. If. IT. Tunu-rau.l Mi. II. F. NYwall. 8 2G. Tin /'/;/"/•/.*•///'/> I'lnni-m (S'irtii-t /'/////•.-• "W .\Y' The glimpses of the corona that I was fortunate enough to get in India in 1898 through a small Savart polariscope convince' I me that that instrument, if properly used, would give just the information that is wanted to decide some of the perplexing points that still survive in the spectroscopic and polariscopic study of the corona. The chief objection is that the phenomena are far too complicated to study by eye observations in the short time at one's disposal in an eclipse. Here is a case in which photographic methods should certainly U> adopted if possible. Shortly after my return from India in 1898 I made some experiments to test the feasibility of photographing Savart's bands, and met with such promising success that no doubt was left in my mind that a photographic record of the distribution of Savart bands over the corona would give good results in supplement- ing the work which Professor Turner has in recent eclipses 1>een carrying out in studying the polariscopic phenomena of the corona. Accordingly when I was asked to take part in the observations of the eclipse at Algiers, it seemed well to make arrangements to prepare a polariscopic camera. Professor J. J. Thomson very kindly put two large Nicol prisms at my disposal. The aperture of these prisms is 1£ inches. It was thus possible to use a lens of a focal length of about 3 feet, if suitable Savart plates could be found. On making the necessary calculations I found that the plates would have to be 15 mm. thick, cut in quartz at 45* to the axis of the crystal. A pair of such plates would give Iwinds of the desired closeness, viz., alxjut 10' apart, instead of the usual 1° or 1° 30'. Fortunately Mr. Hilger was able to cut a slab, from the sloping top of a quartz crystal that I had in my possession, large enough to make two plates, each 14 mm. thick, of circular section, and with a diameter of 39 mm. (1£ inches). The whole slab was worked and polished with plane parallel surfaces, so as to secure equality of thickness, and was then cut into two parts, which were combined in the usual manner. The figure shows diagrammatically the arrangement of the camera with the Savart plates and Nicol prism in front. The lens was a 3^-inch lens of focal length 40 inches. The aperture was reduced to \\ inches, or approximately F/27 for central pencils. The Savart plates were fixed to the Nicol prism so that the bands were parallel to the plane of polarisation of the light transmitted by the Nicol. The whole system was arranged so that it could be rotated on its axis into any desired position, and a pointer was provided so that the position could be read off a large circle. In discussing with Professor Turner the arrangements for the various items in the programme of observations to be carried out, he Total Solar Eclipse 0/1900 (May 28). 363 Pointer. Nicoi. m very kindly suggested that the parts of this apparatus should be put into one of the compartments of the " double tube " alongside of the other polariscopic apparatus which he had himself arranged. I fell in with this suggestion very gladly, and the parts were taken to Algiers to be fitted there. It required very careful arrangement to get the two lots of apparatus into the tube, but in the end it was successfully accomplished, and Professor Turner made the exposures for the Savart camera simultaneously with those for his own polariscopic and other cameras. The pictures obtained with the Savart camera are on the same plates with the pictures obtained with Professor Turner's double image polariscopic camera. The general procedure with the Savart camera was to be as follows : — The Savart and Nicol were to be rotated until the bands due to the plane polarisation of the sky in front of the corona were extinguished, and photographs of the corona were to be taken. But it was not possible to look through the camera itself in order to make the adjustment " to extinction," for this would have interrupted the exposures for all the other instruments in Professor Turner's charge. Accordingly a subsidiary Savart polariscope was provided, which I may call the visual Savart to distinguish it from the camera Savart. The visual Savart was set up in my hut, with pointer and graduated circle attached, and the zero and numbering of the scale were adjusted so that the readings corresponded with those of the camera Savart, account being taken of the fact that the sky was seen in the camera by reflection from the coalostat. The programme of exposures was as follows : — 1 second, 5 seconds, 20 seconds, 5 seconds, 1 second. The first two were made with an arbitrary setting of the Savart, and the setting chosen was approximately that which would correspond to extinction of bands due to vertical polarisation. Meanwhile I had determined the plane of polarisation of the sky in front of the corona by observations with the visual Savart, made immediately after the exposures with the four-prism spectroscope were so far completed that the long exposure for the corona spectrum was begun, viz., 6 seconds 864 lY"f. H. H. Turner and Mi. II. K. N-wall. after the signal " Start." Mr. Potter, standing by me, received the reading resulting from my observations, and carried it to Professor Turner's hut, and Major Kingsley Foster adjusted the large Savart to the corresponding reading, and the third exposure was begun. The camera Savart was left with the pointer at 10° for the rest of totality, no attempt being made to test the permanence in the position of the plane of polarisation of the sky as the total phase of the eclipse passed over. Results. — The resulting photographs show strong bands over the corona. A cursory examination discloses the following results : — No. 1. 1 second. Coronal extensions discernible as far as 10' or 11' from the limb. No atmospheric bands visible, but obvious bands over the corona. No. 2. 5 seconds. Coronal extensions as far as 35' from the limb. The planet Mercury appears on the plate. Atmospheric bands are visible, very faint, on the following side of the sun, extending 4° 40' from the limb, but are not visible on the preceding side near Mercury. No. 3. 20 seconds. Coronal extensions 63' in N/< streamer. „ „ „ „ 52' in S/> streamer. „ „ „ „ 70' in N/ streamer. Mercury very strong. Atmospheric bands visible to the edge of the plate on both sides. Strong bands over^the corona. No. 4. 5 seconds. Coronal extensions 35' from the limb. No atmospheric bands visible on either side. Nos. 5 and 6. Not examined. The existence of the image of Mercury on the plates will l>e of great value in determining orientation in the polariscopic phenomena as well as in the corona. The strong bands over the corona indicate that a considerable por- tion of the light is polarised. There are irregularities in the bands which seem likely to afford interesting study just in the way that was anticipated. The atmospheric bands faintly visible on the plates are almost certainly due to imperfect adjustment of the Savart to extinction, arising from zero errors, &c. ; they might be due to a change in the position of the plane of polarisation of the sky after the initial setting of the Savart. In any case they are very feeble, and it is clear that it would be well, if ever the experiments are repeated, to dm at imperfect adjustment, Total Solar Eclipse of 1900 (May 28). 365 so that the atmospheric bands may be in opposite phase — i.e., with black central band — to the coronal bands. By a very fortunate accident just such an imperfection has arisen in the case of the plate No. 3, for the bright bands on the corona fall on dark atmospheric bands. It might be that the curvature of Savart's bands, which theoretically exists, misleads one ; but a tolerably careful examination of the faint bands shows them to be sensibly straight in the limited field dealt with, and the antagonism of the bands leaves no possible doubt that the bands seen on the corona are due to the polari- sation of the corona. It is difficult to reconcile the marked polarisation evidenced in this investigation with the absence of Fraunhofer lines in the spectrum of the corona. Across the dark moon no atmospheric bands are discernible, and there appears to be no doubt that photographically the dark moon is darker than the sky. These are points that need explanation. An investigation of the real facts would be difficult, but none the less interesting ; for the idea suggested by much of the evidence along different lines is that some of the light which is usually attributed to the sky may come from beyond the moon. For instance, is a milky sky on a moonless night simply the result of starlight scattered by the processes producing scintillation, or are other causes at work 1 § 27. Atmospheric Polarisation. Preparations had been made that a systematic survey of the polarisa- tion of the sky should be undertaken during the eclipse, with a view to determining the plane of polarisation in various quarters of the sky, a more precise knowledge of the general distribution of polarisation being needed for the explanation of some of the anomalies that appear to have been observed with respect to the atmospheric effects in pre- vious eclipses. Nine Savart polariscopes were mounted in similar turning tubes, provided with pointers and graduated circles, and attached to wooden stands. The stands were arranged so that each carried two polari- scopes ; one pointed to the horizon, the other to a point 30° above the horizon. The four stands were fixed on the top of a tall box on the balcony of the equatorial coude which M. Trepied had kindly put at our disposal. The polariscopes were directed towards the four quarters of the sky, N.E., N.W., S.W., and S.E. During the eclipse the sun was at an altitude of 30°, and only a few degrees north of west ; thus the polariscopes were directed to points nearly symmetri- cally disposed with regard to the sun. All the Savart plates were fixed relatively to the Nicol prisms, so that the bands were parallel to the plane of polarisation of the light transmitted by the Nicols. 1'n.f. H. H. Turner an.l Mr. II. K. Xewall. The ninth Savart polariscope was mounted in a turning tube with pointer and circle complete, on a lioard which was screwed to an inclined block on the western doorpost of the hut which contained my spectroscopic apparatus. It was pointed towards the corona, and was in fact the visual Savart used l>y myself in the way described in the previous section (p. 363) for determining the position of the plane of polarisation of the light from the sky in the immediate neigh bourhood of the corona, so that the camera Savart could l>e adjusted accordingly. The polariscope had been left in position with the bands horizontal and the pointer at 90". Six seconds after the beginning of totality I left the spectroscope and looked through the polariscope. The eclipsed sun was slightly (perhaps 5°) to the north of the centre of the field of the Savart. The Iwnds were seen fairly strong over the whole field of view, the central band being black. The Savart was then turned counter-clockwise, until the bands were extinguished. The reading was found to be 9° on the scale arranged to correspond with that on the large Savart in Professor Turner's hut. This reading showed that the plane of polarisation of the sky in front of the corona was inclined at an angle 4" to the vertical read counter-clockwise from the vertex. (There is possibly a zero error ; it has not yet been deter- mined.) When the atmospheric bands were extinguished faint traces of bands were seen over the corona, but much less strong than in the Indian eclipse. I examined the polarisation of the sky in the zenith about 10 seconds after the end of totality, and found that the plane of polarisation passed through the sun. Mrs. Newall undertook the charge of the eight other polariscope*, which were arranged as described above, and her programme for the eclipse was to turn each instrument, so that the Savart bands dis- appeared, paying attention to the direction of turning as follows : — If the band system had a white central band the polariscope was to be turned clockwise. If the system had a black central band it was to be turned counter-clockwise. The instruments were then to be left untouched, and the positions of the pointers were to be written down at leisure after the eclipse. Mrs. Newall devoted herself very dili- gently to setting the instruments under very varied conditions on the days preceding the eclipse, and so expert did she become that she was able to make the necessary settings of all eight polariscopes in about 42 seconds. If the polarisation was weak, about 50 seconds were needed. In the following table, taken at random from her note- book, the figures in the upper line are the readings of the pointers of the various polariscopes when the settings were made in a leisurely manner ; those in the lower line are the readings when the settings were made " racing." " Hor." refers to the horizontal polariscope : " 30" " to that which points upwards : — Total Solar Eclipse of 1900 (May 28). :jG7 S.E. N.E. N.W. S.W. Hoi-. 30° Hor. 30° Hor. 30° Hor. 30° 106° 74 168° 178 170U 121 114° 136 Leisurely. 104 74 164 176 164 122 116 135 Bating. The following readings show that in a leisurely setting the observa- tion of extinction of the bands is satisfactory. 1900. May 24. Ten settings for extinction of bands in the middle of the field— 16°, 15°, 16°, 12°, 16°, 17°, 16°, 17°, 17°, 16°. Mean 158-8. It is thus clear that the " racing " settings give results of about the same order of accuracy as the leisurely ones. In the eclipse itself, the observations were made on the balcony of the equatorial coude, and unfortunately, on account of other noises, the signal, " Stand by," announcing that the beginning of totality was approaching, was not heard by Mrs. Newall, who was standing in the doorway of the balcony with a view of protecting her eyes from the sunlight till the last moment. Nearly half of the duration of totality had passed before she came into the open, and heard the twenty-eighth beat of the metronome being called. Going at once to the polariscopes she began to adjust them; she had set four of them " to extinction," and had nearly completed the setting of the fifth, when sunlight reappeared. With regard to the last observation, Mrs. Newall noted an interesting point. She had nearly completed the setting to extinc- tion when the bands suddenly became bright again, with black centre, and she turned the polariscope counter-clockwise, from somewhere near the reading 20°, and had nearly set again to extinction before realising that totality was over. The reading of the polariscope was then found to be 345°. The actual circle readings recorded immediately after the eclipse were as follows, and it was noted that the bands were very faint : — S.E. N.E. N.W. S.W. Hor. 30°. Hor. 30°. Hor. 30° 105°-0 94°-l 324°-8 340° ?20°± [345° after return of sunlight.] These require small corrections for the index errors, which can only be determined after the instruments return from Algiers, but it may be provisionally stated that the angles made by the plane of polarisa- tion with the vertical, read from the vertex clockwise, are as follows : — S.E. N.E. N.W. W. Hor. 30°. Hor. 30°. Hor. 30°. Over corona. 60° 49° 280° 295° ?305°± 356° [From my own observations.] 368 Prof. II. H. Turner and Mi. H. K. Comparison with the oksei -vatim^ -• -cured on other days at about the same time of day as the eclipse, viz., 4.30, may be summarised graphic-ally ;is in the accompanying figure : — Polarisation i.f the Sky. Algiers, 1900, May 26-28. Plane of Polarisation. During the Tota.L EcLipse of the Sun. •*• *• In Sunshine At same hour of the day. •* >• AUituc/e JO ' Horizon. E. Antisun. The results are of considerable interest in their bearing on the well- known peculiarities in the phenomena of the polarisation of the sky in the neighbourhood of " neutral points." It is a great satisfaction to be able to record these observations, for though they are incomplete, yet they were successfully carried out in spite of circumstances which would have upset many a practised observer ; and the regret is all the greater that Mrs. Xewall had to pay such a forfeit for her resolution, for she did not get more than a glimpse of the eclipse. Of the nine Savart polariscopes used in these observations, four were lent to me by the Council of the Royal Astronomical Society, and three by Professor Lewis, through Mr. A. Hutchinson, of Pem- broke College, Cambridge, who had arranged to come to Algiers, and had volunteered assistance in the observations recorded in Section 2-> of this Report, but was unfortunately prevented at the last momentv by illness, from coming. Total Solar Eclipse of 1900 (May 28). 369 § 28. General Observations. The general darkness during totality was about the same as in India. The dark moon did not appear so strikingly coal black in Algiers as it did in India. This is curious when considered in connection with the fact that the polarisation of the sky in front of the corona was much stronger in India than in Algiers. Round the limb the brightness appeared relatively much greater in Algiers than in India. The breadth of the bright ring was estimated as 2' to 3' ; the decrease in brightness along the radius was very abrupt at this distance from the limb ; at 4' or 5' from the limb a lower level of brightness was reached and thence outwards along the equatorial streamers the decrease in brightness was small up to points about 2° from the centre. My impression of the streamers, recalled from a very vivid memory of the picture in my mind, is that the double streamer on the preceding side of the sun certainly extended beyond Mercury, and there was a similar extension on the following side. The latter extended for some distance as a broad streamer with nearly parallel edges. I used a telescope of 3£ inches aperture and of 29 inches focal length for viewing the corona direct. The instrument was merely clamped to the walls of the hut. I was able to focus the instrument carefully, and devoted some moments to examining the corona imme- diately outside the large prominence in the Sp quadrant. The prominence appeared double, one side having the form of a tapering column projecting radially from the limb, and the other appearing in cloud-like floating forms, both parts being of a wonderful rose colour. In the corona I was disappointed to find no striking signs of arches over the prominence. The only fine structure visible in the corona were a few interlacing wisps crossing one another, presumably in the part where the two streamers on the preceding side of the sun crossed one another in diverging from one another. The shout that is stated to have risen from the Arabs in Algiers was heard by me as I exposed the plate for the spectrum of the corona, that is, as Mr. "Wyles called "six." Algiers lies just a little more than a mile in a direct line from the Observatory. It seems probable that the shout was uttered at the same instant as Professor Turner's signal " Start," and announced that the totality had begun at Algiers. 370 Mr. ,T. Kvi-i->li...l. " Solar Eclipse of May U8, 1900. Preliminary Report of the Expedition to the South Limit of Totality to obtain Photo- graphs of the Flash Spectrum in High Solar Latitudes." I'.y ,T. HvKitsHKit. Head at Joint Meeting of the Iloyal and 1 loyal Astronomical Societies, June 28, 1900. MS. received July 16, 1900. This expedition was one of those organised by the Joint Permanent Eclipse Committee of the Royal Society and the Royal Astronomical Society, funds being provided from a grant made hy the Government (Jrant Committee. The following were the principal objects which I had in view in arranging the expedition : — To obtain a long series of photographs of the chromosphere and flash spectrum, including regions of the sun's surface in mid-latitudes,, and near one of the poles. The photographs to be obtained with a long focus prismatic camera on a large scale, in order to be able to discriminate clearly between high levels and low levels in the chromosphere. The photographs to include as much as possible of the ultra-violet region of the spectrum, for the purpose of verifying the results obtained with a smaller instrument in 1898, and to give more accurate values, of the wave-lengths determined from those results. This report may be conveniently divided into the following four sections, viz. : — 1. Selection of observing station. 2. Instruments, methods of mounting, and general arrangement of camp. 3. Narrative of expedition, and observations made on the day of the eclipse. 4. Results. (1) SrltrJitw of Obterving Station. A consideration of the conditions under which the lowest layers of the chromosphere are presented during a total solar eclipse showed that a very great advantage would l>e gained by selecting a station situated near the limit of the zone of total eclipse, where the two internal contacts would be separated by a small angle on the sun 'a limb. At such a station the motion of the moon relative to the sun is in a direction approximating to parallelism with a tangent to the sun's limb at the points of internal contact, the result being that the exces- sively shallow layer giving rise to the so-ca'led " fln.sh spectrum " is. Solar Eclipse of May 28, 1900. 371 occulted by the moon comparatively slowly. Much more time is therefore available for taking a series of photographs than is the case at stations near the central line of the eclipse, where the moon's motion is at right angles to the layer, and opportunities for obtaining- photographs of the very lowest strata are reduced to a fraction of a second only at each internal contact. I decided therefore to choose some point situated well within the zone of total eclipse, but so far from the central line that the two- internal contacts would be separated by an angle of about 39° on the sun's limb. This would give a duration of totality equal to one-third that at the nearest point on the central line ; and the time available for photographing the flash spectrum would not be less than 30 seconds. At mid-eclipse the moon's limb would overlap the photosphere about 1", so that even at that time the flash spectrum layer would not be entirely hidden. Under these circumstances, also, one of the contacts would take place at, or very near to, one of the poles of the sun, the other being in latitude 51°. A succession of photographs taken during totality would therefore give a series of images of the flash spectrum ranging from solar latitude 51° to the pole. In selecting the most suitable station, dry ness of climate was con- sidered to be the most important factor for securing extension of the spectra in the ultra violet ; I therefore selected Algeria in preference to Spain, although the altitude of the sun would be less in the former country. The best position in Algeria for realising the greatest solar altitude was a point on the coast west of Algiers, and on the southern border of the eclipse track. This region was therefore decided upon at the outset, and in order to realise the favourable conditions mentioned above, two stations were selected provisionally beforehand, and for these Dr. Downing kindly computed for me the durations of totality according to the data used by the Nautical Almanac Office. The first station, near to the village of Zeralda, was found by him to have a duration of 45 seconds, i.e., more than one-half the central line duration. The other station, three miles further south, and near to- Maelma, was computed to have a duration of 29 '5 seconds with a possible error of ± 10 seconds. As the required conditions would, apparently, be very nearly fulfilled at the latter station, I decided to place my camp either at that precise spot, or at some point situated on a line passing through it, and parallel to the direction of the shadow track in that region. The actual station eventually chosen was 6 '5 kilometres distant from the station near Maelma, in a direction bearing West 25° North. Here it was estimated that totality would last 30 seconds. Unfortunately, as the event proved, the value of the diameter of the moon adopted by :V72 .Mr. .1. Kv.-r.slie«l. Nautical Almanac Office is too large, and the limits of error given wriv very misleading. Instead of a duration of 30 seconds, the < •< -lipse at my station was never quite total. (:>) Iii.xfriniH-iil.t, ,1/,/Wx i if Muiinfiii'i, mi'/ General Arrangement of ('" It was my intention originally to take out a fine 18-inch silver-on- glass concave mirror made l>y the brothers Henri, which was given to me by the late Mr. Kanyard. This mirror, having a focal length of 117 inches, would have given images on a scale of 1*08 inch to the sun's diameter Many experiments were made with this mirror to determine the amount of aberration produced on star images at considerable distances from the axis, and with various apertures. It was found that when the ratio of aperture to focal length did not exceed 1/15, good images were obtained 4° from the normal axis, the aberration being very slight. As this would admit of a very wide range of spectrum being photo- graphed with good definition throughout, I decided to adapt my large reflecting telescope for eclipse work. Owing, however, to the difficulty •of obtaining a prism of large angle and not less than 6 inches aperture, I had, most unfortunately, to abandon this scheme and construct a much smaller apparatus. Through the kindness of Dr. Rambaut I eventually obtained a fine 4-inch prism of light flint glass and 45 degrees angle. This prism, which was generously placed at my disposal by Sir Howard Grubb, proved most efficient for the work, although I was unable to utilise the full aperture. Three spectrographs were finally made : a reflecting prismatic camera of 3 inches aperture and 74 inches focus, an ordinary prismatic camera of 2 inches aperture and 47 inches focus, and a quartz prismatic camera of 1 inch aperture and 24 inches focus. These were mounted together inside an observing hut, and were supplied with light from a 12-inch ccelostat. The Reflect i it ij P/'/.-niiiitif Camera. This was an ordinary reflecting telescope with a mirror of 9 inches aperture and 74 inches focus. It was fitted with a strong wooden tube, adapted for carrying two large prisms near the upper end. The prisms used were the 4-inch 45-degree prism, lent me by Sir Howard Grubb, and a 3-inch 60-degree prism lent me by Dr. Common. These were mounted eccentrically within the tube, in such a manner that the incident light, after passing through the prisms, made an angle of about 1£° with the normal axis of the mirror. After reflection from the mirror the rays returned over the upper surface of the 60-degree prism, and came to focus about an inch outside the end of the tube. Solar Eclipse of May 28, 1900. 373 The end of the tube was closed by a block of wood having an aperture 8 inches long by 3 inches wide, a little above the middle. Outside this a long slide was arranged, at right angles to the telescope, and bolted at the upper end to two stay rods attached to the telescope near the mirror. A plate holder, 3 feet long by 10 inches wide, taking two plates 8| inches square and four plates 8| by 4| inches, was arranged to move along the slide by means of rackwork and a pinion wheel. One revolution of the pinion moved the plates 2'13 inches, whereby four images could be obtained on the square plates and two on each of the narrow ones ; the sixteen images all being equal distances apart and symmetrically placed on the plates. The revolutions of the pinion wheel were controlled by a spring catch acting on the crank handle, and holding it firmly in position after each revolution. The whole slide, carrying the plate holder, &c., was attached to the telescope in such a way that the distance of the plates from the mirror could be varied a small amount for focussing. The tube of the instrument was firmly bolted down to the sloping side of a solid pier of stone and cement, built up within the observing hut near the north end. It was adjusted so that the plane of disper- sion of the prisms was in a meridian passing through the ccelostat, and inclined to the prime meridian 68° (the hour angle of the sun at mid-eclipse). The dispersion was therefore in a north and south direc- tion. The internal contacts were computed to occur near to the south point of the sun, and on either side of it. The centre of the flash spectrum arcs was therefore midway between the edges of the spectnim in the photographs obtained at mid-eclipse. The 2-inch Prismatic Camera. This instrument was the same which I employed successfully at the Indian eclipse in 1898, excepting that it was fitted with a specially corrected lens of 47 inches focus instead of the visual objective pre- viously used. The images were therefore on a somewhat larger scale, and larger plates were used. The sliding plate holder, constructed on the same lines as the larger instrument already described, was made to hold three plates, 6^ by 4£ inches, placed lengthwise in the holder ; and the crank handle moving the slide was arranged to stop at each half revolution, moving the plates 1'12 inches between each exposure. The two 60° prisms of this instrument are made of specially selected crown glass, and are exceptionally transparent for ultra-violet rays. The total deviation of the two prisms being approximately equal to that of the reflecting spectrograph (about 80°) the tubes of the two in- VOL. LXVII. 2 E 374 Mr. -T. fitruments were arranged nearly parallel, the 2-inch spectrograph l>eing screwed to the side of the reflector with its aperture alongside that of the latter. The camera end with the sliding plate holder was at the lower end. This was rigged up while in camp, as it was found that a small portion of the coelostat mirror was available to supply light. It con- sists of two double quartz prisms of 60° and 40° angle respectively, each prism having 1^-inch square faces; and a single quartz lens of 24 inches focus. It was screwed on the top of the 2-inch spectrograph, with its apertiire just within the elliptical beam of light from the coelostat. My brother arranged a very convenient exposing shutter, which he was to open near mid-eclipse for a single exposure of 10 seconds. Mrt]i»e poverty- stricken and unpromising. The mayor, whom we found in his mairie busy with the coming elections, was obliging enough to nV our document from the Prefecture ; or rather he got his secretary to do so, being unable himself, apparently, to read or write. Having had our letter duly risAl we abandoned Maelma, and proceeded to the Mazafran River, near the coast, to conclude negotiations already entered into with Alvado. These presented no difficulty, for M. Alvado was " un homme tres brave," and offered us his whole territory, vineyards or cornfields, for our camping-ground. We were, however, limited in our choice to a line bearing West and 24£ degrees North from Maelma, in order to secure the same duration of totality as had been computed for that place. The position finally chosen was near to the mouth of the Mazafran River on the east side, and about 1 kilometre from the sea. The position of the Mazafran bridge, about 400 metres distant, was. ascertained from a recent survey to be North latitude 36° 41' 35" East longitude 2 48 30 The position of the camp, which my brother carefully determined by triangulation from the bridge, was as follows : — North latitude 36° 41' 47" East longitude 2 48 41 It was 17 metres above sea-level, and 6*5 kilometres from the station near Maelma, in a direction bearing West 25° North. As the direction of the shadow track in this region was ascertained to be 24° 39' North of West, we concluded that the above position would be safe for a duration of 30 seconds of totality. Having settled all preliminaries we returned to Algiers to arrange for the transport of the instruments. This was effected without difficulty by means of the light railway recently constructed from Algiers to the Mazafran. On May 9th we returned to Alvado's farm, and the next day the work of erection was begun. Solar Eclipse of May 28, 1900. :.7.) Our hut, which was to serve as sleeping and living room as well as observatory, we had ready for occupation the same evening. During the fortnight preceding the eclipse, our time was fully occupied in erecting and adjusting the three spectrographs, ccelostat, and other instruments, and in taking trial photographs for determining focus. We also made daily observations of the sun with a sextant and artificial horizon for determining time, and checking the rate of a chronometer which we had hired for use in our camp. Being far from any telegraph station in direct communication with Algiers, we were obliged to depend entirely on observation for our time on the day of the eclipse. Working with instruments of very second-rate quality, my brother usually succeeded in determining local time within one or two seconds of error, taking the mean of a day's set of obser- vations (usually employing the method of double altitudes). During the whole time we were in camp, we were ably assisted by our host, M. Alvado, who took a most intelligent interest in all our operations, and was ever ready and at hand to help us in any and every difficulty with which we were confronted. We take this oppor- tunity of expressing our high appreciation of his services, and esteem for his character, and that of his wife, Madame Alvado. The latter attended most assiduously to all our personal wants, and in this way furthered most materially the objects of the expedition. Observations made on tlie Day of tlie Eclipse. Between 6 and 7 A.M. on May 28th I observed in the spectroscope the position angles and approximate heights of all the prominences then visible on the sun's limb. The results were then written out in accordance with a previously arranged code, and sent on to Zeralda to be telegraphed to Mr. A. C. D. Crommelin, at Algiers, for the use of intending observers of the coronal structure near to prominences. The following table gives the position angles and heights ob- served : — Position angle. Solar latitude. Approximate height Ho. 64° + 9° 50" 119 -46 15 217 -36 115/130 236 - 17 25 305 +52 20 The rest of the morning was devoted to final adjustments and rehearsals ; cleaning all lenses and prisms, and in taking more photo- graphs for focus in the 2-inch spectrograph. Soon after noon all slides were filled ready for the eclipse ; and lastly, the 9-inch mirror and the ccelostat mirror were both dusted and carefully polished with rouge to remove all trace of tarnish. 3ftO Mr. .1. Kv«-rsln-d. Fifteen minutes l>efore miil-rdipse the large spectrograph was slightly readj listed for focus by observing with ;i lens the spectrum image of the diminishing crescent. This was effected without any difficulty or uncertainty. I then attempted to focus the 2-inch spectrograph in the same way, using the Fraunhofer lines near G, which were then rapidly becoming sharply defined. As the last determination made photographically appeared to l>e correct, I set it again to the same position. Five minutes before mid-eclipse my brother wound up the coelostat clock, and three minutes later I gave the order " Stand by." The light waned rapidly, and I began the exposures at 4h 16m 58*. At 4'1 17'" 30" I found it difficult to see the seconds hand of the chrono- meter, and a few seconds later I opened for the 10-second exposure, giving at the same time the signal to expose the quartz spectrograph. The absence of any sound from the shutter warned us that the latter had failed to act. At 4h 18m I could again see the chronometer face clearly. I con- tinued the exposures according to the programme, finishing the last at 4h 18°» 18". A minute or two later, after removing all the plate holders from their slides, I observed the large prominences on the south-west limb in the spectroscope attached to the 3-inch telescope. They appeared, of course, exceedingly brilliant in the line Ha. Unfortunately, I was unable to make a critical examination of the spectrum, for at this time a crowd of sight-seers inundated the entire camp, and further observation for the time being was impossible. Later, 1 observed the time of last contact with the spectroscope. This took place at 5h 21 m 34« per clock. + 58 assumed error of clock. 5 22 32 G.M.T. At this moment the moon's limb was seen as a black line projected on the chromosphere. (4) JfaHlt*. Notwithstanding the fact that my station was outside the zone of total eclipse,* the photographs show that there was quite half a minute * From the descriptions given us immediately after the eclipse by M. .Alrudo and others who undertook to determine accurately the duration of totality, it appeared certain that the photosphere never wholly disappeared, a small point of sunlight remaining visible at the moment of mid-eclipse. The edge of the moon's shadow was, moreover, clearly seen traversing the sea and the sand dunes a short distance j.ort)i of our camp, which escaped the shadow by a few hundred metres only. Solar Eclipse of May 28, 1900. 381 available for obtaining good images of the flash spectrum. No. 9 spectrum, for instance, is one of the finest of the series, and shows about as many bright lines as the mid-eclipse photograph, yet it was exposed 15 seconds before mid-eclipse. Several other photographs taken earlier than No. 9 also show a large number of flash spectrum lines. I think this result demonstrates the very great advantage gained at stations near the limit of total eclipse for studying this spectrum. In cleaning the lens of the quartz spectrograph shortly before the eclipse, I unfortunately jammed the exposing shutter in such a way that it would not work at the critical time, and no photograph was obtained with this instrument. Sixteen photographs were obtained with the 2-inch spectrograph, .and sixteen with the reflecting spectrograph. The following table gives the approximate times of exposure, and the plates used in each instrument : — Approximate times. Plates used. Exposure .No. Beginning. Duration. Reflecting spectrograph. 2-inch spectrograph. h. m. s. sec. 1 4 16 58 1 Sandell Triple Sandell Perfect. 2 17 5 1 »> » » » 3 10 2 Sandell Perfect » » 4 14 2- 51 >• » » 5 18 2 Edwards's Ordinary » » Medium 0 23 2 » » Imperial Ordinary (Backed) 7 27 2 > ) 8 31 2 > * 9 35 2 > > 10 40 2 > > 11 45 10 , j 12 18 0 2 Sandell Triple. 13 4 2 Sandell Perfect )> j> 14 9 2 >! )) » >» 15 13 * Sandell Triple » » 16 17 i » »» >j M The images obtained with the 2-inch spectrograph are not in good ifocus. They are very dense in the region near G, but correctly ex- posed in the ultra-violet. The spectra extend from X 3350 to A 5100. Apparently the maladjustment of focus has produced a linear distor- tion of the images ; and at the edges of several of the spectra, where '.the direction of the distortion coincides with the direction of the bright Ml. .1. K lines, the focus appears to be quite perfect through the whole length of the spectrum. This suggests that the lens, which was a thin one, was under some strain in its cell, and it accounts for the difficulty experienced in find- ing the true focus. In the mid-eclipse photograph (No. 11) the bright lines are fairly well defined at the extreme end of the spectrum, and they can l»c traced in this photograph to A. 3320. All the lines between A. 3340 and A. 3500 can be identified with those shown on the best plate ob- tained in 1898. The following table gives the wave-lengths and identifications of these lines as determined for the spectrum obtained in India. In identifying the lines with the elements given in column 4, I received great assistance from Mr. L. E. Jewell, who also supplied me with a revised list of wave-length values for the solar lines given in column 5. The intensities (column 2) are estimated as follows : — Lines just visible but extremely faint The strongest lines in the spectrum . = 0 = 10 Wave-length (flash spectrum) . Intensity. Character. Element. Wave-length, Rowland (solar spectrum). 3326 ± 3330 ± 3333 ± 3335 ± 3340-0 33 42 '3 1 \ i i ik 2 Wave-lengths roughly esti- mated on photograph No. 11 of 1900. First, line on photograph No 3 of 1898 Cr ? Ti Ti 3340-490 4-' -012 3347 ' 0 o Ti 4<>-882 3349 -4 4 Ti 49-558 3354-0 2 \ Sc 53-875 3358-5 3361 -4 2 S 31 Cr Ti 58-649 61 -327 3368 "3 3 Cr 68 -193 3373 '0 4 Ti 72-948 3380'4 3 Short Ti 80'4^ I 338-4 -0 4 Ti 83 '892 3388-1 3 Ti 87 -9H8 3392 '1 U — Zr? 92-109 339-4 -7 3 Ti 94-716 3390-3 1 Short 3403*45 3 Cr 3403 404 3-405 -17 1 Co? 05-217 3407-32 1 Fe? 07-597 3-408-97 3 Cr OS'911 :'410 -24 ::> -94 3* / Cr 22-892 Solar Edii*e of May 28, 1900. 583 Wave-length (flash spectrum). Intensity. Character. Element. Wave-length, Rowland (solar spectrum). 3125 -4tf 3 126 -97 3428-73 3430 -61 0 "1 0 I 0 1 One measure only ; very short, Short Zr? 343 ) -671 3433 -54 i 4 Long Cim Ni [ 33 -453 , i 3438 -40 ' 2 Short Zr 9 1 33 -71:~> 33-376 3440 -93 ! 3 Faintly extended Fe Fe f 40-7H2 3442 -24 4 Long . . \ 41 -155 42 '112 3444-33 3 Faintly extended Ti 44 '467 3446 34 i 2 1 Ill-defined . . r \ 46-406 3452 -80 3456 "55 ; ? 1 M Ill-defined ; short . . 53 -039 56-528 3458 -58 i 3460-53 i I 3 Ni Long Mn 58-601 60 -460 3461 '68 2 Faintly extended Ti Ni / 61 -633 3463 -05 3464 -32 3465 -87 1 1 2 Interrupted ; very short .... Co Visible on continuous spec- Sr ? truin only. Very faintly extended. ... Co Fe \ 61 -801 62 -950 64 -6U8 f 65-900 3467 '46 j 3468 -72 ! 3471-33 3472 -58 0 0 1 I 1 Fe Equal pair, short, and in- terrupt ed Ni \ 66 -015 68 -821 72 '680 3474-28 3 74 -287 3475 • 67 2 Short Fe 75-594 3477 ' 26 ; 3 Lou" " Ti 77 -3->3 3479 -48 1* j 111 -defined- short ^r \ 79-531 3481 -20 3483 -08 3488 -85 3 1 3 I Mn 81 332 83 -047 88 -S17 3491 -16 3493 -22 3 J 2 Ti Ill-defined ' N>' Fe 91 -195 / 93-114 3494 -60 3496 -17 3497 -82 3499 -14 3500 -45 3502-30 1 3 2 0 1 0 Short. Long. Short. Interrupted. Short. (One measure only) Co \ 93 -618 3502 -394 3505 -06 3i 1 Ti 05 -036 3510 -96 3i J 10-98o In column 4 the predominating element in a group is put in italics. The sixteen images of the cusp and flash spectra obtained with the reflecting spectrograph are in good focus throughout.* Each spectrum * The very strong chromosphere arcs, such as II and K, show a faint coma on the more refrangible side. This has since been traced to slight irregular refraction at the base of the 60° prism. The fault is, however, too slight to appreciibly affect the definition of any but the strongest lines. 384 .S"/'?/- K'-iif* '.-/ .1% 28, 1900. is * inches long, ami r\tcnetween X 350 and X510. Good images of the flash spectrum are also impressed on photographs Nos. 10, 12, and 13. (1) In its main features the flash spectrum at the south pole of the sun is the same as in low latitudes. (2) No essential change is shown after an interval of four years ; the spectra photographed by Shackleton, in 1896, and those obtained in 1898 and 1900, all appear to be identical so far as it has been possible to compare them. (3) The flash spectrum, therefore, is probably as constant a feature of the solar surface as is the Fraunhofer spectrum. Observations of the Solar Eclipse of 1900 tiw.de in Spain. 385 With regard to instruments, the reflecting prismatic camera has proved to be a most efficient form of spectrograph for eclipse work. The uniform focus over the entire range of spectrum, and the facility with which the adjustment for focus can be effected, are advantages which those who have worked with prismatic cameras will appreciate. Another important advantage in the use of the reflector is the proximity of the exposing shutter to the plate holder, both of which can easily be controlled by one person. There is no signalling- between the man at the plates and the man at the shutter. There is again the advantage that there is no selective absorption of ultra-violet rays which occurs in lenses, and if the mirror is freshly polished there is no selective reflection for any of the rays which can be photographed. In concluding, I have to acknowledge my great indebtedness to my brother for his untiring devotion to the interests of the expedition throughout. In all the negotiations necessary on- arrival in the country he took a leading part, and was successful in obtaining the goodwill of every person with whom we came in contact. The fine series of photographs which we obtained bear witness to- his skill in carrying out, to the letter, the somewhat troublesome arrangements which I had planned for erecting and adjusting the- instruments. " Preliminary Note on Observations of the Total Solar Eclipse of 1900 May 28, made at Santa Pola (Casa del Pleito), Spain." By RALPH COPELAND, Ph.D., F.R.A.S., F.R.S.E.— Read at Joint Meeting of the Royal and Royal Astronomical Societies, June- 28, 1900. MS. received October 1, 1900. I had again the honour of being nominated one of the observers for the Joint Eclipse Committee, the station allotted to me being at Santa Pola, on the south-east coast of Spain. On the 9th May I left Edinburgh, and sailed from Tilbury on the llth in the Orient steamship " Oruba," accompanied by Mr. Thomas Heath, First Assistant at the Edinburgh Royal Observatory, who was going to Santa Pola to observe the eclipse on behalf of the Royal Society of Edinburgh. My instrumental outfit had preceded me under the care of Mr. James; McPherson, the experienced mechanician of our Edinburgh observatory. This outfit comprised the 40-foot horizontal telescope of 4-inch aperture previously used in India and Norway, together with a small Iceland spar and quartz prismatic camera, with an effective aperture of 1-8 inch. •386 Dr. K. <'n]K'l;unl. Observations of //« ,SVoiind for Santa Pola. Early on the 16th we reached Gibraltar, where we were met by another member of our Edinburgh party, Mr. Franklin-Adams, from Machrihanish, who had most thoughtfully arranged for the transfer of all our eclipse apparatus from the "Oruba" to H.M.S. "Theseus," which the Admiralty had generously placed at the disposal of the Joint Committee. We were most cordially welcomed on board the " Theseus " by Captain Tisdall, who introduced us to his officers, and assigned to us our most comfortable quarters. The few days spent on board the " Theseus " passed most pleasantly. With the greatest interest we followed the various forms of drill, and were greatly struck by the promptitude and precision with which every order was carried out. On landing at Santa Pola on the afternoon of the 17th we were received with the utmost courtesy by the Alcalde and other Spanish authorities, who at once assured us of all possible assistance in the furtherance of our work. The interchange of courtesies being over, we at once proceeded to the camp already laid out for Sir Xorman Lockyer's party. Abundant space had been left for the installation of our appa- ratus, but on closer examination of the ground we found the subsoil too light and sandy to afford the firm foundation required by our heavy instruments. We had therefore to select another site. This we found in the upper part of the town, in a barley-field, from which the crop had been gathered a few days before. Here the solid rock, covered only by a thin layer of soil, afforded an ideal foundation for all our apparatus, while the neighlxmring walls or houses protected the site from the prevailing winds without unduly obstructing the view. To the south-east of the selected spot stood a large barn, which chanced to l>e vacant in consequence of a law suit, and was therefore called " La Casa del Pleito." This barn was allotted to us by the ever- obliging Alcalde, and gave the name to our station. It served in the threefold capacity of a store-place for our empty boxes, a photographic lalwratory, and a most welcome retreat from the burning rays of the noonday sun. While our instruments were being landed and carted up on the morning of the 18th, we commenced laying out and preparing the necessary foundations for them. In this, as in all our work, we were most efficiently helped by a detachment of junior officers and men from the " Theseus." For the first few days there was a good deal of cloud, by night as well as by day, and it was only with difficulty that the exact observa- tions requisite for setting up the 40-foot were secured. Saturday, the 19th, was a red-letter day for us, as well as for our countrymen throughout the world. With his usual thoughtful care, of 1900, May 28, made at Santa Pola, Spain. 387 Mr. Franklin-Adams, before leaving England, had arranged that a concise daily telegram should 1)6 sent to him giving the latest war news. These telegrams were at once communicated to both camps as well as to the " Theseus," and it is needless to say with what keen interest they were received and discussed. We were preparing for lunch at the comfortable little restaurant where we lodged when the telegram announcing the relief of Mafeking was received. Immediately we all rushed into the entrance hall, where we gave three hearty British cheers, greatly to the astonishment of our Spanish friends, who were quite at a loss to understand what all the cheering and excitement meant. By Monday the 21st all our heavier concrete foundations were finished, and we had a clear week in which to adjust and test our appliances. The weather had also become much clearer, particularly in the afternoon, when it was important to check the final adjustments of the long telescope at the hour corresponding to that of the eclipse. Eventually all the adjustments were completed and tested by the 27th, on the afternoon of which day we had the satisfaction of seeing the sun's image traverse the plate-holder of the 40-foot precisely at the computed rate, and at the exact distance from the centre line corre- sponding to the sun's declination at the time. On the 26th we received a visit from the Civil Governor of the Province of Alicante, who was desirous of seeing our apparatus and satisfying himself that everything possible was being done for our comfort and convenience. On the same day a party of French astrono- mers came over from Elche to see our camp and compare notes. We much regretted that time did not permit our returning their friendly visit. Meanwhile, Mr. Heath and Mr. Franklin-Adams had erected the equatorial stands to carry their apparatus. Mr. Heath was provided with a 6-inch photo-visual telescope by T. Cooke and Sons, arranged to photograph the corona in the primary focal plane ; while Mr. Franklin- Adams' equipment consisted of a number of cameras, several of them of large aperture, designed for obtaining pictures of the coronal rays and the sun's surroundings generally. He had also several very amirate thermometers mounted on a suitable screen. The exact duties of each member of the camp were repeatedly rehearsed in accordance with the beats of a metronome, the indications of which were shouted out by a seaman on the plan devised by Sir Norman Lockyer for regulating the numerous operations at his camp. As most of the observers had already practised at their respective instruments, even the first general rehearsal went off much better than we could have expected. The whole credit of this is due to our naval assistants, who, from being trained to act promptly and in concert, readily appreciated the exact nature of the new duties entrusted to them. oftlte In the night all the plate-holders were duly filled and arrangeil in order. For about a week before the day of the eclipse the closely approxi mate time of Greenwich mean noon was signalled to us from the ship- This proved of the greatest value, as it relieved the camps of the necessity of making independent time determinations. From the moment at which the erection of the instruments was commenced four members of the " Guardia Civil " were told off by the Spanish authorities to watch over the safety of our gear. A single wire cord stretched round the area occupied by the instruments served as the line of demarcation, within which unauthorised persons were not allowed to come. On the day of the eclipse this line was thrmvn somewhat farther back at the suggestion of our Spanish friends. The weather on the momentous 28th was all that astronomers could desire. With the greatest care all our apparatus was revised. The end of our barn had been smoothed over with stucco so as to present a white expanse some 30 feet in width by 15 feet in height on which to observe the shadow bands. The azimuth of this wall was very exactly S. 40° W., and as mid-totality occurred in azimuth 92° 10', the position of the wall was very favourable for the observations in question. We also put up a white screen some 14 feet square, projecting at right angles to the northern end of the wall, and whitened the ground in the angle thus enclosed, thereby giving three planes on which we hoped the bands might be seen. Two officers of the " Theseus," who for some days previously had practised marking and recording imaginary shadow bands, were entrusted with the duty of recording the real bands as they appeared on the white surfaces. They were provided with brushes attached to long poles, and with pots of coloured wash — blue for the beginning of totality and red for the end. I undertook to observe the first contact with a small telescope of 2 inches aperture. This occurred 10*'4 before the computed time, but the discrepancy caused no surprise, as the moon's limb was very rough at the point of contact, and there was the chance that our chronometer-time might be out a second or two. In view of the very important work before us, no photographs of the partial phase were attempted. I have a note that at twelve minutes before totality the sky began to darken very rapidly, the darkness increasing more and more visibly during the last minutes l>efore the total phase. Five minutes before totality, at the word " Stations," everyone took up his assigned post. The large crowd of spectators who had collected during the last hour or two pressed closer in to the boundary wire — some of them still expressing their doubts as to whether the eclipse would really be total or not Eighty-three seconds before the computed time of second contact I of 1900, May 28, made at Santa Pola, Spain. 389 gave the signal " Start the clock " to McPherson, who was in the dark room of the 40-foot. At this moment Mr. Franklin- Adams gave a few strokes on a large bell, and called out " Silencio ! " I must here say that this call was immediately obeyed in the most courteous way by the assembled crowds, who maintained a perfect silence until the important phase was over. One minute before totality, at the signal " Chronograph," McPherson registered the position of the moving plate-holder. Sixteen seconds before totality, when, according to Mr. Fowler's computation, the diminishing crescent should subtend an arc of 90°, I gave the -signal " Stand by ; " five seconds before totality, corresponding to 55° of a crescent, the signal " Ready " was called, and at the disappearance of the last glimpse of sunlight I gave the final signal of " Go ! " From this moment the sailor in charge of the metronome announced every fifth second during the first minute, and then every second until the seventy-fifth, when he called " Stop ! " One minute and twenty-four seconds after the signal, I gave to McPherson the final signal " Chronograph," which he again recorded on the moving plate-holder, assuring himself at the same time that it was still moving at the regular speed. While I was giving the earlier signals just mentioned, I noticed a very interesting feature in the diminishing crescent. When the luminous crescent was reduced to a mere line, an exceptionally brilliant bead of light became detached from the rest, continuing to shine like a bright star for perhaps four or five seconds, and probably disappearing nearly at the same time as the rest of the crescent. It was doubtless due either to the passage of the sunlight through a very deep valley on the moon's limb or to the inter- ruption of the crescent by a high range of lunar mountains. What- ever its origin it presented an extreme case of the well-known phe- nomenon of " Baily's Beads." What struck me most, both in the late eclipse and in that of 1898, was the sudden transition from the swiftly changing phenomena attendant on the disappearance of sunlight to the steady unvarying aspect of the corona. During the last few minutes of the partial phase all the phenomena are in a state of rapid change — the light decreases in a swift geometrical ratio, the last shred of the sun's limb disappears, the prominences burst into view, and all at once the corona stands before one fixed and relatively unchanging during the whole of totality. The corona, as seen with the naked eye, presented a striking resemblance to the pictures of the corona of 1878. Below was a broad double streamer, like the outspread tail of a dove, symmetrical to the sun's equator, while opposite to this was a single large .pointed streamer involved in a much fainter dove-tail symmetrical to the one below. The spectrum shown by an excellent direct-vision prism about mid- VOL. LXVII. 2 F Ih. II. < 'ojM>liin,.n. but there is no reason to suppose that it interfered seriously with the observations. The first contact occurred at 2h 6m 20' Lisbon Mean Time, and was observed by Mr. Christie on the ground glass of the coronagraph. The time of commencement of the total phase was not accurately noted ; the duration was observed (by means of a stop-watch) and found to be 84£ sees., during which the programme detailed below was carried out. After totality photographs were" taken for orientation. The fourth contact was at 4h 36"' 13" Lisbon Mean Time. There was a good deal of light during totality, the diminution of light being similar to that occurring during a heavy thunderstorm in summer. The temperature fell about 8° during the eclipse. During totality Mercury and Venus were seen, Mercury especially being very brilliant. The observers had not much opportunity of observing the attendant phenomena of the eclipse, and with the assist- ance which was kindly given them were only just able to provide adequately for the working of the instruments. II. Photographs of the Corona. The programme of observations was composed of two distinct parts: — (1.) Photographs of the corona on a large scale to show structural detail. (2.) Photographs on a smaller scale with rapid lenses to show the coronal streamers with the greatest possible extension. (1) Large-scale Photographs. (These were taken by Mr. W. H. M. Christie.) The instrument used for (1) was the Thompson photographic tele- scope, with object-glass of 9 inches aperture and 8 feet 6 inches focal length, belonging to the Royal Observatory, in combination with a concave telephoto lens by Dallmeyer, of 4 inches aperture and 16 inches focus, fitted as a secondary magnifier, to give an image of the Sun 4 inches in diameter, with a field (for full pencils) of 14 inches diameter. The total length of the coronagraph was 12 feet — the equivalent focal length being about 36 feet. A coelostat with 16-inch plane mirror (made by Dr. Common) was employed to reflect the rays into the coronagraph, which was mounted (on boxes filled with stones) so as to point to the mirror at an angle of depression of about 5", and at an azimuth of about 56° West of South for the day of the eclipse. The camera was furnished with five plate-holders to take 15 x 15 inch plates, or for the shorter exposures 12 x 10 inch plates in a carrier. The five slides for photographs of the corona during totality were Total Eclipse of the 8nn as observed at Ovar, Portugal. '!07 exposed us below, the exposures being given by the observer with the exposing shutter of the plate-holder, and the times noted by him, counting from the commencement of totality. Plate. Lantern 12 in. x 10 in. Empress 12 „ x 10 „ SandeU's Triple-coated 15 „ x 15 „ Special Rapid 15 „ x 15 „ Lantern 12 „ x 10 „ As soon as possible after totality a second plate was put in No. 5 plate-holder, and exposed twice on the sun for orientation (with driving clock stopped for 3 min. between), the exposure being as short as possible (i to | sec.), and the aperture reduced to 2 inches. " Abney squares" were put on Nos. 1, 3 (twice), and 5, after return home. No. 1 300s exposure at 5 ft. to standard candle. No. 3 ...5s and 30 „ 5 „ „ No. 5 300 „ 3 „ „ The plates were all developed after return home, hydroquinone dilute being used. Nos. 1 and 5 unfortunately blistered badly in develop- ment, especially No. 1, though every care was taken, the developer being at a temperature of 60°. It is to be remarked that other plates developed at the same time under precisely similar conditions were free from blistering. No. 2 is to a certain extent disfigured with spots on the plate, which, however, do not materially interfere with the coronal detail. No. 4 shows fine detail in the polar plumes and coronal streamers extending to nearly a diameter of the Sun from the limb. No. 3 shows nearly the same. No. 5 shows very fine detail in the prominences in the S.W. quad- rant, with gradation of brightness merging into the coronal structure close to the limb, thus showing a continuity between the two pheno- mena, and affording fresh evidence of the association between coronal streamers and prominences, which was indicated in the photographs of the 1898 eclipse. It should be mentioned that the coronagraph was carefully focussed in the same manner as for the eclipses of 1896 and 1898,* by means of the image of an object (gauze net) in the plane of the plate reflected from the plane mirror of the ccelostat. The focus was thus obtained with great accuracy after two or three trials, and it was found that the field was remarkably flat. * ' Monthly Notices R.A S.,' vol. 57, p. 105 ; ' Eoy. Soc. Proc.,' vol. 64, p. 8. 398 MCSMX W H. M. Christie ;in«l F. W. Pys.m. (2) Small-scale Plwtoyraphs to sliow Extension of th*1 t',,i,,i,eginning at 80 sees, after first contact, were given for the spectnim of the " flash " at second contact. The conditions under which the " flash " was photo- graphed, as determined by the circumstances of the eclipse at Ovar and Total Eclipse of the Sun as observed at Ovar, Portugal. 401 the instruments used, are shown in the accompanying diagram, which is enlarged about four times. O is the centre of the sun's disc ; AB is the bright arc as seen 10 sees, before totality, and is 70°. The centre of the arc, C, which is about 1° from the equator, is 16° above the point D where the arc is vertical and could be made to touch the slit. The slit, which is repre- sented by dotted lines, cut the bright arc between C and D, the hori- zontal distance between C and D being ^th of an inch. The time for the first exposure, viz., 10 sees, before totality, was given to the observers by the Astronomer Royal from the length of the rapidly diminishing arc as seen by him on the ground glass of the corono- graph. This time appears to have been given correct to about 1 sec. The position of the image on the slit was not changed for the spectrum of the corona, which was obtained near the point of second contact. For the " flash " at third contact the slow motion of the heliostat was used, making the sun's image travel in the direction OC of the diagram, the amount of the displacement being determined by watch- ing the sun in the attached theodolite. The position of the slit relatively to the bright arc is shown in the second diagram ; in this case the slit was not nearly tangential to the sun's limb. The photographic plates used were Ilford " Empress " for the first " flash " photograph with the flint spectroscope and for both the " flash " photographs with the quartz. An Ilford " Ordinary " was used for the second "flash" photograph with the flint. Cadett "Light- ning plates " were iised for both photographs of the corona spectrum. Sjicrtntm of the Sun'* Limb. — The series of spectra of the limb show a large number of lines, but they have not yet been examined in detail. With the flint spectroscope, a spectrum is obtained extending from F to K. This is good from F to h. With the quartz the spec- trum reaches from h to \ 3300, and is in good definition to about A 3450. The photographs taken with the quartz spectroscope at the VOL. LXVII. '2 c; 402 Total Eclipse of the Sun as observed at Orar, Portugal. beginning of totality are an interesting series. They show a long series of hydrogen lines (26 beginning at h), and a large number of iron and titanium lines. The difference in behaviour of these t\\<, metals is shown in a striking manner, the titanium lines, like the hydrogen lines, being bright in the whole series of photos, beginning 10 sees, before totality, while the iron lines are reversed in the earlier photographs. Titanium lines at wave-lengths 3685'30, 3761-46, and 3759'42 are specially bright. A reproduction is given of part of this series of photographs. (Plate 4). The Corona Spectra. — Reproductions of these spectra are given in the accompanying plate (Plate 5). With the flint spectroscope a con- tinuous spectrum is obtained from F to H. Eight bright lines are dis- tinctly shown stretching right across the continuous spectrum, ;m - f • Vi/ \V ^f Christie and Dyson. Roy. Soc. Proc., Vol. 67, PI. 3. ECLIPSE OF SUN. OVAR. 1900, MAY 28. Photograph of Corona, obtained with Dallmeyer rapid rectilinear lens of 4 inches aperture (enlarged 1% times from photograph No. 1). [The planet Mercury is shown on the western side of the photograph.] Photograph of Prominences in S.W. Quadrant (enlarged 2,\ times from photograph No. 5, taken with the large coronagraph). [The spot with cross rays on the right-hand side is a defect in the photographic plate.] Christie and Dyson. Roy. Soc. Proc., Vol. 67, PI. 4. I I an 5 w ~ II o -S Christie and Dyson. Roy. Soc. Proc., Vol. 67, PL 5. o o Oi s O Total Eclipse of the Sun as observed at Ovar, Portugal. 401 the instruments used, are shown in the accompanying diagram, which is enlarged about four times. 0 is the centre of the sun's disc ; AB is the bright arc as seen 10 sees, before totality, and is 70°. The centre of the arc, C, which is about 1° from the equator, is 16° above the point D where the arc is vertical and could be made to touch the slit. . The slit, which is repre- sented by dotted lines, cut the bright arc between C and D, the hori- zontal distance between C and D being ^V-h of an inch. The time for the first exposure, viz., 10 sees, before totality, was given to the observers by the Astronomer Royal from the length of the rapidly diminishing arc as seen by him on the ground glass of the corono- graph. This time appears to have been given correct to about 1 sec. The position of the image on the slit was not changed for the spectrum of the corona, which was obtained near the point of second contact. For the " flash " at third contact the slow motion of the heliostat was used, making the sun's image travel in the direction OC of the diagram, the amount of the displacement being determined by watch- ing the sun in the attached theodolite. The position of the slit relatively to the bright arc is shown in the second diagram ; in thi& case the slit was not nearly tangential to the sun's limb. The photographic plates used were Ilford " Empress " for the first " flash " photograph with the flint spectroscope and for both the " flash " photographs with the quartz. An Ilford " Ordinary " was- used for the second "flash" photograph with the flint. Cadett " Light- ning plates " were used for both photographs of the corona spectrum. Spectrum of the Sun's Limb. — The series of spectra of the limb show a large number of lines, but they have not yet been examined in detail. With the flint spectroscope, a spectrum is obtained extending from F to K. This is good from F to h. With the quartz the spec- trum reaches from h to A. 3300, and is in good definition to about A 3450. The photographs taken with the quartz spectroscope at the VOL. T.XVII. 2 G 402 Proceed i beginning of totality are an interesting series. They show a long series of hydrogen lines (26 beginning at A), and a large number of iron and titanium lines. The difference in behaviour of these two metals is shown in a striking manner, the titanium lines, like the hydrogen lines, being bright in the whole series of photos, beginning 10 sees, before totality, while the iron lines are reversed in the earlier photographs. Titanium lines at wave-lengths 3685-30, 3761 '46, and 3759 '42 are specially bright. A reproduction is given of part of this series of photographs. (Plate 23). The Corona Spectra. — Reproductions of these spectra are given in the accompanying plate (Plate 24). With the flint spectroscope a con- tinuous spectrum is obtained from F to H. Eight bright lines are dis- tinctly shown stretching right across the continuous spectrum, and several shorter lines in the densest part. The line 1474 K is not shown, probably because plates specially sensitive in the green were not used. The wave-lengths of the lines have not yet been determined. The positions of the corona lines are indicated on the plate and can be seen in the top band, though only faintly. With the quartz spectroscope a continuous spectrum is shown which can be faintly traced as far as A. 3600. Strong bright lines are shown at A 3987 and X 3801. November 22, 1900. The LORD LISTER, F.R.C.S., D.C.L., President, in the Chair. Mr. John Home was admitted into the Society. A List of the Presents received was laid on the table, and thanks ordered for them. His Grace the Duke of Northumberland, a meml>er of Her Majesty's Most Honourable Privy Council, was balloted for and elected a Fellow of the Society. In pursuance of the Statutes, notice of the ensuing Anniversary Meeting was given from the Chair, and the list of Officers and Council nominated for election was read as follows : — President.— Sir William Huggins, K.C.B., D.C.L., LL.D Treasurer. — Alfred Bray Ke npe, M.A. Further Note on the Spectrum of Silicium. 403 f Sir Michael Foster, K.C.B., D.C.L., LL.D. Secretaries. — < .rrmi» -A- T ™ A I Professor Arthur William Rucker, M.A., D.Sc. Foreign Secretary. — Thomas Edward Thorpe, C.B., Sc.D. Other Members of the Council. — Professor Henry Edward Armstrong, LL.D. ; Charles Vernon Boys ; Horace T. Brown, LL.D. ; William Henry Mahoney Christie, C.B. ; Professor Edwin Bailey Elliott, M.A. ; Hans Friedrich Gadow, Ph.D. ; Professor William Mitchinson Hicks, M.A. ; Lord Lister, F.R.C.S. ; Prof essor William Carmichael Mclntosh, F.L.S.; Ludwig Mond, Ph.D. ; Professor Arnold William Eeinold, M.A. ; Pro- fessor J. Emerson Reynolds, Sc.D. ; Robert Henry Scott, Sc.D. ; Pro- fessor Charles Scott Sherrington, M.D. ; J. J. H. Teall, M.A. ; Sir John Wolfe Barry, K.C.B. The following Papers were read : — I. " Further Note on the Spectrum of Silicium." By Sir J. NORMAN LOCKYER, K.C.B., F.R.S. II. " On Solar Changes of Temperature and Variations in Rainfall in the Region surrounding the Indian Ocean." By Sir J. NORMAN LOCKYER, K.C.B., F.R.S., and Dr. W. J. S. LOCKYER. III. u On the Restoration of Co-ordinated Movements after Nerve- crossing, with Interchange of Function of the Cerebral Cortical Centres." By Dr. ROBERT KENNEDY. Communicated by Professor McKENDRiCK, F.R.S. "Further Note on the Spectrum of Silicium." By Sir NORMAN LOCKYEK, K.C.B., F.R.S. Received October 26, — Read Novem- ber 22, 1900. In a previous note* I gave an account of some observations on the spectrum of silicium, and showed the relation which exists between the various groups of silicium lines and certain lines prominent in the spectra of some of the hottest stars. Further photographs have recently been obtained (with a 3-inch Cooke spectrograph) of the spectrum of silicium bromide in a capillary vacuum tube, and of the spark spectrum between two poles of metallic silicium which were kindly sent to me by Sir William Crookes. In each case the large Spottiswoode coil and plate condenser were used. The spectra extend from about A 3850 to D, and occupy a length of about 7 inches between those limits. Although all the silicium lines are * ' Hoy. Soo. Fror.,' vol. 65, p. 449. 2 G 2 404 Sir N'oiiuaii Lockyer. o X, . — «. i * s i J 0 i- •E 1~ O •SS V". i? « '3 o ft Li 1-5 . . . L-S »O rH — . -H '^r C 1 0 V V a is v v v 0 A » II a IH « « eo .. .-*-*... 2 H t 9 ll 4 t o *^ o O W t^» 00 ^1 O W •J £ 2 2 " ' I s* « I I I ^- — * O3 CO CO O5 ^ *4* •j S II S c . 4- M " '/ * •— — •* i "c * 11 u *• - -- COOi— iX^Tl^Oi^l XNb- i— I C t» OJ C5 CM • 3 H * IO »O lO CO 06 O C*l ?C • C^ O O • CM CC *O CO t** *-O H oojo^w n w oo •* . ^ V £- - ^< c iHtt-* o I-I*I^HCC --ocxioeo tro 4, 0. ». -. 02 » £, (—1 >• i 2 a . I1 I— i i— i i— I <— 1 i-H * lO «O CO p CC X C"i O ^" ^1 i*» "» ^O t'* *~ i^ OTIOOOO O ^"O-^ ^^— -^— ».Cl^iO ^5^ Plf » Further Nutc on the Spcctm:n of Silicium. 405 common to the two spectra, the relative intensities of the lines differ greatly. These later photographs show all the lines enumerated in my former communication, with the addition of several fainter lines and a strong double in the green, the latter falling outside the region previously investigated. Lines in the spark spectrum of silicium have been recorded by Eder and Valenta* and by Exner and Hasehekf. A comparison of these with the lines photographed at Kensington is given in the following table. Better photographs of the spectra of the type stars have been recently obtained, and we are now in a more satisfactory position to trace the various silicium lines through successive stages of stellar temperature. The lines in the spectra of a. Cygni, /? Orionis, y Orionis, and « Orionis which correspond with lines of silicium are indicated in the table by their intensities. In my former note the lines were divided into three sets, A, B, C, and the behaviour of the different sets in terrestrial and celestial spectra was described. The groupings were then made, roughly speaking, in order of wave-length. For my present purpose, it is important to divide them according to the conditions under which they become prominent lines ; this only involves the changing of the order of the groups, and involves no interchange of any of the lines from one group to another. Reference was also made previously to a line at A 3905-8 special to the arc, but as this is not an enhanced line it was not included in any of the sets A, B, C. Hence, I now add another group consisting of lines most prominent in the arc spectrum. The lines •constituting the various groups will then be as given below, arranged in order of ascending temperature, Group I being the lowest : — f 4089-1- Group IV < 4096 • 9 Set B of previous note. Ull6-4_! |- 4552 -8— Group III J 4568-0- Set C 1 4574- 9- f 3853 -9 3856-1— | ( 3862-7— , Group II ir Yunnan !."< 1 The lines in Group I, although appearing in the spark spectrum, are stronger in that of the arc, and therefore cannot he classed as enhanced lines. They are both given by Kowland in his " Table of Solar Wave- lengths "as being coincident with lines in the Fraunhofer spectrum, and may be considered as the lines of silicium which make their appearance at the lowest of the temperatures we are now considering. It will be seen that only the stronger line of the two is represented in the spectra of the stars included in the table, and that only in a Cygni, which has been placed lowest in order of ascending tempera- ture among those referred to from a previous investigation of lines in its spectrum other than those of silicium. It does not appear to exist at the higher stages of stellar temperature represented by ft, y, or e Orionis. The absence of the other line from the spectrum of a Cygni may be accounted for by its comparative weakness in the silicium arc spectrum. In a Cygni only the very strongest of the arc lines of iron, manganese, &c., are represented, ami then only as very weak lines. The lines in Group II are either absent from the most recent arc spectrum photographed at Kensington or exist there only as weak lines. The members of this group are prominent both in the vacuum tube spark and in the spark between poles of silicium, but are upon the whole more prominent in the latter spectrum. Considering the first five lines in the group, which are the only ones comparable with the Kensington records of stellar spectra, a glance at the table will show that they are at their maximum intensity at the stage of temperature represented by a Cygni, and decline in intensity as we pass to the higher successive stages represented by ft, y, and e Orionis. At the latter stage some of them have disappeared, and the others are on the verge of extinction. With regard to the remaining two lines of this group, those at XX 5042 and 5057, the position of which cannot be estimated more accurately than to the nearest tenth-metre on account of the diffuseness of the lines, it is extremely probable that if better photographs of that region of the spectra of a Cygni and Rigel were available, lines would be found corresponding to these silicium lines. Keeler has recorded* a line in the spectnun of Rigel at X 5056, and this is probably identical with the silicium line at X 5057, which is by far the stronger of the pair. The lines in Group III occur both in the vacuum tube spectrum of silicium bromide and in the spark spectrum. They appear as a well- marked triplet in the latter, but not nearly so prominently as in the former. They first make their appearance in stellar spectra in a Cygni, where, however, they can only just be traced. They are a little stronger in * • Ast. and Ast. Phjs.,' 1894, vol. 13, p. 489. Further Note on the Spectrum of Silicium. 407 /3 Orionis, and are most prominent in y and e Orionis, in which two spectra they are of about equal intensity. The lines in Group IV have never been seen in the spark spectrum of silicium when small coil and small jar capacity are used, but with the spark given by the Spottiswoode coil and plate condenser they appear as weak lines. They are not, like the members of Groups II and III, seen in the spectrum from the bulb when a vacuum tube is used, but in that given by the capillary the strongest ones are very prominent, and vie in intensity with the lines of Group III. None of them appear in stellar spectra until the level of temperature represented by y Orionis, and in the spectrum of that star only the strongest of the three is with certainty present. At the e Orionis stage, however, they have developed enormously in intensity, and are amongst the most prominent lines in the spectrum. The identity of some of the silicium lines — in particular those consti- tuting Group III — with lines in stellar spectra was subsequently but independently confirmed, and the results published,* by Mr. Lunt, Assistant at the Royal Observatory, Cape of Good Hope. The star the spectrum of which he chiefly considered was /3 Crucis, similar to that of y Orionis, the type star in the Kensington classifi- cation. The only enhanced line common to the Kensington and Exner and Haschek's lists, which does not appear to be represented in stellar spectra, is that at X 4030 -0. It is only a weak line in the spark spectrum, and may possibly be due to an impurity, though it has not yet been traced to any other origin. In the Kensington photograph it is a sharply-defined line, and unlike the other silicium lines in appearance. Exner and Haschek, however, record it as a very diffuse line. Of the four additional lines given by Exner and Haschek at XX 3883 '46, 4021 '0, 4103'7, and 4764-20, none appear in any of the Kensington- photographs, nor are they represented in the spectra of any of the stars included in the discussion. With these facts in view, it would appear extremely doubtful whether they are really due to silicium. In a former paper " On the Chemical Classification of the Stars,"! I gave the chemical definition of the various groups. At that time only the stronger lines of silicium included in Group II were known and traced through the stellar genera. We are now in a position to revise the chemical definitions, interpola- ting the various groups of silicium lines as they appear in the stellar groups. * ' Astropliys. Jour.,' vol. 11, p. 262. f ' ROT. Soc. Proc.,' vol. 65, p. 186. Note on the Spectra m <>f Silicium. DEFINITIONS OF STELLAR (}KM:::A. Argonian. proto-magne*ium, proto- Predominant. — Hydrogen and proto-hydrogen. Fainter.— Helium, unknown gas (\ 4151, -4157), calcium, asterium. Alnitamian. Predominant. — Hydrogen, helium, unknown gas (4G49'2), silicium (IV). Fainter. — Asterium, silicium (HI), proto-hydrogen, proto-niagnesiuni, proto- calciuni, oxygen, nitrogen, carbon, silicium (II). Crucian. Predominant. — Hydrogen, helium, asterium, oxygen, nitrogen, carbon. Fainter. — Proto-magnesium, proto- calcium, .-iiirium (III), unknown gas (\4649-2), silicium (II). silicium (IV). C Tauriau. Predominant. — Hydrogen, he- lium, proto-magnesiuin, asterium. Fainter. — Proto-calcium, sili- cium (II), proto-iron, proto-tita- nium, proto-chromium, nitrogen, carbon, oxygen. Biyelian. Predominant. — Hydrogen, pro to- calcium, proto-magnesium, helium, fcilicium (II). Fainter. — Asterium, proto-iron, carbon, nitrogen, proto-titaniumv proto-chromium, oxvgen, silicium (III). Cyonian. Predominant. — Hydrogen, proto- calcium, proto-magnesium, proto- iron, silicium (II), proto-tit aiiium, proto-chromium. Fainter. — Proto-nickel, silicium "i (I), proto-vanadium, proto-manga- nese, proto-strontium, iron (arc), helium, silicium (III), asterium. Acherniaii. Same as Crucian. Algolian. Predominant. — Hydrogen, proto- magnesium, proto-calcium, helium, silicium (II). Fainter. — Proto-iron, asterium, carbon, proto-titanium, pro'.o-man- ganese, proto-nickel. Markabian. Predominant. — Hydrogen, proto- calcium, proto magnesium, sill- cium (II). Fainter. — Proto-iron, helium, asterium, proto-titanium, proto- manganese, proto-nickcl, proto- chromiuin. Sirian. Predominant. — Hydrogen, proto- calcium, proto-magncsium, proto- iron, silicium (II). Fainter. — Tli3 lines of the other proto-metals and the arc lines of iron, calcium, manganese, silicium (I). Solar Changes of Temperature and Variations in Rainfall. 401) Polarian. Trocyonian. Predominant. — Pioto-calcium, Same as Polarian. proto-titanium, hydrogen, proto- magnesium, proto-iron, and arc lines of calcium, iron, manganese, silicumi (I). Fainter. — The other proto-metals and metals occurring in the Sirian v^ genus. Aldtlarlan. Arcturian. Predominant. — Proto-calcium, arc Same as Aldebarian. lines of iron, calcium, manganese, proto- atrontium, hydrogen, siiicium (I). Fainter. — Proto-iron and proto-tita- nium. Antarian. Piscian. Predominant.— Flutings of manga- ! Predominant. — Flutings of carbon, nese. Fainter. — Arc lines of metallic ele- Fainter. — Arc lines of metallic ele- ' ments. rnents. It will be seen that the conclusions arrived at in the former part of the paper as to the different conditions under which the different groups of siiicium lines become prominent verify the order in which the stars were placed on a scale of ascending temperatures. Thus those stars in which Group I occurs prominently are at the bottom, those in which .Groups II and III predominate occupy intermediate positions, and those in which the lines of Group IV are a special feature appear high up in the classification. The photographs of the siiicium spectra were taken by Mr. Butler. Their discussion has devolved upon Mr. Baxandall, who has also traced the siiicium lines through the stellar spectra, and assisted in the preparation of the paper. " On Solar Changes of Temperature and Variations in Rainfall in the Region surrounding the Indian Ocean." By Sir NORMAN LOCKYER, K.C.B., F.R.S.,and W. J. S. LOCKYER, M.A. (Canik), Ph.D. (Gott). Received October 2G— Read November 2i', 1900. The fact that the abnormal behaviour of the widened lines in the spectra of sunspots since 1894 had been accompanied by irregularities in the rainfall of India suggested the study and correlation of various series of facts which might be expected to throw light upon the subject. 410 Sir Xuniiiin Lockycr suitl I>r. NV . .1. S. LM< ; The conclusions already arrived at from bringing together the results of several investigations undertaken with this view may be stated as follows : — (1.) It has been found, from a discussion of the chemical origin of lines most widened in sunspots at maxima and minima periods, that there is a considerable rise above the mean temperature of the sun around the years of sunspot maximum and a considerable fall around the years of sunspot minimum. (2.) It has been found, from the actual facts of rainfall in India (during the S.W. monsoon) and Mauritius, between the years 1877 and 1 886,* as given by Blanford and Meldrum, that the effects of these solar changes are felt in India at sunspot maximum, and in Mauritius at sunspot minimum. Of these the greater is that produced in the Mauritius at sunspot minimum. The pulse at Mauritius at sunspot minimum is also felt in India, and gives rise generally to a secondary maximum in India. India, therefore, has two pulses of rainfall, one near the maximum and the other near the minimum of the sunspot period. (3.) It has been found that the dates of the beginning of these two pulses on the Indian and Mauritius rainfall are related to the sudden remarkable changes in the behaviour of the widened lines. (4.) It has been found, from a study of the Famine Commission reports, that all the famines therein recorded which have devastated India during the last half-century (we have not yet carried the investi- gation further back) have occurred in the intervals between these two pulses. (5.) It has been found, from the investigation of the changes in (1) the widened b'nes, (2) the rainfall of India, and (3) of the Mauritius during and after the last maximum in 1893, that important variations from those exhibited during and after the last maximum of 188S occurred in all three. It may be stated at the same time that the minimum of 1888-1889 resembled the preceding minimum of 1878-1879. ^6.) It has been found, from an investigation of the Nile curves between the years 1849 and 1878, that all the lowest Niles recorded have occurred between the same intervals. (7.) The relation of the intervals in question to the droughts of Australia and of Cape Colony, and to the variations in the rainfall of extra tropical regions generally, has not yet been investigated. We have found, however, a general agreement between the intervals and the rainfall of Scotland (Buchan), and have traced both pulses in the rainfalls of Cordoba (Davis) and the Cape of Good Hope. * This period waa selected becacse the Kensington observations of widened lines only began in 1879, and the collected rainfall of India has only been published to 1886. Solar Changes of Temperature and Variations in Rainfall. 411 (8.) We have had the opportunity of showing these results to the Meteorological Reporter to the Government of India and Director- General of Indian Observatories, John Eliot, Esq., C.I.E., F.R.S., who is now in England, and he allows us to state his opinion that they accord closely with all the known facts of the large abnormal features of the temperature, pressure, and rainfall in India during the last twenty-five years, and hence that the inductions already arrived at will be of great service in forecasting future droughts in India. Addendum. Eeceived November 16, 1900. Since Meldrum and one of us called attention, in 1872, to a possible connection between sunspots and rainfall, there has been a large litera- ture upon the subject which it is not necessary for us to analyse ; it may be simply stated that, in spite of the cogent evidence advanced since, chiefly by Meldrum, and in later years by Mr. Hutehins,* it is not yet generally accepted that a case for the connection has been made out. "What has been looked for has been a change at maximum sunspots only ; the idea being that there might be an effective change of solar temperature, either in excess or defect, at such times ; and that there would be a gradual and continuous variation from maximum to maximum. At the same time, it is possible that the pressure connection, first advanced by Chambers, is now accepted by meteorologists as a result of the recent work of Eliot. The coincidence, during the last few years, of an abnormal state of the sun with abnormal rain in India, accompanied by the worst famine experienced during the century, suggested to us the desirability of reconsidering the question, especially as we have now some new factors at our disposal. These have been revealed by the study, now extend- ing over twenty years, of the widened lines in sunspots, which sug- gested the view that two effects ought to be expected in a sunspot cycle instead of one. The Widened Lines. It will be gathered from previous communications to the Royal Society! that, on throwing the image of a sunspot on the slit of a spectroscope, it is found that the spectmm of a spot so examined indi- cates that the blackness of the spot is due not only to general but to selective absorption, \ and that the lines widened by the selective absorption vary from time to time. * ' Cycles of Drought and Q-ood Seasons in South Africa,' 1 889. t ' Boy. Soc. ?roc.,' TO!. 40, p. 347, 18S6; vol. 42, p. 37, 1887 ; vol. 46, p. 335, 1889 ; vol. 57, p. 199, 1894. I ' Eoy. Soc. Proc.,' Lockyer, October 11, 1866. 41 'J Sir Norman Lot-kyn ;uid Dr. W. .1. S. I...ckyer. Since the year 1879, the selective absorption in spots has been <>!.- served for every spot that was large enough to be spectroscopically examined, the method adopted being as follows : — The regions of the spectrum investigated lie between F — b and b — D, and an observation consists in ol>serving the six most widened lines in each of these regions. These lines are then identified on the best solar spectrum maps available and their wave-lengths determined. An examination of many years' records of these widened lines has shown that at some periods they are easily traceable to known elements, while at others their origins have not been discovered, so the latter have been classed as " unknown " lines. If we compare these two periods with the sunspot curve as constructed from the measurements of the mean spotted area for each year, it is found that when the spotted area is greatest the widened lines belong to the " unknown " class, while when the spotted area is least they belong to the " known " class. The majority of the lines traced to some terrestrial origin belong to iron, but the lines of other elements, such as titanium, nickel, vanadium, scandium, manganese, chromium, cobalt, &c., are also represented in a less degree. It is quite likely that some of the " unknown " lines are higher temperature (enhanced) lines of known chemical elements. In our laboratories we have means of differentiating between three stages of temperature, namely, the temperature of the flame, the electric arc, and the electric spark of the highest tension. At the lowest temperature, that of the flame, we get a certain set of lines ; a new set is seen as the temperature of the electric arc is reached. At the temperature of the high tension spark we again have man}* new lines, called enhanced lines, added, while many of the arc lines wane in intensity. It is found that at sunspot minimum, when the " known " lines are most numerous, the lines are almost invariably those seen most promi- nent in the arc. Passing from the sunspot minimum towards the maximum the " unknown " lines gradually obtain the predominance. As said before, they may be possibly " enhanced lines " — that is, lines indicating the action of a much higher temperature on /.//"//•» sul>- stances. Unfortunately the records of enhanced lines at South Kensington, having been obtained from photographs, are chiefly confined to a region of the spectrum not covered by the visual observations of widened lines in sunspot spectra. We can only point to the evidence acquired in the case of one metal — iron, for which photographs of the enhanced lines, in the green and yellow parts of the spectrum, have been obtained. This evidence quite justifies the above suggestion, for the enhanced Solar Changes of Temperature and Variations in Rainfall. 413 lines of iron can be seen revealing themselves as the number of un- known lines increases. We are, therefore, quite justified in assuming a very great increase of temperature at the sunspot maximum when the "unknown" lines appear alone. The curves of the "known" and " unknown " lines have been ob- tained by determining for each quarter of a year the percentage number of known and unknown lines and plotting these percentages as ordinates, and the time elements as abscissae. Instead of using the mean curves for all the known elements involved, that for iron is em ployed, as it is a good representative of " known " elements, and has been best studied. When such curves have been drawn they cross each other at points where the percentage of unknown lines is increasing, and that of the iron or known lines are diminishing, or vice versa. We seem, therefore, to be brought into the presence of three well- marked stages of solar temperature. When the curves of known and unknown lines cross each other, that is, when the number of known and unknown lines is about equal, we must assume a mean condition of solar temperature. When the un- known lines reach their maximum we have indicated to us a + pulse or condition of temperature. When the known lines reach their maxi- mum we have a - pulse or condition of temperature. The earliest discussion showed that, generally speaking, the un- known-lines curve varied directly, and the iron-lines curve varied inversely with the spot-area curve. The curves now obtained for the whole period of twenty years not only entirely endorse this conclusion, but enable more minute comparisons to be drawn. The " widened line " curves are quite different from those furnished by the sunspots. Ascents and descents are both equally sharp, changes are sudden, and the curves are relatively flat at top and bottom. The • crossings are sharply marked. During the period since 1879 three such crossings have occurred, in- ! dicating the presence of mean solar temperature conditions, in the I years 1881, 1886-7,* and 1892. It was expected that another crossing I with the known lines on the rise would have occurred in 1897, indi- I eating thereby the arrival of another mean condition of solar tempera- ture, but as yet no such crossing has taken place. The following tabular statement shows the years of those crossings, | together with the probable dates, in brackets, of the two previous I crossings, as determined by the time of occurrence of the preceding sun-spot maximum. * According to the observations tbe mean wa: reac-hed in December, 18S6, or January, 1887. 414 Sir Norman Lockyer and I>r. W. .1. S. Luckyer. Rise of Years. Unknown lines . . (1869) 1881 l^UU Known lines .... (1876) 1836-7 P Comparison of Solar and Terrestrial Weather. It has long been known that a cycle of solar weather begins in about lat. 32° N. and S., and in a period of eleven years ends in about lat. 5° N. and S. Just before one cycle ends another commences. The greatest amount of spotted surface occurs when the solar weather-changes pro- duced in the cycle reach about lat. 16°X. and S. It becomes, therefore, of the first importance to correlate the times of mean solar temperature, and of the + and - heat pulses, with the solar weather cycle, in order to arrive at the temperature-history of the sun during the period which now concerns us. This may be done as follows : — Solar cycles. Lat. of spots 19° 12° 18° 10° 19° Heat condition mean + mean — mean + mean — mean + Years .. 1869 1870-5 1876 1877-80 1881 1882-6 1886-7 1888-91 i 1891-2 1892 Connection of the Spots with Prominences. In 1869, when a sunspot maximum was approaching, the promi- nences were classified by one of us into eruptive and nebulous; the former showing many metallic lines, the latter the hydrogen and helium lines chiefly. This conclusion, which was published in 1870, was subsequently confirmed and adopted by Secchi, Zollner, Sporer, Young, and Respighi. In the same year prominences on the sun's disc were also observed by one of us by means of the C and F lines.* The eruptive prominences, unlike the nebulous ones, were not observed in all heliographic latitudes ; but, according to the extended observations of Tacchini and Kicco, had their maxima in the same • ' Boy. Soc. Proc.,' vol. 17, p. 415. Solar Changes of Temperature and Variations in Rainfall. 415 latitude as the spots. This is especially well shown by the diagrams illustrating the distribution of spots, faculse, eruptions, and protube- rances which are given by Tacchini for 1881 — 1887 in the ' Memorie della Soc. degli Spettroscopisti Italiani,' 1882 — 1888. These curves show in the most unmistakable manner that the spots, faculse, and eruptive or metallic prominences have their maximum frequency in the same solar latitudes while the nebulous or quiet prominences are more uniformly distributed, and even have maxima in zones where spots are rarely observed. This is corroborated by what Professor Respighi many years ago stated : " In correspondence with the maximum of spots, not only does the number of the large protuberances increase, but more than this — their distribution over the solar surface is radically modified." In his observations, Professor Young found that the H and K lines of calcium were reversed in the chromosphere as constantly as h or C, and the same lines " were also found to be regularly reversed upon the body of the sun itself, in the penumbra and immediate neighbourhood of every important spot."* This result was confirmed by the early (1881) attempts of one of us to photograph the spectra of the chromo- sphere and spots, and also by eclipse photographs. In the photographic spectrum, the H and K lines are by far the brightest of the chromo- spheric lines, and this fact has been utilised by Hale and Deslandres, acting on a suggestion due to Janssen, for the purpose of photo- graphing at one exposure the chromosphere and prominences, as well as the disc of the sun itself, in the light of the K line. These photographs thus give us in K light the phenomena which one of us first observed by the lines C and F of hydrogen, and thereby I present a record of the prominences across the whole disc of the sun as well as at the limb. In such photographs near sunspot maximum, the concentration of the prominences in zones parallel to the equator is perfectly obvious at a glance. Eruptive or metallic prominences are thus seen to cover a much larger area than the spots, so that we have the maximum of solar activity indicated, not only by the increased absorption phenomena indicated by the greater number of the spots, but by the much greater radiation phenomena of the metallic prominences ; and there seems little doubt that in the future the measure of the change in the amount of solar energy will be determined by the amount and locus of the prominence area. Spots are, therefore, indications of excess of heat, and not of its defect, as was suggested when the term " screen " was used for them. We know now that the spots at maximum are really full of highly heated vapours produced by the prominences, which are most numerous when the solar atmosphere is most disturbed. * ' Catalogue of Bright Lines in the Spectrum of the Chromosphere/ 1872. 410 Sir Norman Lockyer ami Dr. \V. -1. S. Locky»-r The Indian meteorologists have abundantly proved that the in- creased radiation from the sun on the upper air currents at maximum is accompanied by a lower temperature in the lower strata, and that with this disturbance of the normal temperature we must expect pressure changes. Chambers was the first to show that large spotted area was accompanied by low pressures over the land surface of India.* Passing, then, from the consideration of individual spots to the zones of prominences, with which they are in all probability associated, it is of the highest interest to note the solar latitudes occupied when the crossings previously referred to took place, as we then learn the belts of prominences which are really effective in producing the increased radiation. The area of these is much larger, and therefore a consider- able difference of radiation must be expected. The greater disturbance of certain zones of solar latitude seems to be more influential in causing the + pulse than the amount of spotted area determined from spots in various latitudes. It is all the more necessary to point this out localise the insignificance of the area occupied by the spots has been used as an argument against any easily recognised connection between solar and terrestrial meteoro- logical changes.! Assuming two belts of prominences X. and S. 10° wide, with their centres over lat. 16°, the sixth of the sun's visible hemisphere would be in a state of disturbance. Indian Rainfall. S.ir. Monsoon, 1877-1886. It will be clear from what has been stated that our object in studying rainfall was to endeavour to ascertain if the + and - temperature pulses in the sun were echoed by + and - pulses of rainfall. The Indian rainfall was taken first, not only because in the tropics we may expect the phenomena to be the simplest, but because the regularity of the Indian rains had broken down precisely when the widened line observations showed a most remarkable departure from the normal. It was also important for us to deal with the individual observations as far as possible, because it was of the essence of the inquiry to trace the individual pulses if they were found. Hence the S.W. monsoon was, in the first instance, considered by itself, because although Eliot holds that the winter rains (X.E. monsoon) are due to moisture brought * 'Abnormal Variations, ' p. 1. t " So far as can be judged from the magnitude of the sunspcts, the cjclicml variation of the magnitude of the sun's face free from spots is very small compared with the surface itself ; and consequently, according to mathematic principle, the effect on the elements of meteorological observation for the whole earth ought also to be small " (Eliot, ' Report on the Meteorology of India in 1877,' p. 2). Solar Clianfjes of Temperature and Variations in Rainfall. 417 by an upper S.W. current,* their incidence is very different and their inclusion might mask the events it was most important to study. The first investigation undertaken was the study of the rainfall tables published by the Meteorological Department of the Government of India. These were brought together by Blanford down to the year 1886.1 As the widened line observations were not begun at Kensington till 1879, the discussion was limited in the first instance to the period 1877-1886 inclusive, embracing the following changes in solar temperature, occurring, as will be seen, between two conditions of mean solar temperature : — Mean. — pulse. Mean. + pulse. Mean. 1876 1877—1880 1881 1882—1886 1886—1887 Bearing in mind that the intensity of the + pulse may in some measure be determined by the solar disturbances, which for the present are registered by spotted area, it is important to point out that the preceding maximum in 1870 was remarkable for obvious indications of great solar activity.:}: It soon became evident that in many parts of India the + and - conditions of solar temperature were accompanied by + and - pulses producing pressure changes and heavy rains in the Indian Ocean and the surrounding land. These occurred generally in the first year following the mean condition, that- is in 1877-8 and 1882-3, dates approximating to, but followed by, the minimum and maximum periods of sun-spots. * ' Report,' 1877, p. 125. t ' Indian Meteorological Memoirs,' vol. 3. £ " The year 1870 was characterised by an exuberance of solar euergy, which ii* without parallel since the beginning of systematic observations (i.e., since 1825). The number of observed groups far exceeds that of any previous year, and it appears also from a cursory comparison witli the maximum year's observations, as- recorded by Hofrath Schwabe, that the magnitude of the different groups, as well as the average amount of spotted surface during any period of the year, is- unprecedented." (' Monthly Notices,' vol. 31, p. 79, Warren de la Rue, E. Stewart, B. Loewy.) The table which the authors of this paper give shows that during the year, although observations of the sun were made on 213 duys out of the 364, there was no day without spots recorded. In fact, during the whole year no less than 403 new groups of spots were noted, thus showing us that on the average there was more than one new group per diem. The authors further remark. "A very remarkable feature of the groups observed during the year appears to be their extraordinary lifetime ... an exceedingly large number of groups completed three, four, and even more revolu- tions before finally collapsing." VOL. LXVII. 2 H 41S Sir NonnMii L»cky.:r and I>r. \V. .1. S. Lo-kv.-i. Meldrum, a* far hack as !.*>!.* referred in "the eMrrme ox-illations of \\eathcr changes in dinerent pl.n «•-. at the turning points of the <-ir ve- rrpre.-enting the im-rea-i- ami \icwr that the S.W. monsoon depends upon the oscillations of tin- equatorial Kelt of low pressure up to 31° N. lat. at the summer solstice. The months of rain-receipt on the upward and downward swing will therefore depend on the latitude, and these months alone have lieen considered. We Ixjgan by taking elevated stations in high and low latitudes. rThe 1881 pulse (in July) was the heaviest recorded (1*77 inches) save one in 1882 : the rainfall was nearly as Lat. 34 N.< , . , 1 1,500 feet ^^ ^c felfc -n 187g ^ thc highest of a,, Murree rThe 1881 pulse (August) is high, but is followed by a Lat. .'{.'5 X. < higher next year. 6,344 feet (.The 1878 pulse (August) is highest of all. ^ C\VCI**l "*\ " -.-,,. ri Taking the fall in July and August. The 1881 pul conceal the meaning of the separate pulses observed in separate * "On the Relations of Weather to Mortality, mid on the Climatic EiIY-i-t of .Forests." t ' Indian Meteorological Memoirs,' vol. 4, Part V. p. 271. *J '^Nature,' vol. 5(5, p. 110. Sohtr Changes of Tempera/fare nml Variations in Rainfall. 419 localities ; this we found to be the case. But, nevertheless, the table helped us greatly, because it included the summation of results nine years later than those included in Blanford's masterly memoir. Predominant pulses were found in 1889 arid 1893, following those of 1877-8 and 1882-3. So that it enabled us to follow the working of the same law through another sun-spot cycle, the law, that is, of the mean solar temperature being followed by a pulse of rainfall. Mean sun. Bain pulse. 1876 -1878 1881 +1882 1886-7 - 1889 1892 +1893 The main . feature of this table is the proof of a tremendous excess of rainfall in 1893 — by far the greatest excess of all (percentage variation, + 22). This was far greater than the excess in 1882. The next remarkable excess occurs in 1878 (percentage variation, + 15). The pulses in the period stand as follows : — - Min 1878 Percentage variation. + 15 + 6 + 6 + 22 Heat pulse. Years after rise of iron lines. —Max. 1882 —Min. 1889 —Max. 1893 + — I Years after rise of + — unknown lines. The variations in the intensities of the pulses of rain at the successive maxima and minima are very remarkable, and suggest the working of a higher law, of which we have other evidence. But, putting this aside for the present, it should be pointed out that even normally we should not expect the same values for the rainfalls in 1882 and 1893, because the amount of spotted area was so different, 1 1 60-millionths of the solar surface being covered with spots in 1883, and 1430 in 1893. The very considerable variation in the quantity of snowfall on the Himalayas has often been pointed out by the Indian meteorologists. We have, therefore, used the "whole India" curve between 1875 and 1896, to see whether the sun pulses, which we have found to be bound up with the Indian rainfall, are in any way related to the snowfall as might be expected. The Himalayan snowfall beyond all question follows . the same law as the rain, the value occurring at the + and - pulses, as under, being among the highest : — * * ' I.M.M.,' vol. 3, p. 235. 2 H 2 420 Sir Norman Lockyer and Dr. W. -I. S. Inches. -1867-8 134 - +1871-2 110 -1877-8 207 + 1882-3... 81 From these tables it follows that both in rainfall and snow the* quantity is increased in the years of the rise l>oth of the unknown and iron lines. Other llainfall*. Being in presence of pulses of rainfall in India during the south-west monsoon, corresponding with pulses of solar change, it l>ecame necessary to attempt to study their origins. AVe may, add that other pulses were- traced, especially one in 1875, but the simplest problem was considered alone in the first instance. The rainfalls at the Mauritius, Cape Town, and Batavia, were collated to see if the pulses felt in India were traceable in other regions surrounding the Indian Ocean to the south and east. Tlw Mauiitiuft Rainfall. The rainfall of Mauritius has been obtained by utilising the results that have been published in the Blue Books* issued by the Royal Alfred Observatory since the year 1885. The volume for 1886 give* the yearly total rainfall for every station that was then in use from 1861 up to the year 1885, and these values have been employed; since then, the yearly values have been obtained direct from each of the yearly volumes subsequently published, i.e., to the end of the year 1898. It was at first thought that the total Mauritius rainfall could be fairly obtained by employing for the period between 1861 and 18861 the means of several stations as given by Meldrum,t and continuing the values from the observations published in the more recent yearly volumes. It was found, however, that from 1861-1880 the rainfall was obtained from the observations of four stations, while from 1871- 1886, the observations from eight stations were employed. As a study of all the published data showed that more stations * " Mauritius Meteorological Results." t 1861-1880. 'Relations of Weather to Mortality, &c.,' 1881, p. 30. 1871- 1886. 'Annual Report of the Director of the Royal Alfred Observatory for 1886,' p. 18. Solar Changes of Temperature and Variations in Rainfall. 421 might be utilised in determining the total rainfall of Mauritius, it was •decided to discuss all the observations afresh, and make use of as many as possible. To this end the records of twenty-eight stations, situated in six •different districts, were chosen, and the total rainfall for each year obtained. It is only natural that the number of rain-gauge stations in the early year of 1860 was not so numerous as that of more recent years ; the facts may be stated as follows : — Mean yearly Number Mean yearly Number Years. rainfall variation from of stations -,,- rainfall Years. • ,- £ variation, troin of stations normal. used. normal. used. Inches. Inches. 1861 + 26-6 1 1880 -19-3 23 1862 -10-2 4 1881 - 7'3 25 1863 + 9-6 6 1882 + 16-6 25 1864 -12-2 8 1883 1-1-8 26 1865 + 22-6 10 1884 -12-4 25 1866 -18-2 10 1885 - 9-8 26 1867 - 6-6 11 1886 -35-3 26 1868 + 27 -1 11 1887 - 4-2 27 1869 - 3-3 12 1888 -f 22 -3 26 1870 - 3-6 12 1889 + 18 -4 24 1871 -18-9 13 1890 + 1-2 25 1872 - 7-0 13 1891 +1-4 26 1873 + 10-3 13 1892 +5-1 19 1874 + 17-4 17 1893 - 7'4 24 1875 + 3-0 15 1894 - 0-6 24 1876 - 6-5 17 1895 + 10 -0 21 1877 + 31 -4 19 1896 + 17-6 22 1878 - 3-8 22 1897 -19-6 24 1879 - 5-2 22 1898 - 2-1 24 With regard to the general rainfall of Mauritius throughout the year, it may be stated that on the average the most rainy months are from December to April, both months inclusive. The months of November and May are those in which the daily rainfall is increasing and diminishing respectively. Sometimes in •July or August there is a slight tendency for a small increase. The Mauritius Rainfall Curve for the period 1877-1886. In plotting the Mauritius rainfall curve for the period 1877-1886, it was observed that the curve is of a fairly regular nature, showing .alternately an excess and deficiency of rainfall. The highest and lowest points of the curve will be gathered from the following table : — -ll'l' Sir X.niiiaii L<>«-kyri am! I >r. \V. .1. S. l....-ky»-r. Year. Maximum. Excess. Deficiency. 1877 1880 1882 1886 ... 31-4 16 -G 19-3 35-3 Comparing the times of occurrence of the two pulses of rainfall at Mauritius with the times of the crossings of the known and unknown lines, it is found that the Mauritius maximum rainfall of 1877 occurs about a year after the rise of the known lines in 1876. The next Mauritius pulse of rainfall in 1882 follows the succeeding crossing, when the unknown lines are going up, also about a year later. Comparison of the Mauritius Rainfall mth those of Leh, Murrte e expected if the rain at sun-spot minimum comes from the south, as has been surmised. The fact that the pulses at Mauritius, Ceylon, and India in 1882 occur simultaneously, is very strong evidence in favour of an origin in the equatorial region itself for the Indian rain at sun-spot maximum. The pulse at maximum in the Indian south-west monsoon may depend to a large extent upon the action of the excess of solar heat on the equatorial waters to the south of India, and not on an abnormal effect on the south-east trade. * We have found that there was ;i defect of the usual rainfall at Mauritius in 1892-93, and yet the rain supply in India was in excess. KKSl'LT OF THE COMPARISON OF RAINFALL. Tli>' + (tnJ - Pa It seems quite certain that we are justified in associating the 1 >7S pulse of rainfall during the south-west monsoon in India with the rain- fall common to Mauritius, Batavia, and the Cape at that date ; that in Solar Changes of Temperature and Variations in 7.<. The various curves which we have drawn for the purposes of study have been compiled from yearly means, and so far, in these curves the rainfall in months has not been considered. That will have to come later. Hence if the rainfall which most influences the yearly mean occurs in the last three months at one place, and in the first three months of the next year at another, they are shown as being a year apart, whereas they have actually been continuous. With regard to the travel of the pulses over large areas under the influence of the 8.E. trade, it may be gathered from the pressure charts that the + and - conditions of pressure are apt to lie over the centres of land and water areas, and not generally over coast lines. In the case of water surfaces, the effect of a sudden change in the solar radiation on the pressure might be expected to be felt not at the point where the pressure is least or greatest at the time, and of the most general type, but where the equilibrium is most unstable. On the other hand, more time would be required for the new pulse to establish itself where the conditions are more complicated. Hence we should expect the pulses to be felt first in the eastern part of the Southern Ocean, and this seems generally to be the case. Thus after the mean solar temperature of 1876, the - pulse was felt first at .Mauritius, then in India, and the Cape. After the mean of 1881, the + pulse was felt first at Mauritius, then in India, and the Cape. Cor- doba felt both pulses in the same year as India and the Cape. StibsitJuiry Pnlxe*. In a normal sun-spot curve we find a sharp rise, generally taking three or three-and-a-half years, to maximum, and a slow decline to minimum, on which the remaining years of the cycle are spent. The curve on the upward side rises generally regularly and con- tinuously ; on the -down ward portion the regularity of the curve is very often broken by a " hump " or sudden change of curvature. There has not yet been a complete discussion of the number and character of the prominences associated v/ith the spots during the cycle ; Solar Changes of Temperature and Variations in Rainfall. 425 we have found, however, that the " hump " in the sun-spot curve in 1874 was accompanied by a remarkable increase in the number of eruptive prominences. (Acce 3.189 S]£ «3 SQ: i P \ s \ V SE: ooggggoo \ oo 5s. ^> ggg We have already referred, iri discussing the Indian rainfall, to a remarkable intensification of the south-west monsoon in 1874-75, the effect of which is especially noticeable in the rainfalls of the Konkan •and North-west Provinces, and we have come to the conclusion that we must consider all these events as due to a common cause — that is, to 41'tl Sir I.ucky.T aii-l I>i. \V. .1. S. Solar Chaw/ex of 7V///y>ov/V/r and Variations in Rainfall. 427 n subsidiary solar pulse. We propose to return to this subject in a subsequent communication, after inquiries have been completed relat- ing to 1885-86 and 1896-97. Tlif Intern/Is between the Pulses. There will obviously be intervals between the ending of one pulse and the beginning of the next, unless they either overlap or become continuous. The + and - pulses, to which our attention has been chiefly directed, are limited in duration ; and when they cease the quantity of rain which falls in the Indian area is not sufficient without water storage for the purposes of agriculture ; they are followed, therefore, by droughts, and at times subsequently by famines (fig. 2). Taking the period 1887-89 we have— Rain from - pulse No rain pulse ...... Rain from + pulse- No rain pulse 77 78 79 (part) 79 (part) 80 (central year) l (part) :81 (part) 82 83 84 (part) 84 (part) | (central years) 87 (part) C 87 (part) Rain from - pulse < 88 189 The duration of these + and - pulses of rainfall was determined in the first instance by the Mauritius rainfall, which shows both pulses : and later from the Malabar rainfall, which perhaps shows the effect of the south-west monsoon in its greatest purity. All the Indian famines since 1836 (we have not gone back further) have occurred in these intervals carried back in time on the assumption of an eleven-year cycle. The following tables show the result for the two intervals : — 428 Sir Norman Li.<-k\ IT and I >r. \V. .1. S. The Interval between the Pulses, taking 1880 as the Central Year, on the Upward Curve. 1880, Madras famine. N. W.I', famine. 1880-11 = 1869, N.W.P. famine (1868-9). 1869-11 = 1858, N.W.P. famine (1860). 1858-11 = 1847. 1847-11 == 1836, Upper India famine (1837-8). (Great famine.) The Interval between the Pulses, taking 1885-6 as the Central Years, on the Descending Curve. 1885-6 f (1884-5). I Madras famine J 1885-6-11 - 1874-5, N.W.P. famine (1873-4). Bombay famine (1875-6). Bombay famine "I /, 876-7} Upper India famine J 1874-5 - 11 = 1863-4, Madras famine"! Orissa famine J 1863-4-11 = 1852-3, Madras famine (1854). It is clear from the above table that if as much had been known in 1836 as we know now, the probability of famines at all the subsequent dates indicated in the above tables might have been foreseen. The region of time from which the above results have been obtained extended from 1877 to 1886. The next table will show that if the dates, instead of being carried back, are carried forward, the same principle enables us to pick up the famines which have devastated India during the period 1886-97. Same intervals, going Forward. 1880. + 11 1891, N.W.P. famine (1890). Madras famine "| Bombay famine *> (1891-2). Bengal famine J 1885-6. + 11 1896-7, General famine. This result has arisen, so far as we can see, from the fact that the + and - pulses included in the period 1877-1886 were normal ; that is, were not great departures from the average. Solar Changes of Temperature and Variations in Rainfall. 429 Nik Floods. After we had obtained the above results relating to the law followed by the Indian famines, we communicated with the Egyptian authorities with a view of obtaining data for the Nile Valley. We have since found, however, from a memorandum by Eliot,* that Mr. Wilcocks, in a paper read at the Meteorological Congress at Chicago, remarked that " famine years in India are generally years of low flood in Egypt." It remains only for us, therefore, to point out that the highest Niles. follow the years of the + and - pulses. Thus : — 1871, one year after + pulse 1870. 1876, two years after subsidiary pulse of 1874. 1879, two years after - pulse 1877. 1883-4, one and two years after + pulse 1882. 1893-4, after + pulse 1892 (India excess rainfall, 1892-3-4). The Great Indian Famine of 1899. When, in a sun-spot cycle, the solar temperature is more than iisually increased, the regularity of the above effects is liable to be broken, as the advent of the - pulse is retarded. This, as we have already pointed out, is precisely what happened after the abnormal + heat pulse of 1892, following close upon the con- dition of solar mean temperature. The widened line curves, instead of crossing, according to the few precedents we have, in 1897 or 1898, have not crossed yet — that is, the condition of ordinary solar mean temperature has not even yet been reached. We have shown that, as a matter of fact, in a normal cycle India is supplied from the Southern Ocean during the minimum sun-spot period, and that this rain is due to some pressure effect brought about in high southern latitudes by the sun at - temperature. As the - temperature condition was not reached in 1899, as it would have been in a normal year, the rain failed (fig. 3). We may say then that the only abnormal famine recorded since 1836 occurred precisely at the time when an abnormal effect of an un- precedented maximum of solar temperature was revealed by the study of the widened lines. We desire to tender our acknowledgements to Dr. Buchan, F.K.S., and Mr. Shaw, F.R.S., for their kindness in so promptly replying to our appeal for rainfall tables. We wish also to thank Mr. H. Shaw, one * Forecast of S.W. Monsoon rains of 1900. 430 Solar Changet of Temperature and TariM .-;»//. of the teachers in training ;a tin- Royal Coll, i once, for assist- iince in bringing togi-tlier rainfall data and plotting numerous cu: Fio. 3. 1889 1690 1891 EXCESS. DEFICIENCY. MAX 1633 4 MAX. 1696 NCB . ' 18 /MC «*L EXCESS NORMAL * V * N/ •*• N )8 1839 1900 100 60 60 40 20 0 30 20 "j 1892 1693^ 1894 1695 I8i Q HAVE H AF J6 PEN 1897" IR ON. v, iVH^ TO UGH TT ^ ED. (S.V >NS( JON) V 3 ^7^- ^ ^ &» •^^ , ^ > ^ ^ ""- ^ X . Q g x i si N3F OT> X • X i - r^UNKMj OWN _- — • -^^_ •— — N -^ -• — • 1 — --, • — ^ ,.~-— ' ^? *«a ^1 - -t - - — fc, S2 =» 7^ ~- «_ -— — / X 1AUI JITII, S R AINI AL> / X / \ ^ / ^ ^ — ' ^ / 10 " 20 SO \ X ? ^^ - - \ / " \ ^/ s / 100 60 60 40 20 0 30 20 ^ ^^ ^ fVHA \ ^-i* T D IDh API >EN (W HOL E If IDIA ) i nuy s/A ift /I , •' ~- , — X / •--, ++-• -«„ — - -... V - _^« su HS.P yf. s •• •• — i -5=S ~~-— ^ 'UN IN YN.. ^ ~~- — . . —-- ^ '•». -- ^*~n ^^* ^~ \ ^ HAU h ^ »mi r S R . AINf V. s~ ^ \ ALL .— — , — - \ X 10 " \ X EXCESS. DEFICIENCY. NORMAL. GREATEST GOOD (MAX.l89i> < MAX 16961 DEFICIENCY MAINS. ON RECORD OVER LARCEST AN LA. On Co-ordinated Movements after J\V/row joint, and the two points of union were therefore situated one on the outside and one on the inside of the limb, with a bulky muscle between them, which prevented any possibility of confluent reunion of all the divided ends. One of the experiments (Exp. I) was a failure on account of the wound becoming septic, but in the remaining three (Exp. II, III, IV) the animals regained almost completely the power of making voluntary co-ordinated movements of the limb. Thus the leg was used constantly and perfectly in walking and running, and in performing such co- ordinated movements as giving the paw on request, using the paw to hold a bone while gnawing, &c. The recovery of function commenced about the 30th day after the operation, and was almost perfect from the 45th to the 90th day. The physiological examination showed that the nerves which had been crossed had united as they had been placed without one point of union communicating with the other, and that the flexor muscles were thus entirely supplied by the musculo-spiral, and the extensor muscles entirely by the median, ulnar, and musculo-cutaneous. In two of the experiments (Exp. II and III) the musculo-spiral stimulated above the seat of union gave flexion of the paw, and no movement in the extensor muscles, while stimulation of the central segments of the musculo-cutaneous, median, and ulnar gave extension of the paw, and no movement in the flexor muscles. Stimulated on the cerebral cortical centres of the sigmoid gyrus, it was found that on the left hemisphere the centre which normally gives on stimulation flexion of the paw, gave on the contrary extension, and no movement whatever in the flexor muscles. Stimulation of the centre, normally associated with extension of the paw, gave in one of the animals pure flexion of the paw and no contractions of the extensor muscles (Exp. III), while in the other animal the flexion centre was found to lie in the normal extension area, but pure flexion could not be obtained free from extension movements (Exp. II). In the other experiment (Exp. IV) the results of stimulation were somewhat obscure. Stimulation of the central segments of all four nerves gave contractions in the extensor muscles and no contractions in the flexors. Yet the flexors were perfectly healthy in appearance and possessed normal irritability to faradic stimuli. Stimulation of Co-ordinated Movements after Nerve Crossing, ctr. 433 the centres on the left sigmoid gyms showed that the flexion centre had l>ecome an extension centre, but no flexion centre could be discovered. In these experiments (II, III, IV) the centres on the right side of the brain were normally placed. In all the experiments the irritability of the centres on the left side of the brain was increased rather than diminished. In addition to these experiments on nerve crossing, there was also an experiment made on a dog to ascertain if the fact of crossing the nerves delayed the functional recovery beyond what would be expected merely as a result of nerve section. In this experiment the same nerves were divided, but were immediately reunited as accurately as possible. The result was that the course of recovery of function was not materially different from the course in the experiments on nerve crossing. The physiological examination showed that the nerves had united well, and regained their normal irritability and conductivity, and that the muscles of the limb were healthy. Examination of the cerebral cortical centres showed that they were not well defined, but neither were they on the sound side in this animal. II. — Junction of the Peripheral Segment of the Divided Facial Nerve with the Trunk of 'the Spinal Accessory Nerve for the Treatment of Facial Spasm in a Woman. The experiments on dogs having shown that nerve crossing was 'followed by recovery of co-ordinated function, the following operation was undertaken for the treatment of facial spasm in a woman. Faure <'uid Furet had already suggested utilising the branch of the spinal .accessory to the trapezius for the supply of the face in the case of paralysis of the facial nerve, arid Faure* had put the operation in prac- tice, but without success. In the following case the patient had suffered for ten years. The right side of the face was incessantly twitching, the angle of the mouth being permanently drawn up, and the eyelids half closed. The condition had been under treatment at different periods, but without any success. Rather the condition got worse. On May 4, 1899, the facial nerve was divided close to its exit from the aqueduct of Fallopius, and grafted on to the trunk of the spinal accessor}7, just as the latter nerve emerges from under the posterior belly of the digastric muscle. The digastric situated between the central end of the facial nerve and the junction with the spinal acces- sory prevented any reunion of the nerve. Immediately after the operation, the right side of the face was in a condition of complete paralysis, and it remained in this condition for * Faure, "Traitement Chirurgioal de la Paralysie Faciale par 1* Anastomose •Spino-faciale," ' Eerue de Chirurgie,' vol. 18 (1598), p. 1098. VOL. LXVII. 2 1 434 IM. \i. Kennedy. On //•• 'ft»n <>f some time, the muscles losing their faradic irritability. In course of time gradual improvement showed itself, heralded first by recovery of faradic irritability in the muscles. The earliest indications of improve- ment were shown in the orbicularis palpebrarum, which began to recover faradic irritability and movement, thereby enabling the eye to be slightly closed, about the 18th day. The movement of the muscle on stimulation with the faradic current was, however, so slight, that there was a possibility of error, and the slight voluntary movement might have been due simply to the relaxation of the levator palpebrae. By the 49th day, however, there was no doubt, as the contractions to faradic stimulations were well marked, and the palpebral fissure could be voluntarily closed one-half. By the 141st day the faradic irritability of the other muscles began to be recovered, and by the 155th day the faradic current gave, on applying the electrode over the junction between facial and spinal accessory, strong contractions in all the muscles of the face. Improvement gradually continued, and on August 17, 1900, about fifteen months after the operation, the condition was as follows : She experienced no difficulty on account of the condition of the face. There was no return of the spasmodic condition. The conjunctiva of the right eye was quite normal ; there was no increased lachrymal secretion, and she never was troubled with dust getting into the eye, as winking was perfectly efficient. She could shut the eye completely, although not so tightly as in the case of the sound eye. The orbicularis palpebrarum also contracted well to reflex stimuli. The right side of the brow could be wrinkled to a very slight degree only, and movements could be made in the cheek and mouth, although they could not well be co-ordinated. The labial letters could be perfectly pronounced, and the buccinator was efficient to prevent accumulation of food between cheek and gums while eating. There was no atrophy of the side of the face, and in repose there was no appearance of facial paralysis, the muscles having regained their tonus, and the normal sulci being well marked. There was evidence of want of power over the face in the difficulty of raising the eyebrow, or of making a circular aperture with the mouth in whistling or blowing. The muscles, however, of these parts were perfectly sound, as the faradic current gave perfectly normal reactions, both when applied directly to the muscles and when applied to the motor point of the nerve. This motor point lay about 2 cm. lower down than normally, i.e., over the junction of the facial and spinal accessory. The reactions and movements of the trapezius and of the sterno- mastoid were normal. A curious effect resulted when the arm was suddenly thrown up ; for the face at the same time was thrown into contractions, owing to Co-ordinated Movements after Nerve Crossing, &c. 435 the impulses intended for the trapezius being directed to the face. If the arm was continued held up, these contractions of the face passed off. III. — General Conclusions. 1 . In the fore-limb of the dog, the nerve supply of the flexor muscles may be crossed with that of the extensor muscles, with the result that, despite the altered innervation, the animal regains, as before^ the power of performing voluntary co-ordinated movements of the limb. 2. The fact of crossing the nerves does not add materially to the time which would be required for recovery of function of the limb if the same nerves were simply divided and reunited by suture as accurately as possible. 3. The result of crossing the nerve supply of antagonistic groups of muscles is that the nerve centres which formerly innervated the one group now serve for the other group, and this alteration extends to the cerebral cortical centres, which become interchanged in position and retain their irritabilit}'. 4. The cerebral cortical centres which have been made to interchange their positions by the crossing, are able, in response to the will, to emit impulses which can cull forth in the new peripheral terminations move- ments in perfect co-ordination. 5. In man the facial nerve may be detached from the facial centre, attached to the spinal accessory nerve, and the facial muscles thus in- nervated by the spinal accessory centre, with the result that co-ordin- ated movements of the face, both voluntary and reflex, are at least in part restored. 6. In the case of reunion of a divided nerve, it is not necessary to suppose that regeneration restores the old paths for the nervous im- pulses, since, if new paths are formed by the imperfect co-adaptation of the divided nerve ends, with the result of altering the connections between central nerve cells and peripheral endings, the organism has the power of compensating this alteration. 7. In the case of paralysis of a muscle or group of muscles, if the nerve supplying the affected muscle or muscles is grafted on to a neighbouring efferent nerve supplying muscles which are healthy, it is probable that the affected muscle or group of muscles, if not already destroyed by degenerative process, will regain its normal function. 2 I 2 436 Anniversary Meeting and .!/« //',/-/ <>f I>«- ////«•/• 6, 1900. November 30, 1900. Anniversary Meeting. The LORD LISTER, F.R.C.S., D.C.L., President, in the Chaii. (A full Report of the Anniversary Meeting, with the President's Address and Report of Council, will be found in the ' Year-book ' for 1901. The Account of the Appropriation of the Government Grant and of the Trust Funds will also be found in the ' Year-book.') December 6, 1900. Sir WILLIAM HUGGIXS, K.C.B., D.C.L., President, in the Chair. His Grace the Duke of Northumberland was admitted into the Society. A List of the Presents received was laid on the table, and thanks ordered for them. The President announced that he had appointed as Vice-Presidents for the ensuing year — The Treasurer. The Astronomer Royal. Lord Lister. Mr. J. J. H. Teall. The following Papers were read : — I. " The Histology of the Cell Wall, with Special Reference to the Mode of Connection of Cells." By WALTER GARDINER, M.A., F.R.S., and A. W. HILL, B.A. Part I.—" The Distribution and Character of ' Connecting Threads ' in the Tissues of l'in>i< sylvestris and other Allied Species." By A. W. HILL. II. " On the ' Blaze Currents ' of the Frog's Eyeball." By Dr. WALLER, F.R.S. III. "On a Bacterial Disease of the Turnip (Brassica napus)" By Pro- fessor M. G. POTTER. Communicated by Sir M. FOSTER, Sec. R.S. The Histology of the Cell Wall, &c. 437 IV. " The Micro-organism of Distemper in the Dog, and the Produc- tion of a Distemper Vaccine." By Dr. S. MONCKTONT COPEMAN. Communicated by Sir M. FOSTER, Sec. R.S. V. " On the Tempering of Iron hardened by Overstrain." By JAMES Mum. Communicated by Professor EWING, F.E.S. " The Histology of the Cell Wall, with special reference to the Mode of Connection of Ceils."* By WALTER GARDINER, M.A., F.R.S., Fellow and Bursar of Clare College, Cambridge, and ARTHUR W. HILL, B.A., Scholar of Kingts College, Cambridge. PART I. " The Distribution and Character of ' Connecting Threads ' in the Tissues of Pinus sylvestris and other Allied Species." By Arthur W. HILL, B.A., Scholar of King's College, Cambridge Received July 17, — Read December 6, 1900. (Abstract.) The research with which this paper is concerned was undertaken with a view of ascertaining to what extent " connecting threads " are distributed throughout the body of any given plant, and for this purpose the endosperm and the various tissues of the hypocotyl, cotyledons, and root of the young seedling of Pinus pinea, and of the adult stem leaf and root of Pinus sylvestris, were examined. The results show that the presence of such threads can be readily demonstrated in the case of all cells in which the wall retains its cellulose or mucilaginous character, and that in such young tissue as the growing point of the root all the cells are provided with connecting threads. When the lignified or suberised condition has supervened it is difficult or impossible to identify threads, though even in such cases threads may be recognised in certain of the very young elements. In Pinus pinea the tissue of the endosperm, as also that of the germinating seedling, is well connected by threads. In the cotyledon the absorptive side next the endosperm (corre- sponding to the lower side of the leaf) shows a certain histological distinction in that the walls of the cells, both of the epidermis and of the subjacent parenchyma, are more richly provided with threads than are the similar tissues of the upper side. No threads, however, occur * For the preliminary communication on this subject, see Gardiner: "The Histology of the Cell Wall, with special reference to the Mode of Connection of Cells," ' Roy. Soc. Proc.,' vol. 62, 1897. 438 The //{«(•>/•>;/ >/ »/(/"• (.',// »',///, t£r. in the outer or free walls of the epidermis, so that diffusion only (as opposed to direct transference) can take place l>etween the cell con- tents of the endosperm and those of the cotyledon. In the stomata of the cotyledon threads have been seen in a few cases connecting the guard cells with the epidermal cells. The parenchymatous tissue all over the seedling plant shows con- necting threads of a similar character. In the end walls of the cells they occur irregularly scattered, but in the lateral walls they are usually in isolated groups, mainly in consequence of the growth in length which these walls have undergone, and are also "situated in shallow pits. The palisade cells of the cotyledon, which at first are united together in all directions, very soon separate, forming plates of tissue, and the threads in the walls along which separation takes place are very quickly obliterated. A similar obliteration of threads is seen to occur in those walls of pericyclic cells which are situated between the living cells and the young transfusion cells in process of lignification. The living cells of the pericycle, which are richly connected together by threads, form the passage cells from the cortical tissues to the phloem, and between these cells and the sieve tubes come the albu- minous cells, which possess thread groups occurring in localised thick- enings of their walls. The threads, which are long and usually curved, stain in a peculiar manner, and appear to have an important function with reference to the passage of material from the mesophyll to the phloem. The phloem tissues of the seedling of Pinus pinea present a distinct type, the peculiarities of which are treated of at some length. The large cells of the outer portion are characterised by long oblique end walls full of threads; whilst the thick- walled cells of the inner part possess square end walls traversed by numerous long threads, re- sembling the sieve tubes of dicotyledons. As development proceeds, sieve tubes like those of the adult tissues are, however, quickly developed from the cambium. All the sieve tube threads show a characteristic median dot. The root cap of the seedling root shows numerous threads connecting its cells together, and also affording communication both with the free surface of the root as well as internally with the cells of the periblem. The function of the root cap as an organ for stimulus perception and as an absorbent organ is considered with reference to the abundance of the connecting threads. In Pinus sylwslris the characters of the threads in the cortical tissues of the adult stem and root are similar to those of the seedling. Threads occur, however, in the radial and end walls of the cells, but in the cells just under the cork they are distributed in large numbers in the tangential walls, and this change in the main direction of the On the "Blaze Currents " of the Frog's Eyeball. 439 threads points to their value as a means of conducting food material to the developing cork. In the phloem there is a sharp contrast between the starch-containing medullary ray cells and bast parenchyma on the one hand, and the sieve tubes on the other, and no threads can be found directly con- necting the parenchymatous cells with the sieve tubes, but the albu- minous cells of the ray possess numerous thread groups which com- municate with both tissues. The starch medullary ray cells in the phloem and xylem possess numerous threads in the tangential and basal walls, especially in the former, and are also united with the bast parenchyma and albuminous cells. The sieve tube threads which occur only in the radial walls always show a median dot. The existence of threads in the xylem is doubtful. All living parenchymatous cells show them, but it seems probable that they quickly disappear when the cells become lignified. In the case of young bordered pits there is softie evidence that the torus is traversed by connecting threads which are soon obliterated. The leaf of Pinus sylvestris shows a distribution of connecting threads similar to that noticed in the cotyledon. The endodermis is seen to be an important layer connecting the tissues of the stele with those of the cortex by means of thread groups in the tangential walls. In the pericycle there are both dead and living cells, but no threads persist in the walls connecting the dead with the living cells. The albuminous cell thread groups are very well developed, and their function and peculiar properties are discussed. In conclusion, the general distribution of the connecting threads throughout the tissues is considered. " On the ' Blaze Currents ' of the Frog's Eyeball." By A. D. WALLER, M.D., F.R.S. Received December 6, — Read Decem- ber 6, 1900. (Abstract.) The normal electrical response to light is positive. The normal electrical response to every kind of stimulus is positive. The normal response of the frog's eyeball is partly retinal, partly by other tissues. The direction of response is reversed by pressure. The normal " blaze currents " excited by single induction shocks, and by condenser discharges, are comparable with the normal dis- charges of an electrical organ. Their maximum voltage is of the same order as that of the discharge of a single electrical disc (over 440 ])r. A. l>. Wall,-.-. OO3 volt). Their magnitude and duration increase with increased strength of excitation. Summation of stimuli, summation of effects, staircase increase, and fatigue decline are manifested by Maze currents. Stimulation of excessive strength abolishes them completely, hut only temporarily. The energy of a blaze effect may considerably exceed the energy of its exciting cause. The effects are observable for at least five day- after excision of the eyeball ; they appear to be diminished under pro- longed illumination, and increased under prolonged darkness. Fio. 1. o-oot voU. Excitation. Polarisation. " Blaze." mm. Positive response to a single induction shock sent through the ejeball in the positive (upward) direction. The influence of raised temperature and increased pressure is studied, and under the influence of the latter four types of response are recorded. Comparison is made between blaze currents and the responses of electrical organs as described by du Bois Keymond. During and after maximal blaze the resistance diminishes: the diminution is not irreciprocal. If single electrical currents are passed through a normal eyeball and a galvanometer, in a " homodrome " and in a " heterodrome " direction (i.e., with and against the direction of normal discharge), the homo- drome (positive) deflection is greater than the heterodrome (negative) deflection. This inequality is the result of positive blaze current, and is abolished by death or strong totalisation. In the latter case the al)olition is temporary. The normal electrical response to light persists undiminished at ;<. time when blaze currents have been abolished by tetanisation. On the other hand, blaze current may be present in an eyeball giving n<> response to light. The altered state of the eyeball in relation to light does not necessarily run pa-allel with its altered state in relation to electrical stimuli. On the " Blaze Currents " of the Fror/s Eyeball 441 Direction of Direction of exciting — current. organ response. Living. Dead. ^ Dorsuin Torpedo - Venter i Dorsum -4 "I +1 + t ^ Venter Cornea + | + f Eyeball . Fundus Cornea 1 1 -1 -1 ^ Fundus *t + t FIG. 2. G&Lv&nometer. ± " Inducbor-ium. Plan of circuit. 442 IW. M. <'. Potter. On a Bacterial *' Ou a Bacterial Disease of the Turnip (Brassica net-pus)." By M. C. POTTER, M.A., F.L.S., Professor of Botany in the University of Durham College of Science, Newcastle-upon- Tyne. Communicated by Sir M. FOSTER, Sec. R.S. Received November 15, — Read December 6, 1900. In the autumn, when the activity of the turnip plant is mainly devoted to the storage of reserve material and the characteristic roots are increasing in size, it is not uncommon in this neighbourhood to find among the plants still growing in the fields some whose roots are quite rotten and with a highly offensive and peculiar smell. The plants thus affected can be recognised by the drooping, yellow- ish leaves, the older leaves being the first to show any indications of disease. They gradually flag and droop to the ground, at the same time becoming yellow and shrivelled in appearance. The leaves next in age gradually exhibit the same signs of premature decay, and this proceeds until finally the young leaves at the growing point succumb. The time taken for the collapse of the leaves naturally varies with different individuals, but it is usually about two weeks from the time of the first infection. The roots of these plants when examined present a very character- istic appearance. The decaying portion may be of a greyish-white or dark-brown colour, and is quite soft to the touch ; the cell wall has lost its natural firmness and the cells their turgidity, and with the escape of the cell-sap the tissues have been reduced to a soft watery pidp. In the particular disease now treated, the portion attacked remains of a whitish colour, and I have therefore described it under the name " White Rot," as my investigations have shown that this form of rottenness is due to a specific organism producing this particular colour when attacking a root. The brown and other discolorations found in similarly diseased roots are probably due in part to this organism, together with others, but I have not succeeded in cultivating the " Brown Rot," and this awaits further investigation. The disease can be readily communicated to sound roots, it being sufficient merely to make a slight incision and smear a small portion of the rotten mass upon the injured surface for decay to be imme- diately set up. In twenty-four hours the previously healthy cells around the inoculated surface show the characteristic changes of form and colour to a depth of about a quarter of an inch, indicating the progression of the decay. Keeping the plant under observation, with- out further injury, it is noted that the rind bordering on the wound gradually becomes soft and assumes a lighter colour ; the discolora- tion gradually extends ; the older leaves, too, droop and change colour ; Disease of the Turnip (Brassica napus). 443 by degrees the entire rosette of leaves perishes, and the whole root becomes a soft, putrid mass, which eventually collapses, and after a shower of rain almost entirely disappears, exactly the same symptoms appearing as in the case of the plants found decaying in the fields. The most careful microscopic search has failed to detect any trace of hyphae of the higher fungi in the decaying mass, but only a swarming mass of bacteria. The tissues are completely disorganised (see fig. 1), the cells separating from each other along the middle lamella, the cell- walls are soft, swollen, and faintly striated, the protoplasm too has lost its natural colour and become slightly brown and contracted, so that it no longer remains closely in contact with the cell-wall. With a view to determine whether the bacteria are the cause of the rottenness, and if so, to isolate the particular organism which produces it, a series of cultures was undertaken. In the first instance, a nutrient broth made from turnips was em- ployed. Pieces of turnip finely chopped were steamed in a beaker until soft, sufficient tap-water being added to just cover them ; when soft they were pressed through a cloth and the liquid filtered. To the clear light yellow filtrate thus obtained 5 per cent, of gelatine was added, and the mixture was then steamed, filtered, and drawn into test- tubes, which previously had been plugged with cotton-wool and exposed to a temperature of 140° C. for half an hour. These test tubes, containing about 10 c.c. each of the bouillon, were next steamed for half an hour on three consecutive days, and as a further test of complete sterilisation they were incubated at 20° C. for a few days. No colonies were found to develop. (Whenever mention is made of test-tubes containing nutrient gelatine it must be understood that all have been prepared in this manner, and none have been employed which have not been submitted to these tests.) In some cases the broth was neutralised, in others it was allowed to retain the natural acidity of the cell-sap; but subsequently Koch's bouillon, neutralised with sodium hydrate by the phenolphthalein test, was found to give the most satisfactory results, and hence was always used. In separating the various organisms found in the rotten mass a sterile platinum wire was introduced into the turnip (the rotten part practically offering no resistance), and then immersed in a test-tube (A) containing about 10 c.c. of the liquid nutrient gelatine. From this a loop was taken in a similar manner into a second test-tube (B), and so on until a sufficient degree of attenuation was reached. The test-tubes after being well shaken were turned out into petri capsules a, b ... g respectively. These were placed in a cool incubator, and the colonies allowed to develop. In a, and often in b, the entire surface became covered with growing colonies too thickly crowded to be of any use for the purpose of isolation ; but in the others the colonies were less numerous and sufficiently distinct to allow the organisms to be sepa- 444 1W. M. <'. I'. .tier. On <> ]:,',•(• r'ml rated from each other. The most conspicuous colonies were those which liquefied the gelatine ; among others producing no liquefaction Miwtfoccus candicans and a yeast were especially noted for their frequent occurrence ; but no trace of any of the higher fungi was found. The colonies were next transplanted by means of a sterile platinum wire into test-tubes containing about 10 c.c. of nutrient gelatine ; and after numerous trials 1 was satisfied that pure cultures were obtained. The various organisms as isolated were sown by means of a freshly- heated platinum wire upon sterile but living blocks of turnip. To prepare these blocks the turnips were first washed, and then soaked in a 1 per cent, solution of corrosive sublimate to destroy any organisms adhering to the outer surface, the corrosive sublimate being afterwards thoroughly washed away by means of water sterilised by discontinuous boiling. The rind was then removed by a sterile knife, the turnips l)eing cut into suitable blocks on a sterile plate and quickly inserted in the test-tubes. Treated in this way the blocks of turnip, while quite sterile, were composed of healthy living cells, as was shown by three sets of control tubes. In the first set the blocks, prepared as above, were immersed in cooled liquid nutrient gelatine, in the second similar blocks were immersed in sterile water; in neither case were any colonies found to develop either when the blocks were partially or wholly submerged, and after eight days no sign of decay had appeared. In the third set the blocks were simply inserted in the tubes, and kept in a damp atmosphere ; on microscopical examination cell division was observed to have taken place in the outer layers of uninjured cells, and the cell tissues presented a normal and entirely healthy appearance. In the tubes containing the inoculated blocks many showed signs of advanced decay in about twelve hours, and all those in which any rottenness appeared were carefully noted. After repeated experiment and a long series of cultures, I succeeded in isolating a bacterium which liquefies gelatine, and which, when sown on the sterile blocks of living turnip, produced the characteristic '• White Rot " previously described. The isolation of the bacterium in this manner was further confirmed by pricking out the colonies by means of Unna's harpoon. Small colonies of about 15 p growing in a petri capsule were selected and transplanted by the harpoon into petri capsides containing some sterile turnip bouillon. A specially fine harpoon needle was obtained, but the point was still larger than these very small colonies, and it was only after some practice that they could be successfully transplanted. The colonies selected were those growing quite apart, which appeared to have arisen from a single bacterium, to eliminate as far as possible any chance of the needle touching more than one. Lest, however, even these small colonies might have grown from more than one bacterium, Disease of the Turnip (Hrassica napus). 445 a single bacterium was selected and its development watched with the capsule fixed under the microscope until the colony was sufficiently large to transplant. Cultivations were also made by the method of the hanging drop. A drop of gelatine bouillon from a test-tube con- taining a very few bacteria was placed upon a sterile coverslip, and then inverted over a sterile growing cell and examined under the microscope. If the bacteria were too numerous, the preparation was discarded and trials made until a hanging drop was secured with only one or two bacteria. The growing cell was now fixed under the micro- scope, so that a selected bacterium could be observed and the growth of the colony noted. When sufficiently large the coverslip was quickly inverted and the colony removed by the fine Unna's harpoon to a petri capsule. In this way pure cultures were obtained, grown from a single bacterium, which always gave rise to the characteristic " White Eot," and left no doubt that this bacterium is the sole organ- ism concerned in the disease. Pure cultures were also sown upon plants growing in the College garden with exactly the same result. The decay commenced at the point of infection and soon spread through the sound roots, eventually producing the same white putrefying mass of rottenness. The bacterium can live for many generations as a saprophyte without losing its virulence as a parasite. A stock obtained from a " white- rotted " turnip growing in a field near Newcastle on September 10th, 1898, was isolated during that month, and after passing through several cultivations in successive test tubes was finally put aside on April 29th, 1899. On August 23rd two sound turnips were selected in the College garden, and while still growing, the part of the roots .above ground was washed with corrosive sublimate and afterwards with sterile water ; a wound was then made with a sterile knife, and a little of the culture from one of the test-tubes left undisturbed since April 29th was introduced by a platinum wire. The turnips were then covered over with a zinc cylinder, and, upon examination five days after, on August 28th, the rot was found to have penetrated deeply into the tissues, the larger half of the roots having become completely rotten with all the distinctive characteristics of the true " White Rot." In order to ascertain the precise action of the bacterium, and to determine whether it produced any ferment capable of acting upon the cell-wall in a manner similar to those of various parasitic fungi, • the method of precipitation by alcohol was adopted. A litre flask was plugged, sterilised, and then filled about half-full with sterile blocks of turnips, to which was added a small block upon which a pure culture of bacterium had been sown ; a little sterile water was then introduced, the flask closed as quickly as possible, and then well shaken to distribute the bacteria. In twenty -four hours many of the "blocks showed the characteristic action of the bacterium, and in the I it; I'm!. M. ' . l'.<;t«T. On a Ji«,tn-inl course of three or four clays nearly the whole contents had become rotten. The next important step was to separate the bacteria from their products. The contents of the flask were turned out and pressed through a cloth into a glass cylinder to remove the coarser portions, the turbid liquid was then filtered, and afterwards diluted with four to five times its bulk of alcohol. Almost immediately on addition of the alcohol a cloudy precipitate formed, and, at the end of twenty-four hours, a copious flocculent precipitate was deposited. After filtration the precipitate was washed with absolute alcohol, dried, carefully collected, and then digested with distilled water for about three hours. The solution was then passed through a Pasteur-Chamberland filter fixed in a Maassen's bacteria filter. In this manner a clear, pale, straw- coloured liquid was obtained free from bacteria. The liquid when drawn into sterile test-tubes remained clear for any length of time, but when exposed to the air it soon became turbid. A series of ten such sterile test-tubes was prepared, five of which were held over a Bunsen burner, and the fluid allowed to boil ; the other fivQ were left without any exposure to heat. Thin sections cut from sterile blocks of turnip, by means of a razor steeped in boiling water, were taken off in sterile water and quickly introduced both into the boiled and unboiled fluids. The action of the unboiled fluid was very marked. Fig. 1 shows a Fig. 1.— Group of cells from a section of turnip which has been exposed to the action of the cjtnse for twenty-four hours. The cell-walls are swollen and irregular in outline, and the cells are separating along the middle lamella (Zeiss, E. oc. 2). section taken from one of these preparations after twenty-four hours' exposure : the cell-wall is swollen and striated, and so much softened that great difficulty was found in handling the section and removing it to the slide ; it is well seen that the walls have quite lost their Disease of the Turnip (Brassica napus). 447 natural firmness and clear regularity of outline, being bulgy and dis- tended in places ; the dissolution of the cells is very apparent along the middle lamella, and the whole appearance of the section corre- sponds exactly with those taken from turnips found affected by the rot in the fields. The sections contained in the boiled fluid exhibited none of the appearances described above, and the cell-walls remained perfectly normal. It is thus evident that the bacterium secretes an enzyme which dissolves the middle lamella and causes the softening and swelling of the cell-wall. Fig. 2 represents a single cell from a Fig. 2. — Cell immersed for sixteen hours in an unboiled solution of the cytase. Thickness of cell-wall, 2 p. at x x (Zeiss, E. oc. 2). section immersed in the filtered, unboiled liquid for sixteen hours. Fig. 3 shows one after an immersion of forty hours. The thickness of Fig. 3. — Section immersed for forty-two Lours in unboiled solution of the cjtase. Thickness of wall, 5'3 p. at x x (Zeiss, E. oc. 2). the walls was 2 /u, and 5 -3 /A respectively. (I should remark here that these sections were cut out of season from old turnips in which the walls would be more resistant, and this would account for the; rela- I is prof. M. C. Potter. 0,i " Bacterial lively slow development. In sections from more succulent roots the walls have been found to swell from 2 p. to 7 /* in the course of twenty-four hours.) In fig. 4 the cell is drawn from a section im Fig. 4. — Secli >n immersed in solution of cvtase for forty-two hours, whose power had been destroyed by boiling. Cell-walls quite normal (Zeiss, E. oc 2). mersed for forty hours in the boiled liquid. The cell-wall is not per- ceptibly thickened or affected in any way. The activity of the enzyme in the decaying plant was also shown by passing the juice from the bruised pulp directly through a Pasteur- Chamberland filter, when its action on the cell-wall was precisely that described in the case of the watery extract of the alcoholic precipitate. The bacterium also secretes the enzyme when growing in a l»eef solution. Small flasks containing 100 c.c. of beef bouillon, inoculated with a pure culture, became turbid in the course of twenty-four to thirty-six hours. After an interval of eight days, the liquid was filtered and diluted with five times its bulk of alcohol, when a precipitate immediately began to appear. After standing twelve hours the precipitate was collected by filtration, dried, and then digested with 10 c.c. of distilled water. After filtration through a Pasteur-Chamberland filter, experiments were repeated as above with sections of sterile turnip, and the same results were obtained ; tin- liquid was found to possess the property of dissolving the middle lamella, and causing the softening and swelling of the cell-wall. All action of the ferment was destroyed by boiling. To avoid the tedious process of the filtration through a Pastcur- C-hamberland filter, and the necessary sterilisation of the apparatus, various attempts were made to render the solutions aseptic by the use of such re-agents as chloroform, thymol, formalin, &c. But this process had to be abandoned, as in all these cases living bacteria were found -after twenty-four hours, and no reliance could be placed upon it. In the early stages of the investigations, filtration — except when of the T (i r iii ( i (I'>rassica uapus). 449 the elimination of bacteria, was desired— was effected through ordinary filter paper. The quantity of the precipitate being small, the portion of the paper upon which it was deposited was cut out and digested with water. But in order to avoid any possible action of the enzyme upon the paper, kieselguhr was subsequently invariably employed, a few pieces of glass at the base of the funnel covered with a little asbestos serving to prevent the kieselguhr from passing through, the necessary pressure to ensure nitration being derived from an air- pump. The filtered extract from the rotten turnip also contains a diastasic ferment. Two test-tubes, each containing 5 c.c. of the dissolved ferment, were diluted with 5 c.c. of a 1 per cent, starch emulsion, one of the test-tubes having previously been boiled. After twenty-four hours the test-tube with the unboiled ferment showed no starch reaction on the addition of iodine ; but the boiled tube at once gave the character- istic blue. Similar diastasic enzymes are excreted by several other bacteria (Lafar). It will be convenient here to say that, adopting Migula's classifica- tion, I have ventured to name the bacterium I am describing Pseudommas destructans, though the description will be given later. It has been established that P. destructans, both when living in a nutrient solution and on a living turnip, excretes an enzyme which has the power of dissolving the middle lamella and of causing the softening and swelling of the cell-wall. As a further result of the bacterial action, as already described, the protoplasm of the cells is- found to have contracted, become brown, and separated from the cell-wall, showing evidence of the action of a toxin secreted by the bacterium. The same effect was produced in living turnip cells when treated with the boiled pressed juice of a turnip, which had become rotten through the influence of a pure culture of P. destructans. The pressed juice was filtered and about 10 c.c. drawn into test-tubes, which were then plugged and sterilised by discontinuous boiling. Sections cut by a razor, sterilised by boiling, from blocks of sterile, living turnip (p. 444) were quickly transferred to the boiled juice after it had cooled; at the same time similar sections were immersed in test-tubes containing the same quantity of sterile water. After twelve hours a very marked contrast was observable between these sections. In those immersed in the sterile water the cells presented the normal appearance, with the protoplasm pressed close to the cell-wall, while in those in the boiled pressed juice the protoplasm was dead, had assumed a brown tint, and contracted away from the cell-wall. A toxin, therefore, which is not destroyed by boiling is secreted by P. destructans. In his paper " Ueber einige Sclerotinien und Sclerotien-Krank- VOT.. LXVII. 2 K 450 !W. M. ('. I'ntUT. <)n •• heiteu," de Bary has shown th;it oxalic acid is secreted by the hypha' «>f Pi-.i-.n .<>•!> r,>f in, a in when living a- a parasite, and that thi* a.-iil as a toxin in killing and plasmolysing the protoplasm. Wehmer has found that ./>•/» /•/. ,/lini,-,!,,, also form oxalic acid when growing in a sugar-containing solution. Oxalic acid, being an unavoidable product in the metabolism of the higher plants ;i"(l ;dso in some fungi, it seemed reasonable to suppose that it might be found as a similar product in the life of bacteria. With this idea I tested the juice expressed from a rotten turnip, and found, on addition of calcium chloride, a precipitate which proved to be calcium oxalate. Cultures were then undertaken to test for the presence of oxalic acid as a product from P. -tteen kept sterile during the same period, remained quite clear on treatment with the same re-agents. P. 7/W«» /*.•>• thus sets up an oxalic fer- mentation in a sugar containing liquid. It has also been found that carbon di-oxide is given off during the proci When treated with alcohol, the Pasteur solution, in which P. dt /"/(s had been growing for eight days, yielded a white flocculent precipitate which contained the cytase. The oxalic acid, however, remained in solution, and was deposited as the calcium salt on addition of calcium chloride. This calcium precipitate, when mixed with manganese di-oxide and treated with sulphuric acid, yielded carbonic acid, which furnishes a further confirmatory test of the presence of oxalic acid. The precipitation by alcohol affords a ready method of separating the toxin (oxalic acid) from the cytase, and this explains uf f/ic T n i- iii ^j (Brassica napus). 451 why the sections treated with the watery extract of the alcoholic precipitate exhibited no marked plasmolysis. That the oxalic acid formed by P. <.l<'4rndan* in Pasteur's solution acts as a powerful toxin was very clearly shown. Six plugged and sterile test-tubes were prepared, and about 10 c.e. of the fermenting Pasteur solution was poured into each. To three of these (series 1) sufficient calcium carbonate was added to neutralise the oxalic acid. Both series of tubes were then sterilised by discontinuous boiling, during which process the cytase would be destroyed, and into both when cool freshly cut and sterile sections of turnip were placed prepared as described on page 446, and the solutions allowed to act till next morning. The sections in the second series of test-tubes showed a marked contraction of the protoplasm, and it looked brown and dead, and showed no tendency to return to its normal condition when immersed in pure water. In the first, which were treated with calcium carbonate, the protoplasm was quite normal, and exactly resembled a section which had been immersed in sterile water for the same period. A third set of test-tubes were filled with about 10 c.c. of the solution; these were not boiled, and received no calcium carbonate; the sections introduced showed complete dissociation of the cells, the cell-walls greatly swollen and the protoplasm very strongly contracted. This experiment with the Pasteur solution demonstrated the produc- tion of the same cytase, and strikingly illustrated its effect upon the plant cell, as well as the toxic action of the oxalic acid ; even more so than was the case with the same experiment with turnip juice. In considering the effect of the oxalic acid upon the cells, it is important to note that calcium pectate, a salt which is decomposed by oxalic acid with the production of calcium oxalate, enters largely into the composition of the middle lamella. Wehmer has shown that in the cultivation of Asperyillus niyer and Pmidllinni ijlmwnn oxalic acid is formed in saccharine solutions, that the oxalic acid produced acts as a toxin to these fungi, and gradually diminishes their vigour, and that when a certain strength has accumulated no further development is possible; growth, however, is resumed when the oxalic acid is neutralised by a calcium salt. The reaction between the oxalic acid produced by P. destrnctans and the calcium pectate of the middle lamella is preciselly analogous : the oxalic acid would be neutralised, and the pectate replaced by the oxalate, and the continued growth of the bacteria would thus be rendered possible. The oxalic acid* then both acts as a toxin in killing the cells aiid may also play some part in * Since the above account of the formation of oxalic acid by P. destructan.t was written, Zopf lias published a note also describing the formation of oxalic acid by B. xylinum, " Oralsaurebildung durch Bacterien," ' Berichfce d. D. Bot. Gesell.,' Feb. 1900. 2 K 2 452 Pr..f. M. < . h.ttrt. On " i the destruction of the middle lamella and the separation of the cells. Fig. 5 shows a cell swannini; with /'. : (Brassicu napus). 453' organism produces oxalic acid, which acts as a toxin to the protoplasm and, decomposing the calcium pectate, furthers the dissolution of the cells ; and also there is the secretion of a cy tase, which has a destructive action upon the cell-wall and intercellular substance. The question of the parasitism of the bacteria thus stands in these respects on the same platform as that of the fungi, and a complete homology is established between them. At first I experienced considerable difficulty in staining the flagella. Loeffler's method was first tried, but with no positive results ; it enabled one, however, to notice two deeply stained portions at either end of the rod. Van Ermengen's method also failed in spite of the strictest attention to technique, but by gradually increasing the strength of the silver nitrate, and finally iising a 2 per cent, solution, the desired result was obtained, and the bacillus was then found to possess one polar flayelluiu (fig. 6). It should be mentioned that the practice of passing the cover-slip thorough a flame was discarded in favour of drying the cover-slip at 60° C. in a water bath, the latter method being more certain and giving better results. Fig. 6. — Pseudomonas destritctans with single polar flagellum. (Swift's 1 /12th apochvoinatic and compensating ocular 12.) Pmidomonas is aerobic. "A stab culture rapidly develops along the track of the wire, forming a funnel ; the edge of the funnel reaches the sides of the test-tube in about forty-eight hours, and gradually sinks as the gelatine becomes liquid. The gelatine, however, is never wholly liquefied, the liquefaction extending down the sides of the tube only to a depth of about one and a half centimetres. If a layer of gelatine is immediately poured above the stab and the test-tube placed in the incubator, the track of the wire is clearly marked put as before, but the colonies soon cease to develop, and all growth ceases after three days. The tube may be kept for many weeks in this condition. Again, so far as my experiments show, the action of PseudonunuM upon turnips and potatoes only takes place in the presence of oxygen. The following are typical examples of experiments frequently repeated and always with the same results : — A flask holding about 250 c.c., with a tightly fitting indiarubber cork perforated to admit two glass tubes bent at right angles, was sterilised in the following manner. The tubes were plugged at each end with cotton wool, and the plugs pushed well into the tubes, the flask being also plugged with cotton wool, and, together with the glass tubes, sterilised by dry heat. Mean- 4"t Pwrf. M < P.-tti-r. fi,i n ],, while the indiarublH-r cork wa- boiled f,,r half an hour in a 10 per cent. >olution of contrive >ublimatc. The flask having cIock> of living turnip, prepared a> de-"ibrd above, and inoculated with a pure culture of /'. '. The indiarublier cork, after U-ing washed in sterile water, was quickly in>erteil into the Mask, the glass tube l>eing pushed through the perforations and the juiu-tions >ealed with melted wax. The longer tuU- (A) i-eached to the liottoin of the flask, the shorter (B) only slightly protruupply- ing the requisite pressure. When the' longer tulie A was left open, and a sufficient supply of oxygen could diffuse into the flask, carbonic acid was continually given off, and in the course of al»out n week the contents l>ecame entirely rotten and reduced to a watery mess. When, however, in a precisely similar flask used as a control, the longer tul>e after a short interval was closed, and the shorter connected with a tube for collecting any gas given ott', thus cutting oft' the supply of oxygen, the evolution of CO.. soon ceased, and, as far as could l>e observed, the action of P. tkstntcfau ceased also. To carry this |>oint a step further, and to ascertain more definitely whether the action of Ptmdomena* could take place in the absence of oxygen, another series of flasks wa> fitted up with the two tul»es as already descril>ed, the same precautions as to sterilisation* l»eing adopt ei 1, and the prepared blocks of tin nip introduced and inoctila ted as l»efore. The shorter tiuV was now connected with a second flask containing an alkaline solution of pyrogallic acid, and the other with a l>ent tu1»e containing mercury to act as a manometer, and prevent any access of oxygen from the air. The first result noticed was an expul- sion of the air in the flasks, the mercury rising in the distal limb. The mercury continued to rise, bubbles of carlx>n dioxide eventually ex-aping round the l>end. This action, however, cease* 1 in the course of two day*, the available supply of oxygen in the flasks and inter- cellular sjKices Injing exhausted. After a long interval (four months — .lune G to ()ctol>er 5) the flasks were disconnected, and the turnip blocks examined. They still retained their original shape, and were only rotten sujKjrficially ; the pieces had somewhat lost their rigidity, but ottered considerable resistance when stretched. Microscopic examination showed all the cells to be dead, but it was only one or two layers of superficial cells which showed any evidence of bacterial action. The cell-walls on the outside of the block were swollen and striated, and could l>e readily separated along the middle lamella ; the X/' of the. Tt were observed to escape round the bend, and there was no indication of the rot. Character* of Pseudomonas Destructans. Ifn/'if. — On growing turnips producing a " White Kot " in the living tissues. Morphology. — Short motile rods, 3 n x 8 p, with a single polar fla- gellum. Culture* can only be made in the presence of oxygen. Gelatine.- — -Petri Capsule*. Forms circular colonies of whitish-grey liquefying gelatine. Sfal> Culture*. Grows rapidly along the track of the wire, forming a funnel-shaped tube of liquid gelatine, with a white, cloudy deposit in the liquid portion. A(i«c. — -White, glazy growth. Tamil*. — Grows rapidly as a parasite. Potato awl Carrot. — -Same effect as on the turnip. . — No growth as a parasite. -Koch's bouillon and turnip ; becomes cloudy and opaque. Ferine-lit.*. — A cytase, causing the swelling and softening of the cell- wall, and dissolution of the middle lamella. A diastase. A peptonising ferment, producing liquefac- tion of gelatine. Toxiit. — Oxalic acid formed as a product of metabolism in turnip- juice and in Pasteur's solution containing cane sugar. Stain*. — Keadily stained with the ordinary aniline dyes, but not with Gram's method. Jt'eaction. — Kesidual product always acid. Copious evolution of carbonic acid during the fermentation. Among various bacteria at present noted as causing plant diseases, that described by Kramer as attacking the potato (Nassfaule) approaches most nearly to the one which is the subject of this paper. Kramer's i:,i; IW. M. ('. 1 'otter. On " Baatorial bacillu- agin - in liquefying gelatine vei y rapidly, ami it dcy> the middle lamella, and finally the cell-wall. The size of the W-terium as given by Kramer is from 2'5 /x long, and 7-S/t broad, very nearly the same dimensions as those of the turnip bacterium. Kramer, how- ever, has not named his bacillus, and he makes no mention of the Hagellura. He describes two stage* in the decay of the JX-T First, tan acid stage, during which butyric acid and • -ai -Innm- arid are given off; in this stage the sugars, then the intercellular substance, and finally the cell-walls are destroyed : the starch is not attacked. Subsequently, the proteids are broken up with the formation of ammonia, methylamine, trimethylamine, and other products : in this stage the acids are neutralised. In the action of 7'. ikttmrfu H* \\\»\\\ turnips and potatoes carbonic acid is given off, and the reaction of the pulp is always acid. On referring to a chemical friend, he could not definitely state that butyric acid, methylamine, ami trimethylamine are also produced ; he was of opinion that they were present, but that the decomposition is of a more complicated nature. P. es the action of bacteria in the bacteriosis of carnations, to an en/.yme, but without isolating it. Till quite recently I was unaware that any one had isolated from the bacteria an enzyme capable of attacking the middle lamella of living cells, and thus causing a plant disease. Laurent's valuable paper, " Recherches Experimentales sur les Maladies des Plantes," I only obtained in August of this year. It was published in Decemlier, 1898, simultaneously with a preliminary paper I read at the University of Durham Philosophical Society ; but previously, as early as January, 1898, 1 made a brief report to the Royal Society embodying the results of my work, viz., the isolation of the specific bacterium causing the <>J flu 7'nr/ti/i ( Brnssiea juipus). 457 " White liot " of turnips, and the isolation of an enzyme which dissolved the middle lamella and caused softening and swelling of the cell-wall. The pressure of teaching has prevented my publishing the complete paper sooner. Laurent, in his investigations upon the potato and the causes of its greater or less resistance to bacterial disease, also established the existence of a cytase, which dissolved the middle lamella, rapidly softened the cell tissues, and caused the disaggregation of the cells. The organism which was the chief subject of Laurent's researches, B. eoli fommunix, is very rarely capable of living as a parasite upon potato-tubers and other plants. He states that it was necessary for the tubers to be deprived of resistance, by means of exceptional cultures, to enable the bacillus to develop upon the potato. From that point its virulence was increased by successive cultivations upon tubers of slight resistance, until varieties at first highly resistant ended by becoming invaded by the parasite. The virulence disappeared as soon as the microbe ceased to be cultivated on a living tuber, cultures in nutritive solutions sufficed to suppress the aptitude of the parasite, and henceforward it could only be restored after special preparation in alkaline solutions. P. desfrudoMf on the contrary, flourished on nutritive media and even after many cultivations could readily be inoculated from these on to pieces of living turnip, producing all the effects of the rot in about twelve hours ; cultures both on nutritive media and on the turnip also rapidly invaded the tissue of the potato. AVhether, therefore, it has any existence in a saprophitic form or not, it has evidently become strongly established as a parasite attacking the turnip, and probably is not confined to the turnip alone. Wehmer has recently attempted to show that bacteria are not parasitic in the case of the wet rot (Nassfiiule) of the potato, and that their action is only secondary. He maintains that bacteria only attack dead or unhealthy tissue, that the warmth and moisture of the damp chambers impair the health of the cells, and infection is only possible under conditions which renders the tissues morbid. The wet rot, Wehmer says, logins with a maceration of the tissues ; between the separating dead cells numerous small bubbles are to be seen and masses of a small rod-like schizomycete. The initial stage is one of pectin- fermentation, succeeded by cellulose fermentation. AVith these processes are associated two special forms of bacteria. Wehmer's description of the rotting tissues agrees with my own, but he makes no mention of the enzyme nor of cultures of the bacteria. His conclusions that bacteria are not parasitic cannot be accepted in view of the isolation of the special enzyme by Laurent and myself, and of my experiments proving the infection of sound, healthy turnips when growing under perfectly natural conditions. n,, ,< /. ' /' : From nunicrou.- oh.-ervation- in the field-. I have come to the4 coiirln-ion th;it /'. *fr»'f"/i* i- alway- introduced at a wounded surface. Kxcept in case- in which the dccav ha- proceeded to a large extent, the point from which the decay >pread> i- aUvay.- indicated by a \\oiuid in the epidermis and siilijacent ti— lie-. Thi- ol.-ervation is siip|Mirted by the failure to infect sound toots except by tir>t making a small incision, and from iiiuiierous trials it would ap|»ear that 1'. •//•,/, /«///> is ])owerles> to >et up decay unless placed in contact with the |>areiichyma-cells of the cortex. Wounds caused by various snails, slugs, and larva-, l»y which the bacterium could gain an entrance, are frequently to lie seen on the roots, and I have no doubt the bacteria gain an easy entrance by this mean-. That slugs can and do carry the various disease-producing organi.-m.- has l>een shown by Smith in the case of the callage In-own rot by ./'///«////«« >//> and the larva- of 1'lntin /'/v/.svW, and of the tomato In-own rot l»y the larva? of the Colorado hectic, (i. Wagner's e\)ierinient.s also conclusively prove- that the sjMn-esof various pirasitic fungi are very commonly distributed l>y snails. Uacterial disease (»f turnips is much more common than is generally recognised, and the one now descril»ed is often very destructive to the crops, not only in the field Init in store during the winter. On examining numerous specimens sent me for investigation, I speedily found that what is generally known as "finger and toe" or "grub," is liy no means confined to Plafmodiepfofm A/v/.W/v/. Init that many other organisms, either singly or in combination, play a very imjmrtant ]>art in the destruction of living tin-nips and swedes. Finger and toe is everywhere so pievalent that in considering the nature of turnip attack it is often too hastilv assumed that /'/. /»/>/>-.// i> the sole cause of the disease, and that the other effects are merely secondary. In addition to bacteria and /'/. /»>•//»•/'»•//, 1 have found the turnip and swede crops to be attacked by /'//.*//•////// and also by //"//•////>•, and it is probable these do not exhaust the list of vegetable parasites for this crop, I nit further research is necessary l>efore it is possible to separate the various organisms and assign to each its //•/'•. LITERATURE. Arthur, J. C. " Baeh-riosi* of Carnations," 'Bulletin 59, 1896, Agricultural Experiment Station, Purdue I'tiiversitv." de Bary. A. " Ueber einige Sclerotinien und Sclerotienkrankheiten," 'But. Zt-itung,' 1886. Friln-s. I^itar, 'Teehuic-al Mycology,' p. 197. Kruuier. E. " Bakteriologisehe UnU>r»uchungen iibcr die Xassfaule der Kartoffel- knollen," ' Bot. Centralblatt.' vol. 48, 1891. Lafar. ' Technical Myc-ologv,' j>. 192. Laurent, E. " Les maladies des plantes," ' Annales de I'lnstitut Pasteur," Dec., 1898, vol. 13, 1899. The •/// tin' Migula, W. 'System der Bakterien,' Jeint, 1807. Omeiianski. li Sur la fermentation de la cellulose," ' Cornptes Heudus,' 1895. Panimel. " Bacteriosis of Rutabaga," ' Bulletin 27,' Iowa Agricultural College. 1895. Potter. "White Kot of the Turnip," ' Proc. of the University of Durham Phil. Soc.,' 1899. Smith, E. F. " Pseudomonax cam/iesfris : the cause of Brown Rot in Cruciferous Plants," ' Centra Iblatt fur Bakteriologie.' Abt. 11, vol.3, 1897: ' Spread of Plant Diseases.' Massachusetts Horticultural Society, March, 1897. Van Seuus. ' Beitrage zur Kenntniss der Cellulosegahrung,' Leiden, 1890. Van Tieghem. '• Sur la fermentation de la cellulose'," ' Bull. d. 1. Societc Botimique de France,' 1S79. Wagner, Or. Quoted by K. F. Smith in ' Spread of Plant Disease.-.' Wehmer, C. ''Die Bacterienfaule (Nassfaule) derKartoffelknollen," 'Beriehte der deutseh. Bot. Gesellschaft,' vol. 16, 1998; " Ueber den Kinfluss der Tem- peratur auf die Eiitstehung freier Oxalsihire in Culturen von Aspergillus niger," 'Beriehte der deutsch. Bot. Gesellschaft.' vol. 9, 1891. Winogradsky. Lafar, ' Technical Mycology,' p. 197. Ward, H. Marshall. " .\ Lily Disease," ' Annals of Botany,' vol. 2, 1888. " The Micro-organism of Distemper in the I .)o<>-, and the Production of a Distemper Vaccine." By S. MONCKTON COPEMAN, M.A., M.D., F.It.C.P. Coinnuuiicated !>y Sir M. FOSTKI;. Sec. U.S. Ifeceived November 14, — Read December (i, 1900. (From the Brown Institution.) Distemper is so fatal a disease of dogs, more particularly of such as are highly bred, that a method of preventing invasion by the disease has always been a desideratum. As the result of investigations into the bacteriology of this disease, carried out in continuance of those commenced in my laboratory at St. Thomas's Hospital about ten years ago by the late Everett Millais, I find that the specific micro-organism concerned is a small cocco- bacillus, which stains with the ordinary aniline dyes, but is decolorised by the method of (.Ti-a-m. It grows readily on the surface of agar at body temperature • the individual colonies when isolated by the method of plate-culture having a greyish, glistening, semi-translucent appear- ance by reflected light, and a light-brownish tint by transmitted light. The general form is circular, but occasionally, and specially in primary •growths, the edge is somewhat irregular. The microbe also grows well in beef-broth, causing at first a general turbidity. Later on, a deposit falls to the bottom of the tube, and the supernatant liquid becomes somewhat clearer. In cover-glass preparations from broth cultures the bacilli are not unfrequently found united together to form chains, sometimes of considerable length. The bacillus is capable of growing, '/'//' .17 //•/•/.— /v/" ///>//< ••/ iti-it^nifH i- /,/ ///,• //, . though i-i.iuj);ii-;iii\c!y -lo\\ ly, on solidified blood-tenUB, and also in milk, which does not Income coagulated. On j»otato it develops with difficulty, l»nt now and again, after some day-: in- ul.atioii, a moist- looking streak of a jKile liufl' colour may lie observed. If gelatine '»e inoculated, growth occurs slowly at the temi>er;itnre of the room, and, after a time, the medium tends to become liquefied. Growth on agar can be carried on week after week for a great number of generations, but after a do/en removes or so, its morpho- logical and biological characteristics are found to have Income ~<>iur what altered. An account of these variations and of the |«itholo^ical histology of the disease I propose to publish subsequently. In similar fashion the pathogenetic projterties of the micro-organism appear to become gradually weakened, but by related intra-peritoneal inoculations in the guinea-pig its virulence may be regained. The injection lieneath the skin of the abdomen in a dog weighing 7 kilos, of 1 c.c. of a broth culture seven days old, derived in turn from an agar sub-culture, induced an attack of distemper, which terminated fatally in about a week from the time of inoculation. In a large number of other dogs experimented on by ^Jillais or myself a generally non-fatal attack has followed on inoculation of the nasal mucous membrane. Specially characteristic of the disease intentionally produced is the fact that the animal exhibits during the attack a marked and pro- gressive loss of weight. Of other symptoms of the malady so well known to all dog-breeders, those whiub are usually most marked are the result of more or less acute inflammation of the various mucous surfaces. On post-mortem examination I have generally found the whole respiratory tract to be specially affected, the lungs sometimes showing pneumonic consolidation throughout almost their entire extent. The trachea is apt to be congested, ami to contain a quantity of mucus, while the eyes and nose are blocked with a purulent or mueo-pnrulent discharge. By making agar plate-cultivations from the exudation from the lungs, from the tracheal mucus, or from the nasal secretion, the specific bacillus may be isolated — from the first two situations, often in almost pure culture. Examining animals which have died from distemper, whether result- ing from experimental inoculation or contracted in the ordinary fashion, I have never succeeded in obtaining cultures either from the blood obtained from the heart with aseptic precautions, or from tile liver, the gall-bladder, the kidney, or the spleen. Pressure of other work since joining the .Medical Staff of the Local (Government J-Joard. has prevented my having the opportunity of examining even inoculated animals at intermediate stages of the disease in severe forms, or, doubtless, it might have been found possible to isolate the bacillus from one or other of On the Tempering o^Jron haTdened*by O/v/-.-,-/, •"/'//. 461 these situations. In one instance, in which the blood-vessels of the brain were found to be much congested, inoculation of a tube of sloping agar with a large platinum loopful of cerebro-spinal fluid, well spread over the surface of the agar, resulted in the appearance of half a dozen isolated colonies of a pure culture of the distemper bacillus. By heating a broth culture of the bacillus at 60° C. for half an hour, and subsequently adding a small quantity of carbolic acid as a preservative, a vaccine is obtained, which acts in similar fashion to those devised by Haff kine and Wright for use in the prevention of plague and enteric fever respectively. The vaccine may be standardised after the manner originally suggested by Wright in connection with his work on enteric fever. The dose must obviously vary according to the size of the dog, but, as a guide, it may be mentioned that I have found, in three instances, that the injection of 2 c.c. of the sterilised culture of the bacillus is apparently sufficient to protect fox-terrier puppies weighing about H kilos, against attack by distemper, while an unprotected puppy in the same batch contracted the disease on introduction of an affected dog. I find also that guinea-pigs can be protected in this way against the effects of a dose of living culture, which would ordinarily prove fatal in about forty-eight hours. As regards the exact length of time, however, during which such protective effect may last, no definite statement can as yet be made, but a series of tests on a large scale are in process of being carried out by dog-breeders in this country, in Germany, and in America. " On the Tempering of Iron hardened by Overstrain."* By JAMES MUIR, B.Sc., B.A., Trinity College, Cambridge, 1851 Exhibi- tion Eesearch Scholar, Glasgow University. Communicated by Professor EWING, F.R.S. Received July 11, — Read De- cember 6, 1900. (Abstract.) It is well known that iron hardened by overstrain, for example, by permanent stretching, may have its original properties restored again by annealing, that is, by heating it above a definite high temperature and allowing it to cool slowly. Experiments described in the paper, of which this is an abstract, show, however, that if iron hardened by overstrain be raised to any temperature above 300° C., it may be partially softened in a manner analogous to the ordinary tempering or * The work described in this paper is a continuation of that already described in a paper by the present author " On the Recovery of Iron from Overstrain," ' Phil. Trans./ A, vol. 193, 1899. 4 Mr. .F. Muir. •• letting down i.i -H-d which ]};i< IK-CM haidened by «Mienchiiii; I'M mi ;i red heat. This tempering t'rom a condition »] hardne-s induced by overstrain, unlike ordinary tempering, is applicable not only to steel, hut al-o to wrought iron, and jxissibly to other materials which < an l>e hardened by overstrain and softened l>y annealing. The experiments described in the pa])er were all carried out on rods of iron and steel about ;'ths of an inch in diameter and 11 inches long, the elastic condition of the material l>eing in all cases determined by means of tension tests in which the hardness of the material w.-is indi- cated l>y the position of the yield-point. The straining wag performed by means of the 50-ton testing machine of the Cambridge Engineering Lalwratory, and the small strains of extension were measured by an extensometer of Professor Ewing's design, which gave the extension on a 4-inch length of the specimen to the 1 100,000th of an inch. For the purpose of tempering and annealing, a gas furnace was employed 2 feet in length, the specimens being protected from direct contact with the flame by inclosing them in a thick porcelain tube. The temperature inside this tube was determined by means of a Cal- lendar's direct-reading platinum-resistance pyrometer. The method of examining the materials employed is illustrated by the following two diagrams, in which the material examined i.s a J,-inch rod of semi-mild steel (0'35 per cent. C., 1 per cent. Mn).* Curve No. 1 of the first diagram shows that this steel when in the condition ;is supplied by the makers gave a well -defined yield-point at about 38 tons per square inch, the material yielding at that stress by 0'13 of an inch on a 4-inch length. Curve Xo. '2 illustrates the semi-plastic state of the material, pro- duced by just passing this primary yield-point. The specimen was laid aside for If days, then once more tested ; and Curve Xo. 3 shows the progress made during this interval of rest towards recovery of elasticity. Curve Xo. 4 shows the condition of the overstrained material after it had been resting for two weeks. To insure peiiect recovery of elasticity, the specimen was heated to 200° C., but a few minutes at the temperature of boiling water would have l»een nearly as effective in restoring the elasticity lost by overstrain. t After cooling, the specimen was tested by reloading and carefully increasing the load above its previous maximum amount till a well- defined yield-point was obtained at 49 tons per square inch, as shown by Curve Xo. 5, the yield-point having thus l>een raised by the large step of 1 1 tons per square inch. The yielding which occurred at this * Details of the special method adopted in plotting these diagrams will be found in the author's previous paper " On the Recovery of Iron from Overstrain,' ' Phil. Trans.,' A, vol. 193, 1899, p. 12. t Ibid., p. 2-2. On the Tempering of Iron li<(i- steps of 11 tons, fracture occurred at 60£ tons per square inch. The same steel, after annealing at 750" C., is shown DIAGRAM 2. (Steel annealed at 750' C.) Extensions . / Unit Curve 1. Primary test. „ 2. Shortly after 1. „ 3. Ij day's „ 1. 4. 2 weeks 1. of *n inch tJLJ Curve 5. After heating to 3003 C. »> o. ,, ,, ,, >• '• » »• » by Diagram 2 to have yielded at 29 tons per square inch, and finally to have fractured at 59i tons per square inch, after the yield-point had l>een raised four times by a step of about 7A tons per square inch. The i-iuch steel rod when in the condition as supplied by the makers was thus shown to 1x5 in a state of hardness possessing certain dis- tinctive properties. It was found that the steel in the condition as supplied could be tempered or partially annealed by heating to various temperatures lower than the ordinary annealing temperature of alxmt 750° C. — -••. The following table illustrates this tempering from the condition as supplied, the material being a rod of steel very similar to that referred On the Tempering of Iron hardened by Overstrain. 465 to above. The " steps " tabulated in the last column are the amounts by which the yield-points were raised in consequence of overstrain and recovery from overstrain : — Condition of the material. Yield-point. Extension at yield-point. " Step." 361 tons/in 2 0" '16 on 4 inches 11 tons/in." Annealed at 6003 C. . . „ „ 650° „ .. „ 700° „ . . „ 730° „ .. „ „ 780°,, .. 36 33£ 31 28i 24 0"-16 „ „ 0"'15 „ „ 0" '09 „ „ 0" -07 ve for steel were also obtained with Lowmoor iron. The hardening by overstrain and the tempering of soft Lowmoor iron only differed in detail from the analogous harden- ing and tempering of steel. The iron and steel employed in this research were also examined, when in various conditions of hardness, by means of the microscope, and micro-photographs are reproduced in the paper. The ordinary methods of relief polishing and of etching by dilute nitric acid were employed, and a new method of staining steel, by rubbing with ordi- nary moistened cocoa, was made use of and is described in the paper. <>n, the Tcini)c.i'iiij of Iron hardened ft// Overstrain. 465 to above. The "steps" tabulated in the last column are the amounts by which the yield-points were raised in consequence of overstrain and recovery from overstrain : — Condition of the material. Yield-point. Extension at yield-point. " Step." > . 36^ tons/in - 0" '10 on 4 inches 11 tons in.* Annealed at 600" C. 36 „ „ 650° „ 33i 0"'16 „ „ 9} , ., "00° „ 31 0"-15 „ „ 9 „ 730° „ 28i 0" -09 „ „ 8 ., 780° „ -t 0"-07 „ „ 7 Iii order to show the tempering of steel hardened by tensile over- strain, a specimen of annealed steel was overstrained in a manner analogous to that illustrated by Curves 1, 5, and 6 of Diagram 2. The material, after recovery from overstrain, had thus been brought into a condition of hardness, which enabled the specimen to be loaded, to 50 tons per square inch without a yield-point being reached. The specimen was then subjected to a series of tests after being heated successively to various temperatures, the result being to show that 310 J C. produced no softening of the material, 360" C. lowered the yield-point to 47 tons; 500°, 600", and 700° C. lowered the yield-point to about 40, 35, and 30 tons per square inch respectively. It was further shown that the same temperature brought the yield- point to approximately the same stress, no matter what might lie the original hardness of the specimen under test ; and that the harder the material was made by tensile overstrain — that is, the higher the yield- point was raised by permanent stretching — the lower was the tempera- ture which could be shown to produce a slight tempering effect. Thus in the above instance had the material been made harder (by further overstraining) than was shown by the elastic range of from zero to 50 tons per square inch, then possibly the temperature of 310' C. would have produced a slight softening of the hardened material ; a temperature of about 300' C. was, however, found to be the minimum temperature which had a tempering effect on the hardest condition of steel tested. The tempering effects which have been ascribed above solely to tem- perature, were found to be influenced to some extent by time. Thus it was found that by baking a hardened specimen for several hours at any temperature a greater effect was produced than by simply raising the specimen for a few minutes to that temperature. The effect of time was, however, small compared with that produced by increase of temperature. VOL. LXV1I. '2 L /'/•' • All the re.-ult- whii-h are o obtained with Lowmoor iron. The hardening by ovet.-fain and the tempering of soft Lowmoor iron only flittered in detail from the analogous harden- in^ and tem|>ering of steel. The iron ami steel employed in thi> resean-h \\ere al>o examined, when in various conditions of hardnes.>, by mean* of the mici" and micro-photographs arc reproduced in the paj»er. The ordinary methods of relief polishing and of etching by dilute nitric acid were employed, and a new method of staining steel. by rubbing with ordi- nary moistened cocoa. wa> made use of and i> described in the paper. 13, 1900. Sir WILLIAM MUGGINS, K.C.P.., D.C.L.. President, in the Chair. A List of the Presents received was laid on the tal»le, and thanks ordered for them. In pursuance of notice sent to the Fellows, an election was held to fill the vacancy upon the Council caused l»y the retirement of Sir John Wolfe Barry. The statutes relating to the election of the Council, and the statute relating to the election of a Meml»er of Council upon the occurrence of a vacancy, were read, and Professor Dewar and Mr. Godman having l>een. with the consent of the Society, nominated scrutators, the votes of the Fellows present were taken and Mr. Joseph Wilson Swan was declared duly elected. The President made the following statement concerning the Inter- national Catalogue of Scientific Literature: — ••As stated in the Report of Council presented to the Society at the Anniversary Meeting, the President and Council offered to l>ecome the Publishers of the proposed International Catalogue, on l>ehalf of the Intel-national Council, and to advance the capital sum needed to start the enterprise. " I have now the pleasure of announcing that the International Council of the Catalogue, which met yesterday and to-day in the rooms of the Society, has accepted the offers of the Royal Society, and that this great undertaking, which has for several years engaged the earnest attention and demanded the continued labours of the Royal Society, a? well as of other scientific bodies abroad and in this country, is now well on its way. The International Council has laid down all the necessary regulations, and piejKired all the necessary instructions, for carrying Oil t/tr Riwfrniii <>f flic more Volatile (rases of Air.' 467 out the task of collecting and editing ; and it only remains for those who are taking part in the preparation of the Catalogue, to do their best to secure that the Catalogue shall fulfil the hopes which have been raised. " It gives me great pleasure to make this announcement in the presence of several of our foreign brethren, whose co-operation has tended so much to the success of the enterprise." The following Papers were read : — I. " On the Spectrum of the more Volatile Gases of Atmospheric- Air, which are not Condensed at the Temperature of Liquid Hydrogen. — Preliminary Notice." By Professor S. 1). LiVEixo, F.R.S., and Professor JAMES DEWAR, F.K.S. II. " Additional Notes on Boulders and other Rock Specimens from the Newlands Diamond Mines, Griqualand West." By Pro- fessor T. G. BONXEY, F.R.S. III. " The Distribution of Vertebrate Animals in India, Cevlon, and Burma." By W. T. BLAXFORD, LL.D., F.R.S. IV. " Elastic Solids at Rest or in Motion in a Liquid." By C. CHKEE, F.R.S. The Society adjourned over the Christmas Recess to Thursday, January 17, 1901. '•' On the Spectrum of the more Volatile Gases of Atmospheric Air, which are not Condensed at the Temperature of Liquid Hydrogen. — Preliminary Notice." By S. D. LIVEING, M.'A.. D.Sc,, F.R.S., Professor of Chemistry, University of Cam- bridge, and JAMES DE\VAR, M.A., LL.D., F.R.S., Fullerian Professor of Chemistry, Royal Institution, London. Received November 15, — Read December 13, 1900. In August last some tubes were filled at low pressure by an im- proved process with the more volatile gases of the atmosphere.* The air was liquefied directly from that above the roof of the Royal Institution by contact at atmospheric pressure with the walls of a vessel cooled below - 200C C. "\Vhen about 200 c.c. of liquid had * In this paper we describe researches in continuation of those previously com- municated to the Society by one of us, in a paper entitled " Application of Liquid Hydrogen to the Production of High Vacua, together with their Spectroscopic Examination," ' Roy. Soc. Proc.,' rol. 64, p. 231. 2 L 2 I'roi's. S. I). I.iv<-in_f and .1. I>ewar. rondcii-cd. communication with the outer air \\.i- < lo-»-d >ty a stop- i -ock. Sul»equently, communication was ojKMied, Through another stop-cork, with a second vessel coolecl. When about 10 c.c. had condensed in the solid form in the -fund \e--cl. communi- cation with the first vessel was cut off, and a manometer showed a ;re of gas of about 10 to 15 mm. of mercury. This mixture of gases was passed into tubes previously exhausted by a mercury pump, but before reaching the tubes it had to pass through a U-tul>e immersed in liquid hydrogen so as to condense less volatile gases, such as argon, nitrogen, oxygen, or carbonic oxide, which might be carried along by those more volatile. Previous trials with tubes filled in the same way, except that the U-tul»e in liquid hydrogen was omitted, showed that these tubes contained traces of nitrogen, argon, ami com- pounds of carbon. The tubes filled with gas which had ]>assed through the T-tube showed on sparking no spectrum of any of these la-t- mentioned gases, but showed the spectra of hydrogen, helium, and neon brilliantly, as well as a great many less brilliant rays of unknown origin. In addition, they showed at first the brightest rays of mercury, derived, no doubt, from the mercury pump by which they had been exhausted before the admission of the gases from the liquefied air. After some sparking the mercury rays disappeared, probably in . onse- quence of absorption of the mercury by the electrodes, which were of aluminium. In one experiment the mixture of gases in the second vessel, into which a fraction of the liquefied air was distilled as above described, was pumped out without being pa.-sed through the I'-tube in liquid hydrogen and examined. This mixture was found to contain 43 JKT cent, of hydrogen, 6 per cent, of oxygen, and 51 per cent, of other -<••$ — nitrogen, argon, neon, helium, &c. — and it was explosive when mixed with more oxygen. This shows conclusively that hydrogen in sensible proportion exists in the earth's atmosphere, and if the earth cannot retain hydrogen or originate it, then there must l>e a continued accession of hydrogen to the atmosphere (from interplanetary space), and we can hardly resist the conclusion that a similar transfer of other gases also must take place. The tul>es containing the residue of atmospheric gases uncondensed at the temperature of liquid hydrogen we have examined spectroscopically. On passing electric discharges through them, without any condenser in the circuit, they glow with a bright orange light, not only in the capillary part, but also at the poles, and at the negative pole in particular. The spectroscope shows that this light consists in the visible part of the spectrum chiefly of a succession of strong rays in the red, orange, and yellow, attributed to hydrogen, helium, and neon. On tlic Spectrum of tJ«' more Volatile Gases of Air. 469 these, a vast number of rays, generally less brilliant, are distributed through the whole length of the visible spectrum. They are obscured in the spectrum of the. capillary part of the tube by the greater strength of the second spectrum of hydrogen, but are easily seen in the spectrum of the negative pole, which does not include the second spectrum of hydrogen, or only faint traces of it. Putting a Leyden jar in the circuit, while it more or less completely obliterates the second spectrum of hydrogen, also has a similar effect on the greater part of these other rays of, as yet, unknown origin. The violet and ultra-violet part of the spectrum seems to rival in strength that of the red and yellow rays, if we may judge of it by the intensity of its impressions on photographic plates. We were surprised to find how vivid these impressions are up to a wave-length 314, notwith- standing the opacity of glass for rays in that part of the spectrum. The photographs were taken with a quartz calcite train, but the rays had to pass through the glass of the tube containing the gases. We have made approximate measurements of the wave-lengths of all the rays which are sufficiently strong to be seen easily or photographed with an exposure of thirty minutes, and .give a list of them below. These wave-lengths are computed to Rowland's scale, and were deduced from the deviations produced by two prisms of white flint glass for the visible, and of calcite for the invisible, rays. The wave-lengths assigned to the helium lines are those given by Runge and Paschen, and some of these lines were used as lines of reference. In general, the iron spark spectrum was the standard of reference. The tubes when first examined showed the lines of the first spectrum of hydrogen vividly, and the earlier photographs of the spectrum of the negative pole contained not only the violet lines of hydrogen, but also the ultra-violet series as far up as A. 377. In order to get impres- sions of the fainter rays, exposures of half an hour or more were required, and a succession of photographs had to be taken so as to get different sections of the spectrum into the middle of the field, where measurement of the deviations would not be impeded by the double refraction of the calc spar. As the light of the negative pole only was required, the electric discharge was made continuously in one direction only, with the result that the hydrogen lines grew fainter in each Btiocessive photograph, and soon disappeared altogether. Along with the ultra-violet rays, the less refrangible rays of hydrogen also dis- appeared, so that no trace of the C or F line could be seen, nor yet of the second spectrum, so long as the current passed in the same direc- tion as before. lieversal of the current soon made the F line show again, so that it seems that the whole of the hydrogen was driven by the current to the positive pole. The conditions under which this ultra-violet series shows itself are a matter of interest. It appears here in the midst of a brilliant spectrum due to gases other than 470 i >. \>. Livri hydrogen, ami yet it i- very difficult to obtain a photograph of it when no gas but hydrogen is known to be present. or. at lea>t. to i>ec.>me luininoiis in the electric di>clia:_ We have had an opportunity of comparing the <]>c«-t!uni of the volatile residue of air with that of the more volatile part of ga* from the Bath spring. The tulie did not admit of the sejwirate examination of the light from the negative pole, but was examined end-on, so that the radiation probably included rays emitted from the neighbourhood of the negative jxjle. The whole of the hydrogen had been removed from the Bath gas, but not all the argon. In the spectrum of tl the rays of helium are dominant, decidedly stronger than tho-.- ..f neon, although the latter are very bright. In the sj>ectrum of the : esjdue of atmospheric air, the proportion of helium to neon seems reversed, for in this the yellow neon line is as much more brilliant than the yellow helium line as the latter is the more brilliant in the sjH?ctnini of Bath gas. All the prominent lines in the sjn-etrum of the volatile residue of Hath gas were also in that of the residue of atmospheric air, except the argon lines. There were, on the other hand, many lines in the latter not traceable in the former, some of them rather con- spicuous, such as the ray at about A 4(><>4. It is, of course, probable that such rays are the outcome of some material not contained in the Bath gas. A very conspicuous pair of lines appears in photographs of the spectrum of the air residue, at about A .'55S7, which is not traceable in the spectrum of Bath gas. The helium line, A .V»7 -I. is seen in the latter spectrum, but is quite obscure* 1 in the former spectrum by the great intensity of the new pair. This helium ray is really a close double, with the less refrangible continent much the weaker of the two, but the new pair are wider ajwirt, and of nearly equal intensities ; this < -haracter also distinguishes them from the strong argon line at A 3588-6. They are, however, very much more intense at the negative pole than in the capillary, and it will require a good deal more study to determine whether these rays, and many others which we have not tabulated, are due to the peculiarity of the stimulus at the negative pole, or to the presence of a previously unrecognised material. A> our mixture of gases prolwibly includes some of all such gases as pervade interplanetary ami interstellar space, we early looked in their spectra for the prominent nebular, coronal, and auroral rays. Search- ing the spectrum about A 5007 no indication of any ray of alnmt that wave-length was visible in the spectrum of any one of the three tubes which had l»een tilled as al>ove descril»ed. Turning to the other § nebular line at alnmt A 4959, we found a weak rather diffuse line to which our first measure a-ssigned a wave-length 4'.i"»s. The correctness of this wave-length was subsequently verified by measuring with a micrometer eye-piece the distances of the line from the helium lines A 4922'! and A 5015*7 which were in the field of view at the same On the Spectrum of the more Volatile Gases of Air. 471 time. The position of the line was almost identical with that of the ii'on spark line X 4957 '8, and the conclusion arrived at was that the wave-length was a little less than 4958, and that it could not be the nebular line. There remained the ultra-violet line A. 3727. Our photographs showed a rather strong line very close to the iron sp;urk line A 3727'8, but slightly more refrangible. As the line is a tolerably strong one, it could be photographed with a grating spectrograph along with the iron lines. This was done, and the wave-length deduced from measuring the photograph was 3727 -4. This is too large by an amount which considerably exceeds the probable errors of observation, and we are forced to conclude that the nebular material is either absent from our tubes, or does not show itself under the treat- ment to which it has been subjected. Although the residual gases of the atmosphere uncondensed at tha temperature of liquid hydrogen do not show the nebular lines, we found that another lube gave a ray very close indeed to the principal green nebular ray. This tube had been tilled with gas prepared in che same way as the others, with the exception that, in passing from the vessel into which the first fraction of liquid air was distilled, it was not passed through a U-tube immersed in liquid hydrogen on its way to the exhausted tube. In consequence it contained traces of nitrogen and argon, and when sparked showed the spectra of these elements as well as those of hydrogen, helium, &c. The nitrogen spectrum dis- appeared after some sparking, but the tube still shows rays of argon as well as those of the gases in the other tubes. On examining the spectrum of the negative pole in the neighbourhood of the prin- cipal nebular green ray, a weak ray was seen in addition to those given by the other tubes. It was found by comparison with the nitrogen rays A 5002-7 and X 5005-7 to be a little less refrangible than the latter of these rays, and by measuring its distance from the nitrogen rays and from the two helium rays A 4922-1 and A 5015*7 with a micrometer eye-piece, the wave-length A 5007 '7 for the new ray was deduced. This looks as if we might find the substance which is luminous in nebula? to be really present in the earth's atmosphere, and we hope shortly to be able to verify the observation of it. Turning to the coronal rays, our tubes emit a "weak ray at about A 5304. This is not far from the wave-length A 5303-7 assigned by Sir N. Lockyer to the green coronal ray. It is, however, greater than that assigned by Campbell, namely, 5303'26.* Other lines observed by us near the places of coronal lines are at wave-lengths about 4687, 4570, 4358, 4323, 4232, 4220, 3985, 3800. These are all weak lines except that at A 4232, which is of tolerable strength, and that at A 4220, which is rather a strong line. The wave-lengths 4323, 4232, 4220, and 3800 come very close to those assigned to coronal rays, but * ' As.troph. J.,' vol. 10, p. 190. 47l! I'rof>. S. \>. Liveiir.: .ind .1. hrwar. the others hardly conic within the limits of piobablc error. The ray •ll'L'u .>eem- tin. strong in proportion to the other*, but the >treni;th (»f that at »•_'.'!'_' seem- to .-i.-rord with the si i viigt I) <>f the rorresj>oiiding rav in the corona. It will In- >ccn tluit the rays we enumerate above correspond approximately to the stronger rays in Sir N. Lockyer's list.* Further mea.-iirc> of the wave-lengths of the faint Mm needed before we can >ay detinitely whether or no we have in our tubes a sulistance producing the coronal rays, or sonic of them. As to the auroral rays, we have not seen any ray in the sjiectrutH of our tubes near A r)~>71'~>, the green auroral ray. We have ob-er\ed two weak rays at A 4206 and A 4198 which may possibly, one or l>oth. represent the auroral ray A 420. The very strong ray of argon. X 4200*8, would make it probable that aigon was the origin of this auroral ray, if the other, equally strong, argon rays in the same region of the spectrum were not absent from the aurora. Nor have we found in the spectrum of our tubes any line with the wave-length 3915, which is that of another strong auroral line. On the othei hand it seems probable that the strong auroral line A .'!.">,«•; may be due to the material which gives us the very remarkable pair of line^ it about the place of N of the solar spectrum. A :{.r>87, which are very strong in the .spectrum of the negative pole, but only faint in that of the capillary part of our tubes. It may well be that the auroral dis- charge is analogous to that about the negative pole. "We have also a fairly strong rav at A .'5700, which may be compared to the remaining strong ray observed in the aurora A 3700. This, however, is a ray which is emitted from the capillary part of our tubes as well as from the negative pole, and is. moreover, emitted by Bath gas, and may very likely be a neon ray. We hope to pursue the investigation of this interesting sj>ectrum, and if possible to sort out the rays which may be ascribed to sub- stances such as neon and those which are due to one or more other substances. The gas from Bath, even if primarily derived from the atmosphere which is by no means sure — seems to have undergone some sifting which has affected the relative proportions of helium and neon, and a more thorough comparison of its spectrum with that of the residual atmospheric gases may probably lead to some disentangle- ment of the rays which originate from different materials. The arrangement of the rays in series.it' that could be done, would be a step in the same direction. We are indebted to Mr. Kobert Lennox. F.C'.S., for the great help he gave us in the complicated manipulation with liquid hydrogen required to fill the spectral tubes, and to Mr. ,1. W. Heath, F.C> t" kind assistant e. * ' Boy. Soc. Proc.,' vol. 66, j>. 191. On t/tr tfjwfrn.tii <>/'//><• more. Volatile Gas** of Air. 473 st of Approximate Wavelengths of the L'n>/.-; /7>-/V//r a,,// Ultra-violet, flic \r 5031 7247 Tsbe Ne 5852 -7 bHe 5015 -7 7174 sb 5820 wcl 4958 be He 7065 '5 sb 5804 bHe 4922 -1 vw 7058 sbc 5763 v\v 4884 be Ne 7034 be 5747 scH 4861 -5 be Ne (593 J be 5718 vwb 4838 be Ne 6716 be 5680 vw 4819 be He 6678 '4 wbe 5662 vwb 4811 be Ne 6601 be 5656 wd 4791 scH 6563 5592 wd 4754 be 6535 b 5561 bNe 4715 be Ne 6508 5532 bite 4710 Twbe 6446 Twb 5503 bNe 4704 be Ne 6404 vw 5447 w 4687 be IS'e 6382 wb 5432 W 4680 be Ne 6334 vw 5417 4664 be Ne 6304 vwb 5409 wb 4657 be Ne 6266 bNe 5400 w 4647 be 6244 5372 wb 464(» be 6232 5360 w 4636 be- Ne 6217 5355 w 4628 b 6183 be Ne 5341 a pair w 4616 vwbc 6176 be Ne 5330 w 4589 be Ne 6163 w 5304 w 4583 be Ne 6144 w 5298 4570 rwb 6128 5234 w 4570 bNe 6097 b 5222 4540 bNe 6075 5209 453H bNe 6031 bNe 5204 w 4526 b 6001 b 5192 w 452* b 5901 bNe 5188 w 4518 wb 5987 5152 w 4606 be Ne 5976 bNe 5145 w 4500 wb 5964 5122 w 4488 sdb Ne 5945 bNe 5116 vsHe 4471 -6 w 5919 bNe 5080 vw 4460 w 5914 5074 4457 wb 5905 bHe 5047 '8 vw 4438 be 5882 sbe 5038 wb He 4437-7 474 On th> ^jifct ri' in "/ tic. m • •* (if Air. mi I. :- He :- 3! 127 »>x- 8610 8906 w :55<• Hi' - :j be .1 164 MM W h,- 3460 \\ 1370 w 884^ w 84M v\\ 1868 w 8840 be 84M w 4363 oH 8886 Wi- 8451 w 1868 b 3880 sb«- H. :U»7-7 \ \\ •347 .1.,- He 8819*76 w 3421) 1 H 4340-7 we He - :^< vw 3417 \ \\ 1816 3777 w 3407 vw 1806 w H 3770 w 3I<>4 \\ 4290 3766 b 3393 1276 :<7.-.l b 3388 w 4270 3751 I 3378 \\ 1*61 A W 3745 YW :«74 \\ 125S W 373S W 3372 Ittl ? e 3735 !„• 3370 4241 ? 0 8728 33H7 »2:< » w 3722 w :i303 4^2 \v 3721 '5 vw 8861 4220 1 3713 3360 w 421S 3710 3358 w 4206 he He 3705-2 bHe 33547 \ \\ U'.iS \v 3703 1 8845 we 417«5 37»)1 w 3344 wh !!,• !!<»»•! • of 3694 9 3335 « 4151 ,. 3686 8829 h Jl«- 4143 -i) e *;vi 3327 w 4i:<4 •B 3064 a 3324 b 4131 8665 sb 3319 wb 4128 3651 w 33ir, h H. H21 W 3650 w 3313 w H12 3644 a pair w 3311 Se H 4102 «cHc y 3634 a pair 8810 w 4099 vw 8888 1) He ? 32! »7 vw 4080 1.,- II,. :{»J13-8 \ \v 8t64 w 40SW \v 8809 wb 3250 w 4063 be He • 36' MI B 3244 4047 *bc 8598 3233 vw -W43 VS :<."^7 -5 a pair ? puir 3230 vw W37 ;{.'>7"> 3225 robe He 4u2«5-3 H 3571 8 3218 b Ho? 4009 3.v ;:> 3214 w 3996 W 8061 88J08 V W 3985 W 3558 3199 vw 39HO vwb >blfe M87-8 se H 3970 3543 w 3165 1 H, :^M-9 vseb :i521 w 3142 vw 8888 be 3515 Oil Boulder* «ii>ml>er 21, — Road December 13, 1900. The invasion of Griqualand at the beginning of the war with the Transvaal and Free State, put a stop, for a time, to working the New- lands Diamond Mines, some interesting specimens from which were brought to the notice of the Royal Society, on June 1st, in last year.* But shortly before the hurried departure of the employes, another small collection had been despatched to Mr. G. Trubenhach, the Managing Director in London, which he showed to me, early in the present year, most kindly placing the new specimens at my disposal for study. Some represented boulders, some the diamantiferous breccia, popularly called "' blue ground/' in which these occur ; some the " country rock." The first,- though (so far as can be seen) without diamonds, include at least four additional species of rock ; the second throw a little more light on the past history of the matrix. Moreover, they come from a new set of workings to the north-east of the former, where a shaft has been sunk, and galleries driven at a depth of about 46o feet. Apparently two " pipes " are connected by a narrow fissure filled with breccia. t So I have ventured to communicate the result of my investigations to the Royal Society, including with them a short note on a residue obtained- by Sir William Crookes, F.R.S., after dis- solving away almost the whole of a small fragment of the remarkable diamantiferous eclogite which was described in June, 1899. (1.) The BouUtr*. ('?.) Of these one is rudely semi-oval in outline, measuring aVmt •U inches in greatest length and breadth, and H inch thick, being probably a piece broken from an ellipsoidal pebble. The rock is liolo- crystalline, composed chiefly of a pyroxene resembling bastite and of olivine, converted on the older-looking surfaces into a pale-screen serpentinous material. Examination of a thin slice shows the rock to consist mainly of olivine, which exhibits incipient serpentinisation along cracks in the usual manner, and of a very pale brownish-green bastite, with one close cleavage : and possibly one or two small grains of a monoclinic pyroxene ; spinel, and even original iron oxide, being apparently absent. Specific gravity, 3*074. * ' Roy. Soc. Proc.,' vol. 65, p. 223. t The precise depth at which the specimens were obtained cannot be given, as Ihe labels became illegible in the hurried transit. 47G Pr<»f. T. (I. I'.oimey. On Bonlden and othtt lfa£ Specimen* (/'.) Another specimen, apparently alx-ut half «>f a fairly well-worn boulder, is not «|iiite sn large, ("ndcr the mi«-roscopc it is found t<» IK- practically identical in composition, but a little mon- >c; pentinisrd ; a clear i>otropi<- mineral sometimes forming a Imrder to the enstatite. The presence of any original grains of iron oxide is douhtful, l»ut one or two of augite can IK- recognised, Moth sjM-i-inieiis. however, may In- named Saxonites. (''.) Not very much worn, and rather triangular in shajie. alioiit :> indies by I inch, and about O'G inch thick, consisting apparently of garnet, two pyroxenes and jK-rhaps olivine. Microscopic examination shows olivine. almost wholly converted into serpentine, enstatite parti- ally changed to another (the usual) variety of the same mineral ; chrome diopside. a little colourless augite, with a diallagic hal>it. and pyrope (two specimens). As the last-named mineral is not abundant, the rock is more nearly related to the Lher/olites than to the Kuly>ites, and so may lie named a granatiferous hher/olite. ('/. ) A roundish Hat slali alioiit .'{•"> x :.'•"> x UMi inches, containing red garnets, enstatite. and a bright green pyroxene. Micros.-opic examination shows olivine. jKirtly convei'ted into a dull yellowish-green serpentine, chrome diopside. -due enstatite. now altered to a serjHMi- tine. the colour suggesting that it is ehromiferous, and pyrope (not abundant). A little jwdc liniwn mica, probably secondary, occurs about the garnets and the diopside, in one case occupying a crack. The rock belongs to the granatiferous peridotites. and though it con- tains less enstatite than the last one. may also IK; regarded as a variety of Lherzolite. ('-.) The next specimen is evidently a fragment, the angles and edge- of which have lieen slightly worn, as if by water. It measures aliout .'U by •'{ inches, and 1] inch in thickness. The rock in the freshev part consists of pyrope. and two minerals of a dull-green colour, but about half of one surface is affected by decom{Misition. which has penetrated to a depth of about | inch. Here one of the pyroxenic minerals ap]>ears to IK? a pale-coloured bastite with the usual metallic lustre : the other of a brighter green tint. Examination with the microscope shows the following minerals: (1) Olivine in various stages of conversion into serpentine; some grains l>cing traversed as usual by very pale-green strings of the latter mineral, others completely changed into it. and of a yellowish or brownish colour: minute dark -brown needles are sometimes present ( ? rutile). (2) Bas- tite with a well-developed pinacoidal cleavage: sometimes partially or even wholly converted into a fibrous material, which with transmitted light is a rather rich green colour, the usual small brown negative crystals being develoj>ed in some grains. (3) A very pale sea-green augite. probably a chrome diopside. (4) Pyrope : the grains having a kelyphite rim and showing incipient mineral change along the cracks. from tin' Ni'K'Iumlx ])i"iitaiui .!//'/,">. < ! riijtialaiul IVest. As one or two grains with a general resemblance to the bastite appear to give an oblique extinction, a third pyroxene may be present in very small quantities. Mica is wanting as an original constituent, and most of the iron oxide is secondary, but one grain may be primary. The structure of the rock is granular, no constituent being idio- morphic ; hence the order of consolidation cannot be determined with certainty, but I incline to placing the olivine, which is slightly the most abundant mineral, first, and the garnet, which is slightly the least so, last. The rock is distinguished from ordinary Eulysite by the presence of a fair amount of an enstatite, but as this does not indicate any important difference in chemical composition, I prefer calling it Enstatite-Eulysite, to burdening petrology with a new name. (/.) This specimen has a rude resemblance to an oven bottom loaf, measuring full 16 inches in two directions at right angles on the curved surface; the flat side being probably the result of a fracture, apparently produced after most of the rounding had been done. The rock is holocrystalline, its principal constituents being dull red garnets and green pyroxenes. The former have their outer surface worn smooth and flat, the latter a very slightly corroded one. The rock is rnacro- scopically identical with the eclogite described in the last paper, and it proves to be composed of pyrope and chrome diopside with occasionally a few fibres of secondary hornblende, no grain either of olivine or iron oxide occurring in the slice. (#.) This specimen is a rudely trapezoidal block with rounded edges and corners, measuring about '2\ inches each way, apparently rather water-worn, consisting of somewhat rounded crystals of greenish pyroxene, over an inch in length, in a matrix of a similar mineral and felspar. Specific gravity, 3*125. On examination with the microscope, the larger grains prove to be generally diallage, a faint sea-green in colour, with a close pinacoidal cleavage, often made more distinct by the deposit of a little opacite or ferrite. Small brown negative crystals are frequent, one of their longer edges lying parallel with an axis of elasticity. This mineral is altered locally into a pleochroic hornblende (changing from a raw to a burnt umber tint). The diallage is some- times bordered by, and near its edges occasionally encloses, small grains of a slightly browner and more pleochroic mineral, extinguishing parallel with its principal cleavage, and thus representing a rhombic pyroxene,* but it also throws out root-like prolongations in which a cross cleavage is visible. Where the diallage has been replaced by hornblende, the latter often extends some little distance into the mots, which in a few cases suggest the presence of the rhombic constituent. These are embedded in felspar, thus affording a pegmatitic structure which varies in different parts of the slices from incipient to well * These locally are seen to pass into a yellowish serpentanoiu mineral, \vlii.-li •vvith crossing nicols shows a fibrous structure and fairly bright polarisation (ints. 478 1'p.f. T. <;. 15 ..... Btml&n tend ot) - men »• ; small lamelln- of a different tint, arranged in a kind of network with lo/fn-c-.-ha|»ed nie.-hc-. making Pegmatitic association of a pyroxene and felspar (composite), x 21. The "rootlets" and most of the mineral round the central grain is pvroxene. Decomposition shown about a crack.* their appearance. These possibly may indicate an early stage of the conversion of the diallage into hornblende. The grains of felspar vary much in size, even when associated with the " rootlets " of pyroxene. They are generally in good preservation ; exhibit twinning, usually on the alliite type, and are shown by the extinction angles to be ni»»tly if not wholly labradorite. Small grains of iron oxide are present, which are most abundant near the margin of the larger pyroxenic grains. They arc .-oinetinie.- scattered in the pegmatite, and in one or two cas< slightly root-like in shape. Cracks traverse the rock and Imve led to mineral change. They are often lined with small crystals of a In-own mica, similar to that which occurs in some specimens of the "blue ground." These are iml>edded in a rather earthy-looking granular material, which is, no doubt, a decomposition product from the felspar. Pegmatitic .structures, whether macroscopic or microscopic, are fairly common in granites, where the associated minerals are quartz and felspar, but, so far as my experience goes, are infrequent with other minerals. * I am indebted to my friend Mr. C'oo:uara-Swaiuy tor the microphotograph. /row th>- X<'n-iii,iilx Diiiiiiom) Mine.-;, Gru^'nlo n>l JJVs/. 479 Professor Kosenbusch however mentions the occurrence in .some syenites (including those with elseolite), in diorite (very rare), and in a hyper- sthene gabbro (or norite) from P^kersund, on the west coast of Norway, and St. Paul's Island, Labrador.* Thus we may be content to call this rock a pegmatitic hornblendic Gabbro. (/<.) This perhaps represents a pebble rather than a boulder, for it is a fragment only about L! x 1-J x 1 inches, adhering to a piece of " blue ground," the surface in contact with the latter being well rounded. Macroscopic-ally it appears to be a medium-grained diorite ; the micro- scope shows a holocrystalline granular structure; the plagioclastic felspar is in fair preservation, and, perhaps, is labradorite ; the horn- blende is rather strongly pleochroic, ranging from pale brownish-green to deep brown. The mineral, however, is not original, but an alteration product from a pale green augite (omphacite ?). Grains of iron oxide are also present. Slight decomposition has taken place in a narrow zone from the surface inwards. (?'.) The last specimen is a lump of irregular shape. Presumably it is from the blue ground, but there is nothing to prove this. In a compact dark brown to slightly purple ground-mass, a number of irregularly-formed greenish-grey patches are scattered so as to suggest flow brecciation. These, when examined under the microscope, are a very light greyish-brown in colour, exhibiting flow structure, minute devitrification, and some decomposition. The matrix is darker, sprinkled with opacite and ferrite, minutely devitrified, showing an irregular wavy structure, and occasionally ill-defined crystallites of plagioclase felspar. The rock, now a felsite or porphyrite, was prob- ably once either a sanidine trachyte or more probably an andesite, with flow brecciation. This specimen possibly may not represent a boulder, but a dyke or flow associated with the " blue ground." (2.) Diamantiferow Matrix. Specimens of the "blue ground" in which the boulders occurred were also sent. As they came from another part of the mine, and the best preserved exhibited one or two slight differences, I have had a few slices prepared. To the unaided eye the matrix is more of a purple-brown colour, slightly more compact and hard, but more brittle ; the fragments of magnesian minerals, however, seeming more com- pletely serpentinised. A few small, rather crumbling, rock fragments, * ' Elements cler Gesteinslehre ' (1898), p. 221. A case where the structure is more like that of the true graphic granite, from the dolerite of Pouk Hill, is described by Mr, Allport, 'Quart. Journ. Geol. Soc.,' vol. 30, p. 549, and figured by Mr. Teall, ' British Petrography,' PI. XXIII, fig. 2. An instance of micrographic intergrowth of quartz and calcite is described by Mr. Cooniura-Swamy in the afore- named journal, vol. 56, pp. P05, 606. I'l'.f. T. <:. IJonney. <>„ BotUder* and other Bock Specimen* of a rlull white colour, -peckled with green, are present. Mil TO-, »pic examination shows that the larger minerals do not call for any special notice, except that a rudely . rcscentic pyrope has a kelyphite rim on tin- <-onrave as well as on the convex side, proving the fracture to !*• an old one. Hut the small plates of a In-own mica, the occurren, e of whid) ha- lieen already noticed,* are very abundant in the matrix. These plates in some of the specimens are rather irregularly outlined, and rarely exceed O'OOl inch in diameter. Init in others they average al»out doulile that si/,e, and occasionally a few of them may even < Q-004 inch. Then the outline is more, rectangular, and the cleavage more distinct. The smaller flakes often tend to form a /.one around included rock fragments, and scattered granules of iron oxide more common in the .-li--e- containing the larger flakes. t I have now no douht that the miner;*! is a secondary product. The unusual abundance of a minute brown mica in the ground mas- made an analysis desirable. For that annexed T am indebted to Mr. C. .lames, who has executed it in the laboratory at T'niversity College under the supervision of Professor AY. 1 tarn say. Silica 3S-77 Alumina 14'62 Ferric oxide 11-36 Calcium oxide 4-."Jl Magnesia 12-14 Potash 2-63 Soda 1-90 Loss on heating CO^ and H/J 1 •>'•>') 99-4S (The iron was all estimated as Fe-jOsj one specimen gave a trace of nickel.) If we compare this analysis with one given by Professor C. Lewi's.; SiO2 = 33-00, FeO (including Al.O3) = 12-00, MgO = 32-3S, CaO = 0-63, Xa..-O = 0-67, CaCO3 = 16-02, H,.O = 6-0 (total 101-71), and with those of Kentucky " kimberlites " p*'i-imens from >Y\vlaniU opaque granule* (? ilim-nitc) vein to lake the place of the translucent brownish granules (in part pcrof-kite) in the spec-linen- fr«»m De Beers Mine. J ' The Matrix of the Diamond.' p. 17. § ' Kleineute dcr Cotein-lchre,' p. 165. ,| 'The Matrix of the Diamond,' p. d, «;/*. p. «>1. (If1 X'-i'-l ]: d a ml ntti> ,•!;>• .< • metu Another specimen has now been found. Tin- pyrope apparently was rounded in form. about a ] inch in diameter, and surrounded by a kelvphite rim. It is broken across, thus di-ii-lo-ing the diamond, an octahedron, only one face of whirh i- completely expi-ed. This is slightly stepped, and measures roughly one-tenth of an inch along the t-d-c. A small piece of the usual purplish breccia adheres to the pyrope. so the case i- exactly parallel to the former one. In each the perfect form of the diamond shows that it crystallised liefore the garnet, and as the ordinary varieties of the latter mineral seem to l>e produced at a high temperature.* the association may IKJ significant. (3.) Tin ('„„„{,•,, //,*•/.. A fe\\ s]>ecimens of this were also sent, l>ut only two varieties pre sent any feature of interest. One is a greenish conglomerate with calcareous matrix, and rounded pebbles up to al»oui \ inch diameter : the other has a jwde-grey matrix, speckled with .some small angular dark-green fragments and a few sub-angular pebhles up to about an inch in diameter ; one apj>arently a red felsite, the others diabase. In the first specimen the microscope shows abundant sub-angular to rounded grains, mostly diabase, of which there are at least half a doy.en varieties, a microgranite and two or three rocks more fragmental in aspect ; one perhaps a tuff, another apparently a quart/he, affected by pressure, and a third a suit-crystalline dolomite. These are cemented by microgranular calcite, containing probably a little magnesia (the crystals often forming a kind of border to the rock fragments), the interspaces being filled in with clear dolomite. In this cement are embedded some angular bits of quart/, a fragment of altered felspar, and one or two, perhaps, chalcedony. The other specimen shows a fine-grained muddy matrix, in which are scattered angular to siuV angular grains of quartz, with a little decomposed felspar, a little of a green mineral (1 replacing pyroxene), decomposed iron oxide, and perhaps some small rutiles, with rock fragments, generally rather rounded, representing compact diabase, or possibly sometimes andesite, and one or two of a sulHcrystalline limestone. Both these specimens, as Mr. Tnibenbach informs me, represent a rock named "bastard blue" by the miners, and it has l«?en pierced in both shafts, the diamautiferous breccia, or " blue ground," apparently pissing under it in the second shaft. As, however, their is no real relationship between the two rocks, I regard the association as fortuitous. * I do not forget the remarkable gurnets from the Ba*togiie, described by Pro- fe»?«or Renord (' Bull, du Musee Royal d<- Belgique,' vol. 1, p. 9), or that called jirreueite (also one of the andradite group), but both these minerals are very abnormal. [The pene.*i* of the former is disK-usscd by Miss C. A. Raisin, D.Sc., in n paper which will appear in the ' Quarter!} Journal of the Geological Society' "l.j f/'or/i the Ncwhtnds Diamond Mines, Griqualantf West. 483 (4.) Residue from the I)i«nwntiferoux Edofjitc. After the reading of my description of the " diamantiferous eclogite," Sir YY. Crookes kindly offered to examine that rock for microscopic- diamonds. Taking one of the fragments, weighing 130'5 grammes, which had been detached in slicing the specimen, he treated it as follows : — " After being very coarsely broken up, the material was put into a very strong sulphuric acid. The acid was boiled for some time, and, after being allowed to cool, the residue was washed, dried, and then heated for some hours in strong hydrofluoric acid. After it had been well washed and dried the treatment with hot sulphuric acid was repeated. The mass, after a few alternations of these acids, became disintegrated, and all, except a few crystalline lumps, were dissolved. After about ten treatments only a few small crystals remained, and these (with the exception of a sample) were reduced by a few more boilings with the acids to a single small one about half a millimetre in diameter." This was boiled fourteen times in each acid, and appeared to l>e slightly reduced in size. " It sinks in methylene iodide, specific gravity 3'35." This was sent to me with some of the small crystals just mentioned, all being mounted. The solitary survivor of the whole treatment showed on one side curved crystal faces, but on the other appeared imperfect. These faces, so far as I could judge, indicated an isometric or possibly a rhombohedral mineral. Its refractive index is high, the colour a pale smoke-brown, and it apparently produced some effect on polarised light. -That, however, was not conclusive, for diamonds from Newlands, as at Kimberley, are often in such a state of strain as to be anisotropic. Of the survivors of the first treat- ment, the more abundant were colourless, rough in outline, but possibly showing one cleavage surface, apparently at right angles to an optic axis ; polarisation tints bright ; the refractive index high, but inferior to that of a diamond. It appeared to me not improbably corundum. The less abundant granules were more rounded in outline, with rather rough, possibly corroded, surfaces, translucent, of a resin-brown colour, apparently producing some effect on polarised light; on the whole they seemed to bear some resemblance to rutile. But to come to any conclusion about the first mineral it was necessary to detach it from the mount. As I have no apparatus for very delicate work, that not coming within my usual line of study, I had recourse to Mr. L. Fletcher, the Keeper of Mineralogy, and Mr. L. J. Spencer, also of that Depart- ment, at the British Museum. The latter attempted to measure the supposed diamond with the goniometer ; the faces, however, were too curved for the purpose, but both of them regarded the edges as too sharp for the mineral to have suffered appreciably from the acid, as Sir W. Crookes was inclined to think. They consider it to l>e really iso- 2 M 2 4S4 Dr. W. T. I'.l;. nf..,,l. '/'/,, /A\/, ,;,„/;,„(, ,>f tropic, aiid a encer think- tli.a the .•,,]. .nrless binfringent fragments arc perhaps uptically uniaxial, and that they may very well he corundum.! The ln- diamonds. In favour of this identification is the fact that -mall diamonds occur at the Xewlands Mine (I have seen some in Mr. Tru- benbach's hands), rathei1 ovoid in shaj>e, with a roughened surface, some a yellowish-brown, some colourless. But against it we may urge that they aj)]Kiar to have l»een destroyed during the second treatment.* He this as it may, Sir W. Crookes \\^< -ucreeded in showing that micro- scopic diamonds i/e. To conclude : in addition to this residue from the eclogite we have ascertained (1) the existence, in some quantity and variety, of pre- tria--ic dialja-e,.^ (:>) the abundant development of a microscopic brown mica in the ground mass of the so-called kimlxjrlite ; (3) the presence in it, as tnie IwuMers, of at least four more species of holocrystalline rock. The last fact acquires an additional importance, localise, since the publication of my former paper, the boulders therein described have been claimed as " concretions " in the so-called kimberlite.|| With this matter I have dealt elsewhere,*! but the identification of seven species or strongly-marked varieties of holocrystalline rocks, peridotites, eclogites, &c., in which the minerals at the surface are worn as if by the action of water, not to mention the general structure of the so-called kimberlite, must, I think, otter insuperable difficulties even to the most enthusiastic advocate of concretionary action. " The Distribution of Vertebrate Animals in India, Ceylon, and Burma." By W. T. Bi.AXFoiin, LLD., F.K.S. l.'e.-.-iv.-d December •">. — Read December !."», 1000. (Abstract.) Several contributions on the subject of the distribution of Verte- brata, or geographical Zoology, in India and the neighl>ouring countries * On rt-exauiining tin' specimen, now tliat Mr. Sp«Micer has kindly mounted it in a better position, I agree with this determination. + On a final examination of the slides, I find among them one if not two -mall grains which I strongly suspeet to be diamonds. J A final examination and comparison with some bits of " bort " given me by Mr. Trtibonbach lias not made me more favourable to my original identification with rutile. § That is, at any rate, older than the time when the Karoo series was deposited. j| Professor Beck, 'ZeiUchrift fur Praktische Geol.,' December, 1899. •f • Geol. Mag.,' 1900, p. 2 10. Jrei-fi1n'r. W. T. I'.Ianl'ord. Tit 1h*t rtlmin.i of into nineteen tracts, di>lingui>hed liy physical characters -u< h as rainfall, temperature, pre-em e or aW-nce uf fore>t>. ami prevalcn. e »\ hilly ground, ami to ronstruct table- -houing the diltributHW «i each genus of land or fresh- water vertebrate in the tracts. Genera have ln-en selected for (Consideration, localise families an few in numl»er ami I wide in I'an^e, whil-t >pei ies ,ne tun niunerous and t«M» unequal in importance, It is rero^ni-ed that there is much dirt'erenee in the value of genera in ditt'erent grou]»s, the ^enei-ic difl'erences in jxisserine In'rds, for instance, being as a rule of inferior rank to those in some other orders of liirds, or to those gene- rally adopted amount mammals, reptiles, and Katrachians. In the denial-ration of regions and suli-regions. terrestrial mammalia are regarded as of primary importance. The tracts are the following -. A. !iiilfh(i'itii>/i-(i< riniti. 1. 1'unjali. Sind. i>alu<-hisian, and Western Rajputana. 2. (langetie Plain from Delhi to Ixajmahal. 3. Bengal from Kajmahal to the Assam Hills. 1». I ml in H I'i'ii/it.-tiilii. 4. Kajputana and Central India as far south as the Nerlmdda. ~). Deeean from the Nerbodda to alutut 1(5 N. lat. and from the \\fstern (4hats to long. SO' K. 6. Behar, Orissa, £c., from the (langetic Plain to the Kistna. 7. Cai-natic and .Madras, south of o and <>. and east of the Western Ghats. 5. Malaliar Coast, C 'oilcan, and Western Ghats or Sahya. Himaiaytu. 11. Western Tibet and the Himalayas above torest. ll'. Western Himalayas from Haxara to the western frontier of Nepal. 13. Eastern Himalayas, Nepal, Sikhim, Uhutan, >V< •. Vert'i>i-i- IHttril,,'!;,,, Ceylon. l.e regarded a.- ;( -in-l«- -ii'i-iegion, aiiul> region.* The forests of the Sahyadri range ami of the Western or Concan ami Malaliar roast and the hill • r Southern Ceylon have a far richer fauna than the remaining area. l»ut are not sufficiently di>tinct to require sul> regional sepa- ration. The hill fauna of the Sahyadri range. es]>ecially on the highest portions, such as tlie Nilgiri and Anaiinalai Hills, and that of the hill group in South-western Ceylon, contain several Himalayan genera and spei-ies. hut not sufficient to enalile the S. Indian and Ceyloncse areas to l>e .-lassed \\itli tlic Himalayan forest area in a scjKirate Mil>-divisioii or suWegion. The C'isgangetic sul>-iegion is distinguished from the TrUUgaugBtk l»y the presence amongst inaininals of Hya-nida-. Krinacein.t*. (lerKil- lina-, of three peculiar genera of Antelopes and of some other type- ; aim.ngst l>irds l>y the occurrence of Pterodetes (sand grouse). I'hu-ni- copteri (Hainingoes). Otidida- (luistards) and Cursoriiiue ; amongst reptiles l>y the possession of the families Kulilepharida-, Chaina* leontida- and Tropeltida-, together with many ]>eculiar (ieckonida-, Agamida- and Larertida-. and amongst oatrachians l»y alnxit one-half of the genera found in each sinVregion l>eing alisent in the other. The ditterence Between the reptiles and hatrachians l>y itself would justify the classification of the two areas as distinct regions, a view adopted l»y several writer.-. The following figure.- show the total numher of genera recorded from theCisgangetic suli-region and the pen entage of them not ranging into the Tran.-gaiigetic area, the Himalayas and Hurma : Cugangetic. Not Mammal> ............ 6^ 14 or !<*•"• jicr cent. Birds .................. 347 4C, or 1:5 Reptile- ............... 93 :«i or 4-J Matrac-hians ......... 17 i> 01 •">•"> „ Kreshwater fishes ... ">,•< (.» o: l."r."i Omitting l»at.-. the nuinl»er of Cisgangetic mammalian genera is furty-six. of which 14 or .">U per cent, are wanting in the Himalaya* and east of the I Jay of Bengal. The difference Between the Cisgangetic vertebrate fauna ami that inhaliiting the rest of the Indo-Malay or Oriental region is partly due to the alisence in the former of numerous Kastern tyj>e.s, and jwrtly to the presence of two con.-tituent- U-i. and reptiles having * Tin- tc-riii- " C'isgangetic " and " Transgangetic" have already been employed by Professor Gadow, f.t.c. iii Iiulin, Ceylon, and Burma. 489 a distinct relationship with Ethiopiau and Holarctic genera, and with the Pliocene Siwalik fauna. This constituent of the Ciagangetic fauna it is proposed to distinguish l»y the term Aryan. The other con- stituent is composed of reptiles and batrachians, and may l>e termed the PraA'idian element. The latter is well developed in the south of the Peninsula and especially along the south-west or Malabar Coast, and in Ceylon, Imt it gradually disappears to the northward, its northern limit, so far as is known at present, not extending to the 20th parallel of north latitude. It is probable that this is the oldest part of the Cisgangetic fauna, and it may have inhabited the country since India was connected by land with Madagascar and South Africa, across what is now the Indian Ocean, in Mesozoic and early Cenozoic times. The other two elements, the Indo-Malay or Oriental and the Aryan, are probably later immigrants, and its wider diffusion may indicate that the Oriental element has inhabited the Indian Peninsula longer than the Aryan has. There appears some reason for regarding the Oriental portion of the fauna as dating in India from Miocene times and the Aryan from Pliocene, whilst in the Pleistocene epoch the proportion of Aryan to Oriental types of mammals in India, as shown by the fossil faunas of the Nerbudda and the Karnul Caves, was much larger than at the present day. There are some other peculiarities of the Indian Peninsular fauna to which attention may be called. One of these is the presence of genera and sometimes of species which are found on both sides of the Bay of Bengal, but not in the Himalayas or Northern India. A good example is afforded by the genus 7'/v"/"/"-S of which one species inhabits Ceylon and India south of about '2'2 N. hit. whilst two others are found in Southern Tenasserim and the Malay Peninsula. In Pliocene times, the genus inhabited Northern India. Another instance is the lizard JJiilf/f/s ijiiftvtux found in Burma and Arrakan, and also in South Canara on the "NVest Coast of India. Examples amongst reptiles are rather numerous. Moreover, whilst there are numerous alliances between the animals of Peninsular India and those of Africa, there are also some curious connections between India and Tropical America, but these are chiefly amongst invertebrates. Some, however, are found in reptiles. It is probable that such Indo-Aiuerican connection.- are vestiges of older life than the Indo-African. They are of course, generally speaking, instances of animal groups once more widely distributed, but now only preserved in a few favourable tropii-a] localities. IV. The forest area of the Himalayas belongs to the same sub-region as Assam, Burma (except South Tenasserim), Southern China, Toinjuin, Siam, and Cambodia, and to this sub-region the term Transgangetic may be applied. It is distinguished from the Cisgangetic sub-region by the absence of the animals already specified as characteristic of that 4!>n Ih. \V. T. P.lant'ord. 77" area ;uiy the presence of tin- following, whirl) are wanting in tlie Indian PeniiiMila Mammal- : tlic familie- Simiida-, Pro.-yoiiid;e, Talpida-, and Spalacida-, and tin- sub-family ( Jymnuriiui', be-idcs nimierou.o genera siu-li a-> /'//'////#/'///, ///7/W/..-. . //-<7ortion of the Transgangetic region, but not from the Ciagangetic: — Mammals ............ 74 i'tj or :J5 percent. Birds .................. 47o 174 or :{«;•") lieptiles ............... S4 :iO or :»'»•.") ., Batraehism ......... 16 s or 50 ., Freshwater tislie< ... (57 is 01- -21 ., Oinittinic hats, the nuinln'r of Trans^angetir mammals within Indian limits ai-e fifty-four, of which L'li or 40 per <-ent. are not C'isuan^etir. The relations of the Himalayan fauna to that of Assam and Burma on the one hand and to that inhabiting rhe Peninsula of India on the other may l>e illustrated l>y the mammals with l>ats omitted. <)f forty- one genera oei.-urrinjj in the Himalayas, three are not found in the hills south of Assam or in Burma, whilst sixteen are wanting in the Ciflgangetic region. It should l>e remembered that a large numl»er of the genera are widespread forms. As the result is not in agreement with the views of some who have written on the subject, the relations • >t -peeies have Keen examined. It ivsiilts that eighty -one species of mammalia Itelonging to the orders IVimates, C'arnivora, Inseotivora, liodentia, and I'ngulatji are recorded from the forest regions of the Himalayas. Of these 2 are doubtful, '2'2 are not known to occur south of the Himalayan range in India or Burma, 21 are wide ranging forms and are found in both Burma and the Indian Peninsula, 1 only (//if. ff i >.>• h'n,-ii,;i) is common to the Himalayan fore-sts and the Indian Peninsula, but does not range wist of the Bay of Bengal, whilst :i."i are found in the countries east of the Bay of Bengal, but not in the Penin- sula south of the (ianges. Of the 35, rt only range as far as the hills south of the Assam Valley, 16 to Burma proper, and 11 to the Malay Peninsula and Archipelago. Or, in other words, of the 79 Himalayan species, ."id or 70 per cent, are common to the Transgangetic region, and only 22 or 2s per cent, to the Cisgangetic. Of the 22 species nut ranging south of the Himalayas a large majority are either Holarctir species or Injlong to Holarctic genera. The fauna of the Himalayan forest area is partly Holarctic, partly Indo-Maluy. It is remarkably JMIO;, when compared with the C \\'/idj/'ate Animals i/t India, Ceylon, and Bnriiw. 401 getic ;iiid Burmese faunas, in reptiles and batrachians. It also contains but few peculiar genera of mammals and birds, and almost all the peculiar types that do occur have Holarctic affinities. The Oriental element in the fauna is very richly represented in the Eastern Himalayas and gradually diminishes to the westward, until in Kashmir and farther west it ceases to be the principal constituent. These facts are con- sistent with the theory that the Oriental constituent of the Himalayan fauna, or the greater portion of it, has migrated into the mountains from the eastward at a comparatively recent period. It is an im- portant fact that this migration appears to have been from Assam and not from the Peninsula of India. V. Southern Tenasserim agrees best in its vertebrata with the Malay Peninsula, and should be included in the Malayan sub-region of the Indo- Malay region. The continental area of the Indo-Malay or Oriental region is divided into three sub-regions, Cisgangetic, Transgangetic, and Malayan. There are several points left which require explanation. There is the much greater richness of the Oriental constituent in the Cisgangetic fauna to the southward in Malabar and Ceylon, although this is far away from the main Oriental area, and the occurrence also in the southern part of the Peninsula of various mammalian, reptilian, and batrachian genera, such as Lori*, Trn•, which are represented in Burma and the Malay countries but not in the Himalayas or Northern India. In connection with this the limita tion of the Dravidian element to the south of India should also be remembered. Then there is the occurrence of certain Himalayan species on the mountains -of Southern India and Burma, and even farther south, but not in the intervening area. There is also the pre- dominance of the Western, or what I have proposed to call the Aryan, element in the Pleistocene fauna of the Xerbndda Valley, and of Karnul in the north of the Carnatic tract. Lastly we have to account. for the apparently recent immigration of Oriental types into the Himalayas. Whilst it is quite possible that other explanations may be found, it is evident that all these peculiarities of the Indian fauna may have been due to the Glacial epoch. The great terminal moraines occurring at about 7000 feet in Sikhim, first discovered by Sir .1. Hooker.* whose observations have been confirmed by my self t and others, and the occiuTence of similar moraines and other indications of ice action at even lower levels in the Western Himalayas, J clearly -how that the temperature of the mountain range must have been much * ' Himalayan Journals,' vol. ii, pp. 7, «tc. t ' Jour. As. Soc., Beng.,' xl, 1871, Pt. 2, p. 393. J • Manual of the Geology of India,' Ed. 1, p. 373 ; Ed. 2, p. 14, ami references there quoted. 4'.>L' Fff . f in India, Ctylon, and Bvrma. lo\\ei -than ;tt the prc-cnt day when no glacier in Sikhim i* known to de-rclld much bclnW I }.(»(•(» tret. I Miring the coldest portion of the Glacial epoch, a large jwirt of the higher mountains must have Keen covereil by .-now and iee. and the tropical Oriental fauna which had occupied the range, and which ma\ have retembled that uf the Indian Peninsula more than is the < present, must have lieen driven to the ba-e of the mountain- or exter- minated. Tlu- llolatvtic forms apparently survived in larger numU-is. The Assam Valley and the hill ranges to the southward would afford in damp, sheltered, forest-clad valleys and hill slojH-s a warmer refuge for the Oriental fauna than the open plains of Northern India and the much drier hills of the country south of the Gangetic plain. The Oriental types of the Peninsula generally must have lieeii driven south- wards, and .-'inn- of them, such as /////> and 7V<"/"/">, which must originally have Iteen in touch with their Burmese repi-esentatives, have never returned. It was probable during this cold jieriod that the MB feroiis Nerliudda beds and the deposits in the Karnul caves, were accumulated. The tropical damp-loving l>ravidian fauna, if it in- habited Northern India, must have l»een driven out of the country. l"nles> the temperature <»f India and P»urma generally underwent a considerable diminution, it is not easy to understand how plants and animals of temperate Himalayan types succeeded in reaching the hills of Southern India and Ceylon, as well as those of Burma and tin- Malay Peninsula. When the whole country Became warmer again after the cold ei>och had passed away, the Transgangetic fauna appears to ha\e poured into the Himalayas from the eastward. At the present day the compara- tively narrow Brahmaputra plain in Assam is far more extensively fore.-t clad, especiallv to the eastward, than is the much broader Gangetic plain of Northern India, and if. as is. probable, the -anic difference between the two areas existed at the close of the Glacial epoch, it i- to see how much greater the facilities for the migration of a fore-t haunting fauna must have been across the Brahmaputra Valley than over the great plain of the Ganges. This difference alone would give the Transgangetic fauna of Burma an advantage over the Cisganget it- fauna in a race for the vacant Himalayas, even if the latter had not In-en driven farther to the southward than the former, as it probably was during the Glacial epoch. The theory, however, is only put forward as a possible explanation »t -ome remarkable features in the distribution of Indian vertebrate.-. At the same time it doe- ser\e t<> account for several anomalies of which some solution is neces>ary. If thus accepted, it will add to the evidence, now considerable, in favour of the Glacial epoch having affected the whole world, and not having l»eeii a partial phenomenon indti'-ed by special conditions, such as local elevation. Oil flic Intimate Stt'uct»/rc of Ci-i/xfnls. 493 " On the Intimate Structure of Crystals. — IV. Cubic Crystals with Octahedral Cleavage." By W. J. SOLLAS, D.Sc., LL.J)., F.lt.S. Professor of Geology in the Tniversity of Oxford. Received and read March 17, 1898. Revised December 10, 1900. IH4 Dr. W. -I. Soil.-,-. In treating of diamond, tin question of dimorphism naturally pre- M-iit» itself : we proceed therefore to the consideration of graphite. A- the recorded values for the specific gra\ ity of graphite ditlci \\-idelyfroni one another, I made a fresh determiimtion of this con- stant, using the graphite ol.tained as a 1 iye-product in the manufacture of carliorunduni. This \vas kindly given me by Professor Miu>. The graphite was introduced into a diffusion column of methylene iodide and l>en/.ene: it floated at a level of specific gravity 2*286, M given liy small glass indicators. The molecular volume of graphite, a- deduced from this, is .Vi'.j. This value is in close accordance with that found by Petersen, who gives it as 5-3.* The form in which graphite occurs in nature is so closely similar to a hexagonal prism, that for a long time it was referred to the rhomho- hedral system, lint later observations show that it is decidedly oblique or monoclinic. Suppose that a numlier of tetrahedral groups of atoms be placed each with a trigonal axis vertical, the atoms at the base forming a . Macfadyen (A.), Morri* (U. II.), and Rowland (S.) On Expressed V.a-t.ill Plasma (Buchner's " Zymase "), -~>". McClung (R. K.) See Rutherford and McClung. Magnetical Observations at Falmouth Observatory, 1897-1899, 139. Man, Data for Problem of Evolution in (Beeton, Yule, and Pearson), 159 ; (Loe and Pearson), 333. Manson (P.) admitted, 328. Marble, Investigation into Flow of (Adams and Nicolson), 228. Meeting of June 21, 1900, 1 ; November 15, 328 ; November 22, 402 ; November 30, 436 ; December 6, 436 ; December 13, 466. Metals, Crystalline Structure of (Ewing and Rosenhain), 112. Molecular Energy, Distribution of (Jeans), 236. Morris (G. H.) See Macfadyen, Morris, and Rowland. Muir (J.) On the Tempering of Iron hardened by Overstrain, 461. Nerve Crossing, Restoration of Co-ordinated Movements after (Kennedy), 431. Nerve, Mammalian, Demarcation Current and Electrical Resistance of, Effects of Dilute Solutions, &c., on (Macdonald), 310. Newall (H. F.) See Turner and Newall. Nicolson (J. T.) See Adams and Nicolson. North (Sir F.) elected, 1 ; admitted, 323. Northumberland (Duke of) elected, 402 ; admitted, 436. Officers and Council nominated, 402. Offspring, Correlation between Number of, and Duration of Life (Beeton, Yule, and Pearson), 159. Papers published during Recess, List of, 329. Papers read, Lists of, 1, 329, 403, 436, 467.1 Pearson (Karl) On the Kim-lie Accumulation of Stress, illustrated by the Theory of Impulsive Torjion, 222 ; See also Beeton, Yule, and Pearson, and Lee and Pearson. Photography of Coloured Objects in correct Monochrome (Abney), 118. Potter (XI. (.'.) On a Bacterial Disease of the Turnip (Brassica itapus), 442. Psettdomoncu destructans, n. sp., Bacterium of Diseased Turnips (Potter), 442. Rainfall, Variations in: and Solar Changes of Temperature (Lockyer and Lockyer), 409. Rambaut (A. A.) admitted, 1 ; Underground Temperature at Oxford in the Year 1899, as determined by Five Platinum-resistance Thermometers, 218. Ramsay (VV.) and Travers (M. W.) Argon and its Companions, 329. Rayleigh (Lord) On the Viscosity of Gases as affected by Temperature, 137. Resistance in Iron, Change of, produced by Magnetisation (Gray and Jones), 208. Roberts-Austen (Sir W.) On the Diffusion of Gold in Solid Lead at the ordinary Temperature, 101 ; and Rose (T. Kirke) On certain Properties of the Alloys of the Gold-Copper Series, 105. Rock Specimens from Diamond Mines, Griquulaud West (Bouney), 475. 501 Rontgen Rays, energy of (Rutherford and McClung), 245. Rose (T. Kirke) See Roberts-Austen and Rose. Rosenhain (W.) See Ewing and Rosenhain. Rowland (S.) See Macfadyen, Morris, and Rowland. Rutherford (E.) and McClung (R. K.) Energy of Rontgen and Becquerel Rays, and the Energy required to produce an Ion in Gases, 245. Scott (D. H.) Note on the Occurrence of a Seed-like Fructification in certain Palaeozoic Lycopods, 306. Selenafes (Double) of Series R2M(Se04)2)6H2O, Cry stallo graphical Study of (Tutton), 58. Sell (W. J.) admitted, 1. Silicium, Spectrum of ; lines in Stellar Spectra (Lockyer), 403. Skull (Human) Study of the Correlation of (Lee and Pearson), 333. Sollas (W. J.) On the Intimate Structure of Crystals. IV. Cubic Crystals with Octahedral Cleavage, 493. Steam, Thermodynamical Properties of, Characteristic Equations, Tables of Specific Volume, &c. (Callendar), 266. Stoney (G-. J.) Note on Inquiries as to the Escape of Gases from Atmospheres, 286. Stress, Kinetic Accumulation of (Pearson), 222. Sulphur, Boiling-point of; Variation with Pressure (Chree), 3. Sun, Changes of Temperature of, and Variations of Rainfall (Lockyer and Lockyer), 409 ; Preliminary Reports on Observations at Eclipse of May 28, 1900 (Lockyer), 337; (Turner and Newall), 346; (Evershed), 370; (Copeland), 385 ; (Christie and Dyson), 392. Swan (J. W.) elected to Council, 466. Temperature (Underground) at Oxford in the Year 1899 (Rambaut), 218. Tempering of Iron Hardened by Overstrain (Muir), 461. Thermodynamical Properties of Gases and Vapours (Callendar), 266. Thermometry (Platinum), Investigations on; Sources of Change or Error in (Chree), 3. Torsion, Theory of Impulsive, and the Kinetic Accumulation of Stress (Pearson), 222. Travers (M. W.) See Ramsay and Travers. Turner (H. H.) and Bewail (H. F.) Total Solar Eclipse of 1900 (May 28). Pre- liminary Report on the Observations made at Bouzareah (in the grounds of the Algiers Observatory), 346. Turnip, Bacterial Disease of (Potter), 442. Tutton (A. E.) A Comparative Crystallographical Study of the Double Selenates of the Series R,M(SeO4)2,6H2O— Salts in which M is Zinc, 58. Underground Temperature at Oxford in the Year 1899 (Rambaut), 218. Variation, Law of, as regards Reaction to Environment (Vernon), 85. Vernon (H. M.) Certain Laws of Variation. I. The Reaction of Dereloping Organisms to Environment, 85. Vertebrate Animals, Distribution in India, Ceylon, &c. (Blanford), 484. Vice-presidents appointed, 436. Viscosity of Wires, Effects of Temperature Changes on (Gray, Blyth, and Dunlop), 180. Visual Area of Cerebral Cortex, Exact Localisation of (Bolton), 216. .->02 Walker (James) admitted, 328. Waller (A. D.) On the " Blaze Currents" of the Frog's Eyeball, 439 ; — The Electrical Effects of Light upon Green Leaves (Preliminary Communication), 129. Water, Capacity for Heat of (Barnes), 238. (Philip) admitted, 1. WiUon (C. T. B.) admitted, 1. Wires, Effects of Temperature Changes on Elasticity and Viscosity of (Gray, Blyth, and Dunlop), 180. Yeast-cell Plasma, Method of Expressing ; Production of Alcohol and CO; by (Maefadyen, Morris, and Rowland), 250. Yule (GK U.) See Beeton, Yule, and Pearson. " Zymase" of Yeast, Influence of Age on Activity (Macfadyen, Morris, and Row- land), 250. END OF THB 8IXTT-8KTEXTH VOLT7MB. HARRISON AKD SDKS, Printers in Ordinary to Her Majesty, St. Martin's Lane. . MAR 9 1971" Roy^l Society of London 41 Proceedings L718 v.67 Physical * Applied Sci SeriaJi PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY