Se A Oe OO A Th) eG lath Br Orded 8) b-Gam gods i ee ee es soaatgiraten tote @ oe Ve Oe tee te ne t- De tee te hatte Bett lho Mette el Os erence ed eee 10 he he he EA HE: Maer heh he tte t tran Gin Mhethie AE Mew hake Me Meth» nie no Ce Rad te ee rte det Arte ertea dd ay BRM rte Oe Fe teeth ay Ae Pe ed a “ “9 eae Ae We bn eo ee oe oe eet tt Ree tte bet ae Hott mew . Yet ed tree 4 re oer) we ee GD H0eG breed, }.0: 620-6 « Ca Se ee are : Extaea sacra hop Lee oe CO ee | * ‘one Jet he © 2 O44 Greer eta e a ee) “4 ete Te » * BoA eo oe BM he tee : 4 ' ee eee eaes Ure Ormerddotes sours ste by tr a el Efe ah A Meeeretietearie en ha 45 AMEE: home meme ew. tee Com erties: wines Lal 7 cNAAA wa Gee Du Rel bo nev teach Ort mM aerd. Hamner ohne oda * : OF hhh enh ho Ad Or Wh oe he hr he a Wee Hh shee LNT fo kode hn tReet Me emt a Hn Teen wets ae Ft OO Ah MO Hed Orne OW: Melee grerm Leathe Wheel sabe telaictelnte oe thr betes bh hans mot adit epracecseavera nantacgcar sta era ie tintetr ert ae +e eer wy 4 , eee he i hehe ae bee eek ee ey ooh eee artes artes we ee ee ete . 4m BAe Heh Ws O- Om Ome Wn lay ibs 4.9 liv We Moe eh IF =e he hole the > een TEI) ear EE TE ee Ee Hee he | Fe et atl tet Hane etal d aed Oe cee “ ee eo oe ee Pe deherawr tale ents PR ert eee re REG be 8 Mr ternal ae Hotel Cire hr dete Remcg ahistemeirt dl ot: the mg atted ttm: Wu Gs Bebe: Wetenh eee the Ob +e en He ERs ON tee Wolthoha’, Oe thee Ce ee ee es eo oe “we + “ , whaph ahy AA h 4 joes ty Arhebal ee ee ee Peo fee tr tm wey a 1 Orden ke Whe eaters ee ea ee ey Soret wd tethew : n4 OMe meteor ham) ee rotons Hetty a - eer a ee we 1 © deter tere wean oe) 7 eobgalers~n.dee éo4 tee ee ee ee ey Oe ee Ce ee | SE ee eer ee ee eo | Ce ee ee) 7 Ser tte 1 EM ee yO. teeta unde , . te eee ee ee re ee | MG hn tke AG tar & oe rte hs edi os oon Rew esas he o-¢eaeer* “e¢ een ow ‘ LR Rhee ee tt ee ha teen. SARA Se eet Ee He Th he ORR h eRe Reed Oi Hee ell ed & ‘ 4 OPt mi wre gees eter bod tae the eet Oe en ot the rr a Pe OE he thee BP WeTkhe eth W-O.e8 ee ed dee eet Orke 4 eh ee RAY He te oe hike ee atte A Ort ere nett 0.4 @ Astenen Sada GEO a me Niet ee hk oe anther BR a od, beau Woe Mee kt I om Bae wt ee TN See Rr teed: hed Medetted eetbete WM ded ks ted 4, eG ty 1 MCAT On eh en eee ontd Veta totges Men he oo tnt Woy Grete ee ens cr he sy A A et eee te ets oh tt eee nee ais be So Sete Wed 4G: 408s Gre deta th etn a5 heninytin dite da sed @ aCe CEE dee e Oe tenner Beebe ty dae Cas ‘ ‘ SE ee oe ee Peer oe bod tne eh Ret Rete: hte he et BOGE: He Ethe Os GROOT. oh heh 4 ©, aR mee 6 dates bah eh les ee 4 ew de & Grmenwd int Netw tN hot, tat ae LOW Anwa don ee ee ee | et bette We het Oren tant Mehdi Mee dn tt te te wom alg, + Pe er ee er ny ered Re ae ee oe oe eos ee) Cea ee PRT LG ete On oe, bid + oer, ee nn he ee Voth oh tonkond Hk te f= ath the tin he dae om. Rh ot ho Belew H Mediaweek hana UR A rlehye mtd shen AA POE et ed ay Med AR den Oe Que tetn gi nade ue | 6 sts defn dnt stor eee) ee in toe he eee tah Meds hth eed a debate af Mest de bathe ee oe 4 PU Ed ee het ecdotte ORR ve Yea 9 bg ee dk Obed teh ee OO ee ee od tet AON ati tg ee Seba: ested i 140. dee ah sabe tk BoB Reem, Orestes - v tthe te hewn aaah en eee Smt wae . " * a ahi etree ote * +. eh eet os epee Oh ee + Orbe 404 tote ee aehye + arte a ay er ee eer Meet ae eer ee or ee meng ort ther on se teeth He Here to tend S Lenalel oe hats iaTanats aa Tete Sy te tellete tte . 46GB bee ie & id Foe Fh ety the ts oe > OMY OE Hatetbeden Ce ee ee ee TT ee ee ee ee ee re Me etre U cterting ig ty, er ee AA ROEM ae OE ee ee ee - eeerd Re eed re = ete «hte to eee oe ere icar eearee ein 4 Me Aete a Eee Horn ede a omer ent daael se Ae et ohn do we the Hos bat aN Atte te ary ae Cee eee ere eres ak Cee eb eo ee oe | ated ee oe Le ta ae coe? ee ee eo oe ‘ + . me paeearqenk daca is sty" hh nhc rad Qor 4s dy he te 5, pee ee eee re ee Wanted dil doer dating 4 Hee ee ek Cd Ho dette od mths thor fake 4. thu Ce ee ae ee Ns Mmten tirdn Oho inn dent Hee deaee 4 or tetheaek cae tee are 4 ok oe ee, eo ha eG he hs ee, Hey 4 “peggy dot bt % 1 ote de 4-4 4-14.15 wa tt ta ter haphecth Ftal So doer ta tried og 1 Oa a Ot TA Wits Pe tee eee eC oe eee ee rey « Sk EO AL heh Se pedo tl GR te de tel 4. lie eoted. Vs: Bhs Sete tet ett Ried dined eb heme ear Od & 45-4 aod E-8 E h chm t thas Eg hed hd EL ge nce he tee tated hed dete Anti, Hel SM Q04 4 be le sal w ‘ a ae oes eae eet ee ee ea - re aoe ee ea 4 4S ee 4 Chee Ho tes See + teh hehe hed, trige Atrdrne 4 “ue 1-3 uo hi Ac rth oa " 7 Oriel dG | beqea did Se ee ee ee or cer ae eer Heh Bh eee hide m ene dad Ctr wth Ged tet +e 6 OO ew tee “ ot K Ge ee, he oe. ee es - ’ ares HOA athe ter ee ee: ae eee pe vA bed tM, Heke Rate ged eriens Seer ee re eee oe SECS Obagi go Pd dda gedetig ‘ PG Ae eee eH rete 4-88 teas 4 Vo R AS Go ead oD vt Ohl a) PLS Mende Sete wien Ghee Cy Seep ‘ SAE Re BE eRe be nd ao Re de he Gite le trol t | I 4th ted Unto Ades Motel eth Te Ae the Gotemed Goma dae E Ar Ew od i-* i dot ere Gao en ahoh re het Ao E” YS eR hed AE ng de hal Widdeee ds le isi age oy Moh he Re whe Ay net « He AP Ue he te tarde Hee Wi Melek Gaited. a Cee ee ee ee tee hake We SNe ge ete geen Pe eee ae ee 8 Wot dn ds Yow hee Ged wer ON te ee “ tid we FES EN Ee Deb had dnt F dementias ah ae Rat rte Aenean * Hh tae Pr eee a ee « WOO 8 he tet gta) 1d Go Merged actee et es ALE NIRA AEA AMAMA EAM EM Othe teh Es ge ; o 2 Sere Ch ed 4 oe Set Weds 4d erletebede ) 4 due Laine th deiaton ee ck oo a +e We aee Crete dete at dds hate 4<% wards Ae Gee Weed hh Pd. 8 ote a Few SE oh a ete Herbs WH Eee t AS ages $6 Mme. god tne atort, Arte cine te Hd g Ck SR RE tert he yt a< Und qeQee Qetre td edrdeed “+ ~~ + Oe ted oy Ante tet s) On" ee a eee tee Th totowed Cadet Ean a dee eh * sa%e oe - sa *y OM hth WH H ey oe etn ttn eh Oo LOM these ertaen ced tg ak c Meterg eda deaets Pigtail Pry istai LPL yale Ws wd ee brates & hep hetee “ace he OR ee Getdede te etre de FRO NE a Em eee Ge Ee et he ee 19. W d- Feed AP ROWE He eh Rd LG othe ae Coeed ae Pea eh wash tbe Sey Ue Saco Ot ott ok oe OH ee ede te gre 14 A eed-de 4 + Me tet hed Md ett y ee Chiba a Pew eed eae” 4° 9M perhy eed ath 4 doth ogeas ti eh ® Re trae wae te 8 ee bate wk : ‘On the Dependence of the Refractive Index of Gases on Temperature.” By G. W. WALKER. Communicated by Professor J. J. THOMSON, F.R.S. “Solar Prominence and Spot Circulation, 1872—1901.” By Sir NorMAN Lockyer, F.R.S., and Dr. W. J. S. LockYEr. “On the Evolution of the Proboscidea.” By Dr. C. W. ANDREWS. Communicated by Professor E. Ray LANKESTER, ERS. “On the Cytology of Apogamy and Apospory.—I. Preliminary Note on Apogamy.” By Professor J. BRETLAND FARMER, F.R.S., J. G. S. Moors, and Miss L. Dicsy. The Society adjourned over the Haster Recess to Thursday, April 30. 1X April 30, 1903. Sir WILLIAM HUGGINS, K.C.B., O.M., President, in the Chair. A List of the Presents received was laid on the table, and thanks ordered for them. The Croonian Lecture: “The Cosmical Function of the Green Plant,” was delivered by Professor TIMIRIAZEFF, of the University of Moscow. The following paper was read :— ‘Preliminary Note on the Use of Chloroform in the Preparation of Vaccine.” By ALAN B. GREEN, M.A., M.D. Communicated by Wetrowrr, ©.B. ER.S. May 7, 1903. Sir WILLIAM HUGGINS, K.C.B., O.M., President, in the Chair. A hist of the Presents received was laid on the table, and thanks ordered for them. In pursuance of the Statutes, the names of the Candidates recom- mended for election into the Society were read, as follows :-— Bayliss, William Maddock. _ Rutherford, Ernest. Bridge, Thomas William. Sampson, Ralph Allen. Copeman, Sydney Monckton. Stead, John Edward. Darwin, Horace. Strahan, Aubry. Hiern, William Philip. Symington, Johnson. Mallock, Henry Reginald Arnulph. Townsend, John S. Masson, David Orme. Whitehead, Alfred North. Perkin, Arthur George. The following Papers were read :— 1. “On Lagenostoma Lomazi, the Seed of Lyginodendron.” By Dr. Ff. W. Oxiver, F.1.S., and Dr. D. H. Scorr, F.RB:S. x II. “ On the Physiological Action of the Poison of the Hydrophide. By Dr. LEonaRD Rocers. Communicated by Major A. ALCcocK, F.R.S. III. ‘* Preliminary Note on the Discovery of a Pigmy Elephant in the Pleistocene of Cyprus.” By Miss D. M. A, Bate. Com- municated by Dr. HENRY WoopwarbD, F.R.S. IV “ Experiments in Hybridisation, with Special Reference to the Effect of Conditions on Dominance.” By L. DONCASTER, B.A. Communicated by Dr. 8. F. HARMER, F ES. Sa May 14, 1903. Professor CAREY FOSTER, Vice-President, in the Chair. A list of the Presents received was laid on the table, and thanks ordered for them. The following papers were read :— 1. If. IIL. VI. VIL. VIII. “On the Discovery of a Species of Trypanosoma in the Cerebro- spinal Fluid of Cases of Sleeping Sickness.” By Dr. ALDO CASTELLANT. Communicated by the Malaria Committee of the Royal Society. “The Combination of Hydrogen and Chlorine under the Influence of Light.” By P. V. Bevan. Communicated by Professor J. J. THOMSON, F.R.S. “On the Photo-EHlectric Discharge from Metallic Surfaces in Different Gases.” By Dr. W. MANSERGH VaARLEY. Com- municated by Professor J. J. THomson, F.R.S. . “The Hlasmometer, a New Interferential Form of Elasticity Apparatus.” By A. E. Turron, F.R.S. . “ Meteorological Observations by the Use of Kites off the West Coast of Scotland, 1902.” By W. N. SHAaw, F.ES., and W. H. Dinas, F.R. Met. Soc. ‘On the Radiation of Helium and Mercury in a Magnetic Field.” By Professor ANDREW GRAY, F.R.S., and Dr. W. STEWaRT, with R. A. Houstoun and D. B. McQuUISTAN. “A New Class of Organo-Tin Compounds containing Halogens.”’ iby) Professor W. J. Porn, F.R.S., ands. J. PEACHEY. “The Xanthophyll Group of Yellow Colouring Matters.” By C. A. Scnunck. Communicated by H. T. Brown, F.RS. of i i WK 4 eb Nene) Laat yh eer (Parnes Raye He abate a! Po Xl May 28, 1903. Sir WILLIAM HUGGINS, K.C.B., O.M., President, in the Chair. A List of the Presents was laid on the table, and thanks ordered for them. The following Papers were read :— I. “On the Bending of Waves round a Spherical Obstacle.” By Lorp Raytetcy, O.M., F.R.S. | II. “Sur la Diffraction des Ondes Electriques, & propos d’un Article de M. Macdonald.” By Professor H. Poincark, For. Mem. RS. III. “ On the Theory of Refraction in Gases.” By G. W. WALKER, M.A. Communicated by Professor J. J. THomson, F.R.S. IV. “An Analysis of the Results from the Kew Magnetographs on Quiet Days during the Eleven Years 1890 to 1900, with a Discussion of certain Phenomena in the Absolute Observa- tions.” By Dr. C. Cures, F.RS. V. “Ona Remarkable Effect produced by the Momentary Relief of Great Pressure.” By J. Y. BUCHANAN, F.R.S. VI. “Evolution of the Colour-pattern and Orthogenetic Variation in certain Mexican Species of Lizards with Adaptation to their Surroundings.” By Dr. H. Gapow, F.R.S. VII. ‘* Researches on Tetanus—Preliminary Communication.” — By Professor HANS MEYER and Dr. F. Ransom. Communicated by Professor KE. H. Staruine, F.R.S. VIII. “The Hydrolysis of Fats in vitro by Means of Steapsin.” By Dr. J. Lewkowirscn and Dr. J. J. R. Mactzop. Com- municated by Professor E. Divers, F.R.S. IX. ‘On the Optical Activity of the Nucleic Acid of the Thymus Gland.” By Professor A. GAMGEE, F.R.S., and Dr. W. JONES. X11 x. “Note on the Effect of Extreme Cold on the Emanations of - Radium.” By Sir W. Crooxkss, F.R.S., and Professor J. DEwAr, F.R.S. XI “On the Adaptation of the Pancreas to Different Food-stuffs— Preliminary Communication.” By F. A. Barnpripar, M.B. Communicated by Professor STARLING, F.R.S. The Society adjourned over the Whitsuntide Recess to Thursday, June 11. June 11, 1903. Annual Meeting for the Election of Fellows. Sir WILLIAM HUGGINS, K.C.B., O.M., President, in the Chair. The Statutes relating to the Election of Fellows having been read, Major MacMahon and Dr. A. D. Waller were, with the consent of the Society, nominated Scrutators, to assist the Secretaries in the examina- tion of the balloting lists. The votes of the Fellows present were collected, and the following Candidates were declared duly elected into the Society :— Bayliss, Dr. William Maddock. Rutherford, Professor Ernest. _ Bridge, Professor Thomas William. | Sampson, Professor Ralph Allen. Copeman, Dr. Sydney Monckton. | Stead, John Edward. Darwin, Horace. Strahan, Aubrey. Hiern, William Philip. Symington, Professor Johnson. Mallock, Henry Reginald A. Townsend, Professor John S. Masson, Professor David Orme. Whitehead, Alfred North. Perkin, Arthur George. Thanks were given to the Scrutators. >) June 11, 1903. Sir WILLIAM HUGGINS, K.C.B., O.M., President, in the Chair. A List of the Presents received was laid on the table, and thanks ordered for them. X1V The following Papers were read :— I. “The Bending of Electric Waves round a Conducting Obstacle : Amended Result.” By H. M. MACDONALD, F.R.S. Il. “ On the Propagation of Tremors over the Surface of an Elastic Solid.” By Professor H. Lams, F.R.S. III. “The Diffusion of Salts m Aqueous Solutions.” By J. C. GRAHAM. Communicated by Professor W. E. Ayrton, F.R.S. IV. “On the Structure of Gold Leaf, and the Absorption Spectrum of Gold.” By Professor J. W. MALLET, F.R.S. V. “On Reptilian Remains from the Trias of Elgin.” By G. A. BouLENGER, F.R.S. VI. “A Method for the Investigation of Fossils by Serial Sections.” By Professor W. J. SOLLAS, F.R.S. Vil. “An Account of the Devonian Fish, Paleospondylus Gunni, Traquair.” By Professor W. J. Souuas, F.R.S., and Miss I. B. J. SOLLAS. VIII. “The Measurements of Tissue Fluid in Man. Preliminary Note.” By Dr. G. OLIveER. Communicated by Sir LAUDER BRUNTON, F.R.S. IX. ‘Observations on the Physiology of the Cerebral Cortex of the Anthropoid Apes.” By Dr. A. 8. F. GRUNsBAUM and Pro- fessor C. S. SHERRINGTON, F.R.S. June 18, 1903. Sir WILLIAM HUGGINS, K.C.B., O.M., President, in the Chair. Dr. William Maddock Bayliss, Professor Thomas William Bridge, Dr. Sydney Monckton Copeman, Mr. Horace Darwin, Mr. William Philip Hiern, Mr. Henry R. A. Mallock, Mr. Arthur George Perkin, Mr. John Edward Stead, and Mr. Aubrey Strahan were admitted into the Society. XV A List of the Presents received was laid on the table, and thanks ordered for them. The following Papers were read :— I. “Surface Flow in Crystalline Solids under Mechanical Dis- turbance.” By G. T. BEmmBy. Communicated by F. H. NEVILLE, F.R.S. Il. “The Effects of Heat and of Solvents on Thin Films of Metal.” By G. T. Bettpy. Communicated by F. H. NEVILLE, F.R.S. III. “The Forces acting on a Charged Electric Condenser moving through Space.” By Professor TrRouTON, F.R.S., and H. R. Nosie. IV. “On the Discharge of Electricity from Hot Platinum.” By Dr. H. A. Witson. Communicated by C. T. R. WiLson, ELBS. V. “The Bionomics of Convoluta Roscoffensis, with Special Refer- ence to its Green Cells.” By Dr. F. W. GAMBLE and F. W. KEEBLE. Communicated by Professor 8S. J Hickson, F.R.S. VI. ““New Investigations into the Reduction Phenomena of Animals and Plants.—Preliminary Communication.” By Professor J. B. FARMER, F.R.S., and J. E. S. Moore. VII. «“ The Action of Choline, Neurine, Muscarine, and Betaine, on Isolated Nerve, and on the Excised Heart.” By Dr. A. D. WALLER, F.R.S., and Miss 8. C. M. SowrTon. VIII. “ The Physiological Action of Betaine extracted from Raw Beet-Sugar.” Dr. A. D. WALLER, F.R.S., and Dr. R. H. ADERS PLIMMER. IX. “On the Physiological Action of the Poison of the Hydro- phide. Part I].—Action on the Circulatory, Respira- tory and Nervous Systems.” By Dr. L. RoeeErs. Communicated by Dr. A. D. WALLER, F.R.S. X. “The Spectra of Neon, Krypton and Xenon.” By E. C. C. Baty. Communicated by Sir WILLIAM Ramsay, K.C.B., FLR.S. XI. “The Spectra of Metallic Arcs in an Exhausted Globe.” By A. Fow er, A.R.C.Sc., F.R.A.S., and Howarp PAYN, F.R.A.S. Communicated by Sir N. LOCKYER, K.C.B., F.R.S. XVI XII. “The Phenomena of Luminosity and their Possible Correla- tion with Radio-Activity.” By Professor H. EK. Arm STRONG, F.R.S., and Dr. T. Martin Lowry. NII, “Cyanogenesis in Plants, Part I1].—On Phaseolunatin, the Cyanogenetic Glucoside of Phaseolunatus.” By Pro- fessor W, R. DuNSTAN, F.B.S., and Dr. T. A. Henry, XIV. “ The Magnetic Expansion of some of the Less Magnetic Metals.” By Dr. P. E. SHaw. (With an Appendix by G. A. ScHotr.) Communicated by Professor POYNTING, XV. “A Study of the Interaction of Mercury and Nitric Acid.” By Professor CHANDRA RAy. Communicated by Sir Henry Roscog, F.R.S. XVI. ‘Separation of Solids in the Surface-layers of Solutions and Suspensions.” By Dr. W. RAmspen. Communicated by Professor F. Gorcu, F.R.S. XVII. “Some Prelimimary Observations on the Assimilation of Carbon Monoxide by Green Plants.” By Professor W. B. BoTToMLEy and Professor HERBERT JACKSON. Com- municated by Professor REYNOLDS GREEN, F.R.S. XVIII. “On the Oocyte of Tomopteris.”. By W. Watiace. Com- municated by Professor McINTosu, F.R.S. XIX. ‘“ Upon the Bactericidal Action of some Ultra-violet Radia- tions as produced by the Continuous Current Are.” By J. E. BARNARD and H. DE R. Morean. Communicated by Sir Henry Roscor, F.R.S. XX. “The Longitudinal Stability of Aerial Gliders.” By Pro- fessor G. H. Bryan, F.R.S., and W. E. WItLiiAMs. XXI. “On the Synthesis of Fats accompanying Absorption from the Intestine.” By Professor B. Moore. Communicated by Professor SHERRINGTON, F.R.S. XXII. “ Radiation in the Solar System.—Its Effect on Temperature and its Pressure on Small Bodies.” By Professor J. H. POYNTING, F.R.S. XXII. “The Properties of Aluminium-Tin Alloys.” By Dr. W. CARRICK ANDERSON and G. LEAN. Communicated by Professor Miers, F.R.S. XVil XXIV. “The ‘Hunting’ of Alternating-Current Machines.” By BERTRAM Hopkinson. Communicated by Professor Ewine, F.R.S. XXV. “The Theory of Symmetrical Optical Objectives.” ByS. D. CHALMERS. Communicated by Professor LARMOR, Sec. Ras: XXVI. “The Differential Invariants of Space.” By Professor A. R. ForsytTn, F.R.S. The Society adjourned over the Long Vacation to Thursday, November 19. PROCEEDINGS OF THE ROYAL. SOCIETY. NN RNR RNR AAR AAARAARANRAN™ “Maenetic Observations in Egypt, 1893—1901.” By Captain H. G. Lyons, R.E. Communicated by Professor RUcKER, sec. R.S. Received June 6,—Read June 20, 1901. The following Magnetic Observations have been made at various times during the years 1893 to 1899, at first with a Declinatorium, made by Bamberg, of Berlin, the property of the Egyptian Govern- ment, and later with a Kew Magnetometer, No. 73, and Dover’s Dip Circle, No. 99, both kindly lent by the Council of the Royal Society on the recommendation of Professor A. W. Riicker, F.R.S. These observations are most conveniently divided into five groups, each of which includes observations made during a single period and with a single instrument— I. Observations made with a Declinatorium by Bamberg, of Berlin : (a.) In the neighbourhood of Cairo, 1893-—1894. (b.) In the Lybian Desert, near the Kharga and Dakhla Oases, in December, 1893, and January, 1894. (c.) In the Lybian Desert from the Wadi Natrun to the Baharia Oasis, April, 1894. IJ. Observations taken with Kew Magnetometer, No. 73, and Dover’s Dip Circle, No. 99, in the Nile Valley from Cairo to the 2nd Cataract, November, 1894, to June, 1896. (a.) Declination. (b.) Dip and Horizontal Force. III. Observations taken to determine the Diurnal Variation of the Declination. IV. Observations taken at Helwan, near Cairo, in November and December, 1898. V. Determination of the Annual Variation from the above observa- tions and those of various observers in previous years. OE. LXXI, B Z Capt. H. G. Lyons. [June 6, The dechnatorium used at first consisted of a horizontal circle furnished with two verniers reading to 30”, while the magnet was balanced on a vertical steel pivot, in a box which occupied the centre of the horizontal circle. Attached to the magnet was a mirror, in which the reflected image of the cross wires was observed with a small telescope placed excentrically. The same telescope also served to observe the sun or a meridian mark for determining the geographical meridian. The geographical positions were taken from Maps Nos. 740, 662, published by the Intelligence Division, War Office, or from astro- nomical observations taken on the spot; the former are indicated by * and the latter by Tf. In each case the time of the observation is given as Cairo mean time, 7z.¢., 22 5™ 89 fast of Greenwich, since all the observations were made before the time of the 30° meridian E. of Greenwich was adopted as civil time for Egypt. D’Abbadie’s station, beside the Great Pyramid at Giza, was occu- pied on May 10, 1901, to obtain improved values of the secular variation. The observations were made between 1 P.M. and 4 P.M. when the electric tramway was not working. Since there are, as yet, no self-registering magnetic instruments in Egypt, it is impossible to reduce the results obtained to a single epoch with any accuracy ; they are, therefore, given as they were originally observed. Helwan, 20 kiloms. south of Cairo, was chosen for the observations of 1898 and 1899, since the electric tramways of Cairo render obser- vations impracticable even if the amount of iron in the present observatory building at Abbassia did not vitiate all observations taken there. For this reason the value of 5° 36’ west for the declination, given in the ‘ Bulletin Mensuel’ of the Abbassia Observatory for June, 1886, is wholly wrong. In observing with the Declinatorium, the feet of the tripod were firmly pressed into the ground, and this was found sufficient for the precision obtainable with the instrument. With the Kew pattern magnetometer, however, wooden pickets were driven firmly into the ground, and the feet of the tripod rested on these, thus avoiding any errors due to the tripod sinking into soft or sandy soil. All the observations which follow may be considered as satisfactory ones taken under favourable conditions, since all those which were interfered with by high winds, sand storms, &c., have been omitted. The stations where igneous rocks are known to be near enough to affect the results somewhat are marked with an asterisk on page 10. At several places on the Bahr el Abyad granite masses, and occasion- ally basaltic rocks, rise through the sandstone, and the high value for declination obtained at Renk (page 23) is probably due to this. 1901.] Magnetic Observations in Kgypt, 1893—1901. 3 The times given are those of the middle of the observation, 2.¢., for Girga— Mean. Declination observation ...... Sde—9 Ik It 9.5 ESE SVADTALION .....00s So oesacie-ec 9) ir 9.20 DECOMG: VITALION... aea.c.scsses 9.34—9.47 9.40 MEME CULO hicsie sooth caroeiese's OVS 5 —— © The time of the horizontal force value is given as the middle of the time of the vibration observation. Where one value is given for the dip one needle was used; where two values, the value obtained from each is given. Deflection observations as well as the vibration observation were made on each occasion for the observations given in Table II(d); in the observations given on page 22 it is mentioned when it was not possible to take them. Capt. H. G. Lyons. [June 6, ‘PLET ‘uorytpodxy enue, Fo qisuvay, Aq PoulUoJop WOL}ISOg c.6L F hp Be Ne p08 F rO08 F 9.7% F 6.S& F e.1% ¥ 0-41 eA a4? 9.5% Leg jai Nie Gn9c "489M UOTFVUTpOO(, og 8 cy GI eye Sill eh WIL Oc 8L shy OL or Gh ST SU sity “9UUl} UBIUL OLS) 6-8-G¢ €6-O[-&T &6- 8-6 6-6-1 €6-6-06 €6-8-L4T €6-8-VI V6-O1-94G VO-8-VG §6-O1-TT 66-8-TT €6-O1- —- re = = Sivir Se 0g ¢T 8-6L 8Z Ob FT a P-8S 82 OE II IZe. yh LPLEE- 0 0& 0-12 8% 0Z 8 $3 ae ae GaAGee OG Os) oi 60628 -O OP GI z.Ge 9z Oo on & 9.0L 62 OF &T Cc ; , AON DOES WS) Oi Ge etal “ety ; e Sue 5) Ne “OOTOF “OUNTY ‘dt “OUT [@JUOZLLO FY. UBITH OUTRO) sd UvOUT OMB) e+ ee ee oe oe (a9 CC O=GilsGme oe aie: so © oe ROG ler boar kerr ee STO AA FRAN TAT iid Shae Ba 2 he ee “ ULIRYSOP, Tp’ AA. P6-ZI-0Z |° °°" Aatosoq: urIqnyy Qeeug «mg GG-Z-LzZ oe a3 6 6é (14 CG6-3-2 | °° i QYS1t “eZLVAT IpeAA c6-8-Z “ OWIN FO yUCq Foy ‘(wateg) STW C6-2-82 oN Jo yuvq qJoy oquig nqy C6-2-G GO m0 60 a4 VJ] ‘apeuy C6-Z-12 HOH O68 6 3 66 *6-ZI-1 '* *OTINT JO YUR FSA ‘OYso.r0 yy G6-Z-0% ‘airel pia) eis) (a4 “¢ (a4 G6-3-61 Bot si Wo] “enqog Oia pee Se iS Vystn “eperg CG=GsuE OUD 60 Ob (a9 C6 Tun payf C6-Z-9OT seo '** OTINT JO yUvg gjoy = “eyyeq UG | ‘O0UT 10 Capt. H. G. Lyons. [June 6, | Mirgissi Fort*..... | | Wadi Mogharin*. es Description of Stations observed at with Magnetometer and Dip Circle. Place. | Description. | ND OES 556 Gd agar | The same as for Declinatorium. Girga ............ | S.W. of huts, which are 150 metres west of canal head on N. side of steamer landing barge. Dendera.......... | 100 metres E. of point where steamers stop. Luxor ............ | On river bank, 50 metres north of house painted red and white, on bank south of Luxor. Miuallla timer race | 2 kiloms. S. of irrigation resthouse. IBIES oligo Geen rao 100 metres north of landing barge, and on river bank. UNGHENAS S COMA Bi Se _ 40 metres from bank, and 120 metves S. of landing barge. Assuan* .......... | Rest camp, 8S. side of 2nd hut from N.W. corner. Elephantine*...... | 25 metres S.E. from gateway of ancient temple. Ist Cataract*...... | South side of the Bab el Gedid. Phile*.........-.. | Till 29 Nov. E. side of kiosk, on quay, after then on roof | of Temple of Isis, N. end. Dabod...........-. | 200 metres N. of ancient masonry quay. Taifa* ............ | 100 metres from S.E. corner of village, on river bank. Kalabsha* ........ | 830 metres from river bank, and 70 metres north of great | temple: Abu Hor* ........ | On river bank, 100 metres 8. of steamers’ stopping place. Gerf Hussein ..... | At entrance to temple. Wakkkar — | = | 248) abe) — | 25:8 24°9 | 24°6 | 25 °5 | 23-9 | 24°6 | — | 23:8 | 28-7 | 23-8 | 23-5 | 28:3 | 24:2) — | 21-7 | 22:8] — | — | 29-1} 294] 92-6] — | 20°65 | 216 | 21:0 . Se barn fp eee | Soe Rea] BAG) 24-9 | — | 28-8 | 24-8] — | aaa) oae-oroeig |) —— (eg ore | = | 22} a ogy I lor sgii a og |) al | 10 _ -- = 27 3 26 ‘8 — — | 26-2 26°2 (M69) 25-7); — | - _ — — = = = == | 29S) — | 26:2 | 23°1 | 25:3| — | 22:8 | 22°9 | 24-4) ee i jet ao | eon alt Me | caw [eo | 248 | | a6 | 26-6 | azn | oye) = ere | oer] — | — | — bos-8)| pera | 2dee | (| Baee | 28nd | Moon | ee A 26°6 | 277 | 26-4 | PRO | 28°2 | 27°5 | — | 26-4) 26-4] — | 27°7 | 28-2 | 28-2| — | 282/951] — | — | 29-4 | 26-4 | 28-2 | 25-5 | 27-7 | 28-2 | 27-5 | : [_— _ _ 93 } | | 276 | 264 283 2/5273 275) — | — — — BRR Oe HOB Me TROT AG Te [pee ey a egies teas — | — | 26:8} — | . _ = = = | 38) - a coi fe & ear | AOS) | BO) 275) | el BG ral opebal) == olla Gedeleap Bn |) — | 27-7) — | 26"2'| 25:8 |-27-1 | — | 26°0 4 : Phe te Sea eer. | eH Se ae Nae 30d fet ea es ric lll | Se I ee le yday || aa ete eke oe — | 246) 26-1) — | 266 24-6 | 264) 2 sli Ni JAC en a8 Mee ea Ee erga oi ee YY) a1 Yr I RYT porary ome a 2S — | = z Te Oy ee ae | ee hes ae pom | = | 2a) | 2468} =a | See opty a ee ee oprah 120-8 | 2846 | 228 | | | | - Ce te a 14 Capt. H. G. Lyons. [June 6, Woe 4 April, 1896. | Hours. | 1 Mer Ue fare i eS ——————————————— Ba ee en eS ee ee ce = poe | Si | = | 8 | 22-3; — | 22-8 | 21-5 | 21-0) 19°9 | 20-5 | 20-8) 19% 9 — | 244) — | 22:3) — — | — | = 10 — |}26;] — | — | — | — | 21| —, | 20:9 11 | 24°6 | 28-0 | 27-1) — | 23-7 | 23:5 | 24-6 | ==) sieaeae | Noon ||) Bash — =» | 25-5 | 24°9)| 24°8 | 26-0 | 24-6ule24om Reha 3 — Pe ones Pid. | | = | 2 | 24°6 | 28-2 | 26-0 | 24-6 | 24°6 | 24-2 | 25:8 26-0 | 22-9 | 3 —};—-—}]—]- ;— |} —} — |) — | 4 — (246) — | — | — |}; — | — | — | = 5 | 22:9) — | 22:8 | 22-1 | 22-3 | 21-0 | 22:8 | 21-9 | 2266 | | | Hal IV. Magnetic Observations at Helwan, near Cairo. (See table on opposite page.) V. Secular Variation. As many early observations as it has been possible to find in various works have been collected in the following tables for the purpose of determining the average annual rate of change. Generally the decli- nation appears to have been annually decreasing by about 6’ to 7’ in the first half of the century, becoming, however, 3’ to 4’ only in the second half. For the dip the available observations are very few ; but from those at Alexandria and at the Great Pyramid, Giza, the rate of decrease appears to be about I’ to 15 annually. For the horizontal force the observations are too few, and give results which are not very concordant. The observations before June, 1894, were made with the declina- torium, so that those made at later dates are more reliable, since a Kew pattern magnetometer with a unifilar suspension was used. The values obtained cannot be considered as laying claim to a high degree of accuracy, seeing that in most cases the hour of observation is not given in the older observations ; still in most cases the number of years elapsed is sufficiently great to reduce the error thus intro- duced into the value for the annual change to small dimensions. (Tables are printed on pp. 16, 17 and 18.) 15 Magnetic Observations in Hgypt, 1893—1901. 1901.] “sIvIg “PP ap Aq poatosqg » ‘GEST PU QGST 10F ‘oateg ‘quowycedocy AoAING oy} Jo qaodoy [vorsoloLOOJOT Ol} UL UAAIG OAV SMOTZVATIEGO OAOGL OT} JO S[IeZop oI[J, ‘aatzisod A[QVIEVAUT JSOUU[V SEM U014Z001L00 sty, “WOISL04 IOF PoIaLLOOU] (¢) ‘e (s 66 (z) ‘[ o[poou diy (1) "$040 NT LT M/ "MOLYBUI PCT 6g OT Vo cE €¢g 6&1 c& GL ‘um 'Y ‘OME, GEIOE -O L{IZ0E- 0 G8Ll0& O "OOLOF [VJUOZILO FT OE 1g 1 SUC EN I ‘QG681 ‘oltegQ Jevau ‘UeM[OFT 4 SUOeAIESGGO L- 6& ¥*(Z)S: OS (2) L- 86 (2)6- £6 (2)8- 46 (1)0- 6& x(1)G- 1 *(D9- TE / So 2 2S 2&2 2 & { ‘dig OoUsv “AT ee oe or oe 638T LOF SON[CA UuvO TN eo 0.8 6Z 66 ee oe PZ 66 ee PL ce eo oe 6 66 "5" G daquivda(T "* ** T TOQUIOAO NT eeeee [ve 1040}IO “age 16 Capt. H. G. Lyons. V. Secular Variation of Declination. Place. Observer. | Date. | Declination | | west. | Alexandria.... / Quesnot and | 1798 | Le OlaO Nouet. | | F at ? | 1842! 8 24 0 7 : ..| Gissfeldt. 1876 | 5 55 9 e .| Capt. Leslie,| 1890; 5 3 0 y (Ramleh) d’Abbadie. | 1884 | 5 6 2 CANO bocdonos00 oa) | oUBSE eer [aS39)) Ogee Oa » (Old Cay WAbbadie. | 1885 | 2 a a -Abbassiam®.. «|| EGA 8) S899. || 4 50 as , J. Mogattam..| d’Abbadie. | 1885 | 4 56 5 | | | EUG: 1893 | 4 386 0 ,, Great Pyramid d’Abbadie. | 1885 | 4 48 9 is sae EO Ken UT My Bo. is Mh ly JELCh Ibe 1901} 3 48 6 Helwan .. Ly HeGenee: | 1693 nena hs . | Vleanme ror) 1899 sie ATaest ets | 1899. Siwa Oasis ........ | Cailliand. | 1819 12 30 0 ess - | Jordan. | 1874; 7 33 O Mandisha ‘(Gatons | | | Oasis) . | Cailliand. STO || 2 A | Jordan. 1874 6 34 8 | jel s@rl by, 1894.) 25)" 8 a9 | Assiut .... | Cailliand. | 1819/12 0 0 | Jordan. 1874 | 5 42 0 @Abbadie. | 1885 | 5 45 9 FGalic 1893 | 4 49 O | Kharga a arga | Oasis) . .| Cailliand. | 1819 | 12 10 O Jordan. 1874: 6 24 0 | H.G.L. 1894; 5 3 O | *Dr. J. Ball.| 1899) 4 15 0 iL? 3) Change in 93 years . >) per annum ....es.6 eeee0ese#e Edfu Temple— Magnetic bear:ne of axis, 1799 NS Zipeeweeice 92 3) Ghange) iHU9S IveEars is ishe id o.e sos «0 0,0 5 per annum VOL. LX = O + sae approximately, where 6 is the coefficient of expansion of the solution, 4 Mr. W. R. Bousfield and Dr. T. M. Lowry. [June 19, of temperature at points not more than 50° below the freezing point of water. The point of intersection is an important constant ° (a) 4, —40° =e. Fre. 1.—Influence of temperature on the conductivity of very dilute solutions.. for any given solution, and we have been accustomed to refer to it as the conductiwity zero of the solution. At this temperature, if the observed relationship between conductivity and temperature should continue to hold good, the conductivity of the overcooled solution would become zero, and the conductivity zero may therefore be satisfactorily compared with the absolute zero of a gas thermometer at — 273°. | The influence of temperature on the conductivity of aqueous solutions: has been investigated by Grotrian, Kohlrausch, Arrhenius, Déguisne,* and others. In an important paper which has recently appeared,T Kohlrausch has deduced from Déguisne’s measurements the values of the temperature coefficient in aqueous solutions of infinitely great dilution, and has arrived at the important conclusion that the conduc- tivity temperature curves for all such solutions would, if produced, cut the axis of temperature at points lying within a degree or two of — 38°5° C., the differing slope of the lines being compensated by their differing curvature (fig. 1). This temperature, which is almost inde- pendent of the nature of the solute, is evidently a fundamental constant of the solvent, and may be referred to as the “ conductivity zero of the solvent.” The physical meaning of the conductivity zero is a matter of some importance. Kohlrausch states that the viscosity of water may be represented by the formula 7 = 2°989 (¢ + 38:5), which leads to an infinitely great viscosity at — 38°-5, and suggests that at this * } | | | : 10 ” »” ” | ” | Sterile ” »” ” » 3) ” ” Sterile ” Growth Growth ” ” i ” »” | » ” 26 oo» ” » | ” | ” ” ” Sterile ” » ” ” ” ” ” ” ” ” | ” ” ” ” | | | Ne ey 5 | 4 | ey | re of 5 3 Sterile Sterile Sterile Sterile it 2 Sterile Sterile |} - Pe, | rh i Sterile Sterile | | 100 os ” | ” | ” ” ” ” » Growth ” » ” ” ” | ” ” ” ” » ” | ” ” | | | | 1,000 ” » ” ; ” ” ” »” ” Sterile ” ” ” > ” | ” ” ” ” ” ” H ” ’ | | | | | | 10,000» ” » ” ” ” ” ” ” ” ” ” ” ” ” | ” ” | ” ” » ” , | | | | i} | 100,000 ” ” ’ | ” | »” ” ? ” \\ » ” ” ”» ” »” ” ” | ” | ” » ” } o ” = EES es —— u Ul ! | | | A. B. W. had been inoculated against typhoid; W. B. L., F. N. W., and A. B, were normal men; J. N. had recently convalesced from typhoid. The sera were in each case tested within 2—3 hours after the blood had been withdrawn. The culture were in all cases aerobically grown 24-hours-old broth cultures. ‘The serum was in each case allowed to act upon the culture for 18—-24 hours at a temperature of 37° C. cok 1902. | The Colour-physiclogy of Higher Crustacea. 69 “The Colour-physiology of Higlier Crustacea.” By FREDERICK KEEBLE, M.A., Reading College, Reading, and F. W. GAMBLE, D.Se., Owens College, Manchester. Communicated by Pro- fessor 8. J. Hickson, F.R.S. Received July 16, 1902. (Abstract.) The following statement is a condensed summary of the results of a research into the form and physiology of the pigment-bearing organs (chromatophores) of certain Schizopod and Decapod Crustacea with especial reference to the effect of light on these organs and on these animals. The evidence for the statement will appear in a full and illustrated form in the “Philosophical Transactions.” The Grant Committee of the Royal Society allotted £25 for this research. A. The Influence of Light. 1. Under the influence of light the secretory activity of certain organs is modified: an acid substance appears periodically in the ‘“‘liver” and muscle: the appearance and disappearance of acid sub- stance in liver and muscle coincides broadly with nocturnal and diurnal colour-change. 2. In the progressive movements and orientations of the whole animal called forth by light, background is the most important factor : more powerful than change oi light-intensity. By change of back- ground, black to white, the direction of a light-induced movemen’ may be reversed. 3. The response of the chromatophore-pigments to light is two-fold : direct ; and indirect, through the mediation of the eye. The indirect response alone leads to an enduring redistribution of pigment. 4, The ultimate effect of monochromatic light on pigment-movement is the same as that of white light. As with the latter, so with mono- chromatic light, background—white (scattering), black (absorbing), mirror (reflecting)—determines the nature and extent of the pigment- movements. In describing an effect of light, that light must be con- sidered in combination with its background. Neglect to do this must lead to erroneous conclusions. 5. “ Reaction to background ” is traceable to the eye, and is probably a consequence of an asymetrical distribution of retinal pigment brought about not by changes in the amount of light falling on the eye, so much as by changes in the way in which light falls on the eye. B. The Role of Pigments, 6. The phenomena presented by the pigments are not exhaustively explained by any “ protective” hypothesis. VOL. LXXI. G 10 The Colour-physiology of Higher Crustacea. [July 16, The chromatophores are centres of metabolic activity, and from them a nocturnal translocation of a blue substance takes place. There is evidence that this blue substance is produced from, and at the expense of, the diurnal chromatophore-pigments. The blue substance passes from the chromatophore-gentres, persists for a time in the body, and ultimately disappears. C. Morphoiogy. 7. The chromatophore-system of Mysidean Schizopods is built upon « common plan, of which the various genera and species present severally a constant modification. This system we call the primary chromatophore-system. To it the colour-pattern is due. 8. Decapod Crustacea possesses a primary and a secondary system of chromatophores. The primary system appears in the embryo, is completed in the ‘‘ J/ysis-stage,” and persists throughout life, but takes no part in colour-pattern. - The secondary system arises in an early stage of development, increases in extent throughout life, and produces the colour-patterns of the adolescent and adult. 9. The chromatophores of the primary system are profusely branched, few in numbers, segmentally arranged and centralised ; those of the secondary system are sparsely branched, numerous, irregularly arranged and decentralised. D. Histology. 10. The chromatophores of Myside are multicellular organs. Those of the neural group are developed from the epidermis. Losing their connection with the epidermis they acquire a close relation with the central nervous system. The distribution of the primary chromatophore-system follows that of the ganglionic parts of the nervous system. 11. The chromatophores of Decapods are plurinuclear connected structures: their distribution is not confined to the ganglionic parts of the nervous system. E. Taxonomy. 12. The primary systems afford assistance in the determination of genera and species. By their aid, animals in early, as well as in late, stages of development may be diagnosed. F. Jnheritance. 13. The several adult colour-patterns of Palemon and Crangon are constant, and develop directly. The evidence tends to prove that both secondary and primary chromatophore-systems are inherited. 1902. ] Observations on “ Flicker” in Binocular Vision. 71 14. The adult colour-pattern of Hippolyte cranchii is constant, but develops indirectly. The adolescent possesses a special colour-pattern developed in large measure in relation with the primary system of the zoea. Both persist though concealed by the independently de- veloped adult pattern. 15. In Hippolyte varians, several adult colour-patterns occur. They develop indirectly. The primary system is the same in all. In the adolescent, three distinct colour-patterns arise :—“ barred,” *“liner,” and “ monochrome.” These may persist, becoming barred, liner, or monochrome adults. Or either “barred” or “liner” may, by developing superficial or deep chromatophores, become a monochrome. Or, by localised superficial developments either “barred ” or “liner ” may give rise to a “blotched” adult colour-form, under which the adolescent pattern is hidden. The primary system is inherited: the adolescent colour-patterns are possibly inherited ; but inheritance is immaterial since the final goal is reached by any adolescent road; that is, the adult colour-pattern of Hippolyte varians, is the result of environment. “Observations on ‘Flicker’ in Binocular Vision.” By C. 8. SHERRINGTON, M.A., M.D., F.R.S. (Thompson-Yates Labo- ratory of Physiology, University College, Liverpool). Re- eeived July 30, 1902. The connection between the physiological state and reactions of the two retine right and left is close in many respects; this is true particularly and peculiarly for their areas that are conjugate in binocular vision, that is, which receive corresponding images of objects perceived in the binocular field. The observations at basis of the following communication attempt to obtain further information regarding the nature of the tie between these retinal so-called ‘“ identical spots.” ) ssherrineten [July 30, in same way as in 2, continuous steady sensation from the conjugate left retinal spot, And when the intensity of that continuous sensation from left retina corresponds with an intensity of light stimulus more than half. that of each stimulus employed in repetition for the conjugate of the right eye, Then the rate of intermission required for obtaining steady sensa- tion from disc A is higher than that required for obtaining it from the dise B. Of the two discs, seen under speeds sufficient to give steady sensation from both, disc B appears the brighter. | Kaperiment 4.—The dise A representing uniocular vision as before. When the disc B represents intermittent right retinal stimulation, together with, in the same way as in 2 and 3, steady sensation from the conjugate left retinal spot, And when the intensity of that steady sensation from left retina corresponds with that due to a light stimulus of less than half the in- tensity of each member of the series of repeated stimuli employed for the conjugate spot of the right eye, Then the rate of intermission required for obtaining steady sensa- tion from disc A is higher than that required for obtaining it from the dise B. And of the two discs, both seen under speeds sufficient to yield steady sensation, disc A appears the brighter, unless the field offered to the left eye is given by closing that eye or otherwise screening with a homogeneous darkness. Lapertment 5.—The disc A representing uniocular vision as before, And when the disc B represents intermittent right retinal stimula- tion, the intervals of intermission exactly corresponding with periods of ilumination of the conjugate spot of the left retina, and the inter- mittent stimuli being equal in intensity and duration for both right and left conjugate spots ; Then the rate of intermission required for obtaining steady sensation from disc A is higher than that required for obtaining it from dise B. Of the two discs, both seen under speeds of intermission sutticient to yield steady sensation, the disc B appears to be of a brightness not ~ obviously different from that of disc A. Lupertment 6.—With dise A as before. When disc B represents intermittent right retinal stimulation, similar in every way to that applied to the conjugate of the left retina, except that its phases of light and shade precede or succeed those applied to the other retina by an interval of hali a phase length, Then the rate of intermission required for obtaining steady sensa- 1902.] Observations on “ Flicker” in Binocular Vision. 1) tion from disc B does not appreciably differ from that required to obtain it from disc A. Of the two discs, at speed sufficient to yield steady sensation, disc B cloes not obviousiy differ in brightness from disc A. Experiment 7—When disc A represents the binocular combination described for disk B, Experiment 1, and the disk B is as described in Experiment 6, the rate of intermission required to obtain steady sensation from A is higher than that required for obtaining it from B, but at speeds sufficient to yield steady sensation from both discs the two discs appear to be of equal brightness. Kaxperiment 8—With disc A, as in Experiment 7, and disc Be as described in Experiment 6, the frequency of repetition of stimulus required to yield steady sensation from disc A is slightly higher than that required to obtain it from disc B, but at speeds sufficient to yield steady sensation from both discs the two discs appear of equal brightness. Lperiment 9.—With disc A, representing the binocular combination described for disc B in Experiment 5, and with disc B, as described in Experiment 6, the frequence of repetition of stimulus required to yield steady sensation at disc A is slightly lower than that required to yield it at disc B; but at speeds sufficient to yield steady sensation from both discs, the two discs appear of equal brightness. The observations show (i) that Talbot’s law, unimpeachable (over a wide range of ordinary luminosities) for the single eye, is not ap- plicable to combined binocular vision, that is, that if the two eyes functioning together in binocular vision are considered as functionally to combine to a single organ, Talbot’s law does not hold good for that organ as it does for the single eye, right or left. (11) That increase of luminosity of an intermitting image does not always necessitate increase of rate of frequency to extinguish its flicker ; and conversely, They also show that the “Fechner paradox” regarding binocular luminosity makes itself apparent under “ flicker” examination as well as under “ brightness” estimation. It seems that the physiological sum of two luminosities perceived through conjugate retinal areas is of a value intermediate between the individual values of the two component luminosities. Among experimental difficulties incident to the experiments may be mentioned the increased perception of flicker under paracentral as com pared with central locus of stimulus on the retina, as noted by Exner and by Charpentier. Interesting experimental difficulties were also occasioned by the reciprocal and often antagonistic influences exerted by one retina upon another in ways studied and described recently by Dr. W. McDougall.* It was turther observed that. binocular colour mixture did not seem * ‘Mind, 1901. 76 Influence of Temperature of Liquid Air on Micro-organisms. to be either rendered easier or impeded when the components were applied by alternating right and left stimuli as compared with the method of applying them by coincident right and left stimuli. “On the Influence of the Prolonged Action of the Temperature of Liquid Air on Micro-organisms, and on the Effect of Mechanical Trituration at the Temperature of Liquid Air on Photogenic Bacteria.” By ALLAN MacraDyEeN, M.D. Com- municated by Professor JAMES DEwarR, F.R.S. Received August 2, 1902. In previous communications it was shown that an exposure for twenty hours and for a period of seven days to the temperature of liquid air (about —190° C.) had no effect on the vitality of micro- organisms, whilst an exposure of ten hours to a temperature as low as that of liquid hydrogen (about — 252° C.) was likewise without an appreciable effect.* Further experiments have since been made in which the influence of the prolonged action of the temperature of liquid air on organisms was tested for a period of six months. The bacteria employed were non-sporing forms, viz., 5. typhosus, Bh. coli communis, and Staphylococcus pyogenes aureus, along with a Sac- charomyces. The bacteria were directly immersed in the liquid air, either on cotton-wool swabs enclosed in a perforated metal case, or on small loops of platinum wire. The yeast was washed and pressed, then wrapped up in rice paper, and directly exposed. Samples were removed and tested at various intervals up to six months. In no case could any impairment of the vitality of the organisms be detected. The fresh growths obtained were normal in every respect, and the functional activities of the organisms were unaffected. ,The typhoid bacillus retained its pathogenic properties, and responded typically to the agglutination test; the colon bacillus exhibited its normal properties ; the Staphylococcus aureus produced pig- ment on solid and an active hemolysin in fluid media, whilst the yeast exhibited its fermentative power unimpaired. The above experiments show that a prolonged exposure of six months to a temperature of about —190° C. has no appreciable effect on the vitality of micro-organisms. -To judge by the results, there appeared no reason to doubt that the experiment might have been successfully prolonged for a still longer period. * © Roy. Soc. Proc.,’ February 1, 1900; cbid., April 5,1900 ; dbid., May 31, 1900. An Intracellular Toxin of the Typhoid Baerllus. tell The ordinary manifestations of life cease at zero, but at — 190° C. we have reason to suppose that intracellular metabolism must in addi- ‘tion practically cease—as a result of the withdrawal of two of its -cardinal physical conditions, viz., heat and moisture. It is difficult to form a conception of living matter under this new condition, which is neither life nor death, or to select a term which will accurately describe it. In previous experiments it was found that the photogenic bacteria preserved their normal luminous properties after exposure to the ‘temperature of liquid air. On rethawing, a rapid renewal of the photogenic properties of the cells occurred. The light is apparently produced by a chemical process of intracellular oxidation. The ‘feasibility of triturating micro-organisms at the temperature of liquid air has now been experimentally established in the case of the typhoid bacillus and other bacteria.* The effect of such mechanical trituration -at the temperature of liquid air on the luminous properties of the photogenic bacteria has now been tested. The experiments have -shown that the effect of such a trituration is to abolish the luminosity -of the cells in question. _This points to the luminosity being essentially a function of the living cell, and dependent for its production on the intact organisation -of the cell. I am indebted to Professor Dewar for valuable suggestions, and to Mr. Sydney Rowland and Mr. J. E. Barnard for their assistance in the experiments, which were carried out at the Jenner Institute of Pre- ventive Medicine. ‘“ An Intracellular Toxin of the Typhoid Bacillus.” By ALLAN MacraDyEN, M.D., and SypNEY RowLanp, M.A. Commu- nicated by Lorp Lister, F.R.S. Received August 14, 1902. The existence of a specific toxin produced by the typhoid bacillus has hitherto not been demonstrated, although it has been assumed by .analogy with other organisms and by reasoning from the clinical course of the disease. Such a poison must either be intracellular or extracellular. That it does not exist in filtered cultures of the organism is the ‘common experience of bacteriologists. Its absence from such cultures might be due, however, to unsuitability of the soil used for growing the organism. * “The Intracellular Constituents of the Typhoid Bacillus,” Allan Macfadyen cand Sydney Rowland, ‘ Centralblatt f. Bakteriologie, vol. 30, 1901, No. 20. 78 An Intracellular Toxin of the Typhoid Bacillus. Accordingly the first step in the search for the body in question consisted in substituting for the usual broth and peptone media, culture fluids approaching more nearly in constitution the natural body soils which clinically support the growth of the bacillus. For this purpose, the organism was grown on the actual intracellular juices of the following organs and tissues obtained in a fresh condi- tion from the ox or calf :— Intestinal mucous membrane, mesenteric lymphatic glands and spleen. In each case the intracellular juice was brought to the requi- site degree of alkalinity and used as a culture soil under the following conditions :— 1. Aerobically. 2. Anaerobically. 3. With addition of normal. human serum. 4. After heating to 55° C. for 20 minutes. After from 4 to 6 weeks’ growth the organisms were filtered off and. the filtrate tested for toxicity in guinea-pigs. With the possible: exception of one spleen juice, none of the fiuids thus obtained ex- hibited any acute toxic power. It thus became necessary to search: within the typhoid organism itself for the missing toxin. For this purpose the organisms were grown on ordinary beef broth agar, and. after careful washing with distilled water were disintegrated in a mechanical contrivance at the temperature of liquid air (— 180° C.) This course was taken to satisfy the conditions that—(1) No chemical. change should take place during the disintegration, and (2) The organisms could be disintegrated alone, without the addition of any triturating substance, the necessary subsequent removal of which might vitiate the composition of the resulting mass. If such a dis- integrated mass be freed from whole bacilli (f present) and from other: suspended insoluble particles by centrifugalisation, an opalescent fluid results, which on inoculation into animals in small doses invariably proves toxic or fatal. It is therefore concluded that the typhoid. bacillus contains within itself an intracellular toxin. The typhoid cell juices obtained by the above method are being examined for immunising and other properties at the Jenner Institute: of Preventive Medicine, where the above investigations have been. conducted. Fracture of Metals wader repeated Alternations of Stress. 79 “The Fracture of Metals under repeated Alternations of Stress.” By J. A. Ewine, LL.D., F.R.S., Professor of Mechanism: and Applied Mechanics in the University of Cambridge, and J. C. W. Humrrey, B.A., St. John’s College, Cambridge, 1851 Exhibition Research Scholar, University College, Liverpool. Received August 11, 1902. (Abstract.) The paper describes an investigation by means of the microscope of the process by which iron becomes “fatigued” and breaks down when subjected to repeated reversals of stress, as in Wohler’s experiments. It is shown that although the greatest stress is much within the limit of elasticity (as determined by the proportionality of strain to stress: in an ordinary tensile test), it produces rupture after many reversals. The first visible effect is the production of slip-bands here and there on individual crystals. These gradually become more numerous: they also become accentuated and broadened and their edges turn rough and burred, apparently as a result of grinding of one surface on the other over the plane in which slip has occurred. Ata later stage certain of the slip-bands develop into cracks, whose existence can be demonstrated by repolishing the specimen, when the slip bands which have not opened into cracks are obliterated, but the cracks remain visible as actual fissures. As the process of reversals goes on, the cracks spread from crystal to crystal, and fracture ensues. In the particular material dealt with, Swedish iron, having an elastic limit in tension of about 13 tons per square inch and a breaking strength of 23°6 tons per square inch, it was found that a stress not exceeding 9 tons per square inch, when reversed some millions of times, was sufficient to develop cracks and to bring about the fracture of the piece. Stresses of 8 and even 7 tons per square inch were found to. develop slip bands which would probably turn into cracks under a sufficient number of reversals. The paper is illustrated by micro- photographs taken at various stages of the destructive process. 80 Dr. J. Mur. On Changes in Elastic Properties [Aug. 11, “On Changes in Elastic Properties produced by the sudden Cooling or ‘Quenching’ of Metals.” By James Morr, B.A., D.Sc., late 1851 Exhibition Science Research Scholar. Com- municated by Professor Ewine, F.R.S. Received August 11, 1902. (Being part of a Thesis submitted for the degree of Doctor of Science, Glasgow University.) It is well known that when steel is quenched from a red heat, its elastic properties suffer a profound change, the material becoming extremely hard and brittle. It is also known that quenched steel, when tested under tension, exhibits no distinct yield-point, Hooke’s law is departed from quite gradually until abrupt fracture occurs at a high stress. The effect produced on copper by quenching has been considered, at least in some respects, the reverse of that produced in steel. The experiments to be descvibed in this paper, however, show that with mild steel, soft iron, copper, zinc, aluminium, brass, and so probably with all metals, quenching from high temperatures produces effects which are analogous to one another; in all cases there is a marked loss of elasticity produced by quenching, low loads producing appreciable permanent extensions or ‘ sets.” The method of experimenting need not be described in detail here, as it was identical with that described in the paper by the present author on “ The Tempering of Iron hardened by Overstrain.”* The new 5-ton testing machine of the Cambridge Engineering Laboratory was however employed for many of the experiments in preference to the large 50-ton gun machine previously used. Small strains of exten- slon and of compression were measured by instruments of Professor Ewing’s design—extensions by means of the 4-inch extensometer illustrated on p. 2, ‘ Phil. Trans.,’ A, 1902, compressional strains by the instrument illustrated on p. 79 of Professor Ewing’s book on ‘“ The Strength of Materials.” The heating of the specimens was obtained by means of the gas furnace used in the earlier experiments on temper- ing after overstrain, temperatures being measured by a Callendar’s direct-reading platinum resistance pyrometer. The hot specimens were “quenched” by plunging them vertically into a large tank of cold water. The results to be recorded in this paper may be gathered from an examination of the accompanying series of diagrams. ‘The diagrams give with one exception (Diagram 4) the results of tension tests, and it need only be remarked that all the stress-strain curves have been ‘sheared back” in the manner suggested by Professor Ewing, and = Phils (rans, eas! 902 pei 1902. | produced by the sudden Cooling of Metals. Sl fully described in a paper by the present author “On the Recovery of Iron from Overstrain.”* The amount by which the curves have been sheared back is marked at the foot of each diagram. Thus, in extension of the 4-inch length for every 4 tons of stress. For example, the extensometer readings for stresses of 4, 8, and 12 tons per square inch were 120, 240, and’ 360 respectively ; the numbers actually plotted were 20,40, and 60. The origin for the measurement of extensions has been displaced for the various curves of each diagram in order to avoid a confusion of the curves. Diagram No. 1 shows the elastic properties of an annealed specimen of mild steel. The specimen was subjected to a series of tension tests, the load in each test being carried just to a yield-point. Recovery from the overstrains produced by the passing of the successive yield- points was effected by heating the specimen to temperatures of trom DiacRam No. 1.—(Mild steel-annealed.) tons fin? 50 | dee es | | ae | 40 Dee | onmcon | 4 C:Q83) ao : ( MI MOG ZO Pee 7 [ada] a S fs Gen fAboui2'5"¢ i | ormitled, Vi, VL ae ; alae Extensions diminished by i385 of an Wich forever sa Scate.-Wnit=x5,rofaninch.2 1? of stress. Load in tons per square lneh. Diameter of specimen = 07331. Length under test = 4’"00. Fracture occurred at 40 tons per square inch original area. Extension (includmg all yield-points) = 0°38 on 4 inches. 200° to 250° C. The specimen broke at the fifth yield-point, the breaking stress being 42 tons per square inch, or about 40 tons per square inch taking the original area of the specimen. After the pass- ing of each yield-point the diameter of the specimen was of course * “Phil. Trans.,’ A, vol. 193, 1899. 82 Dr. J. Muir. On Changes in Elastic Properties [Aug. 11, shghtly reduced ; this was allowed for in the succeeding loadings, the load being always applied in tons per square inch of actual section. Diagram No. 2 shows the elastic properties of the same steel after it had been heated to 500°, to 650°, and to 700° C., and quenched in water at about 15° C. Each of the specimens employed was broken by a single continuous loading. Curves A and B show that quenching from 500° and from 650° C. had little effect on the elastic properties of the steel. The specimen from which Curve B was obtained had been more thoroughly annealed before quenching than Specimen A, and this may account for the lower breaking load and greater ultimate extension obtained with Specimen B, although all the specimens employed were primarily annealed. Curve C shows that a marked change was produced in elastic properties by quenching from 700° C. DiaGRam No. 2.—(Mild steel-quenched.) 38 fons/ ir: [ara oO . Extensions diminished (as below). tons/in? GO a Extensions diminished by 8, hs ofaninch for every 4tonsofstres Scatle:.-lWnit= 75% ofaninch 2! 2 Diameter of Specimens A, B, and C = 0°37. Lengths under test = 4'"00. The material after quenching showed no range of elasticity. Hooke’s law was departed from gradually from the lowest loads till ultimately fracture occurred abruptly at the high stress of 52 tons per square inch. The ultimate extension was only 07-10 on 4 inches. 1902. ] produced by the sudden Cooling of Metals, 835 Diagram No. 3 shows the effect produced on the elastic properties of Lowmoor iron by quenching from 700° C. Curve A was obtained from an annealed specimen of the material. A very clearly definéd yield-point is shown at the stress of 33,000 Ibs. per square inch. After the yielding at the yield-point had spread throughout the specimen, the load was steadily increased until fracture occurred at the stress of 50,300 lbs. per square inch. The extension produced was 0°95 on 4 inches, neglecting all the local extension which occurred at the point of fracture, or 1’-23 including the local extension. Curve B, which was obtained from a specimen which had _ been quenched from 700° C., clearly shows the loss cf elasticity produced by quenching. A curious recovery efiect was noticed in this test. The load was applied until a stress of 30,000 lbs. was attained, and DiaGRAmM No. 3.—(Lowmoor Iron.) tooo tbs/in? 35 reinch, - N QO Gi 3 Ny 9 G > Load in |000 pounds per sguk O Extensions diminished by pitho%rs of an inch for every 10000 as ofstress Scale.-Wnil=s55 of aninch 2! _? Diameter of specimens A and B = 0°44. Length under test = 47-00. Specimen A.—Annealed at 750° C. Broke at 50,300 lbs. per square inch. Extension 0°95 omitting, or 1/23 including, local extension. Specimen B.—Heated to 740° C., slowly cooled to 700° C., and then quenched in cold water. Broke at 61,300 lbs. per square inch. Extension 0”°61 omitting, or 0’"88 including, local extension. 8+ Dr. J. Muir. On Changes in Elastic Properties [Aug. 11,,. was then removed. The contraction which occurred on the removal of the load was almost perfectly elastic. Had the load been imme- diately replaced, the material would have shown perfect elasticity up to the stress of 30,000 lbs., but immediately the load was increased beyond this amount larger yielding would have occurred, and a smooth continuation of Curve B would have been obtained. The specimen was, however, allowed to rest for about an hour before the load was replaced and increased. This rest proved to have a comparatively large effect, the material showing very perfect elasticity up to the stress of 34,000 lbs. per square inch. At this stress a partial yield- point was exhibited (represented by about 6 units of extension on the diagram), and on increasing the load gradual extension was produced, until ultimately fracture occurred at the high stress of 61,300 lbs. per square inch. The ultimate extension is marked at the foot of the diagram, and was less than that obtained with annealed material. It may be recorded that another specimen of this Lowmoor iron was quenched from 700° C., but, before testing, this specimen was re-heated to about 200° C. in order to see if any appreciable return to the elastic condition illustrated by Curve A, Diagram 3, would be obtained. The behaviour of the specimen was mure nearly elastic for low loads — than is shown by Curve B, but all the main features of Curve B were corroborated ; gradual departure from Hooke’s law was obtained until fracture occurred at 61,500 Ibs. per square inch, the extension being 0-54 on 4 inches omitting, or 0”-9 including, local extension. Before leaving the consideration of iron and steel, the effect pro- duced by quenching iron, as illustrated by compression tests, may next be considered. Diagram No. 4 shows by a comparison of two compression curves: the change in elastic properties produced by quenching mild steel from a red heat. Specimen A, was a short annealed block of mild or . semi-mild steel; the diameter of the specimen was 1156 and its: length 1Z inches. The compression instrument employed enabled the contraction on a length of 14 inches to be measured to the 5.j555 of aninch. Curve A shows that the annealed material was elastic up to: the stress of 21 tons, but 23 tons per square inch had to be applied before really large yielding occurred. A tension test of this material . showed a well-defined yield-point at 224 tons per square inch. Specimen B was exactly similar to Specimen A, but the material in this case instead of being annealed was heated to redness and quenched in cold water. Curve B shows the marked loss of electricity produced by the quenching. The rounding of Curve B at the two points where the load was removed is probably to be accounted for by experimental errors of the nature of back-lash in the testing machine or compression instrument. ) | Diagram No. 5 illustrates the results obtained by experiments with PIOZ. | produced by the sudden Cooling of Metals. 85 copper rods. Curve A of that diagram was plotted from a tensile test made with the material in the condition as supplied. Curve B shows the elastic properties of the material after it had been heated to 630° C., and allowed to cool slowly, while Curves C,, C2, and C3 show the effect produced by quenching the copper from 500° C., from 550° C., and finally from 600°C. Specimen B showed more perfect elastic behaviour for low loads than Specimen A, but large extension is shown by Curve B to have occurred earlier with the annealed material. Specimen C was first heated to 500° C., and quenched in cold water. Diagram No. 4.—(Mild steel under compression.) % 9 ra Load intons persquare inch. Cee ctions diminished by,;£%,,'#° ofan inch forevery4 tors S) i . Scale:-lUnit= a5," ofan inch.2@_?. of stress\ Specimen A.— Annealed. a B.—Quenched from a red heat. Diameter of specimens = 1°16 inches. Total length -= 1% inches. Contraction measured on 14 inches. Curve C, was then obtained by applying and removing a load of 15,000 lbs. per square inch. The specimen was next quenched from 550° C., and Curve C2 shows the slightly greater loss of elasticity which was thus produced. Curve C3 shows the large effect caused by quenching the Specimen from 600° C. The breaking stresses obtained with the three specimens were 35,300, 33,500, and 32,300 lbs. per square inch of original area. These stresses were equivalent to 41,600, 40,800, and 44,200 lbs. per square inch, when allowance was made for the diminutions in area due to the large extensions oi the specimens before fracture. These corrections were made by VOL. LXXI. H 86 Dr. J. Muir. On Changes in Elastic Properties [Aug. 11 calculating the reduced areas from the extensions obtained (omitting the local extensions at points of fracture) and neglecting the small changes in density which are known to be produced by stretching. Copper thus resembles iron and steel in having its breaking stress Diagram No. 5.—(Copper.) S oe 0 Extensions diminished by pe of an inch for ary §000 pounds of stresss Scale -!Unit= == th of an inch °__! ? Diameter of specimens = 0-37. Length under test = 4’”00. Specimen A.—Copper as supplied. Broke at 35,300 lbs. per square inch original area, or after an actual stress of 41,600 lbs. per square inch had been applied to the bar. Extension on 4 inches, 0°74 omitting, or 1-04 including, local extension. Specimen B.—Heated to 680° C. and slowly cooled. Broke at 33,500 lbs. per square inch original area, or 40,800 lbs. per square inch actual stress. Extension on 4 inches, 0°87 omitting, or 125 including, local extension. Specimen C.—Quenched from 500°, 550°, and then from 600° C. Broke at 32,300 lbs. per square inch original area, or 44,209 lbs. per square inch actual stress. Extension on 4 inches, 1/50 omitting, or 1’"98 including, local extension. increased by quenching, but differs from iron and steel in giving a greater extension before fracture when in the quenched condition. The abrupt yield-point which is so striking a feature in the testing of annealed iron and steel, is not exhibited with copper. The ultimate extensions obtained with the three specimens of copper tested were 0:74 on 4 inches with A, 0’:87 with B, and 1°50 with C, omitting 1902.] produced by the sudden Cooling of Metals. 87 the local extensions at the points of fracture, or 17:04, 17:25, and 1”-98 respectively including the local extensions. Diagrams Nos. 6 and 7 may now be given without comment. They illustrate tests made with brass and aluminium, and it is shown in both cases that there is a loss of elasticity produced by quenching. When the quenched material has been once loaded it is brought approximately into the elastic condition, so that from a removal and reapplication of load a straight stress-strain curve is obtained. Diagram No. 6.—(Brass.) jooo Los/in: a uare inch a Ny Ny a 5 or Load in 1000 pounds per sg S) sa) & Se “| ne Rien Extensions diminished by jgc00'=* of a.n inch for every 5000 pounds | of stress.§ Scate-lUnit = a" ofaninch,.o_s_.2. Diameter of specimens = 0°44. Length under test = 4’”000. Specimen A.—Brass as supplied. Broke 4 times in machine grips at about 48,500 lbs. per square inch. Extension from 0’"87 on 4 inches after the first break to 1’°68 after the fourth. Specimen B.—Brass quenched from 600° C. (B,) and from 700° (B,). Broke at 42,500 lbs. per square inch after two breaks in the machine grips at slightly lower stresses. Extension, 1’”33 on 4 inches. Three specimens of zinc, of diameter 0:40, were also tested. The first—in the condition as supplied—broke at 21,000 lbs. per square inch, the ultimate extension being 2’’-04 on 4 inches. There was great local extension, the specimen being drawn to a very narrow neck betore fracture. The second specimen was quenched from 350° C. It yielded rather more for the lower loads than the first specimen, and broke at 20,500 lbs. per square inch with an extension of 0-41 on 4 inches. The fracture was quite abrupt, so that there was little or no 88 Dr. J. Muir. On Changes in Elastic Properties [Aug. 11, local extension. The third specimen was heated to 350°C., and allowed to cool in the air. The behaviour of this specimen was very similar to that of the quenched specimen. Rather less yielding was obtained at low stresses, fracture occurred at 21,400 lbs. per square inch, the ultimate extension was 0”°68 on 4 inches. The fracture was abrupt, so that there was practically no local extension. Two specimens of cast tin were also tested, but owing to the low melting point of tin quenching from 200° C. could only be tried. The Diacram No. 7.—(Aluminium.) ooo Lbs/in’. 25 IC Gu % ve tI S per 8 aH ourids < dir i000 p ss 2 at wv le = aL —t ensions diminished by zgsaq°t* of aninch for every $000 ee l S Pr aS ay a3 € of stress. Scate :-lnit=sg/s5" of ai7 inch.©__t__12. € Diameter of specimens = 0°35. Length tested = 4’-00. Specimen A.—As supplied. Broke four times in machine grips at 28,500 lbs. per square inch. Hxtension 0-14 on 4 inches. Specimen B.---Quenched from 550° C. Broke in the machine grips at ae ,200 Ibs. per square inch. Extension 0’°9 on 4 inches. quenched specimen showed rather greater extension at the lower stresses ; both specimens broke at 5250 Ibs. per square inch; the extensions were 0%5 and 0'°8 on the 4-inch lengths, but local ex- tension occurred in several places before fracture. In conclusion, it is proposed to consider how far the effects produced by quenching described above may be accounted for by the stresses set up in the material by the sudden ccoling and consequent contraction, the material, after SGI es | being no loner in what has been termed its “ state of ease.’ When a long cylindrical rod cools, the cooling takes place radially, and the end effects may be neglected. ‘Taking any cross-section of the rod, the outside ring will cool first and assume its elastic state; the 1902.] - produced by the sudden Cooling of Metals, 89 interior will then contract, and exert a radial pull on the outside solidified layer. This will put the material into a state of circum- ferential compression. If the tangential direction be called the direc- tion of X, the radial direction that of Y, so that the rod considered _ lies along the axis of Z, then the material in the outside layer of the quenched red of metal is subjected to a compressional stress in the X direction. If a layer of material be considered at some distance from the outside it will be found to be subjected not only to a compres- sion in the X direction, but also to a tension in the Y direction. For the outside solidified iayers are able to resist to some extent the radial pull due to contraction. A particle of material at a point such as A will thus be subjected to stresses p and ¢ in the manner illustrated in the sketch. Going nearer the centre of the bar, the pull due to con- traction of the hot material may be more than balanced by the outward radial pull due to the solidified material which has settled down under radial tension, so there may be a resultant outward pull all round the layer considered, and a particle such as B will be subjected to a circum- aT B. Z va Gi D a2 72) / Petre | pe hl Azad + t ferential pull, /’, as well as a radial pull, 4 There will be, of course, a gradual transition from material in the one condition to material in the other. Further, the stresses induced by sudden cooling will probably be severe enough to overstrain many layers of material, and, except in the case of portions which have been overstrained when quite cool, recovery from overstrain will be effected, so that the material will be left in an elastic condition, hardened as regards the stresses in question, and not in the semi-plastic state typical of material which has been recently subjected to overstrain. Now it is well known that when metals are deformed they alter very little in volume, almost the whole strain is one due to change of shape. It is only necessary then to consider the shear stresses applied by the systems of stresses illustrated above at A and B. A pull (¢ or #) is equivalent to a hydrostatic tension (47 or 47’) and two shear stresses in definite directions ; a push (p) gives rise to a hydrostatic pressure (17) and two shear stresses. A circumferential pressure (p case A) gives rise to the following two shears :— | H 2 90 Changes in Elastic Properties produced by Cooling. [Aug. 11, 14, giving contraction in direction of X and extension in direction of Y, 2A 2 ” ” x oe) ” Z. A radial tension (/ cases A and B) gives rise to the following two shears :— 148, giving extension along Y and contraction along Z, Pay NT ae * a Ve ms e Ke A circumferential tension (/ case B) gives rise to the following two shears :— 1p, giving extension along X and contraction along Y, ZIRT are 3 a ee. = . Z. It is necessary then to consider what effect these shear stresses, induced by quenching, have on the behaviour of a bar subjected to a tension (T) or a pressure (P) in the direction of the Z axis. A pull, T, in direction of the Z axis gives rise to the following two shears :— ly producing extension along Z and contraction along X, 27 5 ih lila alee . re ye A push, P, along the Z axis produces the following two shears :— lp giving contraction along Z and extension along X, 2p S) ” Z oy) 2 3 NE 3? It will thus be seen that the shear stresses induced in a bar of metal by sudden cooling have the effect of weakening certain layers of the bar as regards resistance to tension, and certain layers as regards resistance to compression. For the shear stress ly is applied along the same series of parallel planes as the stress 24, and although the stress 27 is directly opposed by the stress lg, the ‘‘ yielding” of the material must be determined by its strength in the weakest direction. Similarly the stresses lp and 2p are in the same directions as the stresses 2p, lap, so that the loss of elasticity exhibited by quenched material both as regards tension and compression has been accounted for. It may, however, be desirable to consider a little in detail what ought to be the behaviour under tension of, say, a bar of iron which has been subjected to the system of stresses described above. At the commencement of the loading the stress due to the applied load will be uniformly distributed over the whole section, but as soon as a very small load is applied, a long cylindrical layer of material (A’), which has been left by the sudden cooling under a stress of type 24 very nearly equal to the “ yielding” stress of the material, will yield. This yield- ing would continue to the enormous extent characteristic of a yield- point were all the material in the condition A’; but the weak layer, being surrounded by stronger material, the yielding is only allowed to continue to a very slight extent. This small yielding will, however, cause a redistribution of the internal stresses set up by quenching to 1902.] Harmonie Tidal Constants for certain Ports. 91 take place, and perhaps also a redistribution of the stress due to the applied load. This alteration in the distribution of the internal stresses must be such as to cause the surrounding strong layers to stretch elastically as far as the weak material has been permanently stretched. The alteration in the internal stresses will remain after the applied load is removed, as the material which has been perma- nently deformed will be unable to relieve the stronger material. The apparent permanent set which is shown with quenched material after the removal of applied load, may thus be due to the real permanent exten:ion only of the weak layers, and to the elastic extension of the strong layers produced by the new distribution of internal stresses. This explanation, however, does not suffice, at least in the case of iron and steel, to explain the behaviour of a quenched rod under applied stress, for Diagrams 2 and 3 show that such a rod may be stretched further than is compatible with elastic extension—even supposing some of the iron to have been overstrained to the maximum in the most favourable direction, without stretching nearly far enough for the yield-point of the iron to have been passed. Hence in the case of iron and steel recourse must be had to the explanations which simply attribute the obseived effects to’ the formation of allotropic modifications of the metal or to the changes caused by the transition of the carbon—always present—from one condition to another. In conclusion, it may be recorded that pieces of the iron and steel specimens used in this research were polished, etched, and examined under the microscope. In the case of the steel specimens the change. from the ferrite and pearlite structure shown with the annealed mate- rial to the martensite structure shown with the quenched steel was very striking. But in the case of the Lowmoor iron no difference was detected by the microscope in the structures of the annealed and of the quenched specimens, although, as shown by Diagram 3, the elastic properties in the two conditions were vastly different. “Harmonic Tidal Constants for certain Australian and Chinese Ports.” By Tuomas WRriGutT, of the Nautical Almanac Office. Communicated by Professor G. H. Darwin, F.R.S. Received August 1, 1902. ) Ballina (New South Wates), Princess Royal Harbour (King George’s Sound), Newcastle (New South Wales), Brisbane (Queensland), and Sydney (New South Wales). _ The tidal observations made at these five ports have been reduced by the aid of certain sums placed at my service by the Government Grant Committee of the Royal Society, and I am indebted to Professor 92 Mr. T. Wright. Harmonie Tidal Constants [Aug 1, G. H. Darwin for the loan of the apparatus he devised to facilitate the summation of hourly tidal heights, and to the Hydrographer, Admi- ral Sir W. J. L. Wharton, who supplied me with the observations. The whole of the observations were reduced by the methods devised by Professor G. H. Darwin.* The observations made at the three ae mentioned ports were derived from copies of continuous diagrams made by automatie tide gauges; those at Brisbane and Sydney were times and heights of high and low water. The observations in every case extended over a period of about 1 year, and were almost complete. The breaks in the continnity of the observations were so short that approximate values could be easily inserted by interpolation with very small risk of error. From the automatic records at Ballina, Princess Royal Harbour, and Newcastle the hourly heights were read off to the nearest one-tenth of a foot. The range of the tide at these ports is small, and an attempt was made to use a smaller unit, one-twentieth of a foot. An experiment with one month’s observations showed, however, that the hourly and daily sums for the month differed very slightly, whether the readings were taken to the nearest one-twentieth foot or to the nearest one-tenth foot only. Besides, when the diagrams for two con- secutive days were placed end to end there frequently appeared to be a difference of at least one-twentieth of a foot between the end of one day’s curve and the beginning of the next day’s. For these reasons it was considered to be sufficient to work to the nearest one-tenth of a foot, and that length was adopted as the unit. The heights being read off, the method followed was exactly that deseribed by Professor Darwin, except in one detail. The S sheet (that is, the sheet which is used for obtaining the hourly sums for the S tides) was not used. These sums were made on the sheets on which the hourly heights were entered from the diagrams, As in Professor Darwin’s method, the hourly sums were made in groups of days which could be built up into the 30-days’ period for the S tides and into the 74-days’ period for the other tides. The daily sums were made through- out the year (they are required for the long-period tides). By forming totals of these daily and hourly sums in appropriate groups they act as a check on each other, and the two sets of sums are settled. The hourly heights for the first 74 days were then entered on the strips, the strips were pinned to the M sheet, and the additions made. The total of the 48 sums was checked against a corresponding total made up from the daily sums and the hourly sums forS, Agreement among these three totals is a check upon the copying on to the strips, and also upon the sums for M. This slight modification of forming the sums for 9 from the original heights as read off from the diagrams sayes one shifting of the strips, and, if all goes well and the totals * ©Roy. Soc. Proe.,’ vol. 48, pp. 277-—340, and vol. 52, pp. 845—389. 1902.) for certain Australian and Chinese Ports. 93 agree, it seems hardly necessary to check. the entering on the strips. In all other respects Professor Darwin’s methods were followed exactly. As already stated, the observations made at Brisbane and Sydney were the times and heights of high and low water. They were reduced by Professor Darwin’s method. The observations were split up into four groups, each covering about one-fourth of a year. Hach of these groups was separately reduced, and means of the four values of x and H for each tide was taken as the final constant. These separate values form a check on the work independently of the systematic method of verification which was adopted in each stage of the computation. In the case of the more important tides, the agree- ment among the four values is very close. In the case of some of the smaller tides the differences are somewhat greater, but not great enough to make any serious difference in predictions based upon the constants. The constants for the five ports are given, with others, below. The small value of H (0-159 foot) for M» at Princess Royal Harbour gave rise to the suspicion that there was some error in the work. ‘This value is, however, borne out by the value of H for Mz at Batavia (Java) given in the American Tide Table for 1900. Batavia is there quoted as a “ Standard Port for Reference” for King George’s Sound, and the height there given is 0°154 foot, or only 0:005 foot different from that obtained for Princess Royal Harbour. Careful examination of the work showed, too, that the value given below is correct. Hong Kong, Swatow, Whampoa, Cooktown, and Cairn’s Harbour. The constants for these ports have been deduced at various times during the past few years. Except in the case of some of the Hong Kong tides, constants for these ports have not yet been published, and the present opportunity is taken to include them with the others. The observations made at Hong Kong, Swatow, and Whampoa were from records by automatic gauges; those at Cooktown and Cairn’s Harbour were observations of times and heights of high and low water. They were reduced by the same methods as the observations at the other five ports. The observations for the Chinese ports were kindly supplied by the Chinese Customs authorities ; those for Cook- town and Cairn’s Harbour by the Hydrographer. There seemed reason to suppose that the observations at Whampoa had not been very good, and the results for the tides See) Ma 25ME Mi ander, seem to be so uncertain that I have thought it best to omit them from the Table of Values. For a like reason the L tide is omitted from the results for Cooktown and Cairn’s Harbour. Harmonic Tidal Constants [Aug. 1, Mr. T. Wright. 94 | | OLY-V tH LV SPT '§ SS YL “‘-GO8T "ANOG. LE TT Smveaau@) “UMO}YOOYD | ZQz vz &48- 1 969: T Qaz Qgz 882-0 SLE. 0 VG9- OT GYD: @ ‘HST SVL Hel OTST "SL ST “8 BS o€8 ‘068T “8881 ‘Koupkg OZT: 2 “0 8ST ‘STS oLZ | ‘9-S98T “AV OUBQSLA UL 96 SIT | NS ‘vodurey AA * GCL: G o8G ‘S-F68T S51 826: O Ive 8&0. O Zz LVE- T gb 610: 0 gS 100- 0 Oz | G<0.- O 98 9TE- O gol G90: 0 989: “Ol 68 OTL ‘N 8% 08% “S-L681 "MOFRMG | | Pze Iz! 6L40-0 | 8940-0 Qze | S61 190-0 | 00-0 £97 7gz Lby- I €80. I OOI | OzZ 090-0 | 00-0 ZT of! 100-0 | {00-0 Le | ore L400: 0 | £00: O 06z SLz L9S-0 91Z- O 6£1 98 S10. O SIO. O 618. G 100: @ “H ,OL oVLL | HSS cee "N SL 6G “8 (GG 86 ‘688T “S68T ‘SIO -‘euryeg suoy | €€z L460: O ght §10-0 60z 86¢- I gh 660: 0 Tee 000: O 6gz 900: O Sgz 168. 0 Sz 8TL0- O St0-€ | “HPP IST | 8 LG Be ‘OO6L CL “MSN ‘OT JSVIMO AT OI — Hy F G00. 6 2 } W 9 | =. od Ute 110: 0 | = ak WN 6E¢ — cli ; 6SI- 0 =ys W ot€ = de €Z0- 0 =y/ NW 662 = HY hs 200. 0 =-H/s® toz = are Z10: 0 ee eel 8 abe = Hy me Z9Z- O =H/s°® 9gi a ate 610: 0 =H 926: I °V “ /0 SIT ‘** apngisuoy ‘S898 | T apngqyeyT | ‘L—9L81 "PUNO 8,09.1094 SUL ‘LNOqae Ay jedoy ssooultg 95 for certain Australian and Chinese Ports. t 1902] | Ports. UW for certa il Constant. 1.4 “PLOMUE i Hu 96 ‘ULOT} JOOTIOM 04 10449q SI AI JBY} JooJAtvdtUl O8 SI SOpTy OYA JO UOIZVUIOMALEJOp aU} JVYY VAPITOY 0F VOSVOL SI OLN] JVY} SOYVOIPUL y UY “poulatoyep you Item sopTy 9804 JO SonTVA oYy YVYI SoyvoIpuL — W—‘aLon ae ee a ———s = aes = a ee ec ey LG1 gf | L6 993 76 96 L6 Loe 10z L6 | = wil | 0S0. 0 Tc0-0 ~~ 800-0 G00. 0 Sél-0 | 882.0 082. 0 £90. 0 PLO. O GE. O =a 6 are | gt g iI | O42 the f of Bn | ae eS G06- 0 9V§-0 860.0 601-0 | *8%0 | Z£9P-0 99P- 0 E1P- O GE: O 8ZE- 0 | = Jet J meee ae, |. aes = ace 0 of Sb off obz bo Ss | ; a = — Wat | HOw | ikO | CHAO | M00 HZO.O. >.< eg —_ == | —— — | * | OZ1 o1t Hic Sol SLI | = i il | pee 2 \ | a FIN | ria a iz * | 690-0 €80- O L60- 0 T¢0- 0 P90. O | = ee | ae = we — He 67 10! 6 Q6I Sica | ao z4 | WAL | — — jf = — eo 0¢0- 0 g40. 0 ZOL- 0 Z80- 0 qn. ) say ™ | ae ae eee “4 i6 Sex E1z git Me owe Lee | a = a aes as 20.0 | 920-0 | 480-0 | 920.0 120. 0 = per ahes — -— —- | — 6S¢ O07 1O€ 661 Cz QgT = a Ga -- — — — PPL-0 | 801-0 | 490-0 | €F0-0 | 290-0 G10. 0 =i pe = = — — Ez ZQ 1Qz Olt 162 St = “lan — | == — — 960- 0 e10- 0 GE0. 0 910: O ZO: O £S0- O ruc! = tL LV Kee “OLS PL | HST o1ST |"H 0 ST |W 92 LTT |W 68 YIT| “a OT PIT | A £8 SI |'A PP TST a MO Stl * epngsuory | Ast geist el $,L6 G1 | S69 .88 | S/TEL6 | N/G GS | N,/€S8S |N 81.66 | 8 ,29.66 |'8 £9 68 S38 o a6 sees Sropngrgery "S-Z68T (06s) s838. ‘9-GO8T "S-P68T ‘8-L68T “6881 ‘8681 ‘CO6T “LZ-9L81 | ) 7 ‘pun, 8.095.100. “ANOGAV Fy eileen eg |, ue , “SUOY, : “ACTON «| Sury “inoque yy PL ced@) | TUMOR ROD] USS oUVG SIL CHORALE | ENORTONNS oUOFL euriied ‘QYSROMON | [RAO SsooUTT | 1902.] On Spherical Harnonic Analysis. 97 “On some Definite Integrals and a New Method of reducing a Function of Spherical Co-ordinates to a Series of Spherical Harmonies.” By ARTHUR ScHuSTER, F.R.S. Received May 50,—Read June 5, 1902. (Abstract.) The expansion of a function /(@) of an angle @ varying between 0 and 7 in terms of a series proceeding by the sines of the multiples of 6 depends on the fundamental theorem, | sin 70 sin g6 dO = 0, J 0 where p and q are integer numbers not equal to each other. Simi- larly if P, denotes the zonal harmonic of degree n, u = cos 0, and Pa : a eee OA) ee dpe ? the expansion of a function of @ in terms of a series of the functions 7 depends on the corresponding theorem, aril | Q2Qrdn = 0, where 2 and 7 are two integer numbers not equal to each other. In many practical applications a continuous function is given by means of its numerical values at certain points, ¢.g., for equidistant values of 6. Such cases present no difficulty when Fourier’s analysis is to be employed, because there is in that case a summation theorem exactly corresponding to the above integration theorem. If @ be replaced by pm/n, where p takes successively the values 1, 2, 3... , the equa- tion =i-— 1 ~ sin (ppm/n) sin (pgr/n) = 0 p=0 will hold true. ‘This allows us to determine the coefficients in the case of problems in which discontinuous values ef the function at equidistant points are known (¢.g., hourly readings of temperature or barometric pressure). If we assume that all Fourier coefficients beyond the nth vanish, n equations are obtained, each of which only contains one of the unknown quantities. If it is desired to expand a function in terms of cosines, a slight VOL. LXXI. I “98 Prof. A. Schuster. [May 30, modification must be introduced, the summation theorem in that case being p=n—l d+icospreosg7+ cos (ppr/n) cos (pgr!n) = 0, p=1 the first and second terms representing half the value of the produc: for p = 0 and p = n respectively. There is no corresponding summation theorem in the case of the functions Q?7, and the application of the method of least squares leads to a series of normal equations, each of which contains all the other coefficients. This has been one of the great practical difficulties in obtaining an expression for the series of spherical harmonics for the earth’s magnetic potential. F, E. Neumann has tried to overcome the difficulty by calculating coefficients a,, @2 . . . dj in such a way that p=F She o o fs a 5 QF (Hp) Q; (Lp) = 0. p=1 Here ;4, #2 - . - Pg ave the quantities for which the values of the function to be represented are known. Neumann’s process is equiva- lent to attaching weights proportional to a, to the different obser- vations, a sraectime against which theoretical objections mighe be urged. 2. The expansion in terms of a series of cosines and sines being so much easier than the direct expansion in terms of a series of the functions QZ, I have endeavoured to obtain the latter series by means of the former. ; It is well known that a function of an angle 6, which is confined to the values lying between 0 and z, may be put either into the form ly +a, cos 9+a2¢08 26+ ... dpcosp@ +... or into the form by sin @+0.sin 26+ ... dbysinpé+ The reduction to the series of spherical harmonics is accomplished by calculating and tabulating the coefficients in the series Cos pO — WACO A Oe oe eee eee sing? = BiQ2 BE sO ese. Seer The choice between the cosine and sine series is open to us, but it appears that great simplicity is gained by taking the former -series when o is odd and the latter when o is even. For in that case the coefficients Az and Bg will all vanish, as long as n is smaller than p. 1002;.° On Spherical Harmonic Analysis. 99 When it is therefore desired to retain only terms as far as the nth degree, the Fourier coefficients need only be calculated as far as p=n+1. The position of the earth’s magnetic axis, eg., only depending on the terms of the first degree, is completely determined by the coefficients 2 for « = 0 and a, a2 for o = 1. 3. The symbolical representation of the results of this paper is much facilitated by the introduction of a separate symbol for the product of alternate factors, n.n-2.n—-4...1, if n be odd,-or n.n—2 . 2if n be odd.- I propose to write n!! for such products, and if a name be required for the product to call it the “alternate factorial ” or the “ double factorial.” Full advantage of the new symbol is only gained by extending its meaning to negative values of n. Its complete definition may then be included in the equations | n!! = n(n—2)"1, | Ae Fie ee From this we may derive when 1 is negative and odd n+l 1 wi! = (=). 2 (=n —2) 0" while for n negative and even, the factorial becomes infinitely large. 4. The calculation of the factors AZ and Bg depends on the values of the definite integrals =1 +1 | Q7 cos pédp, | Q? sin pédp, —1 a | and these may be made to depend on the values of the integrals ) +1 +1 | 7 sin 46d. and | pQ?Z sin dp . =] S| ~ It is proved that +1 : +o—1)! (o+A)!! (W-A—2)!" -, Te AG = (nto 1) (o +A)! @-A—2)1 I aT \, Eee * (na) I (o —X—-2)!! (n+AX4-1) "7 Hoye =o, Beeyen, =o es) if x —o be odd, ii . ee (eo) (oA) (nn — A — 3) ee {va ag DG lA Dah ee Ea ee eee —1 BEG if n —o be even. The factor ¢ is equal to 2 or to z according as 7 +A is even or odd. 5. The integrals +1 | a 2 dp =I 100 Prof. A. Schuster. [May 30, are obtained in the form of a series having a finite number of terms. 6. To find +1 +1 Q’ sin pod» and | QS cos pédp, a = we may either express QZ or the trigonometrical functions in terms of _aseries of powers of sin 6. The second alternative leads to results which in general are more convenient. If we put M SED p-liprti Q=1; Ci =p; O2 = p.53 fo yee 3 Gg 282 (ptrA—2)!2, A! (p- d) 1! By — ls bw oS B, = 2-4-2 -E we find if o be even, p odd, and x even, fo, sin 28s = = ea {one hes _¢, 023 - cob et lots |G obo se ee n= Sn ee” n—3.n—5.n+4.n+6 - ae {3,.0-1.04+1-2 oo m-~4.n+5 and if o be even, p even, and 7 even, (nto)!!(n—4)!! "(n= —1)!!(n +3)! o— 5. oc oo — oes ee \ +B; 5 = = =" eee n—4.n—6.n+5).n+1 If n+o+>7 be even, the integral is zero. Similar equations are obtained for 1 g pO | Qz cos pOdp. =1 7. The final results are expressed as follows :— If gz denote the coefficient of Rzcos o¢ in the series of spherical harmonics when R° = 77Q7, and /7 is a numerical coetticient, ¢ bemg 1902.] On Spherical Harmonie Analysis. 101 the longitude, the results of the investigation may be put into the form p=n+l yg = > afr, when vn is even, o odd, and p odd. p=0 p=nt+l1 => OC ane. NOG i Gr).5. 5-0 PieMeN. p=1 p= ntl > Paes cs. tip, EVER WG even,... po odd. =1 p=n+l ' = SUR ein a Ch ce ee In these equations the factors aglz are the coefficients of the Fourier series (see § 2), and the quantities 7%, n7, m%, sf are numerical quantities, which (as well as their logarithms) are given in tables at the end of the paper as far as n = 12,0 = 12,p = 12. By means of these tables the numerical work is reduced toa minimum, and the coefficients of the series may be obtained as far as terms of the 12th degree. 8. The proposed method is speciaily adapted to deal with problems like that of terrestrial magnetism, in which the function to be obtained as a series of spherical harmonics is not given directly, but by means of its differential coefficients. The force directed to the geographical north may by Fourier’s analysis be obtained as a sum, the terms of which have the form cos c¢ cos p@, and sin of cos pO when o is even, and the form cos o¢ sin pé, sin o¢ sin p? when c isodd. Integrating with respect to 0, the magnetic potential is obtained in a form such that the transformation into the series of spherical harmonics may be proceeded with. A separate expression of the magnetic potential is derived from the force directed to the geographical east. 102 My. G. J. Burch. Contributions to [Sept. 26, “Contributions to a Theory of the Capillary Electrometer. [1.— On an Improved Form of Instrument.” By GerorcE J. Burcu, M.A. Oxon., F.R.8., Lecturer in Physics, University College, Reading. Received September 26,—Read November ZO IO Z: During the sixteen years that I have worked with the capillary electrometer I have had occasion to make between 150 and 200 instru- ments, and have therefore given naturally a good deal of thought to the problem of its construction. J have used eleven different forms, three of which are figured in my little book on the ‘Capillary Hlec- trometer in Theory and Practice,’ reprinted from ‘The Electrician,’ 1896. A fourth is in use in the Physiological Laboratory, Oxford, for the research on .nerve in which Professor Gotch and I have been engaged, and is in fact the improved form referred to on page 9 of my book. : It may be of interest to indicate briefly the points that must be observed in the design of the instrument. It must be simple, easy to adjust and to clean, and with reasonable care not lable to be broken. Above all it must be suitable for use with objectives of short focus and wide-angle condensers. For this reason I adopted in my first projection-electrometers the plan of placing the capillary within a piece of thick-walled burette tubing of 1 mm. bore, half ground away so as to form a trough of semi-circular section. A piece of thin cover-glass serves as a front to this trough, the lower end of which dips into the dilute sulphuric acid, the liquid rising in it by capillary attraction to a sufficient height above the level of the U-tube to enable the microscope to be focussed on the capillary within. In the first instrument of this type the trough was ground to fit the mouth of the U-tube like a stopper, and rested loosely in jit. It was. found, however, that as the acid loses or absorbs water with the changes of weather, the variations of level in the U-tube affect the- adjustment of the’ trough, bringing the capillary sometimes too far from the cover-glass for good definition and sometimes dangerously near it. Hence in the later improved type I fused the upper end of the glass trough on to a glass rod which was fixed firmly with adjusting screws to the brass support that held the capillary, the U-tube containing the acid being independently supported so that the end of the trough dipped into it. The short limb of the U-tube was made wide in order that the trough should not be likely to touch against it in the event of any accidental pressure on the instrument, and also because experi- 1902. ] a Theory of the Capillary Hlectrometer. 103 ence has shown that capillaries are less liable to become sticky with much acid than with very little. The only objection to this type 1s the difficulty of adjusting the capillary in so small a trough. The operation has to be effected under the microscope and is both delicate aud tedious. When it has been completed the definition is perfect, and the instrument not unduly fragile. But in order to clean it the same process has to be gone through again. I therefore determined to try an entirely different plan, on which the safety of the capillary should depend not on the rigidity of the supports by which the trough was fixed, but on their perfect flexibility and on the use of a trough so light that its entire weight might even be borne by the capillary. Fig. 1 is a perspective diagram of the instrument in its final form, and figs. 2, wu, b, c, and d, show the details of the trough which is the essential part. The support A is cut from a solid block ot ebonite 9 cm. long, 5 cm. wide, and 2 em. thick. It is first cut to shape, holes drilled for the binding screws E and F, and the piece B then separated from it by two saw-cuts. V-shaped grooves are cut to receive the capillary C, which is firmly clamped under B by Eand F. The longer limb of the U-tube D passes through a hole drilled lengthways through the lower end of A, which is slit about half way up with a wide saw-cut, so that it may be pinched together by the screw G. Adjustments for setting the capillary at right angles to the optic axis and parallel to the slit of the photographic recording apparatus, are provided for by the stout brass plate K, bent at right angles, one end of which is fastened by a binding-screw at L to the back of A, and the other by a similar screw M to the adjustible stand of the projection microscope. The brass plate K is so shaped that there is a space of about 3 mm. between it and the left-hand side of the ebonite support A, in order to leave room for the adjustment of the latter about L as a centre. The construction of the trough is shown in the full-size diagrams, figs. 2, a, b,c. A piece of mica, such as is used for lamp-shades, is cut to the shape a, with a pair of scissors. Two or three thicknesses may Fig. 1.—Half full size. 104 | Mr. G. J. Burch. Contributions to [Sept 26, be taken if one is not enough. Four holes are drilled with a needle in the positions shown. A thin piece of the best clear mica is then laid on a pad of blotting paper, the piece a placed on it, and four corre- sponding holes pricked through with the needle, the piece being after- Cc. d. Fie. 2.—a, b, and ec, full size; ¢@, about twice full size. wards cut to the shape 0. Finally, a and b are fastened together by four little loops of No. 30 platinum wire, and the whole trimmed to shape with the scissors. The trough is then hung by two platinum chains (made of No. 30 wire with long links as shown enlarged at d (fig. 2), so that the acid may not creep up them) from the hooks shown in fig. 1. These hooks are best made of half-round wire, doubled like a linch-pin, sliding easily but firmly in holes on each side of the capillary, as in fig. 1. The operation of putting in a capillary is as follows: The instru- ment is fixed to any convenient support by the screw M. The milled head G is loosened, and the U-tube D drawn down and turned aside. The whole instrument is then tilted backwards to an angle of 45° from the vertical. In this position the trough H hangs clear of the capil- lary. If the capillary has been already filled and connected with the pressure tube, the nuts E and F must be unscrewed far enough for the tube to pass sideways into the clamps, but a new capillary may be easily and safely inserted from below after merely loosening E and F. It must then be filled to within 2 cm.—not nearer—of the top, with recently distilled mercury from a periectly clean pipette, connected with the pressure apparatus,* and some mercury forced through. The screw M is then slightly loosened, and the instrument raised cautiously to a nearly vertical position. The trough H is adjusted by sliding the hooks from which it hangs up or down, or bending them, until the capillary rests against the centre of it—the apparatus being * Full details of the pressure apparatus, the cleaning of the tubes, and the method of drawing capillaries were given in my book. 1902. ] a Theory of the Capillary Llectrometer. 105 tilted back during each alteration. When these adjustments have been made, the inside of the trough is wetted by touching it with a glass rod dipped in dilute sulphuric acid of 25 per cent., and the apparatus is tilted forwards until the wet trough swings against the capillary, and sticks to it. A piece of thin cover-glass—or of mica if very high powers are to be used—slighly wider than the trough, is picked up with a pair of fine forceps, wetted on one side with the acid, and placed carefully against the trough, to which it adheres, holding it firmly against the capillary, the lower edge of the glass resting against the platinum loeps with which the trough is fastened together ieee 2,.C).; | The U-tube is then turned back into position, cautiously raised until the lower edge of the glass just dips into the acid, and clamped by the screw G. Finally the trough is gently shaken or pushed to and fro in the plane of the mica, until the acid rises in it to the required height and all bubbles are expelled. The trough is held together so firmly by surface-tension that it seems at first sight a difficult matter to take off the cover-glass without breaking the capillary. It may, however, be done with the greatest ease as follows: The screw G is loosened, and the U-tube D drawn down and turned aside. Up in, be “ci tL hy tlo ie =e dae MEN SSL 3s 03090905 (11), iR> 7 V noe Ry ry where 2 denotes intelligence and @, «2 any other characters. Clearly when 7;,, and 7;, are both small #;r,, ,, cannot be large. Let L be length of hee, B be breadth of head, and § be stature. Then in the case of the Cambridge graduates vy, = 31839, fis 10 Lolo Tm — Osos On = 3 2886) eine = 101529, rin = 0:0450, Us == 3°6958, TLE = 0°3448, rs = =): 0CaG: * “On the Correlation of Intellectual Ability with the Size and ee of the Head, ” “Roy. Soc. Proc.,’ vol. 69 (1902), pp. 3583—342. i Pdands rams: GAC eal 187, p. 276. t Lbid.,p.279. (ii) is deducible by simple algebra in the method often indi- eater in this series of papers. * 1902.]| Correlation of Mental and Physical Characters. 107 The v’s and the physical correlations are due to Dr. W. R. Mac- donell,* 71, 73 Were given in our first paper,t and 7g was deduced from the following fourfold table :— (A.) Intelligence. Honours. | Pass. | Totals. é| Over GN” se coouso 244 2985 || 472-5 ell = Wincler OY" G50 50 oc 280 258°5 || . 538-5 Cs} —— {— —_ — v2) Notallseasrter eres 524. | 487 | LOL lf 7;3 were really sensible, it would mean that honours men were shightly shorter than pass men. ‘The only safe conclusion we can _ draw, however, is that stature is not correlated with place in degree examinations. . From the above results we find : Ure, = 41435, URgp = 4£°5530. Hence we have We eS O02F RRS 0:0370. That is to say, the correlations of intelligence with the ratios of length and breadth of head to stature are slightly smaller than the correla- tions of intelligence with the absolute head-measurements. ‘The result predicted from the smallness of 7;g in the discussion on the paper here receives its exact numerical confirmation. (3.) Since our school measurements were started, MM. Vaschide, and Pelletier have published in the ‘Comptes Rendus’{ a statement that although unable to find any relation between intelligence and length or breadth of head, they consider a relationship to hold between intel- ligence and the auricular height of head. Their process was of the following kind. They asked the school teacher to select ten intelligent and ten non-intelligent children, and then measured the heads of these two sets, and found their means. This was done for groups of three ages in boys and two ages in girls. The probable errors of the differ- ence of the means of ten observations are not considered, and by exactly the same process that they reason that the auricular height is greater ‘for the more intelligent children they might have deduced from their statistics that intelligent girls of 11 years have lower heads * * Biometrika,’ vol. 1, pp. 188-9. 7 ‘Roy. Soc. Proc.,’ vol. 69, pp. 335-6. t ‘Comptes Rendus,’ Paris, vol. 133, 1961, pp. 551—558. \ 108 Miss Lee, Miss Lewenz, and Prof. Pearson. [| NWove"3, than intelligent girls of 9 years, and non-intelligent boys of 11 years lower heads than the same class of 9 years! Frankly, we consider that the memoir is a good illustration of how little can be safely argued from meagre data and a defective statistical theory. Taking from our school data the auricular height of 2005 boys, and from the growth table based on the same material, reducing them to the age 12 as standard, we find (B.) Auricular Height of Head and Intelligence. | | | | Intelligent. Slow. Totals. Baia ox. | = Above 127 mm...... 481°5 584 ‘0 | 1065°5 "ap Below 127 mm..... 415-0 ee el 939 °5 o = H = a be a Motals#e. eee 896 °5 1108°5 2005 | | Whence the correlation = 0°0161. There is thus less correlation between auricular height and intelli- gence than between either breadth or length and intelligence ; indeed, it is less than the probable error, and no weight can be laid on it what- ever. The discovery of MM. Vaschide and Pelletier that the auricular height of school children is related to their intelligence seems to us quite incorrect for English boys, and unproven owing to defect of materiai and method even for French children. It has been suggested by a sweeping critic, who clings to the high correlation of intelligence and head size, that our school head-measure- ments are of no value. To this we can only reply that in all cases where the measurements have been in the least doubtful the spanner has been returned and the measurements re-made. Further, if the absence of correlation between intelligence and head-measurements be a proot that the head-measurements have been taken badly or the scale of intelligence wrongly applied, how does it happen that high correlation comes out for the head-measurements of brothers, for all three cases, breadth, length, and height, and that its value is quite in keeping with the correlation between the intelligence of brothers? The existence of careless measurement or appreciation would have reduced these correlations also to near zero, as well as those on the characters on the same individual. We are forced to conclude that while our data give surprisingly consistent and uniform results for collateral heredity when we deal with upwards of twenty characters,* about half mental * Results for seven mental and three physical characters were given in ‘Roy. Soc. Proc.,’ vol. 69, p. 155. These numbers have been more than doubled since that paper was published. 1902. ] Correlation of Mental and Physical Characters. 109 and half physical, they give with an equal weight the definite result that there is no marked correlation between intelligence and the size or shape of head in children. (4.) While it seems desirable later to investigate specially the Cam- bridge data from the standpoint of the subject studied, as well as degree taken, we complete at present the list of other physical correlations with intelligence on the simple basis of honour and pass degree groups. The following are the tables :— Intelligence and Strength of Pull. (C.) First Grouping. | Honours. | Pass. | Totals | Sante a ee ny | || a] Above 84 Ibs........] 251 256-5 || 507 °5 @&,| Below 84 lbs........ | 273 229 °5 | BO) 55 ohtlsnaweive. |) 524 486 | 1010 (D.) Second Grouping. | | | Honours, | 2°28: 2nd, 1st sro drd classes, Totals. =a and Pass. =| Above 84 Ibs. ...... 75 432 °5 507 °5 f4| Below 84 lbs. .... (hs) 424.°5 502°5 Totals ........| 153 857 1010 | Intelligence and strength correlation is from the first grouping — 0:0765, and from the second —0:0199. ‘Thus it would appear that from either grouping the honours men have slightly less strength of pull than the pass men, but as even this small amount is decreased when we group the first class men only together, such inferiority as there is seems to lie in the second and third class honours men. Taking the average, we may say that there is a negative correlation of — 0:0482 between intelligence and strength of pull. The probable error of the result, about 0:035, shows that very little weight can be attached to lt. 110 (K.) Intelligence and Strength of Squeeze. Squeeze. Motallsierneer IMG Woes ooo ds Below 88d lbs. ..... eo | | | Honours. Miss Lee, Miss Lewenz, and Prof. Pearson. [Nov. 3, Pass. 227 °5 255 °5 483 Totals. — 464 5388 1002 The correlation between intelligence and strength in this case — 0:°0242. This result, although negative. (F.) Intelligence and Sight. This is judged in the Cambridge Anthropometric Laboratory by the distance at which the test type can be read. it is less than its probable error, is again Right eye. sai "Over Gl!= aac a Winder Gil? eee Ro tallisaereaceee rd iss) DM wn Totals. Forty-one men on our cards were unclassed—10 in Ist class, 5 in second, 1 in third, and 25 poll-men. This was possibly due to defective sight, or even to the loss of the right eye, because the strength of the left eye was sometimes given; we have not ventured to group these unclassed cases, however, with the short-sighted division. The correlation between intelligence and long sight - 0:0049. This is far less than the probable error of the result, but is again negative. (G.) Intelligence and Weight. : | Horours. Pass. Totals. | Over 10 st.131bs...| 258°5 226 484-5 ©) Under 10 st. 13 ibs. 263 °5 261 526 °d E | oe | = otal Wo creccta 524. 487 1011 The correlation between intelligence and weight = 0:0459, and is thus very slightly larger than its probable error. 1902.] Correlation of Mental and Physical Characters. PV Now, it has sometimes been argued that in any investigation of this kind, it is desirable to take not absolute weight, but its ratio to stature or some power of stature. Let W = weight,S = stature, and nw = any power ; let R, = W/S*”, and v be a coefficient of variation, and r one of correlation, 2 standing for intelligence. Then 2 wy 2.2 Pose 3 UR, = Ow + 7 Ug ay 2nvwvsi SW secccescncnsncsceievs (i), . OW iw — NUSTE5 00 = EN ae Alaiye cca con te AROSE (11). OR, But Je a 36958. rsw = 0°4860, vw = 10°8300, Tw = 0:0459, Ce 0:0058, from results already given for the Cambridge data. Hence, calculating vp, from (i) for n = 1, 2, and 3, we deduce rg, = correlation of intelligence with ratio weight to stature = 0:0540, rR, = es K . (stature)? = 0:0555, sy = - As u (stature)? = 0:0503. There is no substantial difference between any of these correlations and that for intelligence and absolute weight. As they were found indirectly by formule, it seemed desirable to test at least one of them directly. Accordingly Miss M. Beeton found the ratios of weight per inch of stature for 1012 Cambridge men. ‘The resulting table was as follows :— (H.) Intelligence and Weight per inch of Stature. Honours. Pass. | Totals. | | fps ta eRe 1a Over 2°224 lbs. perin. ..... 258 '5 222 480 °5 Under 2°224 Ibs. per inch .. 265 °5 266 | 831°5 Titre aaleee ae | 524 488 | i012 The distribution is sensibly the same as that of the table for abso- lute weights, and the correlation comes out 0-0604, #.c., it differs only by 0:0064, or about one-fifth of the probable error from the. value of the correlation obtained indirectly. _ We may then, I think, conclude that whether we take ‘absolute weights or the ratio of weight to stature, honours. men are slightly heavier than poll-men. Summing up the whole of our examination thus far of the Cambridge measurements we may say that : 112 Miss Lee, Miss Lewenz and Prof. Pearson. __[Nov. 3, The honours men, and presumably therefore the more intelligent class, are slightly heavier and have slightly longer and broader heads ; they are not quite as tall nor as strong, whether strength be measured by pull or squeeze, and are slightly shorter-sighted than the poll-men, or presumably the less antelligent class. In no single case, however, is the correlation between entelligence and the physical characters sufficiently large to enable us to group the honours men as a differentiated physical class, or to predict with even u moderate degree of probability intellectual capacity from the physical characters of the individual. (5.) While the above and the ty published results exhaust the Cambridge data, as long as we preserve the division into honours and poll-men, much more remains to be done on this material when we consider subject. groupings among the Cambridge graduates, or when we turn to the much wider range of both physical and mental characters recorded in our school measurements. A preliminary inquiry may, however, be recorded here as bearing upon a rather vexed question at the present day, namely, the relation of athletics to health and intelligence. In our school measurements we had three categories: Health-—divided into the classes: Very Strong,* Strong, Normally Healthy, Rather Delicate, Very Delicate. Ability or Intelligence—was divided into six classes: Quick Inteliigent, Intelli- gent, Slow Intelligent, Slow, Slow Dull, Very Dull. Lastly, we had the alternative category—Athletic, Non-athletic. By Athletic we understand not only fondness for out-door exercises and games, but good performance in them. There was a control entry in the schedules under the heading Games or Pastimes, in which not only what the children /ked, but in addition what they were good at, had to be entered. We were thus in a position to make that triple correla- tion between health, ability, and athletic power, which seems really needful, if a sane judgment is to be made on the part athletics should play in the school curriculum. The following tables give the relations between health and ability, ability and athletic power, and health and athletic power :— (1.) Health and Intelligence. 2253 Boys. | | | Quick intelligent, | Slow intelligent, slow, Total | | intelligent. slow dull, very dull. ase | Very strong, strong.. | 415 | 453 | 868 =| Normally healthy .. 461 | 542 1003. S| Rather delicate, very | eas Fall eucdlelicabes a seer 128 °5 253°5 | 382 Marbals), fete ate 1004 °5 | 1248°5 | 2288 * Strong in these categories equals robust. 1902. | Correlation of Mental and Physical Characters. 113 The correlation dividing at the Sfrong is 00820. The mean of the other divisions (i) dividing at the Delicate, and (ii) putting the Slow Intelligent with the Intelligent, gave 0°0835. We conclude, therefore, that there is a sensible, but not marked correla- tion between good health and intelligence. Taking, however, health and erhleties we have the table :— (J.) Health and Athletics. 1743 Boys. Rather | Very | Normally Wey 4 | strong. SUSOIE | healthy. | delicate. delicate. heels. | | Athletic ...| 91 | 447°5 AES NAD es a aes 1159 | Non-athletic 9°5 98 °d ae. Deals L665 16 584 | Totals..... 100°5 791 19 1743 | | 546 286 *5 | The correlation between healthy and athletic dividing between Strong and Normally healthy is = 0°4570, a very marked relationship. Next, taking intelligence and athletics, we find :— Intelligence and Athletics. 1708 Boys. | Quick | Slow _ | Slow | Very | ca intelligent. sbugeyone ‘intelligent. Slory | dull. | dull. | Howells. | Athletic ...! 159°5 AM Toge) sshd | LaSe7o 40750) 12 i485 | Non-athletic| 46 163-25 | 187°5 | 99°75| 48-5] 15 || 560 | Totals ..... 2055 | 58 543 | 258:5/89 | 27 || 1708 Dividing between intelligent and slow intelligent we find the corre- lation between intelligence and athletic character is 0°2133. This result may be exhibited also in the percentages of athletic and non-athletic boys who fall under each class of intelligence :— Percentages of Athletic and Non-athletic Boys under each grade oy Intelligence. Quick Slow | Very intelligent. Intelligent. intelligent. Slow. | Slow dell) dull. Athletic... 14 37 ol 13 4 1 Non-athletic 8 29 32 18 9 3 VOL. LXXI. K 114 Correlation of Mental and Physical Characters. [Nov. 3, The relationship between keenness for combined with capacity in games and general intelligence is here manifest. Certain other correlations with the athletic character may be just noticed without giving the tables. The athletic boy is popular (0°3250) and noisy (0°3452), and this although popularity is not found to be directly correlated with noise. He is slightly self-conscious (00761), and is more likely to be fair than dark (0-0391). His temper tends to be quick rather than sullen (0°2207), as the following table, based on 1664 cases, will show :— Percentages of Athletic to Non-athletic Boys for each Temper. | Quick tempered. | Good-natured. | Sullen. | a | laaphletie iceaces Sa 21 | 68 | uW | | J | Non-athletic......' 12 | 74 | 14 —- To sum up, then: While the intelligent are only slightly the more healthy, the athletic are notably the more healthy element in the com- munity. Further, the athletic are considerably more intelligent than the non-athletic ; they are the more popular and more noisy element ; and they tend to quick rather than sullen temper. We may in general terms describe the athletic boy as healthy, quick-tempered, and in- telligent when compared with the non-athletic boy. He certainly under all three headings should make a better soldier than the non- athletic, and it is hard to discover any statistical evidence in school life for such expressions as “‘ the flannelled fool at the wicket,” or “the muddy oaf at the goal.” What happens in later life can only be determined when ample statistics are available for reduction and com- parison. Failing such data, we can argue only from the vaguest of impressions. | 1902.] Descending Spinal Tracts in the Mammaivan Cord. 115 “ Note upon Descending Intrinsic Spinal Tracts in the Mammalian Cord.” By C.S. SHerrineron, M.A., M.D., F.BS., and E. E. LAsLett, M.D. Vict. Received November 5,—head Novem- pene, L902: In the course of experiments upon the paths of nervous conduction in the spinal cord of the mammal, one of us observed* very numerous and wide departures from the fourth so-called “law” of Pfluger. That “law” states that “ Reflex-irradiation in dem Riickenmarke nach Oben, resp. Vorn, gerichtet ist; also gegen die Medulla oblon- gata.”t The observation of the above-mentioned exceptions rendered desirable a search for more detailed evidence of intrinsic spinal paths running in the aboral direction. We therefore set about inquiring into the existence of spinal paths connecting the activity of segments situate nearer the head with segments lying further from the head. Such evidence is obtainable with some experimental difficulty, but it has been eventually forthcoming, and amounts to demonstration of the microscopic course of the channels involved. It is some main features of these latter that we desire to record in the present communication. METHOD. The method employed has been that of the Wallerian nerve-fibre degeneration, but with a novel feature in the mode of application of the method. For the purpose in view the ordinary establishment of a cross-lesion in the spinal cord is futile. The secondary degeneration then produced befalls, in the spinal region under investigation, all nerve-fibres having their perikarya headward of the cross-lesion, whether those perikarya lie in the cerebral hemisphere, basal ganglia, mid-brain, cerebellum, bulb, or cord itself. It is obviously then impossible to identify which particular ones, if any, of the degenerate nerve-fibres are coming from the cord-segments whose nerve-tracts are the special object of inquiry. To obviate this difficulty, we have adopted a method which may be termed a method of “ successive degeneration.” ‘The method consists in producing two or more succes- sive degenerations with allowance of a considerable interval of time between them. In the piece of cord to be examined, a first degenera- tion is allowed time enough to remove all the tracts descending from sources other than those the immediate object of inquiry. This is a procedure which requires in our material, at shortest, 9 or 10 months to complete. When the time is complete, the cord is left, as it were, * “ Crooniar Lecture,’ ‘ Phil. Trans.,’ B, 1897. + ‘Die sensorischen Functionen des Ruckenmarks der Wirbelthiere, nebst einer neuen Lehre tiber die Leitungsgesetze der Reflexionen’ (p. 73), Berlin, 1853. Kee? 116 Prof. C. 8. Sherrington and Dr. E. E. Laslett. [Nov. 5, like a cleaned slate, on which once more a new degeneration can be written without fear of confusion with a previous one. ‘The cord is then ready for receiving the lesion which shall cause degeneration ot the particular tracts whose existence is suspected. After a period suitable for the full development of the new degeneration, the cord is treated histologically by the Marchi method, and the microscopical examination proceeded to. ‘This method resembles in principle a method employed with noteworthy results by Miinzer.* This author performed on the new-born rabbit a first lesion (¢.g., removal of one cerebral hemisphere, injury of mid-brain or cord), and later, in the animal when grown, proceeded to establish a new lesion which was thus uncomplicated by the part already separated: thus ‘“‘ Gudden’s agenesic atrophy” was made to precede the degeneration desired for study. One of the experiments made by Miinzer and Wiener (1895) deals with the problem undertaken by ourselves. After semi-section of the spinal cord of the new-born rabbit at the last dorsal segment, they per- formed total transection two segments further back when the animal was grown. Behind the second section they found “as many fibres degenerate on the semi-sected side as on the intact side.” If not ‘“ decus- sation fibres” these fibres must evidently be of intra-spinal origin in the anterior lumbar region. As to their being decussation-fibres, Miinzer says they are, on the contrary, from the grey matter of the same side as the semi-section, a statement which our own results in the dog in the same and other regions endorse. In our experiments the cord of the dog has been used, and total transection has not been the final, but the first step in the pro- cedure. ‘This course was chosen in order to completely exclude all chance that fibres from sources not the object of inquiry could compli- cate the second lesion. In order to ensure complete transection, we have in almost all our experiments exsected and ablated a short segment of the cord, instead of simply severing it across. The exsection was made immediately in front of those spinal segments whose system of descending fibres in the cord was to be looked for. Then after an interval, which we found by experience must not fall short of 260 days, the second lesion, usually some form of partial section, was performed, and a further period of about 20 days was allowed for degeneration. The procedure of total transection prior to the lower limiting lesion has an additional advantage in the lesser interference with the local circulation of the cord in the final lesion. Better in these respects this plan offers, however, considerably greater difficulties than its converse. We have, in spite of all care, lost a number of experiments in the long intervals necessary to elapse while * EK. Miinzer, with Wiener, ‘Prager Medicin. Wochenschrift,’ 1895; also “Monatschr. f. Psychiat. u. Neurol.,’ vol. 12, p. 241, 1902. 1902.] Descending Spinal Tracts in the Mammalian Cord. jase the cord is ripening for the second operation. We have, however, obtained thirteen successful complete experiments : in these the shortest interval has been 260 days, the longest 568 days. Among these is included two in which, for special reasons, the partial cross-lesion of the cord was made precedent to the total transection. ” RESULTS. The spinal segments examined as sources of aborally-running fibre- systems have been posterior cervical, anterior thoracic, mid thoracic, posterior thoracic, and anterior lumbar. From all these regions our experiments demonstrate that copious aborally-running fibre-systems spring. Thus, the accompanying fig. 1 shows, for instance, tracts of fibres in the 5th lumbar segment which have their origin in cells of the 2nd thoracic segment. Gael LL. R. Cross-section of the cord of the dog at a level in the anterior part of the 5th lhimbar segment; Marchi preparation. The section reveals the topography at that level of the aborally-running fibre-system of the Ist and 2nd thoracic segments. The 8th cervical segment of the cord had been completely exsected and ablated. A partial translesion (rather more than a semi-section) was made the left half of the 3rd thoracic 568 days subsequent to the removal of the 8th segment. The exact extent of this second lesion was determined subsequently by microscopic examination in serial preparations, and its limits will be described and figured in a fuller communication. The dots indicate, in a way mentioned in the text (p. 120), the density and extent of the degenerate tracts of fibres. L = left side; R = right side. Speaking generally, of the fibres composing the aborally-running systems springing from the grey matter of the spinal segments 118 Prof. C. 8. Sherrington and Dr. E. E. Laslett. [Nov. 5, examined, we find there may be distinguished two sets. For physio- logical description it is in some ways convenient to regard the length of the spinal cord as divisible into regions ; thus, a brachial for the fore limb, a thoracic for the trunk, a crural for the hind limb, a pelvic for pelvic organs, a caudal for the tail, and so on. A reflex initiated vidi an afferent path of one such spinal region may evoke its peripheral effect by efferent paths of a spinal region other than that to which the original entrant path belongs. Such a reflex has in a former paper by one of us* been termed a “long” spinal reflex, in contradistinction to reflexes whose centripetal and centrifugal paths both belong to one and the same spinal region. The latter reflex it was proposed to term “short.”t Analogously, in the aborally-running fibre-systems of the spinal segments examined, by our experiments fibres of two cate- gories are found, one a set passing beyond the limits of the spinal region in which they arise, the other not passing beyond those limits. The former we would term “long spinal,” the latter “short spinal” fibres. In each of these main categories there can be distinguished fibres of various intermediate length. Again, the fibres of each of the above two categories may be clas- sified into two sets or tracts, according to their topography relatively to the cross-section of the cord. Fibres of both of the above categories are situate both in the lateral columns and in the ventral columns of the cord. It is useful, at least for descriptive purposes, to indicate this by terminology. We thus recognise in the aborally-running intrinsic spinal fibre systems the following sets or tracts: (a) Ventr al short fibres, (8) ventral long jibres, (y) lateral short fibres, (8) lateral long fibres. Tt must be added that the distinction into lateral and ventral is somewhat artificial, as there exists often, especially in the case of the “ short” fibres, no distinct gap between the ventral and lateral fields of distribution of the fibres in the transverse area of the cord. In regard to the “long” fibres, we find that in all the regions examined by our experiments there is no evidence of decussation of these tracts. This statement does not exclude the possibility that the collaterals or the fine ultimate terminals of these fibres may in some cases penetrate in the grey matter across to synapses in the crossed side of the grey matter. We have at present no reliable microscopical evidence for or against such a possibility. But all our evidence is consentient that the fibres themselves do not pass from the white columns of one side of the cord into those of the crossed side, that is, do not in the ordinary sense decussate. A similar statement seems also to hold true for the ‘‘short” fibres, thus confirming Miinzer:{ it is certainly true of the majority of the C. S. Sherrington, “‘ Croonian Lecture,” ‘Phil. Trans.,’ 1897. Ibid. ‘Prager Medic. Wochenschr.’ 1895. t+ —-k & 1902.] Descending Spinal Tracts in the Mammalian Cord. 119 short fibres ; but analysis of our material makes us hesitate to positively affirm that it is true for all of them. A small proportion of the short fibres may decussate, at least in the sense that short fibres arising in perikarya belonging to one lateral half of the cord may find their way into the white ventrolateral columns of the crossed half. We do not affirm, however, that any of even our “short” fibres do decussate, we simply affirm our present inability to deny that a small proportion of them may do so. Some of the “long” fibres are very long, both in the lateral and in the ventral columns of the cord. Thus, some of those arising from perikarya in the 6th and 7th cervical segments we have traced into the sacral region, 7.¢., through nearly thirty spinal segments, both in the lateral and in the ventral columns. The rule pointed out by one of us in a previous paper,* that the long fibres in the spinal cord tend to le nearest the surface of the cord, is well exemplified in these intrinsic spinal systems. Besides fibres in the ventrolateral columns the aborally-running fibre-systems of spinal origin include fibres in the dorsal columns. INTEh Ay Jip R. Cross-section of the spinal cord of the dog at the level of the 1st sacral segment ; Marchi preparation. The section reveals the topography at that level of the aborally-running fibre-system of the nerve-cells of the Ist and 2nd lumbar segments. A short length of the 13th thoracic segment of the cord had been completely exsected and ablated. A partial translesion (rather more than a semi-section) was then made through the right side of the 2nd lumbar segment in its anterior levels 290 days subsequent to the total exsection. The extent of this second lesion was accurately determined later by microscopic examination in serial preparations ; its exact limits will be described in a fuller communica- tion. The dots indicate, in the way mentioned in the text, the density and extent of the tracts of degenerate fibres. L = left side; R = right side. * ‘Journ. of Physiology,’ vol. 14, p.298. Cambridge and London, 1893. 120 Descending Spinal Tracts in the Mammalian Cord. [Nov. 5, These are less numerous. We defer detailed description of them, together with further detailed description of the ventral and lateral tracts above-mentioned, until a fuller communication dealing with the whole subject. The general features of the topography can be gathered better from the two accompanying figures than from any even lengthy JHE, 33, Cross-section of the spinal cord of the dog at the level of the anterior part of the 4th lumbar segment: Marchi preparation. The section reveals in left half of the cord the topography at that level of the aborally-running fibre-systems arising in perikarya of the grey matter of the left half of the 6th and 7th cervical segments. Total transection of the cord had been performed through the 8th cervical segment 268 days subsequent to left semi-section at the 5th cervical segment. Thirteen days only was allowed for the development of the degeneration after the second lesion. ‘The degeneration may not therefore appear so extensive as it might have done later, but its localisation is probably the more precise. The exact extent of the semi-section was found by subse- guent microscopic examination in serial preparations to amount almost accu- rately to a full section of the left half of the cord ; the detailed limits will be described in a fuller communication. The degeneration in the right half of the lumbar cord figured includes aborally-running fibres derived not merely from the spinal vervical grey matter, but from bulbar and cerebral cources as well; and these are practically inextricably commingled one with another. The dots signify, in the way mentioned in the text below, the density and extent of the degenerations. L = left side; R = right side. textual description. All the figures have been drawn with the camera lucida upon squared proportional paper, and the squares on the paper have been made to correspond with squares in an engraved eye-piece. All the drawings are to exactly the same scale. ‘The dots signity degenerate nerve-fibres, but the number of dots does not of course represent the absolute number of degenerate fibres, but falls far short 1902.) The Lnter-relationship of Variola and Vacernia. 121 of it. Extreme care has been taken, however, to make the number of the dots bear fairly accurately a general proportion to the density of the degeneration, and the same proportion in one drawing as in another. “The Inter-relationship of Variola and Vaccinia.” By 8. MoncK- TON CopreMAN, M.A., M.D. Cantab., F.R.C.P. Communicated by Lorp Lisrer, F.R.S. Received November 13,—Read November 27, 1902. The term “ variole vaccine ” employed by Jenner, as a synonym for cow-pox, has been generally accepted as affording evidence that in so naming this disease, “small-pox of the cow,” he was desirous of placing on record his belief that cow-pox, or vaccinia, was intimately related to human small-pox, if indeed it were not directly derived from it. This theory, however, appears to have found but scanty favour in Jenner’s day, and even at the present time the value of the practice of vaccination is, by some, impugned on the plea that inoculation of one disease—cow-pox—could not be expected to exert any really pro- tective influence against the ravages of small-pox—a disease considered by them of totally different origin. In the hope of obtaining definite information on the subject, many observers, during the long period which has elapsed since the intro- duction of vaccination, have set themselves the task of attempting to solve, by experimental methods, the problem of the true relationship of vaccinia to variola. These attempts have been, for the most part, directed to the possibility of giving rise to cow-pox by the introduction, in one or another manner, of the virus of small-pox into the system of the bovine animal. In the great majority of such attempts, which have been much more numerous than is generally supposed, the results have been entirely negative, although so numerous have been the experimenters, who aon time to time have attacked the problem, that the total number of instances in which an apparently successful result has been obtained, is now considerable. So far as I am aware, the first recorded experiments are those of Gassner of Gunsberg, who, in 1801, succeeded, after no less than ten fruitless attempts, in directly variolating a cow with small-pox virus. The lymph thus obtained was employed for the vaccination of four children, from whom other seventeen were subsequently vaccinated. None of these exhibited any signs of small-pox. 122 Dr. 8. Monckton Copeman. | Novea3; It is impossible, here, to do more than mention the names of other investigators who have engaged in research of this nature at various times from the commencement of the last century up to the present time. I therefore merely append in the foot-note* a list of such names, placed, as far as possible, in chronological order. It is a note- worthy fact that every observer mentioned, with the exception of Chauveau and his colleagues of the Lyons Commission, and Martin, claim to have obtained, on one or more occasions, positive results as regards the production of typical vaccinia, generally after one or more removes from the animal originally variolated. But it must, I think, be admitted that many of the earlier experi- ments, more particularly, are practically worthless owing to the con- ditions under which they were carried out. Some of the main objections are based on the frequently concomitant use of vaccine and of variolous lymph on the same animal, and the want of care as to the cleanliness and freedom from vaccine contamination of lancets and “points” used in the experiments. As regards my own work, carried out on similar lines to those adopted by previous observers, and of which a full account was published in the ‘Journal of Pathology and Bacteriology,’ in 1894, it may here be mentioned that I obtained an undoubtedly successful result in one series only, out of four attempts. In four subsequent variolation experiments, carried out several years later (1901) in con- nection with work, a detailed account of which is set out in the * Chronological list of observers who have carried out variolation experiments — on bovines :— HSOLy Valborg eve. ee ee ol enee eC Open amen S282" MicMichaell tence cl acini aye 1830: sSonderland: 32.7.9... 0) armiens is Numan ier eract ais ie iureciats 1832. Macphail.) i022 eee 0-2) Baltimore: 836.) Whiclom eee ei CUS alee » Martin’ 32.......0......... Attleborough ass: =) MacPherson ee rrcrsrer aac) unela: 183920 Reiter tae sues 2 ose meme be sien MECC] y cte nee amie acco ice ea eae ; 18402 eBadcock. Osseraem evelnsie cele aere 3 V863-CGone Chamveau seein ers) cue) -uiele Lyons. ISG68._"Shorttice peewee ce se. «hee ep una IS7ll Chauveau sates. oc. oe eenOnsE USSE. Violet ey eee erect. e-)ioneten etiam mnie 1886-90.) Miseher Visi aie 2 .ic\- se ee Carlsnulies TSS9. PKCM o jcce velo jefue sails “sake sue Lanai 1890-91. Eternod and Haccius ....... Geneva. IGCEE “Simos oS cabaduooodadaos laalen. Ee 18 hort eee ener: OMA Si incl onoyellewaye tH MGMT re Ay earctel a dee enlorenercte Ree 5 A Oly oo gsuandeadG60 00 dc bs 1902.| The Inter-relationship of Variola and Vacevnia. 123 present paper, my attempts at direct transference of human small-pox material to the calf met with no success. All my earlier experiments were conducted at the Brown Institution, in order to avoid any possibility of contamination with vaccinia. As a further precaution new scalpels were used, which were invariably first carefully sterilised in the flame of a spirit lamp, and, after use, the table was, on each occasion, thoroughly washed with carbolic acid (1 in 20), while during the intervals of use it was kept exposed to the air under an open shed. Similar precautionary measures have been observed throughout the course of my later work. The difficulty experienced by myself and the numerous other investi gators, to whom reference has already been made, in attempts to transmit human small-pox directly to bovines, whether cows or calves, is not infrequently cited as a reason for regarding with distrust the theory expounded by Jenner, that cow-pox, whether carried through the horse as intermediary host or not, was originally derived from small-pox in the human being. But a great deal, at any rate, of the small-pox which was preva- lent at the time that Jenner lived and wrote was of that compara- tively mild variety which, under the name of inoculated small-pox, was intentionally produced in healthy subjects, with the object of thereby conferring protection against subsequent attack by the disease in virulent form. So mild indeed at times were the results of inoculations in the hands of such operators as Adams and the brothers Sutton that, as we learn from contemporary records, in many instances but little obvious effect was observed, with the exception of the local vesicle arising at the site of insertion of the small-pox virus, and the patients suffered but little inconvenience. ‘Thus, more particularly in certain of Adams’ cases, as may be gathered from his own account of the circumstances, the visible effect produced so closely resembled the results then beginning to be known as following on the Jennerian process of vaccination, that numbers of his patients were with difficulty persuaded that he had not, contrary to their desire, intentionally vaccinated rather than variolated them. ‘The gradual evolution of a strain of lymph of such tenuity, according to Adams himself, was obtained by attention to the mode of life and general treatment of persons undergoing the process, together with careful selection of the sources (preferably the primary vesicle) from which the virus was obtained. The majority of persons thus inoculated are not likely to have been incapacitated, as the result of the operation, to a much greater extent than are those who undergo efficient vaccination at the present day, and doubtless, therefore, they would be, for the most part, capable of following their ordinary avocations during the progress of the induced 124 Dr. 8. Monckton Copeman. [Nov. 13, disorder. On the other hand, this would hardly have been possible in the case of persons contracting small-pox in the ordinary way, among whom the disease was apt to exhibit such virulence as to account for the death of perhaps 50 per cent. of those attacked. Not only were the effects following on inoculation comparatively mild, but the disease in this form was intentionally brought into many country districts which otherwise might not have become invaded by small-pox. In the light of these facts, it has for some time past been borne in upon my mind more and more convincingly that it was prob- ably from the znoculated form of small-pox, rather than from the ordinary variety of the malady, that much, at any rate, of the cow- pox, in the pre-vaccination era, was derived. It is not difficult to understand how that the cracks so often found on the udders of cows might become infected by a milker with fingers contaminated by contact with the inoculation sore upon his arm. I determined therefore, if possible, to put the matter to the test, and, learning that in Nubia, in Burmah, and in certain parts of India the inoculation of small-pox is still practised, I made numerous endeavours to obtain the necessary material, but unfortunately without success. In default, therefore, of inoculated small-pox in the human subject, I made trial of the monkey, which, as I have shown in a previous communication to the Royal Society, is readily susceptible to the disease, the various phases of which in this animal closely resemble those observed in man, but in a much milder form ; the occurrence of a generalised eruption being exceptional. The different series of experiments, protocols of which I append, have been carried out at intervals, determined mainly by the possi- bility of procuring the necessary small-pox material. The work was commenced in April, 1898, with a supply of small-pox lymph received from the Medical Officer of Health for Middlesborough, in which town an epidemic of the disease was then in progress. For subsequent supplies I am indebted to the Medical Officer of Health and the Medical Superintendént of the Small-pox Hospital at Glasgow, to the Medical Superintendent of the West Ham Small-pox Hospital at Dagenham, near London, and to the Medical Superintendent of the Hospital Ships of the Metropolitan Asylums Board. The methods employed in the investigation have been briefly as follows :— Collection of Material for Inoculation. In the first instance this was obtained in a manner similar to that formerly employed in obtaining human vaccine lymph. Discrete vesicles, mature, but still containing clear lymph, on one or another por- tion of the body of a patient suffering from small-pox, were punctured 1902.] The Inter-relationship of Variola and Vaccinia. 125 with a sterilised lancet, and their fluid contents received into fine capil- lary tubes, which were subsequently sealed in the flame ofa spirit lamp to admit of transport. This operation, however, 1s a most laborious one, and was subsequently abandoned, at my suggestion, in favour of collection, in the post-mortem room, of vesicle pulp at a suitable stage of the eruption, by means of a small Volkmann’s spoon, after the fashion now invariably used in the Government lymph laboratories in obtaining, from the calf, material for the production of glycerinated lymph. After removal from the body the small-pox pulp is first carefully weighed, and then ground up in a small glass mortar, with the gradual addition of usually four times its weight of a sterilised 50 per cent. solution of pure glycerine in normal saline solution. After thorough emulsification, what is not required for immediate use is stored in tubes, resembling smail test-tubes, which are then corked, sealed with liquefied paraffin to which carbolic acid has been added, and set aside in a chamber kept at a temperature a few degrees above freezing point. Both storage-tubes and corks are sterilised before use. Bacteriological examination by the method of plate-culture often shows a comparatively small number of extraneous micro-organisms in a specimen of small-pox emulsion prepared in the manner described, but whenever possible it has been stored at a temperature of about 15° C. for some weeks prior to using it for inoculation. Snecies and Age of Monkeys Inoculated. For my original experiments on the transference of human small- pox to the monkey, a brief account of which was presented to the Royal Society in 1893, I employed the rhcesus monkey, for the reason that Professor Sherrington and myself had, at the time, a stock of these animals, which had been obtained for other experimental work. Having at that time obtained successful results in every one of my inocula- tions, ! employed the same species of monkey in the greater number of the experiments comprised in the present research. As, however, during the progress of the work I learnt that Dr. Eilerts de Haan, who, - in Batavia, had been working on similar lines to myself, had made use most successfully of the macaque monkey, I also obtained a few speci- mens of this species, in order to compare the results of variolation in these animals with those that I had previously observed in the rhesus monkey. but after two or three inoculations of the macaque with small-pox material, I came to the conclusion that the results fol- lowing on the operation were not ordinarily as typical as in those experiments in which rhesus monkeys had been employed. At the same time the macaque is in this country more expensive and more difficult to obtain than the rheesus, so that I reverted to the use of the latter species in subsequent work. 126 Dr. 8. Monckton Copeman. [ Nov. 13, It would appear also that, as in the human subject, young animals are more susceptible to small-pox than are adults, since it was in those instances in which monkeys probably not more than a year old were variolated that the most successful results were obtained. In one instance, however, in which the monkey was believed to be not more than a few months old, the extremely fine downy hair, after shaving, grew again so rapidly as to render somewhat difficult the photographing of the effect produced by the operation. Mode of Operation and Collection. In the earlier experiments inoculation of the monkey with human small-pox emulsion was carried out by rubbing it well into scarified patches or linear incisions of the skin of the upper arm or of the inside of the thigh, after previous shaving and cleansing of the skin. Subse- quently, however, in accordance with the suggestion of Dr. de Haan, a shaved area on the back of the animal was utilised for inoculation. In this situation the results of the operation were found to be equally good, and there is less liability of damage to the vesicles from the monkey scratching itself. The eruption having arrived at maturity, after the lapse of a period extending from five to eight days from inoculation, the altered epi- thelium was removed either with a smail Volkmann’s spoon or by scraping with a scalpel, after cleansing the inoculation area, between the blades of pressure forceps. The resulting epithelial pulp was then rubbed up in a small glass mortar, with the gradual addition of about six times its weight of normal saline solution, containing, when it was desired to preserve and purify the emulsion, 50 per cent. of glycerine. Experience has shown that in monkeys a year or more old, which have been inoculated, the vesicular stage of the eruption is at its height, as was formerly observed in the human subject, by the eighth day ; but in younger animals the process tends to be hastened, and in some of the later and most successful cases, the eruption was completely vesicular as early as the sixth day (120 hours). The particular breed of monkey does not appear to exert any influence in this respect. Transference to the Calf and Human Subject. The methods employed for transference of the localised disease in the monkey, after one or more passages through that animal, to the skin of the calf need not be set out in detail, being similar to those ordinarily used in the process of calf vaccination But it may here be stated that at no stage of the investigation have these experimental calves been brought into contact with, or even placed in the same room as, the calves used in the current work of the Government Vaccine Establish- 1902.) The Inter-relationship of Variola and Vaceinia. 127 ment. They were fed and otherwise attended to by a man specially detailed for the purpose. All instruments employed for the vaccina- tion of monkeys, calves, or children were previously sterilised by boiling or passing through the flame of a spirit lamp. Attention may perhaps be called to the fact that the skin of the scrotum in the calf affords a specially favourable site for inoculation experiments, especially if, when the incisions are made, the skin is made tense by pressing down the testicles. The first transference from the monkey to the calf does not usually afford a perfect result. Indeed a second, third, or even later passage from calf to calf was usually required before the most typical vesiculation was obtained. In certain cases children were vaccinated with lymph obtained from the experimental calves, and in all instances the resulting vaccination ran a perfectly normal course. With lymph of similar origin I also successfully vaccinated myself. But none of the strains of vaccine lymph, derived originally from human small-vox in the manner described, have been brought into general use. PROTOCOLS OF EXPERIMENTS. Frest SERIES. February 21, 1898.—Glycerinated samples of small-pox lymph received this day from Medical Officer of Health of Middlesbrough. Patients living and aged respectively 20, 27, and 34 years; all had been vaccinated in infancy. April 1.—At Brown Institution, inoculated small rhesus monkey witb small- pox emulsion of 21.11.98 in five linear incisions on left arm, and in fourteen on abdomen, after previous shaving and cleansing of the skin, by means of soap and water, followed by warm boric acid lotion. Monkey isolated in separate room and attendant vaccinated as a precautionary measure. April 5.—All insertions on both arm and abdomen evidently “ taking.” April 8.—Distinct vesiculation at site of all incisions on arm and most of those on abdomen. Monkey etherised and substance of vesicles removed with sharp spoon into small previously weighed and sterilised test-tube. Scrapiags weighed (0°6 gramme) and ground up with six times the weight of 50 per cent. watery solution of glycerine. Resulting emulsion taken up into twelve capillary glass tubes. April 18.—Monkey looks well. All incisions healed up. No sign of generalised eruption. Calf Experiments. April 9.—At the Animal Vaccine Establishment, Mr. Stott inoculated Calf No.1 (No. 4363) on two scarified patches, in twelve incisions on scrotum, and forty-four in perineum and on abdomen, with contents of two capillary tubes of glycerinated pulp prepared from vesicles of monkey. Incisions made with scalpel previously sterilised. April 12 (72 hours).—Practically nothing to be seen. April 14 (120 hours).—All insertions on scrotum appear to have “ taken,” and, in addition, four (not quite so well) on abdomen. Insertions on perineum seem to have failed. Large bullous-looking vesicle on upper scarified patch. 128 Dr. S. Monckton Copeman. [ Novos; From this and from the vesicular lines on scrotum and abdomen collected pap by scraping, after clamping with compression forceps. The same day (April 14), Calf No. 2 (No. 4869 in A.V.E. records) was inoculated on perineum, scrotum, and abdomen with material obtained from Calf Nod: April 19.—All inoculated lines, with exception of two on abdomen, “ taken”’ well, eruption being markedly vesicular. Vesicles clamped and scraped ; pulp being immediately employed for inoculation of Calf No. 3 (No. 4673 in A.V.E. records) in a number of long incisions on the perineum, scrotum, and abdomen. April 24.—All lmes of incision “taken” well; eruption perfectly typical of vaccinia. From this calf, six children vaccinated at A.V.E. same day. May 1.—Children returned for inspection, in ordinary course. All vaccinations completely successful. Photographed arms of two of these children, which pre- sented most perfect eruption. About a month later I hunted up the parents of all six children, when I learnt from the mothers’ statements that in every case the vaccination had pursued a perfectly normal course. SECOND SERIES. March 3, 1900.—At the West Ham Borough Hospital, Dagenham, I removed small-pox vesicles from body of a man, et. 56, who had died 24 hours previously from semi-confluent form of disease. Material removed in test-tube and placed in ice-chest. : March 5.—After removal of some shreds of epithelium for histological pur- poses, the remainder (0°25 gramme) ground up with twice its weight of 50 per cent. solution of glycerine, and the greater portion stored in amber-coloured capillary tubes. The test-tube was afterwards swabbed out, and cover-glass speci- mens made for microscopical examination. With some of the glycerinated emulsion inoculated, at the Brown Institution, rhcesus monkey (young female) on shaved area of back, about 3 inches by 2 inches, previously well washed with warm boric acid solution. Incisions twelve in number made “en échelon.”’ March 8.—Inoculation has evidently “taken,” as tips of each incision are distinctly raised, and whole prospect is that of a typical calf vaccination of about same age (72 hours). March 12.—¥Eruption beautifully perfect; edges of vesicular portion a little irregular, and centre of each line of incision occupied by commencing “ crust.” No general eruption visible. Two photographs taken. Removed lower half of vesicular area with sharp spoon (upper portion left in order to watch further development), and glycerinated resulting pulp. Material used for inoculating Monkey No. 2 and also a calf (No. 606). March 12.—Monkey No. 2 inoculated in fourteen incisions with glycerinated pulp from Monkey No. 1. ‘Technique as before. March 19.—¥ruption not so perfect as in Monkey No. 1. Vesicles not so defined and plump. Monkey very wild, and has made sites of incisions bleed by dashing from side to side of cage, which may be in some degree the cause. Photograph taken. Vesicular pulp removed and ground up with four times its weight of dilute glycerine. With some of this emulsion Monkey No. 3 (young male rhesus) inoculated same day in ten linear incisions on shaved area of back. . March 26.—All isertions “taken” well. Photographed. Material collected and glycerinated, pulp being diluted about fifteen times, by mistake. Some used 1902.] The Inter-relationship of Variola and Vaccinia. 129 for inoculation in twelve insertions of Monkey No. 4 (young male rhesus) same day. April 2.—All places have “‘ taken” well, although material used for inoculation had been so diluted, Photographed. Vesicles scraped and pulp glycerinated. Calf Experiments. March 12.—Inoculated Calf No. 1 in half dozen long linear incisions by method usually employed in current vaccinations at Government Establishment, with glycerinated pulp from Monkey No. 1. March 16.—Lines of incision slightly raised and red. March 17.—Dr. Fremlin found few small vesicles had developed. These he clamped, inserting material obtained thereby on Calf No. 2. March 21.—There were evident signs of “taking” at all po nts of insertion, the lines of incision being elevated and with a tendency to vesiculation. But appearances not considered sufficiently typical to permit of material removed being utilised for vaccination of children. At this point, owing to unforeseen circumstances, this particular series of experiments was discontinued. THIRD SERIES. February 25, 1901.—At the Jenner Institute inoculated medium-sized macaque monkey with small-pox emulsion, made by working up scrapings from P.M. cases of the disease (received from Dr. Thomson, of the Belvedere Hospital, Glasgow), in a small amount of pure glycerine. Technique as in previous experiments. Incisions made on the monkey’s hack with blunt scalpel, which had lost its temper by constant passing through the flame, so that all incisions did not apparently penetrate to the true skin. March 4.—‘ Taken” well, though not throughout all insertions as in first monkey of the last series, but as failure had only occurred where there was no mark of incision, it was probably for reason mentioned above, as the eruption which had appeared was good. Not markedly vesicular; lines of incision which were covered with slight crust being surrounded by a pinkish papular eruption. Photograph taken. Scraped with aid of compression forceps, and rubbed up material in small amount of NaCl 0°7 per cent. solution. Monkey No. 2 inoculated this day, with emulsion of scrapings from Monkey No. 1. Technique as before. March 16.—All insertions “taken”; slightly more vesicular than in No. 1. Photographed. Compression forceps applied and scrapings removed and rubbed up in small mortar with NaCl solution. Monkey No. 3 inoculated immediately, on the back, with emulsion of material obtained from No. 2. Animal very young ; hair downy and not easily shaved. March 18.—All insertions have “taken” and have wide whitish vesicular Margin, but appearance rather spoiled as hair on back has grown so rapidly. Lymph oozed up when compression forceps applied. Photographed. Scrapings rubbed up with NaCl solution, of which small quantity was used immediately for inoculation of Monkey No. 4. Remainder glycerinated, tubed and stored in ice- chest for future trial on calf. March 23.—Monkey No. 4, an old animal, had not apparently taken as well as No. 3, so series was discontinued. April 20.—Monkey No. 3 vaccinated in six incisions on outside of thigh with current vaccine lymph of known potency. VOL. LXXI. L 13 Dr. 8. Monckton Copeman. [Niovemlt: April 27.—No result, although monkeys not previously protected take vaccinia as successfully as in the hnman subject.* Calf Experiments. March 22.—Calf No. 1 (1832), at the Government Animal Vaccine Establish- ment, inoculated with small quantity of glycerinated emulsion of vesicular pulp from Monkey No. 3. March 2'7.—Tiny papules and vesicles which had made ‘their appearance along lines of incisions removed with Volkmann’s spoon and glycerinated. March 29.—Material collected on March 27 inserted into three long incisions on Calf No. 2 (1842). April 3.—Fifth day. Lines of all three incisions occupied by good vesicles, Photograph taken by Dr. Green. Vesicle pulp (0°37 gramme) removed and glycerinated. April 24.—Calf No. 3 (1890), inoculated (with portion of material collected on April 3) in thirty-six incisions on abdomen and scrotum. April 29.—All insertions had “taken” well, vesicles surrounded with slight pink areola. Vesicles scraped and pulp glycerinated. Emulsion stored in ice- chest. October 3.—Four c.c. of this emulsion used for vaccination of calf at the Jenner Institute, by numerous linear incisions extending nearly whole length of abdomen, after manner usually employed at Government stations. October 8.—Appearance indistinguishab!e from normal vaccination. 63 grammes of vesicle pulp collected and glycerinated. The glycerinated emulsion prepared from material removed from tke calf on October 8 was subsequently employed for the vaccination of other calves, a strain of lymph being thus obtained which continued to give excellent results both on children and calves. But the strain was never brought into general use, and all the glycerinated emulsion remaining was eventually destroyed. FourtH SERIES. April 29, 1901.—At the Jenner Institute young rhesus monkey shaved on back, as in previous experiments, and inoculated in a dozen linear incisions with glyceri- nated emulsion of S.P. vesicle pulp received from Dr. Thomson, of the Belvedere Hospital, Glasgow, on March 26, 1901. April 2 (120 hours).—Had “ taken” so well that I decided to collect; lines of incision distinctly vesicular. After taking photo, washed inoculated area, and removed pulp with aid of clamp forceps. Thin lines of altered epithelium came off as in a good calf vaccination. No “ crusting.” Monkey No. 2 inoculated same day with material obtained as above and subse- quently triturated in small glass mortar with small quantity of equal parts of glycerine and normal saline solution. May 9 (120 hours).—Every insertion “taken” successfully. More markedly vesicular in places than Monkey No. 1. Photograph taken, followed by usual process of collection and glycerination af vesicie pulp. Monkey No. 3 inoculated on back in eight diagonal incisions. Emulsion remain- ing over taken up into capillary tubes, of which two given to Dr. Fremlin for trial on calf at A.V.H. on May 15. May 13 (120 hours).—Eruption of perfectly vesicular character along course of all incisions made, the centre in each instance being occupied by thin linear crust. * Copeman, ‘Journal of Pathology and Bacteriology,’ May, 1894. 1902.) The LInter-relationship of Variola and Vaccinia. 131 Drs. Blaxall and Fremlin, on seeing the animal, both described eruption as being equal to that which in case of calf vaccination ‘vould be entered in official records as y.g. (very good). Monkey photographed and vesicle pulp collected and glyceri- nated. Portion of emulsion used same day for inoculation of Monkey No. 4. Another portion handed over to Dr. Fremlin for trial on calf at A.V.E. May 18.—Monkey No. 4 “taken” well. Vesicle pulp collected, emulsified, tubed and stored in ice-chest. Series not continued beyond this stage, as laboratory man had failed in attempts to obtain further supply of young rhcesus monkeys. Calf Experiments. May 8.—Emulsion of vesicle pulp from Monkey No. 1, used at A.V.E. for inoculation of Calf No. 1418. May 13.—Dr. Blaxall noted “ No vesiculation, slight thickening in one line.” May 15.—Material from Monkeys No. 2 and No. 3, inoculated on calves at A.V.E. by Dr. Fremlin. May 20.—Both calves had “ taken” to a certain extent, the result being most marked in calf inoculated from Monkey No. 3. Material collected, glycerinated, and stored. Portion subsequently handed to Medical Director of Jenner Institute for further trial. May 26.—Inoculated calf at A.V.E. with emulsion of second removal fron Monkey No. 3. May 3\.—Perfect vesicular eruption along course of all incisions. General effect indistinguishable from that obtained with the current vaccine lymph of the Government Establishment. In view of successful results following on vaccination of children with former lymph stocks raised in similar fashion, it appeared unnecessary to employ this particular lymph for infantile vaccinations. But on vaccinating my own arm with it direct from the calf, I succeeded in raising by the eighth day a fairly typical vesicle, an effect in excess of that obtained by me on my own person at previous attempts at vaccination. No further transference of this lymph was attempted. My first series of experiments bad not long been concluded, when I came across a reference to an account of similar work which had been carried out by Dr. Hilerts de Haan. ‘The reference occurred in a paper by Dr. Bruno Galli-Valerio,* and on hunting up Dr. de Haan’s original paper, which is entitled, ‘‘ Vaccine et Rétrovaccine a Batavia,” 7 I found that it contained an account of a lengthy series of experi- ments on the variolation of monkeys and on the transference of the resulting affection to calves. Dr. de Haan’s work proved of special interest to me for the reason that, quite independently, we had been able to corroborate one anothers’ work, except as regards the transierence of the strain of variola vaccine to the human subject—a final test which Dr. Eilerts de Haan did not, as he says, feel justified in attempting, in view of the unfortunate experience of Chauveau, in connection with his abortive * ¢Centralblatt fiir Bakteriologie, March 28, 1892, p. 380 et seq. + ‘Annales de |’Institut Pasteur,’ 1896, p. 169. ie, 132 Dy. 8. Monckton Copeman. [Nov. 13, attempts at variolation of the cow. Dr. de Haan’s own words may be quoted :—“‘ Je reconnais quil manque a ma démonstration d’avoir rapporté la variole mitigée du singe sur l’homme ; c’est une expérience que je ne me suis pas cru en droit de faire. L’expérience de Chauveau enselgne a étre prudent, et je ne me croirais autorisé a faire cette tentative que si le vaccin ordinaire dont je me sers me manquait au moment d’une épidémie. Mais j’espére qu’on répétera mes experiences a ce sujet.” For the purpose of his experiments Dr. de Haan made use of the macaque monkey (JMacacus cynomolgus), which is common in the Dutch East Indies, and therefore was readily obtainable. His inocula- tions were, in each instance, made on a portion of the animal’s back, which was first shaved and then cleansed with soap and water, followed by a solution of boric and salicylic acids. In his first series of variolations of the monkey, the small-pox lymph employed was obtained from a Javanese child, no statement however being made by him as to the age of the child, whether or not it had ever . been vaccinated, or at what stage of the disease the lymph was taken. Transferred to the monkey, this lymph gave rise, in a week’s time, to well-marked vesicles at the site of inoculation, while, in addition, afew papules were observed on the lips and the extremities. Seven subsequent primary variolations were, however, successfully carried out on monkeys, in only one of which was any evidence of generali- zation observed. | | From the contents of the vesicles of the monkey first-mentioned, a second was inoculated, which in seven days developed vesicles at the inoculated points only. From this second monkey, in due course, a third animal was inoculated ; from this a fourth, and so on, through a series of seven monkeys. | From the sixth monkey of this series, a calf was inoculated, which five days later presented an appearance indistinguishable from a typical vaccination. From the seventh monkey also, of the series, a calf was inoculated, with the result again, that after an interval of five days, perfect vaccine vesicles appeared at the site of each insertion of the lymph. From this calf another was vaccinated with complete SUCCESS. In a second series of experiments, lymph at the fourth remove, in the monkey, from human small-pox, gave rise to perfect vesicles when inserted on the skin of a calf, and the strain of vaccine lymph thus obtained was carried on successfully, through eight removes from calf to calf. Monkeys and calves all failed to react to subsequent inocu- lation with the strain of vaccine lymph in current use. The results of my own experiments may be briefly summarised as follows :— 1902.] The Inter-relationship of Variola and Vaccinia. 133 In each of the separate series of experiments the human small-pox lymph or pulp was first inoculated directly on calves, and in every instance, so far as could be observed, with altogether negative results. But with monkeys success was as invariably obtained, and when, after one or more passages through this animal, the contents of the local inoculation vesicles were employed for insertion on the calf, an effect was now produced which, after two or three removes in that animal, was indistinguishable from typical vaccinia. Moreover, from the contents of vesicles raised in this manner on the ealf, a number of children have, in turn, been vaccinated, some of whom were afterwards kept under observation for as long a period as a couple of months. Every such vaccination “took” normally, and in no case was any bad result subsequently observed by myself, or reported by the parents of the children; no “ generalisation” of the eruption occur- ring in any instance. In conclusion, I desire to call attention to the somewhat remarkable fact that a mild and strictly localised form of small-pox, such as is induced in the monkey by the inoculation of material from cases of the generalised disease in man, should, when transferred to the calf, ‘“‘take” readily with the production of a vesicular eruption of non- infectious character in that animal, whereas it is well known that successful transference of small-pox direct from man to calf can only be accomplished with the utmost difficulty. The experimental results obtained in the course of the research, an account of which has been set out in this paper, all tend, then, to confirm the view that the vaccinia of Jenner’s time was derived, in all probability, from a comparatively mild form of human small-pox. In addition, I think it will be admitted that the work has afforded conclusive evidence of the essential identity of the virus of small-pox and cow-pox or vaccinia. 134 Sir Norman Lockyer and Dr. W. J. 8. Lockyer. [Oct. 18, “On the Similarity of the Short-period Pressure Variation over Large Areas.” By Sir Norman Lockyer, K.C.B., F.R.S., and Wiuuiam J. S. Lockyrer, M.A., Ph.D., F.R.AS. . Received October 18,—Read December 4, 1902. [Prates 1 and 2. | In a paper presented in June last to the Society,* we pointed out the existence of a short-period oscillation of barometric pressure over the Indian area corresponding generally with a variation in the per- centage number of prominences recorded on the sun’s limb. This oscillation was further shown not to be limited to the Indian area, but to be marked at a far distant station, as Cordoba, in South America. The present paper, which is a continuation of this investigation, was undertaken to extend the research over a larger area. The monthly means of the pressure variations for each station have been divided as previously into two periods, namely, those months in which the pressures are above and those in which they are below the normal, the normal being the mean pressure for the whole period under investigation in each locality. In dealing with large areas, it happens that during the same period of time (that is generally but not invariably six months), the pressure is above the normal in some places, and below the normal in others ; the similarity of the curves representing the variation of the mean for this period, from year to year, indicates therefore that, in one case, a rise in the curve denotes that the pressure is higher, and, in the other, that the pressure is not so low as usual. The accompanying curve (Plate 1) illustrates the variations of pres- sure which have been analysed. Commencing with Indian pressures (as represented by Bombay) the area was gradually extended to Ceylon (Colombo), Java (Batavia), Mauritius, and finally to Australia (Perth, Adelaide and Sydney). In this set of curves about the same months are in question, so that the pressure variations refer in the northern hemisphere to the low pressure (summer) months, and in the southern hemisphere to the high pressure (winter) months. The striking similarity between these curves shows that over the whole of this area, which includes both north and south latitudes, the same kind of variations is in action, and that therefore the whole region is intimately connected meteorologically. It was indicated in our previous paper that the pressure of Cordoba, in South America, was the inverse of that of India for the same period. Since the Indian pressure variations are seen now to extend over a * “On Some Phenomena which suggest a Short-Period of Solar and Meteoro- logical Changes.” ‘ Roy. Soc. Proc.,’ vol. 70, p. 500. “AWIAT#OR AOdsuns Jo VUMTUIM pUR vUTXLUT Jo syDode oyy quosoIded soUT] UYOIG PUR snONUTyUOD [vO1VIOA OT,],—'a0A7 OOS OOS. 0-09¢I 0-028! 0-098! Ose" cwinwusny) ASANGAS : { OSES? scuiWaa HOIH fidas-uydv) ooe 3YNSSIYd oor eewinwuisny) “S0Iv13ag0V Os! ‘SHLW ‘Ud HDIH ee ee ee oN ee | SSON4NIWOUd 0-006! 0-068! 0-08¢! 0-028! 0-098! Lid ‘TL 700 “00g hosp -90n7 “wahyoo'T pun sahyoorT af w Lockyer and Lockyer. 200 PROMINENCES ON SUN. (TACCHINI.) [APR.-SEPT.] 500 LOW PR. MTHS. 400 BOMBAY. °° (NDIA. 200 100 29-7000 PRESSURE. §&® [APR-SEPT]] 60 LOW PR. MTHS. COLOMBO. 29-840 (CEYLON.) PREGSUREG Cs) (UUN-OCT.] HIGH PR. MTHS. 760-00 BATAVIA. (JAVA.) 759:00 PRESSURE. oe [MAY-OcT,] 10 700 PRESSURE. s HIGH PR. MTHS. 100 MAURITIUS. 90 PRESSURE. 70 [APR=SEPT] 50 HIGH PR. MTHS, 30 PERTH. gle (GARDENS.) 20 (W. AUSTA) 70 PRESSURE. é (APR-SEPT] HIGH PR. MTHS. 150 ADELAIDE. (AUSTRALIA.) (10) PRESSURE. 30:0 [APR-SEPT] HIGH PR.MTHS. 59.9 SY DNEY. (AUSTRALIA, 1860-0 1860-0 1870-0 1870-0 1880-0 Sarees: 1880-0 Proc. Roy. Soe., vol. 71, Pl. 1. 1890-0 1900-0 1890-0 2 ja0ha Note.—The vertical continuous and broken lines represent the epochs of maxima and minima of sunspot activity. * 5 fh Salt eatin r Li 1 ‘ : e ’ en ry Pi, ' Le * f 1, te , Vi > ind Dy bot ey nes Datos ae tlt ; ‘i f i ’ 1 te rv Reet Fgh ath hy iy Sey say Sha ai Nee ind . ria a3 FAR ae eat ny Gp Aine rege ahd Aah inc icy Satheass ? 1 * 7 a. i) ob, x . = 7 ; ,, i t y ‘ ral x al ‘ 4j si / a i Mo ' ; : : ; My a wile x - i ‘ 4. i i ‘oe f s : , 9 \ Farka 5 « dat hy wap Po, © raat i a oy > se € wt Fn Se beviieas is Aah Soe WAY et ehh ur= Nh Mech soe, Oana ge eye he el ees: sail nes ' Se io ah hi \ - i = . 1 Miomactiret Rime That ad lena pesecen tabs riage let madsen te nioey el ante gem rales nei: ohm alt Mylan glee ia h Kee gto } a es } ¢ i 4 i i eae ig Ke) t : mf . ¢ ax im if + “ + ms i , - x . iL “ ‘X " . o/s ‘ - . “ - 4 “ - * : ve ¢ 4 iL + > . a Bad ' ‘ ¢ f % ® i 7 : i ~ ‘ -— ‘ Lockyer and Lockyer PROMINENCES. ON SUN. (TaccHINI) =: [°° (SCALE INVERTED.) 200 PRESSURE. 29°860 [ocT-mMAR] 19) -880 HIGH PR.MTHS. & -820 BOMBAY. 200 (INDIA) “910 ISCALE INVERTED%50.990 72-800 PRESSURE. [APR-SEPT]] HIGH PR. MTHS. CORDOBA goo (S. AMERICA) -700 72-500 PRESSURE. 50100 fOcT.-MAR] HIGH PR.MTHS. -050 MOBILE. (U.S.A) -000 PRESSURE. 30-150 ([NOV.-FEB] “100 HIGH PR. MTHS. ‘080 JACKSONVILLE. 06° (FLORIDA U.S.A.) 5.020 PRESSURE. 30-100 [Nov-mMAR] 080 +060 HIGH PR. MTHS. 6 PENSACOLA. .o20 (FLORIDA U.S.A) 30-000 PRESSURE. 350:000 (NOV. APR] HIGH PR. MTHS. .950 SAN DIEGO. (CAL.U.S.A.) 39.900 R fo) een Seay oaSrae (ee Proc. Roy. Sor., vol. '71, Pl. 2. 1880-0 1890-0 1900-0 =a: eRe ' 1 ' 1 i] ' i) 1 1 ! 1 1 ! ! { ! 1 1 ! 1 ! t ! ! t 1 i 1 ! ( 1 1 ! 1 1880-0 1890-0 1900-0 Note.—The vertical continuous and broken lines represent the epochs of maxima and minima of sunspot activity. av a ae * eee ee er t 1902.] Short-period Pressure Variation over Large Areas. 135 region on both sides of the equator, it was important to study the extent of the region in the New World in which pressure variations similar to those of Cordoba had been recorded. As Cordoba represents an area south of the equator, a portion of the United States of America was taken as typifying an area with north latitude and in about the same longitude, and a commencement was made along the lowest available parallel of latitude. This was rendered possible by the kindness of Professor F. Bigelow, of the Weather Bureau, who very generously forwarded proof-sheets of a new reduction of the pressures of many stations. We wish to take this opportunity of expressing to him our best thanks. Treating these pressures in the same way as those formerly in- vestigated in the Indian region, several stations which had the best record were chosen. A graphical representation of the variations of four of these stations (Mobile, Alabama; Jacksonville and Pensacola, Florida; San Diego, California) is given in Plate 2, and for the sake of comparison the pressure of Cordoba, with the inverted curves representing the Bombay pressure and solar prominence variation. This series of curves refers in all cases to the variations of the means of the high pressure (winter) months (October to March in most cases). At Cordoba, which has a southern latitude, the high pressure months extend from April to September. ~ The result of the comparison shows that in this region of the world we have a large area, the pressure variations of which are strikingly similar to, but are the inverse of, those recorded in nearly the anti- podal part of the globe. The facts observed are so suggestive that we are continuing the inquiry by collecting and discussing observations made in other areas. Although the general agreement between the two main sets of curves is most striking, there are minor differences which probably, as stated in the previous paper, will eventually help us to determine those cases in which the prominence effects on pressure are masked by some special local conditions. It may be added that the available observations of prominences refer more directly to their quantity than to their intensity. We wish to express our thanks to Mr. W. N. Shaw, F.R.S., who kindly placed the records of the Meteorological Office at our disposal ; and to Mr. Hodgson, who has extracted the requisite data from the available records of pressure, and constructed some of the curves. 136 Mr. J. H. Jeans. On the Vibrations and [Nov. 8, © On the Vibrations and Stability of a Gravitating Planet.” By J. H. Jeans, B.A., Isaac Newton Student and Fellow of Trinity College, Cambridge. Communicated by Professor G. H. Darwin, F.R.S. Received November 8,—Read Decem- ber 4, 1902. (Abstract.) The first part of the paper deals with the vibrations and stability of a gravitating elastic sphere. The matter is not necessarily homo- geneous, but is arranged in spherical layers. It is pointed out that, in the classical investigation of the displacements produced in a gravi- tating sphere by given surface-forces, the most important of the gravi- tational terms is omitted. The effect of this omission is to necessitate a correction, and this may entirely invalidate the solution when we are dealing with spheres of the size of the earth or other planets. In fact, it appears that for a gravitating solid of the kind we are discussing the spherical configuration may be one of unstable equilibrium, the instability being brought about by the gravitational terms in the manner already explained in a former paper.* Let a be the radius of the sphere, and y be the gravitation constant ; let p, be the mean density, and A, the mean value of A, one of the elastic constants. A general argument shows that the spherical con- figuration will be stable or unstable according as where ¢ is a pure number which must be comparable with unity. If we put in an artificial field of force we can imagine a spherical configuration of equilibrium in which the density and elastic constants. . have uniform values throughout. The artificial field of force is, of course, equal and opposite to the gravitational field produced by the matter of the solid. The stability or instability is determined by a criterion of the form of (i), and ¢ can now be calculated exactly. If we suppose Ay to represent A + 2 in Love’s notation (m + m in that of Thomson and Tait), we find the values ay Ss IIHS Vaan ty = 0), @) il edee nw eM 2) Ne The vibration through which instability first enters is one in which the displacement at every point is proportional to a harmonic of the jirst order. It appears probable that for spheres which are not homo- * “The Stability of a Spherical Nebula,” ‘ Phil. Trans.,’ A, vol. 199, p. 1. 1902.] Stability of a Grarvitating Planet. | an geneous, but in which the density is greatest near the centre, the values of ¢ will be greater than those just stated, but the critical vibration will still be such that the displacement is proportional to a first harmonic term. In the former paper, already referred to, the suggestion was put forward that the instability of a nebular sun or planet, which, upon the nebular hypothesis, is supposed ultimately to result in the ejection of a satellite, may be largely brought about by a gravitational tendency to instability of the kind just described. We take, for the moment, an extreme hypothesis, and imagine that this agency is the preponderating agency, and that the rotational tendency to instability may be disregarded in comparison. We then find that when the ejec- tion of a satellite is taking place, or has just been completed, the value Of ypy°a?/Ay must be nearly equal to ¢. Except for the changes which have occurred since the consolidation of the planets, the solar system supplies material for testing this con- clusion. When a number of planets of varying masses have thrown off satellites, we find (upon our present extreme hypothesis) that the masses ought to be proportional to the squares of the radii. It is found that this law is approximately obeyed in the solar system. It is further found that the absolute values of the masses and radii are approximately such as would be expected. It is interesting to compare two extreme hypotheses, the first refer- ring the phenomena of planetary evolution solely to rotational, the second solely to gravitational, instability. Given the approximate values of the density and elasticity of a planet, and the fact that this planet has thrown off a satellite ; then the former hypothesis leads to a certain inference as to the angular momentum of the system, the latter to an inference as to the radius of the primary. The former leads to no inference at all as to the size of planets which are to be expected—they are as likely to be of the size of billiard-balls as of the size of the planets of our system—while the latter leads to no inference as to the angular momentum of the system, but presupposes it to be small. The contention of the present paper is that the inferences which are drawn from the former hypothesis are not borne out by observation on the planets of our system, while those which are drawn from. the latter are borne out as closely as could be expected. The true hypo- thesis must of necessity lie somewhere between the two extremes which we are comparing, but the evidence seems to show that it is much nearer to the latter (gravitational) than to the former (rota- tional). We next consider a number of questions connected with the figure of the earth. It seems to be almost certain that the present elastic constants of the earth are such that a state of spherical symmetry would be one of stable equilibrium. On the other hand, if we look 138 Prof. H. Marshall Ward. fect of Mineral [Nov. 4, backwards through the history of our planet, we probably come to a time when the rigidity was so small that the stable configuration of equilibrium would be unsymmetrical. At this time the earth would be pear-shaped, and the transition to the present approximately spherical form would take place through a series of ruptures. It is suggested that the earth, in spite of this series of ruptures, still retains traces of a pear-shaped configuration. Such a configuration would possess a single axis of symmetry, and this, it is suggested, is an axis which meets the earth’s surface somewhere in the neighbourhood of England (or possibly some hundreds of miles to the south-west of England). Starting from England, we find that England is at the centre of a hemisphere which is practically all land: this would be the blunt end of our pear. Bounding the hemisphere we have a great circle of which England is the pole, and it is over this circle that earthquakes and volcanoes are of most frequent occurrence. Now, if we suppose our pear contracting to a spherical shape, we notice that it would probably be in the neighbourhood of its equator that the changes in curvature and the relative displacements would be greatest, and hence we should expect to find earthquakes and volcanoes in greatest numbers near to this circle. Passing still further from England we come to a great region of deep seas—the Pacific, South Atlantic, and Indian oceans: these may mark the place where the “waist” of the pear occurred. Lastly, we come almost at the antipodes of England to the Aus- tralian continent: this may mark the remains of the stalk-end of the pear. “Experiments on the Effect of Mineral Starvation on the Para- sitism of the Uredine Fungus, Puccinia dispersa, on species of Bromus.” By H. MARSHALL WarD, Sc.D., F.R.S., Professor of Botany in the University of Cambridge. Received Novem- ber 4—Read November 27, 1902. I have shown in previous publications that the parasitic Uredine Puccinia dispersa, growing on grasses of the genus Bromus, is usually very closely adapted to the species of host-plant selected: that although no morphological differences can be detected between the fungus as met with on different species of Bromus—aA, B, C, &c.—it by no means follows that spores from the parasite, as found growing on A, will infect B or C, or that spores from the fungus as reared on B or C will infect the species A. On the whole, it has so far appeared probable that the fungus | growing on a given species—c.g., B. mollis—infects most readily those 1902. | Starvation on the Parasitism of Puccinia. 139 species which are most nearly related to B. mollis, less and less readily species more remote from £. mollis, but in the same sub-genus, and least readily, or not at all, species in other sub-genera. In the attempt to obtain some insight into what causes are at the bottom of this remarkable phenomenon—the predisposition to, or immunity from, infection of the host by the parasite—it was shown that anatomical differences on the part of the host-plants, such as the sizes and numbers of stomata, hairs, and so forth, do not suffice to explain it, since no relation could be detected between the curves expressing the percentage of infection and those expressing the sizes, numbers, &c., of the hairs, stomata, &c. On the other hand, the evidence suggested some such assumption as the following. The fungus, when growing on a species of Lromus A, may refuse to infect another species B, either because B secretes some body of the nature of an enzyme or anti-toxin which effectually keeps the mycelium of the fungus at bay, or because the fungus habituated to the peculiar nutritive or other conditions afforded it by the host-plant A, cannot immediately adapt itself to the very different conditions offered by the species B. Although the attempts to isolate any such anti-toxin failed, and experiments of a preliminary character to test the effect of differences of nutrition yielded little or nothing of a positive nature, | showed in the discussion of the probable factors concerned that some subtle relations between host and parasite must be assumed to account for the curious facts of immunity and predisposition on the part of the former, and of capacity and incapacity for infection on the part of the latter, in each case in various degrees according to the species of host offered for infection, or on which the fungus has hitherto been reared. During the past year I have attempted to pursue this subject further, and limit myself for the present to the following theme. If the varying infective power of the fungus towards different species of host-plant is derived solely from the “nutritive conditions” afforded it by the host-plant it has hitherto been growing upon, two cases are possible—(1) these ‘‘ nutritive conditions ” may be simply the expression of the power of the tissues to yield certain food-substances to the parasite in proper proportions and in sufficient quantity, or (2) they may imply some more subtle relations between the mycelium of the fungus and the living contents of the host-cells. For instance, it may be not sufficient that the food-substances suitable to the fungus should exist in the cells of the host, but they must be there in a certain superabundance, or presented in a certain manner, and so on; or, it may be that the fungus must be vigorous up to a certain standard before it can obtain a hold on such food, and so on. In order to test some of the possibilities here referred to I planned 140 Prof. H. Marshall Ward. fect of Mineral [Nov. 4, experiments to see whether starving the host-plant of one or other of its necessary food-materials would (1) affect its predisposition to infection, or (2) affect the capacity for infection of the fungus grown on the starved plants, or (3) in any other way affect the fungus or its host. On July 7, 1902, fourteen beakers were selected and filled with equal quantities of a clean coarse-grey sand, carefully washed and dried, but still containing traces of ingredients which the root-hairs of such grasses as the Bromes are capable of selecting. This was not to be avoided or regretted, since my object was not to attempt to grow seed- lings totally deprived of necessary salts, resulting in their premature death, but to bring up plants so starved of certain such ingredients that while they could go on living long enough for the purposes of the experiment, they would nevertheless exhibit the effects of the deficiency, and possibly re-act on the parasite. The beakers of sand were lettered A to O, and treated as follows :— The beaker A received 200 ¢.c. of distilled water only, so that the only mineral supplies available to the seedlings—after exhausting the traces in the endosperm—would be such as the root-hairs could dis- solve from the sand grains. Another beaker received an equal quantity of a cold water-extract of fresh horse-dung, representing a liquid of high manurial value. A third beaker received an equal quantity of a normal nutritive mineral solution containing nitrates, phosphates, and sulphates of potassium, calcium and magnesium ; and a fourth the same with the addition of five drops of a dilute solution of ferric chloric. One each of the remaining beakers received a similar solution of the nutritive salts, but in each case with the omission of one element, viz.: calcium, magnesium, nitrogen, phosphorous, or potassium ; while the other beaker of each pair received the same, together with traces of iron salt. By these means I had prepared a soil in each beaker which was of suitable consistency for growing such sand-loving grasses as Bromes, but which was in each case deficient in one or other of the necessary ingredients for normal nutrition—except in so far as I added such in- gredients—but in no case absolutely devoid of these necessary salts, as otherwise the seedlings could not be expected to live long enough for the purposes of the experiment. On July 8, having allowed the solutions to soak completely into the sand, seven grains of Bromus secalinus, carefully cleaned and selected, were sown in each beaker, and the whole left under large bell-jars to germinate in a suitably lighted position in the laboratory. Germination followed in due course, though somewhat slowly, as the sand was rather wet, and on July 15, from two to five seedlings about 30—50 mm. high were showing above ground in each beaker except 1902. | Starvation on the Parasitism of Puccinia. 141 one, and the whole were now placed in a sheltered situation in the open, the bell-jars being raised on large blocks so that their rims just covered the rims of the beakers ; thus permitting a free circulation of air, but protecting the whole from beating winds or rain. An awning was also provided, and carefully adjusted during hot days as necessary. Growth proceeded fairly rapidly, though of course some retardation was inevitable owing to the impervious nature of the beakers as con- trasted with pots having porous sides, and by July 18 each beaker had from two to seven satisfactory seedlings. On July 23, ten of the beakers had from five to seven excellent seedlings each: three had but four each, and one had only two suc- cessful seedlings. So far there was but little difference observable between the different seedlings, each of which was unfolding its third leaf. Nevertheless, there was evidence that the stores of food- materials in the indosperm were now being exhausted, and that the root-system was beginning to feel the effects of the differences in mineral supplies. Consequently, I fished out the remains of dead or un-germinated grains, and proceeded to infect certain of the seed- lings in each beaker with Uredo-spores of Puccinia dispersa obtained from Bromus squarrosus. This species was selected because I had an abundant stock of vigorous spores at hand, and had already satisfied myself that they readily infect Bromus secalinus. The infection was effected by placing equal doses of the fresh spores on each leaf chosen ; the bell-jars were then lowered in order to keep the plants in a moist atmosphere for 24 hours, and then raised again on blocks for another 24 hours, after which the bell-jars were removed each morning, and only replaced at night or during rain. The results of infection were already evident on one or two plants on July 31 and August 1, increasingly so on August 3, and by August 5 the pustules were as prominent as they ever became during the period of the experiment—~.c., up to August 29—except in so far as some of the larger ones ran together or produced more spores. The results are summarised in Table I, but it should be noted that it has been impossible to compress into the table the details of the observations from day to day as to the stature, robustness, colour, number and breadth of leaves and so on, or as to the rate of develop- ment of the plants and of the flecks and pustules of the parasite. In order that the reader may gather an impression of the appear- ance of these seedlings during the progress of the experiment, how- ever, | append photographs* of one set of the beakers in fig. 1, and of a representative series of the plants themselves, carefully extracted * T have to thank my son, F. K. Ward, for the preparation of these photo- graphs. 142 Prof. H. Marshall Ward. fect of Mineral [Nov. 4, from the loose sand in water and the roots washed and displayed on black paper (fig. 2); while in fig. 3 are appended other representative specimens, similarly carefully extracted and washed and displayed with the roots on black paper, and the shoots on white paper. Fie. 1.—Specimen from Experiment 47 (Table 1), photographed 44 days after sowing, showing differences in stature and breadth of leaf. The plants in A had received distilled water only; in B nutritive solution minus phosphorus ; in E no nitrogen; in K no magnesia, and with five drops of ferric chloride; in N normal mineral solution with iron salts. L had no nitrogen, but five drops of iron solution. The facts recorded in this table and in the legend to these figures speak for themselves, but the following details may be noted. The effects of the deficiency of all salts (A) of phosphorus (B) and of nitrogen (E) are distinctly observable in the poor development of the roots, as well as of the feeble and narrow leaves; but although the opposite effects of the rich manurial action of the horse-dung decoction (O) and the normal mineral solution (G) are plain, the differential results of the solutions minus calcium and potassium are by no means obvious, nor did the addition of the five drops of dilute ferric chloride seem to make much difference. Apart from an error in regard to the calcium (c, Table I), I attribute this to the circumstance that it is extremely difficult to starve such INDEX SLIP. (Procrepines, No. 469.) Burcu, George J.—Contributions to a Theory of the Capillary EHlectro- meter. II.—On an Improved Form of Instrument. Roy. Soc. Proc., vol. 71, 1902, pp. 102-105. Capillary Electrometer for use with High Powers, Burcu, George J. Roy. Soc. Proc., vol. 71, 1902, pp. 102-105. Lez, Alice, LEwenz, M. A., and Pearson, K.—On the Correlation of the Mental and Physical Characters in Man. Part II. Roy. Soc. Proe., vol. 71, 1902, pp. 106-114. Lewenz, M. A., Pearson, K., and Lun, Alice.—On the Correlation of the Mental and Physical Characters in Man. Part IT. Roy. Soe. Proce., vol. 71, 1902, pp. 106-114. PEARSON, K., Len, Alice, and Lewrnz,M. A.—On the Correlation of the Mental and Physical Characters in Man. Part II. Roy. Soc. Proc., vol. 71, 1902, pp. 106-114. Man-——Correlation of Mental and Physical Characters. Lez, Alice, LEwrnz, M. A., and Pearson, K. Roy. Soe. Proc., vol. 71, 1902, pp. 106-114. SHERRINGTON, C. S., and Lasterr, EK. E.—Note upon Descending Intrinsic Spinal Tracts in the Mammalian Cord. Roy. Soc. Proc., vol. 71, 1902, pp. 115-121. Lastett, H. H., and SHerRineton, C. 8.—Note upon Descending Intrinsic Spinal Tracts in the Mammalian Cord. Roy. Soc. Proc., vol. 71, 1902, pp. 115-121. Degeneration descending in Spinal Cord. SHERRINGTON, C. S., and Lastzett, H. E. Roy. Soc. Proe., vol. 71, 1902, pp. 115-121. Spinal Segments—Connections between. SueReineton, C.8., and Laszett, HE. EH. Roy. Soe. Proc., vol. 71, 1902, pp. 115-121. Tracts, Internuncial, in Spinal Cord. SHERRINGTON, C.S, and Laszett, H. E. Roy. Soe. Proc., vol. 71, 1902, pp. 115-121, erowod ik shi 9 oa 708 a01-201 .4q ee Ty Lor ae oh. oe . bo ait to noitorr0 oft nO— HE, ae bane b M aunt Ti ded. fa ry I, % Lov 90rd .o ie BIL-00! qg £00 d + cO—. aii r {iL asad wel at iL gol a4 ae aN Jov oF 208. old to noitslorroD ail , lk estoyit oe = A moassat bos A .M svawal ook a % ‘prr-d0f aq SOeL IV Jor 20a 00k oH AE & avail Bue @ bro) maiscmone i ont at ISi-81l qq S001 IY tow eel 908 uaa hie sient gcibsesee nogas oto oianintal gaibaooeed comp stoh—. ws FSE-88l .qq g0ef LV for oor "208 ong Bo ‘cnet ws aa arsuaait bao 2 5 eoranttrane ‘Fer-alt aq S0ek Lv Jor .2 oH . ae: a & qveuk tae °- eee , ASL-OEL GG. goer a Lov 5 or 508 .yo8 na ti AM sraaeald bos 6.0 ASt-aE a. goe! AY Jor 201 08 col a = ~ aa a Coprman, S. Monckton.—The Inter-relationship of Variola and Vaccinia. Roy. Soc. Proe., vol. 71, 1902, pp. 121-133. Variola— Experimental Transformation into Vaccinia. CoprmMAN, S. Monckton. Roy. Soe. Proc., vol. 71, 1902, pp. 121-138. Vaccinia~— Origin from Variola. CopEMAN, S. Monckton. Roy. Soe. Proc., vol. 71, 1902, pp. 121-1383. Monkey, Variola and Vaccinia in. CopEMAN, S. Monckton. Roy. Soc. Proc., vol. 71, 1902, pp. 121-133 Lockyer, Sir Norman, and Lockyer, Dr. W. J. 8.—On the Similarity of the Short- period Pressure Variation over Large Areas. Roy. Soc. Proc., vol. 71, 1902, pp. 184-135. Barometric Variations, Short Period—Similarity over Large Regions. Lockyer, Sir Norman, and Locxyer, Dr. W. J. 8S. Roy. Soe. Proc., vol. 71, 1902, pp. 134-135. Warp, H. Marshall.—Experiments or the Hffect of Mineral Starvation on ine Parasitism of the Uredine Fungus Puceinia dispersa on Species of Bromus. Roy. Soc. Proc., vol. 71, 1902, pp. 188-151. Mineral Starvation, Effects of on Parasitism of Puccinia, Warp, H. Marshall. Roy. Soc. Proe., vol. 71, 1902, pp. 138-151. Puccinia dispersa, Wffects of Mineral Starvation on Parasitism of. Warp, H. Marshall. Roy. Soc. Proc., vol. 71, 1902, pp. 138-151. Parasitism, Effects of Mineral Starvation on, &c. Warp, H. Marshall. Roy. Soc. Proc., vol. 71, 1902, pp. 138-151. .sidinoe ¥ bas slorisY to gideaoitslet-1ial of T—.aotdonoM .@ B veaneas09 6§1-ISI .qq SOCF LY Lov ..co1l .908% .you . aietioog V odnt soitscriotenerT Istusmineqx ih —slorie .cottosoM .2 .“aMaio9 S8I-ISI .qq S0@L IV for .sotf 508 .yoH sloireY cori a Sab Bl cce 8 .cotdo nol .& MAMAIOD at .661-I8f a « SOCL LY doy ..001L..908 sa af eisioosV bas seit role - wohisnol .4 wamago0 éE{-I${ .qgq S081 {FT lov .sotf .90% .yoH to irene od? cO—2 LW ad satavaood bas ,cearo ue pe 28th ooral revo coftsire YT onssecord boiteq-d1od@ oat BEL-el qq S0@f IY for ,so01L 08 . yo anoiash soared revo ytitelioni@—boired diode ,enoitsraY ointemowd 2.4.0 «aC .asyuood bus .aemr0o 12 aayaood GEI-BEL .qq SOL LV fov .s0ff 908 .yoH So Seerret 2 ter 30 sae ocd] re etagerreoces Mt _. edocs te ae : < to 29199q@ 10 neteqesh Heasoowh ji gnibsr "dat to metticersD adit aor a | EBI-88f .qq £00L LY Lov ..corf .908 .yoH ojeeawS to ceitisteT co to edooR .coitetiet@ ferent j UedeveM A agaWw AIG{-88L ..qq SORL LY lov .sord 908 .yoH jo mettested so comaviedé leveniM to atoohH wersgqssh niwtboow JigdersM .H .gaW 46i-BEI gq SOV LY lov .vorL 0% .yoH 0% 10 coitav wie leteniM Yo stool .eitieated AlsdeteM He .saw 31-661 .qq SO0@L JY tov .sotd 008% .yoH y ne a 1902. | Starvation on the Parasitism of Puccinia. 143 psammophilous plants as the Bromes of the minute quantities of these elements needed for their growth ; the root-hairs appear to be able to Treated as follows:—A distilled water; B minus phosphorus; C minus calcium; D minus magnesium; H minus nitrogen; F’ minus potassium; G normal solution ; O horse-dung decoction, shoot after 44 days. Fic, 2.—Plants of Experiment 47 removed from the sand and carefully washed, showing the development of root and obtain traces of these substances from the grains of sand, and it would be necessary to repeat the experiments with some other medium— perhaps precipitated silica or pure quartz—to obtain through starva- 144 Prof. H. Marshall Ward. Zffect of Mineral [Nov. 4, tion of these elements, a result not wished for in the present con- nection. ; Dminus magnesium; G normal , and then carefully washed and photographed with yy WY YY Zi a BY Y “ Up DL, Wi WLMILYL wy DI, \ roots on black paper. Treated as follows :—A distilled water; B minus phosphorus Fic. 3.—Seedlings of Bromus secalinus 44 days after sowing in sand solution; L minus nitrogen; O horse-dung decoction. But although the differences in size were not obvious in these cases, the magnesium-starved seedlings exhibited some effects which can only have been due to lack of supplies, viz.: the tips of the leaves (aa ; : / .” i aod . ot f P ¥ \ k £ < ¥ i F » : . , ; .s) . 3 ¢ Rae | : : e® cers, to which do: \ ~ w Average number of leaves. 2 narrow 25 narrow 3 broad 23 medium 2—3 narrow 3 medium 831 broad 24 narrow 3 broad 25 rather narrow 2—3 narrow 3 medium 3 broad 3 broad but in different pro) Fl Fl 1902.] Starvation on the Parasitism of Puccinia. 145 were apt to show early yellowing and withering, as well as some degree of stunting. When we come to examine the results of infection, it is clear that no great differences were observable between the various seedlings. The first flecks—+.e., pale patches indicating the presence of the fungus in the tissues—were visible on July 31 and August 1, that is to say, on the eighth or ninth day after infection, and by August 3 pustules were developed on all the plants except those starved of phosphorus, while two days later even these were definitely pustuled in one of the two beakers (h, Table I), though the other remained free of them to the end (b, Table I). As will be seen from further experiments, how- ever, we can lay no stress on this latter case, and must conclude that :—Lack of minerals in no way secured immunity from infection, though seedlings deficient in phosphorus or in nitrogen tended to show retardation of infection. When we look into the matter from another point of view, however, the conclusion seems inevitable that there are considerable differences in the quantity of fungus mycelium, and consequently of Uredo-spores, developed from the latter, which the various seedlings were able to support. Taking the extreme cases, and comparing the sizes of the infected areas on the phosphorus-free plants (h, Table 1) and those on the plants supplied with normal mineral solution (g and n, Table I), the latter were found to be much the larger, and similarly with the pustules themselves: the numerous large well-developed pustules on g and 2 bore many hundred times as many spores as did the few small pustules on f, and similar results were observable in less marked degrees on others. It must therefore be concluded that:—Mineral starvation makes ttself felt quantitatively in the number of uredospores which can be produced by the fungus in the tissues of the starved leaves. Now arose an important question :—Are the spores on the small starved pustules in any way different from those produced by the large well-developed pustules of normal or richly-manured plants ? For instance, can any differences in size, colour, or marking, or in capacity for germination, be detected in spores from the small and few pustules of a, ¢, h, or J (Table I), and those from the large well- developed pustules of g, n, or o ? As regards the morphological features, direct examination showed that no differences were to be detected ; the properly ripened spores in all cases were normal. As regards their capacity for germination, I proved by sowing them in separate watch-glasses of distilled water that in every case the pus- tules yielded spores capable of normal gernunation, and that in proportions which showed no relation to the degree or kind of starvation of the seedliny which bore them (see Table II, col. 4). VOL. LXXI. M Table I.—Exp. 47. Seedlings of Bromus secalinus, sown in coarse cleaned Sand in Beakers, to which doses of Nutritive Salts were added on July 7, 1902. | Exp. No. 47, a 4 c* d afi | =< Treatment. Number of | seedlings. Distilled oe eines 7 oy Minus phosphorus .... 4 een ce Goren 5 lira magnesium .... 5 | » Nitrogen .... ... 5 | 5, potassium ..... 5 Normal ....0cessss05> 4, P-free + Fe..... ‘ 7 Oactrannde HO caer Y/ Mg-free + I'e........ 6 N-free + Fe ....-.... 6 Metrae-F Wines seen / Normal + Ife ........ 5 Tlorse-dung .....+20- 6 Number inoculated. o 4 Condition of plants. Stature. 10—12 em. 1216 ,, 20—22 ,, One = 9 em., others = 16—18 cm. Two Two = 14—15 One 1 cm. 16—18 em. ot dl 19—21 1216 20-22) ,, 16-18 ,, LOW 16160 19= Ole 16—17 ,, cm. 5— 6 cm. » Robustness. Small Medium Stout Slight Slight Stout Stout Medium Stout Slight Small Stout Large Stout and coarse. Colour. Rather pale Slightly pale Good green Fair, but tips of leaves dying off. Pale Good green Bright green Slightly pale Vivid green Good green but tips dying off Good green Good green Deep green Good green Average number of leaves. narrow 3 broad 25 medium medium 3h broad 25 narrow 3 broad 2% rather narrow 2—3 narrow 3 medium 3 broad 3 broad Results on July 31. Flecks ? Flecks ? Flecks ? Flecks ? Flecks ? Flecks ? Results on August 1. Flecks ? One first leaf and three second leayes have pustules One first and one second leaf have pustules 0 Two first leaves and four second leaves pustuled Flecks ? Flecks ? Flecks ? Two first leaves pus- tuled Three first leaves and one second leaf pustuled Results on Number Number infected on | infected on Ist leaf. 2nd leaf. eee ae a 0 0 0 1 1 2 1 0 4 4 3 3 3 3 At 5 2 3 3 3 2 3 2 3 5 6 August 3. Number of seedlings infected. One second leaf 1 shows pustules 0 0 One second leaf 1 shows pustules One first and two 2 second leaves show pustules 1 Two first and three 4 second leaves pus- tuled Three first and two 3 second leaves show pustules 4 Three first and five 5 second leaves show pustules Two second leaves 3 pustuled Three first and three 3 second leaves pus- tuled One first and three 4 second leaves pus- tuled Three first leaves 3 show pustules Five first and three 6 second leaves have pustules Results on August 5. * On July 18 a dose of the K-free solution was given to this by mistake, consequently it received all salts, but in different proportion from the normal. { First leaf dead in one plant also. | Number | infected on both Ist and 2nd leaf. on j On July 23, infected Ist and 2nd Leaves with Spores from Lromus squarrosus. Remarks as to condition at end of experiment.t Two or three small patches of small pustules, yielding fair spores. Leaf pale in infected area, No trace anywhere of pustules. Only one small group of small pustules, rather better than in a. Leaves green. Small but fairly well-developed pustules of spores on one plant; only one pustle on the other. Leaves very pale and tips dead. One very minute pustule on pale first leaf. Extensive infection, and several patches show strong pustules, Leaves rich green. Excellent crop of spores on large pustules, and leaves rich green. Minute pustules, difficult to detect and bearing but few spores. Colour fair. Very good pustules with abundant spores, but few and small areas or patches of infection, Leaves good green. Medium-sized patches of small pustules with not many spores. Pale, and tips dead. Attacked area pale. Pustules small and few, and yield few spores. Few but medium-sized pustules. One leaf has a large patch. Green. Very large areas of well-developed pustules, rich in spores. Leaves deep green. All pustules large and rich in spores, but areas of leaf around pale; otherwise deep green. + On August 23 spores were taken from these plants and infections made on test-tube cultures of Bromus secalinus (see Table Il). i = 146 Prof. H. Marshall Ward. fect of Mineral [Nov. 4, It now remained to test the further ques- tion :—Are such spores equally capable of infecting other seedlings ? In order to test the capacity for infection, or virulence, of the spores developed on the experimental plants, I employed tube-cul- tures of seedlings as described in a previous communication,* with a slight modification in detail, as follows (see fig. 4) :— Picked grains of Bromus secalinus were sown singly in test-tubes of sand, to which the normal mineral solution was added, and sterilised. When the seedlings were well up, and nine days old, on August 14, each having one good green leaf, a sowing was made on each from the uredo pustules on one of the plants of Experiment 47 (Table I), as shown in the annexed tabular summary (p. 147). It will be seen that in every case the results of infection were positive. ‘Those spores which had been reared on seedlings starved of phosphorus (/), and on those starved of all salts (a), were behind the others in their rapidity of effect; but we cannot lay much stress on this, because the same was true of spores reared on what should have been the calcium-starved plants (c), which accidentally received a dose of calcium by mistake. On the other hand, the most vigorous spores were those reared on plants to which horse-dung decoction had been added. Iam not disposed to lay stress on these latter points, however, but am content at present to regard the positive results as amounting to proof that spores even when reared on starved seedlings are capable of normal gernunation and infection when placed on the leaves of other—normal—seedlings. A second experimental series of sand- cultures was started on July 10, 1902, in sixteen beakers, arranged as before, but with the following differences in detail. The sand used was a finer and whiter silver sand, * “On Pure Cultures of a Uredine.” ‘Proc. Roy. Soc.,’ vol. 59, p. 451. = . ae 4 ey Fs .' 78 ne oy Shania ee B oF — nate ge rn te rae nee ge em ane oe ye a oe lene ee et Day capt wereane ny torial E porta derosiyr ; Ae nae cae x ; | Re Re ces, J 7 q Lo A wah elle meen, id Ty hy ries 3% usher eee eae ~ * ; “i cm ea \hp minh oRAAN saab BGA ¥ = iecie : + Yes ‘ 5 a ar BPS fetes ae A= 2 f a iF ' y 5 ' i f Lee ie Pe ly ta oe A ieeer emis aes I 1902.] Starvation on the Parasitism of Pucconia. 149 Control cultures have shown over and over again that these seedlings do not incur the disease unless the spores are placed on the leaves, and I cannot accept these one or two apparent exceptions as evidence for a spontaneous outbreak of the disease. The severity of the attack, as indicated by the size of the pustules and of the infected areas, and by the relative number of spores developed, followed much the same order as in the previous experi- ment, except that the nitrogen-free plants seemed to bear larger pustules than before. The minute and poorly-developed pustules on the seedlings starved of phosphorus (/ and m, Table III) or of all minerals (2) were par- ticularly evident; as also were the severe attacks of the manured specimens. And, again, the magnesium-free seedlings showed pre- mature yellowing and withering of the leaf-tips. Again, also, there was decided retardation of infection pustules in the phosphorus-free, distilled water, and, to a less extent the magnesium-free specimens. But—again concordant—none of the plants were rendered immune from infection; and, as will be séen in the sequel, all the pustules, however minute, yielded spores with normal morphological characters, and perfectly capable of germinating and of reinfecting other seedlings of the Bromus in question. Meanwhile, I had started a third complete series of sand-cultures of Bromus pendulinus, arranged as in the last series, but with two small variations in detail, viz., (1) in order to diminsh the risk of drowning or asphyxiating the seeds, I sunk a tall thistle-head funnel to the bottom of each beaker of sand, and poured the solution in each case down to the bottom of the beaker, allowing it to soak upwards and gradually moisten the upper layers from beneath; and (2) only 5 grains of the Brome (b. pendulinus) were sown in each beaker. The results were very good. The grains were sown July 23. On July 29 seedlings were up in all but three of the beakers. On August 1 these also had seedlings showing, and by August 5 most of the seed- . lings were 25—30 mm. or more high. On August 9 all but three of the beakers had four or five excellent seedlings, each with the second leaves showing, and the effects of the mineral starvation were beginning to be visible. In order to emphasise this effect, if possible, by preventing the further development of the second and later leaves at the expense of materials stored in the jist leaves, I now cut off the latter from each seedling with a pair of sharp fine scissors, and left the seedlings to grow further. During the next week the second leaf in each case attained its full size, and the third leaf began to appear ; and I now proceeded to devote this series of seedlings to test the effect of sowing the spores reared on the seedlings of the last series (see Table III), and which had been affected as to. their numbers by the starving or other 150 _ Prof. H. Marshall Ward. fect of Mineral [Nov. 4, treatment of the seedlings, on the leaves of the seedlings of this series, similarly starved or otherwise. : As we have seen (Experiment series No. 47, Table I, and series 60, Table II), spores from a seedling starved of phosphorus, however few in number, are perfectly capable of infecting the leaf of a normal seed- ling. Can such spores also infect another seedling s¢milarly starved of phosphorus ; and can spores reared on calcium-starved or nitrogen- starved seedlings, &ec., infect seedlings similarly starved of calcwwm or of nitrogen respectively, and so on 2 On August 18, therefore, I infected the third leaf of each of several seedlings in each beaker of series 55 (Table IV) with the spores developed on the corresponding seedlings (which had been similarly treated) of series 50 (Table III). The results are summarised in Table IV. As the table shows, practically all the infections were successful, showing that not only does mineral starvation not prevent the develop- ment of virulent spores on the seedling so starved—if the latter is inoculated with normal spores—but such starvation is also incapable of oncapacitating the corresponding seedling for infection by means of spores grown on semilarly starved seedlings. We must therefore conclude for the present that (1) the starvation of mineral food-substances, although it reduces the size of the host- plant and seriously diminishes the quantity of spores which the myce- lium can give rise to on its leaves, does not affect either the wrulence of such spores or the predisposition to infection of the leaves of the Brome concerned. Moreover (2), in view of the results with the highly-manured seed- lings to which horse-dung decoction or normal mineral solution was added, it seems hopeless to expect that high cultivation of this kind will diminish the predisposition of the plant to infection—or, what would amount to the same thing in practice—increase its resistance or confer emmunaty, The effects of manurial treatment are clearly quantitative only, so far as this question is concerned. If the host-plant is highly fed, its tissues yield more food materials for the fungus; the latter can develope a larger mycelium, and pro- duce a larger crop of spores. But so long as the host-plant is capable of living at all, it is a perfectly satisfactory prey for the fungus in its tissues, so far as quality of fungus food is concerned. | It seems to me that these results throw some new light on the prob lem of infection and parasitism, in so far as they bear out the view that the Uredine mycelium taxes the leaf—robs it of a share of its food-supplies—rather than destroys the protoplasmic machinery, at any rate during the vigorous period of growth and of production of Uredo- spores; and also in so far as they suggest that whatever may be the No. | ss No. of be plates with | pustules. ERS is ss | 3 Two 4 | deve Eee as | 1 | Very s Clleaves ......| 4. | Patche oe 5) Patch 46 t } Patch 5 | Patch 4 | Patch 4 | Patch | 1 || Large resiieisae =: > 2 |, Patche cease 4 | Two p EEN Da sic. 2 | One pe ys a ee 2 || Very s BURR as 3: 4, | Patche 3 | Patche pusti : 2 | Extren August 18 with Spores of Exp Condition on Septemb eld No. 55. m q Table IV.—Exp. 55. Seedlings of B. pendulinus in Beakers of Fine Sand, sown July 23. Infected the 3rd Leaf on August 18 with Spores of Exp. 50 (B. pendulinus). Condition on September Ist. Nowatal NENOsine Source of Germina- ae =F Treatment. plates. | fected. sntceunng tion of Condition on Augu 28. nate No. of : BROLER: EVO, 2 Stature, &e. plates with Sizes, &e., of pustules and patches. plants. pustules. Distilled water only........-- 5 2 50, @ Good ......| Small, but good infection ............ 5 Small, with 2—8 leaves............ 3 Se shea and two rather larger, with well = + Fe. 4. 1 50, a us Very small infection...............+. 4 Small, 3—3% leaves...:............ 1 Very small patch of very small pustules. Ca-free .... BgOMOOO 4 4 50, ¢ Excellent ..| Mach larger infection: good spores.... 4 Medium, 4—4% broader leaves ...... 4 || Patches and pustules large and well developed. Sees viene pangon b) 5 50, d Very good..| Similar, but younger infection......... ) Medium, 4—5 leaves .............. 5 Patches large and pustules fine. Mg-free . 5 4 50, e Poor ......| Between 6 and c indegyee of infection. . bs Large, d—5 leaves. ...eseeeseeeeess 4 Patches and pustules but little inferior to those of d. jy BELO Sa cougqoecavads b) 5 50} 77 FHair.......)| Dike c: good infection............... b) Large, 4—5 leaves...............4% b) Patches and pustules large und well developed. N-free..... 5) 5 50, 9 Very good..} Like d, but very small infection........ 4 Small, 4 leaves... 0.0.05. ..0.-.+5 4 Patches small, but pustules fine and large. i; b) 4 50, h Wair.......| Tike 6, very small infection .......... 5 Small, 33—4 leaves..........++.05- 4 Patches fair, and pustules large. Normal .... 4 4 50, 7 » +++e...| Small infection, but advanced........ 2 Medium, 4 leayes.............+:. “i 1 Large and well-developed patches and pustules. 4, SED eta eoandoon ss 3 3 50, & | Poor.......| Small infection: spores only just ap- 3 Fairly large, 4—44 leaves.....,..... 2 Patches small, but pustules fine. pearing ICED ge crennminm ooo weeT Oo 5} 4 50, p Very good..| Fairly good infection.................) 5 Large, 4—43 leaves..........+1+00. 4 Two patches large, two small: all pustules well developed. Sree chumlilevwicmrsere ss: 5 3 50, 7 Hxcellent#. | Small infection.................-+-.- b) Large, 4—5 leaves. .......0.. 0005 2 One patch small, one fairly large : all pustules well developed. ERG YROE nape atten aerearairs. = sisasi s/s 3 2 50, 2 Good ...... Minute traces only of successful infec- | 3 Small, 3—3% leaves.... 2 Very small patches and minute pustules. tions | » + Fe 5 4 50, 0 TEN Goon cn IDO) cone no0s00v0 on conc en GdOdHoADNS b) Small, 3—83 leayes.............-+. 4. Patches and pustules small, but well developed. | P-free..... 5 3 50, J Excellent ..| Very minute traces of infection........ 5 Medium, 4 leayes.......... 3 Patches extremely small, with poorly developed minute A ae de 5 2 50, m Iles Ombky VARNA 500000 coun on bo neu 5 Medium, 33—4% leaves.,.......... e) | ee es and poorly developed patches and pustules. Very poor.. ioe ripe 4 dae native a i ae See ee aie ee os a er eee ee “ai i ' I i - 1% aR ey Ba) BU! 1902.] Starvation on the Parasitism of Puccinia. Lol causes at work in the living cell which confer immunity or predisposition on the species of host-plant, or which confer virulence or impotence on the spore, they lie deeper than nutrition, reminding us once more of the significant resemblances which, as I pointed out in a previous paper,* exist between the phenomena of infection and those of polli- nation. It is in the highest degree improbable that the pollen-tube of a given species, A, is incapable of growth in the style and the ovule of an allied species, B, simply because the tissues of B do not contain suitable food-materials, while the pollen-tube of a species, C, readily fertilises a more distant species, D, simply because the latter does con- tain suitable nutritive materials, especially as in both cases we may be able to germinate such pollen in artificial sugar-solutions. All the evidence points to the existence, in the cells of the fungus, of enzymes or toxins, or both, and in the cells of the host-plant of anti-toxins or similar substances, as the decisive factors in infection or immunity, although I have as yet failed to isolate any such bodies. Moreover, I regard the results here given as furnishing strong evi- dence, on the whole, against any hypothesis which assumes the exist- ence of a latent or lurking source of disease in the plants themselves, and as supporting the view that every patch of pustules originates from a definite infection spot due to the entrance of a germ-tube from a spore which has there germinated on the leaf. * ©Proc. Camb. Phil. Soc.,’ vol. 11, part 5, p. 307. VOL. LXXI. N 152 On the Variation of the Critical Velocity of Water. [July 16, “ An Experimental Determination of the Variation of the Critical Velocity of Water with Temperature.” By E. G. Coxkkr, M.A. (Cantab.), D.Se. (Edin.), Assistant Professor of Civil Engineering, and 8. B. CLEMENT, B.Sc., Demonstrator of Civil Engineering, both of McGill University, Montreal. Commu- nicated by Professor OSBORNE REYNOLDS, F.R.S. Received July 16,—Read November 20, 1902. (Abstract.) The change from stream-line to eddy motion in water was first xamined by Osborne Reynolds, who in his earlier experiments, intro- duced colour-bands into a glass pipe in which water was flowing, to indicate the change in the motion, and later observed the resistance encountered in pipes over a great range of velocities. The results of these experiments, and a consideration of the equations of motion, enabled him to express the laws relating to the critical velocity of water in pipes by the exceedingly simple equation where 2 is the critical velocity of water, 7 is the radius of the pipe, p is the viscosity of the water, p is the density, and &£ is some constant. In the original experiments the range of temperature was very limited, and it was pointed out that “it would be desirable to make experiments at higher temperature, but there were great difficulties about this, which caused me, at ali events for the time, to defer the attempt.” It does not appear that such experiments have since been made, and although the difficulties were great, it was resolved to test the law through a much greater range than had hitherto been attempted. Preliminary experiments showed that at temperatures beyond 50° C. the losses due to conduction and radiation were large, and that elaborate arrangements would be required to obtain reliable results. It was therefore decided to determine the variation of the critical velocity over a range extending from about 4° C. to 50° C., which seemed to be sufficient as a test of the law, and, with ordinary precautions, only necessitated small corrections for the effects of conduction and radia- tion. The resistance method used by Osborne Reynolds in his later experiments was employed, the arrangement only differing from his in details, such as the use of pressure chambers giving a continuous opening at the ends of the 2-inch pipe examined, the employment of an inverted U-tube for measuring the pressures, and the weighing of the discharge. 1902. ] Jsomerte Change in Benzene Derivatives. 153 Numerous experiments at different temperatures were made when stream-line motion was maintained in the pipe, and the relation of velocity to slope of pressure was determined by logarithmic plotting, giving a series of lines, the “logarithmic homologues” at different tem- peratures. The positions of these lines were found to be in substantial agreement with those calculated from the equations of motion. Similar experiments for eddy motion were made and the logarithmic homo- logues were also plotted, and their intersections with the corresponding ones for stream-line motion determined. These intersections give the minimum critical velocity, and were found to lie very approximately on a straight line in the diagram. The law of variation of critical velocity, v., with temperature was found to be ie |x 14+0°03368T + 0-000156T2, where T is the temperature Centigrade, which agrees very closely with the known variation in the viscosity of water, viz., we 12 1+0:03368T + 0:000221T? ; and it may, therefore, be concluded that over the range of temperature examined the critical velocity of water in small pipes varies directly as the viscosity. “Tsomeric Change in Benzene Derivatives —The Interchange of Halogen and Hydroxyl in Benzenediazonium Hydroxides.” By K. J. P. Orton, Ph.D., M.A., St. John’s College, Cain- bridge, Demonstrator of Chemistry, St. Bartholomew's Hospital. Communicated by Professor H. E. ARMSTRONG, F.R.S. Received December 1,—Read December 4, 1902. In discussing the laws which govern substitution in the case of benzenoid compounds, Armstrong, in 1887, drew special attention to the peculiar behaviour of amido- and hydroxy-compounds, from which he inferred that the phenomena of substitution were less simple than was commonly supposed. He showed that there was evidence that the formation of para-derivatives was preceded by that of an isomeric compound formed by the displacement of the aminic hydrogen or hydroxylic hydrogen, and pointed to the probability that this might prove to be true also of ortho-compounds. Since that time, it has been experimentally demonstrated by various chemists that the radicles, Cl, Br, I, NO2, SO3H, can all be introduced in place of the hydrogen of the amino-group of anilines and of the imino-group of anilides, and that the compounds thus formed can be N 2 154 Dr Keg Py Orton: [Dee. 1, changed into isomeric substances in which the substituent is contained in the hydrocarbon nucleus. The transformation of the aniline derivative appears to take place only in the presence of some other substance: in the case of the phenylacylchloramines, such as CsH;.NCI.COCH., for example, appa- rently the change into the isomeric chloroacetanilides takes place only under the specific influence of hydrogen chloride.* The investigation of such cases of isomeric change is, in fact, of special interest as they are, so to speak, “‘ fermentative” in character, often taking place with remarkable facility, and under the influence of minute amounts of the substance, which apparently provokes the change—the catalyst. Measurements of the velocity at which changes of this type occur, for example, of diazoaminobenzene into aminoazobenzene,j show that a velocity-coefficient of constant value is given by the equation, & = t 1 log a/(a@—2). The isomeric change is, therefore, apparently a so-called mono- molecular change; but, as in other cases, the slowest of the series of simple changes which make up the complete transformation is alone measured ; moreover, the substance which conditions the change is not taken into account. This paper deals with a new case of intramolecular change of a particularly interesting character: that of s-trichloro- and s-tribromo- benzenediazonium hydroxide, CyH2X3.N(OH):N, into hydroxybenzene derivatives, by the interchange of the hydroxyl for one of the ortho- halogen atoms. The change takes place under all conditions under which it is possible for the diazonium hydroxide to be present. Thus in dilute aqueous solutions of the neutral diazonium nitrate, or even of the hydrogen sulphate, chloride is just recognisable by means of silver nitrate after 24 hours; at the same time the solution becomes yellow, owing to the formation of the diazophenol. But as the extent to which the nitrate, and more especially the sulphate, undergoes hydrolytic dissociation is extremely small, the isomeric change takes place very slowly ; and in the presence of a considerable excess of acid no appreciable change occurs during 5 days. On the other hand, using diazonium acetate, 50 per cent. of the material changes within 30—40 hours; and in the case of diazonium bicar- bonate the change is nearly instantaneous. That the change is a transformation of the diazonium hydroxide is further emphasised by the fact that the addition of a solution of the diazonium salt to a considerable excess of an aqueous solution of an alkali carbonate, is not followed by elimination of halogen; the solution remains colour * Compare Armstrong, ‘British Association Report,’ 1899, p. 683; ‘Trans. Chem. Soe.,’ 1900, vol. 77, p. 1053. + Goldschmidt and Reinders, ‘ Berichte,’ 1896, vol. 29, p. 1369. 1902. ] Isomerie Change in Benzene Derivatives. 155 less, the diazonium salt being immediately converted into the alkali diazotate.* The latter change may be represented thus— CyH2X3.NR:N — CsH2X3.N(OH):N — CgHeX3.N:N-OM’. Diazonium salt. Diazonium hydroxide. Diazotate. As the author has maintained on previous occasions, the first step in this type of intramolecular change consists in a transference of the atom or group attached to the nitrogen to the ortho- or para-carbon atom of the nucleus, an ortho- or para-quinone being formed. The complete transformation in the case here described may be represented in the following manner :— N:NOH © N:N N:N.Cl N=N No % o9 ‘ | ° : | 72 Cl” Nat | ay ae | Ce Son cir No) cl =0 | | ag | | | > either | | -~ | | ; or | Vests Ww y | \4 Ty, SF i i | u Cl “a A ie ee Die Ve Ve In other words, the diazonium chloride (IIL) may be represented as changing into the “diazophenol” (IV); or, if, following Wolfff and Hantzsch,{ the diazophenols are considered to be diazoquinones (V), the ortho-quinone form (II) merely loses hydrogen chloride. Closely allied to the intramolecular change just mentioned, are Hantzsch’s observations§ that s-tribromobenzenediazonium chloride changes into chlorodibromobenzenediazonium bromide, and chloro- and bromo-benzenediazonium thiocyanates into thiocyanobenzene- diazonium chlorides and bromides. [The following observations of Meldola|| may also be mentioned, as the author’s results offer a possible explanation. When dinitro-o- or dinitro-p-anisidine is treated with sodium nitrite in the presence of acetic acid, the nitro-group occupying a position ortho or para with respect to the amino-group is eliminated, a diazophenol (diazo- * Professor Meldola has called the author’s attention to a French patent (No. 315,932) of the Badische Aniline Company, dated February 28, 1902, in which the replacement of the nitro-group and halogen by the hydroxyl-group by the action of alkalis on diazonium salts, is claimed as a technical process. The author’s experiments show within what limits this process is applicable, and what is the probable action of the alkali. ft ‘Annalen,’ 1900, vol. 312, p. 119. t ‘Berichte,’ 1902, vol. 35, p. 888. § ‘ Berichte,’ 1896, vol. 29, p. 947 ; 1897, vol. 30, p. 2334; 1898, vol. 31, p. 1253; 1900, vol. 33, p. 508. || ‘ Trans. Chem. Soc.,’ 1900, vol. 77, p. 1172; ‘ Proc. Chem. Soc.,’ 1901, vol. 17, p. 131; ‘Trans. Chem. Soc.,’ 1901, vol. 79, p. 1076; 1902, vol. 81, p. 988. 156 Dre ky ds Py Orton: [ Dec. 1, oxide) being formed. In diazotising in the presence of hydrogen chloride, an ortho-nitro group is replaced by chlorine. | The isomeric changes of the diazonium salts differ, however, some- what from the similar changes of the phenylacylchloramines, phenyl- nitramines, &c. The latter compounds are per se relatively stable, and apparently only undergo change in the presence of some agent (catalyst) ; the diazonium compounds, on the other hand, appear to be intrinsically labile, and in aqueous solution, at least, capable of passing into a more stable configuration. In the latter case the nitrogen atom, bearing the wandering group, the hydroxyl group (or, as in Hantzsch’s instances, the chlorine atom or the thiocyanate group), is pentad, thus, Ph.N(OH):N, whilst in the phenylacylchloramines, &c., it is triad. The idea at once suggests itself that in the case of the last- mentioned compounds, the first action of the catylist is to form an additive product, in which the nitrogen is pentad. The product, in which, it should be noted, more than one negative radicle is attached to the nitrogen, is not a stable compound, and is now capable of passing into the isomeric quinone form, and thus start the transforma- tion. These results are also of interest, inasmuch as Hantzsch* has stated recently that he has obtained s-tribromophenylnitrosamine, C H.Br3.NH.NO, by adding sodium acetate to a solution of a s-tri- bromobenzenediazonium salt. He describes it as a bright yellow amorphous substance, which decomposes at 85°. As far as can be judged, it is this very reaction which has been studied in the course of the author’s experiments. The substance which is precipitated is at first glance a yellow amorphous powder, but close observation shows that long ($—1 cm.) orange crystals are present. These crystals are the 3:5-dibromo-o-diazophenol hereafter described. The powder is probably a hydroxyazo-condensation product. Hantzsch does not appear to have observed that bromine is eliminated. He also affirms (loc. cit.) that he obtained the nitrosamine by the cautious addition of acetic (or other) acid to the alkali diazotate ; in the writer’s experience, however, this always leads to the elimination of halogen. The Transformation of s-Trichlorobenzenediazonium Hydroxide. s-Trichlorobenzenediazomum hydrogen sulphate, Cg>H2Cls.N(SO.H):N, is very easily prepared by diazotising, by means of amyl nitrite, s-trichloraniline dissolved (or suspended) in glacial acetic acid contain- ing sulphuric acid. The salt is precipitated by ether, and is purified by dissolving in methyl alcohol; on adding ether to this solution the sulphate separates in small colourless lustrous prisms, often forming * © Berichte,’ 1902, vol. 35, p. 2964. 1902. | Isomeric Change wr Benzene Derwatives. 157 star-shaped aggregates, which are very soluble in water (SO, found 31:0, calculated 31:4 per cent.). s-Zrichlorobenzenediazonium nitrate, C,H2Cl3:N (NO3):N, prepared in a similar manner, forms colourless needles. When kept for two or three days, both salts begin to show signs of change, which are more marked in the case of the nitrate. The latter becomes noticeably yellow, and when dissolved in water forms a yellow solution. The initially colourless aqueous solution of the pure colourless salts becomes yellow after afew hours ; after 24 hours chloride is just recognisable in the solution. In one experiment, 0°5 gramme of the acid sulphate was dissolved in 100 ¢.¢. of water; after 3 days a very small amount of a yellowish-red solid had separated from the yellow solution ; at the end of 16 days, the chloride in solution was precipitated by silver nitrate in the presence of nitric acid: the silver chloride weighed 0:05 gramme, whereas for the complete conversion of the diazonium salt into the diazophenol, 0°235 gramme of silver chloride should have been found. An aqueous solution of s-trichlorobenzenediazonium acetate, obtained by mixing neutral solutions of the diazonium nitrate and sodium acetate (molecular proportions) rapidly becomes yellow and acid. In a short time the solution becomes turbid, and after 4—5 hours deposits a bright yellow amorphous solid. In an experiment, in which 1°75 gramme of the diazonium nitrate, dissolved in 150 ¢.c. of water, was treated with sodium acetate (1 mol.), and kept at 10—15° for 40 hours in the dark, an estimation of the hydrogen chloride in the solution showed that 54:5 per cent. of the diazonium compound had changed into the diazophenol. When instead of sodium acetate, sodium hydrogen carbonate was used a similar change took place, but far more rapidly. In one experiment a dilute aqueous solution of sodium hydrogen carbonate (3 mol.) was added drop by drop during a period of 1 hour to a cooled solution of 1 gramme of the diazonium hydrogen sulphate. (The bicarbonate was finally present in sufficient quantity to combine with the whole of the sulphuric acid and one-third of the chlorine present in the diazonium salt.) After one-third of the bicarbonate had been added, and the acid converted into the normal sulphate, the solution rapidly became yellow and deposited a yellow solid. Throughout the experiment the mixture was acid. As soon as the whole of the bicarbonate was added, the chlorine was estimated; it represented 54°5 per cent. of the amount which should be obtained were 1 atom of chlorine eliminated from the diazonium salt. In another experiment, using the same quantities of diazonium salt and sodium bicarbonate, the solutions were mixed as rapidly as possible; a copious yellow precipitate at once appeared ; the chloride in the filtrate represented 72 per cent. of | atomic proportion of chlorine. 158 Dr. K. J. P. Orton. [Dee. 1, | DO ee af\=0 , 3: 5-Dichloro-o-diozophenol (3 : 5-dichloro-o-diazoquinone), — | | LAVAS \Y Cl contained in the yellow solutions which are obtained by any of the methods just described. It is best prepared by adding excess of sodium acetate to s-trichlorobenzenediazonium hydrogen sulphate or nitrate. The mixture should be kept during 40—50 hours in the dark ;—in the hght the solution darkens, the diazophenol decomposing. The liquid was then filtered from the amorphous yellow solid, made strongly acid with nitric acid, and extracted four or five times with ether. On evaporating the yellowish-brown extract, a mixture of oil and crystals remained. In order to obtain the diazophenol, dry hydrogen chloride was passed into the ethereal extract ; this caused the hydrochloride of the dichloro-o-diazophenol, O'C;H2Clo:N2,HCl, to separate in small needles. This salt was converted into the diazo- phenol by treatment with a small quantity of water. Thus prepared, the diazophenol is an orange powder, which crystallises from an ethereal solution in flattened orange prisms, which melt at 83—84’, forming a red liquid; at 87° the latter decomposes. On analysis, 0°1068 gramme gave 0°1618 gramme AgCl. Cl = 37-45. 0:2004 gramme gave 25-4 ¢.c. of moist nitrogen at 14° and 766 mm. N = 14°95. C,HsONCl, requires Cl = 37-57; and N = 14°86 per cent. This substance is very readily soluble in all solvents except petro- leum; when dissolved in hot water it rapidly decomposes, the yellow solution becoming brown and turbid. It dissolves in concentrated solutions of acids, forming a nearly colourless liquid, which becomes yellow on adding water. These acid solutions couple with alkaline solutions of B-naphthol. There is no doubt that this substance is an. ortho- and not a para-diazophenol, as p-diazophenols can be easily recrystallised from hot water, whilst the ortho-compounds are decom- posed by hot water. Again, p-diazophenols are easily reduced to p-aminophenols, but the ortho-derivatives are not reduced in a simple manner.* The amorphous yellow substances which are formed in the cases above described appear to be condensation products. The o-diazo- phenol which is produced has the para- and one ortho-position unoccupied, and being formed in the presence of a diazonium salt and sodium acetate or bicarbonate, is under conditions which are very’ favourable to the production of hydroxyazo-derivatives. ~ Ss uN Re tT N | + + te Ve bee eee +-00/ ! | | j i | | | : . 1940. | O 78) 20 JZOMITS. 1902. | “ Blaze-currents” of the Incubated Hen’s Kgg. 189 | Chick, 108 hours. Pressure. | Capacity. Quantity. Energy. | Response. | 6 L. ecole mt. 8:4mc. | 360 ergs. + 0°0005 | us | aie He | Ay — =0°0016 | ; He 2p aad ae 16°8 | 720 +0°0009 | i | ye yy amelie: —0°0025 | oF | '3mf, + 25 °2 ee LOSO / +0°0011 y fe i | e | =0-0032 . |) Qasr 33 *6 | 1440 +0 0015 f | ee ie | | 00040 5 | Bomife = 42:0 | . 1800 | +0°0019 | het " | sl —0:0043 | 21L & mf. + 14°0 | 200 iO .COOOR Ss) i | ate ae re a ‘ —0:0015 | 41. cle ail ur teva | 800 _ +0-0010 | » Geral .- os | —0°0036 | ; | ee 42-0 1800 +0:0019 | 3 | dome ‘s 5 | —0:0042 | July 25. Chick, 108 hours. Influence of Temperature. Exe. by Condenser discharge. 5-6 volts; 4.1L 5 mf. + | Time. Temp. Blaze. Resistance. | | | | [ve ee ele | 0 28° +0017 50,000 w | | 5 28 0016 ~ | i fLO 34 0014 ES | 15 38 °5 0009 = | + . 20 41 °5 0004 35,000 | 25 44° 0000 = | 35 ay 0000 — | 45 Flog 0000 35,000 | Summary. The presence of a blaze-current is a certain sign that development has progressed within the egg. ; ! 190 Dr. A. D. Waller. On the [July 17, In the early stages—when presumably the blastodermic membrane has not yet become folded to form a tubular embryo—the blaze- currents aroused by both directions of excitation are positive or ascending. At a more advanced stage of development the blaze-currents are usually homodrome with the direction of excitation, viz., positive or ascending after a positive or ascending excitation, and negative or descending after a negative or descending excitation. In some cases, but with such infrequency as to deserve to be charac- terised as exceptional, both responses have been observed to occur in a negative or descending direction. This may have been due to attachment of the embryo to the shell. [I have made a few observations on frog’s spawn, and although these have not yet been sufficient to enable me to specify the conditions of the presence or absence of the reaction, I think that the fact of its presence is worth reporting, as well as the further fact that in some cases of undoubtedly living spawn I have failed to detect it. | +0] O 5 /omins. Frog's Spawn.—Two homodrome responses of + and — 0°022 volt to excitation by break current at + and — 5000. Vole. +-03 CU ao rYYYVNVVVVYNVNN 4080.!| 4) 10 £0 30 Mmirs. Frog's Spawn.—Series of homodrome responses to series of single break induction currents 5000 + at intervals of two minutes. +02 +-01 “00 ‘asuodsoa JO | | a0tq[0A APUSIS SUUIN[OD ANOF Oso] UT StoquINU IL, \ | o° 6 ‘ am ‘sup p 40 ¢ f ww pes fees ease i Gumemale ca aithh: ey eae nce | dOfF AOgUqnoUL oly UL stout qgdoy 410 It Ut | U teu de WS i ie - ‘OT | pourny useq you pey s85o osoyy, “G'N | oe i | leg i pee Wg (98st : i aes | jru piu ! ju [tu te a W eee te eS Ode | | [fotus peg JUIUAAO[IAVP ONT 9 : : | | L ta a yu [a Be ee 5 ae hae oT eeu *[BULLO NT Z000- 0 + 9100-0+ | &000-0+ 9000;0+. [°° .— pag 5 Sf SUBS > ‘TeUtON | 0200-0+ | 0900-0+ || 8100-0+ | 0900-0+ |**° “ pag ees _ “PL eee B C B Cc The particular point that aroused my attention in the case of the eyeball was the fact that the anterior half of the eyeball was some- 1902.] « Blaze-currents” of the Crystalline Lens. 195 times found to give a larger response than the posterior half, and the present observations proceed from an attempt to determine the princi- pally effective part in such reaction. And I may state at once, as my chief conclusion, that it is the crystalline lens. The eyes upon which the determination was made, in the first instance, were those of fish—whiting and mackerel—by reason of the fact that these were for a season at my disposal quite fresh from the sea. [subsequently made similar observations on the eyes of octopus, on sheep’s eyes fresh from the slaughter-house, and on the eyes of recently killed cats and rabbits ; also on the eyes of an owl. The point that was most striking in these first observations was the great endurance of the reaction in the crystalline lens as compared with its rapid disappearance from the remaining tissues of the eyeball and from the skin, and with the rapid disappearance of the direct electrical excitability of muscle. I should, as an outcome of these observations, look for the Jast sign of life of a fish by testing the crystalline lens, whereas in the case of man J should test a piece of skin. The reaction—as far as I have yet seen—has been completely absent from frozen fish (salmon) as received from London fishmongers. Its normal direction in the lens is “ negative,” 7.c., from external to internal pole. It is abolished by heat (70°) and by compression. My first experiments with the eyes of fish were to ascertain on the entire eyeball what type of blaze reactions—if any—is manifested The results were as follows :— Hap. 1. Whiting.—Excit. and lead-off through AB. Berne coil. Two Leclanchés in primary circuit. Single break A induction shocks. 1000+ gave —0-0004 volt. LOOO* 5, =050007 5; 5000+ ,, —0:0010 ,, NOOO 52 Ooooh =e Reactions after immersion in hot water. 5000+ gave nil. BROOD ys Ve55 Lup. 2. Whiting —"Excit. through AB and lead off through BC or AC. (The system of notation is ex- plained in the ‘ Roy. Soc. Proc.,’ vol. 69, p. 183.\ B C A 5000 - <———= - 0-00/0 5000 + —————_—> ——=—= — 0:0005 SGC Rela ae 196 Dr. A. D. Waller. On the [ Oct. 25, Response from A to C; no response from B to C. Lens alone. Exe. 1000+ gives —0°0025 1000- ,, -—0-0060 Cornea alone. Exe. 5000+ ,, nil. 5000— ,,. —0-0005 2nd lens alone. Exe. 1000+ ,, —0:0020 1000- ,, -—0-°0050 Completely abolished after immersion in hot water. A similar experiment gave similar results: in the first lens the response was completely abolished by compression, in the second lens it was greatly diminished and modified by tetanisation ; the response to — + excitation being at the outset — —, then — +, then + +. Excitation —_— os hespense ee ——— [lg os 99 $5 ty) ——> —————— Exp, 3. Mackerel-About 5 hours after death. Lens alone. Ist lens. Exec. 1000+ gives nil. 1000 — 2 ”? 10000"); 200007 10000 =) 43795, C00tS 2nd lens. Exe. 1000+ ,, +0-0050 1000- _,, ~==+0-°0020 After strong tetanisation for ] minute. 1000+ ., +0°0015 1000— ,, —0-0020 After compression there is no response at all toYany strength of excitation. N.B.—I was not alive to the orientation of the response in this experiment. The + and — signs may therefore have been transposed in these two experiments. Exp. 4. Mackerel.—2A4 hours post mortem. Isolated lens. 1000+ gives —0-0005 volt. L000 = Fein OF 000Sae. 5000+ ,, -—0°0010 ,, 5000-— ,,- —0°0015 ,, After strong tetanisation for 1 minute. 5000+ ,, +90:0010 volt. |9000—- ,, —90:°0008 1902. | “ Blaze-currents” of the Crystalline Lens. he After compression no further response. There was no appreciable alteration of resistance after tetanisation. Exp. 5. Whiting.—4 hours post mortem. Isolated lens. 1000+ gives +0-0002 Hao 0- 0010 500+ gives | nil nil { 1000+ , | +0-0002 +0-0002 2000-2 0.00251) 00020 5000+ ,, | +0°0050 +0-0040 | OOM, | +0°0055 +0-0050 | \ No further response after compression. Voltage of response —~> Strength of stim. —~ Exp. 6. Mackerel.—48 hours post mortem. Isolated lens. Exe. 1000+ gives —0:°0008 volt. 1000 — a —0:°0004 _,, 10000 + “ —0°0010 ,, 10000 — % —0:0015 ,, The response is abolished by plunging lens in hot water. Kp. 7. Octopus.—15 hours after removal from water. The isolated eye gives no distinct response either to light or to electri- cal excitation. Its isolated lens gives to 1000+ a response of +0-:0040 volt. 1000 — +0-0005_,, 5000 + ‘5 +0°0015 ,, 5000 — 4 +0-0010_,, 198 Dr. A. D. Waller. On the [Oct. 23, Later and with altered position of lens on electrodes. 1000+ a response of — trace. 1000 -— ie — 0-0004 volt. 5000 + 4 =O; OOUnme 5000 — a4 = 0; OOn0N Exp. 8. Dogfish.—Isolated lens. © ' 1000 + sives | — 0 0012 volt, 1000 — ¥ 00020 The response is abolished by pressure. Eup. 9. Cuckoo Fish (white).—Some hours (? 4 or 5) after death. The eyeball gives no response to light or to electrical excitation. Its lens, to 5000 + gives —0-0020 volt. 5000 * -—0°0050 _,, Set up in connection with three electrodes so as to be excited through AB, and led off through AC or BC (as for the entire eyeball). The lens responses are as follows :— Br 1G) ax AB Exe. 5000 - = ~<———-——— AC Resp. — 0035 edad AB Exe. 5000+ | —__-—-—~> AC Resp. — -0025 Zouses AB Exe. 5000-—)— =9<— —__—_ BC Resp. — 0015 <—— AB Exe. 5000 + = BC Resp. — -0006 <——— Exp. 10. Cuckoo Fish (yellow).—Isolated lens. A few hours (? 4 or 5) after death. Exe. 1000 + gives — 90-0030 volt. 1000 + , —0:0050 _,, On returning to London, I first tried salmon’s eyes, and then adopted the cod’s eye as affording a constant supply of suitable material. Ewp. 11. Salmon (from a London shop).—Isolated lens. No response. The fish had been kept in ice. 1000+ gives nil. L000. 10000 ts LOOOOT 55) 55 1902. ] “ Blaze-currents” of the Crystalline Lens. 199 Hap. 12. Codjish (1st).—Isolated lens. Fish said to have been brought to shore on the previous day. 5000 — gives —0:0040, —0:0020 5000 + 35 —0:0010, -—0-:0010 Photo. 4242 is now taken— 5000 + vs — 0-0009 5000 — Ad =) 0022 4000" | £XG | Yoon": + : = Antidrome. Homodrome. fe) Ss 10 /5 20 25 Ns, Photo, 4242.—Codfish Lens. Anti- and homodrome responses. The normal current was — 0-0008. The other lens gave no distinct response. The same lens next day gave to 5000 + — 00005 — —0-0015 abolished by heat. Exp. 13. Codfish (2nd).—Lens. 1 or (?) 2 days after capture of the fish, tested by the ABC method for determination of the seat of the response. (Given 2n extenso.) Eup. 14. Codfish (3rd).—Lens. 4% second day after capture. 5000 + — 0:0007 5000 — —Q:0053 Ob. UXXI. Q 200 : Dr, A. D: Waller. ° On the [ Oct. 25, 10000 + =0:0013 10000 — — 0:0048 10000 — — 0-0052 10000 + —0:0015 6 5000 — —0:0027 5000 + —0-:0008 Moderate compression gives — deflection off scale in consequence of mechanical excitation. Considerable compression abolishes all re- sponse. The other lens gave no response; the eye from which it had been removed was evidently injured, being full of blood. Kap. 15. Codfish (4th).—Supplied as fresh ; neither lens gave any response. | Hap. 16. Mackerel—Lens. Fish reputed fresh. 5000 + — 0:0004 5000 — — —0-°0011 Moderate compression - —— 5000 + + 0:0002 5000 — + 0:00015 Severe compression - 5000 + nil. 5000 — - Exp. 17.—The lenses of four sheep’s eyes brought from' the slaughter-house and tested within 3 hours after death gave no distinct response. Exp. 18. Sheep's Head.—I1st lens removed from the eye 4 hours post mortem with least possible compression. Normal current = +0°003 decreasing. 1000 + nil. 5000 + — 0:0008 — —0:0030 + | cs) S =) i) bS Same lens Next day eee eal, | - OO ae S (ep) =I No response after immersion in hot water. 2nd lens removed from the eye 18 hours post mortem. 5000 + —0:0005 - —0:0012 1902. | “ Blaze-currents” of the Crystalline Lens. 201 Then photo. 4248— 5000 + —Q0:0010 | sateies —0:0014 | » + — 0:0005 5 = —0:0008 Kap. 19. Cat.—Lens. 14 hour post mortem. Ist lens 5000 + — 0°0003 fee — 0°0008 Response abolished by compression. 2nd lens 5000 + — 00002 sabiten — 0:0006 Same lens + — trace _next morning — + trace Exp. 20. Mackerel.—Lens. Reputed fresh. 1000 + + 0:0002 ~ —0:0001 5000 + + 0:0003 ~ —~0:0004 10000 + +0-0004 = -- 0° 0006 Responses homodrome throughout. Hap. 21. 5th Codfish.—Lens. Ist lens 5000 + — 00002 = +0:0001, —0:0002 10000 — — 0:0006 as —0:-0002, +0:°0003 2nd lens 10000 + —0-0001 * +0:0001 Attempts were made to test the last two lenses by lateral eye rota- tion ; the results were uncertain and variable. cases were equally variable. zp, 22. Cat,—Lens. Similar trials in other 5 hours post mortem. Initial current +0:0011 1000+ and — gave nil, nil. Ist lens. 5000+ ,, — 2nd lens 5000+ 5000 — Exp. 23. Brill.—Lens (% 24 hours). Initial current: Exe. by single break 1000+ 5 —0°0004 and nil. ay EOI 5, — >0-0030 (off scale). — 0°0036 — 0:0020 & bo 202 Dr. A. D. Waller. On the [Oct. 25, 1000 — —0-0080 100 + — 0-0003 100 — —0:0015 After compression 1000+ and — roull, vam Electrodes tested by 1000+ ,, — 39 99 Note.—The lens of this fish is rather smaller than convenient. Hep. 24. Frog.—. temporaria. - Entire eyeball. Initial current: = — 10-0030 Exc. by single break 100+ = +off (>0:002) Er ‘ 100 — = + off (> 0-002) Its isolated lens. LOO= and =) =) 5 nls 1000+ | = -—0-:0005 1000 — = —0:0010 Subsequently both responses were observed to be homodrome, viz., + to + exc., and — to — exe. The isolated lens of the other eyeball gave similar results. The lens of the frog’s eye is inconveniently small, nevertheless, with due care, typical effects can be observed upon it, viz., negative responses* to both directions of excitation, the homodrome exceeding the antidrome response. ‘The normal and typical response of the entire eyeball was, as previously stated, positive to both directions of excitation. Rana temporaria has, in my experience, given clearer effects than Rana esculenta. Hap. 25. Cat.—5 hours post mortem. Ist lens. Initial current = —0:0030 Exe. by single break 5000+ = = (I) 000s 5000 — = -off scale (>0°0030) After compression the positive current was nearly doubled, and there was no response to 5000 + 5000 -. . The 2nd lens, less carefully removed, gave no response to 5000 + and —. Of five successive cod’s heads supplied to me in London as fresh, all but one gave responses of typical character, as illustrated by the photo- gram; in every case, however, the lenses of the two eyes were * Throughout this paper, positive current signifies current through the (eyeball or) lens directed from posterior to anterior surface, and negative current the reverse of this. In one experiment (Exp. 13, with reversed zincs) the direction of the readings unavoidably breaks this conventional rule. And, indeed, in other experiments this rule has occasionally been broken, as in Exp. 13, in order to set aside conceivable fallacies of the electrodes, kept in an invariable relation to each other—e.g., an invariable inequality between them, or a gravitation current of liquid from A to B, or a constant difference of area, and therefore of current- density at A and B. 1902. | “ Blaze-currents” of the Crystalline Lens. 203 unequally good ; in three instances one of the lenses gave no response, and in one of these three instances the eyeball was filled with blood. I think the difference between the two eyes must have been due to the fish having been killed by stunning, or it may be that in transit to London one of the eyes had suffered compression. But whatever the real cause of the difference may have been, the lenses of fish obtained in London were far less satisfactory than those of fish directly taken from the sea. In the latter case, both lenses, if carefully removed, were equally effective (provided the fish had not been stunned in the usual way on removal from the hook). Similar effects are obtainable on the.crystalline lens of the mamma- han eye ; but it is essential to avoid any undue compression of the globe. Thus I completely failed to observe any effect on the lens of eyes removed from the orbit of dogs and cats in the usual manner, also on the lens of sheep’s eyeballs brought fresh from the slaughter- house. But with lenses carefully removed from the eyes of a fresh sheep’s head and of a recently killed cat, typical and regular responses were obtained, which were abolished by intentional compression as well as by immersion in hot water (60° to 70°). I think it desirable to give im eztenso one experiment (No. 13) to illustrate the precise nature of experimental evidence and the system on which it is taken down. It is very easy to make sure of the direction of a current used for excitation in relation to a total or bipolar response, but it is not easy without a strictly systematic plan to make sure of this relation when a partial or unipolar response is under investigation. It is advisable for the latter purpose to carefully verify the connections of the ABC key* so that directions of deflection may immediately signify directions of current between the points of investigation, and be noted accordingly in a legible form that can be readily reviewed. Fost. Equal. Ant. B (6, A Pe s000 > throuch DA. << Response from AC —-002 Anti post -kathodic Exc. 5000— through BA <———————-—— Response from AC + -0062 CA. BC. enc +-00/6 All response was from the anterior pole; none from the posterior. The series was repeated with similar results, and now the zines of the electrodes were transposed so that the connections were— Ane. Equat. Post. B Cc A <$ —___. pita BBE TT + -004 i.¢., a8 before response only from anterior and not from posterior pole. Now the lens is turned round—the zines left reversed as before— and the connections thus revert to :— Fosé. - Equal. Ane. ‘oes ~ *0007 oe (ecm “ *00/0 Finally, the zines are replaced so that we have the lens left reversed— Ant. Equa. Post. 8B Cc A ee D 1902. | “ Blaze-currents” of the Crystalline Lens. 205 The lens is submitted to compression, after which there is no response of any kind, either total or partial; it is placed for a few minutes in hot water until coagulated white, and again tested without any response. The temperature at which the first obvious sign of coagulation was observed was 38°. ‘The lens was completely white at 48°. Beyond 50° no further increased whiteness could be seen. I conclude from this and similar experiments-—— 1, That a crystalline iens of suitable size is a good object upon which to study the nature of blaze-currents. 2. That a “ blaze-current ” is a physical sign of the ‘“ living” state. 3. That a blaze-current may be post-kathodic as well as post-anodic, antidrome as well as homodrome. 4, That the direction of blaze-currents in the lens is negative or ingoing, 7.¢., trom external or anterior to internal or posterior pole. ADDED DECEMBER 4, AFTER THE RECEIPT OF Dr. DuRIG’s PAPER. (See p. 212.) A Hap. 25. Nov. 1. Owl.—Positive. responses of the anterior half of the eyeball; negative responses of the lens alone; positive responses of the cornea alone. Anterior half. 1st eyeball. 7 minutes post inortcim. Initial current = —(0:°0008 to —0:0018 volt. Exe. by single break 1000+ + 0° 0005 ” ” e ae 0: 0005 0 5000 + +0-0020 x a — +0°0015 4 a — +0:0015 53 7" + +0:0015 Lens alone. 10000 + —(0-0005 — — 00020 Cornen alone. Single break 10000 + nil, a 10000 — nil. Several 10000 + +0°0012, +0:0010 a 10000 — +0:0020, +0:°0020 Tetanisation 10000+ +0°0015 3 10000 — +0:0025 Cornea plunged in hot water. Tetanisation 10000 — nil. . 10000 + + 0:°0005 206 2nd eyeball. Initial current Exe. by single break 5000+ J) 39 Lens alone. Exc. by single break 5000+. bb) >) 1000 + Cornea alone. Single break 5000+ ‘ 10000 + Several 10000 + Lens replaced, 1000 + 5000 + Single break 29 9) 99 Dr. A. D. Waller. On the [ Oct. 23 Anterior half. 2 hours 10 minutes post mortem. = +0:0005 to nil. +0:0010 volt. +0:0015_,, — >0-0030 (off scale) — >0-°0030 — 0°0004 — 0:0010 nil. HO OO On nil. + 0:0009 nil. +0:0010 —0:0007 —0:-0012 +0:0003 +0°0005 hap. 26. Codfish—Removed from water on Friday, November 7, at 4 P.M. Ist lens. Society.) 1000 + 1000 — Exc. by single break induction shock of Berne coil with 2 Leclanchés in pri- mary circuit 1000 + 1000 — 9) 10000 + 2nd day. Ist lens (Nov. 2nd lens (Nov. 9) 6PM. 500+ 600 + 99 IG) TEAING = oy 2k 33 (November 8, 4.45 P.M. Demonstration at Physiological Lesponse. — 0:004 volt. — >0°:005 (off scale) after nil. compression nil. —0Q:0013 antidrome blaze —0:0010 homodrome ,, = 0: 0002 antidrome rae —0:0018 homodrome ,, —Q°0003 antidrome _,, --0:0018 homodrome ,, =@0-0003 antidrome —0:0017 homodrome ,, ? 1902. ] “ Blaze-currents” of the Crystalline Lens. 207 700 + —0:0013 antidrome blaze = —(0s002 Ss homodromen,, 600 + —0:0008 antidrome ,, 0 0025eehomodrome),. »y + —0:0006 antidrome ,, » 7 —0:0020 homodrome ,, + —0°0004 antidrome ,, — —0:0018 homodrome ,, + —0-°0003 antidrome ,, — —Q-0011 homodrome ,, + hour interval. 600 — —Q:0031 homodrome ,, ard day. 2nd lens (Nov. 10) 600— >-—0:0038 homodrome ,, 500+ =O20011 “antidrome - —0:0038 homodrome ,, 400 + —@0:0003 antidrome ,, - —0:0016 homodrome ,, 99 4th day. 2nd lens (Nov. 11) 400 + —Q0-:0008 antidrome _,, — —Q:0006 homodrome ,, 500+ —0:°0005 antidrome ,, — —0:0005 homodrome ,, 400 + =@0:0001 antidrome ,, = —0:0002 homodrome ,, Sth day. 2nd lens (Nov. 12) 500 + nil. = nil. 1000 + nil. acre nil. 5000 + ee } polarisation i“ + 0-0002 Tetanise 10000 + +0-0010 } ne thE = OOO sa 6th day. 2nd lens (Nov. 13) Tetanise, 12 p.m. 10000+ —0°0005 } hinee tes — 00006 menxday. . ,, 2PM. 10000+ * ooo p Polarisation = =0-0008 2 Sing. shock 10000 + —0-0001 } polarisation = 0 00020) am This is the longest period during which I have followed the response of an isolated lens. 208 Dr. cae DP Waller, "On dhe [Oct. 23, Indubitable response, negative to positive (antidrome) and negative to negative (homodrome), to single shocks, was observed on the 4th day, absent on the 5th day. Response, negative to negative, with strong tetanisation in both pairs of directions, was still present on the 5th and 6th days. But on the seventh day the only visible effects were of polarisation direction, 2.¢., the lens was judged to be dead. Hap. 27. (November 19, with Sir J. Burdon-Sanderson.)—Lens of rabbit eye, removed 4 hours post mortem. Normal current 0°0150 volt positive declining. Ist lens. Excitation by single shock, Coil at 1000+ = -—0-0006 = -—0:0010 followed by positive after- effect. Excitation with the ABC method. By single induction shocks. (1st day). Coil at 5000 : a: = +0:0003 5000 a Qnd lens. S.s. L000 S.s. 5000 Exc. Exe. + 0:°0002 —0:0005 —Q:0012 —0 0002 —0-0015 + 0°0002 1902. ] Exe. Exe. S.s. 5000 Exe. Resp. Exe. Resp. Next day. Ist lens. (2nd day.) S.s. Resp. Eixe..; Resp. Exe. Resp. Exe. Resp. Same lens compressed. S.s. 5000 Exe. Resp. Exe. Resp. 5000 Exe. Same lens coagulated by heat (53° C.— 62° C.) Exc. Ktesp. Exe. Resp. Tetan. Exe. Resp. Exe. Resp. 2nd lens. S.s. 5000 Exe. Resp. Exe. Resp. | > eS ——— — vol. 4, p. 318. 1902. | Sun-spots and Terrestrial Magnetism. 221 Range of temperature. Nis. Ages. - 182° to +15° 0:0972 0°0568 Ly set 00% 0°1248 0:°0737 1D eed. 0°1333 0:°0903 The mean value for the specific heat of silver sulphide is less than that for nickel sulphide throughout, but little can be deduced from the results till the influence of temperature on the specific heat of sulphur is known. “ Preliminary Note on the Relationships between Sun-spots and Terrestrial Magnetism.” By C. Cures, Se.D., LL.D., FBS. Received December 18, 1902,—-Read January 22, 1903. (From the National Physical Laboratory.) I have been engaged during the last two years on an analysis of the magnetic results obtained at Kew Observatory (now the National Physical Laboratory), during an 11-year period, 1890 to 1900. ‘The work has been much interrupted, and is still incomplete. Amongst the points dealt with is the inter-relationship between sun-spot frequency and magnetic phenomena, and, as this has recently been engaging attention elsewhere, I have decided to put certain of my results on record at once. It has long been known from the researches of Balfour Stewart, Ellis, and others, that there is a close connection between the times of occurrence of greatest sun-spot frequency and largest amplitude of the diurnal inequality of magnetic declination and horizontal force. I have investigated whether the numerical relationship between the phenomena can be adequately represented mathematically in a simple way. A convenient basis for the investigation was presented by the publi- cation by Professor Cleveland Abbe in the ‘U.S. Monthly Weather Review, for November, 1901, of a table of sun-spot frequencies as calculated by Wolf and Wolfer for a very long series of years. After I had carried out all the calculations, Wolfer himself published a similar table* embodying his latest corrections. The differences from Abbe’s table are trifling, and mainly confined to two years (1891 and 1892). I judged it best, however, to revise the whole of my arithmetic, so as to employ Wolfer’s own most approved figures. In the following remarks S represents Woifer’s value for the sun-spot frequency. The above-mentioned table gives the mean S for each month and for each year. The magnetic quantity selected for comparison is the mean monthly “range, meaning thereby the difference between the greatest and * ‘Met. Zeitschrift,’ May, 1902 p. 195. 222 Dr. C. Chree. On the Relationships [ Dec. 18 least of the twenty-four hourly values in the mean diurnal inequality for the month in question, based on the five quzet days selected for the month by the Astronomer Royal, Calling this quantity R for any particular magnetic element, I tentatively assumed R= 00S Ak ee eee (iy. with a and d constants. I grouped together the 11 Januarys, the 11 Februarys, and so on, of the 11-year period, and determined a and 0 by least squares for each of the resulting 12 groups. There being only 11 years’ data, the calculated values doubtless are appreciably affected by quasi-accidental irregularities, but there is so striking a resemblance between the more conspicuous features of the results found for the declination, inclination and horizontal force as to justify the conclusion that the phenomena are bond fide. Full parti- culars will be given later. At present it will suffice to record the mean values found for the a and 6 of the formula for three groups of months —V1Z. :-— Winter, comprising November to February, Equinox . March, April, September, October, and Summer 3 May to August. The results are as follows :— Table I. Horizontal force. Vertical force. (Unit ly=10-* (Unit ly=10-° Declination. Inclination. C.G.8.) C.G.S.) (Fk (SiN (tT Bea Cla te a. b. a. b a. b. a. b. Wanter ..., 3 723 0 -0325> .0'263))00105. 10-5) 0216 7°0 0:0382 Equinox... 7°32 0:°0478 1:26 0 -0147 23°5 0°221 17°2 0-026 Summer ..' 8°91 © -°-0428 1:60 "0137 “30°76 70-190 22°7 0:035 Mean..... 6 “49 °O°-0410. “17 00180 2155 TORot 15°6 0-081 As is obvious from (1), a represents the amplitude of the range corresponding to a total absence of sun-spots. During the eleven years dealt with, Wolfer’s mean monthly values for S varied from 0°3 to 129-2, the mean being 41-7. To bring out more clearly the similarity of the results for the declination, inclination and horizontal force, I have represented the mean value of } for the 12 months in each element by 100. The corre- sponding values for the three seasons are, then, as follows :— Table II. Winter. Equinox. Summer. Declination,” 22.5. 79 aay 104 inclination cees-anee 81 113 oo NGS Horizontal force.. ... 85 116 99 1902.| between Sun-spots and Terrestrial Magnetism. 223 In obtaining these figures I have retained a figure in the value of b beyond that recorded in Table I. Tables I and II will suffice to bring out one of the most important points established, viz., that } is certainly different from one month to another, and is, for all the elements except the vertical force, decidedly larger at the equinoxes (more especially it would appear at the spring equinox) than at other seasons. This means that the equinoxes are the seasons at which the amplitude of the diurnal inequality, when considered absolutely, is most dependent on the sun-spot frequency. When we take into account, however, the difference between the ranges of the diurnal inequalities at different seasons of the year, we find that winter is the season when sun-spot frequency is relatively most important. This will be recognised on reference to Table III, remem- bering that a represents the range corresponding to a total absence of sun-spots, while a@ + 41:7 bis the range corresponding to a sun-spot frequency of 41°7, this being, as already mentioned, Wolfer’s mean value for the 11 years in question. Table IIT. Values on lenor a: Horizontal Vertical Declination. Inclination. force. force. NVanaber’.. 225. 6. Orr? 0-69 0:60 0:19 Bigiumimoxe <2. 02: 0°27 0°49 0°39 0:06 SUMMIMMET.:...... 0:20 (0) 2 5) 0:26 0:07 Table III serves also to bring out another important result, viz., that the influence of sun-spot frequency on the amplitude of the diurnal inequality is very much less for the vertical force than for the three other elements considered. A recent interesting paper by Rajna* shows that the idea of a linear relationship between diurnal magnetic range and sun-spot frequency has already been applied by at least two previous investigators, Rajna and Wolfer. They seem, however, to have applied it only to mean annual values, and to have considered declination only. Rajna, deal- ing with declination data, observed at Milan over the long period 1836 to 1901, applies a formula of type (1) to what he calls the ““medie annuali dell’ escursione diurna.” The value he finds for dis 0047. He mentions that in an earlier similar investigation, including declination data from several stations, Wolfer obtained the value 0’:040. I am uncertain as to the precise meaning of Rajna’s “‘ medie annuali,” but it certainly is not quite the same thing as the mean range in Table I, so that the results are not absolutely comparable. * “Rendiconti del R. Ist. Lomb.,’ Serie II, vol. 35 1902. 224 Sun-spots and Terrestrial Magnetism. [ Dec. 18 Another recent and able paper bearing on the subject appears in the last published volume of the French Bureau Météorologique, which has just come into my hands. The author, Mr. Alfred Angot, has anticipated me in applying a formula of type (1) to the individual months of the year; but he treats of the amplitude, not of the diurnal range asa whole, but of that of the coefficients of the several terms of the Fourier’s series into which the diurnal inequality can be analysed. The paper treats only of the declination—dealing with data from ordinary days at Pare St.-Maur, Greenwich and Batavia—but the author expresses his intention of considering in the future the horizontal force. A special feature of the present investigation is that the magnetic data are derived exclusively from magnetic quiet days. This suggests at once a query and a criticism, a query as to why one did not employ corresponding sun-spot data confined to the magnetically quiet days, a criticism that as the two sets of data employed do not absolutely correspond, the comparison actually made may be misleading. As to the query: Wolfer, it is true, publishes at regular intervals in the ‘Met. Zeitschrift ’ provisional sun-spot frequencies for each day. These figures are, however, presumably inferior in certainty to the final figures he has embodied in his table after consulting all available sources of information. The vital consideration, however, is that at certain seasons of the year there are a number of days for which, owing to the absence of observations, Wolfer has no provisional sun-spot data. With information lacking for two or three out of the five quiet days of a month there would have been a very undesirable amount of uncertainty. As to the criticism, it would be difficult to meet it if it could be held that the enhanced magnetic activity existing at the earth’s surface at times of sun-spot maxima is due directly to electrical disturbances in the sun, each disturbance being limited to regions where sun-spots exist, and only those disturbances being effective which happen to be at the moment on the half of the sun visible from the earth. At present I shall only mention the following fact :—I had monthly sun-spot frequencies calculated from Wolfer’s provisional figures, employing only the five “quiet” days selected for each month by the Astronomer Royal. The mean sun-spot frequency thence deduced for the eleven years (1890 to 1900) differed from the corresponding result given by all Wolfer’s days by less than one-fifth of 1 per cent. It is hardly necessary to point out that this fact has an important bearing, not only on the point immediately under consideration, but also on the further question as to the true nature of the connection between sun-spots and magnetic storms. 1902. | On Electric Earth-current Disturbances. 225 “ Characteristics of Electric Earth-current Disturbances, and their Origin.” By J. HE. Taytor. Communicated by Sir OLIVER Lopae, F.R.S. Received December 16, 1902,—Read January BO, US Osy The following notes refer to effects which appear to have a distinct connection with the so-called ‘ionisation ” of the upper regions of the atmosphere by radiations from the sun, and which have repeatedly attracted my attention during the course of recent experiments in wireless telegraphy for the British Postal Telegraphs. In the electronic theory of the causes producing the aurora borealis, it is assumed that by the deflection of the course of the flying ions or electrons towards the poles, due to the earth’s magnetic field, a con- centration results in those neighbourhoods, giving rise to the pheno- menon. The effects classed by telegraph engineers as earth-currents have also, apparently, a direct connection with the ionisation of the atmosphere. As is well known, these are at times, particularly when auroral displays are in evidence, so strongly pronounced as to interfere more or less with ordinary telegraphic working on earthed circuits. In special cases, where sensitive apparatus is used, they are, every day, sufficiently pronounced to cause disturbance, for some hours at least, even under normal conditions. They have been found to be particularly troublesome in the Post-office system of wireless telegraphy, in which a sensitive telephone receiver is connected in a low resistance circuit earthed in the sea at both ends. To enumerate in a systematic manner the various investigations which have been made from time to time on the subject of earth- currents would involve a lengthy paper ; but only the more prominent features which have forced themselves on my observation will here be briefly summarised. The disturbances evidence themselves by producing various charac- teristic noises in the telephone receiver. They have not been con- founded with ordinary telegraphic or other inductive disturbances, as they appear in circuits far removed from any such source of affection. In these latitudes they are always stronger and of more frequent occurrence in summer than in winter. They are daily in evidence for a few hours at or about the time of sunset, 7.¢., whilst daylight is fading. In general they do not evidence themselves to any great extent during broad daylight, but are readily precipitated by atmospheric electrical effects or any tendency to thunderstorms, and rarely, if ever, fail to herald the approach of a storm or gale. 226 Mr. J. EH. Taylor. Llectric Harth-current [Dec. 16, The characteristic noises produced may be divided into five classes resembling— (i) Uniform flowing or rushing of water: this is usually a day- time disturbance, and is occasionally of considerable vigour. (11) Intermittent crackling : an accompaniment of other disturbances. (111) Bubbling and boiling of water: the usual form of nightfall disturbance, but also frequently occurring in the daytime. (iv) Rocket disturbances. These are peculiar and characteristic, having some resemblance to the sound produced by a rocket rising in the air. They commence with a shrill whistle and die away in a note of diminishing pitch. They vary in intensity, but always have a similar duration of from 2.to 4 seconds; are freely heard at night, and only occasionally during the day. (v) Disturbances due to high frequency effects, inaudible on the telephone, but evidenced on the coherer, magnetic detector, or other form of Hertzian receiver. These various disturbances were, for some time, very puzzling to me; but on perusing Professor J. J. Thomson’s paper, read at the Royal Institution on 19th April, 1901, it speedily appeared highly probable that they were due to electrical effects produced in the atmosphere by the ionisation caused by solar radiations and the reaction on this ionisation by electric stresses in the atmosphere. The rocket disturbances, though they are probably not in them- selves due solely to ionisation, furnished the first clue to this explana- tion. They are characteristic of an initial high velocity rapidly damped and ultimately dissipated. They have the same duration as is usually associated with the passage of a meteor across the heavens, and the assumption is that they are actually caused by the passage, in sufficient proximity, of meteoric bodies which set up electrical discharges in the upper rarefied atmosphere, these discharges inducing electric currents in the sea and collected therefrom by the circuit. Assuming this explanation, it might reasonably be asked why such disturbances are not equally evident during the daytime as at night. The answer lies in the screening effect of the ionised (and therefore conducting) air during the daytime and the absence of such screening at night. Professor J. J. Thomson has shown, in a modification of the well- known cloud experiment, how the ionisation of a gas may be cleared up or dissipated by an electric field. Doubtless the electric fields to which thunderstorms are due produce similar effects in the atmo- sphere on nature’s gigantic scale. Hence we may expect, as is presum- ably the case, that the screening referred to above may sometimes be suspended for a time, even during broad daylight, and the rocket disturbances evidenced among others. Now this assumption of a reaction between the electric stresses in 1902.] Disturbances, and their Origin. 228 the atmosphere and the ionisation produced by the sun suggests the source of the other daytime disturbances referred to. It appears highly probable that they are the accompaniment of the clearing up process. Effects analogous to, if not actually, electric currents are doubtless produced in the atmosphere, which are induced in the sea and collected by the circuit. On the other hand, the nightfall dis- turbances are probably due to normal clearing up processes, revealed when the air becomes sufficiently non-conducting to act no longer as ascreen. These suggestions, though by no means complete, are sub- mitted for what they may be worth. It is probable also that the diurnal variations of the earth’s magnetic field are influenced by the same causes. One more point. The periods of maximum disturbances, experi- enced on the earthed circuits referred to, appear to coincide with periods of maximum atmospheric disturbances on the newer Hertzian system of wireless telegraphy, and indicate the same source of trouble. Further, I would suggest that we have here a clue to the true explana- tion of the greater night-time efficiency in signalling observed by Mr. Marconi in recent experiments. With ionised air the electric waves will be partly broken up and absorbed, with consequent abstrac- tion of energy from the transmission. At night, when the ionisation is cleared up, the strength of the radiated waves will be sustained. Some interesting investigations by the aid of sounds produced, in a telephone, by the passage of electrical currents through rarified gases can no doubt be carried out. Professor Righi has already made some observations in this connection, but much more can yet be done. 398 Solar Eclipse of 1900, May 28. [Dee. 17, “Solar Eclipse of 1900, May 28.—General Discussion of Spectro- scopic Results.” By J. EversuHep, F.R.A.S. Communicated by the Joint Permanent Eclipse Committee. Received Decem- ber 17, 1902,—Read January 22, 1903. (Abstract.) In a general way the conclusions arrived at from the discussion of the spectra obtained in 1898 are amply confirmed and extended by the present results. Itis now shown that every strong dark line of the solar spectrum exceeding Rowland’s intensity 7 is found in these spectra as a bright line; and the great majority of the bright lines of the flash spectrum, excluding hydrogen and helium lines, coincide with dark lines of intensity not less than 3. Most of the bright arcs of the flash spectrum are well-defined narrow lines admitting of considerable accuracy in the measures, and the present determinations of wave-length indicate that the coinci- dence of the bright lines with the dark lines is exact within 05 t.m. for all the well-defined lines. As regards the relative intensities of the lines of any one element in the flash and Fraunhofer spectra, my previous results require modification and extension as follows: The relative intensities of isolated lines of an element in the flash spectrum are in general, but not exact, agreement with those of the same element in the solar spectrum, and those lines which are exceptionally strong in the flash are in most cases lines which are enhanced in the spark spectrum of the element. All of the more prominent enhanced lines of iron and titanium, as . determined by Sir Norman Lockyer, are found to coincide with strong lines in the flash, but owing to the compound nature of some of the lines, it is not certain that all of these have abnormal intensities in the flash. There is no evidence of differences in the relative intensities of the lines of an element in the higher or lower regions of the flash layer, and the enhanced lines appear to predominate throughout the entire depth of the radiating stratum. The enhanced lines are equally prominent in the polar regions and in low latitudes, and the flash spectrum generally is now found to be the same in all latitudes and shows no essential change after an interval of five years. An explanation of the abnormal intensities of the enhanced lines in the flash spectrum is now offered, which depends on the assumption of a continuous circulation of the solar gases in a radial direction ; the highly heated ascending gases giving the predominant features to the flash spectrum, whilst the cooler more diffused gases, slowly subsiding, determine the character of the absorption spectrum. 1902.] On the Energy of Magnetisation. 229 The entire chromosphere is supposed to consist of innumerable smalleruptions or jets of highly-heated gases similar to the so-called ‘“‘ metallic” prominences, which are only the more pronounced mani- festations of the same eruptive agencies. Evidence for this is found in the characteristic features of the chromosphere, and in the detailed structure of many of the Fraun- hofer lines, which show wide emission lines underlying the narrow absorption lines. These ill-defined bright lines in the normal solar spectrum are distinctly displaced towards the violet, indicating a strong uprush of the hotter gases, whilst the narrow absorption lines are almost in their normal positions, and appear to indicate a slow and uniform descent of the absorbing gases. The final conclusion is that the flash spectrum represents the emission of both ascending and descending gases, whilst the Fraunhcfer spectrum represents the absorption of the descending gases only. “On the Electrodynamie and Thermal Relations of Energy of Maenetisation.” By J. Larmor, M.A., D.Sc., Sec. B.S. Re- ceived January 2,—Read January 22, 1903. 1. There appears to be still some uncertainty as to the principles on which the energy of magnetised iron is to be estimated, and the extent to which that energy is electrodynamically effective. The following considerations are submitted as a contribution towards definite theoretical views. The electrokinetic energy of a system of electric currents 1,, t2,. flowing in complete linear circuits in free aether, is known to be Z(yN i +%No+...); wherein N; is the number of tubes of the magnetic force (a, 8, y) that thread the circuit 4, and is thus equal to |(le+mB+ny)dS ex- tended over any barrier surface S which blocks that circuit, (2, , y) being circuital (z.2., a stream vector) so that all such barriers give the same result. As under steady circumstances (a, 6, y) is also derivable from a magnetic potential V, which has a cyclic constant 47 with regard to each current, this energy assumes the form OV OV OV’ San B|V(is ms +n jas, in which the integrals are now extended over both faces of each 230 Dr. J. Larmor. On the Electrodynamic and [Jan. 2, barrier surface. This is equal by Green’s theorem to the volume- integral | 4 | (a2 + f+") dr o8) — ae extended throughout all space. This latter integral is in fact taken in most forms of Maxwell’s theory to represent the actual distribution, in all circumstances whether steady or not,* of the electrokinetic energy among the elements of volume of the aether, in which it is sup- posed to reside as kinetic energy. 2. The most definite and consistent way to treat magnetism and its energy is to consider it as consisting in molecular electric currents ; so that in magnetic media we have the ordinary finite currents, combined with molecular currents so numerous and irregularly orientated that we can only average them up into so much polarisation per unit volume of the space they occupy. So far in fact as the latter currents are concerned, the only energy that need or can occupy our attention is that connected with some regularity in their orientation, 7.¢., with magnetisation, the remaining irregular part being classed with heat. If there were no such molecular currents, the magnetic force («, 8, y) in the aether would in steady fields be derived from a potential cyclic only with regard to the definite number of circuits of the ordinary currents. But when magnetism is present this potential is cyclic also with respect to the indefinitely great number of molecular circuits. The line integral of magnetic force round any circuit is 47(2c+ =u’), where =u refers to the practically continuous dis- tribution of magnetic molecular currents that the circuit threads. This latter vanishes when these currents are not orientated with some kind of regularity. If we extend the integral from a single line to an average across a filament or tube of uniform cross-section 6S, with that line for axis, by multiplication by 6S, we obtain readily the formula (adz + Bdy+ydz) = S4riu+ 4a |(Ade + Bdy+Cdz)dS in which (A, B, C)ér represents the magnetisation in volume 67. Thus, after transposition of the last term, and removal of the factor oS after the average has now been taken, we obtain J) | {(a — 4A) dz + (8 — 4B) dy+(y—42C) dz} = 4730 In other words this new vector (2 —4z7A, 6 —47B, y — 4x0), is derived from a potential which is cyclic in the usual manner with regard to the ordinary currents alone. * In the previous electric specification, the fictitious electric currents of aethereal displacement must be introduced when the state is not steady. 1903. ] Thermal Relations of Hnergy of Magnetisation. Zor If we compare this result with the customary magnetic vectors of Kelvin and Maxwell, it appears that («, 6, y) must represent the “induction,” and so will hereafter be denoted, after Maxwell, by (a, b, c). The new vector, which has a potential cyclic with respect to the finite currents only, represents the “ force,” and will hereafter be denoted by (a, 8, y), whose significance is thus changed from henceforth. The “induction” on the other hand has not necessarily a potential, but is, by the constitution of the free aether, always circuital ; that is, it satisfies the condition of streaming flow Ga NOU ce nak a ot Cu Oy CZ Y The expression for the energy now includes terms (Ny + No + aye 5) for the ordinary currents ¢,, 1, ..., where Ni, No,... are the fluxes, of magnetic induction, through their circuits ; this transforms as usual into i ¥ | V (la+mb+nc)dS over both faces of each barrier, which by Green’s theorem is equal to Z | (Gore hG ene oO os ee (i) 87 extended throughout all space. But there are also terms 4 (0y'Ny + t9'No' +...) for the molecular currents; now taking N’ to be the cross-section of the circuit multiplied by the component of the averaged induction normal to its plane, and remembering that «’ multiplied by this cross- section is the magnetic moment of this molecular current, it appears that v'N’ is equal to the magnetic induction multiplied by the com- ponent of the magnetic moment in its direction, and therefore 3Xu'N' is equal to 1 | (Aa+ Bb+Cc)dz. Thus the magnetic circuits add to the energy the amount* 1 | (Ree BBE Oy detent as lens pea (ii) together with : Qa (OC AAED SABE ONC I nese ee Ateiat Went de ane e (iii) * (These energies as here determined are kinetic ; if they are (as is customary) to be considered as potential, their signs must be changed. Cf. ‘Phil. Trans., A, 1894, p. 806.1 VOL. LXXI. Ss 232 Dr. J. Larmor. On the Hlectrodynamic and — [Jan. 2, The formula (i) is usually taken, after Maxwell’s example, to represent the energy of the electrokinetic fieid. It here appears that it represents only the part of the energy that is concerned with the currents, arising from their mutual interactions and the interactions of the magnets with them: that there exists im addition a quantity (11) which is that taken by Maxwell as the energy of magnetisation in the field (a, 8, y), and also a quantity (iii), which is purely local and constitutive, of the same general type as energy of crystallisation. The question arises whether (ii) is a part of the intrinsic energy of magnetisation of different kind from (11), in that it cannot even partially emerge as mechanical work, or on the contrary the usual formula (ii) must be amended. See §§ 5, 8. In any case the dynamics of the field of currents (when there are no irreversible features) involves only that part of the energy function in which the currents operate, thus excluding both (i1) and (iii). 3. The simplest example is that of a coil of m turns carrying a current 4, wound uniformly on a narrow iron ring-core, of cross- section S and length /. On the present basis the energy is made up of an electrodynamic part $on%S and a magnetic part $aBS/; as dani = 41 by the Amperean circuital law, these parts are HBr and sHAv, when v is the volume of the core ; they make up in all ®?/87 per unit volume instead of the usual $4}/87. The former part is mechanically available. The question has been raised by Lord Rayleigh* whether the latter part, which includes the very large term (iii) above, namely 2747v, in the case of iron, has any considerable mechanical effective- ness ; the question can only arise when it belongs in part to permanent magnetism whose ultimate annulment can induce a current,—when the current vanishes the energy of permanent magnetism, in the present case represented by (ii1) alone, is the only part of the energy of the system that remains. The conclusion reached by him is that it cannot be annulled quick enough, when the ring carries a coil, to develope any considerable available electric energy by induction. 4. We may form a rough illustration of the mechanical réle of this purely magnetic energy by considering, as the analogy of the currents, a branching system or network of pipes carrying liquids, in one of which a turbine is located, to be driven by the stream, which will be supposed to be an alternating one. ‘The flow will be directed more fully into this particular pipe, and higher pressure will also be attained, after the manner of the hydraulic ram, if it communicates at the side with an expansible reservoir into which the liquid can readily * “Phil. Mag.,’ 1885: also ‘ Archives néerlandaises, vol. 2, 1891, p. 6, reprinted in ‘ Phil. Mag.,’ 1902, and in ‘Scientific Papers,’ vol. 4, No. 272. 1903. | Thermal Relations of Energy of Magnetisation. 239 force its way, to be expelled again by the elasticity of its walls when the stream begins to set in the reverse direction. This increase of kinetic pressure on the turbine roughly represents the electromotive pressure on a motor due to the increased magnetic flux, and the energy spent in expanding the reservoir as it fills up represents the energy of magnetisation of the iron. If things were perfectly reversi- ble in the reservoir, that is if the iron were perfectly soft, the latter energy would rise and fall concomitantly with the alternations of pressure on the motor, but of course if its temperature remained con- stant it would contribute nothing to the energy driving the motor, which must be introduced into the system from an extraneous source. But if there are frictional resistances involved in filling the reservoir, the operations will not be perfectly reversible, and mechanical energy will be lost in it by conversion into heat ; and moreover on account of the phase of its changes getting out of step—still more by perma- nent delays such as are classed under hysteresis—it will operate less | efficiently in directing the stream of energy towards the turbine. Both these statements have analogical application to the iron in a magnetic circuit. An example is provided by the ring-coil aforesaid. Suppose that when the current has ceased in the coil the core retains permanent magnetism, its energy being the latter term in the formula above. This corresponds to the reservoir becoming temporarily choked, so that it retains its contents after the pressure that drove the liquid into it has been removed. The question arises whether this retained energy is available for mechanical work. The present aspect of the matter appears to lead to the conclusion (Lord Rayleigh’s) that it will. not be available to any considerable extent unless its pressure in the reservoir is considerable, that is, in the magnetic case, unless the iron is not very receptive of magnetisation. "The paradox that energy of residual magnetism, which is outside the electrokinetic system, can on running down affect that system, shows that the circumstances are more general than an analogy of a pure dynamical system of finite number of degrees of freedom can illustrate. In fact the equations of dynamics imply permanent structure of the system; whereas in Professor HEwing’s illustrative model of para- magnetisation, when the displacement is great enough the structure changes by the component magnets toppling over,* and after the general disturbance thus set up has subsided with irrecoverable loss of energy into heat, there remains a new structure to deal with. The only way to estimate the available part that may be latent in the great store of energy of residual magnetism of an iron core is thus by the empirical process of detailed experiment. Lord Rayleigh has inferred * The effective susceptibility dI/dH becoming enormous in the steep part of the characteristic curve. 234 Dr. J. Larmor. On the Electrodynamic and [Jan. 2, from the form of the curve of hysteresis for retentive iron in high fields that the fraction that is directly available at the actual temperature must always be small, and he supports the inference by considerations of the nature of the above analogy; in the absence of hysteresis there would be no such direct availability. He derives the practical result that a complete magnetic circuit is deleterious for induction coils in which length of spark is the desideratum, the increased total induction attained inside the ring-core being more then neutra- lised by the diminished promptness of magnetic reversal. In fact, if the core, laminated so as to have merely negligible con- ductivity, is surrounded by a perfectly conducting coil or sheath, and its permanent magnetism is removed at constant rate —dI/dt by an ideal process applied to it, the intensity of induction in the core will diminish at the rate —47dI/dt; and this defect of induction must be made good by the influence of the current thereby induced in the sheath, as otherwise there would be a finite electromotive power in it, which is impossible on account of its perfect conductivity. This restored induction is of the form H’+47l’, where I’ is the magnetism induced by the force H’ due to the induced current ; thus ad dt ee: , A 7, ! T— = 0, Ti (H + 4’) +42 A and the actual total rate of fall of magnetisation is diminished to ay ay which is only the fraction {7 Se ae aie constrained loss dt dt dt ' dt of retained magnetism dI/dt. In this most favourable case the action of . the coil or sheath thus delays the time-rate of loss of permanent magnetism in the core in the ratio (47x’)!, where «’ is the effective permeability for small additional force under the actual circumstances ; that is, the delay in reversal more than compensates the gain in in- duction. ’ 5. There remains another question, when viscous and other hyste- retic effects are practically absent so that the changes of magnetisation exactly keep step with those of the currents, and the degree of availability of residual magnetic energy thus does not arise ;—whether the energy of the magnetisation comes from the store of heat of the material and is thus concomitant with a cooling effect when no heat is supplied, or whether it is in part intrinsic inalienable energy of the individual molecules merely temporarily classed as magnetic. So far as it may be the latter, it must for each element of volume depend on the state of that element alone, like the part (11) of $2. It has already been seen that no part of (ii) or (111) can be supplied from the electro- dynamic field. This points to the intrinsic energy of paramagnetism, except an unknown fraction of the local part (ili), which depends only on the state of polarisation of the element of the medium, being 1903.) Thermal Relations of Knerygy of Magnetisaiion. 235 derived from purely thermal sources; and the following thermo- dynamic argument* will strengthen this conclusion. If the value of the magnetic susceptibility « for any material is a function of the temperature, we can perform a Carnot reversible cycle by moving a small portion (say a sphere) of the substance in the permanent field of a system of magnets supposed held rigidly magnetised by constraints. We can move it into a stronger region Hy of this field, of varying strength H, maintaining it at the temperature 0 by a supply of heat from outside bodies at that temperature ; we can then move it on further, having stopped the supply of heat, until its temperature becomes 0-60; we can move it back again isothermally by aid of a sink of heat at this temperature until the stage Hj, is reached, when further progress back adiabatically will restore it to its original condition. If « is a function only of the strength of field and of the temperature, this cycle will be reversible. If EH is the heat- energy supplied at temperature 6, and W is the work done on the sphere by external bodies in the cycle, the principle of Carnot gives the relation Bi Cae: d Now se = a) (41H, = 41,H,) 66 = —-4 = (H.? — Hy?) 60, if « is small, when the cycle is taken such that the change of H along the adiabatic * This theoretical deduction of Curie’s law has been already given substantially in Phil. Trans.,’ A, vol. 190, 1897, p. 287. The theory of diamagnetism, which assigns it to modification of conformation in the individual molecule by the inducing field rather than to average spacial orientation of the crowd of molecules, leads to a non-thermal origin as regards that part. The analogous question (loc. cit.) as to whether dielectric polarisation is mainly an affair of orientation of unaltered molecules like paramaguetism, or one of polarity due to internal deformation of the molecule like diamagnetism, is now answered by the experiments of J. Curie and Compan (‘Comptes Rendus,’ June 2, 1902). It appears that the dielectric coefficient of glass, for rapid changes, diminishes, but not very quickly, with fall of temperature, and that at temperatures below —70° C. duration of charge ceases to have influence on its value. The electric excitation is thus analogous to diamagnetism and has no thermal bearing, its energy being self-contained in the molecule; the signs of the susceptibilities in the two cases are different, because the one is of static, the other of kinetic character. The sharpness of the Zeeman magneto-optic effect has already led (‘Aether and Matter,’ 1900, p. 351) in this direction, for it requires that the electric polarisation in the molecule shall be of isotropic type, so that there may be no axis of maximum susceptibility. + This restriction is not necessary for the final result ; if « is not small, W and E have both to be multiplied by the same factor. 236 Dr. J. Larmor. On the Electrodynamic and [Jan. 2, part of the path is negligible compound with that along the isothermal part. Thus Now the experiments of Curie on the relation of « to 6 in weakly paramagnetic materials make «x vary inversely as 6; and this result has more recently been verified down to very low compe atures by Dewar and Fleming. ‘This gives E = $k (HH? = Ee): Thus the movement of the magnetisable material at uniform tempera- ture is accompanied by a supply to it of heat, equal to the mechanical work done by it owing to the attraction of the field; and this heat is just what is wanted to be transformed into the additional energy of intrinsic magnetisation (ii) of §2. It is to be observed that in the actual experiments x was small, and the other part (iii) of this energy therefore negligible: so that no conclusion as to the extent to which its source is thermal can be derived from Curie’s law. 6. The uncertainties of § 4 do not of course affect the estimation of the loss of motive power arising from cyclic magnetic hysteresis, for we have here to do with the mutual energy of the applied field and the magnet, not the intrinsic local energy of the latter by itself. If the applied field is (a, 6, y), the total energy employed in polarising the magnetic molecules in volume 67 is (Aa+ BB +Cy)ar. So long as the polarisation is slowly effected against the resilience of reversible internal elastic forces this is stored as potential energy ; but any want of reversibility involves degradation of some of it into heat, while if the field were instantaneously annihilated the molecules would swing back and vibrate, so that ultimately all would go into heat. Let us pass the magnetic body through a cycle by moving it around a path in a permanent magnetic field («, 8, y). An infinitesimal displacement of the volume 67 from a place where the field is (a, £, y) to one where it is (2+ 6a, 8 +68, y +6y) does mechanical work, arising from the magnetic attraction, of amount (ASa + BSB + Céy) dr. The integral of this throughout the whole connected system gives the virtual work for that displacement, from which the forces assisting it are derived as usual. Confining attention to the element 457 the work supplied by it from the field, to outside systems which it drives, in traversing any path is thus T | (Ada + BdB + Cdy), 1903.| Thermal helations of Energy of Magnetisation. 237 the integral being taken along the path. If (A, B, C) is a function of (a, 6, y), that is if the magnetism is in part thoroughly per- manent, and in part induced without hysteresis, so that the operation is reversible, this work must vanish for a complete cycle; otherwise energy would inevitably be created either in the direct path or else in the reversed one of the complete system of which 67 is a part. Thus the negation of perpetual motion in that case demands that Ada+BdB+Cdy = dd¢, where ¢# is a function of (a, 8, y), involving only even powers, and practically quadratic for small fields. Its coefficients are then the six magnetic constants for general aeolotropic material, no rotational quality in the magnetisation being thus allowable by the doctrine of energy. But if there is hysteresis, so that the cycle is not rever- sible, _ | (Ada + Bdp + Cdy), or in vector product form — 67 | 4d, represents negative mechanical work done, or energy degraded, in the cycle. In addition to this energy concerned with attraction, the external field expends energy in polarising or orientating the individual molecules against the internal forces of the medium, of aggregate amount a br | (ad A + BdB + ydC). In any case, whatever the hysteresis, the sum of this second part and the first reversed is integrable independently of the path, giving or | Aa+ BP + Cy b) namely, the change in the total energy in the element, thus vanishing for a cycle which restores things to their original state, as it ought to do The latter part is purely internal, and of merely thermal value as in §5. ‘The former part represents the averaged waste of direct mechanical energy in moving the iron armature through the cycle, and accounts for the heat thus evolved. It is the expression of Warburg and of Ewing for magneto-hysteretic waste of mechanical energy in driving electric engines ; for a portion of a cycle it repre- sents work partly degraded and partly stored magnetically. 7. Reverting to § 5, we may profitably illustrate by working out into detail a suggestion of Lord Rayleigh (loc. cit.). Consider a ring-coil of nm turns with a flexible open core of soft iron of length / and cross- section 8, whose flat ends are bent round until they face each other at a distance small compared with the diameter of section. We can apply Hopkinson’s theory of the open magnetic circuit to trace the 238 Dr. J. Larmor. On the Electrodynamic and [Jan. 2, transformation of the energy of attraction between these poles of the core, into electrokinetic energy of the coil, as the poles close up together. As frictional waste is not essential to this question, we can consider the coil to be a perfect conductor which will store all the energy without loss. We need not postulate that the iron is of constant permeability or devoid of hysteresis. When the distance between the pole-faces is 7, let the current in the coil bev. The total energy 1s dni.N + energy of magnetisation ; and part of the latter may remain sub-permanent when « vanishes. The principle of the magnetic. circuit gives, as 4S = N, the formula NN 1+ —% = 4rm, fis) 1S) assuming as usual that the lines of magnetic force are conveyed straight across the air-gap between the pole-faces. Thus the electromagnetic energy T, equal to dN, is nS Qa ; we Up and when « is diminished by —- 62, its increment is approximately In this displacement the work expended from the electrodynamic system in mechanical attraction between the poles magnetised to intensity I, equal to «N/p, is — IS27I6r, which is eerie -(1-2)2r0 So (i) \ (1/p)? Comparing (11) with (1) it appears that in cases for which p is great, to which alone the principle of the magnetic circuit can be applied, the work of mechanical attzaction by which the pole-faces can transfer potential energy to a spring placed between them, by compressing it, is concomitant with equal increase of the electrokinetic energy if the current do not change. As there is no source of energy, the current must therefore vary, and so that the total change of electrokinetic energy given oy © and ae vanishes ; that is, it must diminish by — 6, given by 6vu= 3 + (1 +(1- er which is practically equivalent to OL = me On. l This result may be immediately verified by the Lagrangian process. L ae As there is infinitely small resistance, the electric prossire == t Ou 1903. | Thermal Relations of Energy of Magnetisation. 239 the coil must be infinitely small; hence «a %+//p, so that a4 wl when # is negligible. The energy of oe of the core, which is 44/8, where §=*#/p», and therefore is $ me ane i IS, is not included in this con- Goalie servation. Its increase for a change oz is — ae Fn) Se which is wm \ Uf large compared with the quantities above. The fraction p~! of it, which is comparable with the other variations, is compensated thermally, by absorption of the heat of the system, and has, therefore, only the limited availability of thermal energy. The remainder belongs intrinsically to the magnetisation, constituting mutual energy of con- tiguous molecules ; how much of it, as above expressed, is of thermal origin remains undetermined in the absence of calorimetric experiment. 8. The main points that it has been sought to bring out are as follows :— (i) In an electrodynamic field there exists the usual specification of electrokinetic energy, but also in addition the energy of magnetisation of magnetic material. (ii) This energy of magnetisation appears as made up of a part given by the ordinary formula, which (when paramagnetic) is derived from thermal sources, and so in the absence of hysteresis has the limited mechanical availability of thermal energy ; together with a local part which is to some extent thus available, but is also in part permanent intrinsic energy of the molecules, regarded temporarily as magnetic energy. (ii) The law of Curie, that the susceptibility of weak paramagnetic substances 1s inversely proportional to the absolute temperature, is involved in these statements. (iv) The extent of the direct (non-thermal) availability of retained magnetism can be inferred only by empirical procedure, for example, in general features by inspection of the hysteresis diagram, as pointed out by Lord Rayleigh. 240 The Spectrunr of y Cygni. “The Spectrum of y Cygni.” By Sir Norman Lockyer; K.C.B., F.R.S., and F. E. BAXANDALL, A.R.C.Se. Received Decem- ber 11,—Read December 11, 1902. (Abstract.) The paper gives an account of the investigation of the Spectrum of y Cygni im relation to other celestial and terrestrial spectra. It is pomted out that this spectrum—which is of the Polarian type in the Kensington classification—is the connecting link between the spectrum of the Aldebarian stars, in which the arc lines of the metallic elements predominate, and that of a Cygni, chiefly com- posed of the enhanced lines of some of the metals. These two sets of lines are of about equal prominence in the y Cygni spectrum. It is also shown that, in regard to the relative intensities of the metallic and proto-metallic lines, there is a close resemblance between the spectrum of y Cygni and that of the chromosphere, thus indicating that the temperature and electrical conditions prevailing in the two light sources are nearly identical. A comparison, however, of the intensities of the stronger arc and enhanced lines in the two spectra tends to show that the chromosphere is, if anything, of a slightly higher temperature. It is claimed that the majority of the lines of y Cygni—and therefore, also, of the chromosphere—are due to metallic vapours, and that there is little evidence to support Professor Dewar’s suggestion that most of the 339 chromospheric lines, recorded by Humphreys in his eclipse results, can be accounted for as probably being due to the rarer atmospheric gases. The relation of the y Cygni spectrum to that of a Cygni is also discussed in detail. At the end of the paper a table is given showing the wave-lengths, intensities, and probable origins of the lines which occur in the photographic spectrum of y Cygni, taken with a 6” Henry prism of 45° angle. The lines are compared with those of the chromosphere, § Canis Majoris (Pickering’s record), and « Cygni. Dielectric Properties of Solid Glycerine. | 241 “Some Dielectric Properties of Solid Glycerine.” By ERNEST WILSON, Professor of Electrical Engineering, King’s College, London. Communicated by Sir WILLIAM Preece, K.C.B.. F.RS. Received January 7,—Read January 22, 1903. At high frequency, obtained by resonance, Thwing gives 56-2 for the specific inductive capacity of glycerine at 15° C.* The specific inductive capacity of pure liquid glycerine at 8° C. was found to be about 60, whether the frequency of experiment was of the order 2 million or 100.f At a frequency of 120 the specific capacity of glycerine varies from about 60 to 4-6, as the temperature is varied from — 50° to—100° in platinum degrees.t The present paper deals with the specific capacity of glycerine at temperatures varying from about +10° to — 80°C. The methods of experiment and the apparatus are those described in connection with the earlier experiments.t The platinum plates forming the electrodes of a condenser were placed in pure glycerine, supplied by Messrs. Hopkins and Williams, and after a preliminary experiment, made to find its conductivity (see table), the glycerine was frozen by aid of carbonic acid snow. The temperature of the glycerine was observed by aid of a platinum thermometer. At —48° C. the specific capacity at about 2 million periods per second is 3°97. At —44°C. the specific capacity at about 100 periods per second is 54, which is of the same order of magnitude as that given by Fleming and Dewar. The conductivity of this condenser at — 59° C., and with times of contact varying from 0:00002 to 0:009 second, was found and is shown by the curve No. 1 in the accompanying figure. If K be the instantaneous capacity and C the electric resistance, then the total capacity of the condenser at time ¢ is aah 1 K —di-— ~—t. c+ [a cz Now the capacity of this condenser at—48° C. at about 2 million periods per second is 0:00032 micro-farad, and the value of Basel och 1 : | om It — t, as given by the curve, will only account for -) 0-00002 20 0°00018 micro-farad. Assuming that it will be the residual charge which comes out in one-sixth of the period which produces the effect on the capacity, a large proportion of the residual charge comes out between the times #4, x 107° and 20x 10-* second. If the refractive index of the glycerine when frozen is taken to be 1-47 for short times, then Maxwell’s law would indicate a specific capacity of 2°17; and * ‘ Zeitschrift fir Physikalische Chemie,’ vol. 14, p. 293. + Hopkinson and Wilson, ‘ Phil. Trans.,’ A, vol. 189 (1897), pp. 109—136. ~ Fleming and Dewar, ‘ Proc. Roy. Soc.,’ vol. 61, p. 316. web. UXXI. x 949 | Prof, E. Wilson. . | [Jan. 7, 3°97 was observed at 2 million periods per second. It is possible, therefore, that some residual Una 2 has already come out at 2 ; million periods per second. | When the glycerine was warmed up to atmospheric temperature 10°C., it still remained solid, and its dielectric properties were examined at this temperature. The resistance of the condenser was 216000 ohms, whether the time of contact was 0°00002 or 0-011 second, and this is about 34 times the resistance at 10° C. when the glycerine was in the liquid state. No residual charge comes out between the above times. At 2 million periods per second the specific capacity of this solid glycerine at 10° C. is 6°67. By the method described previously* the specific capacity is 16 at 50 periods per second. The specific capacity in the liquid and solid states at 10° C. varies approximately as the conductivity for long times. The residual charge comes out at times less than 0:00002 Benall whether the glycerine be liquid: or solid ‘at 10° C, Finally, the condenser was frozen in carbonic acid | snow in ether, and the con- ductivity at — 81° C. is showir by the curve No. 2 in the figure. The final conduc- cael tivity is less, and the area is less than given by the curve No. 1. This difference is due | to the lower temperature for curve No. 2. Fleming ‘and | Dewar have shown that. the specific capacity of glycerine for long times changes very rapidly with its temperature between — 50° and — 100° Pt. | The specific capacity at | —81° C. was found to be | 3°8 at 2 million periods per iy second. — | | - During the freezing pro- cesses above referred to the conductivity of the glycerine was observed at different temperatures. The results. are given in the accompany- -005 ‘Oo += ing table. It is noteworthy UID EP ED NAEIID Ie] See eS. that for time 0:00002 second. * Hopkinson and Wilson, ‘ Phil. Trans.,’ A, vol. 189 (1897), p. 118. oO nN esi Conductivity & in 10.’Ohms”!. 1903.] - the conductivity falls and then rises, as the temperature is varied from +13° to —59° C. when freezing from the liquid to the solid state. Drelectric Properties of Solid Glycerine. analogous effect has been observed in soda-lime glass.* 243 An When the time of contact is 0-006 second, all the points in the table lie well on the same curve between — 26° and —81°C. It is when freezing from the liquid to the solid states that such serious changes occur at the short times. | i 1 Le | | Time of contact in seconds - after application , _ of force. 0 «00002 0 :00017 0 :00035 0 00099 0°0028 0°0060 0 :0090 0:011 —0:018 | Freezing from liquid to Conductivity in 10~-® ohms}. ] | | solid state. eeaago @. —26° C. 1°72 Freezing from solid state. —59° CO. +12°C. —48° C., 8-08 0-172 0:0708 | 0°0481 0°0319 0°0287 | 0 -0249 | 4°63 0 -0160 | Curve No. 1 0 °388 | ee 0134 | 0°191 ee 0-012 a — 75° Cc. —81° C.| 0-197 0:0301 0 -0201 0°0120 0 0109 * Hopkinson and Wilson, ‘ Phil. Trans.,’ A, vol. 189 (1897), pp. 109—136. bo | | | | 244 Sir Norman Lockyer and Dr. W. J. 8, Lockyer. [Jan. 14, “The Relation between Solar Prominences and Terrestrial Mag- netism.” By Sir Norman Lockyer, K.C.B., F.RS., and WILuiAM J. S. Lockyer, M.A., Ph.D., F.R.A.S. Received January 14,—Read January 29, 1903. [PLATES 4 AND 5.] It has been stated in a previous communication* that a preliminary reduction of the Roman observations of prominences, observed on the sun’s limb by Tacchini, indicated that, in addition to main epochs of maxima and minima of prominences coinciding in time with those of the maxima and minima of the total spotted area, there are also prominent subsidiary maxima and minima. One of us has pointed out in a recent communication to the Académie des Sciences+ that a comparison of the frequency of prominences visible in each solar latitude with the frequency of the most intense magnetic storms, indicated that (a) magnetic storms classed as ‘‘ great” by Ellis, and the greatest prominence activity near the poles of the sun occurred at the same time ; and (0) that the curve of general magnetic activity was nearly the same as that of the prominences observed near the solar equator. — The object of the present communication is to give a more detailed account of the research so far as it has gone. The Observations of Prominences. The fine series of observations, made by Tacchini, of the numbers and latitudes of prominences seen on the sun’s limb was used as a basis for the curves discussed. These observations were commenced in 1872, and have been continued up to the present day, so that we have a valuable continuous record. They have been published{ from time to time in full detail, thus rendering it possible to deal with them in any manner that may be desired. In the reduction of the observa- tions each zone of 10° was examined and discussed by itself. The observations were divided in the first instance into groups of three months, and the percentage frequency of the prominences was deter- mined by dividing the number observed by the number of days on which observations were made in this period. In this way a set of eighteen curves, nine for each hemisphere, was made, showing the variation from year to year of the percentage frequency of prominence activity in each ten-degree zone. * “Roy. Soc. Proc.,’ vol. 70, p. 502. + ‘Comptes Rendus,’ vol. 135, No. 8, 25th August, 1902. t ‘Societé degli Spettroscopisti Italiani,’ vol. 1, 1872; vol. 29, 1900. 1903. ] Solar Prominences and Terrestrial Magnetisn. 245 In the curves accompanying the present communication (Plate 4) the above-mentioned set, except those for 80°—90° north and south, have been grouped in pairs, thus representing the percentage frequency of prominences in each hemisphere for zones of 20° of latitude, 0°—20°, 20°—40°, &c., since it was found that this reduction could be made without losing any of the characteristic variations. An examination of these curves shows that they differ very con- siderably one from the other as we proceed from the equatorial to the polar zones. Generally speaking the curves representing the variations for each of the zones, 0°—20° north and south, cenform with the sun-spot curve; that is, the maxima and minima occur at about the epochs of sun-spot maxima and minima. Those for the two zones, 20°—40°, in both hemispheres conform also in the main to the general sun-spot curve, but in addition they display subsidiary maxima or changes of curvature superimposed on the main curve. The curves for the two zones, 40°—60° north and south, have, on the other hand, hardly any likeness to the sun-spot curve, but are made up of a series of prominent maxima representing special out- bursts of prominence activity. Passing to the curves corresponding to the next zones, 7.¢., 60°—80° north and south, these indicate two prominent outbursts lasting for a short period, showing that this region of the sun is, as a rule, practi- cally free from prominence activity ; in the remaining zones, 80°—90° north and south, the variation is small, and is a faint echo of the condition of affairs in the neighbouring zone 60°—80". The Magnetic Curves. The data regarding the magnetic phenomena employed in this com- parison are those brought together by Mr. William Ellis, in two published papers on magnetic phenomena.* We may take the opportunity here of thanking Mr. Ellis for kindly communicating to us a continuation of the data published in these two papers, which information he has brought down to the year 1899. Two classes of magnetic phenomena were there dealt with, namely, the variations from year to year of the diurnal range of the declina- tion and horizontal force, and magnetic disturbances. As regards the former, Mr. Ellis has shown7 that the curves indi- cating these variations are very similar to that of the general sun-spot curve; in fact, the curves were found to be almost identical in all their smaller irregularities. * © Phil. Trans.,’ Part IT, 1880, “On the Relation between the Diurnal Range of Magnetic Declination and Horizontal Force, as observed at the Royal Observatory, Greenwich, during the years 1841 to 1877, and the Period of Solar Spot Fre- quency”; ‘Monthly Notices, R.A.S.,’ December, 1899, vol. 60, No. 2, “On the Relation between Magnetic Disturbance and the Period of Solar Spot Frequency.” t ‘Phil. Trans.,’ Part IT, 1880. 246 Sir Norman Lockyer and Dr. W. J. S. Lockyer. [Jan. 14, The second class of phenomena, namely, the magnetic disturbances, which are more irregular in occurrence, has been classified by Mr. Ellis into five groups, and tabulated by him under five separate sub- heads. For the present paper, reference will only be made to one of these classes, namely, that described as ‘ great,” this group represent- ing the largest disturbances. The curve representing the variation in number of these disturbances indicates short intermittent crests, out- bursts, in fact, with rapid rises to maxima and falls to minima, and comparatively long intervals of quiescence. Comparison of the Curves representing Pronunence Frequency and Variation of Diurnal Magnetic Range. Mr. Ellis, as already has been pointed out, has indicated the close resemblance between the sun-spot curve and that representing the variation of the magnetic elements; and it has been shown in the earlier part of this paper, that the curves representing the percentage frequency of prominences near the solar equator, conform in the main to the general sun-spot curve. There is therefore an apparent connection between phenomena oc- curring in the equatorial regions of the sun (as represented by zones of prominences near the equator, and sun-spots which are practically restricted to these zones), and the ordinary diurnal magnetic varia- tion. The accompanying set of curves (Plate 5) illustrates the great similarity between those showing the frequency of prominences in a zone about the equator (0°—20° north and south) and the variations of the mean daily range of magnetic declination ; for the sake of com- parison, three other curves are added, showing the variation of the mean daily area of sun-spots for the whole, and the two hemispheres of the sun separately.* * In referring to the curve representing the variation of the mean daily areas of sun-spots, it may be noted that this is obtained by combining the mean daily areas of both hemispheres of the sun. A closer analysis shows, however, that this variation is not the same for both hemispheres. From the year 1862, when such a division of the sun’s disc can be easily investigated, the northern hemisphere, about the time of the two last maxima, displayed double maxima occurring in the years 1881 and 1884, and in the years 1892 and 1895. About the time of the maximum of 1870 this duplicity is not so marked, although when compared with the curve for the southern hemisphere for this period, there is a slight indi- cation of a subsidiary crest in 1872. In the case of the curve representing the mean spotted area for the southern hemisphere alone, at all the three epochs of maximum, the curves are single-crested and indicate sharply-defined maxima in the years 1870, 1883, and 1893. From the above it will be seen, therefore, that the actual epochs of sun-spot maxima, as determined from the northern and southern hemispheres respectively, 1903. ] Solar Prominences and Terrestrial Magnetism. - 247 Comparison of the Proninences with the Magnetic Disturbance Curves. If a comparison of the curve representing the number of days of the “‘oreat” magnetic disturbance is made with those representing pro- minence frequency (Plate 4), it will be seen that the former is as unlike the curves representing the prominence frequency about the solar i equator as it is like those near the poles; in fact, the polar prominence outbursts, and great magnetic disturbances occur almost simul- taneously. The peculiar form and general similarity of the curves can be best seen from the accompanying illustration (fig. 1). In the figure com- parison is made between the epochs of the crossing of the known and unknown lines, the percentage frequency of prominences about the _ solar poles and Ellis’ “great” magnetic disturbances. Two curves representative of prominence frequency are given, one to indicate the abrupt nature of the curves representing the frequency in a zone near the pole 10 degrees in width (in this case 60°—70° north), and the second to illustrate polar action as a whole; this latter was obtained by making a summation of prominence frequency for the _ two zones 60°—90° north and south. The simultaneous occurrence of the maxima suggests that, when the prominence action takes place at the polar regions of the sun, one effect on the earth is that we experience our greatest magnetic disturbances. Further, according to Mr. Ellis,* “ unusual magnetic disturbance is ~.. frequent about epochs of sun-spot maximum, and nearly or quite absent about epochs of sun-spot minimum.” We find that not only do these “ great” disturbances occur at the _. Same time as the polar prominences, but the spectroscopic observations __~ of sun-spots show that they take place not only “ about” the times of spot maximum, as stated by Mr. Ellis, but when the sun-spot curve is _ approaching a maximum and at the dates of the widened line crossings, T _. when the curve representing the “unknown” lines is on the rise, and crosses the “known” line which is descending. At the other epoch of “crossing,” 7.¢c., when the curve showing the “ known” lines is on the rise and the ‘unknown ” is falling, there is practically no magnetic disturbance recorded. Attention is again drawn to these crossings, as it is desired to indicate that it is only at those particular times when _ the sun is increasing his temperature that these disturbances occur. _ are not the same, and in dealing with the curve representing this variation for the whole hemisphere, this fact should be borne in mind. It may further be noted that the epochs of minima may be practically con- sidered the same for both hemispheres. * “Monthly Notices R.A.S.,’ vol. 60, p. 148. tT ° Roy. Soc. Proc..’ vol. 67, p. 412. tiie — val 2 . m a (‘Ajaatjoodsoa vultutm pur surrxeur yods-uns Jo syooda oy oyBoIpHT SOUT] [VOI4I0A UAayOLG puL snonuTyuOD oy y,) ‘soUl] PaUepIA Jo ssuIsso.10 pus ‘soomoutuosd avjod ‘sourqanystp oreuseur .,yeotd ,, Jo s{up Surmoys Uostavduog— [ ‘HIT 0-006! 0-068! 0-0¢9l 0-029 0-098! e) (sinng) _LWau5, G SAONVEYNLSIG JILANOWW E Te) all SAVd fe) CINIHODOWL) “LAWTON o0L -09 ‘"SAONANIWO Yd O O02 (INIHOOVL) ‘1V1°S 8°N 906 -,09 SP TUNANIWOS Og O Ol Cr SSN TAI GaNacim oz. ce Ke) OG o9 SDNISSOND OZ {J 0:006I 0-068! 0-088! 0-029! 0:09] Solar Prominences and Terrestrial Magnetism. 249 The facts in this paper explain why it is that magnetic storms some- times take place when there are no spots, or no very large spots, on the surface of the sun. Since the occurrence of magnetic storms is fe) eo) > 1.) Fria, 2.—Comparison between horizontal magnetic force (Bigelow) and solar prominences (Tacchini). (Continuous and broken vertical lines as in fig. oO o 2 \O oO Pa Cera | a re) Oo re) . Ti 9x i mee. oe pPOUEe 28a 2 o#2u2 Eos vz Ze Onn a (Ie) Be Za t= oO Se ane = >8r-e? ew >t =e fr (ade Rs Oz iss De Ge Spe go ira 1S Re ea lga~-u O37 z = —m— shown to be very closely connected with the solar prominences, there may be prominences and magnetic storms when there are no spots.. Prominences may also sometimes be associated with large spots, and as. the latter can be seen while the former can not, the resulting magnetic storm is generally attributed to the spots. 250 Solar Prominences and Terrestrial Magnetism. [Jan. 14, Further, the magnitude of magnetic storms appears to vary accord- ing to the particular position as to latitude of the prominence on the sun’s disc. The nearer the poles (either north or south) the prominence occurs, the greater the magnetic storm, and these are the regions where no spots exist. In this paper we have shown that the variations of the general magnetic phenomena, as given by Ellis, synchronise with the occurrence of prominences about the solar equator, while his “ great” magnetic disturbances occur, in point of time, with the appearance of prominences in the polar regions of the sun. Professor Bigelow has recently* investigated the variations in the horizontal magnetic force, and finds that the curve representing these changes exhibits subsidiary maxima which synchronise with those recorded in the curve representing the mean variation of prominences for all latitudes. Thus, to use his own words, “the remarkable synchronism between the curves cannot escape recognition, except after the year 1894, when an extra minor crest is developed in the horizontal force.” The accompanying diagram (fig. 2) gives Professor Bigelow’s curve, which represents, as he says, “the series of minor variations which were found in the horizontal magnetic force ... . after the 11-year cycle curve has been eliminated,” together with the percentage frequency of prominences in all latitudes obtained by us from Tacchini’s observations. * “Monthly Weather Review,’ vel. 30, No. 7, July, 1902, p. 352. (poatqoodsox vurrurit pue viarxent yods-uns so syoodo oy OFROIPUE SOUT] [LOTFA9A UOYOIG PUY sNONTL|OD oT]F)) "Q puv “NI 0UIOZ OZ Yovo oF soovoUtmMoAd «eos Jo Kousnboay osvqzuoodod oy} SULMOYS soAang 0-006! 0:068I 0:0¢89I 0:O0Z8I 0-098] ioe ! a Pasion een Heh (Boas ere fea og | eer a = | ) l MLAS l OGZ | 06-08 | O0G | | | | | | J I O Vas | OG2 | | 08-09 ! 00S i ’ ' i ' O ‘LV1'S i OG2 09-0r | i OO0G ' | ' \ ' | i = = Sa . m DAS S | l SESE SES [BASS TEST FS A) i a] al PP 0-006! 0-068) 0-08¢I 0-0Z8I 0-099! FAD] “TL, 100 “oon “v0g ‘hoary “uahyoorT pun waliyaorT Lockyer and Lockyer. 80-90° N. LAT. 60-80" N. LAT. 40~-60° N. LAT. 20-40 N. LAT. 0-20 N. LAT. 0-20 S. LAT. 20-40 S. LAT. 40-60 S. LAT. 60-80 S. LAT. 80-90 S. LAT. 250 QO N (e) 10.) oO Oo Sc 1860-0 1870-0 1880-0 1890-0 ee et 1 4 | i I | | | | | I I | I ! | ! | | | | I | | | ! ! ] ! | i | ( | | | | ' ' { ' | | { ' | | ! ' ! 1 \ | 1 \ | i \ { ! ' { ' ! 1 | { ! | ! | ' | i} | ' | | ' ! I 1 | ' ! ' | ' ! I I ! ! | 1 | ! I ! | I tS sre det | I | 1 oe Laas el eee Sy a 1860-0 1870-0 1880-0 1890-0 Curves showing the percentage frequency of solar prominences for each 20° zone N. and 8. (“he continuous and broken vertical lines indicate the epochs of sun-spot maxima and minima respectively.) Koy. Soc. Proc. vol. 71, Plate 4. 1900-0 1900-0 ME “Su Su ME Lockyer and Lockyer. Roy. Soe. Prov, vol. 71, Plate 5. 1860-0 1870-0 1880-0 1890-0 1900-0 MEAN DAILY !500 2500 SUNSPOTS. 229° WHOLE HEM. MEAN DAILY AREA. 1200 1000 SUNSPOTS. go S.HEM. 1400 1200 SUNSPOTS. 1000 MEAN DAILY 809 AREA. 600 N. HEM. 400: 200 le) PROMINENCES.’-~ O-20° 500 N.&S. EATS. 250 (TACCHINID MEAN DAILY '& RANGE OF MAGNETIC DECLINATION. & (ELLIS) 1860:0 1870-0 1880-0 1890-0 1900-0 Comparison of curves representing variations of magnetic declination, solar prominences (0°—20° N. and 8.), and sun-spot areas. (Continuous and broken yertival lines as in Plate 4.) 1903. ] Electric Waves round a Conducting Obstacle. 251 27 “The Bending of Electric Waves round a Conducting Obstacle. By H. M. Macponat.p, F.R.S., Fellow of Clare College, Cam- bridge. Received January 21,—Read January 29, 1903. 1. The mathematical theory of the formation of a shadow when waves impinge on an obstacle rests on an application of Huygens’ principle, which may be stated in the form that, if a closed surface S be drawn, enclosing all the sources of the waves, the circumstances that obtain at any point outside this surface at a definite time can be expressed in terms of the state of affairs at the surface S at previous times. For waves of sound, the usual analytical expression involves a knowledge ot both the velocity potential of the motion and the velocity normal to the surface S at each point of it for all time ;* for electric waves, which may be taken to include waves of light, it requires a knowledge of both the electric and magnetic forces tangential to the surface S for all time.t In the application to the theory of shadows,t{ the surface $ that is chosen coincides in part with the surface of the obstacle, the remaining part being chosen so as to simplify the calculation as much as possible ; for example, in the problem of the passage of waves of light through an aperture in a plane screen, the surface 8 is taken to be the plane of the screen. It is then assumed that the part of the surface S which coincides with the surface of the obstacle contributes nothing, and it follows that, when the wave-length is small, a shadow is formed, whose boundary is determined by the extreme incident rays that meet the surface of the obstacle. The assumption thus made is equivalent to assuming that the obstacle is perfectly absorbing, and for waves of light incident on opaque bodies this is known to be approxi- mately true, with possible exception for the case of opaque bodies whose surfaces are polished. For waves of sound incident on an approxi- mately rigid obstacle, and for electric waves incident on an approxi- mately perfectly conducting body, this theory does not apply: as, in the first case, the condensation does not vanish at the surface of the obstacle; and, in the second case, the tangential magnetic force does not vanish at the surface. In what follows the behaviour of electric waves incident on a. per- fectly conducting body will be discussed, and the conditions necessary for the formation of a shadow in this case will incidentally appear. The results for waves of sound incident on a rigid obstacle are very similar. | 2. It will be sufficient to consider a comparatively simple case, * Lord Rayleigh, ‘ Theory of Sound,’ vol. 2, § 293. + Macdonald, ‘ Electric Waves,’ § 14. t~ Cf. Stokes, ‘Camb. Phil. Trans.,’ vol. 9, 1849, p. 1; Lorenz, ‘Pogg. Ann.,’ p. 111, 1860; Kirchhoff, ‘ Berlin Sitzungsberichte,’ vol. 2, 1882, p. 641. 252 Mr. H. M. Macdonald. The Bending of —[Jan. 21, which is of some importance on account of its application to the propa- gation of electric waves along the surface of the earth. The case is that of a Hertzian oscillator placed outside a perfectly conducting sphere. Let the radius of the sphere be a, and let the oscillator be at the point C, whose distance OC from O the centre of the sphere is 7, the direction of the axis of the oscillator being along the line OC. The lines of magnetic force are circles, whose centres lie on OC, and whose planes are perpendicular to OC. If y denotes the magnetic force at any point P whose distance from OC is p, yp satisfies the differential equation £7) ~ = ye) a0) + ep = 0, pP where 2 is the distance of the point P from some plane of reference perpendicular to OC, and 27/« is the wave-length of the oscillations. Transforming to polar co-ordinates (7, 6), where 7 = OP and @ is the angle COP, this on becomes uc 5 Pepe Pie En Spi yp 1) HOR (1), in which » = cos @. The general solution of this equation, which is applicable to the space external to a sphere, is 2 s 9 Nar Nes m2 Andn+s (x7) + Br J —n—3 (xr) } (1 — p*) Fale ) pe in which J,, (x7) denotes Bessel’s function of order m and P, (4) the zonal harmonic of integral order n. It is, therefore, first necessary to express the magnetic force due to the oscillator in this form. If yj; is the magnetic force at the point P due to the oscillator, y; is the real u(R—Vt) . i f part of of “in which R is the distance CP and V is the r : velocity of radiation.* Writing 4 y 0 e—kR and remembering that R? = 7?+7,?- 277ryp, this is equivalent to yy =p ay, he (21 + 1) GURU Syl) (eK7"1) Jn+h (Kr) lig (2) > Pp when 7<7, where in Gee) = * Hertz, ‘ Electric Waves, Hng. Trans., p. 141. + Macdonald, ‘ Proc. Lond. Math. Soc.,’ vol. 32. ee | Ln («?’) ~ emmy, (Kr )]. T 2 sin ma 1903.] Electric Waves round a Conducting Obstacle. 953 Now 6) oe, = (i — »a[re a “1g. OPn+1 ol ee 1 2 1 Vi PNT ein A eg (Qn+1)P, = oh i, (2n + Dig. = no + (n+ +1, *. Te making these substitutions and rearranging the series, it becomes 2. Te 0 2 Wy, = ry 2e™4* > [ee DE Kn y (oer) ent) 1 r neat A 4 ng ihe Oe d at a (n we l)r 2 Jn—3 (7) zm ee ) 2 Ky+s (Kk?) T Bh one s(Kr) . atte 2) 7 Ti (a) b Ja- a = that is Beek J py (1 ) oP n 3 4 Val i = Galt i) a+3(Kr) ( — ps”) Be ti <7; Kefoelsieterelepeteks (2), where In (71) SES = Kr 73 em/4 {ee >K,,_3(uxry) + ent V5 Ki 42(ex7;)}. If, then, a solution ¥ of equation (1) can be found, which is such that w becomes infinite as 7, at the point (7,, 0) and a vanishes when 7 = a, the real part of Cye*V! will be y, the required magnetic force, for then : (yp) will vanish when 7 = a, that is, the electric force ‘ tangential to the sphere vanishes. The solution required will be of the form Ye ae ~ gn{7) [Jn+3(K7) + AnKn+3(x7r)] @isaH2) aE ’ Op where 7; > 7 >a, and the constants A, are determined by the con- dition that CL = 0, when + = a; hence Se 2 | {a In+4(Ka)} : a P= PS galery) Insalo)— G——— Kaa (oor) [1-08 Pa, aC Kia, (exa)} Of where 7) >7 >a. bo 54 Mr. H. M. Macdonald. The Bending of — [Jan. 21, : The calculation of the electric forces at a point not on the surface of the sphere presents difficulties, but when the wave-length is small compared with the radius of the sphere, the electric force at the surface of the sphere can be obtained in a simple form. The electric force at the surface of the sphere is normal to the surface, and denoting it by F, Ae is the real part of or whens =a. Now V2 dt a2 Op Le OC 2 WR 8 K42(cxa cage weve — ee > In (71) Jin 4-1 (KC) - ae ) 0 a ay {at I n43(Ka)t 12 ; a Oey, cee a Raat) ee | f] | and, when the wave-length is small compared with a, this becomes oP, 14 («a)} | (1 — p?) a C Tz ants Soar) | dT n+3 (Ka) that is, writing Y, = f(r), ‘ 1 22 [rosie rot] a Cua [= = From the above 0 ‘ eK t() = Ohi AR ? 0 ee where Ro? = a? +7)? — 2ar;p, that is E OR, ORo 0 eK, Fo) = 0-0) (0 So - ee ae BA), Cer Rp. <-ORy Ao or, fw) = therefore A pO gn zt alu a fe 0 ¢ cag 0a oO 31 a Ghar a ral ze Ro ORo Ro uk Ot Ro OR, Ry a. 1 oF 2 @ a? © cos x(R,- Vé) V2 O° a? op fa ri) eae aces i 0 a2 sin eV) K ou RS at th. 1903.] | Lilectric Waves round a Conducting Obstacle. 250 and therefore wilt wl cv ce) [ 22) a? 0 sin x(Ro -— Ve) Ka Ro OR» vo 10 @ d cos ve = Vi) | Sa ae | iiivell de nasa (3). K Ot Ro OR» At points on the surface of the sphere for which Ro is great com- pared with the wave-length, this becomes, retaining only the most important terms, P= - ov (1 25H a 2) sin «(Ro— Va), 0 ie ( 1. sal ih eG = GOS! Birra ee aia 0.) (4), “0 where x is the angle subtended by OC, and F, is the electric force along the normal, which would be due to the oscillator if the sphere were absent. From this it follows that the ratio F/F, gradually diminishes as m decreases, until » approaches the value — 1, when it becomes com- parable with the terms which have been neglected. Hence, when electric waves are incident on a perfectly conducting sphere, there is no true shadow near the surface when the wave-length is small compared with the radius of the sphere. It can, therefore, be inferred that, when electric waves are incident on a perfectly conducting body whose surface is convex, and has its radii of curvature everywhere great compared with the wave-length, there is no true shadow near to the surface. It is known that, when electric waves of small wave-length are incident on a perfectly conducting wedge, the disturbance does not sensibly creep round the corner, but shoots out so that there is a shadow which coincides the more closely with the geometrical shadow as the wave-length diminishes.* It therefore appears that the condition for the formation of a distinct shadow near the surface of a perfectly conducting body, whose surface is convex, when electric waves are incident on it, is that there should be a line on that part of the surface inside the geometrical shadow along which the radius of curvature of the surface in the plane of incidence of the waves is small compared with the wave-length. 3. The electric force normal to the surface of a sphere, which is a fairly good conductor, may be obtained by an analysis similar to that given above. The result is lp OG a a @ Orn @ 148 Op Wor < Ao); * Spninferteld. ‘Math. Annalen,’ vol. 47, 1896; or Macdonald, ‘ Electric Waves,’ 1902, p. 187. ty p 256 Mr. H. M. Macdonald. The Bending of —[Jan. 21, where « = ox/47V, and yjp is the real part of f(r), and o is the specific resistance of the material of the sphere, the other symbols having the same meaning as before; it being assumed that ¢, which for ordinary metallic conductors is of the order 10~" when the wave-length is about 10 cm., is small. If F’ now denotes the electric force normal to the surface, EF’ differs in phase from F (the electric force normal to the surface when the sphere is a perfect conductor) by a small amount, and the ratio of the amplitude of IF’ to F is 1—Je?. The effect of imperfect conduction is therefore to diminish the electric force normal to the surface, but only by an inappreciable amount when the obstacle is as good a conductor as an ordinary metal ; for sea water, taking x = 10), the correction is less than one part in a thousand.* 4. The effect of a rigid spherical obstacle on the waves of sound sent out from a source, when the wave-length is small compared with the radius of the sphere, can be obtained by an analysis which s almost identical with that given above. The result is that at any point on the sphere at a distance from the source great compared with the wave-length, b = $i(1-“ 3) = $1 - c08 3) 0 where ¢; is the velocity potential at the point due to the source, and is the actual velocity potential there. There is, therefore, no true shadow near the surface of the sphere. Lord Rayleigh? has discussed the effect of a rigid sphere on the waves sent out from a source close to the surface, and found that there was no indication of the forma- tion of a shadow for wave-lengths greater than half the circumference. The above statement completes the investigation. The conditions for the formation of a distinct shadow, when waves of sound are incident on an approximately rigid obstacle, follow; they are of the same type as those already stated for an approximately perfectly conducting body on which electric waves are incident. 5. The results of § 2 have an immediate application to the question of the propagation of electric waves around the surface of the earth. Let C be a place on the earth’s surface from which waves are being sent out; these waves may be supposed to be due to an oscillator placed vertically. ‘The electric force acting on a receiver at a place, whose angular distance from C measured along a great circle is 6, will be F, given by equation (4) § 2, when the distance of the receiver from the oscillator is great compared with the wave-length. In this case 7; is nearly equal to a, and may ‘be put equal to it for values of 6 for which equation (4) applies; thus F = F,(1—sin3@). It is * For the waye-lengths actually used the correction is less than one part in a hundred millions. + ‘Theory of Sound,’ vol, 2, § 328. 1903. ] Electrie Waves round a Conducting Obstacle. 257 convenient for purposes of comparison to substitute for F, in terms of another quantity ; let F, denote the electric force due to the oscillator at a point in its equatorial plane at a distance @@ from the oscillator, which is the same as the arcual distance of the receiver ; then F _ Geos? Z9(1—sin 30) _ FE, 2 sin $6 k, where F is the amplitude of F, and F, the amplitude of F2; the ratio of the intensities in the two cases is 47. The following table shows the manner of variation of the amplitudes and the intensity (£?) near the sphere as @ increases from 20° to 120° :-— @ 1—sin g k ke s 20° 0 82635 0 °80551 0 64885 25° 0 °78356 0 73567 0 °54121 30° 0°74118 0 -69949 0 °48929 | 35° 0 °69929 0 °64546 0 °41661 40° 0 °65797 | 0°59297 0 °35162 45° 0°61731 | 0 *54070 0 29235 50° 0°57738 0 -48964. 0 -23975 | 55° 0 °53825 0 -44019 0°19377 60° 0°5 0 °39269 0°15421 | 65° 0° 46270 0 34745 0°12072 70° 0 *42642 0°30473 0 °09286 75° 0°39127 0 26477 | 0 :07010 | 80° 0°35721 0 °22766 | 005183 85° 0: 32440 0 °19343 0 -03741 90° 0 29289 016266 | 0 :02645 95° 0 26272 0°13483 | 0°01818 100° 0 °23395 0:11011 0 -01212 105° 0 -20664 0 °08844. 0 -00782 110° 0:°18084 0 -06972 0 -00486 | 115° 0 -15660 0 :05379 0-00289 120° 0°13397 0 ‘04050 0 00164 For example, when 6 = iz, that is for the case of the earth at a distance of rather more than 3000 miles, the amplitude of the electric force acting on the receiver is more than half the amplitude of the electric force that would be directly due to the oscillator at that distance, and the intensity nearly three-tenths. These results will apply when the two places are separated by good conducting material such as sea water, the effect of the imperfect conduction of such substances being by § 3 negligible. They explain why wireless telegraphy is more effective over the sea or wet soil than over dry soil; from § 3 it follows that a badly-conducting obstacle diminishes the effect. It is also to be expected from § 2 that the influence of a VOL. LXXI. U 258 Prof. F. O. Bower. Studies in the [Jan. 30 ridge of some sharpness between the places is to create a distinct shadow, to such an extent that the effect would be inappreciable ; the same result would be produced by an intervening headland; this agrees with the experience of Captain Jackson.* “ Studies in the Morphology of Spore-producing Members.—No. V. General Comparisons, and Conclusion.” By F. O. Bowrr, Se.D., F.R.S., Regius Professor of Botany in the University of Glasgow. Received January 30,—Read February 12, 1903. (Abstract.) This concluding Memoir contains a general discussion of the results acquired in the four previous parts of this series, and of their bearing on a theory of sterilisation in the sporophyte. The attempt is made to build up the comparative morphology of the sporophyte from below, by the study of its simpler types; the higher and more specialised types are left out of account, except for occasional com- parison. It is assumed for the purposes of the discussion that alter- nation of generations in the Archegoniate is of the antithetic type, and that apogamy and apospory are abnormalities, not of primary origin. After a brief allusion to facts of sterilisation in the Sporogonia of Bryophytes, the similar facts are summarised for the Pteridophytes. It has been found that examples of sterilisation of potentially sporo- genous cells are common also in vascular plants, while occasionally cells which are normally sterile may develop spores. Hence it is concluded that spore-production in the Archegoniate plants is not in all cases strictly limited to, or defined by, preordained formative cells, or cell-groups. A discussion of the archesporium follows, and though it is found that in all Pteridophyta the sporogenous tissue is ultimately referable to the segmentation of a superficial cell, or cells, still in them, and indeed in vascular plants at large, the segmentations which lead up to the formation of spore-mother-cells are not comparable in all cases ; in fact, that there is no general law of segmentation under- lying the existence of that cell or cells which a last analysis may mark out as the “archesporium”; nor do these ultimate parent cells give rise in all cases to cognate products. Therefore it is concluded that the general application of a definite term to those ultimate parent cells which the analysis discloses has no scientific meaning, beyond the statement of the histiogenic fact. Further, it is shown that the tapetum is not a morphological constant, = Roy SOc wbroe aL O02: 19035. ] Morphology of Spore-producing Members. 299 but varies both in occurrence and origin; that even the individuality of the sporangium is not always maintained. All that remains then as the fundamental conception of the sporangium in vascular plants is the spore-mother-cell, or cells, and the tissue which covers them in, for such cells are always produced internally. The definition of the sporangium may then be given thus: “ Wherever we find in vascular plants a single spore-mother-cell, or connected group of them, or their products, this with its protective tissues constitutes the essential of an individual spor- angium.” From the point of view of a theory of sterilisation such sporangia may, at least in the simplest cases, be regarded as islands of fertile tissue which have retained their spore-producing character, while the surrounding tissues have been diverted to other uses. It will be seen later how far this view will have to be modified in the more complex cases. In a second section of the Memoir the variations in number of sporangia in vascular plants are discussed; the methods of variation may be tabulated as follows, under the heads of progressive increase and decrease :— I.—-Increase in Number of Sporangia. (a.) By septation, with or without rounding off of the individual sporangia. ().) By formation of new sporangia, or of new spore-bearing organs, which may be in addition to, or interpolated between those typically present. (c.) By continued apical, or intercalary growth of the parts bearing the sporangia. (d.) By branching of the parts bearing the sporangia. (e.) Indirectly, by branchings in the non-sporangial region resulting in an increased number of sporangial shoots; this is closely related to (c) and (d). Il.-—Decrease in Number of Sporangia. (f.) By fusion of sporangia originally separate. (g.) By abortion, partial or complete, of sporangia. (h.) By reduction or arrest of apical or intercalary growth in parts bearing sporangia. (.) By fusion of parts which bear the sporangia or arrest of their branchings. (j.) Indirectly, by suppression of branchings in the non-sporangial region, resulting in decreased number of sporangial shoots ; this is closely related to (h) and (2). We are justified in assuming that (subject to the possibility of other factors having been operative of which we are yet unaware) u 2 260 Prof. F. O. Bower. Studies in the [Jan. 30, the condition of any polysporangiate sporophyte as we see it is the resultant of modifications such as these, operative during its descent. | The problem will, therefore, be in each case to assign its proper place in the history to any or each of these factors. It is pointed out that in homosporous types, which are certainly the more primitive, the larger the number of spores the better the chance of survival, and hence, other things being equal, increasing numbers of spores and of sporangia may be anticipated; but in the -heterosporous types reduction in number both of spores and of sporangia is frequent. The former will accordingly illustrate more faithfully than the heterosporous forms the story of the increase of complexity of spore-producing parts. The general method put in practice here is to regard homosporous forms as in the upgrade of their evolution, as regards their spore-producing organs, unless there is clear evidence to the contrary. The onus probandi lies rather with those who assume reduction to have taken place in them. A summary of evidence of variation in number of sporangia by any of these methods is then given for ,the Lycopodiner, Psilotacez, Sphenophyllex, Ophioglossacez, Equisetinez, and Filicinee ; followed in each case by a theoretical discussion of the bearing of that evidence on the morphology of the spore-producing members. The general result is that all of them, including even the dorsiventral and megaphyllous types, are referable to modifications of a radial strobiloid type ; progressive elaboration of spore-producing parts, followed by progressive sterilisation, and especially by abortion of sporangia in them, of which there is frequent evidence, together with the acquire- ment of a dorsiventral structure, may be held to account for the origin of even the most complex forms. But the vegetative organs once formed may also undergo elaboration, and differentiation parr possu with the spore-producing organs, a point which has greatly com- plicated the problem, especially in the higher forms; all roots are probably of secondary origin; facts of interpolation of additional sporangia, especially in Ferns, and of apogamy and apospory, are also disturbing influences, which have probably been of relatively recent acquisition. A comparison is drawn as regards position, physiological and evolutionary, in the sporophyte between the fertile zone in certaim Bryophytes and the fertile region of certain simple Pteridophytes, v.g., the Lycopods; though no community of descent is assumed, the relation of the reproductive to the vegetative regions is the same. In the Bryophytes that region is regarded as a residuum from progressive sterilisation ; it is suggested that the same is the case for a strobiloid Pteridophyte, such as Lycopodium. The theory of the strobilus, based on this comparison, is that similiar causes would lead to 1903.] Morphology of Spore-producing Menbers. 261 the decentralisation of the fertile tissue in the primitive Pteridophytes as in the Bryophytes, and result in the formation of a central sterile tract, with an archesporium at its periphery ; that such an archesporium, instead of remaining a concrete layer as it is in the larger Musci, became discrete in the Lycopods; that the fertile cell-groups formed the centres of projecting sporangia, and that they were associated recularly with outgrowths, perhaps of correlative vegetative origin, which are the sporophylls. Whether or not this hypothesis of the origin of a Lycopod strobilus approaches the actual truth, comparison points out the genus Lyco- podium as a primitive one, characterised by more definite numerical and topographical relation of the sporangia to the sporophylls than in any other type of Pteridophyta. Then follows, as a consequence of conrparison, the enunciation of a theory of the sporangiophore, a word which is here used in an ex- tended sense to include not only the spore-producing organs of Psilotaceze, Sphenophyllee, Ophioglossacez, Equisetacez, but also the sori of Ferns. The view is upheld that all these are simply placental growths, and not the result of ‘‘metamorphosis” of any parts or appendages of prior existence ; that the vascular supply, which is not always present, is not an essential feature ; that they are seated at points where in the ancestry spore-production has been proceeding on an advancing scale; hence they do not occupy any fixed and definite position. It seems probable that at least a plurality of sporangia existed on primitive sporangiophores, and that where only one exists that condition has been the result of reduction. The above theories are then applied to the several types of Pterido- phyta. The Lycopods, Psilotacez, Sphenophyllez, and Ophioglossacez ‘may be arranged as illustrating the increased complexity of the spore- ‘producing parts, and of the subtending sporophylls; the factors of the advance from the simple sporangium to the more complex spor- angiophore are, septation, upgrowth of the placenta with vascular supply into it, and branching, with apical growth also in the Ophio- -glossaceze. But even in the most complex forms the sporangiophore may be regarded as a placental growth, and not the result of transfor- mation of any other member. In the case of Helminthostachys the marginal sporangiophores are regarded as amplifications from the sunken sporangia of the Ophio- glossum type ; in Lquisetum they are regarded as being directly seated -on the axis, and having originated there by a similar progression : they would thus be non-foliar. It is pointed out that though a foliar theory would be possible for Equisetwm itself, it is not applicable to the facts known for the fossil Calamariez, which are so naturally related to it. Thus the strobilus of the Equisetinee is of a rather different type from that of the Lycopods, Psilotacew, or even the 262 Prof. F. O. Bower. Studies in the [Jan. 50, Ophioglossacez, in all of which there is a constant relation of the spore-producing parts to the leaves ; in the Equisetineze no such con- stant relation exists ; the leaves and sporangiophores may be in juxta- position, as in Calamostachys, without exactly matching numerically ; or the sporangiophores may occur in larger numbers and in several ranks, between successive leaf-sheaths, as in Phyllotheca and Borna ; or without any leaves at all, as in Hgwisetwum. Thus, on a non- phyllome theory the latter may be held to be only an extreme case of what is seen in certain fossils. The Ferns, notwithstanding their apparent divergence of character from other Pteridophytes, may also be regarded as strobiloid forms, with greatly enlarged leaves; the primitive sori of the Simplices resemble the sporangiophores of other Pteridophytes ; the more com- plicated soral conditions of the Gradatze and Mixtz were probably clerivative from these, the chief difference being due to the interpolation of new sporangia, an innovation which is in accordance with biological probability, as well as with the palzontological record. The effect of the results thus obtained on the systematic grouping of the Pteridophytes is then discussed; it is pointed out that the Liycopods, Psilotacee, Sphenophyllee, Ophioglossacez, and Filices illustrate lines of elaboration of a radial strobiloid type, with increas- ing size of the leaf. The division of Pteridophyta by Jeffrey, on anatomical characters, into small-leaved Lycopsida, and large-leaved Pteropsida is quoted ; but it is concluded that the anatomical distinc- tion of Jeffrey does not define phylogenetically distinct races, but is vather a register of such leaf-development as differentiated them from some common source. It is contended that the Ophioglossacez and Filices, which constitute Jeffrey’s Pteropsida, are not necessarily akin on the ground of their large leaves, and consequent phyllosiphonic structure ; but that they probably acquired the megaphyllous character along distinct lines. The opinion of Celakovsky is still held, “ that the Lycopods are probably of living plants, the nearest prototypes of the Ophioglossacex.” The more recent investigations of Jeffrey, and of Lang, have shown, however, that in the gametophyte of the Ophio- glossacee, there is an assemblage of “ Filicinean” characters, which differ from those of Lycopodium itself. But Celakovsky’s comparison is with the Lycopods, not with the genus Lycopodium ; so far as the facts go, increasing “ Filicinean” characters of the gametophyte follow in rough proportion to the larger size of the leaf; thus from Jsoctes we learn that a combination of cross characters is found in a mega- phyllous Lycopod type. What we find in the Ophioglossacez is that in conjunction with their more pronounced megaphyllous form, still retaining, however, the Lycopodinous type of the sporophyte, they show more pronounced ‘ Filicinean” characters of the gametophyte, and of the sexual organs. Jt is unfortunate that the facts relating to 1903. ] Morphology of Spore-producing Members. 263 the gametophyte of the Psilotaceze and Sphenophyllez are not avail able in this comparison. It is not obvious what the meaning is of this parallelism between leaf-size and characters of the sexual organs; a further difficulty in its interpretation lies in the fact that for the Equiseta the parallel- ism does not hold; there ‘“ Filicinean” characters of the gametophyte accompany entirely non-Filicinean characters of the sporophyte, the latter showing nearer analogy to the Lycopods than to the Ferns. Such cross characters are difficult to harmonise with any phylogenetic theory ; on account of them, the Equisetinez are placed in an isolated position, and in the same way, though with less pressing grounds, a separate position should be accorded to those types which lie between the extremes of Lycopods and Ferns, in proportion as the characters are more or less pronounced. On this basis the Isoetaceze would probably best take their place as a sub-series of the Lycopodiales, Ligulate ; the Psilotaceze and Spheno- phyllez would constitute a series of Sphenophyllales, separate from, but related to, the Lycopodiales. The Ophioglossaceze would form an independent series of Ophioglossales, more aloof than the latter from the Lycopodiales, but not included in the Filicales. The actual con- nection of these series by descent must remain open; it is quite possible that some or all of them may have originated along distinct lines from a general primitive group, which may be provisionally designated the Protopteridophyta ; these were probably small-leaved strobiloid forms, with radial type of construction, and with the sporangia disposed on some simple plan. The grouping arrived at in these Memoirs may be tabulated as follows :— PTERIDOPHYTA. I. LYCOPODIALES. (a) Hligulatee. Lycopodiace. (L) Ligulatee. Selaginellaceze. Lepidodendracex Sigillariaceze. Isoetacez. II. SpENOPHYLLALES. Psilotacez. Sphenophyllacex. III. OPHIOGLOSSALES. Ophioglossacex. 264 Dr. N. H. Alcock. On the Negative Variation [Jan. 17, TV. FILICALES. (a) Simplices. Marattiacee., Osmundacee. Schizaeaceze. Gleicheniace. Matoniner. (0) Gradatee. Loxsomacer. Hymenophyllacee. Cyatheacee. Dicksoniez. Dennstaedtiine. Hydropteridez (?). (c) Mixtee. Davalliew. Lindsayeee. Pteridez, and other Polypodiacer. V. EQUISETALES. Equisetacee. Calamariex. “On the Negative Variation in the Nerves of Warm-blooded Animals.” By N. H. Atcocx, M.D. Communicated by A. D. Water, M.D., F.RS. Received January 17,—Read February 12, 1908. , (From the Physiological Laboratory, University of London, 8.W.) Introduction. The negative variation in the nerves of warm-blooded animals has already been the subject of several researches.* While the nerves are still in connection with the tissues it has been the experience of most observers that there is no difficulty in examining the negative variation, * Valentin, ‘ Pfliiger’s Archiv,’ vol. 1, p. 423; Fredericq, ‘Du Bois Archiv,’ 1880, p. 70; Hermann, ‘ Physiologie,’ vol. 2, p. 120; Gotch and Horsley, ‘ Phil. Trans.,’ “ Croonian Lecture,” 1891, p. 267 ; Macdonald and W. Reid, ‘J. Physiol.,’ vol. 28, p. 100; Waller, ‘Animal Electricity,’ London, 1897; Boruttau, (a) ‘Centralbl. f. Physiologie,’ vol. 12, p. 317, 1898, (>) ‘ Pfltiger’s Archiv,’ vol. 84, p. 309, 1901. 1903. ] in the Nerves of Warm-blooded Animals. 265 but with regard to isolated nerves contradictory statements have been made, and it was to ascertain if possible the reason of this discrepancy that the present research was undertaken. Methods. The following method was employed in all the experiments here quoted. The animal was killed by decapitation, and the body left undisturbed for 30—45 minutes. The nerves were then dissected out, placed in a 1:05 per cent. NaCl solution at about 30° C., and kept at this temperature for about half-an-hour or more. They were then allowed to cool to room temperature (17—19° C.), and it was found that, as a rule, the negative variation of nerves so treated was of the order of 1 millivolt (wide infra). This, of course, is not an absolute value of the true P.D. between active and inactive parts of nerve, but only a fraction of it, depending upon the amount of internal derivation in a nerve trunk by indifferent conducting tissue. From about 2 to 6 hours post-mortem this value remains at a fairly constant amount, for instance, in Experiment 5b‘, the sciatic of the rabbit was used 5 h. 30 m. post-mortem, and gave a negative variation of 000083 volt; in Experiment A*, 2 h. 40 m. post-mortem the value was 0°00076 volt, and these are typical instances. The table on p. 277 gives the result of twenty-two experiments, in which the values were taken for heat determination, which illustrates this. Certain points may be here noted. The practice of placing nerves in salt solution for some time before use has been employed by Waller,* Gotch,t and Boycott, in the case of the frog.§ The effect of changes in the composition of the salt solution is the subject of another research which I hope to publish at a future time ; however, I may here state that small differences in the concentration of the solution—e.g., +0:1 per cent. NaCl—make no apparent differ- ence in the condition of the nerve, and the same is true in the main of small differences in the reaction, and of small differences of tem- perature. Waller|| has pointed out that the presence of lactose in the solution is of advantage, and taking a greater value of the negative variation for a sign of greater real eee. the same appears to ie true in mam- malian nerve for maltose and glucose, though I make the statement at present with some reserve. For instance :—Experiments B® and BY. * Waller, ‘ Brain,’ vol. 73, 1896, p. 43, et seq. + Gotch, ‘J. Physiol.,’ vol. 28, p. 32. ag Boycott, loc. cit. K See also Gotch and Horsley, Joc. cn Macdonald and Reid, loc. cit. |) Waller, loc. cit. (Lectures), p. 73. i i| 266 Dr. N. H. Aleock. On the Negatwe Variation [Jan. 17, Young Rabbit. Nerves kept for 3” circa post-mortem. | Volt. of Volt. of Experi- Temp. of neg. var. neg. var. NEES ment No. nerve. at excit. at excit. nes 100. 30. Bre 30°C. | 0:00033 | 0:00023 {1°05 p. ce. NaCl + 0°5 p. c. maltose. R. sciatic. | | 1:05 p. c. NaClonly. Bt | 380°C. | _0:00023 0:00016 | L. sciatic. The two nerves of opposite sides are here compared, and the com- parison is in favour of the R. sciatic which had been treated with maltose. No account is here possible of the precautions used to. exclude fallacy, but many further experiments support the one quoted, and it is at least very probable that the maltose is the active variant. The method employed for determining the negative variation and the action of anzsthetics was in all cases that of Waller. An addi- tional larger box was used, with a false bottom of wood, on which the nerve chamber rested. Below this was a layer of water, so that the nerve was kept in an atmosphere nearly saturated with water vapour at whatever temperature was desired. The standard “ Berne” coil, worked with two Leclanché cells, was. used to give the excitation, which always consisted of tetanising shocks. for a period of 13 seconds, repeated once a minute. The number of units used was 500, when not otherwise stated. The usual precautions against current escape and electrotonus were carefully observed ; this. was found to be particularly necessary in the case of bird’s nerve. Half-grown rabbits were found very suitable animals to use, as the small amount of connective tissues permitted the nerves to be dissected out with a minimal amount of injury. Experiments were also made with cats, kittens, guinea-pigs, hedgehogs, pigeons, and frogs. The technique is considerably easier inthe case of young animals ; adults, however, answer well if care be taken. In the latter, and especially in the nerves of birds, it is advisable to work at rather higher tem- peratures than those given above. Voltage and Strength of Lactation. The voltage of the negative variation varies with the animal, the nerve employed, the temperature and condition of the nerve, and, within certain limits, with the strength of excitation. 15% )\. i eRe 1905.) = «inthe Nerves of Warm-blooded Animals. 267 The question of temperature will be considered later. Assuming for the present purpose that the conditions of experiment were equally favourable throughout, and taking the sciatic nerve as a standard, different animals gave the following values :— Millivolts. Maximum observed. Mean. IRADDIG) stele. siecie oe" Leal 0-69 Mean of 11 experiments. EEO 2600) ose scie-| 0°66 0°50 S 5 Bs ENSECOM. 66 800s ss 1°05 0°42 ma 5 fs iL e@ 0°89 i 2 i | Guinea-pig..... | | All these are very much less than the frog, which gives 2 millivolts or more. Different nerves in the same animal often show individual inequali- ties, but as a rule the larger nerves give a smaller negative variation than those of less diameter. The sciatic gives commonly the least, but is the most resistant to adverse influences. The median and ulnar nerves are more delicate, but give larger variation under favour- able circumstances, ¢.g., the median nerves of the pigeon gave a mean value of 0°54 millivolt (five experiments) as against 0-42 for the sciatic. The greatest value I have yet measured was in a branch ot the anterior crural of the rabbit, which on the right side gave 2°5 milli- volts and on the left 2:3. Boruttau* found the vagus in the rabbit to give a larger negative variation than the sciatic, and obtained only very small responses from the nerves of hens, ducks, or pigeons. The larger number of fibres not in contact with the longitudinal electrode would appear to act as a deriving circuit of less resistance in the larger nerves, and so less current passes through the galvano- meter, and the greater amount of connective tissue in the sciatic would have the same effect. It is possible that there are other causes in addition to these ; there is at present no evidence for or against such a possibility. Similar reasons probably also explain why a stronger stimulus is necessary for the warm-blooded nerves than for the frog. The difference is, however, not very great. The minimal effective excita- tion I have so far observed is 6 units of the “ Berne” coil, 500 units is commonly a maximum, 1000 nearly always so. The smaller electrical * Boruttau (7), loc. cit. 268 Dr. N. H. Aleock. On the Negative Variation [Jan. 17, resistance of the mammalian nerve between the exciting electrodes must be borne in mind in these comparisons. Gotch* has recently stated that in determining the sub-maximal response of frog’s nerve, the excitation of a smaller number of fibres is a far more potent cause than the varying response of each fibre, and it seems very probable that the higher threshold and wide range of excitation in mammalian nerves is due to the failure of the exciting current to reach the more distant fibres, protected as they are by inter- vening fibres and connective tissue, and not to any essential difference in the nerves. The negative variation commonly persists without great alteration under ordinary conditions for at least 4—8 hours post-mortem. ‘The longest time I have seen was in Experiment C“ (internal popliteal of the hedgehog, 28 hours post-mortem) ; the right and left median nerves the kitten in Experiments C* and C* gave a small and rapidly ‘diminishing response 19 hours post-mortem. The earlier observers (Valentin, Fredericq, Hermann) have stated that they have found the negative variation to persist for days, and to last longer than in frog’s nerve. I am unable to confirm this; even in the hedgehog the nerves are much more short-lived than in the frog under similar conditions, and the phenomena referred to were probably of a different nature to those examined here, viz., electro- tonic spread or ordinary diffusion. Mi | ‘| | | aM nr i ; V. Fre. 1.—Rabbit. R. ext. popliteal. Normal series of negative variations. Exp. B*, Reads from left to right. . * Gotch, ‘J. Physiol.,’ vol. 28, p. 40. 1903.| wthe Nerves of Warm-blooded Animals. 269 Action of Anesthetics. The following is the summary of the observations made on this: subject :— Chloroform. Hxperi- | Negat. : | | ment and ite mae Cl After. | Notes. | plate No. | before. ee ey 1 i000 volt. 204: 0°48 | 0°42 to O co) to 0°42 Kitten. Sciatic. Temp. | 3 min. _ of nerve chamber, 32°. | _ Complete abolition and | recovery. 207 0°38 | 0°38 to O O to 0°22 Kitten. Sciatic. Temp. 5 min. | = 30°C. After CHCl, | a positive variation appeared, changing | again to a negative. Fair recovery. A° 408 0°95 | 0:70 to 0°40 | 0:21 to 0°13 Rabbit. L. sciatic. 6h. 3 min. _ post-mortem. Temp. | = 20°. Gradual pro- | | gressive diminution and | norecovery. FIG. 3. Cee 424 0°46 | 0°46 to 0°15 | 0°18 to 0°36 | Kitten. L. median. 4 min. | Temp. = 32°. CHCl, dilute at first, stronger | after first minute. B! 417 0°47 | 0°6 to O71 | 0:21 to 0°40 | Pigeon. R. median. 3h. 5 min. | post-mortem. Temp. | 37°. Recovery. | | | | | Ether. i i During ether. | 201 0 -40 0°25 to O 0 to 0°28 Kitten. R. sciatic. | 3 min. Temp. 34°. 37 m. posé- | | mortem. Imperfect | recovery. 203 0°66 0°25 to O 0 to 0°46 Kitten. i. sciatic. 4 h. 3 min. post-mortem. Temp. = 28°. Recovery. BP 413 0°54 0-22 to O 0 to 0°54 | Rabbit. Ext. popliteal. 3 min. 2h. 30m. post-mortem. Mempe —130r- Re- | covery. FIG. 2, Les 4:6 0°83 — to 0 0 to 0°51 Pigeon. 2h. 30 m. post. | | 2 ymin. mortem. Temp. = 39° | C. UL. sciatic. Re- | | covery. | 270 Dr. N. H. Aleock. On the Negative Variation [Jan. 17, es Fie. 2.—Rabbit. Same nerve as fig. 1. Ether vapour. Exp. B®. Fie. 3.—Rabbit. LL. sciatic. Chloroform vapour. Reads from right to left. ix. SA'S. 1903.] -00/ Vv. -O005 Vv. in the Nerves of Warm-blooded Animals. GO. Fie. 4.—Rabbit. R. sciatic. CO,. Exp. A’. Tec. Fic. 5.—Rabbit. R. sciatic. Tetanisation for im. Exp. --OO/ -*- 272 Dr. N. H. Alcock. On the Negative Variation [Jan. 17, CO». | : . |Neg. var | [eaten ihetore | Mee) Neg. var. . | | mentand |. 7. | during a Notes. Panay in 0°001 CO. | after. | plate No. | volt | 9° | | | | ee Ii 2 | | ( | [3 8202 Og59 Opt 0 3, 0°66,0°53| Kitten. L. sciatic. 3h. | | 3 min. post-mortem. Temp. | | = 80°. Primary dimi- | nution and secondary | | augmentation. | A' 406 | 0°39 0:29 to 0°15 | 0°16 to 0°61 | Rabbit. R. sciatic. 3 h. | 4 min. | 45 m. post - mortem. | Temp. 22°C. As 202. Marked secondary aug- | | mentation. FIG. 4, 1} AS42405 =|" O70 0°56 to 0°3 | 0°35 to 0°97 | Rabbit. L. sciatic. Lh. 5 min. 10m. post - mortem. | - As 406. | Bm 418 | 0°32 0°3,0-42,0°23 0-64 to 0°33 | Pigeon. L. sciatic. 3h. | | 6 min. | post-mortem. Temp. | = © Tetanisation. A‘ 409 2°5 | Tet. 5 min. 2°4 Rabbit. Branch of R. i ! ant. crural. Temp. = 27°5. 1 h. post-mor- tem. | 1°8 Rabbit. Do. do. L. | Temp. = 21°°5. A’ 410 2-3 | ,Tet.6 mim: PLT ee eae A 407 | 0°90 | Tet.5 min. ; 0°85 to 0:96 | Rabbit. L. sciatic. | Temp. = 20°. At 411 0°86 Tet. 5 min. 0-92 Rabbit. R. sciatic. | : Temp. = 27°. FIG. 5. | | Note.—0 0005 volt com- | pens. sent in after Tet. | 2h. 30 m. post-mortem. | 0°26 to 0:19 | Kitten. L. ulnar. Temp. } | = 36°5C. 3h. post- | mortem. Cee 428 | 0:26 | Tet 6 min. The values are taken from photographs. The action of these drugs on the nerves of warm-blooded animals is very similar to their effect on the nerves of the frog. Chloroform, ether, and carbon dioxide all produce diminution of the negative variation, followed by recovery in the case of the latter, with recovery or not in the case of CHCls, and the details of the process are clearly to be seen in figs. The only difference that may be detected is that while in the frog the negative variation is increased by “little” CHCl; or Et,O, and the abolition by “much” CHCl; and Et.0 is commonly followed by recovery to or beyond the normal; this increase has only been 1903. ] in the Nerves of Warm-blooded Aninvals. 273 observed in the bird (Experiment 5') and not in the mammal. In the case of CO. this increase is seen in all the warm-blooded nerves ; primary and secondary augmentation are shown in Exp. B™ from the pigeon, and the latter, in fig. 4, from the rabbit. I do not here enter upon any discussion of the actual mechanism of this increase, it may be due to either of two causes—increase of E.M.F. or increased duration of electromotive change. The latter explanation has been suggested by Gotch as being the true one. The effect of tetanisation (fig. 5) has not been marked. Three experiments proved negative, and two gave a slight increase, so that, this question is still undecided. It appears, therefore, certain that neither in the voltage of the negative variation, in the strength of excitation, or in the action of anesthetics is there any marked difference between the warm-blooded and amphibian nerves, and that all the facts ascertained for the latter under these heads can be applied en bloc to the former. Temperature of Extinction by Heat. Three series of experiments on frogs, mammals, and birds were undertaken to ascertain the precise point at which the negative varia- tion was abolished by heat. Jethod.—The nerve chamber was kept at a constant temperature throughout, ¢.g., 30 C. The nerve itself was placed on the electrodes, and when it had reached the temperature of the chamber, the value of 0-001 volt was determined on the galvanometer scale, and then the values of the first six negative variations. The nerve was then removed, placed in 1:05 per cent. NaCl solution (containing Ca salts, &c.) at the desired high temperature (¢.g., 49° C.), left for exactly 5 minutes, placed in cool (18°C.) salt solution for 7 minutes. A fresh transverse section was made, the nerve was replaced on the electrodes, and the value of 0:001 volt and the second set of six nega- tive variations determined. This method fulfils several desiderata. (1.) It is possible to keep the beaker of hot saline solution at any given temperature with an error of less than + 0:1°C. A standardised thermometer was placed in the bath close to the nerve, and with 5 minutes immersion all parts of the nerve reach the temperature of the solution. (2.) Any alteration in the resistance of the nerve is readily detected by means of the standard deflection with 0:001 volt, and as both readings are taken at the same temperature, this alteration must be a permanent one, and not the temporary alteration always seen when a nerve is heated or cooled. (3.) Using a concentration of salt solution*® that had been found to * No carbohydrate was added to the solution in any of the heat experiments. The solution was neutral, VOL. LXXI. ; xX Dr. N. H. Alcock. On the Negatiwe Variation [Jan. 17, work equally wel] with all classes of nerve, and carefully preserving similar conditions of experiment, the results in the different series are strictly comparable not only with each other, but also with Halli- burton’s* researches on the heat-coagulation of the nerve proteids. The excitation was a maximal one throughout. Series I. A. With Laboratory Frogs. B. With freshly-caught vigorous Frogs. Sciatic nerve. Excitation 30, except in A®. Temp. Neg.var. Trrita- Experi- | Hours | of nerve| Temp. | initial | Neg. var.) bility | ment post- | cham- | of hot ee final | quotient | Notes | No. mortem.| her, hath. | 9-001 v.| = 2. NO | GQ | po lio en | | i | Ae 1 0 room 43°35 1-017 |) 10 O \ Af Bona 40°7 0-8 “0 0 | ABs 62.00) ee 40:0 1-2 0 0 LA. PAV 5 30 is ie SoRO LS) O56 0°31 | A& 2 30 i 38 °5 120 | ey, 1°0 J | A’ 6 0 ley Gs 42-1 2°5 8) 0 7 AY 2 40 . 42-0 2°3 0 0 % | AY 3 30 ¢ 411 18 | 0-91 0°51 f | A) 2150: | 0 AOR eames 1:7 series II. Rabbit. Sciatic. Excitation 1000 for first three experiments, last four, 500. | | . e Trrita- | Experi- | Hours | pone Temp. ney va. Neg. var.} bility of nerve initial a ane ment post- | yam. | oF hot ae, final | quotient Notes. No. mortenr. Ries bath. Owns | = b. fe b a oy A VR Sal ze, Ele os | isamas i mS Bfa 6 30 33 *O 48 °5 0°53 0°12? 0°22? | Doubtful re- Bea 3 30 29:0 48 °O 0°57 ®) O turn. Ba 5 30 30-0 Uh 0 0°83 0°25 0°30 Au 4 0 20°95 46 °O 1-1 0°45 0°41 AP 6 35 21°0 44 °3 0°52 0°34 0°65 Ap |b 0 |) 100 | 428) O48 | 0-52) ae i? ae | 245" ago | iae-5 i eoc76) no.ce | O-o1 | | * Halliburton. ‘““Croonian Lecture,” 1891, and below. 1903. ] in the Nerves of Warmz-blooded Animals. 275 Series ITI. Pigeon. Experiments B" and B* Sciatic, all the rest Median. | | Temp Neg. var ) Tene Experi- | Hours | \¢ verve Temp. | + tial Neg. var. bility | l ment | post- | io. | of hot oe final quotient | Notes No. | mortem.| ) | bath. | 9-991 y.| = 2 aay | a | | elngeeants 4 » | ef setae | 25 53°6 | 0°30 0 1 @ | Cold bath accidentally omit- | | | ted. Thenerve was much | | contracted longitudinally | | | after heating. Cre 4 0 30 | 53-0 0°75 0 (pan O Nerve contracted. No “ne- | | gative variation” after heating, but large posi- | | tive current escape ob- | | ) served, not abolished by crushing. Be 4 15 38 52 °5 0°25 O12 | 0°48), Neg. var. rapidly diminish- Jays 4 30 38 52 °0 0°56 OLA ‘Or30 ing. Bi 3 0 37 50°0 0°61 0-18 | 0°30 Bé 74 {U) 30 45 °8 0°30 E05" Fla 3-5 The experiments can be summarised thus : | | | | Normal temp. | Temp. of incr. | Temp of dim. | Temp. ofabol- | | of animal. neg. var. neg. var. | ished neg. var. | en pS 2 ae ee | | Frog — S94 | 39—41° 40—42° | SS | a | Rabbit..... 37—4.1°* 42 -3° 44, °3—47 *7° | 48 —49° | — —— — ——_|—— [Pigeon.....|| 40—42 5+ | 45 °3° 50° 52—53° | | | | It is seen that the effect of heat occurs in three stages. In the first, at a temperature of 1—2° above that of the animal, the negative variation is increased. In the second there is diminution, recovered from at the lower temperatures (4° over normal) if the nerve is cooled longer than the standard time, not recovered from at the higher (6—7° over normal), and finally the negative variation is permanently abolished, 8° over normal in the rabbit, 10° in the pigeon. While the mammalian and avian nerves show quite small individual * Pembrey (Schifer’s ‘ Text-book,’ vol. 1, p. 790). The higher limit for the rabbit is from unpublished observations of Dr. Pembrey, which he has very kindly furnished me for this paper. + Corin and Van Beneden, ‘ Arch. de Biol.,’ Gand., 1887, vol. 7, p. 265. bie: 276 Dr. N. H. Alcock. On the Negative Variation [Jan. 17, differences in different animals as regards their reaction to heat, the frog’s nerve varies a little according to the condition of the animal, and so the observations have been arranged in two divisions. Here one also notices that the “injury range” is very much smaller than in the warm-blooded nerve, 2° at most separating a temperature that has no ill effect for one that finally kills the nerve, as against 5—6° in the mammal and bird. Observations. A summary of the previous work on the effect of temperature on nerves is to be found in Howell’s* paper, and in that of Boycott (loc. cit.). Howell, from his own researches, gives 41—44° as the temperature at which conductivity is abolished in frog’s nerve, the other authors give 45—50°. The difference appears to be due to the methods em- ployed. Hitherto, there has been some difficulty in ensuring that all parts of the nerve shall have the same temperature, and this tem- perature has in most cases been ascertained indirectly, further, the time during which the temperature is kept up and the conditions of moisture, &c., greatly influence the results.T Another explanation is possible. The majority of observerst have examined the conductivity of nerve as opposed to the excitability, and if the two processes are supposed to be distinct, it might be said that the excitability was extinguished before the conductivity. In view of the considerations stated above, and also of the relation to the coagu- lation point of the proteids, this hypothesis does not seem to be well founded. The relationship of the extinction point given above and the coagu- lation point of the proteids in the body of the animal is a very close one. In the frog, the first coagulation of extracted muscle proteid occurs at 40°C,$| the first step in heat rigor of the muscle itself at 38—40°,9 the electrotonic currents are abolished at 40°,** and the extinction point of the nerves as determined above, 40—42° C. In the rabbit the proteid coagulation occurs at 47°,S|| the muscle * Howell, Budgett and Leonard, “J. Physiol.,’ vol. 16, p. 298. + Some earlier experiments I have made under different conditions lend sup- port to these remarks. ~ Except Edwards’ ‘J. Hopkins Lab. Studies,’ vol. 4, 1887, p. 18 (45—48° —55°?); and Moriggia, ‘ Maleschott’s Untersuchungen,’ vol. 14, p. 382 (46—47°). § Halliburton, loc. cit.; also Halliburton and Mott, ‘ Archives of Neurology,’ vol. 2. || Von Fiirth, ‘Arch. f. Exper. Path. u. Pharmak.’ Leipzig, 1895, vol. 36, p- 281, and zdid., vol. 37, 1896, p. 389. © Vincent and Lewis, ‘J. Physiol.,’ vol. 36, p. 445; see also Brodie and Richardson, ‘J. Physiol.,’ vol. 21, 1897, p. 353, and ‘ Phil. Trans.,’ B, vol. 191, 1899, p. 127; and also Vernon, ‘J. Physiol.,’ vol. 24, p. 239. ** Waller, ‘Roy. Soc. Proc.,’ vol. 60, p. 384. 1903. | in the Nerves of Warm-blooded Animals. 277 rigor at 45—50°C,* and the nerves die at 48—49°. The proteids of the cat’s brain coagulate at 47° C.t No data for the bird are available, the nerves die at 53°. In table form. | Frog. Mammal. | Bird. | ' i a pra eee i lor ee ees ep ema | Muscle proteid (Halliburton and von Firth), 40° | 47" = | Muscle rigor (Vincent and Lewis)......... ie 38-40" | 45— 43") — | Nerve proteid (Halliburton).. vee! —~ te — | Nerve electrotonic currents (Wailer) . | AO? too os — _ Nerve (present experiments).............. | 40—42? | 48—49° Soe It is reasonable to conclude from these figures that the extinction of the irritability of the nerve is due to the coagulation of the proteids which enter into its composition, and I venture to forecast, that when the proteids of the frog’s nervous system are examined one will be found to coagulate at 40°, and that the two proteids coagulating at 40° and 47° are absent from the nerves of the bird. It is possible, therefore, to make a nearer approach to the analysis of actually living nerve substance than has been practicable hitherto. Temperature of Extinction by Cold, Method.—Alongside the nerve in the nerve-chamber, was placed a junction (A) of konstantan and iron wire, and the nerve was arranged so as to touch this. The junctions konstantan copper (B) and iron-copper (C) were placed in glass tubes and immersed in water at room-temperature, the two copper terminals led to a key-board, with connections to a sensitive Kelvin-type galvanometer of low resistance (16 ohms), and a compensating circuit arranged as shown. The wire rheochord marked 1 ohm was of the ordinary du Bois- Reymond type, and with the voltage and added resistance as marked 1° difference between the junction A, and B C, was represented by about thirty-five scale divisions. The compensating current was furnished from an accumulator of large capacity. It was found after careful tests that no perceptible alteration (within 0:05°) of the * Vincent and Lewis, ‘J. Physiol.,’ vol. 36, p. 445; see also Brodie and Richardson, ‘J. Physiol.,’? vol. 21, 1897, p. 333, and ‘ Phil. Trans.,’ B, vol. 191, 1899, p. 127; and also Vernon, ‘J. Physiol.,’ vol. 24, p. 239. + Halliburton, loc. cit.; also Halliburton and Mott, ‘ Archives of Neurology,” vol. 2. I Demant, ‘ Zeitschr. f. Physiol. Chemie, vol. 3, p. 241, and Kithne and Chittenden, ‘ Zeitschr. f. Biol.,’ N. F., vol. 7, p. 358, 1889, have made some observa- tions on this point, but I have been unable to consult the papers. 278 Dr. N. H. Aleock. On the Negative Variation [Jan. 17, temperature of the fixed junction took place, if it did, a correction could be readily applied to the figures obtained. To guard against current escape from one circuit to another all the wires leading to the nerve chamber were placed within rubber tubing, and the konstantan- iron junction (A) was coated with rubber “ tyre-repairing” solution, which on drying left a thin and even coat of rubber on the surface, insulating it from any nerve currents and from any possible mutual action from or to the nerve. It was found by experiment that no such _ action occurred. | Constane. 5,000 ohms. Fie. 6. To graduate the instrument the junction A was placed in melting ice, the thermo-current compensated till the galvanometer spot stood at zero, and the compensator scale read off. This was repeated with het water when it was desired to use a temperature higher than the constant of B C. This graduation was performed as a matter of precaution at the commencement of each day’s experiments. The whole apparatus proved simple and convenient to use, and of an accuracy much in excess of what proved necessary in these experiments. And as the junction A actually touches the nerve, there can be no doubt that the actual temperature of the latter is observed. The readings are given to 01°C. The nerve chamber was cooled by being placed in a tin box, outside which was a layer of ice and salt, o ed 2% 4 im the Nerves of Warim-blooded Animals. 1903.] “UOTVULLOGIY FO 9FVIS V UL STLUTU oso} JIM ow poystutuy Atpiry Atoa Lorquioog “Iq x ee 80 oe 66 “os "TOOT oer cee “4rqqry oo 66 o° 66 ce * SOlO9 Pet] | 6-9 + unIpeyy GL. 0 100 Cree eee. Z.8+ UIP oF. 0 20 8: + jeoptdod “yur 09-0 | qvO | 8-o8+ OFF 4+ Lect jeoytdod 4x7 Pe. 0 mye) Pol + peottdod juz 68: 0 qn i esl= : Zé. 0 | 08) ‘spQ SB oATeU | | @oI- as FF-O | wo OUIVS Tf} YA quot Fr oI- FO + 81 : Kas ie -1aadxa pug V SUM ap) [= Seer: See ee Ore 200 ae P Ot+ teeqidod 4x4 ZIL-0 wo 0¢. .e-— =) GeeG ITYVIOG “Eb 0. Z ag ‘9[O4 TOO. O : eee a fl “quiod ae ae) ON SIJON Sy T, HOTOUIAX TT DATO NT ae qrowriedxyy | | NS “‘QUv.IASq V “SLNGINTU A XO oes oe ve SOL] 280 Dr. N. H. Alcock. On the Negative Variation [Jan. 17, using the same apparatus that Waller* employed in his researches of the effect of temperature on the electrotonic current of frog’s nerve. The negative variation was observed in the way before mentioned. The value of the galvanometer deflection was ascertained by taking the scale value of 1 millivolt at intervals. It was found, as is well known, that the resistance of the nerve and electrodes gradually increased as the temperature was lowered, and this causes a small error in the strength of excitation, though this was annulled as far as possible by an added resistance of 100,000 ohms in the exciting circuit. Comments. The limits determined are for the temporary abolition of the negative variation, not for its permanent abolition. There is a gradual rise of the extinction point through the four classes of amphi- bians, hibernating mammals, mammals and birds. The limit varies a little in each experiment ina manner that is not accounted for by either the apparent condition of the nerve or by its anatomical character. That some variation was to be expected was clear from the researches of Howell,t who found that the vaso-constrictor fibres in the cat’s sciatic were paralysed by cold (+4° C. q.p.), while the vaso-dilatator fibres were paralysed 1° lower, and even greater differences were observed between the cardiac and respiratory fibres in the vagus. Taking the experiments as they stand, it is evident that those upon which most reliance can be placed, are where the nerve has reached Mittivolts. SS Degrees Centigrade. Fie. 7.—Exp. C*%. Value of the negative variation in nerve of hedgehog at different temperatures. * W aller, * Proc: Physiol. Soe. in “Sd. Physiol; yelse: + Howell, /oc. cit. To 60 Boa Sle O° -/° -2° 1903. ] in the Nerves of Warm-blooded Animals. 281 the lowest point before extinction of the negative variation took place, and both the frog and hedgehog agree in giving a measurable response below O°. In experiment B on the frog the negative variation reached a maximum at +3°8 C. In the experiment C* on the hedgehog an exactly similar maximum was observed at +7°:9 C., and plotting out the “ cooling” and “ warming ” curves the latter gave also a maximum at the same point (fig. 7). No such maximum was certainly observed in the mammal or pigeon. There were traces of a maximum at 25° C. in the former (in experiment B" and an earlier experiment on the kitten not recorded above), but the experiments C*%, C%, C*, C*" showed no sign of this. Several explana- tions are possible, but it seems preferable to await the result of further experiments before insisting too strongly on any of them. One, however, seems well established, that the negative variation follows the tempera- ture with a certain “lag.” This is seen to a small extent in the nerve of the hedgehog (fig. 7), in the rabbit and bird it is larger in amount, and tends to obscure curves taken in this way. I have not yet determined the permanent extinction point, recovery took place in experiment B? in the frog after a tempera- ture of —3°5 C. had been reached, and in Experiment B"? on the rabbit (— 2°-5), expe- riments are in progress in this direction. The range of temperature through which the nerve can function is obtained by com- bining the figures here observed with those of the former series, and it is found that this range is the same for all the nerves examined, 45°°5 for the frog, 45°:2 for the rabbit, and 46°1 for the pigeon, one step higher in the temperature scale in each case (Fig. 8). Conclusion. (1.) It is possible to examine isolated mammalian and avian nerves under the same conditions as frog’s nerves. } 55 Gee o. Re) _ NP (ie y y 5 SP © -¥ © oO ~~ Qo % . 8 gies S : ~S ro) cles ~ S Se ’ acl ~S s 2S 282 Miss 8.C. M. Sowton and Mr. J.8. Macdonald. [Dee. 10, (2.) There is no essential difference between the nerves of frogs, mammals, and birds as regards their negative variation, excitability, and reaction to anesthetics. (3.) There is a marked difference in the extinction point for heat. ‘The negative variation in frog’s nerve is abolished at 40—42° C., in rabbit’s nerve at 48—49°, in pigeon’s nerve at 53°. (4.) This extinction point corresponds closely with the first coagula- tion point of the body proteids, where these are known, and thus coagulation is probably the cause of the permanent loss of irritability of the nerve. (5.) The point at which the nerves are paralysed by cold is - 3°°5 in the frog, — 1°-4 in the hedgehog, +3°°8 in the rabbit, and+6°:9 in the pigeon. _ It gives me great pleasure to acknowledge my indebtedness to Dr. A. D. Waller for his great kindness and assistance in everything connected with this paper. ‘*On the Decline of the Injury Current in Mammalian Nerve, and its Modification by Changes of Temperature. Preliminary Note.” By 8. C. M. Sowron and J.S. Macponatp. (From the Thompson-Yates Laboratory of Physiology, University College, Liverpool.) Communicated by Professor C. 5. SHERRINGTON, F.R.S. Received December 10, 1902,—Read February 12, 1903. The sciatic nerve of a freshly-killed frog led off from transverse section and longitudinal surface to a galvanometer, gives a current of injury which, as Engelmann* and others have pointed out, is greatest immediately after the section has been made. If tested at frequent intervals, it is found that from the outset the HE.M.F. rapidly diminishes. A continuous record of the decline of the current may be obtained, by photographing the movement of the galvanometer spot, using for the purpose the method devised by Dr. Waller, and fully described in his papers.— The nerve with its electrodes being inclosed in a moist chamber, such an observation may be prolonged almost indefinitely. The curve is convex to its abscissa, the decline being rapid at first and gradually diminishing in speed. If the injury current of fresh mammalian nerve be examined in a similar way, the records show in many cases a marked difference of curve, the decline being often very gradual; some records even * Engelmann, ‘ Pfliiger's Archiv,’ vol. 15, pp. 116—148. + S.C. M. Sowten, ‘ International Congress,’ Cambridge, 1898. Oo 1902. | On the Injury Current in Mammalian Nerve. 28 exhibiting an actual rise, so that the highest value of the injury current is only attained when the nerve has lain for some time upon the electrodes. Several observers* have noted the initial rise, and explanations have been offered which accord with the generally held ‘alteration ”” theory of the nerve current. In face, however, of facts recently brought forward in a series of papers by one of us,f it is difficult to accept the prevailing theory as satisfactory. The attempt was therefore made to study this particular phase of the injury current from the point of view set forth in the papers alluded to. That view, which may be called the “concentration cell” theory of the injury current, is based upon the hypothesis of the core-model structure of nerve,t and lays stress on certain pre-existing peculiarities of constitution. These may briefly be described as (a) a separation of the solutions of electrolytes of the nerve into internal and external solutions by a membrane which permits only imperfect diffusion to take place between them; () a difference in the quantitative distribu. tion of electrolytes in the solutions, the internal one being of small volume, but of great concentration and high specific conductivity. Such a difference between the solutions must give rise on' rupture of the membrane, as by section or other injury, to diffusion processes, and consequently to differences of potential. If such electrical differences as are found in the phenomenon of the injury current arise from the source indicated, they should be capable of modification in just the same manner that a process of diffusion can be modified. The value of a diffusion process depends primarily upon the concentration ratio of the two solutions in contact, and may be increased by diluting the weaker solution. In the experiments already reported upon,§ the extreme case of this dilution was exemplified by immersing the nerve for a short time in water; the result was an increase of the injury current due to the enhanced value of the diffu- sion process. In relation to the experiments recorded here, it may be said that the temperature of the solutions, or the fact of any difference of tempera- ture existing between them, is hardly less important than the concen- tration. In frog’s nerve, the solutions are already approximately at the temperature of the laboratory. But with mammalian nerve the case is very different. Removed immediately after death, such a nerve has a temperature presumably not far removed from that of the * Waller, “‘ Croonian Lecture,” ‘ Phil. Trans.,’ London, 1896. + J.S. Macdonald, “‘ The Source of the Demarcation Current considered as a Concentration Cell,’’ ‘ Proc. Roy. Soc.,’ vol. 67, p. 315, &c., 1900, &e.; “The Injury ~Current of Nerve. The Key to its Physical Structure,’ ‘The Thompson Yates Laboratories’ Reports,’ vol. 4, part 2, 1902, pp. 218—350. } Griinhagen, ‘ Konigsberger Med. Jahrb.,’ vol. 4, p. 199; Strong, ‘Journal of Physiology,’ vol. 25, p. 427; Boruttau, ‘ Pfliiger’s Archiv,’ vol. 68, p. 154, &c., &e. § J.S. Macdonald, ‘The Thompson Yates Laboratories’ Report,’ loc. cit. 284 On the Injury Current in Mammalian Nerve. [Dec. 10, mammalian body. Cooling must, therefore, continue after the nerve has reached the electrodes ; and this cooling is not a perfectly simple matter, for the fatty sheath of the nerve is not only, probably, a bad conductor of electricity, but also undoubtedly an indifferent conductor of heat, we should expect, therefore, the cooling process to be differential, the temperature of the external solution falling sooner than that of the internal solution or axis cylinder; and since cooling is in a manner equivalent to dilution, the concentration ratio of the two solutions should be at first increased, and with it the value of the E.M.F. : , The theory then affords an explanation of the increase observed when a nerve just out of the body is examined at a lower tempera- ture. It remained to further exemplify this fact by attempting to reproduce the phenomenon at will. An experiment was so arranged that the proper temperature of the nerve was artificially maintained, and then at a given moment allowed to fall gradually. The resulting curve shows that at or near body temperature the injury current declined rapidly ; in this particular case it fell in half-an-hour 44 per cent. of its original value. In the second portion of the record during the period of cooling, the rapid fall was arrested, there was a period of hesitation, and then a slow rise of the current. On the temperature being again raised, the current resumed its rapid decline. In such an experiment the alteration of condition obtained by fall of temperature is comparable to that produced by dilution of the external solution, and the results are in complete agreement. It is not surprising that to obtain such an initial rise 2f 7s necessary to restrain the fall of temperature within certain limits, outside of which the effect is masked by a more powerful factor working in the direc- tion of diminution. Where, for example, a nerve fresh from the body is transferred to a temperature of 0° C., there is no initial increase ; for here the difference of temperature between nerve and surroundings being great, the rate of cooling will be rapid, and internal solution as well as external will be quickly affected, and to cool the internal solution is equivalent to diminishing its con- centration and hence to a lowering in value of the diffusion process. Extreme cooling should, indeed, annul the production of the injury current, by arresting the processes of diffusion upon which its mani- festation depends. Yet, in a cooled nerve, the source of the pheno- mena, the concentration ratio, is preserved, and a raising of the tem- perature ought again to develop the injury current. This expectation is fulfilled experimentally. At a temperature near 0° C. the current declined rapidly, but it regained its value to a great extent as the tem- perature was increased. With a nerve maintained at body tempera- ture the injury current fell rapidly. A high temperature favours diffusion (7.¢., the equalising of the solutions), and the phenomenon i : | a 1902.] On the Formation of Definite Dust-figures. 289 is reduced by exhausting the value of the source. A low tempera- ture reduces the value of the injury current by checking the process upon which it depends. Both extremes, then, reduce the value of the current, but by quite different means. ‘This being so, there will be a mean temperature around which these two effects are balanced ; at which the value of the source decreases less rapidly than at the higher temperature, and the value of the diffusion process is greater than at the lower temperature, at which, therefore, the injury E.M.F. is best maintained. This consideration led to the systematic study of the injury current at different steady temperatures, and the sought-for point of best maintained E.M.F. was found to le between 14° and ors. In the experiments hitherto considered, we have dealt with current, no allowance being made for changes of resistance brought about by altered temperature. ‘The error, in some cases, was of no great moment, in others, those, namely, where a lowering of temperature gaye an increased injury current ; a correction for altered resistance would but have accentuated the point it was sought to establish. But there are instances in which the error might be serious—and it seemed, therefore, desirable throughout the inquiry to supplement the photographic records by a series of observations, in which measurements of potential by compensation should be taken at frequent intervals. Such measurements have been undertaken by us, and completely confirm the statements made above. “On the Formation of Definite Figures by the Deposition of Dust.” By W. J. Russeni, Ph.D., F.R.S. Received January 29,—Read February 19, 1905. (Abstract. ) The author shows that when a plate of glass or other material is slightly warmed and allowed to cool for 6 or 7 minutes in a dust- laden atmosphere, a clear and definite figure is formed on the plate. The figure is determined by the form of the plate on which it is deposited. Ii a square plate is used then a simple cross is formed, a ray of deposit proceeds from each corner of the plate to the centre. If the plate be triangular, a ray again proceeds from each corner; and with an octangular plate an eight-rayed star is formed. In every case the number and position of the angles of the plate determine the form of the figure. The dust generally used was that produced by burning magnesium ribbon, but any fine dust acts in the same way and produces the same figures. 286 Dr. W. J. Russell. On the Formation of [Jan. 29, With regard to the plate on which the figure is deposited, its com- position is not of importance except as a back ground for the dust. A glass plate for many reasons is best, but the figures form with equal certainty and sharpness on one of copper, or mercury, or ebonite, or India-rubber, or card-board, &e. In order to heat the plate it may be passed several times over the flame of a lamp, warming it as uniformly as possible, and, if it be a glass plate, until the moisture condensed on the under side has disappeared ; or the plate may be heated by laying it on a copper plate heated to about 120°C. for 30 minutes, or it may simply be warmed in an air or water bath. The plate is best supported on three pieces of wire about 1} inches long, and a receiver filled with the dust, inverted over it and allowed to remain there for 6 or 7 minutes. In order to obtain symmetrical figures the plate on which they are deposited must be perfectly horizontal, and as they are very sensitive to heat, there must be no unequal heating either of the plate or the surrounding atmosphere while the deposition is taking place. As long as the plate and the surrounding atmosphere are nearly of the same temperature only very imperfect figures form, but as the tem- perature rises a more and more nearly perfect figure appears. If the plate be above 17°, indications of pictures are produced when the plate is ataslightly lower temperature than the surrounding atmosphere, but when the difference is 6° or more, these indications cease altogether. Very good pictures are produced by having the plate at 12° or more degrees above the dust-laden air, and even when the plate is 100° or 120° above the air, distinct but thin pictures are produced. The effect of a slight heat below the plate, while the deposit is taking place, is shown to thicken the figure, and distort it in a curious manner and is illustrated by photographs. Also the effect of radiant heat on these figures is shown by the action of a Bunsen burner at distances of 12 and 26 inches, and of other sources of heat at considerable distances from the plate. Some singular and complicated effects are 1903. ] Definite Figures by the Deposition of Dust. 287 produced by placing a source of heat above the plate instead of below it. A large number of experiments are also recorded and illustrated showing the effect which different bodies in the immediate neighbour- hood of the plate have on the figures which areformed. Taking only one case, that of a pin. When it is placed in contact and at right angles to the plate a definite deposit is produced, and this varies as the pin is moved further and further away, and as it is placed either on a level with plate, or above or below it. Even when the pin is 6 mm. below the level of the plate, and 2 mm. away from it, a distinct effect is produced. Again, these dust currents may be influenced in a remarkable way by suspending glasses of different sizes, and at different heights above the plate on which the figures are depositing, and photographs of the figures produced are given. The effects pro- duced by obstructions of different sizes laid on different parts of the plate are also shown. It was also found that a current of dust drawn through a tube will form a characteristic figure on a plate, which need not be warmed, as it passes over it. If the magnesia dust be allowed to settle on a surface of water, about the temperature of 17° C. or, on water containing a very small amount of alcohol or glycerine, the deposit which forms on the surface breaks up, by the powder sinking, into a figure of cellular form. Magnesia dust, which was generally used, undergoes some strange changes. When first deposited it is removed by the slightest touch, but if the plate be kept for a week or fortnight it may then be softly brushed over without damage to the figure. Another change which this powder undergoes is shown by collecting it immediately it forms, and examining it under a microscope, when it will be found to consist of irregular shaped and separate particles, but if the collection of the dust be made a few minutes after its formation, it is then seen that the particles are strung together, forming small and irregular fibres. In the various figures that have been produced the magnesia seems to have assumed this form. 288 Prof. K. Pearson. Mathematical Contributions [Jan. 20, “ Mathematical Contributions to the Theory of Evolution.—On Homotyposis in Homologous but Differentiated Organs.” By Karu Pearson, F.R.S., University College, London. Re- ceived January 20,—Read February 19, 1903. (1.) In the paper on ‘ Homotyposis in the Vegetable Kingdom,”* I defined homotypes as “undifferentiated like organs.” In the course of that paper, I endeavoured to indicate that I was not unconscious of the influence of age, local environment, and position upon organism in modifying homotypic correlation. The object of my memoir, how- ever, was to obtain some general appreciation of the average intensity of individuality in living forms, and to see if it approached the average value of fraternal heredity in plant or animal life. For this purpose I selected such material as was readily available, indicating the series where I thought differentiation of a sensible amount was present owing to the age, the situation, or the environment factors. From the standpoint of theory, however, we are not compelled to adopt a mere indication of this kind. As soon as we can correlate between: («) age and the quantitative character of the homologous organs, (0) situation on the organism and this same character, or (¢) local environment and the character, we can allow for the differentiation of homologous parts, or reduce them to pure homotypes. In other words, homotyposis can be deduced from differentiated homologous parts, if we correct for the differentiation due to (a), (0) or (¢). The test for the existence of such differentiation is simply the presence or absence of the corresponding correlation. We have accordingly the following problems to find solutions for :— G;)) Lo find the correction to be made to the apparent homotypic correlation, when the pairs of homologous parts are differentiated from each other by their periods of growth. (ii.) To find the correction to be made to the apparent homotypic correlation, when each pair of homotypes is differentiated by a common period of growth from other pairs of homotypes. (iii.) To find the correction to be made to the apparent homotypic correlation when the pairs of homologous parts are differentiated from each other by situation on the organism. (iv.) To find the correction to be made to the apparent homotypic correlation when each pair of homotypes is differentiated by the environment of its organism from other pairs of homotypes. It will be seen that in problems (ii) and (iv) we are dealing with true homotypes, but that the homotypic factor requires modifying for the influence of age or environment on the organism. In (i) and (111) * ¢Phil. Trans.,’ A, vol. 197, pp. 285—379. 1903. ] to the Theory of Evolution. 289 we are not dealing with homotypes at all, but with homologous parts, and we wish to reduce them to homotypes by correcting for differ- ences between them due to growth or to situation on the organism. I propose at present to deal only with problems (1) to (iii), not because (iv) does not admit of theoretical treatment, but because we have not thus far obtained data to illustrate satisfactorily the correlation between character and the immediate environment of the individual organism. Experimental determinations of homotyposis in plants, when the individuals are subjected to a graduated environmental scale, e.g., in depth of soil or quantity of moisture allowed would be fairly easy to carry out, and most interesting in result. I hope it may be possible to arrange experiments of this kind for the coming sum- mer. We can then illustrate the fourth proposition from actual obser- vation, and the publication of its theoretical solution will be of greater value. (2.) To find the correction to be made to the apparent homotypic correla- tion when the pair of homologous parts are differentiated from each other by their periods of growth. Let z and y denote the characters in the two homologous parts quantitatively determined, and 4, f2 their respective periods of growth. Then we have four variable quantities a, y, t), 4, no one of which fixes absolutely any other, for individuals will have different charac- ters even with the same period of growth. The proposition accord- ingly reduces to this: What is the correlation R between « and y for constant values of the variables, 7.¢., selected values of, ¢, and fy 2 This problem is answered in formule (lviii), (lix) and (lx) of my memoir: “On the Influence of Natural Selection on the Variability and Correlation of Organs.”* Let us write in those formule 7, for the subscript 1, f for 2, x for 3, and y for 4; we have at once soy mee . ° 2 x = Ox iT aye 51 tc eS Gis ie OT Gr A Be (1), tyts — Yr ia» i Des 7 2 %) 7 7 a : SD Td ara (ii), ]— Tee ae . . . . . * Try(1 its ) — Tat, T yt, — Vaty Nyt T Tht, (Tet Tyty TTyty rt) (iii) Ty 9 ° 1- Red 3,5,R I Ox Now if we deal with direct and not cross-homotyposis, .¢., with the correlation of the same character in two homologous parts, we can put these results more simply. We in this case render our correlation tables symmetrical by entering each one of a pair of homologues first as anzand then asavy. We may then write * ‘Phil. Trans.,’ A, vol. 200, p 30. wen. LXXI. Y 290 Prof. K. Pearson. Mathematical Contributions [Jan. 20, Ot es yp tere tat = yt, = i. Voy = Ps aes = 3 C= Ox = Ty and we find >? NRO Ls — —Yr— ‘G — 724 2rrr’ —2 ey z 1—-r? 1—rYr? Qrr —Vr(724-72 : R ( ) ha (iv). rates 1—r—p— 72+ 2rrr’ 1 Pr? 72 — 72 4 Orr’ This is the full solution of the first problem. We see that in order to solve it, it is necessary : (i.) To find the correlation p of the homologous pairs as if they were simple homotypes. (ii.) To find the correlation r between the growth periods of each pair of homotypes. (iii.) To find the correlation 7 between the character and the period of growth. (iv.) To find the correlation 7’ between the character of one homo- type and the period of growth of its fellow. Now these correlations can be found at once by the usual statistical processes, if the data are forthcoming. (3.) I propose to illustrate this on material, which, although not homotypic, is so analogous that it brings out all the important features. We will determine the correlation between the head-length of brothers, such length being measured on school boys of all ages, from 4 to 19.* It will be clear that we have here all the difficulties of the homotypic problem—resemblance due to common origin obscured by differences in the period of growth of each individual. Table I gives the correlation of pairs of brothers without regard to their differences of age. Table II gives the correlation between age and length of head in the same individual. Tables IIIA and IIIB gives the correlation between the age of one brother, and the length of head of the second. Table IV gives the correlation between the ages of pairs of brothers. These tables have been prepared by taking off from the brother- brother data papers of my school measurement records all the avail- able pairs of cases falling into each series. ‘Thus in some cases the ages of both brothers were given, but not the head measurement of one or other ; in other cases the head measurements of both, but the age of one or other would fail, or again the age of one and the head measurement of the other might be all the information available. Thus the total number of cases and the frequency distribution varies slightly from one table to a second. * The measurements form part of the material obtained with the assistance of a grant from the Royal Society Government Grant Committee. 1903. ] to the Theory of Evolution. 291 A few remarks must be made on these tables. Table I gives the following values of the constants :— Mean length of head of elder brother = 186-7508 in mm. y) » VOUNMPeL 5) — leo S200, Standard deviation of elder brother = 7:°5027 _,, > 9 younger ,, = (°s036 , The correlation is, then, found to be 0°601,751,* and the regression, younger on elder brother, 0°5897. These give the intensity of heredity, uncorrected, for the growth factor. Now, the most noteworthy part of this result is, as we shall see later, that taking brothers at different ages tends to exaggerate the apparent intensity of heredity. If we were to take pairs of boys at ages from 4 to 19, each pair having no hereditary relationship, but being, on the average, within a year or so of the same age, we should find a spurious correlation due to the mixture of material, each pair having approximately-like head-lengths because the members of it were, approximately, of like age. On the other hand, if the boys were blood relations of very different ages, their apparent relationship would be weakened, because we should be correlating the same organ at different stages of its growth. We have thus two factors: one tending to exaggerate, and: the other to weaken the apparent strength of hereditary resemblance. It is of great interest to note that the former factor in the present case is the more effective. In Table II we have what I term a growth table, 2.¢., a correlation table between period of growth and the quantitative measure of a character. The constants of this table are as follows :— Me Ame Per Ol DOW). oe..ct eee: = 13:0394 years. Standard deviation of age............ = 28207... We anenead-lemebh (004.5...) ..0...-.- = 185°4516 mm. Standard deviation of head-length = 7°4991 ,, Correlation of age and head-length = 0°453,496 The regression coefficient for head-length on age = 1:205676, and we have the probable head-length Hpfor observed age A given by Hi 16987308 12057 A lee. (ey Thus, on the average, boys’ heads grow in length 12 mm. a year. My results are based on 1637 cases entirely taken off the brother- brother data papers. Dr. Alice Lee at an earlier stage also worked out a growth table. We had not then so many brother-brother data papers filled in. She used in addition all the brother measurements: on the brother-sister papers, and so reached 1856 boys, of which, I * Six figures have been kept in the correlation coefficients, as we require to’ calculate the regression coeflicients from the differences of products and powers. Yew 292 Prof. K. Pearson. Mathematioal Contributions [Jan. 20, think, we may safely assert that 400 at least are not included in my series. She found: Mean age,* 12-7177; mean head-length, 184-8182 ; and slope of regression line, 1:2040, giving the formula Hp = 169-5061+1:2040A, a result in substantial agreement with mine. DIAGRAM 1. | CECA AEE SEA A rh L f2@s 456 67 8 SDN HB 4S 67 ee Age o& Boy. In Diagram 1 the formula (e) is represented with the observed mean values at each year of life. The results for the 4th, 5th, and 19th years of life ought not to be considered, for they are based on only 2,10, and 12 observations respectively. It will clearly hardly be possible to express the growth curve better than by a straight * The mean age is less, because brother-sisters are obtained chiefly from primary, not secondary, schools. 1903. ] to the Theory of Evolution. 293 line, until the range of data is very largely extended. The regres- sion is sensibly linear. Table IIIA and Table IIIs give the following results :— Mean age of elder brother.. = 14 °1249 Mean age of younger brother = 11°7149 S.D. of elder brother’s age.. = 2 °5124 S.D.of younger brother’sage = 2°7221 Mean head-length, younger Mean head-length, elder brother .. . = 183 °8578 brother... 5606 tn boon Sadlelo Sea S.D. of head Jength, Sounaer S.D. of head: Jength, elder brother . : = 7°2806 brother . rere .= 7:°5005 Correlation of age wat lder ae Correlation ai due obs cidueer head-length of younger... = 0°396,598 and head-length of elder. = 0°379,326 We see accordingly that within the limits of the probable error, the correlation between younger brother’s head-length and elder brother’s age is the same as that between elder brother’s head-length and younger brother’s age. This result might, to some extent, have been anticipated, but actual proof of this type of cross-relation is of value. In Table IV we have the correlation between ages of brothers giving the constants :— Mean age of elder brother ............ = 14-1508 Mean age of younger brother ......... == JUL (ekssrr SD0f elder brothers age ......2.)-.. = 2-°5080 S.D. of younger brother’s age ......... = 2 a 220 Correlation of brothers’ ages | oO oo oe) t= = ee) ler) The first four results are in good agreement with those of Tables IfJa and IIIs. The last result shows how nearly there is an approxi- mation to a constant difference in age between brothers in schools. Very closely we have— Probable age of younger brother = 0-96 x (age of elder brother) — 1°83. When the elder brother is 6, his younger brother is probably 2:1 years younger than he is; when the elder brother is 12, the younger brother is probably 2°3 years younger, and when he is 18, 2°6 years younger. The explanation of this is that when the elder brother is very young only his near or second brother will, as a rule, be at the same school, but in the secondary schools, which he reaches at a much later age, it is possible for a much younger brother to be at the same school. Now let us substitute the correlation values, found in equations (1) to (iil), of page 290. We have I 0-601,654, fey ee kot Ne litte) = 0; 379,326, Tit, = 0°884,186 7 “ay ryt, = 0°396,598. 294 Prof. K. Pearson. Mathematical Contributions [Jan, 20, Whence we find S,/oz = 0°890,051, S,/ey = 0°891,209, and R = 0°5446. This is a very reasonable value of fraternal correlation, agreeing quite well with results obtained for horse, man and dog. It is worth noting that Txt, X Tit, = Tyt,X Tet, = 0°4010, and, therefore, either equals 7, or 7,:, fairly closely ; in fact, within the probable error of their difference. Hence, it would appear highly probable that the cross-relation between one brother’s head length and a second brother’s age is solely due to the correlation of the ages between the two brothers. If such a result as Vet, x Vt t, = Tat, belete ee ode 6 oo ee ch ee cee eee (7) should be verified on the reduction of further data, it will enable us to much simplify our formule. Thus we easily find for this case aS: / ps Nv ee =z = Jz 1 — "ates SN oy 1 ie and S a . 2ne pi Tay — Txt, Tit, . OB) 1 ap Txt, Or, we require to find only the uncorrected correlation (p) the growth correlation (r), and the correlation between periods of growth (r). The correction to be made to the apparent correlation is then the subtraction from it of r(r—p) l-?r I hope shortly to ascertain whether relations like 7 above hold also for other head-measurements on growing children. a "P—E8L put E—ZET SUUIN[OO 9Y4 U9IM4Oq PaA|LIT SLA 41 G. EQT sv ONTBA w TONS UO A[JOBXO N T[°F YISus]-pvoy v Ud AA “WO O8 pU G. EOT ULI] ssoT pu G. TOT UY AoQvard syysua]-pvay YIM skoq [Te sureyu0o e—zgT dnoas oyy, x 986|/I | T [116 LI| @8] LF G.g9] ¢. TZ] S.6 arte. ze FOL | S8| 64] 8S] 48) OZ 9T ¢& 61] ¢-6/F | oO]T | stron, sete Rs BA rk can ee ee Ne eee eS os Neue Be! | |e | ed el es Es a | = ae Se ene Te | es PS) Se a Se) oe) a | | ong Se a | er ; cal (tne ame |) a aie Spa = |B ee ee ae eh comes eet ma Pf a S|) eae ieee higal onc: Te oe pe Te) NS Ns en ea ee Ciel 16) 7) ie) uh | es) | — | ol ee | ee eS | eee See bales) Pe | 2-0) Se) Seta he ee | ae | | are eel] | baie S| S| S| S-Se) eRe ess a Dea ig | ee ci Coe ah iS T | 2k | Onl Bul Gee) GOT] eee @ le | Teles | — | — | | eae S OF} | |6 || &| 8) Ol G-00) 24) 9:6) col igs ice) 1} ¢)—|—]/—j—| — | —|—|—|_| poer is S) Cie ee | Le Ga Cae GOs ReecmeGrer Orie 2 (| een valec. Ge Go let ye |e oon & See OGn || | Sa | BGR eC Sm IGA ZEiccsen | Bale 2 \Goet) Gels | ie ee We |) | = ees gon AF 2 ll | Sich) GC \ce et iecnra lec. LelezoS C2 GuGyast enh 20 seg. wht bm | =i) 9 |e eer = CON eee) lh \Gcah vam Cop OLS. IL) ol \Sa8 2.7L) Sell: Gilg sy isha liz @ | = | |=" | 6 ces ea ieee |e) ee SRG Se Case ReSOn) rT | Ober UGnetl) 2 iy | ie iGo k ig lle) tages S SS ee eT ee eC hes ee ot ey aan eee) Ksuid ie CF as | ae ean) Ca UR Be ss0e)|—|—|—|=| —|¢0)-- | t|—|-2| ¢ | #1 orleotegie | ¢| 911. t| 2 lel—ol =| soon S Se | — || — ||) | — if G | @) 8 8] 6)¢ | | —/7 (82-8 \eZ-011 |—| —| e—get | & eG ee al a T |} Tj%e| @| @|t | €| B1e ([gz.h |ey.1|,—|—|—| e—ezt | 8 = j|¢-s¢)—|—|—|—| —| — |} — Pies BS Sos ey er Ie f= Ie 2 ae aie eoer |S CaS | 62h Gas Ol le es Se Ge Oe ISS aor | SONA lle elle) | de NCO) (GC Op ya | a er Tt lexouin T |= | 2 oonaes CM Needle ae ice | ee geass Re NR ae nd reer ee ect eee Oia BP a i a a a a ecg ge era eee eee ag Ig ee Se rey yi SUES PS ESS Bes ee) is Pere [teases coded) sah seitse |) slide ole et eRe i= SO a eee ie a eam Wy cat ney Be ce wrlieaiee ore) ost (ele ae Be | ae eeelbee| eee en eee e|als| ¢ és | yuu wm wayjoug sepia fo yybuaj-poary — ‘OSV OF popottooun ‘suoyJorg Ut yysuary prox, Aqtpoaopy [es9ye][OO—"T ory, 1903.] Ir —— "LT 91GB, 07 OJON 90g 4 = ‘suvok G.g+u = shoq eA yons JO ose uvaut oy} Lo ‘uvaNToo YYW eyy UL peoed eq pynom Aepyyarq meyy Jo Arestaatuus yy(T +) oy 07 YYQw OYA Wo«rZ sLoq IV x* S — == = = — es oo eee — — — : ie a fae} ‘ oy LE91 Gl| 6¢ 8 621 9h) C81 LUZ 622 S9I | SPI ZIT 99 PP CT | OL | 2 | * ‘872904, 8 g -- — i _- iL — — — —- — it — — — |—|— | 6—802 8 g — I — Ti — — == — it = = = = =e | lee ee 8 raul — i z 9 Z — I — — = == = = — |—|—| s—#o0z S G.6 — I | Get T T i T — == ar == = a i role © & G. 6T — I @)2 |Rexe G z — -— z — Ti — == — |—|— | I—002 S G.ZP i [ rah || Oey 8 L g g g — = — a — | —|—| 6—86I 3S Gg. Fg iS Pe) v 6 L | Gh € g z = ee Seiden Fe 1 Seer = Lb it v G.9 | &.¢L | ¢.8T ¢ G.8 IL L = 1 — it — |— | — | ¢—PeI S rAleie ih te bliz 9 G.9 | ¢.OT Cl 2 Sale Sagi ZI Ot Z G — z — |—|— | ¢—zZ6r a3 G.28T | 9-0 P 6 Of | G61 | $61 | ¢-8T Se | So. PT 9 L — = t | —- | — | 1-061 iS 2.291 I 6 6 | ¢. 6 lie NG ek 02 £3 CT Oy Gag g 1 1 |— |— | 6—88T S 8ST — I 8 |ae Gn aGae ro |G. bo | G.9T 4 €L. | ¢.9 G Te) ee alate 0S iS G. 261 i 1 9 | @.&T 8% GZ 9g 1g 81 ie | Goce | Ges V t | — |— | S—FS8T iS G. LST S g @ 9 91 Lt GZ | 8.92 ZZ GT 9 6 v TL | Ss) see eee eS =~ TOL — — G 8 G.6 oH We LI ii €Z re, Wo 2) G Z |—|— | I—O8st SOL -— — ii G G 6 @. II ral SI OL Oa ecxOn z Z |c¢ | —(|6=9821 dj LL aa aa = 1 yp | oP ZI €L 9 21 OT P Or | ¢-0 |—|— | 4-921 2 6F — — it I ¢ g.¢g G.Z 8 P 6 g.g¢ 9 yr |¢O |L | — | S—PAT fz G. 8 — — — — r4 if G ERs ¢ G.P Pe N enat — | LT | §—-Z21 D G. 68 — = 1 — 1 g G.g 9 Gae P P Gan cieg 2 | | t | tot Py 91 —-) = _ I — | 29 I Z Z V7 Caalecen 2 |—|— | 6—89T bd G.9 — — — — — —~ — iL — it 1 Zz | 3:0 L990 Z — — — _ — -— —~ I I —— = ~ — — |—|— | ¢—F9T 5 g — —~ _— —~ T — — — — —- I I — — |—|— | ¢— SS ee COM | tole ase Peet ee OL ‘CT “PT 451 “el ain ‘Ol 6 zs) L 9 Ghar, | yhog fo aby fo wax ‘sAOG UL PROT] JO YAsueT pur osV Jo UOlVIOLIOQ—'TT 9TqG¥ q Ul pee] fo ¥ age V $ ye[attoO— TT P1981. 296 pum ur poa fo yzbuaT to the Theory of Evolution. 297 1903.] 19 19.19 18 t We) 1D 1 19 9 oS cl po 18 H i ‘S[RIOT, ler) re Pilea eateaeal slop nmgme. | octal Sil claay ‘61 Ol a Laity TMs Te ele ass ies eal acai lal oul oO on ‘LT 7q"Z, 07 oyonN 90g 4 tl fl 19 19 19 19 19 19 18 19 19 19 lHannwoanoune “LT scrim | | |e | at | | | OO 20 —] — Z| G8 v OT ZI | 9-8 Ol ALS G-| 8 meee oe S Ue el eee GP er Ole. 4 onCculeecie te PL \erch al ictes| ~ Olgl ) el) |) ee epCeeret ~ (tees | |e i z g | 9. FT 8 11 PL |&-PL | 6 G | 2 i — |)Sa=cell S TL g 9 8 6 91 |9-9 | 9.6 q y | z == | i Omh = gs pee | g € 9 8 9 OL | 8 Ve|| 26 z T | su ee ie fl aie [eee Teng 25 q g | 328 ig a = I =| Ovi A Ocoee le eel Camalce’ an g | a8 z aie = = ea 5 Oy Sr eee ae =| ale z z Copal ec Bo iy Bea tage Sa = | Saht Pech = | ee eee Ty ose SG |&s (98 | F Le) to =" | TO S G.8 = eh SNR SO) Mi ot I Go) To eee PG Sor a v See eal as eT = = = z S| ea eee 1 a —" | i901 0 me pes eo ee 3 as = ae a wees oe | 0 rs i Ea [ = = = = | Rae ee Ol a4 | | (i>) — | —— — | — — ST SS Pe | Se |S SS | — | | | | | | | aa ‘s[@]0L | 6T | “8T | ‘AT Elis eI Se ect tg | “Ol 6 Bela ec | aiGaealle ey, dw ur sayjzorg sappy fo yzbua)-poaye xy daypoug sabunox fo abp fo “va. 298 "LOYQOIT 1osun0 XK JO ody pur Joply Jo yasuary prox] Jo uolyrjo10g—a][] ale I, 299 ‘ose ours 043 Fo aveodde FYSTUr ‘suIMy Jou “sAoyJorq OM9 qeIy os “KepyqaTg yQUEY 0119 oAofoq Lep oy 07 Lepygarq YQUIU oY vLOAT “9°? “Avot YUM OYY UIYIIM SUTT[VF [Te Suvout “G sv “ou AUY “AoY}O Toto Jo s|UOU V9} UIA UtOg Udeq Zuravy sxoyjorq Jo ared ovo ‘surg Jo sured uas_zy ATUO surezM0 01quq BING, IL PIL OF OJON 908 » C86 Tee (OVe || RISE) “Wht |. Gale Zit) vee = cry | 29 a) 169-\\ cccuelliee cee eee are z — | speqoy, a See ats oe ge eo Se ole ol gee pe ee ee 61 S = i at I Se ie ee meee Se i ee ee pee 81 ‘3 I G G [ — ga S|. |S ie ib a ee ees LI S i Pe z ie | ii ea a al ae ee ee ee ee 91 c S €8 G Gh | CSeleecee leg - Sie | =) SS ee ale GT Ke SS cor G L Sie er | lc apace Sele eh | Se Se val & ng PIT é z 9 Ve e007) ARP OC eee leo a | eI S S LIT I Sees 1) One) 6 a)etre |e = oS See eS a ra = 621 a ea ae | Li ween) =89s 89 | a ee lee | = Oe eee IL Ny S e1L =a Gag we € 8 eee | ee eat Ale cite e | Meee | Ol 8 ~ SIT Says qe I z L Ome ie GG CC sey Ties ar Se ie 6 cS S GL SSS ee | ae 9 OL | Si i Oe | Ot | = (>= 2) =| | 8 ¥ ~ g¢ Se eee eB é 8 8 PL | OL | OL | + 2 | eee 4 oy = | 02 S| Geen eh eee | oT G ¢ 7 v = | 7 = | = 9 S a) el = — — — — a= = z z g == g == z I — G S 7 ao ogee | eee ee A eee | I =e TE See elon = U7 Z eT lenGle |) a8 | GyTes Ole cgi al pier lee ien| oz | iT | (Ol, S6R = sree 0) sO. slaty x ayjoug saply fo abp fo savax ‘SLOUJOLT JO SUING JO Sosy JO UOKYR[V1I0Q— AT IVI, 1903.] 300 Prof. K. Pearson. Mathematical Contributions [Jan. 20, (4.) To find the correction to be made to the apparent homotypic correlation, when each pair of homotypes is differentiated by a common period of growth from other pairs of homotypes. The solution of this problem may be deduced at once from equation (i111) of the preceding problem by simply putting 4;= f. In this case f= 7. r — I and we find > 9 Stee (v). = eae 1 er This equation was given by me in a note in Biometrika, vol. 1, p. 404, and its use illustrated on Dr. Simpson’s data for Paramec- ium caudatum. (5.) To find the correction to be made to the apparent homotypic correlation when the pair of homologous parts are differentiated from each other by situa- tion on the organism. We have only to put in formula (iii) on p. 290, 4 = py; and ty = pr, the positional co-ordinates of the first and second homologous parts, to make that formula available for position instead of age differentiation. If we denote by ¢ and cy the characters of the parts in the positions p; and 2 respectively, our solution takes the form below, where we have confined our attention to the same character, R = : p(l a pps) SF en ee eee 1 'DP2 pyc, lpyco a Alp pl per! pycn De pe ap pee a =! pie! pce pips (Tpjc, + T pcs”) (vi) ie a ae a ae 5 a ci Le TpyP2 pe I pycy ae 21p,pol Dey! Prey His followsisimce 75. — aan a tea eee We have again, therefore, to find four correlation coefficients. But this formula simplifies immensely if we observe the following con- ditions : (a.) Take the same number of homotypes or homologous parts from the same positions in each organism. (b.) Enter each one of these homotypes or homologous parts with each other on the same organism, so as to obtain a symmetrical table, 1.é., ¢, 18 first entered with cy and then cy with q, These conditions are or can be usually satisfied in any homotyposis investigation. (6.) Further, the positions will, as a rule, be arranged in series and may be numbered 1, 2, 3, 4,..m, if m homotypes or homologous parts be taken from each individual organism. The position scale is, of course, perfectly arbitrary, and has nothing to do, for example, with the actual distances between positions on the organism. We can make it a uniform numerical scale, which for convenience we can take to be the same serial order as that of positions on the organism. 1903. ] to the Theory of Evolution. 301 Let p = mean position, o, = standard deviation of positions on the arbitrary position scale. Let there be n organisms, and suppose that Sin denotes a summation of all m homologous parts on an organism, and S a summation for all n organisms. Then, if o, = op), = o>, am (m — 1)Tp,7.%p? = Sn {Sim (Pi — P) (p2 - Prt Sn {Sin (pi —P) x Sin (pe —P) an Sin (PA Oy But Sn(~i-p) = 9, hence, since Sn(pi—p)? = may’, 2 a See nm(m—1)%p»,%p? = —nmo,?, or eee as om Saaa hs ° Senco 22 ese eee (vil). Further nm(m—1)Tp.c, pF = Sn {Sm(pi —P) (C2 — €)} = Sn {Sin (¢2-2) x Sin (pi — P) — Sm (1 - 2) (71 — p)} - $1 {8n (1-2) (1-7) }. But Sn {Sm (4 — @) (p1 —f)} = 2M Oop Teo, Hence I Teyp a emeaa si caa 1 CKD RR DD, m—1 1P} PiP2 pyc a a relation precisely similar to that discovered in the case of growth periods for brother’s head-lengths from the actual numbers on p. 294. Substituting we find Us Type. i Tpyey” a T poy 1° 21 pp, "pyc Tpye2 = (1 ri ‘pip. ) dl at Tp,e,") 21 p,¢, pyc, — Tp,Pe (Tpre.” a pyc») = Typ, Tre" (1 a T pips’) Then substituting in (vi) and using (vil) we determine the simple formula for homotyposis corrected for positional differentation R i Tree = . Pp + pe % : 1| jee fone mM — 1 il oe Prac? see e cece c eee sere ssece (Gx). where 7, stands for the correlation of character and position on the organism. An exactly similar formula might be found for the correction for the age or growth factor, if the m homologous parts dealt with had the same distribution of ages or growths in each organism. (7.) Now the equation just found has the serious disadvantage that it is based on the linearity* of the regression relation between position * The reader should note that this condition does not involve any assumption of normal frequency, or the Gaussian law. The latter applies only to a very special case of linear regression. 302 Prof. K. Pearson. JMlathematical Contributions [Jan. 20, and character. But while organic and homotypic correlations give for a surprising variety of cases sensibly linear regression relations, the relation between position and mean character is far more rarely linear. We obtain, as a rule, remarkably smooth curves. We, therefore, require some modification of equation (ix). Still supposing the regression of character and position linear, we should have, if c’ be the mean standard deviation of an array of organs in the same position, oe = ae (h— ine) But if oy be the standard deviation of the means of the arrays, we have from first principles o? = oy2+0” Hence te Se Pag We can now write equation (ix) in the form nr On * R= Ds ae Y On = 1) (2 eres 2) a a's ave oreereetereaie (x). This is quite free from 7p-, and, what is more, although we have deduced it from (ix) and the relation o? = o?(1—v7,-") peculiar to linear regression, it is now free of any limitation as to the nature of the relation between position and mean character. Thus (x) is a far more important formula than (ix), and should always be used, until we have shown that the relation between position and mean character is sensibly linear. If anything, it involves less arithmetic than (ix). We can show this ad initio as follows :—Let the individuality of the organism in any homologous part be measured by its excess above (respectively defect below) the mean value of the character for the homologous part in that position.* Then, if c’ = element of character due to individuality, and ¢, be the mean character in any position for the n individuals dealt with, | Gy = G —C 0 Sn (Gi) =e, and snl Gee Hence we easily find Dn (6,7) = Sn (Gi?) — ney? Smn (C17) = SmSn (C12) —Sm(Nep?), * It might be considered better, if the standard deviations of the homologous parts vary very considerably with position, to measure the individuality by the ratio of this excess to the corresponding standard deviation. Not only, however, does the use of such a ratio immensely increase the arithmetical labour, which is a possibility, which of course, we could face, but there is also a question as to whether the ratio is really a ¢rwer measure of individuality. A full discussion of this important point must for the present be deferred. 1903. ] to the Theory of Evolution. 303 Or, noting that Eon G1) / (AML) = Sin (Gp), we: have « (Oe ee 2 C= — 1G; Ons where o’ is the standard deviation of the character-individualities free from the position factor. We see that it is precisely the same quantity as we have previously used for the mean standard deviation of the arrays for given positions. Next taking the correlation of characters c’; and c’2 in positions 7; and p. we have Sin (1?) + Sin (61'¢2') = Sim (617) + Sin (162) — 25in (01) ME + 7C?. To get this result we have multiplied every quantity like c’; = c; —¢,, by every other quantity like ¢’2 = cy —¢p, and by itself, and then added such quantities together for every position on the one organism. Thus on the left hand side there are m terms in the first, m(m—1) terms in the second summation ; on the right hand side there are m terms in the first, m(m—1) terms in the second and m terms in the third sum- mation. Now sum for each of the » organisms, and we have nino’? +nm(m—1)Ro? = mn(o? +c?) +nm(m— 1) (pa? +) — 2m?ne? + mnc?. ; a? —o? Whence R = po?/o? + (m= 1)e2’ or, as before oe Cine R = [Pp = a Tu? == (m a 1) (o2 = oy?) S srette sn sevayevolle repeats (a): Now while this proof is independent of the theory of partial cor- relation coefficients, involving only simple algebra, and is further independent of any consideration of linear regression, it yet wants something of the width of the former theory, which allows us at once, for example, to correct for a combination of factors, such for example as for both growth and position influences simultaneously. The difficulty lies entirely in the extent within which it is legitimate to assume the relation between position or age, and the mean value of the character at that position or age to be linear. It is therefore clearly advisable to start by plotting this relationship,* and fitting, if possible, such position or growth graphs with appropriate curves. If, for the series of positions dealt with or the period of growth taken, we find that a straight line7 is a close approximation to the relationship, then we * In the case of some animals and many plants the relationship is in itself of much interest, for it expresses a law of development or growth in serial parts. + The analytical consideration of this point is very simple. If the regression 304 Prof. K. Pearson. Mathematical Contributions [Jan. 20, may use the general theory of partial correlation, otherwise we must fall back on results like (x). For example, in head growth in boys, we cannot much improve on a straight line; in positional influence on the branches in the whorls of Hquisetum arvense we need at least a third order parabola. (8.) Although material for several investigations on the homo- typosis of serial homologous parts has been collected, the progress in some of these cases is slow, as it involves rather laborious micro- scopic measurement. I content myself at present with an illustration from the vegetable kingdom. © I collected in the autumn of last year, 126 plants of Equisetwm arvense in Raydale Side, an offshoot of Wensley Dale; the plant was growing on a lane side high up above Semmerwater. This Hquisetum grows from the top with a single stem, and I counted the number of branches to the whorl from the root upwards. As a rule, there will be one or two whorls close to the soil which have never developed any branches at all; then we have what I shall term the first whorl in which some branches have developed, but the number is irregular and obviously subject to some cause of variation, other than the growth law of the plant. The number of branches to the whorl then increases uniformly and steadily up to the 4th whorl, after which it falls almost equally steadily to the 10th whorl. Beyond this the results becomes somewhat irregular again, for very few plants will be found—at any rate in the locality considered—with more than 12 or 13 whorls, and even in these whorls there is a certain amount of forking or irregularity which it is difficult to deal with. The plants were certainly fully developed be linear, the means of the arrays all lie on the regression line, and the mean standard deviation of the arrays about their means is o./1—r?. If the regression be not linear, the means of the arrays will have a mean square deviation 2? from the regression line. The mean square deviation of the arrays from the regression line, but not from their means, is still o?(1—r°). he mean standard deviation (deviation of mean square from means) is now given by a = a (1 —r) —(oy?—77o"), since o> = o”+oy"- But we easily find ly? => ou’ —ro. Hence = is a good measurement of the deviation of regression from linearity, or of cy from ro. If we take 7 = om/o, we have o? = g2(1—n?), Su? = (7-3) 0. Clearly n? must lie between 77 and 1. Further, » can only vanish when the cor- relation is zero, or become +1 when the correlation is perfect. Between these values it gives the mean reduction in variability of an array as compared with the whole population. Further, the deviation of » from 7 isa good measure of the devia- tion of the system from linearity. Thus 7 is a useful constant which ought always to be given for non-linear systems. It measures the approach of the system not only to linearity but to a single valued relationship, 7.e., to a causal nexus. 1903. ] to the Theory of Evolution. | 305 when gathered at least as far as the 12th or 13th whorl, and I doubt whether even beyond this so late in the season, any further branching would have taken place. A few branches were broken off, and these were of course counted ; there was no difficulty, however, in easily ascertaining whether a branch had in any case been developed or not, and the peculiarity of the lst whorl was certainly not due to missing, but to undeveloped branches. Table V gives the relation between branches to the whorl and position for the whole of the 126 plants. In two columns to the right are given the means and variabilities of the branches for each whorl. Now, whether we judge by mean or standard deviation, we see a perfectly gradual change from. whorl to whorl, which absolutely precludes us from considering the number of branches to the whorl as a pure homotypic character. We see a marked differentiation due to position of the whorl on the plant ; the whorls are homologous but not homotypic parts. Suppose, however, that we disregard our test for differentiation,* and proceed to find a correlation table for the whole material as homotypic. We have Table VI, for which I have to heartily thank Dr. Alice Lee. The value found for the homotypic correlation from this table is p = —0:0064+0-0185, or, there isno sensible homotyposis at all. But we might have gone to the other extreme and taken only the 3rd, 4th, and 5th whorls, which have more nearly the same means and standard deviations as homotypes. The result is Table VII, giving p = 0°7918+0-0129. It will be perfectly clear, therefore, as these two results ought to be the same, if the whorls were true homotypes, that we may get any result at all if we neglect differentiation.t The answer to this is that no trained biometrician would call these whorls “ undifferentiated like organs” with the two right hand columns of Table V before him. * On the test for differentiation, see ‘ Biometrika,’ vol. 1, p. 334. + Bateson, ‘ Roy. Soc. Proc.,’ vol. 69, p. 200. VOL. LXXI. Z | [Jan. 20, Mathematical Contributions Prof. K. Pearson. 306 ei i~ BA Oo 2 cc a es) © > . . e . . e e . 1 NAS ARR ANNAN €98- "MO1YeTANp piepuryg Se ve e. . . - . ~ NY e e . . e e rAaArnranDrmonantnnnae | | | “UvO TL *STBJOT, “él Wein h\ x = SST | ple | 096| I91 | ZIT | *9 | TS | ee a See | ee fe | — — — IT — =: =e te dL T g G — | G te ee | ie ¢ DE f= fira| Bi |) vl | Gi s| (eGee ecole sie ache) =e Ages ecen enece S| cole 2-9 z el Gs Gir g Z ji? |) 48 Ol eg = Gy | SeGan nO wieo = = Gi Gear aaG 6 —|— 7 | fe AL Wee. € i lem! aeons, \eeeles a8 6 ‘OL | ‘6 8 ‘k 9 ¢ “LT Sepa ere papery | | oer “7404 M4 947 07 SayouDLG JO LOQUINAT Ns SLED "14oy M f9 worpsog to the Theory of Evolution. 1903.] 199 SSE1 66ET "8[@40], €1 ni] Yo OZ¢ ii OF SéV OVE OST IIT I T908 G 9 66S “Wg fO 140Y 44 ST uo sayounug fo Laquinyy TASTE NE 199 661 SESL L 8I O9T LES 186 661 6681 € GG YtT VLé 186 el S[BqOT, €I él “avg {0 }40Y YY pug uo sayounug fo saquinyy A 2 308 Prof. K. Pearson. Mathematical Contributions [Jan. 20, Table VII.—Whorls, 3rd, 4th, and 5th only. Number of Branches to 1st Whorl of Pair. < 7 8 9 10 iat! 12 13. | Totals i) an a 28 8 8 1 il = = 46 ss 8 8 | 28 19 3 — — | — 58 6 > 9 8 ig) 146 26 6 = i 206 oS 10 1 3 26 196 aD dl = 262 eoN 11 1 - 6 35 HO Wi ®) 1 162 mae 2 — | — 1 = ; lo | — 20 ss 13 — -- — 1 i — —- 2 Ree ie koe es Pe 2a betel | a ai m Totals | 46 08 206 262 162 20 2 756 Now let us consider how to handle the material, allowing for the differentiation of the whorls. To begin with, our formula requires the use of the same number of homologous parts for each organism, _ and it is, on account of the value of the probable error of the random sample, undesirable to use fewer than 100 individuals. This leads to our cutting off Table V at the 10th whorl. In this way we get rid also of the forking, which certainly begins in many individuals at the 11th or 12th whorl. Table VIII gives us the data of Table V recon- stituted for 110 plants, with ten whorls apiece. The only serious difficulty now remaining is that which I have referred to as arising from heterogenity in the first whorl. A glance at the mean and standard deviation of the branches in the first whorl given in Table V DIAGRAM 2. ~~ Mean Number of Branches. Position Of Whort. tori ie L£quation: Yy=9-45)443 --549,4302 2 --179,988/ x *_.008,4206 x3. Regression curve. —Whorl branches and position. 309 to the Theory of Evolution. 1903. | | | | “STBJOT, “ST O&T 9VG | 606 SOL | 8 Os ae a 1 f Il vI ar g Ol WHE 61 al 14 FT VS OG 61 ie oT 86 LG oL 6 c 6% L& 1g L g € 86 6€ 86 ye; 9 =< GG &P LG L V ss 1G GV 66 9 L a €L 8& GE oi V g IL LT 06 TG 6 9 IL | ‘OL 6 8 Z ) “70Y 4 UL sayounsg fo laquin xT 1& 66 I él Gl ek ¢ g ge —, 8 € WV € G OL—é 81840, Bole) Soy) Ney tay Ss TS: Le 0) 2 eR) NAN I [LOY AA JO UONISOg pur soyourig Jo roqunyy TOOMYoq UOTLIOy—TTT A qe yz, "140ym fo uorpusod 310 Prof. K. Pearson. Mathematical Contributions [Jan. 20, will suffice to demonstrate this heterogeneity. Certain individuals have the normal number of about 8°5 branches to this whorl, but about a fifth of the total number of individuals only develop about half this normal number of branches. To illustrate this I have in Diagram 2 plotted the mean number of branches to the whorl, and fitted these means with a parabola of the third order,* using only whorls 2 to 10. The equation to this parabola is y = 9451443 — 0°549,4302% — 0°179, 988127 — 0-008, 42062", the origin being at the 6th whorl, and y giving the mean number of branches for x whorls from the 6th. We have the following results :— Observed number of Position of whorl. branches. Calculated. 1 TELS Sr to2 2 O-32% 9-308 3 9°718 9:707 4 9-827 9-898 5) 9-746 9-829 6 9-564 9-451 7 8-891 8:714 8 7°491 7: 565 9 5°736 5-956 10 3°964 3°835 A much worse fit was obtained by striking a cubical parabola through all ten points. It will be seen that the excellency of fit fully justifies the use of this curve. But that there is a large deviation from the observed mean of the Ist whorl, when we calculate its value from the curve thus obtained. Somewhat reluctantly, therefore, I felt compelled to omit the consideration of the 1st whorl from my investigations. Had I possessed a sufficient number of specimens I should have separated my material into two classes, those plants with normal 1st whorl and those with abnormal 1st whorl. But with my available material I should have had considerably less than 100 individuals to deal with, and accordingly I settled to take nine homologous parts only, namely, the 2nd to the 10th whorls, in which the differentiation appears to be solely due to position on the plant. Above the 10th whorl, the phenomenon of forking obscures the determination of branches to the whorl, while below the 2nd whorl the full or partial development of branches to the whorl seems to be determined by the local lower vegetation round the stem. Taking Table VIII, I found for the mean of the means 8:2515 branches, and for the standard deviation of the means oy, oy? = * By the method indicated in ‘ Biometrika,’ vol. 2, p. 11. 1903.] to the Theory of Evolution. oll 3,938,354. Further, if o be the standard deviation of the frequency distribution of branches, as found from the bottom row of Table VIII, we have o? = 6'721,083. Hence for use in formula (x) we have, since m = 9, Ree 280, eS HO WG OU ee (xi), a? — Cy" m— 1 o%—o,? Table IX gives the uncorrected homotyposis for the nine whorls treated as simple homotypes. From this I find, PO B58. cd WN Te (xii). Substituting (xi) and (xii) in (x), we find for the homotyposis of the number of branches in the whorls in Equisetwm arvense, when corrected for differentiation due to position, R = 0:4939. This result it must be admitted is extremely satisfactory, and indi- cates how it is quite possible to correct a result like (xii) by allowing for the differentiation of the homologous parts due to serial positicn.* I hope before long to publish other results dealing with homotyposis in serial parts, where the differentiation has every variety of intensity. I think they will suffice to show that differentiation is not a subtle and evasive quality beyond the appreciation of the naturalist who is provided with the training requisite for modern biometric research. (9.) The values of R as given by (ix) and (x) may be illustrated from the actual numbers for Hquisetum arvense. We have seen in the footnote, p. 304, that n = Cy/o. This in our case gives 7 = 0°76549. Bat by direct calculation on Table VIII, using whorls 2 to 10, Dr. Lee finds 7, = —0°64616. Hence with the notation of the footnote referred to o = 076326, Dy = 0°4104c. * The value obtained for the crude homotyposis oi the members of the whorls in Asperula odorata in my first memoir was p = 0:1733 (‘ Phil. Trans.,’ A, vol. 197, p. 326). I have little doubt that when we are able next summer to calculate the correction for differentiation in position of whorl, we shall find R for woodruff in good accordance with other homotypic results. My remarks about it were: “In counting the members on the whorls I soon found evidences of differentiation in position, the whorls towards the top of the spray having, as a rule, fewer members than those lower down” (loc. cit., p. 325). Unfortunately I have not kept my records of position. [Jan. 20, Mathematical Contributions Prof. K. Pearson. 312 | | ; M 10 Osos) te | 20 | vere forces con gol | +8 | 0g Woes le 8| 6G Se te Se eee ee Ss pee re | tee a ee eee | ee ee | ee ———|-——-|- __-| __|-___--__}—_____| § oz6L | 8 9&1 | OFOT | 896T ZL9L rads) | ZL9 | 00F GU SEG a) (CEG Oc. |e cen = ce cero) Ss 8 ee | z ot I Z = = I ~- = — ET Ss 98T See ve a 6s 1é 6 6G g P g = | ¢ alt w OFOT I ee | gee 161 eal 96 1g OF 8g ZB el | &T OL Teil 8 S961 | Z ig Yen OLE FOS 691 6ST EIT GG 6 9¢ PS 68 OL 5. ZZ9T T 6 IZt FOe Z9¢ SPL TSI 96 6 ge ae | @2 | pe 6 Ss res | (CT Ze | 96 691 ral PPL 02 6g BZ ce ee 6 ra 8 = B19 z e Le 6e1 11 02 OLT ge 1g Ov 1Z 1g 61 p SS Or — | ¥ OP ell 96 6E ge z eI CT Zi er cL g = aie | — | 8 ge eg 62 rea 1¢ el 9 ! c 6 F G = Sra | — | 2 6 9¢ ee og GT L r 9 el OL F aS Bes — | & PL g¢ 6S ee 1Z ral G 9 9 L IL e S N= | = SI PS cL 62 1G ST 6 Cl ez FL 8 tw = 261 | — G OT 6E Lg rae Gl eT t OL IL 8 ai | it fe SS = -———|——- S SICIOL | “Sh | SI “Tit ‘OL 6 g iy ‘y re "'p g % Ti ‘WD f0 140 YY IST Ur sayounag fo saquenay ‘asuanww unyjastmby 10y sisodAqowoFT poyoo1t09u—" XT [qe 1903.] to the Theory of Evolution. allio Thus 7 diverges much from 7, and = from zero. Indeed, a glance at the diagram shows how far we are from true linear regression. If WE USE /'pe aS above instead of 7, 2.¢., (ix) instead of (x), we have R = 0:3511, a value very much below the actual value. This illustration will suffice to emphasise the importance of testing the actual curve of regression before we assume it to be linear and use equation (ix). (10.) The subject of differentiation due either to position or age 1s, of course, a difficult one, but it does not seem at all beyond biometric treatment. The greatest difficulty which it seems to me will have to be encountered is not that of discovering and allowing for differentia- tion due to serial position, but in ensuring that when this has been allowed for, there is not remaining an organic correlation due to the necessity of adjacent parts “fitting.” On this account it is most desirable that as large a number of homotypes as possible shall be taken, so that the part of the correlation due to the homologous parts having to fit, or, indeed, to serve a common end, should be reduced to as small a quantity as possible. For example, if we suppose adjacent whorls to have their number of branches influenced by an organic: relationship, this result will only bias nine out of the forty-five pairs we should form in dealing with ten whorls. The question of separating organic from homotypic correlation is one that I hope to return to at a later date. Meanwhile the present paper will suffice to indicate how partial correlation coefficients enable the biometrician to iree himself from the differentiation between individuals due to different periods of growth, or to different positions on the organism. In conclusion I should like to thank Dr. Alice Lee and Mr. F. E. Lutz for aid readily granted at one or other stage of this investiga- tion. ol4 Prof. J. T. Wilson and Mr. J. P. Hill. [Jan. 21 “Primitive Knot and Early Gastrulation Cavity co-existing with Independent Primitive Streak in Ornithorhynchus.” By Professor J. T. Witson, M.D., and J. P. Hitt, D.Sc., Univer- sity of Sydney. Communicated by Professor G. B. Howes F.R.S. Received January 21,—Read February 12, 1903. Amongst the material at our disposal for the investigation of the earlier stages in the development of the Monotremes is an egg of Ornithorhynchus, measuring 10 x 95 mm. The stage of development represented by this egg is of such interest and importance that we have deemed it deserving of a brief preliminary communication. The stage of embryonic organisation would appear to fit in as immediately succeeding the oldest of the early embryos described by Semon,* viz., his “EK,” (fig. 15, Taf. 8). His “Ey,” appears, indeed, to have been older than “Eg,” but with the exception of one figure (fig. 39, Taf. 9) representing the structure of the extra-embryonic blastoderm, Semon gives no indication of the conditions met with in that ovum. This is the more to be regretted, as we see reason to believe that ““E;” must have corresponded pretty closely with the stage now to be described. Semon notes that the embryonic area of this egg was injured through an unfortunate accident. The general structural arrangements of the Monotreme ovum in its early stages of development have been described and figured by Semon. In all the stages dealt with by him, illustrating the develop- ment from segmentation up to the first indications of gastrulation, the yolk had retained its original arrangement as a tolerably coherent solid or semi-solid spheroidal mass with alternating strata of white and yellow yolk-spheres. As he states, however, the egg in the course of further development increases considerably in size during its sojourn in the uterus through absorption of fluid. By this process of fluid absorption, the yolk-mass is disintegrated and its spherules dissemi- nated throughout the interior of the growing blastodermic vesicle, though many of them remain adherent to the deep surface of the blastodermic membrane. | This conversion of the solid or semi-solid yolk-mass into the fluid contents of a large blastodermic vesicle renders the investigation of the structure of the blastoderm, from the period of the commencement of gastrulation up to the formation of a distinct embryo, an exceed- ingly difficult one. The loss of Semon’s “EH,” may very likely be attributable to such difficulties in the way of manipulation as we have encountered in dealing with a delicate blastodermic membrane sur- rounded by a thick, tough, and opaque shell as well as a vitelline * Semon, R., ‘ Zool. Forschungsreisen,’ &c., Bd. 2, Lief. 1 (1894). 1905. ] Primitive Knot, ete., in Ornithorhynehus. old membrane, and distended by a considerable bulk of fluid material. It is quite impossible to remove the shell without serious damage to the delicate blastoderm. Fixation with the shell intact is imperative, and, as a matter of fact, the result in the way of preservation proves quite satisfactory, as evidenced by the condition of the cellular blastoderm, in which mitotic figures are well preserved. But even after fixation, the opening up of the ovum is attended with no little risk. The inevitable evacuation of the contained fluid allows of crumpling of the blastodermic membrane with possibility of injury to the embryonic area. It is impossible, owing to its size and osmotic difficulties, to treat the ovum throughout unopened. LHven were that course practic- able, the impossibility of orientation would be an insuperable difficulty. After fixation and subsequent dehydration in graded alcohols, the 10 mm. egg was cleared in origanum oil and opened. The blastoderm still remaining zn setw in relation to the shell was examined from the interior aspect, and was found to possess at one spot a small more Opaque area, somewhat oblong, but rather irregular. The portion of the blastoderm containing this small opaque patch was photographed by transmitted light at a magnification of 6°5 diameters for the purpose of orientation. Our surmise that the area in question was of the nature of an embryonic or primitive knot was afterwards confirmed by the exami- nation of serial sections. No differentiation in way of an embryonic area in the wider sense is recognisable in the photograph, nor was any such discovered in the course of examination of the wall of the blastodermic vesicle in toto under low magnification. We naturally concluded that the very evident knot represented the earliest and only differentiated area, and for a time devoted our attention solely to this area and the blasto- derm in its vicinity. The portion of the blastoderm containing the knot was separated from the remainder of the wall of the vesicle, and was then imbedded and cut in serial sections, Examination of these did in fact show that in the neighbourhood of the knot, and for some distance from it, the wall of the vesicle was destitute of any indication of further differentiation. But towards the periphery of the portion sectioned and comparatively remote (nearly 2 mm.) from the knot, itself, we found the commencement of a region of thickened ectoderm with underlying mesoderm. Our attention thus being directed to other manifestations of developmental activity in the blastoderm in addition to the primitive knot, we found, in the portion of the vesi- cular wall originally put aside, a quite extensive area showing im- portant changes. These amount to no less than the establishment, quite away from the region of the knot, of a distinct linear primitive streak formation, surrounded by an area over which the ectodermal 316 IProf. J. T. Wilson and Mr. J. P. Hill. [Jan. 21, layer is thickened and. cubical, and within which a mesodermal sheet has already undergone a wide extension. It is peculiarly unfortunate that, owing partly to the difficulty of dealing with the delicate and originally collapsed and torn vesicular wall after the cutting open of the egg, partly to the failure to detect any visible differentiation apart from the knot, and our consequent conviction that the latter was the sole trace of embryonic organisation yet present, the orientation of the distinct portions into which the vesicle was separated was not adequately determined and preserved. It thus became impossible for us to guarantee that the planes of sectioning of the remaining portions should be accurately co-ordinated with each other, or with that of the important piece first sectioned. This failure, not wholly blameworthy, when the conditions of the task are appreciated, has introduced an element of conjecture into our sub- sequent attempt to determine the precise relation to the knot of the primitive streak area, of whose existence we later on became aware. Nevertheless, we think that we shall be able to establish these rela- tions with at least a high degree of probability. In this preliminary paper we propose to restrict ourselves to an account of the highly interesting area which first attracted our atten- tion, and which we have already referred to as a “ primitive knot.” Tt will, we think, be admitted that no possible doubt can be entertained of the justice of employing for its designation a term which would stamp it as the homologue of the well-known structure in many Sauropsida to which the same name has been applied. It is a true primitive or gastrula knot, in the Sauropsidan sense, possessing a transversely elongated gastrula-mouth or blastopore and an invagina- tion cavity, which both in appearance and minute structure resembles in the closest manner the structure described under these names in various Reptilian forms. But if this be admitted, we come face to face with this extraordi- nary fact that, in addition to this reptilian-like gastrula, there is to be found in Ornithorhynchus, quite distinct and even remote from this gastrula knot, a region in which there is being differentiated independently a primitive streak of quite ordinary and _ typical mammalian character. The oblong portion of the wall of the vesicle containing the primitive knot near its centre, and measuring about 8 x 10 mm., was left adhering to the portion of the shell covering it. The direct observation and photography of the piece were thus confined to its deep aspect so as to avoid undue disturbance and injury. The whole piece was then double imbedded in photoxylin and paraffin and cut into serial sections 10 micra thick. The sections were stained in hematéin and eosin. The plane of section chosen was approximately at right angles to 1908. | Primitive Knot, ete., in Ornithorhynehus. 317 the larger diameter of the embryonic knot, and parallel to its pre- snmable axis, on the supposition that a clear spot near one of its margins represented the position of the gastrula opening or blasto- pore, thus determining the true posterior margin of the knot. Examination of the series did ultimately show that the sections were approximately longitudinal sections through the gastrula cavity. The general character of the egg at the stage under consideration must be briefly referred to. At the period now dealt with, the formerly yolk-laden ovum has become transformed into a large blastodermic vesicle with fluid contents, amongst which are large numbers of dispersed yolk-spheres of the original yolk-mass. . One can, without hesitation, homologise the interior of the vesicle with the subgerminal cavity of a Sauropsidan egg, extended so as to include by liquefaction the whole of the yolk itself. Ornithorhynchus indeed may be said to afford an actual demonstration of the trans- formation of a Sauropsidan subgerminal cavity, such as is figured in its first beginnings in Semon’s figures 36 and 38 of his “Os,” into the cavity of a mammalian blastodermic vesicle, thus supporting Keibel’s view of the correspondence of these cavities. The establishment of the vesicular stage has been effected by the extension of the cellular blastoderm completely around the yolk, a condition representing a considerable advance on that existing in Semon’s figures “‘H;” and “‘Q3.” He does not record the condition in this respect of his ‘“‘ Ky.” He was also unable fully to determine the con- dition of ‘“‘ E;” in this respect. We believe that the latter must have been very similar to our present stage, judging from the figure he gives of a portion of the extra-embryonic blastoderm. Not only is the cellular wall of the blastodermic vesicle complete in our specimen, but it is already bilaminar throughout, and trilaminar over a not inconsiderable area. A complete layer of yolk-entoderm (‘ second- ary” entoderm) has been differentiated lining the cavity, except under the small area of the primitive knot, with whose tissue it is continuous. The yolk-entoderm cells are large and swollen, being distended by yolk-spheres of various sizes and somewhat different staining capacities. Owing to their yolk-laden character their protoplasmic contents are relatively greatly reduced. The nuclei are, as a rule, large and vesi- cular. In the least successful sections the entoderm cells may be broken up and more difficult of recognition. We are firmly convinced that it is owing to the imperfection of the section shown in Semon’s figure 39 from his “ E;” that such a careful observer has been unable to recognise the yolk-entoderm as such, and figures it as a zone of vacuolated coagulum next the surface of the yolk. The imperfect layer of angular-looking cells which he figures and regards as the entoderm in “ E;” is not really such, but forms an intermediate layer of mesodermal cells. 518 Prof. J. T. Wilson and Mr. J. P. Hill [Jan. 21, This layer of mesoderm is well represented in the stage under description. But it by no means extends throughout the whole extent of the blastodermic vesicle, nor is it met with at all at or in the immediate vicinity of the primitive knot. It is found most fully developed in and around the primitive streak area, and here it is continuous with the ectodermal thickening of the primitive streak (i.e., paraxially), where it attains greater thickness and consists of several layers of cells. Outwards from the line of the primitive streak* it thins out gradually into a single layer, and further out becomes patchy and incomplete. It is throughout distinct and in- dependent of the underlying yolk-entoderm, which can be followed through the whole wall of the vesicle. We have several pre- parations of small portions of the wall of the vesicle stained and mounted in toto, which give clear demonstration of the huge yolk-laden entoderm cells forming a continuous lining of the vesicle wall, precisely as figured by Hill and Martiny in a considerably later stage. The ectoderm of the vesicle wall throughout the greater part of its extent forms a membrane consisting of very thin flattened cells closely applied to the vitelline membrane. This character it retains in the vicinity and over a portion of the primitive knot. But over the region already referred to as the primitive streak area, remote from (posterior to) the primitive knot and extending outwards so as to be practically co-extensive with the mesoderm of this area, the ectoderm shows a marked change in character. Here its cells are no longer flattened and squamous, but thickened and cubical, and here also their developmental activity is often betrayed by the existence of frequent mitotic figures. In the posterior part of the primitive streak thickening, the cubical ectoderm gradually becomes more attenuated, the mesoderm still continuing beneath it for some distance, even after the ‘‘ extra-embryonic” character of the ectoderm is assumed. We now come to the “ primitive knot ” itself, whose occurrence and characters it is the special object of this brief paper to record. The oblong area of the blastoderm constituting the knot measured 0-42 mm. in one diameter (antero-posterior axis of gastrula) and 0:49 mm. in the other diameter (transverse). It is therefore compar- able in general form with the area from an Echidna egg (Keg) figured by Semon in his fig. 15 (loc. cet.). Unfortunately Semon gives no clue to the magnification of this figure, which is on a different scale to the other illustrations. One cannot therefore compare the area of EK, * The plane of section through the primitive streak area of the vesicle wall, though its orientation was determined in rather haphazard fashion in default of any visible guide, is obviously approximately transverse to the axis of the primitive streak, showing the characteristic bilateral symmetry of organisation of this important region. + Hill and Martin, ‘‘On a Platypus Embryo from the Intra-uterine Egg,’ ‘Proc. Linn. Soc., N.S.W..,’ vol. 10, figs. 31—33. 1903. ] Primitive Knot, ete., iv Ornithorhynchus. o19 in size with that now described. It is to be regretted that Semon’s specimen was lost through accident, so that no further comparison can be instituted. There is, however, a general resemblance between fig. 15 of Semon’s Ky and his figs. 14 and 18 showing areas from other eggs of Echidna (E;) and Ornithorhynchus (O3). The dimen- sions of these can be calculated and are both found somewhat to exceed those of our primitive knot, in spite of their relatively younger ages. It is therefore difficult to say to what extent the areas referred to (of Es and E; and O3) correspond to our primitive knot. We are, however, of opinion, that Ej, though probably younger, must have pretty closely corresponded, and there can be little doubt but that the centre of the other areas, at least, is later on developed into an actual primitive knot. The sectional figure of O; shown in fig. 38, inter- preted by Semon as showing the commencement of gastrulation, may possibly bear this interpretation, but we cannot feel satisfied that the granular coagulum layer between the cellular layer and the surface of the yolk shown in this figure has been adequately interpreted, especially in view of our complete conviction that in fig. 39 the layer represented in somewhat similar fashion is beyond all doubt really yolk-entoderm. We do not wish to be taken as suggesting that here in O; the coagulum represents differentiated yolk-entoderm. We are simply not convinced beyond all doubt that all the formed cell elements of this area in Os are disposed in the form of a superficial continuous membrane as shown in fig. 38, more especially in view of the totally different condition of the nearly related stage H;, as illustrated in the sectional fig. 33, and of the condition described below in our own somewhat later stage in Ornithorhynchus. The general form of the embryonic knot has already been noted. A schematic surface projection of its outline is shown in fig. 1. This Fia. 1. figure represents a plane reconstruction of the area of the knot, com- piled from the serial antero-posterior sections through the region. One of the more typical of these sections is also figured semidiagra- re ——— 320 Prof. J. T. Wilson and Mr. J. P. Hill. [Jan. 21], matically in fig. 2. More adequate illustration by photomicrography and otherwise will be forthcoming in a future communication. nre. 2: ech.V.W. Ue . ‘ol pee ae xe) ay Wet) PSS % 0 SOLOS S Po 8. roo Of Or © ° 05 8% 0092 NY : yer. FO ce Tae, ed, SENS oe ( 1S) © “ ones oy h Oy) ® C) \ aye "8 OBO So ald, ¥y. SP. . LC * % AC OETA) Fe ° rs) ° ese Oo ° \ * \oe a SoS COG OF iES Casio XO) Shs ence aS \Sy u SPOS Noy CEO Ou 2 O.58O = ote ES) CH — S je * 7 ote = | See 8 SENS AEM ool er oe (ho 99090 ee ay dy 2 b° % * Lie fe wo 9} 8 b e Pat 3 cal ew Yisphe pee eh et sie Ou ees ! b hese =e g eee ? ay Paes \ j Few. 1 tees De PTET ea a He siyet \ THB Ss Ee omamay \ eCL.V.W. j ve < Soest I \ z | SPOR FCO \ / pee ’ yl Y.-S. ! y C.W.S hiuc. In mesial section, the knot is seen to form a thick and prominent lenticular mass projecting into the cavity of the blastodermic vesicle. It is largely composed of a loose reticular tissue in which nuclei are only sparsely distributed and cell outlines are for the most part invisible. This tissue is thickly dotted with minute yolk-spherules and small vacuoles, and is not limited towards the cavity of the vesicle by any very sharp or clear-cut boundary. This reticulum of the knot is continuous peripherally with the yolk-entoderm of the bilaminar blastoderm around the knot. Penetrating the interior of the knot is the archenteric or gastrula- cavity, opening on the surface at the blastoporic aperture near the hinder part of the knot and appearing in sagittal section as a curved canal passing from the blastopore at first deeply, and then forwards, to end blindly in the more anterior part of the knot. This cavity is lined throughout by a very definite cellular wall. Both in front of and behind the knot, the blastoderm is simply bila- minar, with thin ectoderm closely applied to the deep surface of the vitelline membrane. ‘The entodermal cells are large and contain yolk- spherules of varying size and staining reaction and loose yolk-spheres are also found adherent to its deep surface. The thin ectoderm is continued over the knot from the region in 1903.] Primitive Knot, ete., iv Ornithorhynchus. a21 front without change of character as far as the transversely elongated blastopore seen in fig. 1 and in section in fig. 2. Thence it is con- tinued round the anterior lip of the blastopore so as to line the roof of the blastoporic (archenteric) canal or gastrula-cavity. Posteriorly to the knot the thin ectoderm is continued forwards, but on reaching the posterior edge of the knot, its cells lose their markedly flattened character, and become continuous with a mass of cuboidal or even rounded cells, which forms the prominence of the posterior lip of the blastopore. The superficial cells of this mass do not, however, wholly surrender their epithelial arrangement, and are continued forwards in the hinder wall and floor of the gastrula-cavity to complete the cellular lining of this cavity. This cellular lining consists, for the greater part of its extent, of large and cuboidal cells, but in the hinder part of the roof, and extending from this upwards around the anterior lip of the blastopore, the cells are more flattened, ultimately becoming continuous with the thin ectoderm overlying the anterior part of the knot. Figs. | and 2 together enable one to realise the form of the gastrula- cavity as a wide, but dorso-ventrally flattened cavity opening posteriorly by a transversely-elongated blastoporic aperture. _ The posterior lip of the blastopore is formed by a mass of cells which, if not wholly indifferent, at least show a less clear differentia- tion into superficial and deep than is found elsewhere. As was indicated at the outset, the resemblance between the primitive knot in Ornithorhynchus and that found in a number of reptilian forms is very striking. We would particularly draw attention to the figures by Mitsukuri* of the primitive knot in Chelonia, and more especially to his figs. 9 and 13 on Plate 8, which show a remarkable similarity to the condition here described, in all essential features. In Ornithorhynchus we lack entirely the columnar arrange- ment of the ectoderm over the knot, and the anterior lip of the blasto- pore there shown, otherwise the characters both of the knot and of the gastrula-cavity and its wall appear to be almost identical with those fivured, especially in Mitsukur’s fig. 9, Plate 8. In a future more extended communication we hope to illustrate more adequately the points above set forth, and in addition to add something in the way of elucidation of the fate of the primitive knot in somewhat later stages. We propose also to describe and illustrate the con- dition of the primitive streak area, which at the period now dealt with already co-exists with, though independently of, the primitive knot, but which later by extension comes into more intimate relationship with the knot. * K. Mitsukuri, ‘On the Process of Gastrulation in the Chelonia,” ‘Journ, Coll. Sci., Japan, vol. 6. Of. also Gertrude C. Davenport, ‘ Radcliffe Coll. Monograph,’ No. 8. Boston, 1896. Dendy, ‘Qu. Jl. Micro., Sci.,’ vol. 42, p. 18, 1899. VOL. LX XI. DIN 322 Prof. G. Elliot Smith. [Jan. 15, : The mere fact of its co-existence at this stage with the knot, necessarily occasions some reconsideration of the morphological relationship of the mammalian primitive streak to the process of gastrulation. “The Brain of the Archeoceti.’* By G. ELuior Suitn, M.A., M.D., Fellow of St. John’s College, Cambridge, Professor of Anatomy, Kgyptian Government School of Medicine, Cairo. Communi- cated by Professor G. B. Howes, LL.D., D.se, F.RS. Re- ceived January 15,—Read February 12, 1905. So far as I have been able to ascertain, nothing whatever is known of the form of the brain or, more strictly, of the cranial cavity in the Archeoceti. Hence no apology is needed for presenting even this imperfect account of two cranial casts representative of this sub-order, which have come into my hands. Among the Eocene remains found in the Faytm region of the Egyptian desert by Mr. H. J. L. Beadnell and Dr. Charles W. Andrews, in 1901, there was a broken skull of Zeuglodon,t+ from which it was possible to obtain a mould, representing the form of the greater part of the dorsal and lateral aspects of the brain. A plaster cast was made in the British Museum at the instance of Dr. Andrews, who kindly placed it at my disposal for description. In the following winter (1902), Mr. Beadnell found in the same locality a natural cranial cast of the same size and general form as the artificial cast of Zeuglodon. It is obvious at a glance, if the two specimens be placed side by side, that the natural mould belongs to some member of the Archeoceti, but whether to the same species or even genus as the other specimen must at present remain an open question. Mr. Beadnell kindly placed this specimen at my disposal. The size and relative proportions of the different parts are almost identical in the two casts. Nevertheless, there are a considerable number of differences, some features being displayed in one and not in the other, and vice versd. Many of these differences are obviously due to the imperfections of the casts, and especially to the failure of the plaster mould to represent the true form of the brain. But there are * These notes were originally intended for the Report on the Survey of the Fayum, to be issued by the Egyptian Survey Department, and are now published separately with the permission of the Under Secretary of State for Public Works and Captain H. G. Lyons, Director-General of the Survey Department. + C. W. Andrews, “ Extinct Vertebrates from Egypt,” Part II. (Extracted from the ‘ Geological Magazine,’ N.S., Decade IV, vol. 8, 1901, p. 437,—Zeuglodon Osiris, Dames’.) 1903.] — The Brain of the Archeoceti. 323 several peculiarities of the natural cast—such, for example, as the form of the caudal part of the cerebellum and the shape of the cerebral hemispheres—which are difficult to reconcile with the artificial mould, even if we admit that the inner face of the cranium (from which the latter was made) is damaged or imperfectly cleaned. ‘The differences, nevertheless, are sufficiently pronounced to indicate a generic dis- tinction between the two specimens; and in this connection it is interesting to recall a statement made by Dr. Andrews in his first reference to Zeuglodon, as “including apparently Dames’ Z. Osiris, and perhaps a second species.”* It would, however, be very unwise, because it would serve no useful purpose, and possibly lead to error, to found a new genus or even a new species on the evidence of this natural cranial cast, when our source of information concerning the known genus (Zeuglodon) is as unsatisfactory as that obtainable from the artificial one about to be described. More especially so, when it is remembered that in the case of the only indisputable facts (.¢., size and general form) the two casts are agreed. I shall therefore merely describe and attempt to explain the meaning of the form of the two specimens, and leave the question of the specific identity open for future research. The general appearance of the brain is extraordinarily peculiar (figs. 1 and 2). The shape of the anterior part of the natural cast (fig. 1, a and 6) closely resembles the cerebrum of a Lizard greatly magnified. An anterior prismatic stalk (a), representing the pedunculi olfactorii, suddenly expands into a plump, broad, smooth mass (0), showing the form of the chief part of the cerebrum. The maximum breadth of the two hemispheres (fig. 1, 5) is 95 mm.; the greatest length of each (measured in front from the point where the ventral surface of the olfactory peduncles appear to expand into the chief mass of the hemisphere) is 47 mm.; and the maximum depth is 54 mm. Kach cerebral hemisphere (exclusive of the olfactory peduncle) is slightly broader than it is long. The two olfactory peduncles are represented in the natural cast by a single prismatic process. This extends forward for a distance of 37 mm. (measured along the dorsal edge) in front of the point where the expansion to form the hemispheres commences; and as the peduncles are broken across there, it is not possible to estimate their total length or the shape and size of the olfactory bulbs. The coronal section formed by its anterior (broken) surface gives an isosceles triangle with a base measuring 8°5 mm. and sides of 10 mm. each. It expands as it passes backward, so that at its junc- tion with the rest of the hemisphere its sides are each 19 mm. and its base 16 mm. in length. In the artificial cast (fig. 2) all that represents this extensive * ‘Geological Magazine,’ September, 1901, p, 401. ( 324 Prof. G. Elliot Smith. [Jan. 15, olfactory stalk is an irregular rostrum with two small boss-like projec- tions, one above the other (a anda’). The cerebral hemispheres in the natural cast have a broad base, from which the sides extend up- ward toward the narrow dorsal surface with a gradual slope. In the artificial cast, however, the lateral parts of the hemispheres seem to be expanded into full rounded swellings. Then, again, the antero-posterior diameter of the hemisphere is much shorter (being about 13 mm. less) than it is in the natural cast, although the breadth of the two specimens is approximately the same. Fra. 1.—Dorsal aspect of the natural cast described in the text. 2 natural size. a, olfactory peduncles; 6, cerebral hemisphere; c, d, cerebellum; e, e’, fragments of skull. Tt may be that the anterior parts of the skull, from which the artificial cast was made, are so damaged that little reliance can be placed upon the mould as an indication of the exact form of the brain. In fact, if this artificial cast even approximates to the form of the brain, it is quite certain that /it did not belong to the same genus as the animal from which the natural cast was derived. In other words, as we know that the artificial cast belonged to Zeuglodon, the probability is that the natural cast furnishes the first evidence of some hitherto undescribed genus of Archeoceti. Behind the part 6, which I have just described as the cerebrum, 1903. ] The Brain of the Archeocett. 325 there is (in the natural cast) a large, irregular mass of a very peculiar shape, not exactly comparable to the condition occurring in any other brain known to me. Immediately behind the hemispheres (/) there is a great transverse bar (c) measuring 125 mm. in the transverse direction—+.c., extending on each side 15 mm. beyond the lateral margin of the cerebrum (0). Hach lateral extremity of this mass (c) is expanded to form a large buttress. In the natural cast these buttress-like masses are practically Fie. 2.—Dorsal aspect of the artificial cranial cast of Zeuglodon. 2 natural size. a, b,c, d, asin fig. 1. a’, the dorsal rostrum, and 4’, an irregular boss on the cerebral hemisphere. (These are probably due to imperfections in the cranium.) vertical, and of uniform thickness ; whereas in the artificial cast they are obliquely-placed, and expanded ventrally. In the natural cast the mesial continuation of these thick lateral masses (each of which measures 30 mm. antero-posteriorly) becomes reduced to a bridge measuring only 5 or 6 mm. [the exact figure cannot be stated, because a piece of bone (fig. 1, ¢) partially covers this region]. In the deep concavity behind the narrow bridge of the area c (in the natural cast) two rounded, irregular, walnut-like bosses project, one on each side of the middle line (fig. 1, d). Each of these is 26 mm. in 2A 2 326 > “Prof. G. Elietsmith, ~ [Jan. 14, diameter, and is placed so obliquely that its surface looks almost directly backward. Shallow but clearly defined furrows separate these two bodies from each other and from the areac. In the artificial cast there is only a very faintly-marked indication of these bodies (fig. 2, d). . At a first glance it might seem that they represent the whole cere- bellum, in which case ¢ would be part of the cerebrum! But careful examination of the natural cast renders such an interpretation highly improbable, and comparison with the artificial cast seems to finally establish the belief that the whole of the region marked c forms part of the cerebellum. It is extraordinarily difficult to accurately interpret this peculiar form of cerebellum. A comparison with other primitive types of cerebellum* points to the probability that the lateral buttresses of the mass ¢ represent the floccular lobes, and that the walnut-like mass (d) represents the cerebellar lobule which I have called “area C ” (op. cit., ‘Catalogue,’ p. 211). Tf it be objected that the lateral buttress-like mass is much too extensive to be entirely “ floccular,” attention may be called to the fact that in the large aquatic Sirenia, which have retained an exceedingly primitive type of brain, the floccular lobes are enormous in comparison with those of other mammals (op, cit., ‘ Catalogue,’ p. 346). It would perhaps be difficult to find elsewhere in the mammalia a greater contrast than that presented by the smooth, reptilian-like cerebral hemispheres of these casts and the highly complicated, ultra- mammalian neopallium of the recent whales, both Odontoceti and Mystacoceti.t And yet, if we inquire into the nature of the factors which have moulded the form and determined the size of the various parts of the brain in Eocene times and at the present, the contrast between the brain of Zeuglodon and the-modern Cetacea loses much of its significance, and becomes much less peculiar, even though it may not be wholly explained. In most Eocene mammals the cerebral hemispheres were exceedingly diminutive in comparison with those of their modern descendants and successors. Moreover, the bulk of the primitive mammalian hemi- sphere was composed of those parts (hippocampus and lobus pyri- formis), which are pre-eminently olfactory : in other words, the neo- pallium (é.¢., that part of the pallium which is neither hippocampus nor pyriform lobe) is especially insignificant. It is a well-known fact that the sense of smell loses much of its importance in mammals of aquatic habits (¢.g., Ornithorhynchus, the Sirenia, the Pinnipedia, and especially the Cetacea), and in these animals the olfactory parts of the * Compare, for example (‘Catalogue of the Royal College of Surgeons,’ 2nd edition, vol. 2, 1902), Armadillo (p. 211), Tapir (p. 311), Manatee (p. 346). + Vide ‘Catalogue of the Royal College of Surgeons,’ op. ci¢., pp. 8348—359. 1903.] The Brain of the Archwocett. 327 brain dwindle to very small proportions. In the Odontoceti the olfactory bulb and its peduncle actually disappear. The Archzoceti, therefore, are subject to two factors, which will account in some measure for their small cerebrum. For, in addition to the smallness of the brain to which most Hocene mammals are subject, there is their aquatic mode of life. This causes a reduction in size of just those portions of the pallium which form the greater part of the Eocene hemispheres. In the modern Cetacea the neopallium attains to the greatest absolute size which it ever reaches in any mammal. This fact cannot, however, be considered fatal to the belief in the close affinity of the Archeeoceti and the Cetacea, because the extraordinary dissimilarity between the brains in the two sub-orders is such as we know to: have been produced by the operation of well-recognised causes in the long lapse of time which separates the dawn of the Tertiary period from the present day. In all mammais which lead a life ‘in the open” it has become a condition of their survival that the neo- pallium must increase in size in each successive generation: failing this, the creature must either adopt a “retired and safe mode of life” or become extinct. Numerous examples might be quoted in support of this hypothesis. But the case of the Sirenia shows us how little we really know of the factors which determine the size of the brain. These creatures began the struggle for existence in Hocene times with relatively large brains, in spite of their aquatic mode of life; and they have been succeeded by generations of descendants whose latest progeny at the present day have a brain-equipment only slightly superior to their earlier Tertiary ancestors (vide ‘Catalogue,’ op. cit., p. 344, ef seg.). Even if we admit that the modern Manatees and Dugongs- lead an eminently safe and retired life, which is in marked contrast to the venturesome and “open” life of the whales and porpoises, much still remains to be satisfactorily explained. Perhaps the most striking feature of the brain of Zeuglodon is the extreme disproportion between the size of the enormous cerebellum and the diminutive cerebrum. In this respect the fossil brain presents a most marked contrast to that of all recent mammals, and especially to that of the Cetacea. This relatively great size of the cerebellum is not peculiar to the Archeoceti, but is common to other extinct mammals of large size. In my memoir on the brain in the Edentata* the difficulty presented itself of adequately explaining a similar phenomenon in Glyptodon ; and it must be borne in mind, in even attempting to do this, (1) that the obtrusive greatness of the cere- bellum presents itself only in large mammals and not in lowlier vertebrates, and (2) that the size of the cerebellum is not proportionate * “The Brain in the Edentata,”’ ‘ Linnean Society’s Trans.,’ 2nd series, Zoology, vol. 7, part 7, 1899, p. 381. 328 Prof. G. Eliot Smith. [Jan. 15, to that of the cerebrum. In the case of Glyptodon I four years ago attempted to explain these facts in this manner. The development of the neopallium in mammals opens up the possibility of the performance of many more complex muscular acts than are possible in the Amphibia or Reptilia: these acts require a co-ordinating mechanism, the size of which will be largely determined by the bulk of the muscular masses, the actions of which are to be harmonised, and the extent of the sensory surfaces which send into the cerebellum streams of controlling impulses. A large cerebellum is being demanded by a large mammalian body, even if the cerebrum is small. I cannot offer any more satisfactory explanation of the magnitude of the cerebellum in Zeuglodon than this. It is clear from the foregoing that the extraordinarily great contrast in the appearance of the brain of the Archzoceti and that of the Cetacea cannot be urged as a reason against their kinship, when it is remembered that the operation of known factors is quite sufficient to explain the transformation of the one type into the other in the time which has separated the Kocene period from the present. Having disposed of these negative arguments, we may consider the positive evidence for Cetacean affinity in the brain of Zeuglodon. The shape of the cerebrum, and especially its relatively great breadth, is peculiar. In fact, this form of hemisphere rarely or never occurs among mammals, other than the Cetacea. I have elsewhere* attempted to explain the shortness of the Cetacean hemispheres by the fact that the abortion of the basal (olfactory) parts of the cerebrum limits their longitudinal extension. This, however, is not the whole explanation, because in many microsmatic Sirenia (Halicore), and Pinnipedia (Ofaria, Phoca) the hemispheres are not especially broad. The disproportionate breadth seems, in fact, to be to some extent a characteristic of the Cetacea ; and, in this respect, Zeuglodon agrees with them. The peculiar elongation of the olfactory peduncles beyond the anterior extremities of the hemispheres is rarely found in mammals, though it is common enough in Reptiles and the Ichthyopsida. In fact, the exact parallel to the condition found in Zeuglodon occurs “among recent mammals only in the Cetacea.t An analogous condition is found in the extinct Lemuroid Megaladapis [described by Forsyth Major (op. cit.)| and some Amblyopoda. It is not without interest to note that the two outstanding features ot the cerebral hemispheres of the Archeoceti, even if their value as indices of kinship be slight, both find their nearest parallel in the * “Catalogue of the College of Surgeons,’ op. cit., p. 350. + Full references to this are given by Forsyth Major, ‘‘On the Brains of Two Sub-fossil Malagasy Lemuroids,” ‘ Roy. Soc. Proc.,’ vol. 62, 1897, p. 48, second footnote. 1903.] The Brain of the Archeoceti. 329 Cetacea. There are no characters of the brain of the modern Cetacea which can be regarded as certainly distinctive, if we put aside such features as the extreme dwindling of the olfactory apparatus, and the enormous development of the neopallium. Both must be regarded as late acquisitions, not to be expected in an Eocene mammal. Under these circumstances these slight points of positive evidence of the relationship of the Archeoceti and Cetacea must be allowed some value, as reinforcing the testimony of the skeletal parts. _Ifiwe seek to institute closer comparisons between the brain of Zeuglodon and of the Odontoceti and Mystacoceti with a view to the determination of its relationships, we are not unnaturally doomed to disappointment. It might, perhaps, be supposed by some anatomists that the absence of an olfactory bulb in the Odontoceti might point to a closer affinity of Zeuglodon to the Mystacoceti, in which a small olfactory apparatus is retained. But there is every indication that the olfactory apparatus of the Odontoceti has become aborted quite recently. Thus in a specimen of the embryonic brain of the Narwhal (Monodon), which was given to me some years ago by Professor Howes, the remains of the olfactory bulb (fig. 3, 6.0.) are still quite visible as a small umbili- Fie, 3..Ventral aspect of brain of an early foetus of Monodon, Natural size. a.d., locus perforatus (desert region); 4.0., bulbus olfactorius; /.p., lobus pyriformis. ; cate area in part of the “desert region” of Broca (fig. 3, «a.d.), wherefore it follows that in the early embryo the olfactory bulb and peduncle develop as in all other mammals. Moreover, in all Odontoceti traces of the pyriform lobe are found even in the adult ; and in the brain of Kogia Greyi the rhinal fissure and the typical (macroscopically only) pyriform lobe are retained’in a form as clearly defined as that of any macrosmatic mammal (fig. 4). Professor Haswell, in describing this brain* emphasises the fact that ‘the most remarkable feature of * W. A. Haswell, ‘On the Brain of Grey’s Whale (Kogia Grey?),” ‘Linnean Society of New South Wales Proc.,’ vol. 8, 1883 (publ. 1884), pp. 437—439, preext ~- - SSS SSS SSS | |! i } 330 The Brain of the Archeocett. [Jan. 15, the [basal] region, and perhaps of the whole brain, is the great depth of the ectorhinal fissure, a feature marking off the present form very strongly from Delphinus” (p. 438). Since his illustrations do not properly delineate this interesting conformation, Professor Haswell kindly permitted me to examine his specimen ; and Mr. J. P. Hill has made me an excellent photograph (of its ventral surface), roughly reproduced in the accompanying drawing (fig. 4). It shows the com- plete and quite-typical rhinal fissure and the characteristic pyriform lobe. In its anterior part the rhinal fissure is-fully a centimetre deep. Fra. 4.—Ventral aspect of left hemisphere of Cogia Greyi. Reduced approxi- mately one-half. a.d., corpus striatum (desert region); d.0., place occupied by bulbus olfactorius in fetus; f.r.@., fissura rhinalis anterior; f.7.p., fissura rhinalis posterior; /.b., lobus pyriformis. The exact reproduction of these characters of the rhinencephalon in an adult anosmatic Cetacean, and the presence of the olfactory bulb in the foetal Narwhal, show that these toothed Cetaceans were certainly (and probably quite recently) derived from ancestors presenting the normal mammalian type of olfactory apparatus. The absence of the olfactory bulb and peduncle in the Odontoceti cannot, therefore, be considered a just reason for adopting the utterly-improbable suggestion of a nearer affinity of the Archeoceti to the Mystacoceti than to the Odontoceti. Gian Kstimated by the amount of sand which it displaced, the bulk of the natural cast (ineluding that of a considerable quantity of matrix attached to the base of the brain and some small fragments of bone) 1903. ] The Differential Invariants of a Surface, ete. oa l is 410 c.c. If the necessary corrections and estimations be made from this gross cubic capacity, the weight of the brain in the Archeoceti must have been considerably less than 400 grammes, and perhaps nearer 300, as against that of the recent Cetacea, which ranges from 455 grammes in Kogia (Haswell) to 4,700 grammes in Balenoptera (Guldberg). “The Differential Invariants of a Surface, and their Geometric Significance.” By Professor A. R. Forsytu, Se.D., F.R.S. Received February 14,—Read March 5, 1903. (Abstract.) The present memoir is devoted to the consideration of the differ- ential invariants of a surface; and these are defined as the functions of the fundamental magnitudes of the surface and of quantities con- nected with curves upon the surface which remain unchanged in value through all changes of the variables of position on it. They belong to the general class of Lie’s differential invariants ; and some sections of them were obtained about ten years ago by Professor Zorawski, who, for this purpose, developed a method originally outlined by Lie. Earlier, they had formed the subject of investigations by a number of geometers, among whom Beltrami and Darboux should be mentioned. Professor Zorawski’s method is used in this memoir. In applying it, a considerable simplification proves to be possible ; for it appears that, at a certain stage in the solution of the partial differential equations characteristic of the invariance, the equations which then remain unsolved can be transformed so that they become the partial differential equations of the system of concomitants of a set of simultaneous binary forms. The known results of the latter theory can then be used to complete the solution. The memoir consists of two parts. In the first part, the algebraic expressions of the invariants up to a certain order are explicitly obtained ; in the second, their geometric significance is investigated. An invariant, which involves the fundamental quantities of a surface HK, F, G, L, M, N (these determine the surface save as to position and orientation in space) and their derivatives up to order n, as well as the derivatives of functions ¢, ¥%, of position on the surface up to order n+1, may itself be said to be of order n. The invariants up to the second order inclusive are obtained. It appears that, if two iunctions ¢ and y occur, all the invariants that occur up to the second order can be expressed algebraically in terms of VOL. LXXI. 2B 332 Mr. W. C. D. Whetham. Electric Conductivity of [Feb. 14, 29 algebraically independent invariants ; while, if only a single function ¢ occurs, all the invariants that occur up to the second order can be expressed in terms of 20 algebraically independent invariants. The significance of these respective aggregates of 29 and of 20: invariants is obtained in connection with curves p=0,~ = 0, drawn upon the surface. The investigation reveals new relations among the intrinsic geometric properties of a curve upon a surface. In particular, up to the second order, four such relations exist for a single curve ; and their explicit expressions have been constructed. “The Electrical Conductivity of Solutions at the Freezing-point of Water.” By W. C. D. WHETHAM, F.R.S., Fellow of Trinity College, Cambridge. Received February 14,—Read March 3, 1903. The following paper contains an account of experiments which bring to greater concentrations the series of measurements on the conductivities of dilute solutions at the freezing point, communicated to the Royal Society in February, 1900.* The work has been carried on at intervals during the last two years, and was made possible by the kindness of Professor Ewing, who placed a room at the Cambridge Engineering Laboratory at the disposal of the writer and his wife. The earlier experiments were originally planned in connection with the observations undertaken by Dr. E. H. Griffiths on the freezing points of corresponding solutions ; they were therefore conducted in a platinum cell of design similar to that used by Griffiths, with the object of eliminating any solvent action of glass. Any such action would be quite inappreciable at the concentrations used in the experi- ments now to be described ; resistance cells of glass were consequently used, and the labour of observation was much reduced. The structure of the cells is shown in figs. 1 and 2; that illustrated in fig. 1 was made of Jena glass, and reserved for the more dilute solutions. ach cell is so arranged that by applying a slight exhaus- tion, the whole of the contents can be drawn up into a bulb; by this means, when water or stock solution is added, complete mixture can be easily secured. The method of preparing the solutions was usually the same as that formerly employed; the pure solvent was placed in the cell, and its weight and resistance determined; weighed quantities * *Phil. Trans.,’ A, vol. 194 (1909), p. 321. 1903. ] Solutions at the Freezing-point of Water. O90 of stock solution were then added successively, the resistance being measured after each addition. When, however, the solutions ap- proached saturation, this procedure was reversed ; the stock solution was first examined, and was then diluted by a weighed quantity of water. ay PR a X! Fia. 1. Fie. 2. (4 full size.) (4 full size.) The resistance cell was placed in a tin bath filled with melting ice. At first, a thermometer was inserted in the cell, and the observations postponed till its readings became steady at 0° centigrade ; but it was soon found that the resistance itself gave a more sensitive and con- venient means of thermometry. Measurements were taken at intervals till they became constant ; the cell being always completely immersed in ice, the result thus obtained shows the resistance at 0°. As an example, the following numbers may be given :— Time. Resistance. 1D). Aud: 2179 WDD Z1OT We I 2202 eahG 2202 ers 2202 It was intended to make up the solutions in the resistance cell while it was immersed in the ice, but condensation of water from the atmosphere was found to occur on the inside of the cold vessel when stock solution was added. Before each addition, therefore, the cell and its contents were raised to some temperature near that of the 78 2 334 Mr. W. C.D. Whetham. ilectric Conductivity of [Feb. 14, room by. standing the cell for a few minutes in slightly warmed water. sp The water used as solvent was first boiled, then distilled in glass from alkaline permanganate, and finally redistilled in a platinum still with a trace of acid potassium sulphate. It was kept in a large stop- pered flask of Jena glass till required for the experiments. The samples of salt used to make the stock solutions of potassium chloride and copper sulphate were the same as those prepared for the earlier experiments; details will be found in the account of those - experiments to which reference has already been made. In the cases of barium chloride, potassium bichromate, and magnesium sulphate, the best salt, sold as chemically pure, was obtained, the two salts first mentioned being recrystallised before use. Any probable impurities would not affect the results to an amount equal to the other errors of experiment. The measurement of the electrical resistance was performed exactly as in the earlier set of experiments. The current from one or two dry cells was alternated by means of a revolving commutator, which was driven by a hand wheel and cord, the connections of a D’Arsonval galvanometer being simultaneously alternated by the same instrument. The alternating currents were passed through a Wheatstone’s bridge, in one of the arms of which was inserted the electrolytic cell. This apparatus worked with perfect success; its sensitiveness was enough to enable readings to be taken to an accuracy varying from one part in one thousand to one part in thirty thousand, according to the resistance in the circuit. The results which were obtained are collected below. In the table headed Potassium Chloride I, and in the tables for all the other salts, the concentrations, given in the first column under m, are calculated in terms of the number of gram-equivalents of salt per thousand grams of solution, while, for the sake of comparison, under Potassium Chloride II, the conductivities are reckoned per thousand grams of water. In the second column under m*, for convenience in plotting curves, are tabulated the cube roots of the concentrations ; the third column, R, shows the measured resistances of the solutions; next comes k/m = p, the equivalent conductivities ; and finally, in the last column, are placed the ratios of » to its value at infinite dilution, which was estimated from the former set of experiments. It was thought advisable, in view of the freezing point observations now being conducted in Dr. Griffiths’ apparatus by Mr. T. G. Bedford, to include magnesium sulphate in the investigation. This substance was not examined in the platinum cell, so that it was necessary to extend the experiments to very dilute solutions. This was done in cell No. 1, which is made of Jena glass, and therefore is not likely to react appre- ciably with a solution of the nature of magnesium sulphate. 1903. | Solutions at the Freezing-point of Water. 330 Potassium Chloride I. KCl = 74:59. In Cell No. 2, the water used had a resistance of 8:3 x 10° ohms at 0°. mM. ms. R. Keli = [. M/Mon« 0-02393 0-288 7894 755 0-936 0-04261 0-349 4504 744 0-922 0:07924 0-430 2475 728 0-902 0-1459 0-526 1374 MA 0-885 0-244] 0-625 839°4 701 0:868 0°4268 0:753 484°3 691 0°857 0°6379 0-861 325°5 688 0-853 0-8924 O7963 232°3 689 0:854 1 O78) 1-025 Oa 9 692 0°857 Potassium Chloride II. KCl = 74:59. mM. ms. 1. Klin = p. ES eae 0°02394 0-288 1894. 754. 0-934 0:°04275 0-350 4504 741 0-919 OOG97 I 0°430 2475 724 0°897 0:1475 0-528 1374 705 0:°873 0: 2486 0-629 835°4 688 0°852 0-4408 0-761 484°3 669 0-829 0°6697 0°875 329 °5 655 0-812 0°9575 0-986 232°3 642 0:796 131697 1:054 Iles 636 OSS Barium Chloride. 3BaCl, = 104:1. In Cell No. 2, the water used had a resistance of 8:5 x 10° ohms. Mm. ms. R. kim = p. H/o: 0-02040 0°273 10856 644 0°863 0:-05112 Ora 4584 609 0°816 Opals 0-485 2178 577 0°773 Ooi 0-619 1095 549 0:°735 0:4705 0:778 580°4 525 0:700 0°8710 0-955 329°5 504 0:674 17741 1-203 172°2 476 0-638 Copper Sulphate. $CuSO,5H,O = 124-87. Cell No. 2. The water had a resistance of 6°6 x 10° ohms. mM. ms. R. kim = p. BD 0-02174 0:279 17173 382 0-548 0: 07607 0-423 6397 293 0-421 0-2681 0-645 2359 226 0-324 0-5608 0-825 1313 19379 0-279 0-8440 0-945 Joli:5 Lol! 0-231 1°954 1-250 eat) ° 7 136°4 0-196 336 Mr. W.C. D. Whetham. lectric Conductivity of [Feb. 14, Potassium Bichromate. £KoCr20, = jb 2B, In Cell No. 1, the water used had a resistance of 7°26 x 10° ohms. mM. 0°002278 0° 006047 0°01475 0-04019 0-1010 0°2577 Magnesium Sulphate. Mm. *00000644 “00001398 -00003218 -00007707 °0003438 °001423 -005030 °02678 Io oo Ce eS Magnesium Sulphate. Me *04995 -1158 -2136 *3574 °6413 -189 °812 °642 "463 woreeF OCC CO oO 1SU) ms . "0186 "0241 "0318 "0426 O701 lal mci 2299 So oe 2 © © © 0-368 0°487 0°598 Ooo 0-862 1-060 1-220 1°383 1°513 R. 3004 1138 472°2 5240* 22028 Sule kim = p. 710 706 698 678 642 604 HB + 0-865 0-860 0° 850 0-826 O-782 0:736 In Cell No. 1, the water used had a resistance of 9°75 x 10° ohms. R. k/m = p 511000 702 330000 697 179200 687 84710 680 21230 652 5054 598 1881 513 456°8 397 = 8 SSSOSS6 6 © Ne) or OU 1MgSO,.7H,O = 123-26. In Cell No. 2, the water used had a resistance of 8°7 x 10° ohms. R. 8067 4083 2494 1654 1052 682° 540° 488° 524° em H CO kb Ol kim = p. oD4 302 268 241° 211° I) 145: 110° hs" AOnNNTeE GS O = ae zs 8 498 "4295 ‘377 "340 °298 "248 "205 *1556 DIOS SOO eo eee © In order to obtain the most probable results for the ratio of the equivalent conductivities to their values at infinite dilution, curves were drawn on squared paper between m and k/m, and the smoothed readings taken at the required places. It is usual to call this ratio the coefficient of ionization, but at the high concentrations here dealt * In Cell No. 2. 1903. ] Solutions at the Freezing-point of Water. 337 with, we cannot assume that it really gives the fraction of the number of the molecules which are at any moment ionized; in the light of probable changes in the ionic fluidity of the liquids, and of the possible existence of complex ions, such an assumption is clearly unjustified. For the sake of convenience, the results previously obtained, as well as those of the experiments now described, are collected in the follow- ing table :— Equivalent Conductivities at 0° referred to the Limiting Value as Unity. m=Number of Gram-equivalents of Solute per thousand grams of Solution. Mm. my. KCl. zBaCl,, | ¢K,Cr,0-. | xCuSO,. *MgSO, | O-O0001 | 0°0215 1 -000 1 -000 0-991 0-998 0-983 0:00002 | 0 °0272 1 000 1-000 0-930 0 °993 0°976 | 0°00005 | 0 -0368 1 -000 0 °998 0 °952 0-981 0°963 | O-0001 | 0°0464 0 -999 0°995 0 °929 0-967 0 °950 | 0:0002 | 0-0585 0-998 9 -990 0-902 0 °94:7 0 932 | 0.0005 | 0°0794. 0-996 0-980 0-880 0 °908 0-899 0-001 0-100 0-992 O- 969) (O28 70 0°863 | 0°864 0-002 0-126 0-987 0°953 | 0°864 0° 807 0 °814 0°005 0-171 0 976 0°925 0 °863 0°717 0°720 0°01 0°215 0-962 0 896 0 °858 0 638 0° 659 0°02 0-271 094.4: 0 °864 0° 84:7 0°d557 0 °587 | 0°03 0°311 0-932 0 °843 0 °834, 0°512 0 °545 | 0°05 0 °368 0-917 0°813 0 °815 0°468 0 °497 0°10 O °464: 0-896 0°778 0°783 0-405 0°435 0-20 0 °585 0 °874 0 °742 0°749 0 °348 0°380 0 -40 0°737 0-858 OL7O = 0-294. 0 °322 0°50 0-794 0 855 0-699 ase 0 °275 0°313 1) 1-000 0 °856 0 665 = 2 230 0 °264 We 1 063 0°860 0°657 — 0° 218 0-248 1°5 1°145 a 0 645 — 0-208 0 +227 2°0 1 +260 a= 0 °632 aaa 0 °194 0-192 3°0 1 °4.42 = | == = = 0-133 In the earlier set of experiments, approximate values only were ob- tamed for the absolute equivalent conductivities, changes in the adjustment of the platinum cell between the experiments on each salt causing a slight uncertainty in the cell constant, From the values of the constants of the glass cells now used, it is possible to calculate throughout the whole range of concentration of both sets of observa- tions the exact equivalent conductivities of the salts investigated. The results are given below, and are expressed in Kohlrausch’s units, in which the electrical conductivities, measured in reciprocal ohms, are divided by the concentrations of the solutions, measured in gram- equivalents per cubic centimetre. 08: Dr. Kohlrausch. The Resistance of the Ions [Feb. 17, Hquivalent Conductivities at 0° in Absolute Units. m = number of Gram-equivalents of Solute per thousand grams of Solution. | om me: KCl. | $BaCl,. | $K.Cr,0,.| 4CuSO,. | 4MeSO,. |’ sl il En es) Vans Ea iN a aE 000001 | 0°0215 | 807 746 813 696 699 0-00002 | 0-0272 | 807 746 804 692 694 | 0:00005 | 0-0358 | 807 745 781 684. 685 0-0001 | 0-0464 || 806 | 742 763 674 676 0-0002 | 0:0585 | 806 739 740 660 663 00) 0: 0005 OO 798803 731 722 633 6389 | 0-001 0-100 || 800 723 714 602 614k 0 002 OWS Fi Pes yal 709 563 57 OR nan 0-005 OL nasa 690 708 | 500 lp | 0-01 O:2ilan | 776) “| 1669 704 | 445 468 | 0-02 OL | | WG 645 695 | | 388 ALT 0:03 Osi) 752 629 6857 7) Sor 25 eae 0-05 0°368 || 740 607 669 | 326 Soca 0-10 0-464 | 723 581 643 282 309 0-20 07585 || 705 554 615 243 270). not 0-40 0-737 || | 692 53n — 205 229 0°50 0-794 || 690 522 a 192 2220 1-0 1:000 || 690 | 496 = 160 188)" | iLeQ 1:063 || 694 | 490 — 152 176 1°5 EP) een ABIL = 145 161 2-0 T2600 te ar eee 135 136 3-0 TA IP a ef _ 94,°5 “The Resistance of the Ions and the Mechanical Friction of the Solvent.” By Frrepr. Kouirauscu, Foreien Member R&.S. Received February 17,—Read March 5, 1903. (Translated into English for Dr. Kohlrausch by Dr. L. Austin.) Messrs. Bousfield and Lowry in their interesting paper, “ The Influence of Temperature on the Conductivity of Electrolytic Solutions,”* have discussed a hypothesis recently advanced by me. In this I stated the probability that the conductivities of all aqueous solutions approach, with decreasing temperature, a zero value at about the same temperature, and that the cause of this phenomenon is to be looked for in the disappearance of the fluidity of water. This hypothesis was very briefly mentioned, as it were, in parenthesis, in the midst of the discussion of the numerical data which formed the main portion of the paper.T * Bousfield and Lowry, ‘ Roy. Soc. Proce.,’ vol. 70, p. 42, 1902. + Kohlrausch, “ Uber den Temperatureinfluss auf das elektr. Leitvermogen von Lésungen, insbesondere auf die Beweglichkeit der einzelnen Ionen im Wasser,” ‘Sitz. Ber. Berlin Akad.,’ 1901, p. 1028. 1903.] and the Mechanical Friction of the Solvent. 339 I beg the honour of laying before the Royal Society the following more complete consideration of the subject. The above-mentioned article had for its object the study, from careful measurements made at my request by Mr. Déguisne,* of the influence of temperature on completely dissociated, that is, infinitely dilute aqueous solutions of strong electrolytes, and the deduction from this of the temperature coefficients of the single ions. Mr. Déguisne expresses the influence of temperature on the con- ductivity x, starting from 18° as a mean temperature, in the form of the quadratic interpolations equation Ke = Kig/ l+a(t— 18) + 6(é- 18)?] and shows that this closely represents his observations between 2° and 34°C. I shall confine myself to the consideration of this formula. (1.) Numerical values.—In the table, under A is given the equivalent conductivity of infinitely dilute solutions, then the coefficients « and ( for 1/1000 and, for the neutral salts, observed by Déguisne, « also for | a for | | ees | cog ! 1/10900. | 1/1060. | | } | j | KCl.. pe 130°1 0:0216 | 0:0217 | +0:000066 | KF ,. aa Petites 2 | 226 069 KANO. «. i se bz Pi) | 211 062 KeSO,.. si 133 °*4 2255 223 077 TRO 1308. eee c239 io 190 | 033 NH,Cl . i 30 219 219 068 NaCl .. ial 109-0 228 | | 084. NaF ... cate 90" 2 Ges | 240 100 NaNO,.. . | 105°3 220 221 075 Na.SO,. Occ rece sere ee ces | 112-2 234 234, 097 NaC SEOs... . io eS 8 6 242 | 110 H NaC,H,O, 69 = H 244; natal BAC! 0: | 98:9 e | 233 091 beac lo... : be By DI} 226 083 Ba(NO;)5: . + [este 22m 220 076 AgNO; . ; | 115°8 216 216 + 0000067 UO Meee os ohana sass s[ 883 a 165 —Q 000015 [EUG cae 5 ae el | 380 = 163 — 0 -000017 1/10000 normal solutions, corrected by me for the impurity of the water. Mr. Déguisne has kindly furnished me with the data regarding the water in each of his solutions. After applying these corrections the values for 1/10000 and 1/1000 normal do not continue to show the systematic changes, which from the observations appeared to * Déguisne, ‘‘Temperatur-Koefficienten des Leitvermégens sehr verdiinnter wassriger Lésungen,” ‘ Dissertation, Strassburg,’ 1893. t Kohlrausch u. v. Steinwehr, ‘Sitz. Ber. Berlin, Akad.,’ 1902, p. 581. J40 Dr. Kohlrausch. The Resistance of the [ons [Feb. 17, make the reduction to completely dissociated solutions very uncertain, but now show only differences of irregular sign within the limits of the errors of observation. From this we may assume that the true values of the coefficients for the concentration 1/10000 would be practically the same as those for complete dissociation. But, as a matter of fact, in the case of such dilute solutions considerable uncertainty underlies the observations themseives, on account of the variations of the solutions with time, as well as the corrections for the conductivity of the water, on account of its magnitude and somewhat uncertain theoretical basis. These facts are especially true in the case of the acids. Therefore, the values found for the concentration 1/1000, in which these two sources of error need scarcely be considered, must be looked upon as experimentally much more accurate. Then too, the condition of ionization of very dilute solutions of weak salts is always somewhat uncertain on account of possible hydrolytic action. In order to treat all electrolytes alike, I shall therefore choose the coefficients « and £ of the concentration 1/1000. The errors which may arise from the fact that these solutions lack from 2 to 5 per cent. of complete dissociation cannot at present be avoided. Since the values of the constants themselves differ but little from those for a concentration of 1/10000, we may assume that these errors are not great. The material is not yet complete enough to allow the application of the corrections proposed by Mr. Whetham. The values may indeed be capable of improvement, but I am of the opinion that the following conclusions drawn from them are in the main correct. In order to fill certain gaps in the series of the coefficients, | have added to the electrolytes investigated by Déguisne my own observa- tions on 1/1000 normal solutions of potassium and sodium fluoride, sodium acetate and lithium chloride. The coefficients here are derived only for the temperatures 10°, 18°, and 26°, and will therefore be less accurate than in Déguisne’s results. (2,) Values of the Conductivity at Different Temperatures—My con- clusion, that by extrapolating the quadratic formula below 0°, the conductivities of all electrolytes approach a zero value within a narrow range of temperature, rested upon an empirically established connec- tion between the coefficients « and PB. I will introduce here, however, the more easily followed graphic representation (fig. 1) plotted from the single coefficients. The extrapolated portions of all sixteen curves converge toward the same region and pass through zero between — 35° and — 41°.* * NH,Cl would nearly coincide with KCl; MgSO, on account of the large changes in its coefficient with increasing dilution is uncertain and therefore omitted ; it would probably lie near Na,SOx,. 1903.] land the Mechaiucal Friction of the Solvent. 341 ENG ale a, eee Vey ee ae ee ae a ae eee tt | VI Le SLCC Lae me | VV ove eo AT a. Coy Oy SER MAMD 20's VaD: Zo 4m CHA | | 10° ZO 30" 40° 50” i aztrapolated. metod 8-02 20° Or OF In regard to the lowest curve, cf. 4. The fact that the greatest differences exist for KOH and KF may be connected with the smaller accuracy of these curves. But on this we will lay no stress. 342 Dr. Kohlrausch. Zhe Resistance of the Ions [¥Feb. 17, (3.) Discussion of Results—We may derive from the above fact the certainty that the individual differences of the electrolytes come under a common law, the degree of accuracy of which must, however, remain unsettled. This law I have stated as follows: for dissociated aqueous solutions the coefficient 6 of the quadratic member can be approxi- mately expressed in terms of the coefficient « of the linear member, in the form B = C (a—A), where C and A are constants common to all electrolytes. One sees at once that this law is identical with the other ; all curves of the expression (1 + af + (7?) pass through the same - point, having for its abscissa — 1/C.* The proposed constants have the following values, taking 18° C. as the point from which the temperature is reckoned, C = 0:0163, A = 00174. The convergence takes place at the point where ¢-18 = —1/0:0163 = —61, or ¢ = — 43°. On account of the small difference between C and A, this point lies not far from the zero axis. If C and A were identical, the extrapolation according to the quadratic formula would show that the conductivity of all electrolytes becomes zero at the same temperature. Introducing this critical temperature fo, all electrolytes could be nearly represented by a formula containing but two individual! constants, Be ORO G sae On the one hand, I consider it impossible that the inequality of A and C, and the resulting deviations from a common point of converg- ence on the zero axis are produced by errors of observation. Even the circumstance that we have no completely dissociated solutions can scarcely have so great an influence. On the other hand, it appears very improbable that the approximate equality of the constants A and C is purely accidental. The deduction that the extrapolated curves all have a nearly common point of convergence appears to me especially worthy of notice in that this point hes approximately at the zero value of the conductivity. ‘The importance of this is still more increased by the fact that if the mobility of the water particles be extrapolated according to the same formula, it becomes zero at about the same temperature (cf. 4). (4.) Variations of the Fluidity of Water with Temperature. — The relations which have just been mentioned concerning the motions of the ions in water assume a greater interest when they are compared with the mobility of the water particles themselves. The fluidity (the reci- procal of the viscosity) of water when calculated in the same way as the conductivity, with the quadratic formula, is represented by the lowest curve. That the best observations on the fluidity of water agree excellently * The coincidence in the drawing differs a little from this, because each of the expressions is multiplied by its corresponding A. 1903. ] and the Mechanical Friction of the Solvent. 343 at moderate temperatures with the quadratic interpolation formula, already applied by Poiseuille, has long been known to me. The literary priority regarding this observation belongs, however, to Messrs. Bousfield and Lowry. In my calculations I have made use of the values chosen by Mr. Heydweiller,* as the most probable between the limits of 0° and 30’. These values were selected from different observers, especially Messrs. Thorpe and Rodger. Since the constants were calculated for this range of temperature, they have practically the same signification for the fluidity of water as the former constants for the conductivity. The formula , = 55°68 +1-981t+0-0105# C.G.S. or do, = 94-74[1 + 0-0249(¢ — 18) + 0-000111(¢- 18)] C.G.S. represents the fluidity within the limits of the table, 0° to 30°, with a maximum error of 1/1000, that is, with about the same degree of accuracy as Déguisne’s formula for the conductivity. As far as 90° the error would not exceed 1/100. The curve represented in the figure instead of having the factor 94-74 (which has no reference to the conductivity), was given the arbitrary factor 67:0 in order to give the curve the desired position close to the lowest curve of conductivity, that of sodium valerate. The coincidence of the two curves is striking, indeed the coefficients 0:0249 and 0:000111 differ little from those of sodium valerate 0:0244 and 0:000111.7 The curves of conductivity are cut by the fluidity curve approximately in the same region to which they converge. The curve of fluidity passes through zero at — 34° C.t Messrs. Bousfield and Lowry calculated from the measurements of Thorpe and Rodger the coefficients 0°0251 and 0°000115. This curve differs very little from mine, especially in the neighbourhood of the crossing point. * Mr. Heydweiller calculated at my request the table for the ‘ Lehrbuch der Praktischen Physik’ (Tab. 20a, 1901). 7+ The same would apply for Déguisne’s observations on Na,HPO, (0°0241 and 0000105) and on NaHC,H,0, (0°0241 and 0:000109), which, however, on account of the unknown constitution of these salts in solution, I have left out of account. Messrs. Bousfield and Lowry further called attention to the fact that the tem- perature change which I have found for ordinary distilled water (practically a very dilute solution of CO) corresponds with the temperature change of fluidity. £ It is also to be mentioned that Messrs. Lyle and Hosking, from their interest- ing observations on the viscosity and the electrical resistance of 0:1 to 4 normal solutions of NaCl between O° and 100° draw the conclusion: “The curves so arrived at are remarkable, in that they indicate that for solutions of the strengths experimented with, both the fluidity and the sp. mol. conductivity vanish at a temperature of —35°5C.” The manner of extrapolation is not dealt with.— ‘Phil. Mag.,’ May, 1902, p. 496. O44 Dr. Kohlrausch. Zhe Resistance of the Ions [Feb. 17, The fact is therefore established that the temperature change of the fluidity of water is nearly the same as that of the conductivity of dissociated aqueous solutions of electrolytes which have a large tem- perature coefficient. Even if nothing more was known than this fact, the question of a connection between the electrolytic and the mechanical motion in water must be considered a matter for serious cliscussion. (5.) Discussion of the Hautrapolation—HExtrapolation of an empirical formula over a wide range can never be considered as necessarily repre- senting the truth. This is especially true in a case like the present, where at low temperatures it is applied to a state of matter other than that for which the formula was originally deduced. It is a priori impos- sible for the formula to hold where its extrapolation gives to the con- ductivity or the fluidity a value zero. Since these quantities are from their very nature positive, negative values are physically impossible. Therefore the cutting of the zero axis by the curve at an acute angle is «@ priori inadmissible, just as, for example, the assumption is inadmissible that the Joule heating effect is proportional to the current strength, or that the kinetic energy is proportional to the velocity. A quantity from its nature positive can, as it becomes zero, have a finite differential quotient as function of another quantity, only when the other quantity cannot vary beyond the critical point. This can be considered identical with the impossibility of negative values. In reality the conductivity and the fluidity must reach the zero value in a curve which is tangent to the axis of temperature. (Becoming zero through a discontinuous process as in freezing is, of course, something entirely different.) Therefore the quadratic formula, in spite of the fact that it shows such a remarkably wide range of applicability, must be replaced by another expression before the zero value is reached. The above explanation shows that my view of the “critical tem- perature” of the fluidity and the conductivity of water as derived from the quadratic formula, does not materially differ from that of Messrs. Bousfield and Lowry. This temperature is only a quantity by which one constant of the ordinary formula can be replaced ; but the importance of the constant now introduced is verified, in that now the individualities of the ions, if they do not entirely vanish, at any rate disappear except for small differences. Further, the remarkable fact follows, that approximately the same constant may be intro- duced in the temperature formula of the viscosity of water. This number, entering as a temperature, may therefore be called a funda- mental constant of water, of course with the reservation which follows from the fact that it varies by several degrees in the different cases.* (G3) * The objection that the use of such a constant may be responsible for the 1903. | and the Mechanical Friction of the Solvent. 345 (6.) Other Formule.—Among the former attempts to derive an empirical formula from the behaviour in ordinary temperatures, which does not “ ~priore lose its significance as the conductivity or the fluidity approaches zero, that of Slotte, 6=¢o(1+0t)", must be given the first place. This has recently assumed a greater importance on account of the extensive work of Thorpe and Rodger. Here, in the case of the fluidity, the condition that dp/dt = 0 where ¢ = 0 is fulfilled, for here everywhere »>1. For the conductivity, however, nothing is gained, since in the case of the acids n<1, and there dx/dt = © for x=0; there is, theretore, no object in recalculating our results according to this formula. However, in a former paper I have given the preference to Slotte’s formula in a remark concerning fluidity. To try to bring the coefficients of this formula into a relationship with those of the quadratic formula was indeed not allowable. (7.) Haperimental Indications—The attempt to draw conclusions in regard to the region near the zero value from the course of the phenomenon in the region where the fluidity &c. have values of considerable magnitude, would have, of course, very little prospect of success. But perhaps the attempt to follow the fluidity of water or the conductivity of dilute solutions down to lower temperatures would be more successful than we think, if made in closed vessels of small dimensions. Up to the present time we are acquainted only with phenomena from which uncertain conclusions from analogy can be drawn. The idea may be pretty certainly held as probable, that the viscosity and the electrical resistance finally increase more slowly than the quadratic formula extrapolated would indicate. In regard to viscosity, I remember the investigations of Tammann on the freezing of over- cooled liquids, and Ostwald’s observations on salol. The phenomenon of the gradual solidification of alcohol at low temperatures also leads to the same conclusion. Numerous observations of the same kind in regard to electrical behaviour are also recorded. The very slow increase of the conduc- tivity of glass with increasing temperature is well known. A quanti- tative determination of this has been made by Messrs. Bousfield and Lowry. But how far a heterogeneous mixture like glass can be considered parallel to a dilute ionised solution is doubtful. In the same way, experiments such as those of Lehmann on the electrolysis introduction of errors is without much weight ; it is, at any rate, practically the same case as that of the gas formula. This is written, making use of the absolute zero, vp = const. (f)+¢), where ¢) = 2738, although it is certain that the result v = Ofort = —f, is false, so that the formula must assume another form before v approaches its zero value. ‘The constant of the gas formula loses its significance at the point where the gas passes into the liquid state. The same is true for the temperature constant of the fluidity, which we have introduced, when it passes out of the liquid state. 346 Dr. Kohlrausch. Zhe Resistance of the Ions [Feb. 17, of solid silver iodide, or those of W. Kohlrausch on its conductivity at different temperatures, scarcely form a basis for conclusions in regard to solutions in water, particularly as the behaviour of silver chloride and silver bromide is quite different from that of silver iodide. In addition, it is difficult to compare substances which are so highly concentrated, in comparison with ordinary electrolytes, with dilute solutions. All these conductivities reduced in the ratio of the con- centration, as, for example, the measurements made on hot glass of large surface and small thickness by means of the potentiometer, would probably have to be considered zero, that is, smaller than the errors of observation, in comparison with our dilute salt solutions at moderate temperatures. Of the greatest interest for our problem are the recently published measurements of electral conductivity as far as — 70° C., which Mr. Kunz* was led to make through my remarks on the relation between conductivity and temperature. He was unfortunately unable (as was I also) to sufficiently over-cool dilute solutions, and the measurements were therefore made on strong solutions of sulphuric acid of at least 4 er. equiv./litre. Notwithstanding the fact that no certain conclu- sions can be drawn in regard to dilute solutions, Mr. Kunz’s values are of sufficient interest to represent graphically. ‘The curves are marked with the percentage concentration of the solutions (fig. 2). Bre. 2. The conductivity of the strongest solutions sinks gradually with the temperature, and reaches at — 70° a relatively small value. It is im- possible to follow the more dilute solutions to so low a temperature on account of their freezing. But it is evident from the observations that the more dilute they are, the more rapid is the rate of change of the * Comptes Rendus,’ vol. 135, 1902, p. 788. 1903. ] and the Mechanical Friction of the Solvent. 347 conductivity, so that it may be supposed that the 19 per cent. solution would reach a relatively negligible value at a much higher tempera- ture than the stronger solutions, if it were possible to follow it. Mr. Kunz adopts the view that the electrical resistance is due to friction. It is his opinion, indeed, that the conductivity would disap- pear only at the absolute zero of temperature; this conclusion can hardly be supported by his observations, as his lowest temperatures are still + 200° absolute. (8.) The Temperature Coefficients of the Single Ions.—For these* I have recently published values. It is of importance to us that the tempera- ture coefficients of univalent monatomic ions appear to be functions of mobility, decreasing as the mobility increases. Complex and multi- valent ions as groups deviate from this series, so that in this relation- ship we have a new criterion for univalent elemental ions. The largest temperature coefficients of the ions approach that of water. (9.) The Electrolytic Resistance Considered as Friction of the Solvent.— In the common view concerning the motion of the ions, an assumption is tacitly made which in other cases we do not consider justifiable. In the relative motions of adjacent particles we assume a discontinuity only in the case of friction between two rigid bodies where this by definition must occur. Even here it is impossible to deny that on the actual surfaces of contact there may be connected with the motion a rubbing away of particles which produces a continuous variation of velocity from one to the other. When a fiuid is in question, whether in contact with another fluid or with a solid, we concede no finite variation of velocity in two points at an infinitesimal distance from each other. The primitive assumption, until recently held to be correct, in the case of mercury on glass, that the fluid in actual contact with the solid moved with a finite velocity, would demand that the external friction be infinitesimal in comparison with the internal. This is now, to the best of my knowledge, entirely given up.t The idea of discontinuity, however, we employ in regard to the ions when we think of them as moving through the solvent without connection with it. In addition to the objection of discontinuity there exist also the following difficulties in this assumption. In the first place, it is diffi- cult to see how the electrical energy passes into the solvent in the form of heat, unless the latter takes part in the motion of translation. Wurther, it seems probable, from the fact of the ionisation of the salts, * ‘Sitz. Ber. d. Berlin. Akademie,’ 1902, p. 572. The values here given are strengthened by the fact that a linear connection between a and 8B (cf. 3) appears @ priori in the case of the single ions as well as in the case of electrolytes. + Comp. Warburg, ‘Pogg. Ann.,’ vol. 140, 1870, p. 379. The “slipping”’ of rarefied gases on solid surfaces, established by Kundt and Warburg, being a separate phenomenon, need not be considered here. VOL, LXXI. 2 C 348 Dr. Kohlrausch. Zhe Resistance of the Ions [Feb. 17, that forces must exist between the 1ons and the water. Ciamician,* twelve years ago, concluded that this must lead to the hypothesis of a water-shell about the ion. The attempt was also made to measure this hydration, by using a method proposed by Nernst. We may, therefore, look upon it as probable that the moving ion earries with it amass of adhering solvent, just as a moving immersed solid carries with it a portion of the liquid, and we will endeavour on this basis .to construct a new representation of electrolytic resistance. According to this the mechanics of electrolysis assume an appearance quite different from the old hypothesis of isolated ions. The resist- ance to motion takes place not directly between the solvent and the ions, not between H,O and K or Cl, &¢., but it is a phenomenon of friction between the particles of the solvent itself, modified by the fact that the accompanying shell of solvent may be thin enough to allow the ion to act through it, to a certain extent, upon the outer liquid. It is impossible at first to make anything more than this rough sketch of the hypothesis, and the more so, as the expressions “ continu- ous” and ‘‘discontinuous” must be especially defined if we are to apply them to molecular processes. The expression ‘ continuous” stands from its very nature in contradiction to the atomic or molecular hypothesis ; and in the case of solutions, and especially in electrolysis, a molecular representation seems to be the only one which is scientific- ally thinkable. I hardly need after this to say that in our hypothesis we shall not claim to be able to differentiate strictly between the outer portions of the solution and those parts which have separated themselves from the rest as the accompanying atmosphere of the ion. A continuous change in the condition of motion from the moving ion to the solvent con- tradicts a strict differentiation. Fundamentally, however, the same is true for all atmospheres, even for that of the earth. For the sake of brevity, we will retain the expression that an atmosphere of the solvent takes part in the motion of the ion. In the light of this hypothesis, I believe that all the phenomena which have been here described become so much clearer that this fact itself serves as a remarkable experimental support of the hypothesis. (10.) Hypotheses and Conclusions—The hypotheses are: About every on moves an atmosphere of the solvent, whose dimensions are deter- mined by the individual characteristics of the ion. The atmospheres of multivalent or compound ions differ from those of monatomic ions. Data are at present lacking for a more detailed representation. The electrolytic resistance of an ion is a frictional resistance that increases with the dimensions of the atmosphere.j The direct action * Zeitschr. f. physikal. Chem.,’ vol 6, 1890, p. 405. + The resistance of a sphere is proportional to its radius. Kirchhoff, ‘ Vorles. ub. math. Physik.,’ 1, 380, 1897 (herausgegeben von W. Wien). 1903.] and the Mechanical Friction of the Solvent. 349 between the ion and the outer portion of the solvent diminishes as the atmosphere becomes of greater thickness. Conclusions.—(a.) ‘The electrical resistance of an ion, expressed in mechanical units, must be of the same order of magnitude as the mechanical frictional resistance of a molecule of the solvent; a law whose assumption, as ] some time ago showed, “leads to an expression for the distance between the molecules which is comparable with the usually accepted views in regard to this quantity.”* (b.) The empirically discovered law that the temperature change of the resistance of the most sluggish ions is very like the temperature change of the viscosity of water, becomes now understandable. For ions of large resistance we must assume that the atmosphere is of considerable thickness, and hence the action of the ion itself on the ‘outer portion of the solvent will be small. As a limiting case, for a very sluggish ion there will be only the friction of water against water, and the electrolytic resistance will have the same temperature coeffi- cient as the viscosity of water, provided that the atmosphere itself does not change its dimensions with the temperature. If, however, the atmosphere become, for example, smaller with increasing tem- perature, the temperature gradient of the conductivity might be greater than that of the fluidity. According to the observations now at hand, this would seem to be the case for the slowest moving univalent ion Li. Even here, however, the differences scarcely exceed ‘the errors of observation. (c.) I now come to the remarkable relationship between the mobility ‘of the ions and their temperature coefficients, which was mentioned in Section 8. This first led me to seek a general explanation for the electrolytic resistance in the idea of a water atmosphere, in order to ‘escape being compelled to explain this otherwise unreconcilable funda- mental characteristic of the ions as a deus ex machina. Assuming as the single fundamental characteristic of each univalent monatomic ion the formation of a water atmosphere which varies according to the nature of the ion, the mobility of this complex on the one side, and its temperature coefficient on the other, will be functions of these atmospheric formations, and therefore both quantities must hold functional relations to each other. We know too little of the molecular forces at present to attempt to describe this -connection more exactly. But for the case in which the water shell is so thick that the jon exerts no force beyond it, the resistance to motion becomes simply a matter of water friction, which explains the fact ‘that the most sluggish ions have nearly the same temperature coefficients as the viscosity. In the case of smaller aggregations, we must remain content with the fact that we have at least the possibility of a fundamental explanation. * ‘Géttinger Nachrichten,’ 1879, p. 1. RS) Q io) 350 Resistance of the Tons and Friction of the Solvent. [Feb. 17, There are two experimental questions which are of importance although difficult to answer : first, whether the functional relationship is: exact or only approximate, and second, whether the positive and negative ions are fully identical in regard to this relationship. That the non-elementary ions also show as their greatest temperature coefficients that of water friction, but that they as groups differ from the elementary ions, is to be expected. The latter fact cannot be quantitatively explained. Here also it will be necessary to wait for more exact experimental data to settle the question. (d.) Finally, the indication of the temperature formula that the mobility of all the ions converges towards zero (cf. 2 and 3) at about the same temperature, is a logical result, if the electrolytic resistance is in reality a mechanical friction. The fact that the formula for the fluidity of water takes part in this convergence, gives the hypothesis further support. - It does not seem at all impossible that the deviations from a strictly common zero point, found in extrapolating the different formule, have- a systematic cause. ‘These divergences seem to indicate that the more mobility an ion shows at ordinary temperatures, the more slowly relatively it loses the residue of its mobility as the solvent becomes. more viscous. ‘The mobility of the water molecules themselves becomes small at a comparatively high temperature, where such ions as K, Cl, NOs, SOy, and even more, OH and H still possess a con- siderable residue of electrolytic mobility. Such a relationship does not seem at all improbable. In the foregoing pages I have sought to find a cause for the electro- lytic resistance in the single fundamental characteristic of the ions, their hydration, that is, their ability to form atmospheres from the solvent. ‘These views form a hypothetical sketch for the completion of which much is still wanting. It appears to me, however, complete enough to invite one to its experimental or theoretical continuation. 1903.) Lininwinising Effects of Contents of Typhoid Bacillus. 351 “Upon the Immunising Effects of the Intracellular Contents of the Typhoid Bacillus as obtained by the Disintegration of the Organism at the Temperature of Liquid Air.” By ALLAN MacraDyEeN, M.D. Communicated by Lorp Listrr, O.M., PLS. Received February 19,—Read March 12, 1903. In a previous communication® it was shown that it was possible to disintegrate mechanically the typhoid bacillus at the temperature of liquid air, and to obtain the cell-juices of the organism. The typhoid cell-juices obtained by this method on inoculation into animals proved toxic or fatal. It was, therefore, concluded that the typhoid bacillus contained within itself an intracellular toxin. It remained to test the typhoid cell-juices for immuninising and other properties. The preliminary experiments in this direction, which form the subject of the present note, were made upon the monkey. The monkey was selected as an animal most likely to furnish data of possible application to man. For this purpose the typhoid cell-juice was administered subcutaneously to the monkey. The injections did not produce any general symptoms beyond a transient rise in temperature, whilst the material was quickly absorbed after each injection without any traceable local effect. In this manner doses of 0°5 to 1 ¢.c. of the material were injected at intervals. An immediate result was the agglutination of the typhoid bacillus by the serum of the blood of the treated monkeys, whereas no such effect was produced by the serum of monkeys which had not been treated. This furnished useful evidence that the animals were under the influence of cell-juices derived from the typhoid organism. ‘The injections were repeated at intervals of three to four days, and after a lapse of four to six weeks the animals were bled. The serum obtained was then tested for immunising properties. The test objects were (1) a virulent culture of the typhoid bacillus, and (2) the intracellular toxic juice of the same organism. A varying amount of the virulent bacilli and of their toxic cell-juice was mixed with a varying quantity of the serum. The respective mixtures were then injected into the peritoneal cavity of the guinea-pig. The broth cultures of the typhoid organism used in the experiments were per s¢ lethal in doses of 0°1 c.c. in five to ten hours. The typhoid cell-juices were fatal in doses of 0-2 and 0:1 ¢.c. in three to five hours, and in doses of 0°05 ¢.c. in about twelve hours. The serum was thus tested for (1) specific antibacterial and (2) specific antitoxic properties. The experiments showed that the serum of the monkey, after injec- tions of the typhoid cell-juices, possessed antibacterial and antitoxic *- “Roy. Soc. Proc,,; supra, p. 76. 302 Immunising Effects of Contents of Typhoid Bacillus. (Feb. 19, properties, inasmuch as the serum protected the experimental animals: against the bacilli, and also against an intracellular toxin obtained from them. A simultaneous injection of (1) serum with the bacilli, and (2) serum with the toxic cell-juice produced no lethal or toxic effects. The control animals, on the other hand, invariably succumbed. It was further investigated whether the serum possessed preventive and curative properties. The serum from treated monkeys was injected into guinea-pigs, one injection being made in each instance, and the same animals received at an interval of 12 to 24 hours lethal doses of the typhoid bacillus and of its toxic intracellular juice respectively. The treated animals survived the test, whilst the control animals. succumbed. It was therefore concluded that the serum had protective: properties. A third series of guinea-pigs received lethal doses of the typhoid bacillus and of its toxic cell-juice respectively. The serum was then injected at various intervals into individual animals. It was found that the lives of the animals could be saved by one injection of the serum from a fatal infection or intoxication, even when half of the lethal period had elapsed in each instance. The serum, therefore, possessed curative properties. In view of the above results it appeared desirable to test the effect of the typhoid cell-juices upon animals larger than the monkey, in order that a greater amount of serum might be obtained conveniently and tested quantitatively as regards antibacterial and antitoxic pro- perties. These experiments are at present in progress, and the results will be communicated in due course. From the experiments made upon the monkey it would appear :— (1.) That by the injection of the intracellular juices of the typhoid organism into the monkey it is possible to obtain a serum with both antibacterial and antitoxic properties. (2.) That such a serum possesses curative and preventive properties as regards the typhoid bacillus and an intracellular toxin present in the same organism. ] fo} Result.—Geotropism (average)...... Cool 47° Hot 2 Hichotropisml sy), meee sh fl 7 9) i These and other similar experiments showed us that we had no right to conclude that the loss of geotropic capacity depended on the absence of the special mechanism (statoliths), but rather that the loss of the starch may perhaps be no more than a symptom of exhaustion which shows itself both geo- and heliotropically. Némec has shown that in decapitated roots, 7.¢., roots from which 1903. ] The Statolith-theory of Geotropisn. 365 the tip containing the statocytes have been removed, the capacity tor geotropism returns with the regeneration of the statoliths. Hven here—though the fact is a striking one—the argument seems to require the heliotropic test. For obviously the regeneration of the tip may mark the recovery of a generalised sensibility, and not merely the rehabilitation of the special gravitationai mechanism. The same objection holds to some extent with regard to Haberlandt’s* experi- ments on plants found to be starchless in winter. It should have been more definitely shown that they are heliotropically active though incapable of gravitational reaction. The Tuning-fork Method. In the autumn of 1901, I began a series of experiments by a method which was at the time a new one, but has been in the meanwhile published by Haberlandt.— It seems to me that Haberlandt’s argu- ment is open to the objection above set forth, and as it is an objec- tion I have tried to meet, my results seem to be worth giving. My point of view was that if gravitational sensitiveness is a form of contact-irritability (which must be the case if the pressure of the statoliths on the plasmic membrane is the critical event), then it might be possible to intensify the stimulus by vibration. I hoped, by applying vibration in a vertical plane to a horizontal seedling, to make the starch grains dance on the lateral walls, and by such repeated blows on the protoplasm to produce more active geotropic response. The experiments were made with seedlings of Sorghum, Setaria, and Panicum. Inthe earlier trials entire seedlings were used, but they were found difficult to fix horizontally with sufficient accuracy, and I consequently employed cut hypocotyls cementel by means of melted cocoa-fat on cork supports, and kept damp in small metal boxes, each containing a strip of wet filter-paper. _ The vibration was supplied by means of a tuning fork driven by an electric escapement. ‘The fork was fixed in a horizontal plane so that the vibration was vertical. The amplitude of the vibration varied in different parts of the fork from 4 mm. to less than a millimetre. The rate was about 47 vibrations per second. The general plan of the experiments was to attach a pair of metal boxes, one to each limb of the fork, each box containing four to six seedlings fixed approximately horizontal. Control boxes were placed on a support a few centimetres from the fork. It was found essential to insure an identical temperature for the experimental and control plants. The fork is set in motion and the experimental plants sub- * Haberlandt, ‘ Pringsheim’s Jahrbiicher,’ 1903, p. 472. + Loc. cit. (1903), p. 489. 2D 2 366 Mr. F. Darwin. [ Mar. 6, jected to vibration for about 8—20 minutes. The control and experi- mental boxes are then placed on a klinostat to avoid further gravitational stimulus, and the angular curvature estimated after a few hours. The general result is clear, viz., that the plants subjected to vibration are more strongly curved. In other words, that vibration increases the gravitational stimulus. Precautions and Sources of Error. In performing the experiment it is necessary to take scrupulous care that the control and experimental plants receive similar treat- ment. In the case of both it is necessary to cement the plants into their boxes without allowing any possibility of geotropism being induced before the boxes are fixed to the tuning fork. In all the later experiments the plants were cemented into their places vertic- ally and received no gravitational stimulus until horizontal on the fork, or on the place prepared for the control plants, as the case might be. Any error arising for the want of this precaution in the earlier experiments was equally divided -between the experimental and control plants. The most serious source of doubt and error is the tendency in the grass seedlings to nutate in various directions ; all that can be said is that the error in question is equally applicable to the experi- mental and control plants. Sorghum saccharatum is especially lable to nutation, and had to be discarded as material. A theoretical source of error arises from the fact that the movement of a vibrating rod being part of a curve there must be a generation of centrifugal force parallel to the rod and towards its free end. If the force were sufficient 1t would cause a displacement of the starch in the cotyledons of the grass seedlings, and this would affect those seedlings whose apices were directed towards the free end of the vibrating rod, precisely as if they were in an oblique position, the apices being down- ward. Now this position is known* to be more favourable to geo- tropism than horizontality, therefore the plants on the vibrating rod could not be fairly compared with the motionless specimens. But we were unable to see any displacement of starch in seedlings exposed for short periods to the vibration of the tuning fork. By a rough method I estimated the centrifugal force in my experiments as about 0-2 g. This is a force quite sufficient to affect the starch grains, if enough time is allowed; but it is hard to believe that for periods of 8—20 minutes it could have any serious effect. However this may be, I avoided, in the later experiments, all possible favouring of the experi- mental plants, by placing them with their apices towards the base of * Czapek, ‘ Pringsheim’s Jahrbiicher,’ vol. 27, 1895, p. 283. 1903.] The Statolith-theory of Geotropisnr. 367 the tuning fork. In this case any acceleration in the line of the fork would be in favour of the “ still” specimens. Table I requires a few words of explanation. The third column gives the number of the experiment and simply refers to the original notes. The next column gives the time in minutes during which the plants were left horizontal. Then follow the temperature and the length of time during which the plants were left on the klinostat. When two readings occur under this column it means, of course, that the curvature was measured more than once: thus in experiment 74 the curvature was read after 4 hours’ rotation, and again after 6h. 33m. from the time at which the plants were originally placed on the klinostat. The column ‘‘ Shaken” gives the angular deviation from the horizontal of the plants which had been on the fork ; their average curvature follows in a separate column. The last two columns give in the same way the actual observations and the average for the control plants. For the sake of brevity we use the terms “Shaken” and “Still” for the experimental and control plants. In columns “ Shaken and “Still” the letter z occurs occasionally in brackets, thus (x): this means that nutation downwards or sidéways had occurred. In the same way (5) indicates a downward or lateral nutation of 5°. In striking averages, | have counted the nutating specimens and also those which showed no geotropism, i.¢., remained horizontal. Thus, if the readings were 13, 15 (z) 17, the average would be 2 =i lla: 2 or if it had been 13, 15, 0, 17, the average would have been the same. When the cases of nutation were very frequent, or where the curvature occurred very slowly, as in experiments 131, 140, I have omitted the average curvature from consideration, although I have allowed the readings to remain in the table. The first thing that strikes one is that the shaken specimens show a greater curvature than the still ones in a large majority of cases. Thus, taking the whole of the thirty cases in which the average curvatures are calculated, we have :— Shaken, curvature greater in ............-.- wy ,, equals or practically equals stl 3 Still, greater than shaken ........ parapet 2 Out of these thirty cases, seven are second readings, 7.¢., records ot curvature in experiments in which readings had been taken a few hours previously. If these are omitted, we get :— Shaken, curvature greater......... 1 je ME QUIGES SUZDUF be. tots Biee ose y Sill, curvature greater ...........- 2 Finally, there are four cases in which the plants were leit on the fork {r. F. Darwin. i oe Si, Ol SY a8 by. 6r ee "" 86 ZL 2 O ie ee = Oe ON tle a ae EROS OG oR “OR | eo. ee i tS Oe OL BO ad) Bf Wears. Oo “O. -OL Me Oee O88 a es Sere 02 OL | ek oe G. 61 Z | ISL |‘ ‘hvu buoy |***at =“ 1 SEL oe Ce ah) eo be.s 1 ap aOR de G eS Tach obec ‘ eee SO Se | OF & G. 61 GQ | OSL |° ‘baw buoy |**'9T “ tn eR re Oey bs ag Ae oO GT. Or Ze ee ae RT ee 0 j ee = men son. Chemo TZ GZ | 82l |°° 2072 ‘buoy | °° "TT ‘ave ‘ZOGL ee i SOC One et ee ee Opes oe. \—- Ole ZE—1e | 09 | ZIT | ‘4yna buoy |etg 91 (02 8t) OL 9 9 6 ee eye Ao WA (a) OL YP 0% 28. | GIT |°"bjna shuog js" 'og OGD) HE MOI) (0) 2D ye Koni i 0h =) OF a0 SesOn ‘at he eS) Oe TOR Ie | ken €Z—Ze | O@ | 98 "Byna Bioy |G ms oe Oe Gr Reo } oz sea "* 8 G9 82 8I -_— 3 ee Tei BO, 208) = OL : a NO SRee Cp ee On|. 08, 2 uD OZ eS eee ask Gn Ke see OV Gin Ol ae ce OL - 9 OL OL (08 OL ot) 6 3 ne os | 64 ‘yoons ‘buog |\***g * (18199%[ 06-08) FI gz be Of (8) AOS Oc cr = le " 02 | 94 |" byna sbtogy \e°"g “08g : ra te, ip e Oliaeee. 10: HR STO ee! ap Oto BeOk: | @ Vase ie of 92 S& SI ST! gs 9 ge 87 ; OF 0 cl ae 0 2009291 PA eps rate MOS eA CO | (0) iy 61 02 | FL CED SEDLO SK 2h Oa aa, : : co lh G2 Ze O08 &F Te EO! oe me Pale oe 13: Ob Zp se se of] 9 9 | Or | ge | Oyna Bing) | gg ES a pow OF Sh &% O OF O08 | ag I ia a ee Sl sion GL oc = or 0% ot | 19 | * Byna ‘Buoy |**+yg *aonr ‘um Uy “ul "LOGI 1108 "AV “TOBY “OUT Mf VD Ung | “XOL ‘quel ‘mstdo1oo0gj—'T 914], 19 co) Yea > ina) — ‘opism. The Statolith-theory of Ceoti g0 kt aan eo ao oO @ f(s) & CO O 0 ee 8G) (8) @) ¥ (8) 0 0 (OT) (83) @)2 0. 8 SE O- £ OL 8 Qe 02-66 Ot 7) Cc Py OL SL StI (S(O pf 9 ST 9 6— 6. =o @e-6 il @) ie @ @ (@) er Oe Ge © LZ OL (® ®) Tt Of ce 19 91 9% &% 6I QO FL &L PI Oo oO g @) (2) toed 19 02 O& OF PL ZL ZL 2S 06 89 89 4&6 08 O08 SF O08 sz 0 89 06 8Z O08 ZS &9 TIS it SS ee eee Gataan\ Guana Gama Ae OO Oo N ies) = | ond [e-2) rie) ae) OD oD i N i~ Sc) Ge Gi Ce | Lie @ See Ge) ar 0 (©) @) ¢ or g LT Oo “panuruor—J eT, oe 009 © 6 @ 708) 2 e@ ©8 +e o WEYC) subiu “b40s + pan “yasy "Dyna “bog OHTA, “bina *bwoy “7272 728) DEE ENS "eo "]PR 798! BS LO TEKS) **7ppn Jag) ee, 66 2) eel (5 02 (79 ‘8b Use “GOGL [ Mar. 6, = = : = - | -— a = ae, re MB so ike Le Ol ue St its Ree) YL 18 | an) = Eh Oe SOAS EGE ier wi SO aeceOnmeceml: aba: | 5G ieee ee octee 6 2 eS =) So ak lesa eee . FH ea epee TE OI 6.cL se & Nt e AS |p 26, ell GL ae S| mae ELI 8 ©? GE a1 ee Se eG) in). SI = 8.8 ee ST ORT ee ‘3 Ei aot Gr ce. @ a Sergi ce op Poo A) SOT Wn 38 ge ce Seale OIE 28 Us = el et Si | Cr t Ae lee | Ose Cone tay pecter « We bee le Ope ica ee ee a a (ey “OL 0 lpia. Oh Gr tel de AGryy us to eT On OL | el Gas i of S G 4 & es . 0 9 8 0 oe "16 een? ° jae Oe OP 6 ai Oe er Gr Sls ee SI OL | SST |*° ‘2972 789) PL oO | mie TU) "Ul "2061 A oe TINS “AW "WOYBYS ‘OulTy | | SE esc aul 370 “panwuyuoor—J 9G T, 1903.] The Statolith-theory of Geotropism. o71 for long periods, .c., 40-—82 minutes ; these are, perhaps, not strictly comparable with the other experiments, though in three out of the four the shaken specimens were clearly more curved than the s//// ones. If these are omitted, we have :— Shaken, curvature greater ......... 15 Pee MICOMANS SHOU aeh so ce soe 2 Siipacumyature greater......5.... 2 The amount of curvature is not merely greater in a large majority of cases, but also differs by w considerable percentage. Thus, taking the sum of the average curvatures in all thirty cases, Werte. Still. Shaken. 403°°7 ? 600°°8 or 100: 148°8 Omitting the seven second readings (which give the sums—siil/ 62:1, shaken 108°2), we get :— Still. Shaken. 341°°6 AOD Cuero lOO) 144-2 Finally, omitting the four cases of long exposure (which give—sévll 71:1, shaken 103-7), we get :— Still. Shaken. 270°°5 . 300-0 non lOO = F438 In another series of experiments the difference between the shaken and still plants was much smaller. But we have reason to believe that the failure depended on the small amplitude of vibration employed, for when, in the last four experiments of the series, a more ample vibration was adopted, there was once more a well-marked increase in geotropism in the shaken specimens. In none of our experiments have we seen such striking results as those obtained by Haberlandt ; it must, however, be remembered that his apparatus differs from, and is apparently more effective than, ours. Control Kaperiments (Heliotropism). The experiments were made in the same way as the last, except that the seedlings were vertical instead of horizontal, so that the starch would be made to vibrate on the basal instead of on the lateral walls of the cells. The boxes had glass lids to admit the light, which was given by incandescent gas. Care was taken that the distance from the light of the still specimens was the same as that of the shaken specimens. ae u ost a Sa nee OED CN Ge eR OE, eS 3 = z. = cen. Mate » = i | 692 6-626 joo" TTL GOL ‘col ‘sdxy ar Surpvoa ys] SarqqrUto “yezO], Tint { Ve el NL Se Bs { LT B80 &% => Ph oe 8 Oc 6-02 ol 2 8 BL eg ¢ ra SE | LoS |S Pee UT | °"* OL oe 1 OL 9808 LI 2 { cg @h 86 8P = E se 99 84 82 G. LS ch ZL OL 06 6I ¢ 12 G6 | 966 | = 2 poss Gee aoe { PL 8 ST (g) a { (¢) 6L IL 22 | ‘ 0 OLse ¢ Oat 1 2 ee Gwe SI ¢ cad DS S| EES ED SO P73 2. Fe { 08 8% GZ 8é es | 96 O€ ST 02 6 (8)9% 2 §- 8Z es Gl G7 si Ov 17 1% Oe eg 22 22 IAA EIS [Pf = HOO) a 3-8 6 tm OLOr © OL || C6 wl 2 a F ec ‘s SSS 5 I Oa rg z 6 11 Bik GO FAL OV 3-9 OL O SI () sg ¢ = Z-8 Bec 4 8 8. ¢ Li SS SiO 8G 6 re 09 | SLT | °° ° Apna biog | +" 3 ‘qa a g. AT SNe Ol €. 2B - CL 08 “G2 ty rae GQ] | Len | bier boys eae 9. LE = 8 0¢ =cy &: OF "* 22 0g 19 Or ¢ g SG — |Ogn | 9 > 7072 pos | oT ue a ‘OST = 3. SP Sh 08 oh OL 0. GE 82 08 Z& 8 oF 3-11 OL © 08 § 8. ¢ OO Se -3 OO) 30 ec =cee 0S = ele eee Sines bon,| Seaeien G- OL © Oe Be -G 8. 91 CL 08 (¢) 32 of F | 0. ¢ OS 40 0) 0-81 Ol “28 Bi QO ©. | eek 06 = 60m =| SP)Mo Dog | yes 0-11 92 0 F PFI 8-8 (9) ZL Gi-¢ eG <4 ¢. BT ST OL 0% ¢ 0: OT a © Sie Ol 0° = 20 Gel OL | BOL |°°* “byna “biog |\+* HT -00q ut “Y “TO6T “AV Hus “AV “WayeY[Y ‘OUP aU TWO | “XO UU ‘wistdorjorapy{y—"T] eTq RT, NI J|~ 1903. ] The Statolith-theory of Geotropism. O19 In three of the experiments (102, 109, 111) readings were taken before the plants were plaved on the klinostat, and these readings are omitted in summing up the results :— Curvature of shaken plants greater... 7 cases a still hs ANON » Seat: | ISMC MMC OUIAL SOLU, Nes os oe age aueee es 1 Thus there is a majority of wins in favour of the shaken plants, though it is not nearly so striking as in the geotropic experiments. When the amount of curvature is taken into consideration, the result is decisive :— Still, Shaken. Sum of averages 259-9 273°9 or as 100 105-4: The conclusion at which I arrive is that vibration does not materially increase heliotropic curvature, whereas it does increase geotropism. The following figures show in round numbers the differ- ence between the geotropic and heliotropic experiments :— Geotropic Curvature. Still. Shaken. 100 144 Heliotropic Curvature. Stull. Shaken. 100 105 The inference I draw from this result is the same as that of Haberlandt, viz., that the increased geotropism of the shaken specimens is due to the increased stimulus produced by the vibration of the starch grains on the protoplasm of the lateral walls of the cells. I desire to express my thanks to Miss Pertz, and to my assistant Mr. Elborn, for much valuable help. o74 Mr. W. BR. Carr. On the Laws governing [Feb. 11, “On the Laws governing Electric Discharges in Gases at Low Pressures.” By W. R. Carr, B.A., University of Toronto. Communicated by Professor J. J. THomson, F.R.S. Received February 11,—Read March 5, 1903. (Abstract. ) The experiments described in this paper were undertaken with the object of determining the potential difference required to produce cischarge in a number of gases over a wide range of pressures, and especially of ascertaining if the law enunciated by Paschen* was generally applicable, provided the electric field in which the discharge took place was uniform. The paper is divided into the following sections :— (1.) Introduction. (2.) Description of apparatus. (3.) Experiments in air. (4.) Experiments in hydrogen. (5.) Experiments in carbon dioxide. (6.) Spark potentials with different electrodes. (7.) Minimum spark potentials. (8.) Connection between spark lengths and spark potentials. (9.) Minimum spark potentials in different gases. (10.) Summary of results. Paschen’s experiments showed that when a given potential difference was applied to two spherical electrodes whose distance apart could be varied, the maximum pressure at which discharge occurred in a gas » varied inversely with the distance between the electrodes. ‘The range of pressures covered by his experiments did not extend below 2 cm. of mercury. While Paschen has shown that as the pressure of a gas diminishes the difference of potential necessary to produce discharge between electrodes in a gas, a fixed distance apart, also diminishes, Peacet has shown that a critical pressure is finally reached when the spark potential reaches a minimum value, and that below this critical pressure the potential difference required to produce discharge rapidly increases as the pressure is lowered. Peace’s experiments were conducted with air, and his electrodes consisted of a pair of large parallel plates supported in the gas. The values of the spark potentials recorded by him led to the conclusion that Paschen’s law did not hold for electric discharges at and below the critical pressure. In this paper it is shown that with the apparatus used by Peace, the * Paschen, ‘ Wied. Ann.,’ vol. 37, 1889, p. 79. + ence, “Roy. Socsleroc. ci voluo2.up soo: 1903. | Hlectrve Discharges in Gases at Low Pressures. O19 discharges at low pressures in all probability did not take place along the shortest path between the plates, and it is inferred that the failure of his numbers to establish the applicability of Paschen’s law at all pressures is due to his having always taken this shortest distance between the electrodes as a measure of the spark length. In the present paper an account is given of an investigation on the potential difference necessary to produce discharges in a gas with a form of apparatus which ensured the passage of the discharge in a uniform electric field at all pressures. With this apparatus the spark potentials were determined in air, hydrogen, and carbon dioxide, for different distances between the electrodes, over a range extending considerably above and below the critical pressures. Hlectrodes of brass, iron, zinc, and aluminium, of the same size, were in turn used in the apparatus, but the readings obtained showed that the spark potentials were not influenced at any pressure by the size of the electrodes, provided the discharge took place in a uniform field. The result of the investigation not only confirmed the truth of the law enunciated by Paschen for discharges in a gas at high pressures, but also demonstrated the applicability of the same law to the critical and to lower pressures. This law is summarised in the statement “that with a given applied potential difference, electric discharge in a uniform field in any gas is dependent solely on the constancy of the quantity of matter per unit cross-section between the electrodes.” It is shown that a general application of Paschen’s law demands that the minimum spark potential must be a physical constant for each gas. A special set of observations gave the following values of this quantity for a number of simple and compound gases. Minimum spark Gas. potentials in volts. ek cee y. eis 278 O, e aeeN oe HoN couse take Oot, o iedekolayn: sifeaie 455 GS eee thins tc cee aes ini. 414 CON ee a vies ata 419 IND Ore ots tor tociet sgt: 420 DOD eae. fac. goseh ac 457 Oaleley yea siee nay eca a qe 467 Adopting Strutt’s value of 251 volts for nitrogen, the conclusion is drawn that the minimum spark potential is a property of the atom rather than the molecule of a gas, and it is shown that if H’, N’, O,, etc., represent the spark potential constants in volts, corresponding to atoms of the gases H»2, No, Oz, etc., respectively, the minimum spark potential for any compound gas whose formula is H,N,0., etc., will be given by zH’+yN’+2O' + etc. volts. It is pointed out that oxygen forms an exception to the general application of this law. 376 Prof. A. Gamgee and My, A. Croft Hill. ern, oil, The latter part of the paper deals with the extension of Paschen’s law to spark lengths much shorter than those actually used in the experiments, and evidence is adduced in support of the conclusion that the law is applicable for discharges in a uniform field in any gas, as long as the spark length is greater than the diameter of the sphere of molecular action. “On the Optical Activity of Hemoglobin and Globin.” By ARTHUR GAMGEE, M.D., F.RS., Emeritus Professor of Physiology in the Owens College, Victoria University, and A. Crort Hitt, M.A., M.B., late George Henry Lewes Student: in Physiology. Received January 31,—Read February 12, 1903. Introductory Observations. All observations hitherto published concerning the optical activity of the albuminous substances have led to the conclusion that the bodies thus designated, whether derived from the vegetable or the animal kingdoms, without a single exception, deviate the plane of polarisation to the left, no case having hitherto been known either of a dextrogyrous, a racemic, or an otherwise inactive albuminous sub- stance.* There is one group of albuminous substances which, notwithstand- ing the fact that it includes bodies of paramount physiological and chemical interest, has hitherto been completely neglected, in so far as the investigation of the optical activity of its members is concerned. The group to which we refer is that which has been designated by German writers the group of the “ Proteide.” This group comprises. those complex albuminous substances which can, with greater or less ease, be split up into, or which yield as products of decomposition, on the one hand, albuminous bodies, and on the other, such bodies as colouring matters, or nucleins and nucleinic acids and the purin- bases which result from the decomposition of the latter. The best * Whilst this paper was being printed, it has come to our knowledge that the late Professor Alexander Schmidt, of Dorpat, described under the name of Cyto- elobin, what was certainly a mixture of impure nucleoproteids which he separated from the soluble constituents of many animal cells. He definitely recognised the dextrorotatory properties of this product. For information on A. Schmidt’s work, the reader is referred to the “Supplementary Bibliographical Note”’ at the end of the paper by Gamgee and W. Jones “On the Nucleoproteids of the Pancreas’ Thymus, and Suprarenal Gland, with especial reference to their Optical Activity.” Infra, p. 385.—March 5. 1903.] On the Optical Actirty of Haemoglobin and Globin. — 377 characterised and the most striking members of this group are: Firstly, the hemoglobins and their compounds. Secondly, the nucleo- proteids and the nucleins. In hemoglobin, we have the example of a complex proteid, which differs from all other members of the albuminous group of bodies by its colour, by its marvellous power of forming easily dissociable com- pounds with oxygen and certain other gases, by the facility with which it admits of being crystallised and recrystallised and obtained free from all foreign mineral matters, by the startling manner in which its solutions fail to furnish any one of the reactions characteristic of albuminous substances in solution, so long as the reagent has not effected a fundamental decomposition which has liberated the albuminous and coloured residues. ‘The researches of one of us have moreover lately shown that whilst hemoglobin is a diamagnetic body, the iron-con- taining products of its decomposition by acids are not merely para- magnetic, but probably the most powerfully “ferromagnetic” organic bodies known to science.* So complete a divergence thus exists in physical and chemical pro- perties between hemoglobin and the substances which are the imme- diate products of its decomposition, and which are doubtless linked together in it, that it appeared in the highest degree interesting to ascertain whether or not, in respect to optical activity, hemoglobin would behave as an albuminous body proper and prove to be ‘“levo- gyrous.” Having, if possible, determined this point, the subsequent step in the research would naturally be to determine the optical activity of the albuminous and coloured products of the decomposi- tion of the hemoglobin molecule. 1.—Determination of the Optical Activity of Hemoglobin. So far as the authors have been able to ascertain, the optical activity of solutions of coloured organic bodies has not yet formed the object of serious investigation. Landolt,t in the last edition of his authoritative work, which contains all the reliable results relating to the optical activity of organic bodies up to the date of its publication, mentions only one colouring matter as having been investigated, viz., the vegetable colouring matter hematoxylin, of which the alcoholic solution is said to be dextrogyrous. Nor is this neglect of the study * A. Gamegee, “On the Behaviour of Oxy-hemoglobin, Carbonic-oxide-hxemo- globin, Methzemoglobin, and certain of their Derivatives in the Magnetic Field, with a Preliminary Note on the Electrolysis of the Hemoglobin Compounds,” * Roy. Soc. Proc.,’ vol. 68, p. 5038. A. Gamgee, The Croonian Lecture for 1902, “‘ On certain Chemical and Physi- cal Properties of Hemoglobin,” ‘ Roy. Soc. Proc., vol. 70, p. 79. + Dr. H. Landolt, ‘Das Optische Drehungsvermégen Organischer Substanzen, &e., Zweite ginzlich umgearbeitete Auflage,’ Vieweg u. Sohn, 1898. O78 Prof. A. Gamgee and Mr. A. Croft Hill. [Jan. 31, of the optical activity of coloured solutions surprising when we con- sider the much greater difficulties which encounter the observer, in comparison to those attending the examination of colourless solutions. The Method Employed in the Present Research. When a powerful beam of white light is passed through a stratum 1 cm. thick of a solution of Oxy- or CO- hemoglobin containing 0-9 per cent., the only region of the spectrum which is unabsorbed is that which extends from B to a little distance on the red side of D. It was therefore clear that the only light which could be employed in the work before us was monochromatic red light, and that in the place of one of the polarimeters commonly employed in our laboratories and whose adjustments only permit of their being employed with lght of a definite wave-length (the half-shadow polarimeters of the Laurent type being adjusted for use with monochromatic sodium light), an instrument should be employed, the arrangements of which permit of observations with light of any desired wave-length. In our first observations, we attempted to employ the lithium flame as our source of light, but we were unable to secure by this means either a sufficiently powerful or steady illumination. We subse- quently employed, however, as a source of practically monochromatic red light, the light of an arc lamp which had traversed Landolt’s filter for red rays. This light filter consists of a double cell, each compartment of which has a depth of 20mm. One compartment is filled with a solution of hexamethylpararosanilin, a substance sold commercially under the name of “Crystal Violet 5 BO.” 0°05 gramme of this compound is dissolved in a small quantity of alcohol and the solution is then diluted with water to the volume of one litre. When light is made to traverse a stratum 20 mm. thick of this solution, its spectrum consists of a narrow red band and a broad blue-violet part. If, however, the second compartment of the double trough contains a solution made by dissolving 10 grammes of potassium chromate in 100 c.c. of distilled water, the blue-violet is entirely absorbed and the spectrum of the light which has traversed the two compartments of the light-filter consists of a narrow strip, extending from A718 to » 6394p, where it ends abruptly. The mean wave-length (“ optischer Schwerpunkt ”) corresponds to 665yup, the wave length of C being 656°3uy.* By means of the above method we secured a beam of red light having a mean wave-length approximately the same as that of C and of sufficient intensity to allow us to make observations on solutions of hemoglobin containing + 1 gramme in 100 c.c. of distilled water, the tubes employed in different sets of observations being 100 mm. and 200 mm. in length. * Landolt, op. cit., pp. 387-390. 1903.] On the Optical Activity of Hemoglobin and Globin. 319 The polarimeter employed in these observations was a magnificent Lippich’s ‘“ Halbschatten-Polarimeter,” with tripartite field of vision, made by Schmidt and Haensch, of Berlin, and belonging to the Davy- Faraday Laboratory of the Royal Institution of Great Britain. The Hemoglobin Employed.* The solutions of hemoglobin employed for the determinations of which the results will be given below, were prepared with oxy-hemo- globin of remarkable purity which had been obtained from the blood of the horse by following the best of the methods (the third method) described by Zinoffsky.7 | Two preparations of hemoglobin made on a large scale and at the interval of some months one of the other were employed. The preparation employed to make the solution of Oxy-hemoglobin had been crystallised three times, the product of each successive crystalli- sation having been many times washed with ice-cold distilled water of which the purity was controlled by determining its electrical resistance. ‘This solution contained 2°446 grammes of hemoglobin in 100 ¢.c. For polarimetric observations this solution was diluted with an equal volume of distilled water, the dilute solution examined con- taining, therefore, 1:223 gramme of oxy-hemoglobin in 100 c.c. The preparation employed to make the solution of CO-hzemoglobin had been crystallised four times. The crystals of each successive crystallisation had been subjected to washing with pure distilled water as stated above, the solution of the washed crystals of the fourth crystallisation having been saturated with CO. This solution contained 1°84 grammes of dry CO-hemoglobin. For polarimetric measurement this solution was diluted with an equal volume of distilied water ; the dilute solution contained, therefore, 0°92 gramme of CO-hemoglobin in 100 c.c. Haemoglobin, whether in Combination with Oxygen or Carbonic Oxide, is Dextrovotatory. A. Oxy-Hemoglobin. The diiuted solution of Oxy-hemoglobin, previously referred to, was employed. This solution, containing 1:223 gramme of hemoglobin in 100 c.c., was thoroughly saturated with oxygen before, being subjected to polarimetric observation. * The part of Section 1 of this paper which follows has been recast, and the oservations described under A and C added, since this paper was submitted to the Royal Society.—March 5. + Zinoffsky, O., “‘Ueber die Grisse des Hemoglobinmoleciils,” ‘ Zeitschrift ¢. physiol. Chemie,’ vol. 16 (1886), p. 23. aces UX XT. 25 380 Prof. A. Gamgee and Mr. A. Croft Hill. [Jan. 31, The tube employed in all the sets of observations measured 1 decimetre. Three sets of observations were made. Observed Specific angle. rotation (@)c. 1. Mean of first set of observations... +0°:12 +9°°8 2. - second set of observations +0°:125 +10°:2 3s :, third . . +0°°1225 +10°:0 From the above observations we conclude that the specific rotation of Oxy-hemoglobin for light of the mean wave length of C @)o— 100 0-2: b) B. CO-Hemoglobin. The diluted solution of CO-hemoglobin, previously referred to, was employed. This solution contained 0°92 gramme of CO-hemoglobin in 100 e.e. Two sets of observations were made with this solution; in the first set a tube 1 decimetre long, and in the second a tube 2 decimetres long being employed. | Length Observed Specific of tube. angle. rotation (a)c. 1. Mean of first set of observations 1decim. +0°098 +4+10°:65 Te 4 second ,, a Deianyer +0°°203 +117:03 Taking the mean of the two series of observations we obtain as the specific rotation of a solution containing 0°92 gramme of CO-hemo- globin in 100 c.c. (a)o= +1078. When the feeble rotatory power of hemoglobin is considered, the agreement between the results of the investigation of the rotatory power of Oxy- and CO-hemoglobin must be considered satisfactory and as pointing to the conclusion that the molecule of oxygen or carbonic oxide in combination with hzmoglobin does not influence its specific rotation. The correctness of this conclusion has been established by direct experiment. C. The same Solution of Hemoglobin saturated with O and with CO Compared. With the object of determining by direct experiment whether the dissociable combinations formed by O and by CO with hemoglobin had any influence on its specific rotation, the solution of Oxy-hemoglobin which served for three sets of observations recorded under A, and which contained 1:223 gramme in 100 c.c. of water, was again experi- mented with. One portion of this solution was saturated with 1903.] On the Optical Activity of Haemoglobin and Globin. 381 oxygen; another portion was agitated with pure CO so as completely to expel the oxygen from its combination with hemoglobin and replace it by carbonic oxide. In this way were obtained two solutions of hemoglobin identical in so far as the quantity of colouring matter which they contained, but differing in the fact that in the one case the hemoglobin was in combination with oxygen and in the other with CO. The solutions were examined in tubes of the same length under the same conditions of illumination. The result was to show that the rotations were identical in the two cases, having the mean value represented by the specific rotation («),= +10°:0. It is to be remarked that the observations recorded under A and C were carried out subsequent to those on CO-hemoglobin recorded under B. In the case, particularly, of observations A, the intensity and steadiness of the monochromatic red light employed was, in con- sequence of the experience previously acquired, more satisfactory than in observations B. We are therefore inclined to consider the numbers expressing the specific rotation of hemoglobin which we have obtained as the result of observations A to be most worthy .- of confidence. We do not pretend that these numbers may not need slight modification as the result of future work, though we believe that they are a very close approximation to the truth. 2.—Deternunation of the Optical Activity of Globin. Preyer gave the name of Globin to the albuminous product of the spontaneous decomposition of hemoglobin, without, however, being able to furnish any precise account of its properties, its chemical com- position, or its relationship to other albuminous bodies. A compara- tively recent investigation which we owe to Fr. N. Schulz,* and the results of which have been substantially confirmed by Ivar Bang,t has placed us in possession of valuable and suggestive facts concerning the main albuminous product resulting from the decomposition of hemoglobin. He has shown that when a solution of hemoglobin is decomposed by the addition of small quantities of hydrochloric acid, it yields, as main products, 4:2 per cent. of hematin and 86°5 per cent. of a characteristic albuminous substance for which he retains the name of globin. He has shown that this substance belongs to the class of “the Histons,” so that it would have been preferable, in our opinion, if Schulz had applied to his new body such a name as “ Hemato-Histon,” which would have indicated both its origin and its affinities. Schulz’s method of preparing globin, as described by him, is essen- * Schulz, Dr. Fr. N., “Die Eiweisskérper des Hemoglobins,” ‘ Zeitschr. f, physiol. Chemie, vol. 24 (1898), p. 449. 7 Bang, Ivar, “Studien tiber Histon,” ‘Zeitschr. f. physiol. Chem.,’ 1899, p. 463. 2H 2 382 : Prof. A. Gamgee and Mr. A. Croft Hill. [Jan. 31, tially as follows: to a solution of crystallised hemoglobin, either pre- pared by Hoppe-Seyler’s method or by the ammonium sulphate method, dilute hydrochloric acid is added in extremely small quantities, until a flocculent brown precipitate falls which is immediately dissolved by the slightest excess of acid. The solution then no longer exhibits the beautiful red colour of hemoglobin, but has assumed a brown colour. Not only, remarks Schulz, has its colour changed, but a complete separation has occurred between the albuminous and coloured con- stituents of hemoglobin. If to the solution, which has now a faint acid reaction, about one-fifth of its volume of 80 per cent. alcohol be added and the mixture be shaken with ether, the whole [sic] of the colouring matter is taken up by the ether, whilst the subjacent aqueous-alcoholic, perfectly clear solution contains the decolourised albuminous matters. Schulz gives particular directions as to the precautions which must be taken in order that the separation of the ethereal solution of the colouring matter should be complete, stating that a certain relation must exist between the proportions of water, alcohol, and ether, which must be experimentally determined in each case. By the above process there is obtained a more or less brownish- yellow solution, containing both alcohol and water and having a faintly acid reaction. On neutralising this solution with ammonia, a faintly yellow, coarsely flocculent precipitate falls. The latter is rapidly separated by filtration and then washed with water. When the excess of ammonia has been removed, the precipitate commences to dissolve in the wash water. At this stage, the precipitate is dis- solved in water with the aid of a few drops of dilute acetic acid. Solution occurs rapidly and completely. The excess of acid is now removed by dialysis continued for some days, the dialyser being sur- rounded by distilled water. ‘There is thus obtained a clear, odourless and tasteless solution of globin the reaction of which is perfectly neutral. : It is not our object to examine in this place the reactions presented by solutions of globin, and which have led Schulz to place it among the ‘“ Histons.” Before describing briefly the methods we employed to prepare the solutions of globin which we investigated optically, we desire to make certain observations on certain points in Schulz’s statement. In dis- cussing the quantity of dilute hydrochloric acid needed to effect the decomposition of hemoglobin, he merely remarks that it is extra- ordinarily small (‘“‘Die zu der Spaltung erforderliche Menge von Saure ist ausserordentlich gering, &c.”). We have determined the quantity of decinormal hydrochloric acid required to effect the decom- position of a solution of CO-hemoglobin of known composition. As a result of very careful experiments with a solution containing 1:84 grammes dissolved in 200 c.c. of water, there were required 1905.] Onthe Optical Activity of Haemoglobin and Globin. 383 20 ¢.c. of decinormal hydrochloric acid to effect the complete separa- tion of globin from the colouring matter. We found that agitation with ether, unless repeated several times, fails to remove all the colouring matter which is capable of removal in this way. Further, we found that even when, as a result of agita- tion with ether, the aqueous-alcoholic solution of globin is of the faintest straw colour, on being neutralised with ammonia the preci- pitated globin, which is at first colourless, assumes a somewhat reddish tinge, and when subsequently dissolved in water faintly acidulated with acetic acid the solution is much more deeply coloured than the original aqueous-alcoholic solution. The following is the precise method which we followed in preparing the solutions employed in our polarimetric determinations :— 100 ¢.c. of a solution of four times crystallised hemoglobin, con- taining 1°84 grammes of the substance, was diluted with 100 c.c. of distilled water and treated with 20 ¢.c. of decinormal hydrochloric acid. 44 ¢.c. of absolute alcohol were then added to the liquid, which was placed in a stoppered separating funnel and thoroughly agitated with its own volume of ether. The aqueous-alcoholic liquid having been separated from the supernatant ethereal solution of colouring matter was twice more agitated with fresh quantities of ether. By proceeding as we have described, the separation of the solution of globin occurred completely after the first agitation with ether, and the solution after the third agitation only possessed a faint straw coloura- tion. In certain cases, the globin was separated according to the method of Schulz by precipitation with ammonia, the flocculent pre- cipitate being subsequently dissolved in very weak acetic acid. In this manner was prepared the solution of globin which served for the first set of determinations recorded below. As it was impossible to obtain in this way solutions sufficiently colourless to allow of their rotation to be determined satisfactorily for light of the wave-length of D, this was done as in the case of hemoglobin for light of the mean wave-length of C. In the second set of observations, the rotation of the aqueous-alcoholic solution resulting from the decomposition of hemoglobin, after thorough agitation with ether, was determined. Globin a Levorotatory Substance. Preliminary observations having shown that solutions of globin are optically active and levogyrous, the following sets of observations were made with the object of determining the specific rotation of solutions of the substance. 1. A solution of globin in distilled water, but containing a little acetic acid, was examined with the arrangement for red light, as was used in the case of hemoglobin. The solution contained 2-4 grammes 384 On the Optical Activity of Hemoglobin and Globin. [Jan. 31, of globin in 100 ¢.c. It exhibited in the most characteristic manner the reactions of globin. The tube employed measured 1 decimetre. The angle of rotation (mean of many determinations) was — 1°:30. From the above data, it follows that in the case of this feebly acid solution of globin, containing 2°4 per cent., the specific rotation [a|o= —54°°2. 2. The faintly straw-coloured solution obtained by the decomposition of hemoglobin by means of dilute hydrochloric acid, the addition of alcohol and repeated agitation with ether, was placed in a shallow capsule in a current of air for some hours and afterwards on the water-bath at the temperature of 40°C. In this way all the ether and some of the alcohol were expelled. The perfectly clear straw- coloured solution, which had a density of 987-4 at 16°C., contained 0-98 gramme of solid matter in 100 c.c. Monochromatic sodium light was employed in the polarimetric observations. The tube employed measured 1 decimetre. The angle of rotation (mean of many determinations) was — 0°-64. From the above data, it follows that in the case of this feebly acid, aqueous-alcoholic solution of globin, containing 0°98 per cent. of solids the specific rotation, [«]p= —65°°5. It may be pointed out that the greater part of the discrepancy between the results of the polarimetric measurements of the solution of separated globin and of the solution now under discussion is to be explained by the difference in the wave- length of the light, of which the rotation of the plane of polarisation was determined in the two cases. General Conclusions. The following are the conclusions to which we have been led by the experiments described in this paper :— 1. Hemoglobin is a dextrogyrous albuminous body. 2. Globin, which is the principal, or as we are inclined to believe, the only albuminous product of the decomposition of hemoglobin by highly dilute hydrochloric acid under the conditions determined by Schulz and confirmed by our own observations, behaves as a normal albuminous substance, in respect to its influence on the plane of polarisation of light, i.c., it is a levogyrous body. Whilst the conclusions above stated are beyond question correct, we wish it to be understood that the numbers expressing the specific rotation of the bodies which we have examined must be looked upon as very close approximations and may need revision in the case of hemoglobin by determinations carried out with a more perfectly mono- chromatic and intense light than that which we have employed, and in the case of globin by working with the substance in a purer condi- 1903. On the Nucleoproterds of the Pancreas, ete. 389 tion than is possible in the actual state of our knowledge of this body. We hope to be able to carry out these further investigations, and to direct our attention to the optical activity of the coloured products of the decomposition of the hemoglobin molecule, especially hemochromogen and hematin and their coloured derivatives. In conclusion, we have to express our thanks to the Managers of the Davy-Faraday Laboratory of the Royal Institution for the facilities which they afforded us in carrying on the optical part of our work. «On tne Nucleoproteids of the Pancreas, Thymus, and Suprarenal Gland, with especial Reference to their Optical Activity.” By ArTHur GAMGEE, M.D., F.RS., Emeritus Professor of. Physiology in the Owens College, Victoria University, and WALTER JONES, Ph.D., Associate Professor of Physiological Chemistry in the Johns Hopkins University. Received February 9,—Read February 12, 1905. Part J.—BIBLIOGRAPHICAL AND CRITICAL. Ina research in which one of us was associated with Dr. A. Croft Hill, it was discovered that Hemoglobin is a dextrorotatory body, whilst the interesting Histon-like albuminous substance Globin, which is obtained by the splitting up of Hemoglobin under the influence of highly diluted hydrochloric acid, and of which the characters, no less than the mode of preparation, have only been known since the researches of Fr. N. Schulz, is a normally levogyrous albuminous body. These interesting observations naturally suggested the probability that the Nucleoproteids might, like Hemoglobin, prove to be dextro- gyrous, and the research of which the first results are contained in this paper is the outcome of this idea. The hypothesis has been fully confirmed, as will be shown in the sequel, and it has thus been proved that some of the members of a group of albuminous bodies of great importance in the life-history of the organism, are dextrorotatory bodies. | The preparation of nucleoproteids of such purity and especially so free from contaminating colouring matters as to yield solutions sufficiently transparent and colourless for polarimetric work, was a necessary preliminary to our special researches, and has led to the discovery of many facts of interest in relation to the chemistry of the nucleoproteids. 386 Profs. A. Gamgee and W. Jones. [ Feb. 9, Preliminary Remarks concerning the “ Nucleoproteids” and ‘* Nucleins” and the Sense in which the Latter Term is used in the Present} Paper. By the term Nucleoproteids, we designate complex, or rather compound, albuminous substances which are the constituents of the nucleated protoplasm of all the organs of the animal body, but especially of the ductless, as well as of the secreting, glands. These bodies are characterised by the large quantity of phosphorus which they contain, by the constant presence of iron, and by the fact that under - the influence cf heat, by the action of acids, of alkalies, but especially of pepsin and hydrochloric acid, acting at temperatures favourable to their action, they split up into albuminous matters, and into so-called true (to distinguish them from pseudo-) nucleins. The latter differ from the mother nucleoproteids which yielded them, by the fact that they result from the splitting-off of a fraction of the albuminous molecules which these contained in their pristine and native con- dition. These secondary, or we may say, degraded nucleoproteids, “the nucleins,” contain all the phosphorus originally present in the mother substance. . By the action of caustic alkalies and heat, the nucleins yield as products of decomposition, albuminous matters, and the so-called “nucleinic acids,” bodies which vary in composition in the different nucleoproteids, but which are characterised by the fact that when heated with certain mineral acids, they yield as products of hydro- lysis (Kossel), one or more of the purin-derivatives long known as ‘the xanthine bases,” Adenin (Amidopurin), Guanin (Aminooxypurin), Hypoxanthin (Oxypurin), and Xanthin (Dioxypurin), as well as in many cases a base called Thymin, C;H yN2O2, a derivative of Pyrimi- dine.* At the same time, the phosphorus is separated as phosphoric acid. Kossel, to whose fine researches we owe the greater part of our knowledge of the nucleinic acids, advanced the hypothesis (based on the great variation in the quantities of the xanthine bases which result from the hydrolysis of nucleinic acids of different origins) that there are four nucleinic acids, each of which yields one of the bases only. This theory of Kossel appeared to gain important support from Ivar Bang’st discovery of guanylic acid, a nucleinic acid obtained by the action of solution of potassium hydrate on the nucleoproteids of the pancreas, and which, as its name indicates, yields on hydrolysis one only of the purin-bases, viz., guanine. This hypothesis does not * Walter Jones, ‘ Zeitschrift f. physiol. Chem.,’ vol. 29 (1900), p. 26; H. Steudel u. A. Kossel, ‘ Zeit. f. physiol. Chem.,’ vol. 29 (1900), p. 303; H. Steudel, ‘ Zeitschr. f. physiol. Chem.,’ vol. 30 (1900), p. 539; vol. 39 (1901), p. 241. + Ivar Bang, “ Die Guanylsiure der Pancreasdriise und deren Spaltungspro- dukte,” ‘ Zeitschr. f. physiol. Chem.,’ vol. 36 (1898), p. 133. 1903.] On the Nucleoproterds of the Pancreas, ete. 387 appear to be in unison with the facts known to us (Schmiedeberg, Levene, W. Jones and G. H. Whipple, T. B. Osborne and I. F. Harris).* Hammarsten,t to whose researches on the nucleoproteids and their relations to the nucleins we owe much of our knowledge of these bodies, would restrict the term “ Nucleins” to the albuminous com- pounds of the nucleinic acids which remain undissolved after prolonged digestion with pepsin and hydrochloric acid. But this limitatiom appears to us undesirable and unphilosophical, and we think that the term nuclein, which it is convenient to retain, both for historical and descriptive reasons, should be applied to designate all those bodies resulting from the splitting-off of some, but only some, of the albuminous molecules originally forming part of the more complex nucleoproteid mother substance. It is in this sense that we shall in this paper employ the term nuclein, it being understood that every nuclein 1s to be considered a nucleoproteid, inasmuch as it is a compound of an albuminous body with a nucleinic acid or acids. The Researches of Hammarsten on the Nucleoproteids of the Pancreas. In his most interesting and suggestive paper published in the year 1894, Hammarsten gave an account of two nucleoproteids which he had obtained from the pancreas. The first of these bodies he designated proteid-z. He ascertained that this body which, being soluble in water, is present in cold aqueous extracts of the pancreas, is precipitated by acetic acid, and that when boiled its solutions yield a coagulated albuminous substance, the sub- stance remaining in solution being presumably nucleoproteid-f. Although Hammarsten fully recognised that the first or o-body was the mother substance, and that proteid-8 was only a product of its decomposition, he devoted his attention to the latter, being actuated by the following reasons :—In the first place, his object being at that time to study the non-albuminous products of the pancreatic nucleo- proteid, it appeared to him more simple and wiser to take as the starting-point of the investigation a material containing less albumin. The chief ground, however, for leaving the more interesting nucleo- proteid provisionally uninvestigated was the great difficulty of obtain- * O. Schmiedeberg, ‘ Archiv f. experiment. Path. u. Pharmak.,’ vol. 48 (1899), p.57; P. A. Levene, ‘ Zeitschr. f. physiol. Chem., vol. 32 (1901), p. 541; W. Jones and G. H. Whipple, ‘Amer. Jour. of Phys.,’ vol. 7 (1902), p. 4238. See particularly the recent paper by Thomas B. Osborne and Isaac F. Harris, “ Die Nucleinsiure des Weizenembryos,” ‘Zeitsch. f. physiol. Chem.,’ vol. 36, Heft 2 (September, 1902), p. 85. ft Olof Hammarsten, “ Zur Kenntniss der Nucleoproteide,” ‘ Zeitschr. f. physiol. Chem., vol. 19 (1894), p. 19. 388 Profs. A. Gamgee and W. Jones. [Feb. 9, ing it in any degree pure, attempts at purification being attended with such loss that the yield was too small. 7 Hammarsten remarked that among the impurities most difficult to. separate was the blood-colouring matter, as well as another colouring matter which he believed to be produced by the action of the air on the nucleoproteid itself. Further, another impurity adhering to the nucleoproteid was found by Hammarsten to be trypsin, which he was. unable to separate from it. He remarks, indeed, that the proteolytic activity of the substance is so intense that in no other way could he - obtain so powerfully acting a trypsin. Having, for the reasons above stated, abandoned the study of the interesting mother-substance, his nucleoproteid-z, Hammarsten then directed his attention to the 6-body. This body, he did not seek to obtain by the decomposition of the mother substance, of which it is a product, but by adopting the following method :—he boiled the finely comminuted and perfectly fresh pancreatic gland of the ox in water and obtained, after filtration, a perfectly clear, faintly yellow solution, to which he added, after cooling, from 1 to 2 parts of hydrochloric acid, or from 5 to 10 parts of acetic acid per 1000 parts of the liquid. In this manner he obtained an abundant, white, flocculent precipitate. He dissolved the substance thus precipitated in water, with the aid of the least possible quantity of alkali and reprecipitated it by adding an excess of acid. By repeating this process several times, the body originally precipitated was purified, so far as such a method can effect the purpose. It must be clearly insisted upon that, as Hammarsten himselt pointed out, the so-called nucleoproteid-B does not represent an original proximate principle of the pancreas, but is a nuclein pro- — duced from the original mother nucleoproteid (or nucleoproteids ?) by the action of boiling water. It is certainly in no spirit of detraction or want of respect for the eminent Swedish chemist, that we add the remark that the study of a nuclein to be satisfactory should, if possible, take as its starting point the pure mother substance, of which it is a product of decomposition, rather than the animal tissue which contains that substance. In the case of Hammarsten’s nucleoproteid-6, one can at present only assert that it is a nuclein or a mixture of nucleins produced by the action of boiling water on the nucleoproteids, properly so called, existing preformed in the tissue of the pancreas. These strictures notwithstanding, we have to point out the remark ably interesting facts which were discovered by Hammarsten in the course of the investigation under review. He made a series of ulti- mate organic analyses of different specimens of this nuclein, and showed that whilst its solutions when boiled with Fehling’s solution gave no trace of reduction, the body when heated on the water-bath with dilute sulphuric acid, furnished a highly reducing substance. £903. ] On the Nucleoproteids of the Pancreas, cte. 389 Although unable to separate the reducing substance in a state of purity, he succeeded in preparing an Osazone of constant melting point and the characters of which agree with those of the osazone of a pentose, an observation which absolutely coincides with the researches of Kossel and Bang, which establish the presence of a carbohydrate nucleus in the nucleinic acids and the formation of pentoses when they are subjected to the hydrolytic action of dilute mineral acids and heat. Further, Hammarsten showed that when his nuclein was decomposed by heating with a: 3 per cent. solution of sulphuric acid on the water bath, a erystalline sediment often separated which, after being purified, was analysed and shown to consist of guanine sulphate. Later, at Hammarsten’s instigation, Ivar Bang, continuing the investigation, prepared from Hammarsten’s nuclein, the nucleinic acid to which he ascribed the name of Guanylic Acid. PART 2.—HXPERIMENTAL. On the Nucleoproteid of the Pancreas and on Certain Characters of the Nucleens which are associated with, or derived from, tt. A. The Nucleoprotecd. Method of Preparation. The finely divided pancreas of the pig was treated successively with 50 per cent. alcohol, 75 per cent. alcohol, and 95 per cent. alcohol, and finally subjected to the action of absolute alcohol and ether, with the object of dehydrating it. The material thus obtained was extracted with successive portions of a 5 per cent. solution of ammonium acetate, the united extracts were filtered, and the perfectly clear fluid was poured into four times its volume of weak alcohol. The precipitate thus formed was washed by decantation with a large amount of dilute alcohol, and finally dried with absolute alcohol and ether. The object of this series of procedures was to remove the colouring matter of the gland, which is somewhat soluble in dilute alcohol, more so in an alcohouc solution of ammonium acetate, but soluble to a very slight extent in an aqueous solution of ammonium acetate. These manipula- tions also remove a large amount of inorganic salts, and render the coagulable albuminous substances insoluble. A 2 per cent. aqueous solution of this raw material had only a pale yellow colour, and it was found that it could easily be examined in a tube measuring 220 mm. with the polarimeter, monochromatic sodium light being employed. The polarimeter was a ‘“ Halbschatten-Polari- meter” made by Schmidt and Haensch of Berlin. The result of the examination was to show that the solution contained a dextrorotatory substance. The solution, moreover, failed to give any indication of the presence of a reducing substance, even by prolonged boiling with Fehling’s solution, and 390 Profs. A. Gamgee and W. Jones. [Feb. 9, was tound to be rich in material which yields xanthine bases on hydrolysis with sulphuric acid. The main portion of the gland substance, purified by the processes above described, was treated with 20 parts of water, and to the filtered solution acetic acid was added, drop by drop. When a quantity of acid had been added sufficient to bring the amount of acid in the entire solution to 1 per cent., a well-defined white, flocculent, precipi- tate separated. This precipitate of nucleoproteid was separated by the centrifuge, suspended in water, and treated with an extremely - dilute solution of ammonia, drop by drop, and the reaction of the liquid continuously tested with litmus. A very small amount of alkali was needed to neutralise the adherent acetic acid, when the solution became neutral and remained so until approximately twice as much ammonia had been used as had been required to completely dissolve the nucleoproteid. Evidently, the nucleoproteid is, at least, a dibasic acid, whose acid ammonium salt is soluble in water and neutral to litmus. Purification of the nucleoproteid was efiected by alternate solution in ammonia and precipitation with a minimal quantity of acetic acid. The final solution was poured into five volumes of 95 per cent. alcohol, washed repeatedly by decantation with excessively large quantities of 95 per cent. alcohol and ether, and then placed in an exsiccator over sulphuric acid. Optical Properties. 1, A weighed amount of the nucleoproteid was suspended in water and dissolved by the addition of a trace of ammonia. The solution was made up to a definite volume with water, and examined polari- metrically : Weight of substance (W)... 1°006 gramme. Volume of solution (V) ...... 25 €.C. Observed angle (a)............ +3° 4 Lenethiot tube (isc eee 200 mm. alee Seale 2. The results of the above observation were confirmed by the ex- amination of another preparation of nucleoproteid. Weight of substance ......... 0°500 gramme. Volume of solution.......... fi 2bicies Observed angle ......... ae +1° 30’ Lenethiof tuber 2.2.4: 200 mm. The solution was treated with an excess of acetic acid and the pre- cipitate filtered off. The filtrate was found to be inactive. 1905. | On the Nucleoproterds of the Pancreas, ete. O91 B. Nuclein accompanying, and probably resulting from, the Nucleoproterd. Method of Preparation. The aqueous extract of the purified gland substance to which acetic acid had been added until it contained 1 per cent. of the latter, and from which the nucleoproteid had thus been separated, was treated with 20 per cent. acetic acid added a drop at a time. When the liquid contained 2 per cent. of the acid not the slightest precipita- tion had occurred. Continued addition of acetic acid, however, soon caused a turbidity, and when the acidity reached 5—6 per cent., a well-defined flocculent precipitate fell. This precipitate, which we shall call nuclein, was separated by means of the centrifuge and, at a great cost of material, was twice washed with water, the washings being separated by centrifugalising. The washed nuclein was suspended in water, and solution of ammonia added cautiously, one drop at a time ; when the nuclein was completely dissolved, the reaction of the liquid was still acid to litmus. This solution was poured into four volumes of 95 per cent. alcohol, and the precipitated nuclein washed and dried by the methods described in the case of the nucleoproteid. The fluid from which the ‘“nuclein” had been precipitated, as has been stated, was now poured into four volumes of alcohol, and the precipitate thus thrown down was washed and dehydrated by the action of alcohol and ether. This preparation, which is necessarily very impure and especially rich in organic salts, will be described and referred to as ‘‘ residual material.” Thus, by fractional precipitation with acetic acid, in the presence of inorganic salts, we have obtained three preparations. The nucleo- proteid, which is doubtless the body which Hammarsten denominated Proteid-z, is almost insoluble in pure water, but may be dissolved by minute quantities of ammonia and caustic soda. The body which we have termed nuclein, to indicate our opinion of its relation to the first substance, is soluble in water with the greatest ease. By the addition of a trace of copper sulphate to a solution of the nucleoproteid in caustic soda a fine pink colour is produced, but not a shade of violet makes its appearance until a comparatively large amount of copper solution has been added, a reaction which resembles closely “the biuret reaction” with the proteoses. The ‘“ nuclein” by similar treatment gives only the faintest pink colour, the violet shade being observed even when a very small amount of copper sulphate is used, while the “residual material” produces a violet colour from the beginning. It has recently been shown by one of us* that the nucleoproteid of * Walter Jones and G. H. Whipple, “The Nucleoproteid of the Suprarenal Gland,” ‘Amer. Jour. of Phys.,’ vol. 7 (1902), p. 423. 392 Profs. A. Gamgee and W. Jones. [Feb the pancreas, prepared in substantially the same manner as the pre- parations employed in the present research, yields, when subjected to hydrolytic treatment, two of the xanthine bases, viz., guanine and adenine, and in a ratio which closely approximates four equivalents of the former to one of the latter. The “nuclein” and “ residual material” of the present research were also found to yield xanthine bases on hydrolysis with sulphuric acid. All the three preparations under discussion contain phosphorus, all are completely precipitated from aqueous or faintly alkaline solutions by the addition of a trace of hydrochloric ecid, and all yield precipitates when their neutral solutions are boiled. Optical Properties of the Nuclein. We had convinced ourselves by the following experiment that the specific rotation of the substance which we have denominated “‘nuclein” would be found to be greater than that of the nucleo- proteid, before we had the opportunity of making a careful optical examination of the former substance. A perfectly neutral solution of the neucleoproteid was prepared by treating some of the substance with water and an insufficient amount of ammonia to effect complete solution. The filtered fluid, examined with the polarimeter in a 200 mm. tube, gave a rotation of 1° 46’. The solution was heated to boiling, and the coagulated albumin filtered off. The filtrate polarised in a 200 mm. tube gave a rotation of 1° 49’. Now, as is well known, the process of boiling, removing a portion of the albuminous matter previously forming part of the com- plex nucleoproteid molecule, converts the latter into a nuclein. As the length of the tube was the same, and the angle of rotation re- mained sensibly constant in our experiment, a decrease in the amount of matter in solution (equal to the coagulated albumin removed from it) must mean an increase in the specific rotation. The following direct determination of the specific rotation of the nuclein was made. The body was dissolved in water, and as the fluid was somewhat coloured, it was examined in a shorter tube than those which we have usually employed :— Weight of substance............... 1:009 gramme. Volume ob solution” 2s ee-cee 50 e.c. Obsenvedtang lee se eee nee +1° 18’ ene thot mules yore err ae 100 mm. [|p — = GAmAR The solution was treated with hydrochloric acid to precipitate the nuclein, and the filtered fluid examined in a tube 200 mm. long. ‘The rotation was slightly negative (0°9’). In reference to this observa- POs: | On the Nucleoproterds of the Pancreas, ete. 399 tion, we have to remark that we noticed several times that very slight levorotatory filtrates were obtained when hydrochloric acid was used for precipitating the proteid, and especially when the acid fluid was allowed to remain in contact with the precipitate. Presumably, the negative rotation is due to an optically negative acid albumin being formed, which is soluble in the dilute hydrochloric acid. As in the case of the nucleoproteid, a solution of the nuclein yields a coagulum on heating, and the rotation of the solution is not appre- ciably changed. This would lead one to assume the existence of a nuclein of which the specific rotation 1s greater than +64°-4. It can easily be proved that such a substance exists in the preparation which we have designated ‘“ residual material.” A weighed amount of this substance was dissolved in a measured volume of water. The solution was examined with the polarimeter, treated with hydrochloric acid, and the amount of matter determined in the filtrate, which was found to be optically inactive. The follow- ing data were obtained :— Weight of substance taken ............... 0-520 gramme, Weight of optically inactive matter ... 0°269 __,, Weight of optically active matter ...... 0-251 ei Woltrnte of SOlItOM yoo oot iek. se necnnsnes 25D 6.C. CIDRSICE IO BHOed Gaiey, ayee Bens SO MOM FORA SHORE + 1° 38’ LLGINGAN OM BTC OSes R an seca Ne Ae DOr See 200 mm. ial Slt C. Hammarsten’s Preparation. As we have already explained, Hammarsten’s so-called nucleopro- teid-6, whichis obtained from an extract of pancreas made by boiling the finely comminuted gland in water must, zpso facto, be a nuclein. The results which we had obtained and which have been described, made it highly desirable that we should make an optical examina- tion of this substance also. By slight departures* from the method described by Hammarsten, which were absolutely necessary to remove the colouring matter, but which cannot possibly have exercised any influence on the chemical nature of the product, we were able to pre- pare a nuclein which must have been identical with Hammarsten’s preparation (nucleoproteid-6), The substance which we obtained is soluble in water, and gives a violet biuret reaction. Its solution was comparatively highly coloured, but possessed so great a rotatory * We used ammonia for redissolving the nuclein, instead of a fixed alkali employed by Hammarsten. We also finally poured an aqueous solution of the nuclein into 95 per cent. alcohol, and washed by decantatiou with absolute alcohol and ether. 394 Profs. A. Gamgee and W. Jones. [ Feb. 9, power that fairly satisfactory polarimetric observations could be made in solutions of great dilution. The substance is dextrorotatory. The following data were obtained :— Weiehtvolgsulbscamce rey: ce ease eee 0°200 gramme. Voltmerotjsolintionyey -earpe Fela rere on aa: 2D; Cc! Observed angle (mean of eight readings) ... 0° 47’ Meng thot tube...) se noua. aaer ese eee 100 mm. (ailp =) hOne ot On the Nucleohiston of the Thymus Gland. It would seem quite easy to obtain this substance in any desired quantity by following the very simple method which Lilienfeld described in twenty lines.* This method leads, however, to a product whose solutions are highly opalescent, and an optical examination could not be thought of. The cloudiness is so persistent, that for a long time we were inclined to believe it to be a property inherent in the substance. We finally suc- ceeded, however, in obtaining solutions almost as colourless and transparent as distilled water. It is only necessary to extract Lilien- feld’s preparation with a 5 per cent. solution of ammonium acetate and filter. The fluid filters very slowly, but perfectly clear and con- tinuously. The solution was poured into 95 per cent. alcohol, and the precipitated proteid washed and dried with alcohol and ether, as described in connection with other preparations mentioned in this paper. The substance thus obtained was submitted to polarimetric examination, the solution being made with the aid of very dilute ammonia. The following data were obtained :— Weight of substance ......... 2°023 gramme Volume of solution <22..---.... 50 ¢.¢. Observedsaneley Sy aes: + 3° 20’ iene thvon tulbecspwecnesn 9. -ce 220 mm. On the Nucleoproteid of the Suprarenal Gland. In a research carried on conjointly with G. H. Whipple,t one of us lately described the nucleoproteid of the suprarenal gland, and showed that this body is a thymo-nucleoproteid. Ultimate analyses showed that the nucleoproteids of the suprarenal gland of the ox and the * Leon Lilienfeld, ‘“‘ Zur Chemie der Leucocyten,” ‘ Zeit. f. physiol. Chem.,” vol, 18 (1894), p. 473. + Walter Jones and G. H. Whipple, op. cit., p. 423 (Sept. 4902). 1903.] On the Nucleoproteids of the Pancreas, ete. 395 sheep are identical, and scarcely differ in chemical composition from the nucleoproteid of the pancreas prepared substantially by the same method as that which has served for the researches described in this paper. The following table exhibits the results of the ultimate analyses of these bodies, and, for purposes of comparison, the analyses made by Hammarsten of his preparation is also given :— Nucleoproteid | Nucleoproteid §Nucleoproteid of suprarenal of suprarenal of pancreas Hammarsten’s gland of sheep. gland of ox. of pig. preparation. Ore... 46°22 46°81 45°83 43°62 lal ee 6°10 6°38 6°26 5°45 lee 4-70 4°72 5°05 4°48 Niece peo 2 17°85 17-42 Wes) As closely as the analytical processes at command could determine, the nucleoproteids of the pancreas and the suprarenal gland yield guanine and adenine in the same relative proportions, and these appear to indicate that one molecule of a nucleinic acid, or of a nucleo- proteid, may yield two different xanthine bases. We must refer the reader to the paper quoted above for a descrip- tion of a method of separating the nucleoproteid of the suprarenal gland. As is well known, the characteristic physiologically active con- stituent of this gland forms a dark brown pigment when exposed in aqueous solution to the oxidising action of the air. Aqueous extracts of the gland are therefore always highly coloured, and this colouring matter places great difficulties in the way of the preparation of sub- stances from the gland which are intended for optical examination. While, therefore, the work on the nucleoproteid of the suprarenal gland is not as satisfactory as we could desire, it can nevertheless be stated most positively that this nucleoproteid also is dextrorotatory. The method of isolation which we employed does not differ essen- tially from that employed in the research already referred to, except that the gland was extracted several times with acetic acid before removing the nucleoproteid. A substance was finally obtained which is too highly coloured for accurate polarimetric determinations, but which, even in the necessarily high dilutions which could alone be used, could easily be shown to be dextrorotatory. The following data were obtained :— Weight of substance ......... 0-199 gramme. Nolume of/solution. \:2....5-- 25 ¢.¢. Observed angle a... -2.-..: . +0° 23’ MemechpO ty tiloemene:. 2.40042: 100 mm i) ty VOL. LXXI. 396 Profs. A. Gamgee and W. Jones. [Betis The value of this rotation is lable to revision, but its direction is beyond question. Before formulating the general conclusions which, it appears to us, may legitimately be deduced from the researches of which an account has been given in this paper, we may sum up our work in the following manner :— | Summary. We have, in this paper, described six substances obtained from ~ _ various glands and have given methods by which several of these may be isolated and obtained sufficiently free from colouring matters to admit of exact polarimetric determinations. All six of these substances yield on hydrolysis, alouminous bodies, phosphoric acid, and purin derivatives, and all contain iron in stable combination ; they are, therefore, all nucleoproteids in the wide sense of the term. The methods of preparation were such as to exclude all dextro- rotatory substances which are not of a proteid nature, and all prepara- tions were shown to be free from substances which reduce Fehling’s solution even on prolonged boiling. Nevertheless, all these substances were found to be dextrorotatory, having specific rotations for light of the wave-length of D which vary from 37°:58, that of the nucleohis- ton of the thymus gland, to 97°-9 that of Hammarsten’s nuclein obtained from the pancreas and described by him as proteid. General Conclusions. 1. The nucleoproteids (employing this term in its wider sense, as including the compounds of the nucleinic acids with albuminous sub- stances) which are contained in the pancreas, the thymus, and the suprarenal gland are dextrorotatory albuminous compounds. 2. When a nucleoproteid, by the splitting-off of albuminous mole- cules, which in its original condition formed part of its more complex molecule, becomes converted into a nucleoproteid of the ‘‘nuclein” type, its specific rotation increases. 3 3. It is legitimate to infer that not only the well characterised and typical nucleoproteids which we have subjected to examination, but all the nucleoproteids, including in this term the so-called nucleins, form a class of dextrorotatory albuminous substances. Whilst the facts which have come under our notice appeared to us so full of interest that it would not have been wise to defer their pub- lication, we are perfectly alive to the importance of answering with the least possible delay a number of most interesting questions sug- gested by them. We are already actively engaged in the investigation of these questions and hope shortly to publish the results of our 1903. ] On the Nueleoproterds of the Pancreas, ete. 397 enquiries. We trust, therefore, that during the next few months we may be permitted to work out so far as we are able, the problems which have been suggested by the new facts recorded in this paper. Supplementary Bibliographical Note. Since the above paper has been in print, it has come under the notice of one of us that the late Professor Alexander Schmidt, of Dorpat, in his published researches on the coagulation of the blood,* drew attention to the fact that among the soluble constituents of protoplasm was a body to which he gave the name of “ Cytoglobin,” and which he found to be dextrogyrous. So far as we are aware, this observation of A. Schmidt has never been noticed or quoted, either by systematic writers on physiological chemistry or by those who have devoted their attention to, or written upon, the subject which formed the life-work of the Dorpat professor. There can be no question that Schmidt’s cytoglobin was an exceedingly impure mixture of nucleo- proteids, an opinion which is based upon the fact that his substance contained 12°52 per cent. of ash, and that, on ultimate organic analysis, the amount of carbon found was 56°36 per cent., as compared with 45°83, the percentage of carbon in the nucleoproteid of the pancreas. Still the fact remains that this indefatigable worker, whose suggestive writings have been too little read, left data which prove that the so-called ‘“‘ Cytoglobin” was nucleoproteid in nature, though in no sense a definite proximate principle, and that this impure mixture of nucleoproteids was characteristically dextrogyrous. Murch 4, 1903. A. G. * Alexander Schmidt, ‘ Zur Blutlehre, Leipzig, Verlag v. F.C. W. Vogel, 1892. Refer to the chapter entitled ‘“‘ Ueber den in Wasser léslichen Bestandtheil des: Protoplasmas,” &c. (pp. 127—142) ; ‘ Weitere Beitraege zur Blutlehre’ (nach des Verfassers Tode herausgegeben). Wiesbaden, J. F. Bergmann, 1895. Refer to the chapter entitled “Zur Kenntniss des Protoplasmas und seiner Derivate” (pp. 201—249). eS) R WOE. LXXtI. 398 Prof. J. A. Fleming. Magnetic Detector for [Feb. 11, ‘““A Note on a Form of Magnetic Detector for Hertzian Waves, adapted for Quantitative Work.” By Dr. J. A. FLEMING, F.R.S., Professor of Electrical Engineering in University College, London. Received February 11,—Read March 5 1903. ? The known power of electrical oscillations to demagnetise iron or steel was first applied in the construction of a detector of Hertzian waves, as far as the author is aware, by Mr. EK. Rutherford.* The power possessed by electrical oscillations to annul the magnetic hysteresis of iron was discovered by Mr. G. Marconi and applied by him in the construction of his ingenious and extraordinarily sensitive Hertzian wave detector, for use in connection with wireless tele- graphy.t The following note describes a form of magnetic Hertzian wave detector, which has been constructed by the writer for the purpose of quantitative experiments in connection with Hertzian waves. Every one who has experimented with a Hertzian oscillator, or electric wave radiator in any form, involving a spark gap, is well aware of the immense difference in the radiative power produced by slight alterations in the nature of the spark or the spark balls, and has felt the want of some instrument which shall indicate and measure exactly the intensity of the radiation. As a receiving instrument, the coherer or sensitive imperfect contact is of very little use quantita- tively, because its indications are influenced by very slight accidental changes at the contact or contacts. Thus, the sensitiveness of the metallic filgs coherer depends upon the manner in which it was leit after its last use, and by the mode in which it is tapped or shaken, and the change in the conductivity which it experiences on the impact of an electric wave, is variable and uncertain. Hence, although sensi- tive as a mere wave detector, the coherer is of little or no use in quantitative work. On the other hand, the magnetic detector is not -only superior to the coherer in sensitiveness when properly con- structed, but is capable of being used as a measuring instrument. In the form in which it was constructed by Mr. Rutherford, an extremely fine bundle of iron or steel wires was magnetised by means of a magnet, or by being placed in the interior of a solenoid, and then demagnetised by an electrical oscillation passing through another coil * See Mr. E. Rutherford, “On a Magnetic Detector of Electric Waves and some of its applications,” ‘ Roy. Soc. Proc.,’ 1896, vol. 60, p. 184; see also ‘ Phil. trans. .AL 1897, vole 189g0 i + Mr. G. Marconi, “ Note on a Magnetic Detector for Electric Waves which can be employed as a Receiver for Space Telegraphy,”’ ‘Roy. Soc, Proc.,’ 1902, vol. 70, p. 341, 1903.) Hertzian Waves, adapted for Quantitative Work. 399 surrounding it. The amount of demagnetisation was detected by means of a magnetometer. In this form, it has been much used in experimental work, but it was not a telegraphic receiver.* In the sensitive telegraph receiver invented by Mr. Marconi the change in magnetisation of the iron, due to the temporary abolition of hysteresis, is detected by the production of a sound in a telephone connected to a secondary coil surrounding the iron. After trying various forms, the writer has found that a convenient magnetic detector for Hertzian waves can be constructed in the follow- Ing manner :— On a pasteboard tube, about # of an inch in diameter and 5 or 6 inches long, are placed six bobbins of hard fibre, each of which contains about 6000 turns of No. 40 silk covered copper wire. ‘These bobbins are joined in series, and form a well-insulated secondary coil, having a resistance of about 6000 ohms. In the interior of this tube are placed seven or eight small bundles of iron wire, each about 6 inches in length, each bundle being composed of eight wires, No, 26 S.W.G. in size, previously well paraffined or painted with shellac varnish. Hach little bundle of iron is wound over uniformly with a magnetising coil formed of No. 36 silk-covered copper wire in one layer, and over this, but separated from it by one or two layers of gutta-percha tissue, is wound a single layer of No. 26 wire, forming a demagnetising coil. This last coil is in turn covered over with one or two layers of gutta-percha tissue. The magnetising or inner coils are connected in series with one another, so that when a current passes through the whole of them, it magnetises the whole of the wires in such a manner that contiguous ends have the same polarity. The outer or demagnetising coils are joined in parallel. Associated with this induction coil is a rotating commutator, consisting of a number of hard fibre dises secured on a steel shaft, which is rotated by an electric motor about 500 times a minute. There are four of these fibre discs, and each dise has let in its periphery a strip of brass, occupying a certain angle of the circum- ference. These wheels may be distinguished as Nos. 1, 2, 3, and 4, The brass sector of No. 1 occupies 95° of its circumference; the brass sectors of Nos. 2 and 3 occupy 135° of their circumference ; and that of No. 4 disc 140° of its circumference. Four little springy brass brushes make contact with the circumference of these wheels, and therefore serve to interrupt or make electric circuits as the disc revolves. The function of the disc No. 1 is to make and break the * Note added March 7th, A general term seems to be required to include all forms of wave-detecting devices, The author suggests the word kumascope (from xvpa, a wave) for this purpose, Uncouth phrases, such ag anticoherer or selfs decohering-coherer, which have crept into use in connection with Hertzian wave telegraphy, would be rendered unnecessary. 2. G2 400 Prof. J. A. Fleming. Magnetic Detector for [Feb. 11, circuit of the magnetising coils placed round the iron bundles, and thus by applying a magnetising current to magnetise them during a portion of one period of rotation of the disc, and leave them mag- netised in virtue of magnetic retentivity during the remaining portion. The function of discs 2 and 3 is to short-circuit the terminals of the secondary coil of the bobbin during the time that the magnetising current is being applied by disc No. 1. A sensitive movable coil galvanometer is employed in connection with the secondary coil, one terminal of the galvanometer being permanently connected to one ~ terminal of the secondary coil, and the other terminal connected through the intermittent contact made by the disc No. 4. This dise No. 4 is so set that during the time that the secondary coil is short- circuited, and whilst the battery current is being applied to magnetise the iron wire bundles, the galvanometer circuit is interrupted by the contact on disc No. 4. The operations which go on during one complete revolution of the discs are as follows :—First the magnetising current of a battery of secondary cells is applied to magnetise the iron bundles, and during the time this magnetising current is being applied, the terminals of the fine wire secondary coil are short-circuited and the galvanometer is disconnected. Shortly after the magnetising current is interrupted the secondary bobbin is unshort-circuited, and an instant afterwards the galyanometer circuit is completed and remains completed during the remainder of one revolution. Hence, during a large part of one revolution, the iron wire bundles are left magnetised, but the mag- netising current is stopped and the galvanometer is connected to the secondary coil. If during this period an electrical oscillation is passed through the demagnetising coils, an electromotive force is induced in the secondary bobbin by the demagnetisation of the iron and causes a deflection of the galvanometer coil. Since the interruptor dises are rotating very rapidly, if the electrical oscillation continues, these intermittent electromotive impulses produce the effect of a continuous current in the galvanometer circuit, resulting in a steady deflection, which is proportional to the demagnetising force being applied to the iron, other things remaining equal. If the oscillation lasts only a very short time, the galvanometer will make a small deflection ; but if the oscillation lasts for a longer time, then the galvanometer deflection is larger, and tends to become steady. By means of such an arrangement it is possible to verify the law according to which variation falls off with distance. The instrument can be employed also as a telegraphic receiving instrument, but its chief use will be for comparing together the wave-making power of different radiators. For this purpose the oscillation coils must be con- nected to two long connecting wires, or one end may be connected to the earth and the other to a vertical aerial. This detector serves, for 1903.] Hertzian Waves, adapted for Quantitative Work. AOL instance, to show in a very marked manner the great effect of slight differences in the surface of the spark balls. If a steady series of sparks from an induction coil is passed between the spark balls of a Hertz linear radiator, it will produce a steady deflection on a galvano- meter connected with the above-described receiver placed at a distance. If the balls are then polished, the galvanometer deflection immediately increases considerably. If, on the other hand, the balls are slightly smeared with oil, the galvanometer deflection decreases. Jf the radiator is approached to the receiver, or withdrawn from it, corre- sponding variations in the galvanometer deflection take place. Such an instrument will probably be found of great use in connec- tion with the design of radiators and transmitters for Hertzian wave wireless telegraphy. Up to the present it has been generally difficult to ascertain whether an improvement in the signalling is due to an accidental increase in sensitiveness in the coherer, or to any alteration or change made in the transmitter. Similarly, the instrument promises to be of considerable use in the investigation of the transparency or opacity of various substances to Hertzian waves, not merely qualitatively, but in the determination of a coefficient of absorption. Preliminary experiments of this descrip- tion made with the above-described instrument seem to promise for it a field of practical utility, both in the laboratory and in connection with Hertzian wave telegraphy. 2 In the numerous experiments which finally resulted in the construc- tion of the above-described form of wave detector, it was found to be essential to have the iron core in the form of a number of small bundles of iron wire, each wound over with its own magnetising and demagnetising coil. No good results could be obtained when the iron core was in the form of a large bundle, say half an inch in diameter, and enveloped by a single magnetising and demagnetising coil. Another condition of success is the short-circuiting of the fine wire secondary coil during the time of magnetisation of the core. ‘The core ean be indefinitely increased in size, provided the augmentation of mass is obtained by multiplying small individual cores, each consisting of not more than eight or ten fine iron wires, and each wound over with a separate magnetising and demagnetising coil. ‘The electro- motive force in the secondary coil can in this manner be increased as much as is desired, and a very sensitive wave detector produced. The commutator can be driven either by an electric motor or by any other source of power. In conclusion, I have pleasure in mentioning the intelligent assistance rendered to me by Mr. A. Blok in the experiments conducted in con- nection with this appliance. 402 A New Form of Self-restoring Coherer. [Mar. 18, “A New Form of Self-restoring Coherer.” By Sir OLiver LopGE, F.R.S. Communicated verbally March 5,—Received in Manuscript March 18, 1903. On the general subject of the detection of Hertzian waves the writer took the opportunity of the discussion on Dr. Fleming’s paper (p. 398), to describe briefly the latest form of coherer, which Dr. Muir- head and he had brought out and always now employed in connection: _ with space telegraphy, and which their assistant Mr. E. E. Robinson had helped to work out. It might be described as a development of the mercury form of coherer described some years ago by Lord Rayleigh, and again in a modified fashion by Mr. Rollo Appleyard. In Lord Rayleigh’s form this consisted of a pool of mercury cut across with a paraffined knife, and the two half pools connected to a battery and key. As soon as the key was depressed so as to throw a few volts on to the intervening film of ol, the electrostatic pressure seemed to squeeze the oil out, and the pools of mercury became one. Needle points dipping in oil and mercury were tried as practical coherers, the points being pulled out electromagneticaily every time a signal arrived. Rotating forms of contact for automatic decoherence were also tried in various forms, and ultimately the method took the form of a rotating sharp-edged steel wheel, about half an inch in dia- meter, constantly touching a pool or column of mercury on which was a thin layer of oil. No effective contact occurs between the wheel and the mercury, notwithstanding the immersion, because of the film of oil ; but the slightest difference of potential applied to the two, even less than one volt, is sufficient to break the film down and complete a circuit, which, however, the rotation of the wheel instantaneously breaks again. The spark is so sudden that for its purposes the wheel is for the instant virtually stationary, and yet the decohesion is so rapid that signals can be received in very rapid succession. The definiteness of the surfaces and of the intervening layer make the instrument remarkably trustworthy, and the thinness of the insulating film makes it very sensitive. In fact a single cell of a battery cannot be employed as a detector, because it is of too high a voltage for the film to stand. A fraction of a volt is employed by a potentiometer device—usually something like one-tenth of a volt—and it is adjusted to suit circumstances. The battery acts through the coherer direct on a low resistance recorder, and the record on the strip shows every character of the arriving pulses, and exhibits any defect in the signal- ling. Provided that every joint and contact, except the one intended to be filmed, is thoroughly good, the coherer in this form is so definite and satisfactory that it becomes safe to say that the only outstanding defects are those which occur at the sending end. The signals are 1903.] On Central American Karthquakes, ete. 403 picked up and recorded precisely as they are emitted, as has been tested by intercalating a ‘siphon recorder in a much diluted tapping circuit at the sending end, so as to get a record with which to make comparison. The traces obtained at the two ends are identical to a surprising degree. The mercury level has an adjustment which is easily made. One precaution is to keep the rim of the wheel clear of dust, which is done by a cork or leather pad pressed lightly against it by a spring. The instrument is not at all sensitive to tremor, and requires no particular delicacy of adjustment. The wheel has to be positive, the mercury negative. . A telephone in circuit, through a transformer or otherwise, affords an easy method of discriminating the signals by ear. The speed of the wheel gives another convenient adjustment to suit various circum- stances. “On Central American Earthquakes, particularly the Earthquake of 1838.” By Admiral Sir Jonn DatrymMpLe Hay, :Bart., G.C.B., F.R.S. Received March 6,—Read March 19, 1903. The very interesting report of Dr. Tempest Anderson and Dr. J. S. Flett on “Recent Volcanic Eruptions in the West Indies,” induces me to suggest that inquiries should be made in Colombia and in Central America as to disturbances in those regions, in order to trace the connection existing between the earthquakes and volcanic erup- tions which are there so prevalent, and those in the West Indies. Doubtless Mr. Welby, who has just returned to his post at Sta. Fé de Bogota, might be able to obtain thence some information on this matter. In the British Association Report, of 1858, upon Harthquakes, those of 1835 and of 1841 are given ; but, unfortunately, no information is given as to the effect of those disturbances from any British source. The British Association in both cases seem to have been mainly indebted for their information to the Académie, and hence the effects in Martinique and Guadeloupe are quoted, but nothing is reported from our West India Islands. One of the most terrible disturbances in its effects, that of 1838, is not even alluded to. The only record of it is in the work of EH. G. Squier, late Chargé d’Affaires from the U.S. of America to the Repub- dics of Central America (2 vols., Appleton, New York, 1852), who was commissioned by his Government to report upon that region in reference to the Nicaraguan Canal. Vol. 1 has the map; vol. 2, pp. 114, 115, and p. 162 alludes to the eruptions of Cosequina in 1838, 404 On Central American Earthquakes, ete. [ Mar. 6, and in describing the earthquakes and disturbances and eruptions of 1835, speaks of their effects in New Granada, Nicaragua, Popayan, Bogota, Sta. Marta, Caracas, Hayti, Curacao, and Jamaica. I think it may be of interest to record my personal observations of the earthquakes and eruptions of 1838 as given in the log of H.MS. “‘Tmogene,” Captain (afterwards Sir Henry) Bruce, in which | was then serving. On Friday, November 16, 1838, noon, Lat. 6° 2’ N., Long., 82° 9° W. Rio Lejo (to which we were bound from Panama) bearing N. 38, W. - 490 miles. At 10 P.M. we felt the shock of an earthquake. Tuesday, November 20, 1838, at 3.5 A.M., noon of that day, Lat., 9° 17’ N., Long. 85° 2’ W. Three heavy shocks in succession. Calm c.r. The topsails were lowered on the cap, and the ship lay till daylight, uncertain whether she had struck on a wreck or whether she was damaged. She rolled over to the limit of her capacity on three successive occasions. At daylight (6 A.M.) all-hands were sent for to see an extraordinary spectacle. ‘The sea was white, as if it had been mixed with marl, and, as far as the eye could reach, the sea was covered with fish of various kinds stunned by the concussion. The boats picked up a sufficient supply for the ship’s company, principally albacore, bonita, and dolphin. Some turtle were also captured. We anchored in Rio Lejo on November 25, and remained there till November 30, 1838. On the Ist December, we communicated with H.M.S. “Sulphur,” Captain E. Belcher, then surveying the Gulf of Fonseca. Cosequina was in eruption, but El Viejo was quiescent. Belcher had ascended El] Viejo on February 10, 1838, without difficulty, but Cosequina was in active eruption. No record, so far as I know, has been published of this seismic disturbance ; and although newspapers, received sometime after at Guaymas, reported synchronous disturbances at Tobago and elsewhere, no information except that contained in Mr. Squier’s report is to be found on the subject. It would seem desirable that the inquiry, so weil carried out by Dr. Tempest Anderson and Dr. Flett should be extended into Colombia and Central America. The disturbance to which I have alluded above is traceable for more than 1500 miles. The Colonial Office informs me that an earthquake was reported at St. Lucia on January 11, 1839, doing much damage. _ No information is recorded as to Tobago. 1903. ] The Emanations of Radium. 405 “ The Emanations of Radium.” By Sir WILLIAM Crooxkes, F.E.S. Received March 17,—Read March 19, 1903. A solution of almost pure radium nitrate which had been used for spectrographic work, was evaporated to dryness in a dish, and the crystalline residue examined in a dark room. It was feebly luminous. A screen of platinocyanide of barium brought near the residue glowed with a green light, the intensity varying with the distance separating them. The phosphorescence disappeared as soon as the screen was removed from the influence of the radium. A screen of Sidot’s hexagonal blende (zine sulphide), said to be useful for detecting polonium radiations, was almost as luminous as the platinocyanide screen in presence of radium, but there was more residual phosphorescence, lasting from a few minutes to half an hour or more according to the strength and duration of the initial excite- ment. The persistence of radio-activity on glass vessels which have con- tained radium is remarkable. Filters, beakers, and dishes used in the laboratory for operations with radium, after having been washed in the usual way, remain radio-active; a piece of blende screen held inside the beaker or other vessel immediately glowing with the presence of radium. The blende screen is sensitive to mechanical shocks. A tap with the tip of a penknife will produce a sudden spark of light, and a scratch with the blade will show itself as an evanescent luminous line. A diamond crystal brought near the radium nitrate glowed with a pale bluish-green light, as it would in a “ Radiant Matter ” tube under the influence of cathodic bombardment. On removing the diamond from the radium it ceased to glow, but, when laid on the sensitive screen, it produced phosphorescence beneath, which lasted some minutes. During these manipulations the diamond accidentally touched the radium nitrate in the dish, and thus a few imperceptible grains of the radium salt got on to the zine sulphide screen. The surface was immediately dotted about with brilliant specks of green light, some being a millimetre or more across, although the inducing particles were too small to be detected on the white screen when examined by daylight. In a dark room, under a microscope with a 32-inch objective, each luminous spot is seen to have a dull centre surrounded by a luminous halo extending for some distance around.’ The dark centre itself appears to shoot out light at intervals in different directions. Outside the halo, the dark surface of the screen scintillates with sparks 406 | Sir W. Crookes. [ Mar. 17, of light. No two flashes succeed one another on the same spot, but are scattered over the surface, coming and going instantaneously, no movement of translation being seen. The scintillations are somewhat better seen with a pocket lens magnifying about 20 diameters. They are less visible on the barium platinocyanide than on the zine sulphide screen. A powerful electro-magnet has no apparent effect on the scintillations, which appear quite unaffected when the current is made or broken, the screen being close to the poles and arranged axially or equatorially. A solid piece of radium nitrate is slowly brought near the screen. The general phosphorescence of the screen as visible to the naked eye varies according to the distance of the radium from it. On now examining the surface with the pocket lens, the radium being far off and the screen faintly luminous, the scintillating spots are sparsely scattered over the surface. On bringing the radium nearer the screen the scintillations become more numerous and brighter, until when close together the flashes follow each other so quickly that the surface looks like a turbulent luminous sea. When the scintillating points are few there is no residual phosphorescence to be seen, and the sparks succeeding each other appear like stars on a black sky. When, however, the bombardment exceeds a certain intensity, the residual phosphorescent glow spreads over the screen, without, however, inter- fering with the scintillations. If the end of a platinum wire which has been dipped in a solution of radium nitrate and dried is brought near the screen, the scintilla- tions become very numerous and energetic, and cease immediately the wire is removed. If, however, the end of the wire touches the screen, a luminous spot is produced, which then becomes a centre of activity, and the screen remains alive with scintillations in the neigh- bourhood of the spot for many weeks afterwards. “ Polonium ” basic nitrate produces a similar effect on the screen, but the scintillations are not so numerous. Microscopic glass, very thin aluminium foii, and thin mica do not stop the general luminosity of the screen from the X-rays, but arrest the scintillations. I could detect no variation in the scintillations when a rapid blast of air was blown between the screen and the radium salt. A beam of X-rays from an active tube was passed through a hole in a lead plate on to a blende screen. A luminous spot was pro- duced on the screen, but I could detect no scintillations, only a smooth uniform phosphorescence. A piece of radium salt brought near gave the scintillations as usual, superposed on the fainter phosphorescence caused by the X-rays, and they were not interfered with in any degree by the presence of X-rays falling on the same spot. During these experiments the fingers soon become soiled with radium, err 1905.] The Emanations of Radiwi. 407 and produce phosphorescence when brought near the screen. On turning the lens to the, apparently, uniformly lighted edge of the screen close to the finger, the scintillations are seen to be closer and more numerous ; what to the naked eye appears like a uniform “ milky way, under the lens is a multitude of stellar points, flashing over the whole surface. A clean finger does not show any effect, but a touch with a soiled finger is sufficient to confer on it the property. Washing the fingers stops their action. It was of interest to see if rarefying the air would have any effect on the scintillations. A blende screen was fixed near a flat glass window in a vacuum tube, and a piece of radium salt was attached to an iron rocker, so that the movement of an outside magnet would either bring the radium opposite the screen or draw it away altogether. A microscope gave a good image of the surface of the screen, and in a dark room the scintillations were well seen. No particular difference was observed in a high vacuum ; indeed, if any- thing, the sparks appeared a trifle brighter and sharper in air than in vacuo. A duplicate apparatus in air was put close to the one in the vacuum tube, so that the eye could pass rapidly from one to the other, and it was so adjusted that the scintillations were about equal when each was in air. The vacuum apparatus was now exhausted to a very high point, and the appearance on each screen was noticed. Here again I thought the sparks in the vacuum were not quite so bright as in air, and on breaking the capillary tube of the pump, and observing as the air entered, the same impression was leit on my mind; but the differences, if any, are very minute, and are scarcely greater than might arise from errors of observation. It is difficult to form an estimate of the number of flashes of light per second. But with the radium at about 5 cm. off the screen they are barely detectable, not being more than one or two per second. As the distance of the radium diminishes the flashes become more ire- quent, until at 1 or 2 cm. thev are too numerous to count. | Added March 18.—On bringing alternately a Sidot’s blende screen and one of barium platinocyanide, face downwards, near a dish of “polonium ” sub-nitrate, each became luminous, the blende screen being very little brighter of the two. On testing the two screens over a crucible containing dry radium nitrate, both glowed ; in this case the blende screen being much the brighter. Examined with a lens, the light of the blende screen was seen to consist of a mass of scin- tillations, while that of the platinocyanide screen was a uniform — glow, on which the scintillations were much less apparent. The screens were now turned face upwards so that emanations from the active bodies would have to pass through the thickness of card before reaching the sensitive surface. Placed over the ‘“‘ polonium ” 408 The Emanations of Radium. [Mar 17, neither screen showed any light. Over the radium the platino- cyanide screen showed a very luminous disc, corresponding with the opening of the crucible, but the blende disc remained quite dark. It therefore appears that practically the whole of the luminosity on the blende screen, whether due to radium or “ polonium,” is occa- sioned by emanations which will not penetrate card. These are the emanations which cause the scintillations, and the reason why they. are distinct on the blende and feeble on the platinocyanide screen, is that with the latter the sparks are seen on a luminous ground of general phosphorescence which renders the eye less able to see the scintillations. Considering how coarse-grained the structure of matter must be to particles forming the emanations from radium, I cannot imagine that their relative penetrative powers depend on difference of size. I attribute the arrest of the scintillating particles to their electrical character, and to the ready way in which they are attracted by the coarser atoms or molecules of matter. I have shown that radium emanations cohere to almost everything with which they come into contact. Bismuth,* lead, platinum, thorium, uranium, elements of high atomic weight and density, possess this attraction in a high degree, and only lose the emanations very slowly, giving rise to what is known as “induced radio-activity.” The emanations so absorbed from radium by bismuth, platinum, and probably other bodies, retain the property of producing scintillations on a blende screen, and are non-penetrating. | It seems probable that in these phenomena we are actually wit- nessing the bombardment of the screen by the electronst hurled off by radium with a velocity of the order of that of light ; each scin- tillation rendering visible the impact of an electron on the screen. Although, at present, I have not been able to form even a rough approximation to the number oi electrons hitting the screen in a given time, it is evident that this is not of an order of magnitude incon- ceivably great. Hach electron is rendered apparent only by the enormous extent of lateral disturbance produced by its impact on the sensitive surface, just as individual drops of rain falling on a still pool are not seen as such, but by reason of the splash they make on impact, and the ripples and waves they produce in ever-widening circles. * T have been quite unable to detect any lines but those of bismuth (and of known impurities) in the spectrum of the strongest and most active “ polonium” salt I have been able to procure. + Radiant matter, satellites, corpuscles, nuclei; whatever they are, they act like material masses. 1903. ] Constitution of the Oopper-Tin Series of Alloys. 409 BAKERIAN LecturE.—“On the Constitution of the Copper-Tin Bemes of Alloys: By C. T. Huycock, F.R.S; and) F. H. NEVILLE, F.R.S. Received February 26,—Lecture delivered. February 26, 1903. (Abstract.) This paper is an attempt to fill a very serious gap in the study of alloys. As a rule, an alloy begins to be interesting when the tempera- ture of the liquid alloy has fallen to its freezing point. This point, which records the moment when solid first appears in the liquid, is easily observed on account of the evolution of latent heat that occurs on the formation of solid, and if the freezing points of all the alloys of a series are determined, we can plot the freezing-point curve.: Many such curves have been traced in the last ten years: that of the copper-tin alloys is given by the upper line in our diagram. The curve consists of several branches cutting each other in angular points. The one thing that these curves record without ambiguity is the number of different solids that can crystallise out of the liquid alloys, for each branch corresponds to the crystallisation of a different substance. But this is almost all that such curves tell us with certainty. They do not tell us whether the solids forming are the pure metals, or pure compounds, or crystalline solid solutions of the metals. Other experiments are needed to decide such questions. The other great branch of the study of alloys consists in the microscopic examination of the solid alloys after they have cooled to ordinary temperature, that is to say, after they have, in general, ceased to undergo change. Between these two series of experiments there is _ an enormous gap of temperature, it may be 1000 or 500 degrees, and it is in this range of temperature that the whole life-history of the alloy, regarded as an organism, is to be found. The only fruitful experiments we know of dealing with this intermediate region are the cooling curves of Roberts-Austen and Stansfield. These observers traced auto- matically the whole of the cooling of the bronzes and obtained some remarkable results. They found that the evolution of heat at the freezing point was often succeeded at much lower temperatures by other evolutions of heat, and that many of these must have occurred ° after the alloy had wholly solidified. These thermal changes point to important chemical or physical changes, though they do not tell us what these changes are. They suggest, however, that the final patterns found by the microscope in the solid alloys are likely to be very complicated, as they may contain several records superposed the one on the other. We found these patterns very beautiful, but hopelessly complicated. 410 Messrs. C. T. Heycock and F.-H. Neville. [Feb. 26, It occurred to us, and this is the method of the paper, that we could simplify the phenomena by a systematic chilling of the ingots at selected temperatures. A number of small ingots of the same alloy were placed in separate tubes in a bath of tin, together with a recording pyrometer, the temperature was raised above the freezing point of the alloy, and the whole very slowly cooled, the slow cooling being an essential feature of the experiment, Ingots were then extracted at selected temperatures and rapidly chilled by immersion in water. The microscopic examination of these chilled ingots showed that it was quite easy to distinguish the large crystals, that had formed during the slow cooling preceding the chill, from the matter that was liquid when the ingot was withdrawn from the furnace. Successive chills of an alloy exhibit the solid growing in amount as the temperature falls, and finally show the ingot completely full of solid. We thus obtain, with very reasonable accuracy, the temperature of com- plete solidification of the alloy ; and by applying the method to alloys with different percentages of tin we have traced anew curve, the “solidus,” or curve of complete solidification. The solidus of the bronzes is the second line of the diagram. It is a remarkable line, made up of sloping, horizontal and vertical branches. As in the freezing-point curve, each branch corresponds to the crystallisation of a different solid. In the notation of Professor Roozeboom, the upper curve ABC...I, is called the liquidus, because all wholly liquid alloys le above it; and the solidus, Abledef E,E3;H’H"I, is so named because all wholly solid alloys lie below it. The solidus of the bronzes is remarkable for the very narrow range of temperature within which some alloys pass from the wholly liquid to the wholly solid state. According to Roozeboom’s theory, each sloping branch of the solidus, . and there are four such in the diagram, corresponds to the crystalli- sation out of the liquid of a different series of solid solutions, each vertical part to the crystallisation of a pure body, and each horizontal part to the case of the solid alloy at temperatures immediately below the solidus, being a complex of two substances. Our examination of the chilled ingots has completely verified all these statements. The evolutions of heat observed by Roberts-Austen and Stansfield at temperatures corresponding to the point C, D, Gand H are due to definite chemical transformations in which one solid is decomposed and another is formed. Chills taken immediately above and below these critical temperatures reveal the nature of each change most clearly. The transformations at C, D, and especially at H, are very slow and do not become complete unless the temperature is maintained constant for hours or days at a point slightly below the transformation tem- perature, but all these changes can be made to agree exactly with re 411 Constitution of the Copper-Tin Series of Alloys. 1903.] Yaa T Inog Ear Ne NIL dO SLNDOUAd DINOLY g OV 412 Prof. A. Agassiz. On the Formation of [Feb 7, theory if time is allowed for them. The change at the G temperature is the breaking up of a solid solution into a mixture of the compound Cu;Sn and liquid, and is instantaneous: here we have a case of a solid partially melting as it cools. | The curve IXE’f forms with the part of the solidus immediately above it an area, roughly triangular, within which all the alloys appear to be uniform solid solutions, but, as soon as an alloy cools to the curve, it becomes saturated and a new body crystallises out of the solid solution. One branch of the curve XE/f corresponds to the crystal lisation of a body rich in copper, the other to the crystallisation of a body rich in tin, which is probably the pure compound Cu,Sn. The angle X (or rather C’), is the eutectic angle at which both bodies cr sacallse together, the whole nhcapienen being exactly like crystallisation out of a liquid. All the results obtained from the study of the chilled alloys are in harmony with the pyrometric work of Roberts-Austen and Stansfield, and many of the changes we have examined correspond to an evolution of heat recorded by them. | The paper is an extension of a short paper published by us in the ‘Proceedings’ of December, 1901. “On the Formation of Barrier Reefs and of the Different Types of Atolls.” By ALEXANDER AcassIZ, For. Mem. R.S. Received February 7,—Read March 19, 1908. The results here presented are bocca upon abaeies ations carried on during the past 25 years in Florida, the Bermudas, Bahamas, Cuba, Jamaica, ana the West Indies in the Atlantic. They include in the Pacific the Galapagos, the Hawaiian Islands, the Great Barrier Reef of Australia, the Fiji Islands, and the Coral Reefs and Islands of the tropical Pacific, from the Marquesas to the Paumotus, the Society Islands, the Cook Archipelago, Niue, the Tonga, Ellice, Gilbert, and Marshall Islands, the Carolines and Southern Ladrones, and the Maldives, in the Indian Ocean. Recognising that Darwin’s theory did not explain the conditions observed, my reports were limited to descriptions of the different types of Coral reefs and of the causes to which they probably owed their formation, and no attempt was mace to establish any independent general theory. Beginning with the Barrier ee we find that those of Fiji, the Hawaiian Islands, and the West Indies usually flank volcanic islands and are underlaid by volcanic rocks. Those of New Caledonia, 1905.] Barrier Reefs and of the Different Types of Atolls. 415 Australia, Florida, Honduras, and the Bahamas, are underlaid by outliers of the adjoining land masses, which crop out as islands and islets on the very outer edge of the Barrier Reefs. Some of the Barrier Reefs of the Society Islands, of Fiji, and of the Carolines, show that the wide and deep lagoons, separating them from the land mass, have been formed by erosion, from a broad fringing reef flat. Encircling reefs, such as characterise especially the Society Islands, hold to their central island or islands the same relation which a Barrier Reef holds to the adjoining land mass. Denudation and submarine erosion fully account for the formation of platforms upon which coral reef and other limestone organisms may build, either barrier or encircling reefs, or even atolls, rising upon a volcanic base, of which the central mass may have disappeared, as in Fiji, the Society and Caroline Islands. We may next take the type of elevated islands of the Paumotus, the Fiji, the Gilbert, and the Ladrones, many composed only of tertiary limestones, others partly of limestone, and partly volcanic. We can follow the changes from an elevated island, like Niue, or Makatea in the Paumotus, to an island like Niau, through a stage like Rangiroa to that of the great majority of the atolls in the Paumotus. The reef-flats and outer reefs flanking elevated islands, hold peculiar relation to them, they are partly those of Barrier Reef and partly of Fringing Reef. We may also trace the passage of elevated plateaux like Tonga, Guam, and islands in Fiji, partly voleanic and partly lime- stone, to atolls where only a small islet or a larger island of either limestone or volcanic rock is left to indicate its origin. Atolls may also be formed upon the denuded rim of a volcanic crater, as at Totoya and Thombia in Fiji, as well as in some of the volcanoes east of Tonga. In the Ellice and Marshall group and the Line Islands, are a number of atolls, the underlying base of which is not known, and where we can only follow the formation of the land rim of the atoll, as far as it is due to the agency of the trades or of the monsoons in constantly shift- ing the superficial material (prepared by boring organisms) which goes to form its rim. Many of the atolls in the Pacific are merely shallow sinks, formed by high sandbanks, thrown up around a central area. Throughout the Pacific, the Indian Ocean, and the West Indies the most positive evidence exists of a moderate, recent elevation of the coral reefs. This is shown by the bosses, pinnacles, and undermined masses of modern or tertiary limestone left to attest it. The existence of honeycombed pinnacles of limestone within the lagoons of atolls, as shoals, islands, or islets, shows the extent of the solvent action of the sea upon land areas, having formerly a greater extension than at the present day. Signs of this solvent action are to be seen every- where among coral reefs. Atmospheric denudation has played an VOL. LXXI. 2 H 414 Formation of Barrier Reefs and Atolls. [eber, important part in reducing elevated limestone islands to the level of the sea, by riddling them with caverns and by forming extensive sinks, often taken to be elevated lagoons. Closed atolls can hardly be said to exist; Niau in the Paumotus is the nearest approach to one, yet its shallow lagoon is fed by the sea through its porous ring. Sea water may pass freely into a lagoon at low tide over extensive shallow reef flats where there are no boat passages. The land area of an atoll is relatively small compared to that of the half-submerged reef flats. This is specially the case in the Marshall Islands and the Maidives where land areas are reduced to a minimum. The Maldivian plateau with its thousands of small atolls, rings, or lagoon reefs, rising from a depth varying from 20 to 30, fathoms, is overwhelming testimony that atolls may rise from a plateau of suitable depth, wherever and however it may have been formed and whatever may be its geological structure. On the Yucatan plateau similar con- ditions exist regarding the formation of atolls, only on a most limited scale. The great coral reef regions are within the limits of the trades and monsoons and areas of elevation, with the exception of the Ellice and Marshall Islands and some of the Line islands. The extent of the elevation is shown by the terraces of the elevated islands of the Paumotus, Fiji, Tonga, Ladrones, Gilbert, and West Indies, or by the lines of cliff caverns indicating levels of marine erosion. In the regions I have examined the modern reef rock is of very moderate thickness, within the limits of depth at which reef builders begin to grow, and within which the land rims of atolls or of Barrier Reefs are affected by mechanical causes. This does not affect the existence of solitary deep-sea corals, of extensive growths of Oculina or Lophohelia at great depths, or in any way challenge the formation of thick beds of coralliferous limestone during periods of subsidence. The Marquesas, Galapagos, and a few islands in the Society and West Indies have no corals, although they are within the limits of coral areas. Their absence is due to the steepness of their shores and to the absence or crumbling nature of their submarine platforms. Coral reefs also cannot grow off the steep cliff faces of elevated, coralliferous limestone islands. Corals take their fullest development on the sea faces of reefs ; they grow sparingly in lagoons where coralline alge grow most luxuriantly. Nullipores and corallines form an important part of the reef-building material. 1903. ] Hlectrical Conductivity of a Vacwwin. 415 “The Electrical Conductivity imparted to a Vacuum by Hot Conductors.” By O. W. Ricuarpson, B.A., B.Sc., Fellow of Trinity College, Cambridge. Communicated by Professor J. J. THomson, F.R.S. Received February 28,—Read March ae, 1903. (Abstract.) The experimental part of this paper is an Investigation of the electrical conductivity of the space surrounding hot surfaces of plati- num, carbon, and sodium, at low pressures. In addition, the first portion of the paper is occupied in deducing a theory by which the experimental results are explained. Some of the results that have been obtained with platinum were described in a preliminary note read before the Cambridge Philosophical Society on November 25, 1901. The present paper is subdivided as follows :— A. Theoretical investigation. 1. Calculation of the saturation current. 2. Equilibrium of corpuscles near a hot conducting plane of infinite area, 5. Experimental investigation. 1. Experiments with platinum. 2. Experiments with carbon. 3. Experiments with sodium. . Conclusion. The experiments show that the negative leak from hot wires at low pressures is a definite function of the temperature of the wire, and increases very rapidly as the temperature is raised. Professor McClel- land* had previously found that this leak was independent of the pressure at pressures less than 0°04 mm., whilst Professor J. J. Thom- sony had shown in addition that the current was carried by corpuscles or electrons. The theory here put forward to explain these facts, and those to be described later, is based on the corpuscular theory of conduction in metals. On that view a metal contains a very great number of free corpuscles whose mean free path is comparable with that of a molecule in air at atmospheric pressure. The corpuscles must, there- fore, be moving with a distribution of velocity given by the Boltz- mann-Maxwell law. Since the corpuscles do not escape from the metal at ordinary temperatures, it is evident that there must be a discontinuity of potential at the surface which prevents their escape. * “Camb. Phil. Soc. Proc.,’ vol. 10, p. 241, and vol. 11, p. 296. ft ‘Phil. Mag.,’ vol. 48, p. 547, py 8 ye 416 Mr. O. W.. Richardson. Electrical Conductinity [Feb. 28, On raising the temperature of the metal, the energy of the cor- puscles is increased, and at high enough-temperatures some of them will be able to shoot through the surface into the surrounding space. The number of corpuscles which escape at any temperature has been calculated on this view. The saturation current, which corre- sponds to the number emitted per second, is given by the equation C= 8 yer" 2am where | n is the number of free corpuscles in 1 ¢.c. of the metal, « the charge on an ion, S the area of the hot-metal surface, R the gas constant for a single corpuscle, whose mass is 7, 6 the absolute temperature, and ® the work done by an ion in passing through the surface layer. The rate at which energy is emitted, due to this cause, is also caleulated. Owing to the important part which the ionisation from hot bodies has played in certain recent cosmological theories, the equations which determine the equilibrium of corpuscles near a plane surface of hot metal of infinite extent are also given and solved. The chief problem which has been attacked experimentally is the way in which the saturation current from a negatively-charged hot metal surface to a neighbouring electrode varies with the tempera- ture of the metal. In the case of platinum, the hot metal consisted of a fine wire spiral passing along the axis of a surrounding cylindrical electrode. The temperature was obtained from the resistance of the wire. In the case of carbon, the leak was measured from a small lamp filament to a surrounding cylinder. The temperature was estimated in two ways: (1) by fastening a platinum and platino-iridium thermo- couple of very fine wire round the filament, and (2) from the resistance of the filament. With sodium this method could not be adopted. The metal was distributed on the inner surface of a steel cylinder, and the current from it to an insulated wire inside the cylinder was measured. The temperature was obtained by a thermo-couple of copper and nickel. Owing, doubtless, to the peculiar shape of the electrodes and the some- what high pressure of the gas, the current with sodium was never saturated. For this reason the current under a given voltage was. measured instead of the saturation current. Incidentally, it was found necessary to examine the relation between the current and the applied electromotive force. Current H.M.F. curves are given for all three substances, and, in the case of carbon,. 1903. | emparted to a Vacuwm by Hot Conductors. 417 for a considerable range of pressures. ‘To account for the results at the higher pressures, it is necessary to assume that ions are produced by collisions. The variation of current with temperature is examined over the following range :— For platinum from 10~!° to 10~° ampére per sq. cm. of surface. For carbon ee el Oe Ne x5 Me Hi For sodium LO 2 Oss ampeKey total cumremb, The corresponding ranges of temperature for platinum and sodium are roughly from 1000°C. to 1600° C., and from 100° C. to 450° C. re- spectively. ‘The small currents from sodium were measured with a quadrant electrometer, but as a general rule, a sensitive D’Arsonval galvanometer with suitable shunts was used. Perhaps the most striking result of the investigation is the rela- tively enormous currents which have been obtained. The biggest leak measured was 0:4 ampere from a carbon filament to an electrode placed near it; this corresponded to a current of 2 amperes per sq. cm. of the carbon surface, the potential on the filament being — 60 volts. The pressure in this experiment was only ,i,th mm. This current and some of the largest currents from sodium were registered on a Weston ammeter. In all these experiments the potentials employed were too small to maintain a discharge between the electrodes. Throughout the range given above, the relation between the saturation current and the temperature was found to be represented by a formula of the type Op a where A and 0 are constants for each conductor. The values which have been found for these constants are— oma lacimiana yA — 102° 4°93 x10, oir Garcon, A == WO = fee lO Ber s< MOR: les 9S AKOe Horcodmmm A= 0 G3: 16% 10 The value of A varies very rapidly with the value found for J, so that only its order of magnitude is given. The number x of free corpuscles in a c.c. of the metal is calculated from A. For platinum this gives » = 1071, whereas Professor Patterson* found 1022. In the case of the other conductors, the number found is absurdly great compared with Patterson’s values. * © Phil. Mag.’ [6], vol. 3, p. 655. ee oa — - ——— Se eg 418 Electrical Conductwity of a Vacuuin. [Feb: 287 The discrepancy can be made to disappear by assuming a small tem- perature variation of ), This assumption is shown to be consistent with the general nature of 0. The work required to drive an ion through the surface layer is calculated, in each case, from the value of 0, to which it is propor- tional. Dividing by the charge on an ion this yields the discontinuity of potential at the surface of the conductor. The values found for this are: for sodium, 6¢ = 2°45 volts, for platinum 6¢ = 4:1 volts, and for carbon 6¢ = 6:1 volts. These numbers are inversely propor- tional to the cube roots of the respective atomic volumes. This leads to the conclusion that the work required to force a corpuscle out of a. metal varies, approximately at any rate, inversely as the cube root of the atomic volume of the metal. In all these experiments, the current when the hot wire is charged positively is small compared with that obtained with the metal negatively charged. Only in the case of sodium was the positive: current large enough to deflect a sensitive galvanometer. The results which have been obtained are shown to furnish a com- plete explanation of the phenomenon known as the Edison effect. The fact that such enormous currents are obtained at such very low pressures confirms the conclusion that the ions are not produced from the gas by its interaction with the metal. Calculation shows that to obtain the currents registered with carbon, each gas molecule would have to give rise to 25 ions each time it collided with the hot metal surface. The energy lost owing to the escape of the corpuscles is compared with the energy emitted in the form of ordinary electromagnetic radiation. The former is shown to be smaller than the latter at the temperatures at which measurements have been made, but it increases more rapidly with the temperature. 1903.] New Series of Lines wn the Spectrum of Magnesiwin. 419 “On a New Series of Lines in the Spectrum of Magnesium.” By A. Fowuer, A.R.C.Se, F.R.AS., Assistant Professor of Physies, Royal College of Science, South Kensington. Com- municated by H. L. CaALLENDAR, F.R.S. Received March 9,— Read March 26, 1903. Although the spectrum of magnesium has been the subject of many investigations, certain lines which occur in the are spectrum appear to have hitherto escaped notice. The lines in question are comparatively feeble, but on account of their theoretical interest it seems desirable to draw attention to them. The new lines make their appearance in the spectrum when the arc is made to pass between poles consisting of magnesium rods, but they do not always appear with equal intensity. They are somewhat nebulous, especially on their less refrangible sides, so that their positions cannot be determined with great accuracy ; but as nearly as they can be ascertained with the instruments at my disposal, the wave-lengths are (in air), 4511:4, 4251-0, 4106°8, and 4018°3. A mere inspection of the photographs suggests that these lines constitute a regular series, associated with the much stronger series described by Rydberg*, having wave-lengths 5528-75, 470333, 4352°18, 4167-81, 4058°45, and 3987-08, according to the measures of Kayser and Runge. This view seems to be sutticiently confirraed by calculation. Rydberg found that neither his own general formula nor that of Kayser and Runge could be applied with sufficient accuracy to the stronger series, and employed a combination of the two formule, namely, Zeal b tue C i (m+)? (+p)? where 7 is the oscillation frequency, m has the values 3, 4,5, &c., and a, 0, c, and » are constants to be determined from four lines belonging to the series. For the magnesium series, the equation calculated by Rydberg is v= 111,856-92 | 147,764-05 a) COU Vi Ll ae rae ’ Bree me isan Goda: Gm <0406). n being the oscillation frequency am «zr, and m having the values 3, 4, 5, 6, 7, 8 for the six lines named above. Using the same formula for the new series,and calculating the con- stants from the four lines, the equation for frequencies 7” vacuo is _102,496°6 fi 168,840°5 (m+0°618)2 © (m+0°618)* * «Ofversigt af Kongl. Vet. Akad. Forhandl.,’ 1893, Stockholm. m = 26,595°4 — 420 New Serres of Lines in the Spectrum of Magnesium. [Mayr. 9, Another formula* which may be conveniently employed is C * Gin +p)? Sing § This formula gives for the two magnesium series the equations : i : 107,071°37 66 ; 1] Oo , rs ih = Sy) : = 2 p) Rydberg ” series, 1 6,601-49 (i + 11-2304)? 4 9°13982 100,033°6 Tf “ hav. Aig a nee ~ (iv + 0°495)? + 2°38919” n being the oscillation frequency zn vacuo in each case. Tt will be seen that the convergence frequency of the new series is as nearly equal to that of the Rydberg series as can be expected with the comparatively rough wave-lengths employed, and it may be added that in each case the constant mp is of unusual magnitude. These facts, in conjunction with the general characters and relative intensities of the lines, render it highly probable that the new series is associated with the Rydberg series as second and first subordinate series respectively. Applying the formula to the calculation of the members of the new series for which m=3 and m=2, the corresponding wave-lengths in air are 5065-0 and 6674:5. The first is probably represented by a line having an approximate wave-length 5067, which is not so readily observed in the photographs as the others, because the plates employed are comparatively slow for this part of the spectrum, and if the expo- sure be lengthened, the banded spectrum of magnesium becomes strong enough to almost mask the line. The line 6674:5 is perhaps too far in the red to be conveniently observed, seeing that it is probably feeble and not well defined. It may be reasonably concluded that the arc spectrum of magnesium includes two subordinate series of single lines in addition to the two well-known subordinate series of triplets. No such combination of series appears to have been previously noted in the spectrum of a metal, but two sets of series, each set consisting of a principal and two subordinate series, are well-known in the spectra of helium and oxygen. The author desires to express his obligations to Mr. Herbert Shaw for valuable assistance in making the computations involved in investigating the various formule which have been suggested for series, as applied to the series which forms the subject of the present paper. * After much labour, this formula was arrived at by Mr. Herbert Shaw and the author as the one giving the most consistent results for series in general, but it was afterwards found that a similar formula, expressed in wave-lengths, had already been employed by Mr. Rummel (‘ Roy. Sce. Victoria Proc.,’ vol. 10, Part I, 1897, p. 75). 1905.] Relative Ainounts of Krypton and Xenon vir Arr. 42] “ An Attempt to Estimate the Relative Amounts of Krypton and of Xenon in Atmospheric Air.” By Sir WituiAmM Ramsay, K.C.B., F.R.S. Received March 9,—Read March 26, 1903. When Dr. Travers and I isolated krypton and xenon from air, we had very little idea of the total amount of liquid air from which, by its evaporation, these gases had been obtained. And we were then more concerned with the isolation of the gases in a pure state than in the determination of the proportion in which they exist in the atmo- sphere. Our knowledge of the composition of the air, however, is not complete until the total yield of krypton, xenon, neon, and helium has been determined. An estimation of the two last is being undertaken by Dr. Travers. In our experiments on these gases* we did not measure the total quautity of air evaporated. We used liquid air for various purposes, and for some months we collected the dregs, allowing them to evaporate into a large gas-holder. We guessed (but it was only the roughest estimate) that we had accumulated in this manner the residues from about thirty litres of liquid air; and on this assumption we thought the following estimates not improbable :—Helium, 1—2 parts per million of gaseous air; neon, 1 per 100,000; krypton, 1 per million ; xenon, | per 20,000,000. But they rested on a very insecure founda- tion of fact. The first preliminary experiment was made to ascertain how much of the air which passes through the Hampson liquefier is converted into liquid. The results, however, were inaccurate, and I would rather cite the conclusions given by later experiments on a much larger scale. The compressor was run for several hours every morning and afternoon during six and a half days; the liquefied air was weighed after each run ; and the escaping air passed through a large gas-meter, where its volume was registered. The air escaping had a somewhat lower density than ordinary air, owing to the partial removal of oxygen and argon; but the experiments were not sufficiently accurate to make it worth while to take this into account. The volumes given are, however, corrected for alterations of pressure and temperature. In all, 179-7 kilos. of gaseous air passed the meter, and 10°8 kilos. of liquid air were collected. During the collection, about 6 per cent. of the liquid air evaporates ; adding this, 11°4 kilos. must have been the total weight of air liquefied. The total weight of air taken in by the compressor, therefore, and delivered to the liquefier was conse- quently 179°7 + 11:4 = 191-1 kilos, and the percentage liquefied a little under 6 per cent. The number may be taken without sensible error as 6 per cent. liquefaction. * “ Argon and its Companions,” ‘Phil. Trans.,’ A, vol. 197, pp. 47—89. 422 Sir W. Ramsay. Lstimate of the Relative [ Mar. 9, During these runs the pressure was kept at 190 atmospheres ; and the delivery of air through the escape valve of the liquefier remained fairly uniform. Separate readings, taken at different times, of the amount of air passing the meter gave 0°244, 0230, 0:258, 0-261, 0-248, and 0°239 kilos. per minute; these figures testify to fairly uniform working. It was of interest to see whether argon concentrated itself in the liquefied portion of the air, or whether most of it escaped as gas. The boiling pomts of the principal atmospheric gases are, on the - absolute scale :—nitrogen, 77°54; argon, 86°°90; oxygen, 90°°5 ; the differences are 9°°36 and 3°:6 respectively ; it was therefore to be expected that the argon would concentrate in the liquid. An experiment was therefore made in which 12°73 grammes of freshly collected lquid air was allowed to boil in a double-walled vacuum-tube, and the gases were led directly over metallic copper and magnesium lime; the resulting argon measured 165 c.c. at 0° and 760 mm. pressure. The weight of this argon was 0:2943 gramme, or from 1 gramme of liquid air, 0°02312 gramme. Now 1 gramme of atmospheric air contains 0°0129 gramme of argon; hence the process of lquefaction nearly doubles the content of argon in the air. It is, therefore, very advantageous to prepare argon from air which has been liquefied. Acting on this suggestion, it appeared w fortior: probable that if the percentage of argon in air were doubled by liquefaction, the krypton and the xenon would be practically wholly removed and liquefied. And by submitting liquefied air to a second hquefaction, by boiling it off through the compressor, it appeared to hold that a concentration of the krypton and xenon would result, and that they would be found wholly in the liquefied portion. An experiment was therefore begun in which about 100 kilos. of gaseous air were passed through the liquefier ; the liquefied portion, amounting to about 6 kilos., was again passed through the liquefier, somewhat added to by gaseous air, drawn in by the compressor. By an unfortunate accident, however, nine- tenths of the air collected during the second liquefaction was lost ; and the amount of xenon and krypton in the remaining tenth, repre- senting about 10 kilos. of gaseous air, did not appear to justify repetition of the troublesome experiment. I have no doubt, however, that had these experiments not had for their object the determination of quantity, but only the preparation of krypton and xenon, they would have effected the separation well. The liquid air, resulting from the liquefaction of 6 per cent. of 191-1 kilos. of gaseous air, was sucked, several litres at a time, into a large glass balloon of about 5 litres capacity, fitted with an india- rubber cork, through which a wide tube passed, connected with a doubie-acting Fleuss pump, driven by an electric motor. Through 1903.] Amounts of Krypton and Xenon in Air. 423 another hole in the cork there passed a siphon which could be closed by means of a brass stop-cock ; this tube served to admit liquid air to the balloon. A manometer was also connected with the interior of the flask so as to register the pressure under which the liquid air was boiling. The air boiled at a pressure of about 250 mm., corresponding to a temperature of about —195°. The boiling was quite quiet, without bumping; it was sometimes necessary to warm the balloon gently in order to accelerate the evaporation. The object of distilling at a low temperature was to lower the vapour-pressure of the krypton and xenon in the liquid air, and so to lessen, or in great part to prevent their evaporation. The total liquid air was thus reduced to about 200 c.c. The balloon containing this air was coupled with a large iron tube, holding about 20 kilos. of reduced copper heated to bright redness. The liquid air residue, naturally, consisted largely of oxygen, for the more volatile nitrogen had in great part evaporated. over the copper the volume of gas was about 50 litres. It may be contended that during the evaporation of the air, even at — 195°,a large portion of the krypton and xenon may pass away as gas. It is not possible to estimate the amount lost in this manner ;. but at — 195°, the vapour pressure of krypton is 2°8 mm., and that of xenon 002mm. These figures have been arrived atin the following man- ner. Relying on the vapour pressures of mercury by Ramsay and Young,,. given in the Trans. Chem. Soc., 1886, p. 50, and on the measurements made by Ramsay and Travers* of the vapour pressures of krypton and xenon, ratios were found between the absolute temperatures of mercury on the one hand, and of krypton and xenon respectively on the other, between the pressures 300 and 3000 mm., with two addi- tional data—the temperatures of krypton corresponding to pressures of 9 and of 174mm. They are as follows :— After passing | Temp. of Meni p aon | ALeea ps Oi) | mercury. krypton. : xenon | 5 pe aes Dae Desrecs Ratios. | Hewes | Ratios. | absolute. absolute. | absolute. | | mm. i 4 | s 9-0 454.8 94°2 O218d5i | — | — 300 582 °2 110-4 Ops 9am 14898 nO e220 400 596 °4 113 °8 0°1909 | LSS Zee a On2a69 |} 500 609 °O IL) 1. USO) 4 156 °8 0 -2575 | 600 617 °9 118 *35 OMGIS = | Used 072583 k ©2700 | 626 °5 120-2 0°1918 162 °0 0-256 760 | 631 °2 121°3 0 °1922 163 °9 0 2597 3000 | 726°8 142 -2 0°1957 192 -4 0° 2647 SS ORihiveirats: oA vol VOT wOOL, p72. A24 Sir W. Ramnsay. Lstimate of the Relative [ Mar. 9, These ratios were mapped against the absolute temperatures of mercury, and were found, as usual, to give straight lines. The lines were extrapolated to lower temperatures, and the vapour pressures required were calculated from the extrapolated curves. The portion of interest is given in the following table :— Temp. C. Vap. press. of krypton. Vap. press. of xenon. — 205° 0-27 mm. 0:0005 mm. ~ 200 0-7 1. 00m eae ~ 195 DB. 0-02 44 - 190 eB O04 E365 uy Dos 0. ae -- 182-4 rf cA bake (OE lyf 35 The last two data for krypton are the results of direct measure- ment. It may be mentioned here that the melting point of krypton is about — 169°, and that of xenon — 140°; and the boiling points at atmospheric pressure are, krypton — 151°:7, and xenon — 109°:1. The nitrogen was removed from the 50 litres of gas by passing it over red-hot magnesium-lime mixture. The resulting crude argon measured 12°5 litres at 16° and 770 mm. ; its weight is calculated from its known density as 21:3 grammes. This argon was iguanas 7 in a bulb immersed in liquid air, boiling under reduced pressure, so as to reduce the vapour pressure of the krypton and xenon; and the major part was re-transferred through a Topler pump to the gas-holder from which it had passed to the liquefying bulb. About 1500 c¢.c. of the last portions to distil away were collected in five mercury gas-holders, each of a capacity of 300 ¢.c. The argon was now methodically fractionated according to the accompanying scheme. The 1200 ¢.c., numbered (1) was, as described, distributed in five gas-holders. ‘The contents of the first-——the one first filled is termed A —were liquefied, and half the amount replaced in gas-holder A. The contents of B were liquefied, and A was filled, by allowing the liquid argon to evaporate under reduced pressure. ‘The contents of C were liquefied along with what remained of B; and B was filled in like manner. D was liquefied, and C filled; and, finally, E, and D filled. The residue in the liquefying bulb, which evaporated very slowly after the argon had boiled away, was removed through the pump, and collected in a tube over mercury. The contents of A, B, C, D, and E are labelled (2), (3), (4), (5), and (6). The process was repeated and with this explanation the scheme can be understood. 1905. | Amounts of Krypton and Xenon vir Arr. 425. Residue of 1500 ¢.c. :— A -2— 7—12—17—, —22—27—32 — —37 Rejected. B 3% sg Rejected a Rejected —38 Trace Kr. (1) c a wae ane 9499“ 34, _—39 Strong Kr. D Be” s0 — 2035 Had teal Strong Kr. E 621 —_-—- ---— epee —41 Kr. and Xe. Residue. Residue. Residue. No residue. No. 41 was mixed with the residues, which contained much xenon, along with krypton. Nos. 38, 39, and 40 were mixed, and the frac- tionation continued in a small apparatus. The gases were now sparked with oxygen over soda, so as to remove traces of air; for the operations had now to be conducted with great care, ant the spectra of the two samples showed traces of nitrogen. After with- drawal of oxygen from both sets of gas, by means of phosphorus, the krypton mixture was fractionated into three portions; 42, containing much argon; 43, rich in krypton; and 44, a non- volatile residue. The xenon residues, 41, were also solidified with liquid air, and placed for a few seconds in connection with the vacuum of the Tépler pump. The gas which was pumped off was added to 43, and the residue to 44. The bulb was now left for a quarter of an hour, so that equilibrium might be restored, and the stop-cock opened a second time. A bubble or two was removed with the pump. On making communication with the pump a third time, no gas escaped ; and this is not remarkable, for the vapour pressure of xenon is little over 0°1 mm. at that temperature. It was therefore assumed that the final residue was pure xenon; and, indeed, its spectrum showed no trace of the krypton lines. And it may also be taken for granted that very little xenon was added to No. 45. The densities of Nos. 42 and 43 were determined; and from the known densities of argon and krypton, their relative proportions were calculated. The density of 42 was found to be 21°31, corresponding to the per- centage composition, argon, 93°5 per cent. , krypton, 6°D per cent. ; the density of 43 was 39°43, implying a mixture of argon, 6°6 per cent., krypton, 93-4 per cent. The ee of No. 42, reduced to 0° mad 760 was 22-0 ¢c.c.; it therefore contained 1°43 c.c. of krypton; that of No. 43 was 6°5 c.c., and it contained 6-1 c.c. of krypton. The lod total volume of krypton was therefore 7°5 c.., and its weight 0°0028 gramme; the volume of the xenon, reduced to 0° and 760 was 0°87 c.c., and its calculated weight, 0°0005 gramme. These results are reproduced in the following tabular statement :—— Air passed through liquefier............... 191-1 kilogrammes. 6 UR ee ee 11-3 Argon obtained (11°8 litres at 0° & 760) 21:3 grammes. 0 ) . = 5-91 p.ec. Percent. Ol gaseous airs! 22.20. 0: 118 ; of liquid air, 0°1885. 426 Relative Amounts of Krypton and Xenon in Air. [Mar. 9, Motaltkny pton obtamede 5. -sete ass ee 0°0028 gramme. oy) xenon Jol tame cll). Wem cae 0:0005 ie Percentage krypton in gaseous air ...... 0-000014 by weight. aH xenon He ly CNA 0:0000026 & Krypton equal to 1 part by weight in about 7 millions of air; by volume, 1 part in 20 millions. Xenon equal to 1 part by weight in about 40 millions of air; by volume, 1 part in 170 millions. As before remarked, it is not maintained that all the krypton and all the xenon have been separated; it is likely, however, that the separation of the xenon was more perfect than that of the krypton. The results are merely brought forward as the result of a careful experiment to quantitatively isolate these gases. I have to express my cordial thanks to Mr. E. C. C. Baly and to Mr. Inglis for aid in carrying out part of these operations. As a quantity of pure krypton, sufficient for determination of density, had been collected, occasion was taken to redetermine the density of that gas. It was submitted to careful fractionation; a considerable portion was rejected as possibly containing argon, and the «dregs were set aside as possibly having contained xenon. The sub- stance weighed had a low vapour pressure,—about 15 mm. at the temperature of the liquid air used in fractionating. The separation of the lighter and heavier portions was repeated four times, the density having been determined on each occasion, with only small differences. Finally, a very careful determination of density was carried out, with the following results :— Volume’ of density-bullb) gas cee 7268 Ge Tem perait ure ilinin ticles. Vln a Ae eee Ld oe: Eressure.oneas, scone cue dasa sare eraereree 754°0 mm NAC ed OMe ieee RPO MEE Cn SEEDERS science yao bn 0-02488 gramme. Hence density, compared with O = 16... 40°81. Previous determinations with two samples of gas, one fractionated from argon, the other fractionated from xenon, gave 40°82 and 40°73 respectively as the density. The result given above is in perfect concordance with these figures. ‘The chief cause of error is in the weight ; I think it would be fair to regard two units in the fifth place as the limit of error, which gives a possible divergence of about 1 part in 1200. The atomic weight of krypton would accordingly be 81°62; the mean of former determinations is 81°28. This is in accordance with its position in the periodic table, which lies between bromine, 80, and rubidium, 85. 1903. ] Some Physical Properties of Niche Carbonyl. 427 “Some Physical Properties of Nickel Carbonyl.” By Jams Dewar, M.A., Se.D., LL.D., F.B.S., Jacksonian Professor in the University of Cambridge, and HUMPHREY OWEN JONES, M.A., Fellow of Clare College, Jacksonian Demonstrator in the University of Cambridge. Received March 3,—Read March 26, 1908. The properties of nickel carbonyl have until recently been the subject of but few investigations. Dr. Mond and his collaborators in the discovery of this remarkable substance determined some of its physical properties, including its boiling point, specific gravity, and vapour density. Subsequently Dr. Mond, in association with Pro- fessor Nasini, made observations on its molecular refraction and thermal expansion. A substance of the peculiar molecular structure of nickel carbonyl seemed to call for further study. The investigation described in the present paper was carried out in the winter of 1901; the authors’ in- tention being to make a complete study of the stability of the com- pound both in the gaseous and liquid conditions. While the work was in progressa paper by Mittasch was published* containing an account of an admirable and exhaustive investigation of the velocity of the reaction between nickel and carbon monoxide, including the heat of formation and vapour tensions of the compound, covering part of the ground which we had examined. A number of interesting problems examined in the course of our inquiry however remain, which have not been touched upon by previous investigators, and to some of these the present paper is devoted. The vapour density of nickel carbonyl was determined by Mond, Langer and Quincke7j in air, at 50° C. by Victor Meyer’s method. ‘The value obtained was 86-7, while theory requires 85. The vapour density at this temperature is quite normal, and there is no evidence of association even at this temperature only some 7° ©. higher than the boiling point of the compound. It was found that the vapour exploded at 60° C. with a flash of light, and carbon dioxide was detected among the products of decomposition. Berthelot explained the explosion as being due to the production of carbon dioxide by the intermolecular reaction :— 2CO=CO.2+C, which was observed to take place when carbon monoxide acts on nickel at 350°—450° C. Later§ it was found that the action only proceeded in this way to a small extent, when * © Zeit. Phys. Chem.,’ 1902, vol. 40, p. 1. + ‘Jour. Chem. Soce.,’ 1890, vol. 57, p. 749. + ‘Compt. Rend.,’ vol. 112, p. 1343. § ‘Ann, Chim. Phys., 1901 [7], vol. 22, p. 204. 428 Prof. J. Dewar and Mr. H. O. Jones. [ Mar. 3, nickel carbonyl vapour was decomposed. ‘The reaction 2CO=CO,+C liberates 38°8 kilogramme calories, so that if the heat of formation of nickel carbonyl is less than 77:6 kilogramme calories per molecule, this explanation would be valid. This condition is satisfied since Mittasch* has shown that the heat of formation of nickel carbonyl is between 50°6 and 55°6 cals. Hence the detonation observed by Mond might result from the reaction : Ni(CO), = Ni+2CQ,+4 2C. The fact that nickel carbonyl was thus reported to be explosive, together with the explanation of its explosibility offered by Berthelot, strengthened the belief in its great instability and deterred experi- menters from working with it, in any case at temperatures above its. boiling point. The authors observed that when nickel carbonyl was suddenly heated in an atmosphere of some inert gas, such as hydrogen or nitrogen, the vapour decomposed quietly with deposition of metallic nickel; there was no explosion or flash of light, and the quantity of carbon dioxide produced was so small, in most cases, as to be almost negligible. This was found to be the case even when the temperature used was as high as 130° C. Mittasch* states that he could detect no carbon dioxide when the compound decomposed below 100° C. The explosion “observed by Mond, Langer, and Quincke must therefore have been due to the presence of oxygen, and doses not occur inits absence. The amount of carbon dioxide produced by the quiet decomposition is very small, so that the Berthelot reaction, though possible, only takes place to a very slight extent. It was therefore clear that vapour-density determinations of the compound could be made in atmospheres of inert gases at temperatures much higher than 50° C. As vapour-density determinations of such a unique compound as nickel carbonyl under varying conditions would. have a special interest and might well be expected to repay careful study, a large number of vapour-density determinations were made by Victor Meyer’s method at temperatures between 63° C. and the boiling point of naphthalene (216° C.), in order to ascertain the effect of increasing temperature on the dissociation of the compound. The vapour-density apparatus was filled in different experiments with various dry inert gases, viz., hydrogen, nitrogen and ethylene, all care- fully purified and especially freed from oxygen. In order to trace the effect of the rapidity of gaseous admixture on the dissociation various forms of the vapour-density reservoir were employed. An atmosphere of carbon monoxide was employed to * Loc. cit. t ‘ ; 1903. ] Some Physical Properties of Nickel Carbonyl. 429 investigate the effect of the gaseous product of dissociation on the stability of the carbonyl. It was found that the rate of admixture of the vapour with the gases had a marked effect on the dissociation, as shown by the differ- ence in the vapour densities, when taken under similar conditions, in the various inert gases; and, further, that the presence of carbon monoxide produced the expected diminution in the amount of the compound decomposed. In order to further confirm this, two reser- voir tubes of different bore were used, one having a cross-sectional area about three times that of the other (the latter will be referred to as the narrow tube). In the latter, admixture could take place much less readily than in the former; consequently the surrounding gas would be expected to have a smaller effect on the extent of the dissociation. ‘This was also confirmed by the experimental results ; the vapour density in the narrow tube is almost independent of the gas employed. The effect of the nature of the surface on the extent of dissociation was tested by using the tubes coated internally with a film of metallic nickel deposited from the vapour of nickel carbonyl by heating. The film of nickel seemed to bring about a state of equilibrium more rapidly, so that the vapour densities determined in these tubes were lower than those in the same tubes not covered by nickel (a similar effect was observed by Mittasch). It was found that the rate at which the liquid evaporated, as would be expected in the case of a substance which readily dissociates, had some effect on the extent of the dissociation. Hence, it was necessary, in order to get comparable results, to arrange that approximately the same tite was taken for vaporisation in all the experiments made at the same temperature. A definite end-point could be observed in each case at which the gas displaced by the vapour ceased to come off, and a much slower evolution of gas took place. The experiment was stopped when the more rapid evolution of gas gave place to the slower. The results obtained are appended in the following table (Table I). In the fourth column the percentages of nickel-carbonyl molecules 85 — D 3D 100, where 85 is the theoretical vapour density of nickel carbonyl and D is the observed value. Unless otherwise stated, it is to be understood that the deter- mination was made in a Victor Meyer’s apparatus of the usual type, occasionally referred to as the wide tube. From the figures in the above table, it is seen that the value of the vapour density, deduced from the experiments in the wide tube, is greater in ethylene than in nitrogen and hydrogen. It was also VOL. LXXI, 2 dissociated are given, calculated from the formula P = SS ee ee 430 Prot. J. Dewar and Mr. H. O. Jones. [ Mar. 3,. Table I—Vapour Densities determined by Meyer’s Method. | | | fe cars| bqee eeu eee & T : vy | 5 | hempente mningthe| HF CO), Roark ulb. Gest | dis- nC: ~ ~7*| sociated. eS sees eee eee | ay E | | 63 | Nitrogen .} 83°3 | 0-7 | Very slight deposit of nickel. 78°2 | 2:7 | Slight deposit of nickel. 64 | Hydrogen| 79:2 -| 66 | Carbon 83:9 | 0°15 | Nickel covered tube. monoxide] 85°2 | 81 Nitrogen.| 71°0 | 6:2 | Slight deposit of nickel. (Benzene) | Ta As) 100 | Hydrogen} 67°3 | 8°S | Nickel deposited over several | 67:0 | inches in the tube. | 53 56:9 16°7 | Wide tube covered with nickel. BO a 9 750 2°4 | Narrow tube. Nickel deposit | | 798 extended about 32 cm. | Re | Nitrogen .| 70°8 6°7 | 70 °9 .. | Wide tube. | . 75 °6 4°1 | Narrow tube. : >) Ethylene.| 70°6 6°8 | Wide tube. Nickel deposited | | Osa over 2cm.from bottom of tube. | | Und, 4°3 | Wide tube. Moist nickel car- 76°7 bonyl. Nickel deposited irreg- | ularly over about 3 cm. 77°8 | 2:7 | Narrow tube. Nickel deposit 49 3D) a | extended 22 cm. 5 | Carbon 85°0 | 0°39 | Wide tube. No visible deposit monoxide} 83°71 | of nickel. 82°6 0°6 | Wide tube, covered with nickel 84°35 | carbonyl, dried over phos- phorus pentoxide. 76°4 | 8°87 | Narrow tube. Slight deposit of 76°0 | nickel. | 110 Nitrogen .| 48°4 | 25°4 | Distinct deposit of nickel. (Toluene) AS 1 | Carbon 75°2 | 4:4 | Deposit of nickel scarcely visible. monoxide| 74°8 | 129 | Nitrogen .| 25°] 76°5 | Extensive deposition of nickel | (Amy alcohol) 26 °4 over tube. Carbon Bib 7 54°5 sa as monoxide] 32°7 135 Carbon 26 °9 72 °0 . (Acetic monoxide | anhydride) 155 Nitrogen .| 22°5 94 °3 3 3 (Turpentine) 22°5 | | Carbon | 23°2 ste <5 Ee | monoxide! 23°16! 88°8 | | | 182 Nitrogen .| 23°38 89-0 | i 5 | | | (Aniline) 23 °0 Carbon 24. °4, 82°8 sh ‘S monoxide! 23°3 | 88:2 216 Nitrogen.| 22-4 | 93:0 ‘ F (Naphthalene) 22 44 | Carbon (2059) monoxide 21°3 g)e7/ 1903.] Some Physical Properties of Nickel Carbonyl. 431 noticed that the deposit of nickel in the latter two gases extended higher up the tube than in the former, and was higher in hydrogen than in nitrogen. In carbon monoxide, on the other hand, the vapour density is higher than in the gases hydrogen, nitrogen, or ethylene, the values at 100° C. being nearly normal, and the dissociation was incomplete even in aniline vapour. This demonstrates very clearly the effect of the presence of one of the dissociation products on the amount of the dissociation. In the narrow tube, however, the values obtained at 100° C. do not seem to depend, to any great extent, on the nature of the surrounding gas, the values in carbon monoxide and in the inert gases being almost identical, which shows the great effect of rate of admixture and diffusion on the dissociation. The amount of dissociation increases rapidly with the temperature ; in nitrogen at 155° C. it is practically complete. The rate of increase in carbon monoxide is distinctly slower, the difference between the vapour densities in nitrogen and carbon monoxide at 129° C. being quite marked. Above 155° C. the results obtamed are somewhat irregular ; but dissociation seems to be nearly complete at atmospheric pressure, since only a small deposit of nickel could be obtained when the tube was placed horizontally and a clear part heated with the blowpipe. A few vapour-density determinations were also made by Hoffmann’s method at temperatures between 17° C. and the boiling point of aniline (182° C.) in order to observe the dissociation of the undiluted vapour. Complete dissociation is practically reached at 182° C., but even then the application of a pointed flame to a clear portion of the tube produced a slight deposit of nickel, so that traces of nickel carbonyl were still present. The results are given in the appended table (Table I1). The results of the experiments in the narrow tube are given in the fifth column for the sake of comparison, the phenomena in this case being practically the dissociation of the vapour in contact with its own dissociation products. The dissociation is clearly greater under reduced pressure, as might be anticipated. The general results of the vapour-density determinations are readily seen from the curves in fig. 1 (p. 433). Having thus found that the vapour of nickel carbonyl was much more stable at elevated temperatures than had hitherto been suspected, we resolved to examine the stability of the liquid under pressure, and if possible make observations on it as far as its critical point. Small sealed tubes, from one-half to one-third full of nickel carbonyl, were heated to 200° C. without bursting. A small quantity of nickel was deposited on the first heating, and it was found that its quantity was not appreciably increased on repeating the operation. On 2 2 432 Prof. J. Dewar and Mr. H. O. Jones. [Mar. 3, standing a sufficient length of time at the ordinary temperature the nickel gradually dissolved again, although it was covered by the excess of liquid. This proved in an interesting manner the ease with which the reaction between the deposited nickel and carbon monoxide is reversible. The amount of the nickel deposition was notably greater when the liquid was heated in contact with mercury, so that the use of the Cailletet pump for the examination of the critical phenomena was not practicable. ; Rough measurements of the critical temperature were made by observing the temperature at which the meniscus disappeared and Table I1.—Vapour Densities of Nickel Carbonyl by Hofmann’s Method. Percentage dissociated at | Pressure | Percentage atmospheric pressure. Temperature | Vapour | inmm. | of nickel of bath. density. of carbonyl mercury. |dissociated. Tae | In carbon n nitrogen. . | monoxide. “On | 17 85°6 67°7 0:0 35 81°8 161 °4 0°13 (Ether) | 78 62h 650) F21758 11-9 4°1 at 100°C. | 3°87 at 100° (Alcohol) in narrow tube | in narrow tube 129 25 +2 389 °4 gga ; ie (Amylalcohol)| 23°8 365 °6 85 °7 a ant ee hae 182 21°5 | 349-0 98 “4 a (Aniline) 20°8 320 °6 S60 Practically complete in 2a 356 °0 95 °7 wide tube. reappeared, when a quantity of the liquid was alternately. heated and cooled in an exhausted and sealed off piece of glass tube of 2—3 mm. bore. The tube was heated side by side with a thermometer in a glycerine bath, and was so arranged that it could be inverted at will. The temperatures at which the meniscus disappeared and reappeared in the course of many repeated observations with different samples of material, ranged between 191° C. and 195° C. It appears from the observations made with such tubes that the most reliable results were obtained the first time they were heated, the meniscus being better defined and disappearing more sharply than on subsequent occasions. ‘The blurred effect noticed on a repetition of the experiment is due to a somewhat greater amount of carbon mon- oxide being present. 1903. ] Some Physical Properties of Nickel Carbonyl. 433 oO S ae il | | ° S eee |: | ie) x or: ro) © T — oO : l © | cS) IN | Ny iw OS / / ~~. S ai \ AK cc | {e) > 7) 4S a. = ae a o vie Sg aol TSBs i | | S 08 S | Ve Ses S ° geo ws : x Sa a 8 2 . e. 98 Were : 5 Sas ¥ = SSC io) S | ic 8 0 S 5 ql See eEo So a) KO S lee SUT me Q S) 9 9 9 re) Q Q Om aw N SANMND UOIZEIDOSSI 9 AZISUAGT The presence of carbon monoxide would be expected to cause a lowering of the temperature at which the meniscus disappeared, and this expectation was realised in the course of the experiments. Hence the temperature observed in the first experiment is probably below the 434 Prof. J. Dewar and Mr. H. O. Jones. [Mar. 3, true critical temperature. ‘The presence of 4—5 per cent. of carbon monoxide with its low critical temperature 128° abs. would be expected to lower the temperature at which the meniscus disappears by about 5°C. That this quantity of carbon monoxide might be present at any time is shown by the following experiment :— Volume of tube, 1°35 c.c. 55 nickel carbonyl left after heating to 193° C. = 0°437 c.c. Amount of nickel deposited = 0°0147 gramme. if carbon monoxide liberated = 0°0284 gramme = 5:02 per cent. of the residual nickel carbonyl. Hence it is very probable that the true critical temperature is about 200° C. Comparative experiments made with pure ethyl ether in similar tubes, gave a critical temperature of 193° C. The accepted value for the critical temperature for ether being 194°:5 C., the observed tem- perature for nickel carbonyl cannot be far removed from the correct value. , The formula 'l, = 0°66 T., where T, is the absolute boiling point and T, the absolute critical temperature, should be applicable to the case of nickel carbonyl since it is applicable to ether, a liquid which has a critical temperature of approximately the same value. Taking the boiling point to be 43°°3 C., the critical temperature calculated from the above relation would be 201°:4 C., which agrees very well with the value which was found as a probable minimum critical tem- perature. Although it had not been surmised that nickel carbonyl could ever stand heating above 100° C., nevertheless Mond and Nasini* calculated the hypothetical critical temperature of nickel carbonyl from the results of their experiments on its coefficient of expansion by means of a formula given by Thorpe and Riicker and found the value 151° C. In a similar manner Ramsay and Shields} calculated the critical temperature from the temperature coefficient of the molecular surface energy. The value deduced in this way is 182°°8 C. It therefore appears that the hypothetical critical temperature calculated by either ot these methods falls considerably below its actual value. Rough indications of the critical pressure were obtained by intro- ducing into the sealed tube containing the nickel carbonyl a small tube of very fine bore, closed at one end and having a small globule of mercury introduced at the other, to act as a manometer. ‘The posi- tion of the globule was observed at the ordinary temperature, and again at the critical point. The volume of the air in the small tube occupied about one-thirtieth of its original volume at the latter * “Zeit. Phys. Chem.,’ 1891, vol. 8, p. 150. * ‘Jour. Chem. Soc.,’ 1893, vol. 63, p. 1108. 19035. ] Some Physical Properties of Nickel Carbonyl. 435 temperature, so that the pressure in the tube seems to be rather more than 40 atmospheres at the critical point. On cooling, the indicating globule remained permanently displaced some distance up the tube, showing the existence of a pressure developed by the decomposition of the nickel carbonyl. On standing for some time the whole of the nickel disappears, and the carbonic oxide pressure disappears. The pressure on cooling seemed to be about ten atmospheres, hence the critical pressure would be about thirty atmospheres. Later it will be shown that this is near the actual value. The critical constants of the compound being known, together with the boiling point, it is possible to calculate a vapour-pressure curve. it was, however, thought better to determine the vapour pressure at a number of temperatures below the boiling point of the liquid, by the static method, and from this curve by extrapolations to deduce the values for higher temperatures. A wide barometer tube (about 0-7 cm. diameter) was carefully cleaned, dried, and drawn off to a fine capillary tube at one end. The tube was then placed upright in a vessel of pure dry mercury and exhausted thoroughly with a Fleuss pump. A small tube full of nickel carbonyl was now introduced at the bottom of the tube and the whole then exhausted again, while surrounded by a freezing mixture, in order to get rid of all adhering air, and finally sealed off rapidly at the fine capillary. By this method of procedure only a very small amount of decomposition took place during the sealing off, as indicated by the very slight deposit of nickel. The pressure was then read off by means of a kathetometer, while the tube was surrounded by a hath kept at various constant temperatures. The results are appended below, together with those obtained by Mittasch* by the dynamic method. Dewar and Jones. Mittasch. — 9 C. 94-3 mm. RB (05) (OU Barol jan, = 7 104°3 f° 11740) 5) — 2 Ae Lb) Qe 238 ° 2 0) ayy DO m2, 294°3 +10 215-0 DAT 6 349-7 +16 283°9 29 52 444-9 + 20 DA) 25) Syl ODM) Hayy 20 + 30 461-9 39 -97 647 °2 The values for — 9° C. and + 30° C. give the following Rankine formula for the relation between the pressure p in millimetres of mercury and the absolute temperature T. Log p = 7:°355 — 1415/T. At 200° C. (about the critical temperature) the pressure calculated from this equation is 30°4 atmospheres. Taking the results obtained * Loc. cile 436 Prof. J, Dewar and Mr. H. O. Jones. [ Mar. 3, by Mond and Nasini, viz., the boiling point 43°C. 751 mm. and the pressure at 20° C. 338°7 mm., a similar expression, log p = 7-281 — 1392/T, is obtained. Taking Mittasch’s pressures for 2° C. and 40° C. the Rankine obtained is :— logp = 7:781 —1555:7/T. The boiling points calculated from these formule are given below :— Nori oe re Sa ee eer Abs" 93}(Ch, INV RECON Wo wee hs see ale ba 44° +A Deira einch JOUGS ssvssseac Meo) In order to further confirm our vapour-pressure determinations, we made a determination of the boiling point of some carefully dried and redistilled nickel carbonyl. With the barometer at 769 mm. the liquid boiled at 43°:2 — 43°:33 C. It would therefore appear that our vapour tension curve is the more accurate. The following curves, fig. 2, illustrate very clearly the extent of the deviations at different temperatures. 800 7O0O Y. g S 10) S S : Fressure in ‘Millimetres of Mercuri 1005 -/0°C Ox /o° 20° 30° 440° 507 Temperature. INTE, Ze 1903.] Some Physical Properties of Nickel Carbonyl. 437 The latent heat of vaporisation of Ni(CO),, is 38:1 calories per gramme, and the Trouton constant (molecular latent heat divided by the absolute boiling point) is 20°6, its value for ether being 22. The: number obtained by dividing the absolute critical temperature by the critical pressure, which is proportional to the volume of the molecule, Van der Waal’s constant b, is 15°5; the similar number for carbon monoxide is 3°7, so that nickel carbonyl according to theory has a molecule 4°2 times larger than carbon monoxide. The molecular volume of nickel carbonyl at its boiling point is 136, as compared with 110 for ether. The critical density appears to be about 0°46, while that of ether is only 0°25. If the liquid densities of Mond and Nasini* are taken with the critical data, then the Waterston formula v = 2:0398 — 05667 log (198-7 — ¢) fits in very well with the results.. A similar formula for ether given by Avenarius is :—- v = 3:19-—0°802 log (191 — 2). The molecular volume of nickel carbonyl at its boiling point is 136, subtracting from this 7-2 for the nickel atom, we have 32:2 as the volume of each molecule of carbon monoxide in the molecule. Now liquid carbon monoxide at its boiling point has the molecular volume of 35, so that contraction would take place if liquid carbon monoxide could combine with nickel. The heat of formation of nickel carbonyl is about four times greater than that of the liquefaction of the equi- valent amount of carbonic oxide under normal conditions. The experiments described above show clearly that nickel carbonyl is a substance admirably suited for the demonstration of the phenomena of dissociation. Great care must, however, be taken in handling the substance, owing to its poisonous properties when inhaled. It also forms an excellent illustration of a reversible reaction, and the following experiments serve to illustrate the way in which it may be used for this purpose. A number of carefully dried tubes were exhausted by means of a Fleuss pump, and were then filled with a mixture of 10 per cent. nickel carbonyl vapour and 90 per cent. carbon monoxide, at pressures of 50, 100, 226, 304, 396, 504, and 624 mm. of mercury respectively. These tubes were then heated in a bath until nickel began to deposit ; the tube under observation was then kept at that or a slightly higher temperature for about half-an-hour, and afterwards tested for the presence of nickel carbonyl by heating a clean portion of the tube with a fine pointed flame, and in this way the presence of even a very small trace of nickel carbonyl was immediately detected by the formation of a bright mirror of nickel on the hot part of the tube. The tube at 50 mm. pressure did not deposit nickel when placed in alcohol vapour, * Loc. cit. 438 Some Physical Propertees of Nickel Carbonyl. [Mar. 3, but did so at 100° C., and, after heating for a few hours at this tem- perature, no nickel carbonyl could be detected in the tube. After standing for a few days, however, nickel was deposited when a clean portion of the tube was heated with a pointed flame, thus showing that nickel carbonyl had been regenerated. In the 100 mm. tube nickel was still deposited after heating for two days at 100° C.; all the nickel carbonyl was found to have disappeared after the tube had been heated for some time to 130° C. Nickel was deposited in the tube at 226 mm. pressure at 130° C., in the 301 mm. tube at 158° C., and in the other tubes at slightly higher temperatures. In the two tubes at the highest pressures there was a considerable quantity ot the carbonyl present after heating for an hour at 160°C. All the tubes in which the nickel carbonyl had been so far destroyed that no visible deposit of nickel could be obtained on heating a clean portion of the tube with a small flame, after standing for a few days contained enough of the carbonyl to be readily detected by the above test. Another form of experiment suitable for demonstration proved the reaction proceeded rapidly at the ordinary temperature, and with a measurable velocity at low temperatures, even when the pressure of the carbonic oxide atmosphere was below 200 mm. A large bulb of about 200 c.c. capacity was connected to a mercury manometer of small bore (so that the movements of the mercury in the manometer were proportional to the changes of gas concentration in the bulb). The bulb was highly exhausted and then filled with pure nickel carbonyl vapour to a pressure of 51 mm. of mercury at 15° C. After heating for about an hour to 100° C. the pressure, measured after cooling, had risen to 143 mm., corresponding to a decomposition of about 60 per cent. of the nickel carbonyl present. Heated in a glycerine bath to 154° C. the pressure reached 198 mm., corresponding to practically complete decomposition, which ought to develope a total pressure of 204 mm. On rapidly cooling the bulb and allowing it to stand at the ordinary temperature the pressure fell, at first, about 3-2 mm.in an hour, then after two days it had fallen to 120 mm., or about 55 per cent. had recom- bined, after four more days the pressure was 97 mm., or about 60 per cent. had recombined, after standing four weeks some of the deposited nickel remained unattacked. The bulb was again heated to 150° C., so as to deposit all the nickel on the lower part of the tube, and the pressure now rose again to 200mm. ‘The lower part of the bulb where the nickel had deposited, vas now immersed in liquid air, when it was observed that still a small but distinct diminution in pressure took place after some hours. Liquid carbonic oxide did not, however, appear to react with nickel reduced from the oxide by hydrogen. 1903.] On the Variation of Angles observed in Crystals. 439 The volatile iron carbonyl has been made the subject of a number of similar observations, dealing with its physical properties and chemical stability, which will be discussed in another communication. “An Enquiry into the Variation of Angles observed in Crystals, especially of Potassium-Alum and Ammonium-Alum.” By Protessor H. A Mirrs, M.A. D.Sc. FBS. Received March 10,—Read March 26, 1908. (Abstract.) Corresponding angles measured on different crystals of the same substance usually differ slightly. On cubic crystals the theoretical angles are known. Pfaff professed to have established that only those cubic crystals which display birefringence exhibit divergence from the theoretical angles, but Brauns showed that in lead nitrate, ammonia-alum, and spinel, for both isotropic and birefringent crystals alike, the octahedron angle may differ by as much as 20° from that of the regular octahedron. The author has endeavoured to trace the changes of angle upon one and the same crystal during its growth by measuring it at intervals without moving it from the solution in which it is growing. This is accomplished by means of a new telescope-goniometer in which the crystal is observed through one side of a rectangular glass trough, and the changes in the inclination of each face are followed by watch- ing the displacements of the image of a collimator slit viewed by reflection in it. The crystal is held by a platinum clip which it envelopes as it grows. Small movements of the image are followed by means of a special micrometer-eyepiece which accurately measures the magnitude and direction of the displacement. Examined in this way an octahedron of alum (ammonium or potassium) is found to yield not one but three images from each face ; and closer inspection shows that the crystal is not really an octahe- dron, but has the form of a very flat triakis octahedron. It often happens that of the three faces which nearly coincide, one is large and the remaining two very small, so that of the three images one is bright and the others are very faint, and can only be discerned with difficulty ; in such a case the crystal as measured in the ordinary way would appear to be an octahedron whose angle differs from the theo- retical value by a few minutes. When a growing crystal of alum is watched for several hours or days, it is found that the three images yielded by an apparent octahedron face continually change their position ; one set fades away and is replaced by another set, which are generally more widely 440 = On the Variation of Angles observed in Crysials. [Mar. 10, separated than those which they succeed. The images move in three directions inclined at 120° to each other, and indicate that these faces always belong to triakis octahedra. The point in which the lines of movement intersect within the field of view of the telescope would, therefore, be the position of the image reflected from the true octahe- dron face. Measured in this way the octahedron angle of alum is found to be the theoretical angle 70° 312’. The images do not move continuously, but per saltwm, indicating that the reflecting planes are vicinal faces which probably possess - rational indices, and must, therefore, be inclined at certain definite: angles to the octahedron face; but the indices are very high numbers.. Observations upon sodium chlorate, zine sulphate, magnesium sulphate, and other substances, show that other crystals exhibit the: same behaviour. ‘The faces of a crystal are in general not faces with simple indices, but vicinal planes slightly inclined to them, and they change their inclinations during the growth of the crystal; they also change their inclinations when the crystal is immersed to a greater or tess depth in the solution. . Every point within a crystal has at some time been a point on the surface, and has been subject to the conditions of equilibrium between crystal and solution which prevail there. It is believed by the author that a study of the vicinal planes and of the liquid in contact with them, may lead to some understanding of these conditions. In order to ascertain the composition of the liquid, attempts were made to determine its refractive index by means of. total reflection within the crystal. This appears, indeed, to be the only method which can give direct information concerning the ultimate layer in contact with the growing face, and it is somewhat remarkable that it has not been applied before. Considerable difficulty was experienced in making this measurement, but ultimately good readings were obtained, which gave the value 1°34428 as the refractive index in sodium light, at 19° C., of the liquid in contact with a growing crystal of alum. ‘The refractive indices of a series of solutions of known strength, ranging from dilute to supersaturated, having been previously measured, the above index was found to correspond to a liquid con- taining about 10°80 grammes of alum in 100 grammes of solution. A saturated solution at 19° C. was found to have the refractive index 1:34250, and to contain about 9°01 grammes of alum in 100 grammes of solution. Sodium chlorate was examined in the same way ; it was found that the liquid in contact with a growing crystal has at 19° C. the index 1-38734, and contains about 47°8 grammes of salt in 100 grammes of solution ; a saturated solution of sodium chlorate at 19° C. has the index 138649, and contains about 47-2 grammes of salt in 100 grammes: of solution. 1903.] Dependence of Refractive Index on Temperature. 441 The liquid in contact with a growing crystal of sodium nitrate has at 19° C. the index 1:38991, and contains about 48°45 grammes of salt in 100 grammes of solution; a saturated solution at 19° C. has the index 1°38905, and contains about 48:1 grammes of salt in 100 grammes of solution. In each case the liquid in contact with the growing crystal is slightly supersaturated. It was not found to exhibit double refrac- tion even in the case of sodium nitrate. No experiments seem to have previously been made upon the nature of this liquid. G. Wulff has suggested that vicinal faces are due to concentration streams in the solution. Jn order to test this view, crystals of alum were measured after growing for several hours in solution kept con- tinually agitated in order to eliminate the action of the concentration streams. Almost no effect was produced upon the angles of the vicinal faces. In sodium chlorate and sodium nitrate the solute is about 45 times more dense in the crystal than in the adjacent liquid. Now planes with high indices in a space-lattice contain fewer points in unit area than planes with simple indices. The author suggests that vicinal faces grow upon a crystal in preference to simple forms because the erystallising material descends upon the growing face in a shower which is not very dense. “On the Dependence of the Refractive Index of Gases’ on Tem- perature.” By Gzorce W. WALKER, M.A., Fellow of Trinity College, Cambridge. Communicated by Professor J. J. Tuomson, F.R.S. Received February 26,—Read March 26, 1903. (Abstract.) The investigations of Professor Mascart on this subject are perhaps the most extensive of any up to the present time. He examined the effect in several gases, and found that in general the tempera- ture coefficient exceeded the theoretical coefficient given by Gladstone and Dale’s law. The range of temperature was, however, compara- tively small, and his results for air do not agree with those of Lorenz, von Lange, and Benoit. In fact these four observers dis- agree. Lorenz ard Benoit found a coefficient agreeing with the above law, while von Lange obtained a coefficient less than the theoretical value. A repetition of the measurements therefore seemed desirable. The gases examined were air, hydrogen, carbon dioxide, ammonia, and sulphur dioxide. The range of temperature was from 10° C. to 442 Dependence of Refractive Index on Temperature. [Feb. 26,. 100° C. The method used was the well-known one of Jamin, but special precautions were taken to obtain accuracy, and to be sure that the gas had not changed in composition during the various. changes of pressure and temperature to which the containing tubes were subjected. An accuracy of about one part in 600 has been obtained. The results are briefly shown in the following table, and it will be observed that the temperature coefficients obtained are substantially less than those obtained by Mascart. Absolute Value of » for the D line at 760 mm, and 0° C. | Carbon dioxide. Sulphur Ammonia. es dioxide. Observer. Air. Hydrogen. Mascart......| 1°0002927 | 1:000139 | 1:000454 1-:000379 | 1:0007038 | | | | } | Morenz.. 6 =| — | 1-000139 | — 1°000373 _ Ketteler .....| _- 1000143 | 1:°000449 — _ 1-000686 Dulong...... | 1:000294 | 1°000138 1:°000449 1°000385 = 1 000665 Walker SG SSS | 1 °0002928 | 1°0001407 | 1°0004510 1-°0003793 | 1°0006758 | +3 +15 +5 +5 +4 | Temperature Coefficients of Refractive Index. WVIEGr ca ack settee | le, SiWhie ota Hydro- = Carbon |Ammonia.| Sulphur | gen. dioxide. dioxide. | | oe | Coefii. of vol. ee igs ‘00367 0-00366 0 -00371 | 0-00382? 0 -00390 | Mascart . ss be as q 0 00382 | 0:00378 0:00406; —_ | 000460 | 000360 0 -00350 0-00380 0:00390 0 -00416 | i | i 48 +3 ey, 23 | = 1903.] On the Evolution of the Probosculea. 443 “On the Evolution of the Proboscidea.” By C. W. ANDREWS, D.Se. Communicated by Professor E. Ray LANKESTER, F.R.S. Received March 5,—Read March 26, 1903. (Abstract.) Until the author’s recent discoveries of primitive Proboscidea in the Middle and Upper Eocene formations of the Fayum, Egypt, the oldest known members of this mammalian order were Dinotherium cuniert and Tctrabelodon angustidens, from the base of the Miocene in France. The new Egyptian fossils not only reveal for the first time the early history of the order, but also provide more satisfactory material for the discussion of its evolution than has hitherto been available. The most important changes in the Proboscidea occur in the skull, mandible, and dentition. Owing to the increase in the size of the tusks and to the presence of the proboscis, the facial region of the skull becomes shortened, and at the same time the premaxille become wider. The presence of the proboscis also accounts for the position of the external nares. The demand for a greater surface of attachment for the muscles supporting a skull rendered heavy by the tusks and trunk, is met by the great development of the diploé in certain of the cranial bones,, resulting in the enormous expansion of the forwardly sloping occipital. surface. The maxille become greatly enlarged concomitantly with the increase in the size and degree of hypselodonty of the molars. At the same time the zygomatic arch becomes weaker and the jugal takes a smaller share in its composition. The mandible is at first short and stout, with a massive symphysis. Afterwards it becomes more and more elongated as the stature of the animals increases ; and this elongation is for the most part effected by the lengthening of the symphysial region, though the backward rotation of the ascending ramus tends to the same end. The pro- longation of the mandible beyond the premaxille must have been covered by a proboscis-like structure composed of the upper lip and nose, probably more or less prehensile at its extremity. The length- ening of the mandible seems to have reached its maximum degree in the Middle Miocene, after which it again became shortened by the reduction of the symphysis, while the fleshy and now mobile proboscis was left behind as the sole organ of prehension. In the upper dentition the chief changes are the loss of incisors Nos. 1 and 3, and the great increase in size of incisor No, 2, which eventually forms the great tusk characteristic of the later Proboscidea. The canines are soon lost. In the earliest forms, some at least of the cheek-teeth (milk-molars) are replaced by premolars in the usual Ata Drs. D. Hepburn and D. Waterston. (Jame 23. manner, and these teeth remain in wear simultaneously with the true molars ; but in later forms no vertical succession takes place, and as the milk-molars are worn they are shed, being replaced from behind by the forward movement of the molars. Of these also the anterior may be shed, until at length in old individuals of the later types the last molar is alone functional. The gradual increase in the complexity of the proboscidean molars is one of their most striking characteristics. All stages can be traced between the simple, brachyodont, bilophodont (quadritubercular) molars of Moeritherium (Middle Eocene) to the extra- - ordinarily complex type of tooth found in Elephas. Thus in Palwo- mastodon (Upper Eocene) the molars are trilophodont, and the same is true of the first and second: molars of Yetrabelodon (Miocene), in which, however, the last molar is complicated by the addition of further transverse crests. In the Stegodonts of the Siwalik Hills (Pliocene) a further increase in the number and height of the crests takes place, and the whole crown of the tooth is more or less.covered with a thick coat of cement. Still later, the transverse crests become highly com- pressed laminze united by cement, and these are!as many as twenty- seven in number in the Pleistocene Elephas primigenius and the recent E.indicus. The evolution of the lower molars corresponds with that of the upper molars. Of the lower incisors the middle and outer pairs (Nos. land 3) are soon lost, but the second pair remains functional for a long geological period. When the symphysis becomes shortened, these incisors are sometimes retained as vestiges (¢.g., in Mustodon americanus), but in the genus LHicphas they have completely dis- appeared. «A Comparative Study of the Grey and White Matter of the Motor Cell Groups, and of the Spinal Accessory Nerve, in the Spinal Cord of the Porpoise (Phocena comnvwunis).” By DAVID Heppurn, M.D., and Davip WaTErSTON, M.A., M.D. Com- municated by Sir Wm. TuRNER, F.R.S. Received January 23, —Read March 12, 1903. (Abstract. ) Recent advances in our knowledge of the arrangement of the motor cells in the anterior cornua of the spinal cord of man have been made almost entirely by the study of the changes produced in these cells by the division or removal of limbs or ‘parts of limbs in the human subject, and very little has, as yet, been done to elucidate this subject by the comparative method of investigation. On the Spinal Cord of the Porpoise, 445 The authors considered it probable that much information might be obtained by the careful study of the cell groups in the spinal cord of a mammal differing markedly in its musculature from man, and as no previous observations on similar lines had been made in the same fulness on the spinal cord of any of the Cetacea, they describe the results of an examination of the cell groups in each segment of the cord of a member of this class, Phocena communis. The investigation was carried out by obtaining a very recently captured specimen and at once preserving its tissues by injecting into its blood-vessels a solution of formalin, a method which has the advantage of preserving the natural configuration of the enlargements of the cord. A number of sections were prepared for the microscope by different methods from each segment of the whole cord, and typical sections were selected and photographed. The principal features in which the musculature of the porpoise differs from that of man are the almost entire absence of a hind limb, the reduced musculature of the upper limb, and the possession of a large and flexible tail acted upon by powerful muscles, with some other differences noted in the text. The segments of the cord giving origin to the nerves supplying these parts were compared with corresponding segments of the human cord. The groups of motor cells were found to be clearly differentiated from one another, and striking changes were found to occur in the shape of the grey matter and in the cell-groups as we passed from one segment of the cord to another, pointing clearly to a connection between the character of the part supplied with motor nerves from any segment, e¢.g., limb, trunk muscles, genital muscles,—and the arrangement of the anterior horn cells in that segment. The area of the grey matter and of the different columns of white matter was also determined at each segment. The authors also describe some hitherto unrecorded features in the minute structure of the cord of this animal, especially the position of the nucleus of the spinal accessory nerve, and a detached mass of grey matter, probably corresponding to the vesicular column of Clark, in the lumbo-sacral region of the cord. Lo A VOL. LXXI. 446 Sir Norman Lockyer and Dr. W. J. 8. Lockyer. [Mar. 17, “Solar Prominence and Spot Circulation, 1872—1901.” By Sir Norman Lockyer, K.C.B., F.B.S., and WILLIAM J.S. LOCKYER, Chief Assistant, Solar Physics Observatory, M.A. (Camb.), Ph.D, (Gétt.), F.R.AS. Received March 17,—Read March 26, 1903. | PLATES 6 AND 7. | In our former communications* referring to the connection between solar, meteorological, and magnetic changes, some of the results obtained by the reduction of the solar prominences, as observed by Professor Tacchini at Rome, were described. It was stated that the curve representing the variation of percentage frequency of the prominences for the whole limb of the sun indicated that in addition to the main epochs of maxima and minima coinciding in time generally with those of the maxima and minima of the total spotted area, there were also prominent subsidiary maxima and minima. Further, dividing the sun’s limb into zones of 20° in width from the equator, with a polar zone of 10°, and dicussing each zone separately, the variation of the prominence percentage frequency about the equator was found to be very different from that in the higher lati- tudes, the former changing with the spots, and the latter exhibiting sudden outbursts just previous to the epochs of sunspot maxima, followed and preceded by comparatively long intervals of quietude. In the present communication, the prominence observations have been discussed from a different point of view, in order to trace out, if possible, the heliographic latitudes of the chief centres of action of prominence disturbance. In this way it could be determined whether such movements are subject to some periodic law, in which case it would be possible to increase our knowledge of the circulation of the solar atmosphere in regions outside those in which sunspots alone have, up to the present, been employed. The changes of latitude of the zones which contain the centres of sunspot disturbances were first pointed out by Carrington,7 whose fine series of observations led him to discover ‘‘a greater contraction of the limiting parallels between which spots were formed for the two years previous to the minimum of 1856, and soon after this epoch the apparent commencement of two fresh helts of spots in higher latitudes north and south, which have in subsequent years shown a tendency to coalesce, and ultimately to contract as before to extinction.” The study of the subject was taken up later by Spoerer,t who * “Roy. Soc. Proc.,’ vol. 70, p. 502; vol. 71, pp. 184 and 244. + ©Observations of the Spots on the Sun from November 9, 1853, to March 24, 1861, made at Redhill,’ p. 17. t ‘Beobachtungen der Sonnenflecken yon Oct., 1871—Dec., 1873, und von Jan., 1903.] Solar Prominence and Spot Corculation, 1872—1901. 447 corroborated Carrington’s results and extended the discussion of the observations up to the end of the year 1879. The result of these two investigations showed that at sunspot maximum there was only one zone in each hemisphere in which spots were situated, the centre of this being about 18° N and 8, while at minimum there were two zones existing simultaneously in each hemi- sphere ; the older cycle dying out in the zone, the centre of which was situated in low latitudes, and the new one commencing in high latitudes, its centre being about latitude + 30° to + 35°. Later observations extending up the present year have further cor- roborated these general deductions, for each hemisphere, and we are now quite familiar with this cycle of sunspot latitude variation. In the present investigation, the fact has been brought out that the prominences also undergo an apparently regular variation of latitude throughout a period of about eleven years concurrently with the spots. For the purpose of our inquiry, the object of which has been stated above, we have discussed independently of each other, two fine series of prominence observations, one made by Tacchini at home extending from 1872 to 1900, and the other by Ricco and Mascari at Catania from 1881 to 1901. Both these series were handled in the same way, and both indicated similar changes of latitude of prominence action, showing that the variations recorded were real and not due to any personality of the observer or difference in the method of observation. The data for the discussion of the solar prominences as observed by Tacchini have been taken from the same source as before,* while those of Ricco and Mascari are published in and have been extracted irom the same volumes. We may here take the opportunity to express our thanks to Professor Ricco,. with whom we have been in communication, and who has very kindly forwarded for our use some unpublished data con- cerning his prominence observations and reductions. The method of reduction adopted was to determine for each year the percentage frequency of prominence activity for every 10 degrees of solar latitude north and south, A series of curves was next drawn, one for each year, the abscisse representing the latitudes of prominences north and south, and the ordinates their percentage frequency. It was then found that the centres of prominence activity, or, in other words, the maxima of the curves, were sometimes single, sometimes double, and in one or two cases even triple in each hemisphere. This suggested that just as sometimes there are two 1874—Dec., 1879.’ ‘ Publicationen des Astrophysikalischen Observatoriums zu Potsdam,’ Band I and II. * “Societa Spettroscopisti Italiani,’ vol. 1, 1872, to vol. 29, 1900. Ze ke2 448 Sir Norman Lockyer and Dr. W. J. 8. Lockyer. [Mar. 17, zones of spots existing at one time, so there might be one, two, or occasionally three zones of prominences In existence in each hemisphere simultaneously. 600 1 | | 1 : ! ' I { | 200 S00 e400 325007 0 100 200 300 400 I 100 i | 500% 400) 500 0 { 200 Fig. 1.—Curves showing the variation of the percentage frequency of prominence activity for every 10° of latitude for the years 1879—1881, from Tacchini’s observations. Further, a close examination of the whole set of curves with reference to these points of maxima made it possible not only to study the changes of latitude of these points from year to year and their positions when commencing to develop or about to disappear, but the in- tensity of these centres in rela- tion to each other. The accompanying illustration (fig. 1) shows the curves drawn for the years 1879, 1880, and 1881, from the observations of Tacchini, and serves as an ex- ample of the curves that have been discussed ; they exhibit the change from a single to a double centre of activity in each hemi- sphere. Thus in 1879, there was a prominence maximum in each hemisphere at latitudes + 50°. In the next year (1880), both the maxima had retreated further away from the equator, namely to latitudes + 60°, while another centre of disturbance began to make itself apparent at latitudes + 30°. In the year 1881, both centres in each hemisphere were strongly marked and became of about the same intensity, their mean latitudes in each hemi- sphere being about +30° and +60°. These curves thus indi- cate that during these three years, the direction of motion of these centres of activity tends pole- wards or away from the equator. By examining both series of observations made by Tacchini and Ricco and Mascari, and analysing the positions of the principal and 1903.) Solar Prominence and Spot Circulation, 1872—1901. 449 subsidiary maxima for the whole period covered by the observations, the results illustrated graphically in Plates 6 and 7 were obtained. In these figures the facts are brought together for each hemisphere separately. The medials of the lines (curves A and B) show the heliographic latitudes of the centres of prominence action; the thick- ness of these lines represents the relative percentage frequency of prominence action. For the sake of comparison, three other curves for each hemisphere are given. ‘The first (curve C) shows the mean heliographic latitude of spotted area for each hemisphere. For the construction of these, the values, since 1873, have been extracted from the Greenwich Reductions,* but previous to that date the values have been obtained from Mr. Marth’s reductions,y and those completed at the Solar Physics Observatory from measures supplied by Professor Backlund, of the Wilna Observatory. The next curve (curve D) illustrates the variations of the per- centage frequency of prominence action for each hemisphere taken, as a whole, and is similar to those given in our former papers. The last curve (curve E) shows the variation of the mean daily area of sunspots from year to year, also for each hemisphere. Referring now to the changes of latitude of the prominence centres of activity, both series of curves for the north as well as for the south hemisphere exhibit the same general features. The first conclusion illustrated by the curves is that prominence activity in the main has a poleward drift, that is, the change of position of the zones of activity is in the direction from low to high latitudes. In some years, the centres of activity appear to form two zones in each hemisphere at about latitudes + 24° and + 50°, which eventually amalgamate at about latitude + 40° and move polewards, fading out in about + 70° to + 80°. As this zone disappears in high latitudes a new zone at about latitude + 20° begins, and this after a few years becomes associated with another zone in about latitude + 50°, and eventually amalgamates with it. The epochs at which these different zones come into play in relation to the general curve of prominence activity for the whole hemisphere are as follows: From a little after the maximum of prominence activity to just before the minimum, two zones in the latitudes + 24° and + 50° are in existence and of decreasing intensity. Before the minimum is reached these two zones amalgamate in about latitude + 40°. At the minimum there is only one zone, and this of small intensity. Between the minimum and the following maximum this zone rapidly takes a northern movement, increasing in intensity; a new outburst * “Spectroscopic and Photographic Observations made at Royal Observatory, Greenwich, 1884,’ and after. + MSS. at Royal Society. 450 Sir Norman Lockyer and Dr. W. J. 8S. Lockyer. [Mar. 17, occurs In a zone nearer the equator (latitude + 24°), which also increases rapidly in intensity. After these general statements, we now refer to soime details showing that there are some variations from the above generalisa- tion. For these details the curves deduced from both sets of observations made by the different observers are so very similar that it does not matter which are examined. | Attention may first be drawn to certain differences between the curves representing the latitude variation for the two hemispheres. It will be noticed that for the period 1872—1882, the curves for both hemispheres are very similar. We next consider the period 1880— 1893. Here there are differences between the two hemispheres. The curve for the northern hemisphere resembles very closely that for the preceding period, but it differs somewhat from its corresponding curve for the southern hemisphere. The corresponding northern zone in latitude 45° is missing from the southern hemisphere, while a zone of activity nearer the equator about latitude 24° is present. Further, the polar zone for the southern hemisphere continues to be prominent for two years longer than that in the opposite hemisphere. In the succeeding curves, which extend from 1891—1901, both hemispheres are more or less similar, and both resemble in a greater degree those for the southern hemisphere for the period 1880—1893 than those for the period 1872—1882. Although the Roman and Sicilian observations give nearly identical curves, hemisphere for hemisphere, the apparently regular cycle of change of latitude which was operative for the northern hemisphere 1872—1893, and for the southern hemisphere 1872—1882, does not seem to have been so exactly maintained in late years; more irregular still perhaps is the last cycle commencing in the year 1892. Hence, there seems reason to believe that the prominence circulation is not quite the same for each cycle, and this may in some way be due toa longer solar period such as that of about 35 years. But it is important to state that our deductions may be partially incomplete owing to the difficulty of determining sometimes whether a new centre of action has been formed or the position of an old one changed. Further, account must be taken of the fact that the material discussed does not represent the record of the percentage frequency of prominences determined from observations made on the dise of the sun (now rendered possible by the Janssen-Hale-Deslandres method), but one obtained from observations of the phenomena occurring only at the limb of the sun. The close agreement between the observation of the different observers shows nevertheless that this latter method is of great value. Another important series of prominence observations is that made 1903.] Solar Prominence and Spot Circulation, 1872—1901. 451 by Father A. Fényi, 8.J., who has published* the individual observa- tions, and the reductions of the positions and frequency of promin- ences observed at Haynald Observatory for the years 1884 to 1890 inclusive. He gives curves constructed somewhat after the manner adopted in the present enquiry, as illustrated above, in fig. 1. A com- parison of the points of maxima from his curves with those of Tacchini and Ricco and Mascari for the period common to all three sets of observations is made in the following tables, each hemisphere being given separately. The vertical columns show, for each year, the heliographic latitudes of the points of maxima, and an asterisk (*) is placed against the one which is the more or most prominent in each hemisphere ; when there are two, and they are of equal intensity, this symbol is attached to each, while in the case of only slight indications of maxima the latitude is enclosed in brackets. Northern Hemisphere. 1884. 1885. | 1886. | 1887.| 1888. 1889. | 1890. | SS | ava * * | * * | | Tacchini... . 50, 25 | 43, 25 | 45, 20 | 30 30, 15 | 40 | 45 | Ricco and 520 55, Lai) 4b, 25 | 30) | Bb) Ad 45 | Mascari % % | % * | | | Kényi «.....| 65, 45, 15 | 45, 25 | 45, 2 | 45 | 40, 20 | 438, 25 | 45 | | | | | Southern Hemisphere. | : aay | | 1884. 1885. 1886. 1887. 1888. | 1889. | 1890. | | Tacchini..| (75), 25, 5 25) 85,2045, 25 | 45,25) 4b 45 | % * x | | Ricco and | (83), 25, (5) Doo 25 | 50425) 45) ul SON Mascari x % Ea Ea. Es ES * | * | Fényi ....| (75), 35, 1 | (0), 85 10 | 85,10] 50, 80, 15/50,25) 40 | 50, 20, | | | It will be seen that for these seven years, Fényi’s results are in very close accordance with those deduced from the other two series of observations, thus generally endorsing those portions of the curves in Plates 6 and 7 covering this period. It was mentioned in a previous papery that the mean prominence curve for each hemisphere exhibited subsidiary maxima and minima. In the light of the present investigation, it is interesting to compare * *Publicationen des Haynald-Observatoriums, Kalocsa,’ Heft VI, 1892, und WV Eir 1902. + ‘Roy. Soc. Proc.,’ vol: 71, p. 244. 452 Solar Prominence and Spot Circulation. [Mar. 17, this curve with that representing the changes of latitude of the zones of prominences. In every case, and for each hemisphere, the subsidiary maxima are coincident in time with the presence of two zones of prominences, each well-developed, while at the principal minima only one zone is in evidence. We have already explained the fact that spots are restricted to a zone having its limits at latitudes + 5° and + 35°, while prominences occur all over the sun’s disc, even up to the poles, and also that spots always commence their cycle in high latitudes (about + 35°) and gradually approach the equator until within 5°, when a new cycle is commenced in high latitudes. Prominences on the other hand begin in comparatively low latitudes (about + 24°), and finish their cycle near the poles. 3 A glance at the Plates 6 and 7 brings out the interesting fact that at sunspot minima, when two zones of spots are in evidence, there is only one zone of prominences, while when only one zone of spots exists the prominences are for the most part confined to two zones. The conclusions arrived at in the present communication may be summarised as follows :— 1, The centres of action of prominence activity undergo an appa- rently regular variation. 2. The direction of motion of these centres is from low to high latitudes, the reverse of that of spots, which travel from high to low latitudes. 3. At epochs of prominence minima (which are concurrent with sunspot minima) these centres of action are restricted to one zone (about latitude + 44°) in each hemisphere, while those of the spots occupy two zones in each hemisphere. 4. At nearly all other times these centres are apparent in two zones, while those of the spots occupy only one in each hemisphere. 5. The subsidiary maxima exhibited by the curves representing the percentage frequency of prominence activity for each entire hemi- sphere, are due to the presence of two well-developed centres of pro- minence activity in each hemisphere. Lockyer and Loc I4-0C SUNSPOTS. MEAN patty 2° AREA. | N.HEMeE 60C 20C 2S PROMINENCES “FREQUENCIES .)3, NORTH LATS. (TACCHINI) 2 MEAN HELIOGRAPHIC LATITUDE OF SUNSPOTS NORTH OF EQUATOR. lO O 8C LATITUDE OF 6C PROMINENCES N.HEME 4¢ (RICCO AND +20 MASCARI) a 8 LATITUDE OF PROMINENCES,, N. HEME (TACCHINI) 45 @ PLATE 6.—Curves Note.—The contin * pees The Papi re «+ ec immat iea aeae aaa Lockyer and Lockyer. ‘Roy. Soc. Proc., vol. 71, Plate 6. 1890-0 1900-0 rap oid a db dah pak on ld na nay 1860-0 140 SUNSPOTS. MEAN patty [20° AREA. E N.HEMe 600 20 “FREQUENCIES .|3 NORTH LATS. 23 PROMINENCES | (TACCHINI) 3 SUNSPOTS NORTH OF EQUATOR. 8 LATITUDE OF 60 PROMINENCES N.HEME 404. (RICCO 0 (0) 0 MEAN 2 HELIOGRAPHIC LATITUDE OF 10 (@) AND +2 MASCARI) 8 LATITUDE Gh 7c PROMINENCES,, N. HEME (TACCHIN!) 45 O es ae : . , Puare 6.—Curves showing the Relation between the Positions of the Centres of Action of Solar Prominences (A and B) and Spots (C), Percentage Frequency of Prominences (D), and Mean Daily Areas of Spots (E), for the Northern Hemisphere of the Sun. Note—The continuous and broken yertical lines represent the epochs of sunspot maxima and minima as determined from the mean daily areas of the whole = solar clisk. ” Ld old dwg eee ag an Bal x + “" f < ly a ‘ : f =f ‘ = ee Wiha ‘ nm &@ t A > ¥ \de , i ns ees a - Nes ; > Ss an . Ns . 5 - i \ x = ~ ~ = 7 ‘ - ee | c aw 3 1 ‘ cng * a A ” » 2 7 ” i et ye r? , ee i te | ‘ » ? A y he = a A MSD phe aay ER Ue od OP Cae a Lockyer and Lo LATITUDE -2¢ OF | PROMINENCES saa S.HEME | (TACCHINI) 6¢ 86 LATITUDE OF —2( PROMINENCES S HEME AC CRICCO AND 6C MASCARI) 8C MEAN HELIOGRAPHIC LATITUDE OF IC SUNSPOTS SOUTH OF 2C EQUATOR . >ROMINENCES<’ eFREQUENCIES. _ SOUTH LATS. (TACCHINI) I20C SUNSPOTS MEAN DAILY AREA. S.HEME 806 40¢ Pxrate 7.—Simil: ’ i S ’ x 2 ) . ~ . ie n 5 sre ‘ 2 vee c i . 2 Lockyer and Lockyer. Foy Soc. Proc., vol. 71, Plate 7. O LATITUDE-20 OF PROMINENCES 40 S.HEME (TACCHINI) 60 80 0 LATITUDE On =20 ya PROMINENCES S.HEME 40 (RICCO AND 60 MASCARI) 80 MEAN 0 HELIOGRAPHIC LATITUDE OF SUNSPOTS SOUTH OF EQUATOR . 10 20 FREQUENCIES. |_ SOUTH LATS. (TACCHIN1) | a MEAN DAILY AREA. S.HEME 70° 1200 =. 0 FORGE LO Poo Peg r oer Pree nea dea dy dee ee bat pa at 1860-0 1870:0 1880-0 1890-0 1900-0 PLATE 7.—Similar Curves to those on Plate 1, only in this case the Southern Hemisphere of the Sun is referred to. Vertical lines same as on Plate 1. ¥ y . % F ; ‘ } 1903. ] On the Cytology of Apogamy and Apospory. 453 “On the Cytology of Apogamy and Apospory.—1. Preliminary Note on Apogamy.” By J. B. Farmer, F.B.8., J. HE. S. Moore, and Miss L. Dicpy. Received March 24,—Read March 26, 1903. The phenomena of apogamy and apospory have always been regarded as “short cuts” in the life history of ferns, and the fact that apparently either generation may be directly produced from :the other, without the intervention of oosphere or spore respectively, has been taken to indicate that the gametophyte and sporophyte are homologous phases in the life histories of these plants. But since the cytology of the two generations has been carefully studied, it has become recognised that the prothallial generation is composed of cells, the nuclei of which possess only half the number of chromosomes that are characteristic of the alternate sporophyte generation. In normal cases the doubling of the number of the chromosomes is effected during the transition of the gametophyte to sporophyte by the addition of the chromosomes belonging to the spermatozoid to those of the oosphere, and this double number is retained until they once more become reduced to one half in the formation of the spores that introduce again the gametophyte stage of the life-cycle. It is obviously therefore of considerable theoretical interest to ascertain how the irregular transitions known as apospory and apogamy are effected. It is with the facts of apogamy that we are here concerned, the details relating to apospory being reserved for a future communication. Certain species of nephrodium (¢.g., NV. pseudo-mas, var. polydactylum) are known to produce prothallia on which the apogamous formation of sporophytes is of normal instead of rare occurrence.* By the kindness of Dr. Lang we received a number of prothallia in all stages of growth in which the special apogamous developments could be perfectly traced. An examination of them has resulted in the dis- covery of remarkable nuclear changes that appear to be obviously related to the apogamy of the prothallia in question. If very young prothallia are examined before any apogamous growths have begun to manifest themselves, it will be seen that cells not unfrequently occur in which two nuclei are present. This fact was recorded by Lang in the case of older prothallia, and was also figured by Heim7 in the case of Doodya caudata, but he makes no mention of it in the body of his paper. * Lang, ““On Apogamy and the Development of Sporangia upon Fern Pro- thallia,” ‘Phil. Trans.,’ series B, vol. 190, 1898, p. 214. + Carl Heim, ‘“ Untersuchungen ti. Farn Prothallien,” ‘Flora,’ vol. 82, 1896, p. 338, fig. 7. 454 Messrs. Farmer and Moore and Miss Digby. [Mar. 24, Furthermore, it is to be observed that whenever this state of things is seen, there 1s always, so far as we have observed, at least one con- tiguous cell which is destitute of a nucleus (see figs. 1, 2, 4). We convinced ourselves of this highly important fact by examining entire prothallia that had been carefully stained, as there was always the Fra. 1. oe sas arrestee) = i oe ORI Ss To Sares a tn 8 Ni tig PACE! Pee O Pin i 4 enone ; Ak WS anc ro ce as "he ae ‘ " : tisk that in erections the appearance might be due to the displace- ment of the missing nucleus. We were further able to trace the migration of the nucleus from one cell into that of its neighbour in a sufficient number of instances to convince us that this affords the explanation of the peculiar circum- stances just mentioned (figs. 3,4). In several instances the nucleus was seen in the act of passing through the wall, and in others the path through which it had traversed was plainly visible as a pertora- 455 ia ae * On the Cytology of Apogamy and Apospory. ; : i ‘3 H : 456 Messrs. Farmer and Moore and Miss Digby. [Mar. 24, tion through which a strand of cytoplasm was still visible connecting the two cells. When the migrating nucleus has passed into the neighbouring cell it sometimes fuses at once with the nucleus already present there, but often the two nuclei remain more or less separated for an appreciable interval of time. It appears then to be clear that the presence of the pair of nuclei is not to be regarded in these cases as resulting from a division of the nucleus proper to the containing cell, but has merely been arrested at a stage short of producing a wall. Such a state of things is very common in the tissue-cells of many plants, but we think that the facts enumerated above suffice to prove that no such simple explana- tion will fit in the present case. The migration of the nucleus, as above described, goes on discon- tinuously in a growing apogamous prothalliun. In this way there is provided a cellular aggregate that may possess no very homogeneous character, nor can one cell, or even isolated groups of cells, be defined as the sole parent tissue from whence the apogamous out- growth may have sprung. And this is quite in harmony with the 1903.] On the Cytology of Apogamy and Apospory. 457 irregular growth of the new tissue, and with the sporadic appearance of sporophytic members, on the prothallium already described by Lang. When the nuclei of the cells in the apogamous regions are examined in course of karyokinesis, they are seen to possess a much larger number of chromosomes than those of the ordinary tissue cells of the prothallium. Owing, however, to the manner in which the very numerous chromosomes are distributed on the spindle, an exact estima- tion of actual numbers is a task of considerable difficulty. There appear, however, to be forty and eighty in the respective classes of nuclei. We regard the whole process as a kind of irregular fertilisation. The doubling of the chromosomes receives an explanation strictly analogous to that afforded by the normal fusion of oosphere and spermatozoid. But instead of one cell only (the oospore) serving as the starting point for the new generation, a number of such units loosely co-operate to produce it. And in this connection it is perhaps significant that the young plantlet is commonly borne on, and produced from, a special sporophytic outgrowth, of which the constituent cells may have become homologously differentiated into a sort of pro-embryo. Instances are not wanting amongst lower animals to show that close cellular in-breeding may occur, and form part of the normal sexual series of events. It is not desired to press the analogy of such cases; but the case of Actinospherium shows that a process indis- tinguishable from normal sexual fusion may occur between sister cells that have only lately arisen from the division of a parent cell. We do not propose to enter on a full theoretical discussion of the bearings of these observations at the present time, but we hope to do so when we are in a position to deal completely with the corresponding cytological features associated with apospory. 458 Miss H. Chick. Ona Unicellular Green Alga, (Feb. 28, “A Study of a Unicellular Green Alga, occurring in Polluted Water, with especial Reference to its Nitrogenous Meta- bolism.” By HarrietrE Cutck. Communicated by Professor RuBERT Boyce, F.R.S. Received February 28,—Read March 1}, GOR: (From the Thompson Yates Laboratories, University College, Liverpool.) [PLATE 8]. A small, unicellular, green alga was noticed to be frequently present in sewage, and sewage materials, when these had been kept for some time. The same alga was also found to have seeded itself in dilute ammoniacal solutions. This somewhat peculiar habitat seemed to promise that something of interest might repay a study of the physiology of the plant. On examination, under the microscope, it was found to be an extremely small, unicellular, non-motile alga, with a well-defined chromatophore, which latter contained a distinct pyrenoid. The general characteristics of the alga pointed to its being closely allied to Chlorella (Beyerinck),* and I have provisionally placed it in that genus. Chlorella vulgaris was originally described by Beyerinck as possessing either no pyrenoid, or an insignificant one ; Chodat,f on the other hand, has described the species as having generally a pyrenoid, and has figured it accordingly. In a culture of C. vulgaris, obtained from Kral, I have been unable to detect a pyrenoid, and, since the alga here treated of possesses a very conspicuous one in its chloroplast, I have ventured to attach a new specific name, and to call the organism Chlorella pyrenoidosa. ‘Though in the higher forms, such variation in cytological character between the species of a genus is practically unknown, yet the same difference occurs between different species of Stzchococcus,t of Raphidiuwm,§ and of Pleurococcus,|| all genera of unicellular green alge, showing very simple cell structure, and more or less nearly allied to Chlorella. The peculiar natural habitat of the plant, and its appearance in ammoniacal solutions alluded to above, both pointed to a preference for ammonia in its nutritive fluid. In fact, in one instance, this alga was found to have grown in a dilute solution of ammonium chloride, and it was thought worth while to make an estimation of the ammonia. This was done by directly Nesslerising, and it was found that the ammonia had disappeared. By distilling the solution with alkalh, however, the ammonia was obtained in a roughly quantitative yield. * ‘Bot. Zeit., 1890, p. 730. + ‘ Beitrage z. Kryptogamenflora d. Schweiz.’ {~ Matruchot and Molliard, ‘Rev. Gén. de Bot.,’ vol. 13, 1902 (May). § Chodat, ‘Mém. de l’Herb. Boiss.,’ 1900, No. 17. || Snow, ‘Annals of Botany,’ 1899; Chodat, ‘ Bull. de l’Herb. Boiss.,’ 1894. 1903.) with especial Reference to its Nitrogenous Metabolism, 459 The algal material also grew exceedingly well when stroked upon the following ammoniacal medium :—2-0 grammes NH,Cl, 1:0 gramme NayCO3, 0°5 gramme K2HPO,, 15-0 grammes agar-agar, in 1 litre water. The growth was, however, contaminated with numerous bacteria. This evident ready assimilation of ammonia pointed to the possibility of the organism’s playing some part in the “purification ” of sewage or further natural “ purification ” of sewage effluents, and its nitrogenous metabolism was studied with this view. Mode of Preparation of a Pure Culture. The first step necessary towards a study of the chemical physiology of the alga was the preparation of a pure culture. As the separation of alge in pure culture is not of very common occurrence, few as yet having been obtained, it may be permissible to describe in detail the method adopted. The following methods were tried without success. Sterilisation, as regards bacteria, was attempted by exposure of liquid cultures to sunlight, and hydrogen peroxide was ‘also tried as a means of killing the adherent bacteria, but the alga appeared to be susceptible also. More purely bacteriological methods were employed, viz.:—by stroking in great dilution over the surface of potatoes, and by “ pouring” plates of ammonium agar, or ammonium gelatine, and incubating in sunlight, ata low temperature. All these | methods failed for one reason or another, but the organism was finally separated in the following way. It seemed advisable to employ an ammonium-containing medium, and ammonium agar was therefore selected. Plates of this medium were poured and allowed to set. An ammoniacal solution was also made and sterilised, having the following composition: —0°5 gramme Na,COs, 0°5 gramme K.HPOQ,, 0-1 gramme NH,Cl, 1 litre tap-water or distilled water. This solution wil] in future be alluded to as “solution A.” A few drops were allowed to drop upon the surface of the poured plates, then a small portion of the impure material was added to the first plate, and the whole was brushed over the surface. The same brush was then used to brush the liquid over the second plate, and so on, to six plates or more, without the addition of fresh algal material, so that a considerable dilution was obtained. The plates were kept in the light, in as much sunshine as was possible, protected from dust, and in a dampatmosphere. In seven to fourteen days, green growth was generally visible, and after three or four weeks, definite algze-colonies were to be distinguished among the colonies of bacteria. In this way pure cultures of Chlorella pyrenoidosa were obtained from (1) Material of Ducat Filter Bed (Hendon), (2) Sludge wash-water (Leeds). 460 - Miss H. Chick. On a Unicellular Green Alga, [Feb. 28, This method may be relied upon to yield pure cultures, for it has since been repeated, and again proved successful. The following is a fairly comprehensive diagnosis of Chlorella pyrenoidosa. Chlorella (Beyerinck)* very minute, unicellular, non-motile, spherical or elliptical green alge; chromatophore single, parietal, with or without a pyrenoid; cells isolated, showing no tendency to form colonies ; multiplication by division of the cell-contents to form a number of daughter cells. C. pyrenoidosa (sp. n.):—cells spherical, 3—5 p» in diameter, some- times attaining 11 »; chromatophore single, parietal mantle-shaped, covering nearly the whole of the cell-wall; pyrenoid conspicuous and single (see Plate 8, fig 3) ; reproduction by successive division of the cell-contents to form within the mother cell as many as eight daughter cells, which subsequently become free of the mother cell wall (see fig. 5). Physiological characters.—Showing a marked preference for ammonia and ammoniacal compounds (¢.g., urea) compared with nitrates in its culture fluids; growth and multiplication largely increased by the addition of glucose to the culture fluids, causing the cells and cell- contents to assume a changed and characteristic appearance. Chlorella pyrenoidosa resembles C. protothecoides (Krugert) very closely, both in its physiology and morphology ; it differs, however, in the possession of a pyrenoid. This pyrenoid with few exceptions is very conspicuous, and can often be stained blue with iodine, being the only part of the cell which shows this staining. C. pyrenoidosa differs also from C. vulgaris (Beyerinck){ in many of its physio- logical properties, as well as by the possession of the well-marked pyrenoid. Its morphology bears, however, a very close resemblance to that of a green alga isolated and studied by Kossowitsch.§ This alga itself resembled C. vulgaris (Beyerinck) and. Cystococcus (Nageli)|| very closely, and Kossowitsch decided to call it “‘ Cystococcus,” in spite of certain small differences which existed. It has seemed to me the wisest plan to attach the alga, diagnosed above, to the genus Chlorella, and to add a new specific name ‘ pyrenoidosa,” although it resembles the Cystococcus figared and described by Nageli, and it seems not unlikely that it may be identical with Kossowitsch’s Cystococcus. C. pyrenoidosa grows exceedingly well, in pure culture, upon nutrient gelatine, nutrient agar or ammonium agar, and also in many liquid media. In media containing glucose the growth is much * “Bot. Zeit.,’ 1890, p. 725; ‘Cent. f. Bakt.,’ vol. 13, 1893, p. 368. y Zopt’s ‘ Beitrage z. Morph. u. Phys. Nied. Org.,’ Leipzig, vol. 4, 1894, p. 92. t ‘ Bot. Zeit.,’ 1890, p. 725; ‘Cent. f. Bakt.,’ vol. 13, 1893, p. 368. § ‘Bot. Zeit.,’ 1894, vol. 6, p. 97. || ‘Gattunger einzeiliger Algen,’ Ziirich, 1894, p. 84. 19035.] with especial Reference to its Nitrogenous Metabolism. 461 stimulated, and this is particularly rapid when grown upon nutrient agar or gelatine to which glucose has been added. The liquid medium par excellence for studying this organism, is sterilised sewage, in which it preserves its normal form and characteristics. In solution A, it also grows well, while if to the liquid a small quantity of glucose is added, the growth is much stimulated, and the individuals assume the appear- ance, described below, characteristic of growth in glucose media. Accompanying the much more abundant development, the nitrogen assimilation is also increased, as will be seen later. Glucose-containing media (fig. 4). When grown upon media containing glucose, while the general growth is much improved, and the individual cells are also larger in size upon the whole, the green colour of the cell-contents is much paler, and sometimes almost disappears, while the chromatophore is much disorganised. The contents of the cell appear to be segregated into a variable number of slightly refractile bodies, usually pale greenish-yellow in colour, which are apparently free in the cell. Ii when mounted in water under the microscope, a slight pressure is applied to the coverslip, the cell envelope is easily ruptured, and these bodies are liberated, and float freely in the surrounding liquid. The protoplasm also, in the case of glucose cultures, shows small granules, and the pyrenoid is no longer to be traced, but a certain amount of staining with iodine may be noticed in the granular protoplasm. These changes closely resemble those noticed respectively by Kriiger* in the case of his Chlorotheciwm saccharophilum and Chlorella protothecoides, and by Matruchot and Molliardy in the case of Stichococcus bucillaris, when these algee were grown upon glucose containing media. The results of growing C. pyrenoidosa upon media containing lactose or saccharose also resembled those of the above authors. These two sugars appear to possess a far lower nutritive value than does glucose, while the algal cells preserve their normal appearance. When grown upon ammonium agar, the growth is comparatively slow, the individuals appear larger in size than when grown upon ordinary nutrient agar, their green is darker and the cell-contents are of the normal type. Quantitative Chemical EHxperiments. The evident preference of this organism for ammoniacal culture solutions seemed to be a reason for studying its nitrogen assimilation. The absorption of ammonia by alge is not a new discovery ; it is well known, from the researches of Letts and Hawthorne,{ that Ulva latissima * Zopt’s ‘ Beitrige z. Morph. u. Phys. Nied. Org.,’ Leipzig, vol. 4, 1894, p. 91. t ‘Rev. Gén. de Bot.,’ vol. 13,1902. _ { Letts and Hawthorne, ‘ Roy. Soc. Proc., Edin.,’ 1901, p. 268; Letts, “‘ Report on the Scheme of Sewage Purification proposed for Belfast, and its Probable Effects on the Lough.” VOL. LXXI. 2 462 Miss H. Chick. On a Unecellular Green Alga, [Feb. 28, absorbs ammonia with extreme rapidity, and, consequently, will grow and develop to an extraordinary extent in sewage-polluted water, for such water contains a comparatively large amount of ammonia. C. pyrenoidosa also absorbs ammonia in a marked degree, and this fact lends practical importance to the study of its nitrogenous metabolism, for it also may have some bearing upon the purification of sewage. Although the organism is itself very small, and the absolute quantities it tackles also insignificant, yet the results of the following experi- ments, themselves upon a very small scale, tend to show that if C. pyrenoidosa were growing in great quantities (¢.g., in the bed of a river polluted with sewage), very important changes might be effected in the composition of the water in which it grew. The method of pure cultures was adopted in every case, unless otherwise stated. ‘‘ Pasteur flasks” were used to contain the cul- tures; these were sterilised and then filled with the sterilised culture fluid by means of sterilised pipettes. One flask was always kept sterile, as a control, while a second was inoculated with Chlorella pyrenoidosa. ‘The flasks were kept, often for months, in a sunny place, protected from dust, and in a damp atmosphere in order to prevent evaporation. These Pasteur flasks can be obtained ° to hold 300 ¢.c., and this size, when containing 150 c.c. of fluid, was found to be extremely convenient. To see if the cultures and controls remained in a sterile condition as regards bacteria, they were tested from time to time by abstracting a few drops and adding them to tubes of ordinary bouillon. These tubes were incubated both at the ordinary temperature and at blood heat, and were watched to see if any bacterial growth took place. ‘The risk of con- tamination was found to be extremely small, and its occurrence very rare. The figures in the following tables refer in every case to cultures that were successfully maintained in a state of purity. It was found, in a few cases, as will be seen in the following tables, that difficulty was experienced in maintaining constant the analysis of the sterile control, when the experiment lasted over a long period. In these cases it would seem as though the ammonia present suffered a slight volatilisation, but this does not vitiate the analysis of the corre- sponding cultures, because the changes in the control are, in all cases, insignificant when compared with the changes taking place in the inoculated fluid. In such cases, it seems a fair procedure to subtract the change in the control from that in the culture, and to consider the difference as due to the activity of the alga. Methods of Analysis——Samples for analysis were removed from the flasks under examination from time to time by means of sterile pipettes. These pipettes were always carefully cleaned, plugged, and then sterilised by heating in a hot-air steriliser at 150° C. for one hour. To protect them from any dust they might attract during this 1903.] with especial Reference to its Nitvogenous Metabolism. 465 operation, their ends were fixed by means of cotton wool in the mouths of test tubes, and the whole wrapped in filter paper, and left so until actually used. When the samples were removed for analysis, the liquid was at the same time tested for sterility in the way already described. The culture liquids employed were sterilised sewage and sewage effluents, and also sterilised artificial liquids whose composition was arranged to resemble that of sewage as nearly as possible. The sewage was usually allowed to stand for several days, then filtered, and after- wards sterilised by successive heatings at 100°C. The artificial culture media were the solution, A, referred to above, and also the same solution to which a small amount of glucose has been added (about 0:25 per cent.). The methods of analysis were those usually adopted in water and sewage analysis. In the estimation of the free ammonia a small portion of the liquid to be analysed (3 to 10 c.c.) was diluted with about 500 c.c. ammonia free water and distilled. Three successive portions of 50 c.c. were distilled off, and the ammonia contained in them esti- mated by means of adding Nessler’s solution and comparing the yellow tint with that given by definite amounts of standard ammonium chloride solution. ‘“ Albuminoid ammonia” was afterwards estimated by adding a fixed amount of “alkaline permanganate” solution,* distilling as long as ammonia came over in the distillate, and esti- mating these amounts in the same way. It is possible that in some cases the quantities of albuminoid ammonia given in the analyses in the following tables may be slightly too low. For, in certain instances, towards the end of the distillation, the ammonia, came over, in the distillate, very slowly and in very small quantities. It was considered more accurate in these cases to discontinue the analysis when these amounts fell below a certain very small value, since the error in estimating such very small quantities becomes comparatively great. The presence of nitrates or nitrites was detected by means of diphenylamine-sulphuric acid and metaphenylene-diamine, while quantitative estimations were made by means of the “ copper-zinc couple” method in which the nitrogen present as nitrates or nitrites is estimated as ammonia.f To allow for any ammonia that might have been originally present in the solution or introduced as traces during the analysis, a control estimation was always made. This analysis was carried out in every way like the real one, except that no “couple” was introduced, and the ammonia thus obtained was sub- tracted from that found in the actual estimation. The analysis of the cultures of this alga by these methods is by no * © Volumetric Analysis,’ Sutton, 8th ed., p. 512. + Lbid., p. 482. 2 2 464 Miss H. Chick. On a Unicellular Green Alga, [Feb. 28, means easy, for the quantities treated are exceedingly small, the methods of analysis adopted demand a comparatively large margin of experimental error, and the length of time, over which many of the experiments must continue, is also a drawback. The sterilisation of flasks and pipettes must also introduce a slight error where such small quantities of ammonia are concerned. The estimation of the albuminoid ammonia is frequently rendered troublesome by the tendency of the liquid to bump while distilling. It has been found that this difficulty is usually obviated if the ammonia-free water used for dilution is well aérated, and if, from time to time, small quantities of fresh aérated, ammonia-free water are added during the distillation. The absorption of ammonia, as this alga grows in ammonia-contain- ing culture liquids, is well shown hy the results of the analyses in Tables I—VII. Atthe same time, the amount of albuminoid ammonia present steadily increases, and, on the whole, a fairly even balance is maintained, which can be seen by comparing the quantities of total nitrogen. The evident preference of Chlorella pyrenoidosa for its nitrogen as ammonia, rather than in an oxidised form, is seen from Tables II, III, IV. In Table III an impure culture of the alga was allowed to grow alongside the pure culture, and it will be seen that similar changes had been taking place in the former, although they were not so well marked. In the case of the pure culture, nearly all the free ammonia was absorbed in six weeks, a considerable amount of albuminoid ammonia was produced, while the nitrates and nitrites appear to have remained free from any attack on the part of the alga. In the case of the impure culture less free ammonia was absorbed, less albuminoid ammonia formed, and some of the oxidised nitrogen present in the solution had been assimilated. The presence of a small proportion of glucose (0°25 per cent.) has a very remarkable effect in stimulating the nitrogen assimilation of the plant (see Tables VI and VII). In these two cases, cultures were started in exactly similar solutions, except that to the one set the above small amount of pure glucose was added. In this case (Tables VI 6 and VII 0), practically all the free ammonia was absorbed, and largely converted into albuminoid ammonia in an extremely short time, when compared with the cultures grown in the solutions containing no sugar. The glucose cultures showed evidence of much more grow th, and the ool algal cells were also much paler in colour. It appeared to be a point of interest whether this albuminoid ammonia, so invariably formed in these cultures of Chlorella pyrenoi- dosa, corresponded to nitrogenous substances formed within the plant cell and kept there, or whether such substances were formed, and then 1905.] with especial Reference to its Nitrogenous Metabolism. 465 allowed to go free in the liquid.- With the object of settling this point, cultures which had mannfactured a good deal of albuminoid ammonia, were subjected to centrifugalisation, and analyses: made of the clear solution (see Tables VIO (analysis No. 3) and V). The clear solution in either case contained no more albuminoid ammonia than the control, proving that the elaborated nitrogenous substances were entirely retained within the cell body. In two cases, however (Tables VIO (analysis No. 4) and VII), older cultures were examined. The word “older” is here used to imply that these cultures had passed through many more cell-generations, judging from the appearance of the growth and the size of deposit present. In these cases, a considerable amount of albuminoid ammonia was found to exist in the clear liquids. This seemed to prove that the algal individuals, under certain conditions, yield up to their culture liquid somewhat complicated nitrogenous matters in a soluble form, and a probable explanation would be that they exhibit this phenomenon when the cells are in a dying or dead condition. This fact assumes a certain significance when it is remembered that the natural habitat of this alga is ammoniacal liquids, hence notably polluted waters. ‘The actual amount of assimilation and excretion is extremely small, and it would be difficult to assert that the presence and growth of such an alga as Chlorella pyrenoidosa was of very great importance in nature. But the effect of this alga, if present, would probably be to leave the water, in which it has grown, in what would be termed a “more impure” condition. It is evident from the above experiments that the organism possesses the faculty of converting saline ammonia into albuminoid ammonia, which, under certain conditions, is discharged from the cells into the culture fluid. “ Albuminoid ammonia” is a name given to a certain class of sub- stances on account of facts connected with their analysis; these substances are quite unstudied, and, in this instance, may be perfectly harmless compounds. But the presence of albuminoid ammonia to more than a very small degree has always been considered, perhaps without sufficient foundation, to show evidence of dangerous pollu- tion, and hence to be most prejudicial to any water. The formation and excretion of such compounds by a green alga appears to be a new phenomenon, and is not without especial imterest in this connection. . An attempt was made to compare the nutritive value for Chlorellu pyrenoidosa of various nitrogenous substances, and with this object a series of nutrient fluids were prepared of the same composition as those used by Kriiger* in his experiments with Chlorella protothecoides and Chlorotheciwm saccharophilum. : * Kriger, Zopt’s ‘ Beitrige z. Phys. u. Morph. Nied. Org.,’ Leipzig, vol. 4, p. 103. +66 Miss H. Chick. On a Uncellular Green Alga, [Feb. 28, A stock solution was made having the following composition :— Glucose 1:0 per cent., K2HPO, 0-2 per cent., MgSO, 0-04 per cent., CaCl, 0°02 per cent., distilled water 100 c.e. With this stock solution eight different culture solutions were pre- pared containing respectively 0-1 per cent. of the following :— 1. Asparagine. 5. Peptone. 2. Aspartic acid. 6. Xanthin. 3}, (Uneees 7. Hippuric acid. 4. Uric acid. 8. Stock solution, without further addition. Of these solutions, the cultures grown in those containing urea and uric acid were found to flourish exceedingly well, and by far the best, while of the others, the culture in the liquid containing xanthin gave evidence also of being very well nourished. Uric Acid Cultuwre—An analysis was made of this culture in the ordinary way. In the case of the control liquid, it was found that no free ammonia was obtained, and comparatively very little of the nitrogen of the uric acid was yielded as albuminoid ammonia. It was found, however, that when Chlorella pyrenoidosa had been growing in the liquid, the yield of albuminoid ammonia was much increased, showing that much of the nitrogen of the uric acid had been assimi- lated by the organism and converted into albuminoid ammonia. Urea Culture.—In the case of the control urea-containing liquid, about 25 per cent. of its nitrogen was yielded, on analysis, as free ammonia. In the case of the inoculated fluid, about 30 per cent. of this ammonia was found to have disappeared, that is, to have been assimilated. After addition of alkaline permanganate, distinctly more ammonia was yielded from the liquid in which alga had grown than from the sterile control. It is difficult to draw very positive quantitative conclusions from the results of the above two experiments, but it seems evident at least: that this alga is able to assimilate with very great ease and without any previous decomposition the nitrogen both of urea and uric acid. The former substance has also been found by other observers* to be a useful source of nitrogen for plants in certain instances. In the case of Chlorella pyrenoidosa, however, these facts are of special interest, when it is remembered that the plant naturally grows in water that has been contaminated by sewage. * Cameron, ‘Trans. Brit. Assoc.,’ 1857; Ville, ‘Compt. Rend.,’ vol. £5, p. 32. i a ee ee 1903.] with especial Reference to its Nitrogenous Metabolism. 467 Table I.—Showing Assimilation of Ammonia. Culture in Solution “A.” Started October 21, 1901. | Culture of | Control. Chlorella pyrenoidosa. No. of analysis. | Lele © | — 2 SAG, an Se Naar este | analysis, | ea Per cent. Per cent. | nitrogen. nitrogen. 1 21.10.1901 | Free ammonia] 6 -00209* ai 2 26.10.1901 | Free ammonia) 0 -00183* 0 -00144# 3 | 211.1901 | Freeammonia) 0 -00152* 0 -00114* ena / | hes aif Sahat 4 28.11.1901 | Free ammonia| 0-00194+ 0 000714 | | Albuminoid 0 00062 0 -00081 | ammonia | Nitrites 0 0 | Nitrates O 0 * Hstimated by direct Nesslerisation. + Estimated, after distilling, by Nesslerisation. Table II.—Showing the Assimilation of Ammoniacal Nitrogen in preference to Oxidised Nitrogen. Culture in a Sterilised Sewage Effluent (Land). fal | | | Control. Culture. | { | | H Date of analysis.) — | | Percent. | Per cent. | nitrogen. | nitrogen. 31.10.1901 resmammoniarneedecse sso. O 000142.) | — Albuminoid ammonia........| trace —- 712.1901 | Free ammonia............-.| 0°00009 | 0 Albuminoid ammonia .......| 0:°00007 | 0700025 INMBEMIEE! Sonn da omod cameo voln) O) (Oral 0°00213 INMATE 6 6 05 Uo CO oa go.od0e Oe 0 | Faint trace | | Nitrogen— | Total (by addition) ........ 0 -00231 | 0700238 468 Miss H. Chick. On a Unieellular Green Alga, Table II].—Showing the Assimilation of Ammoniacal Nitrogen in Preference to Oxidised Nitrogen, and Formation of Albuminoid Ammonia. Culture in Sterilised Sewage, which had stood for some time. November 8, 1901. Started Date of analysis. | Free ammonia | Albuminoid ammonia 8.11.1901 18.12.1901, to 20.12.1901 | Albuminoid | ammonia | Nitrites | Nitrites and | nitrates ‘Total nitrogen | ee (by addition) Free ammonia | | Culture of Chlor. pyrenoi- Pure culture ! dosa contami- Control. of Chlorella | nated with the pyrenoidosa. bacteria which | usually accom- | . | pany it. | | | Per cent. Percent. | Percent, | nitrogen. nitrogen. nitrogen. 0 °00124 _- — 0 -00023 — — nik ! | | | | 0 °00141 0 :00080 0°:00058 | 0 °C0025 0°00105 0 :00097 Small reaction.| Small reaction) Small amount | 0 :0014.2* 0 :00135+ | 0 :00119 | 0 00270 0:60274. | | | 0-00308 | | * Free ammonia was got rid of by boiling before the addition of the ZnCu couple. + Probably under estimated. [Feb. 28, 1903.] with especial Reference to its Nitrogenous Metabolism, 469 Table 1V.—Showing the Assimilation of Ammoniacal Nitrogen in Preference to Oxidised Nitrogen, and the Formation of Albuminoid Ammonia. Culture in Sterilised Sewage. Started October 31, 1901. \ | i | | Pure culture | | Control. of Chlorella | _ pyrenoidosa. me. ae | Mi | _ pyvenoidosa a ae aaa | analysis, | | [Eiger | Per cent. Per cent. nitrogen. nitrogen. il 31.10.1901 | Freeammonia! 0°00247 ae | Albuminoid - 0°00041 — | ammoma 2 20.10.1901 | Free ammonia _ 0 °00231 0 :00237 and Albuminoid 0:00072 =| 0 -00059 21.10.1901 ammonia | | | | 3 | 20.12.1902 Freeammonia| 0:00276 | 0:00082 21.12.1902 Albuminoid 0 :00063 | 0 :00230 | and ammonia | 22.12.1902 | Nitrites Small reaction | Small reaction, | | | | but greater. | Nitrites and 0:00128 | 0 °001381 | nitrates | | | Total nitrogen | 000467 000443 | (by addition) | | A470 Miss H. Chick. On a Unicellular Green Alga, [¥Feb. 28, Table V.—Showing the Assimilation of Ammoniacal N itrogen, Forma- tion of Albuminoid Nitrogen, and the Results of Centrifugalisation Experiments. Culture in Sterilised Sewage, containing no Oxidised Nitrogen. Started August 4, 1902. | | Culture of Chlorella pyrenoidosa. | | | Control. Clone | No. of Date of Whole solutionafter| analysis. | analysis. et culture. centri- | fugalisation. | | eee af Per cent. Per cent. Per cent. nitrogen. | nitrogen. nitrogen. i 4.8.1902 | Freeammonia| 0 :00251 Albuminoid 0°00033 ammonia Nitrites and 0-00 — — nitrates Total nitrogen 0° 00284: (by addition) | eye oe eB alae Ca a ee ose 2 | 9.10.1902 | Freeammonia! 0:00258 | 0-00145 | Albuminoid 0:°00032 | 0O-°O00068* ammonia | Nitrites and 0-00 | 0-00 rue nitrates Total nitrogen 0 -00290 0 -00213 3 {14.10.1902| Free ammonia} 0-00233 0-00131 0 00115 and Albuminoid 0 -00029 0°00108 0 -00041 15.10.1902 | ammonia Nitrites and. 0:00 0:00 0°00 nitrates Total nitrogen 0 -00262 0 -00239 0 °00156 rf Bape eos a [see | iis ten ee a 4 411.1902 | Free ammonia} 0-00212 | 0-00049 0 -00041 | 5 18.12.1902 | Free ammonia 000013 0 :00012 | | Albuminoid 000182 0:00058 | ammonia Nitrites and ea 0:00 0-00 | nitrates | | Total nitrogen | 0 00194 0 -00070 * Probably under estimated. = CT i eet L<. Se ——eeEeEeEeEeEeEeEeEeeee - ~ 1903.] with especial. Reference to its Nitrogenous Metabolism. 471 Tables VI (a) and (b).—Showing increased Nitrogen Assimilation when Glucose is present in the Culture Fluid, and the Result of Centrifugalisation Experiments, in the latter instance, both in the case of Younger and Older Cultures. Table VI (a).—Culture in Solution “ A.” Started November 3, 1903. | | | Culture of | Control. | Chlorella . Dhiovet | | | pyrenoidosa, | o. of analysis. ; —— | | analysis. | | Percent. | Per cent. | nitrogen. nitrogen. 1 3.11.1902 | Freeammonia| 0 -00216 | Albuminoid 0 -00042 | ammonia Nitrites and 0-00 | — nitrates _ Total nitrogen 0 00258 | (by addition) rs 2 a dee ia L 2 21.11.1902 | Free ammonia} 000231 0 -00202 and | Albuminoid 0 -00040 0 -00039 fez. 1902 | ammonia | | Nitrites and 0-00 0-00 | nitrates | Total nitrogen 0 ‘00271 | 0:00241 Miss H. Chick. Ona Uiucellular Green Alya, [Feb. 28; Table VI (6).—Culture in Solution “A,” to which a small amount of Glucose had. been added (about 0:25 per cent.). ber 3, 1902. ——— — — 1 | 3.11.1902 | 25.11.1902 | 22.12.1902 | Albuminoid 21.11.1902 24.11.1902 | and Started Novem- Free ammonia Albuminoid ammonia Nitrites and nitrates Total nitrogen (by addition) Free ammonia Albuminoid ammonia Nitrites and nitrates Total nitrogen (by addition) Free ammonia | Albuminoid ammonia Nitrites and nitrates Total nitrogen Free ammonia ammonia Nitrites and nitrates Total nitrogen Culture of Chlorella pyrenoidosa. | Control. Glee | Whole {solution after) culture. centri- | fugalisation. | a Per cent. Per cent. Per cent. nitrogen. nitrogen. nitrogen, 0 -00241 0 -00045 0-00 — = 0 -00286 000046 J 0 -00122 o o 2 0-00 = o po) z x 0 00168 BO see Sel ao 0 00021 0 -00025 =n 0 -00182 0 -00025 Se 3s so 0 00173 0°00050 Be 000084 | 0 -00032 =| 0 -00160 0 00058 s 0-00 0-00 = 000194 0 :00090— 1903.] with especial Reference to its Nitrogenous Metabolism. 473 | ! Tables VII (a) and (b).—Showing increased Nitrogen Assimilation when Glucose is present in the Culture Liquid, and the Result of Centrifugalisation Experiments in the latter case. Table VII (a).—Culture in Solution “ A.” Started August 5, 1902. | | | Culture of | | Control. Chlorella | hae | pyrenoidosa. | No. of analysis. = analysis. | | | Per cent. Per cent. | | | nitrogen. nitrogen. | | } ! 1 5.8.1902.' | Free ammonia 0 -00268 | Albuminoid 000012 (?) ammonia | | | Nitrites and 0:00 _ | | nitrates | | Total nitrogen 0 -00280 | (by addition) | 2 | 19.11.1902. | Freeammonia} 0°00221 000166 | Albuminoid 9 00089 0 -00091 ammonia | Nitrites and 0:00 0-00 | nitrates | Total nitrogen 0°00260 0 °C0257 474 Miss H. Chick. Ona Unieellular Green Alga, [Feb. 28 Table VII ().—Culture in Solution ‘ A,” to which a small amount of Glucose had been added (about 0°25 per cent.). Started August 5, 1902. Culture of Chlorella | pyrenoidosa, | | foe Control. | | Glens _ No. of of Dy | Whole | liquid after analysis. | . .alvsi | culture. | centri- er | fugalisation | chee ; | ie | | Percent. | Percent. | Per cent. | nitrogen. | nitrogen. nitrogen. | | | | | | | | | | i 1 5.8.1902 | Free ammonia 000254. | Albuminoid 0 -00015 (2), | | | ammonia . a | Nitrites and 0-00 nitrates Total nitrogen 000269 | 2 23.10.1902] Free ammonia) 0:00281 | 0:00070 | Albuminoid 0:00022 | 0:00141 ammonia | | : | Nitrites and 0-00 r 0:00 ee ! nitrates ! | Total nitrogen 0°00253 | 0-00211 3 |20.10.1902| Free ammonia}; 0°00210 — 0:00074 | 0-00063 | Albuminoid 0°:000384 = =0°00166 = = 0:00081 | ammonia | | | | Nitrites and 6:00 |’ 30)-00 0:00 nitrates | | | Total nitrogen, 0°00244 = 0:00240 0-00144. | General Conclusions. It appears to be generally true that most plants containing chloro- phyll prefer the nitrogen of their food in the form of nitrates. On the other hand, many observers have shown that certain plants can assimilate nitrogen in the form of ammonia, and in fact prefer it. Of these latter experiments, many were not performed with the precautions necessary to prevent the access of bacteria (¢.g., nitrifying organisms), and hence must be considered inconclusive. Some, on the contrary, were most carefully carried out with every precaution, and among these may be mentioned those of Kriiger,* who showed that his Chilo- rella protothecoides and Chlorothecoum saccharophilum (two alge nearly allied to Chlorella pyrenoidosa) could both assimilate their nitrogen * Zopt’s ‘ Beitrige z. Phys. u. Morph. Nied. Org.,’ Leipzig, 1894, vol. 4, p. 115. we 1903.] with especial Reference to its Nitrogenous Metabolism. 475 when offered in the form of ammonium salts, while the latter alone could assimilate nitrate nitrogen. Artari,* working with the gonidia of two lichens, isolated in pure culture, showed that, after peptone, asparagine and ammonium sulphate were the useful sources of nitrogen for these alge, and other observerst have made equally careful experi- ments in the case of some of the higher plants, and have shown that not only simple ammonium salts, but also compound ammonias, such as methylamine, ethylamine, can readily be assimilated. From a physiological point of view Chlorella pyrenoidosa must be included in this second class, for it has been shown, as the result of quantitative as well as qualitative experiments, that this alga prefers its nitrogen to be presented to it in the form of ammonia or ammo- niacal compounds. Among the latter urea, uric acid, &c., rank high in nutritive value. It would also appear from the foregoing chemical experiments that this ammonia, after being absorbed by the cell, is elaborated into albuminoid ammonia,{ a term used to describe certain nitrogenous bodies of ammoniacal nature, which yield ammonia when boiled with alkaline permanganate of potash; in fact, almost all the nitrogen assimilated would appear to remain in this comparatively simple form. ‘This nitrogen, for example, is more easy of attack than that contained in uric acid. These elaborated nitrogenous compounds appear to be retained wholly within the cell body, but under certain conditions, only observed in the case of “older” cultures, they seem to escape from the cell, and can be traced free in the liquid. The presence of glucose in a culture liquid frees the alga from the necessity it would otherwise experience of manufacturing carbohydrate for itself. The algal cell, being thus relieved of a certain part of its ordinary work, appears to be enabled to reproduce itself much faster, and its nitrogen assimilation is also much increased, though, owing to the increased multiplication of cells, it would be impossible to say that the nitrogen assimilation per cell was increased. The chlorophyll body of the cell, at the same time, gives evidence that its function has been interfered with by a most striking change in form and in the amount of chlorophyll. It is distinctly noteworthy that neither cane- sugar nor lactose can be substituted for glucose in this connection. The effect of glucose in causing a definite change in the chlorophyll body, and in generally stimulating growth, would not appear to be an isolated fact. It has been shown by other observers in the case of three green alge (Chlorothecium saccharophilum, Chlorella protothecoides, Sticho- coccus bacillaris), upon whose nutrition the effect of glucose was * *Bull. d.1. Soc. Imp. des Nat. de Moscou,’ 1899, p. 39. + Laurent, ‘Ann. de l’Inst. Past.,’ vol. 3, 1899; Lutz, ‘Comptes Rendus,’ vol. 126, p. 1227. ~ Wanklyn, ‘Chem. Soc. Journ.,’ 1867, p. 59, A76 | On a Unicelluiar Green Alga. [Feb. 28, studied in pure culture, that there was a marked increase in growth when this carbohydrate was added to their culture medium, though no quantitative experiments were made to measure any difference in assimilation. The general features of the assimilation of nitrogen, on the other hand, and its subsequent history displayed by Chlorella pyrenoidosa, may bea specialised characteristic of this plant, having reference to its continual occurrence, in nature, in waters which contain a compara- tively large amount of ammonia, notably in sewage and sewage- polluted waters. It is, however, impossible to express a definite opinion upon this point until the nitrogen assimilation of a series of many other plants has been studied in a similar manner. DESCRIPTION OF PLATE 8. Fig. 1.—Culture on nutrient agar (1 month). »» 2.—Culture on glucose nutrient agar (1 month). ,, 8.—Cells from a culture on nutrient agar of 10 weeks. 5, 4.—Cells from a culture on glucose agar of 3 weeks. », 5.—Cells dividing, from a culture on nutrient agar of 10 weeks. 3 : g ~ E x 4 x is > a Ss) d S iG a) . a Sy S A fioy. Chick del. 1903.] On Lagenostoma Lomaxi, the Seed of Lyginodendron. 477 “On Lagenostoma Lomaai, the Seed of Lyginodendron.” By Pay OnE Doe. KTS. and D. Hs. Scorm MA. Ph.D., F.R.S. Received March 19,—-Read May 7, 1903. The existence in Paleozoic times of a group of plants (the Cycado- filices of Potonié) combining certain characters of Ferns and Gymno- sperms, has been recognised for some years past by various palzeo- botanists.* The group in question embraces a number of genera, among which Medullosa, Heterangiwm, Calamopitys and Lyginodendron may be mentioned; the fern-like foliage of these plants is placed according to its external characters in the form-genera Alethopteris, Neuropteris, Sphenopteris, and others. The evidence for the intermediate position of the Cycadofilices is extremely strong, but at present it is drawn entirely from a detailed comparison of their vegetative organs, especially as regards their anatomical characters. In no case, as yet, 1s the fructification of any member of the group known with certainty ; such indications as have hitherto been detected are still in need of corroboration. Thus, the suggestion has been made that the large seed, Vrigonocurpon olivwforme, may have belonged to some member of the genus Medullosa ;— and in the case of Lyginodendron itself there is fairly strong reason to believe that one form of fructification (in the light of the observations to be described below, presumably the male) may have been of the Calym- matotheca type,t a type, however, of which the organisation is not yet fully understood. In the absence of satisfactory data as to the fructi- fication, so high an authority as M. Zeiller has expressed a doubt whether the Cycadofilices were, after all, anything more than a specialised group of Ferns.§ A re-examination of the seeds, placed by Williamson in his genus Lagenostoma, has revealed unexpected points of agreement between the structure of the envelopes of certain of these seeds, on the one hand, and that of the vegetative organs of Lyginodendron on the other. Two species of Lagenostoma (L. ovvides and L. physoides) were * Williamson, ‘“‘ Organisation of the Fossil Plants of the Coal-measures, Part XIII,” ‘Phil. Trans.,’ B, vol. 178, p. 299: 1887; Solms-Laubach, ‘ Fossil Botany,’ 1887, Engl. ed., pp. 141 and 163; Williamson and Scott, ‘‘ Further Observations on the Organisation of the Fossil Plants of the Coal-measures, Part III,” ‘ Phil. Trans.,’ B, vol. 186, p. 769, 1895; Potonié, ‘ Lehrbuch der Pflanzenpaleontologie,’ p. 160, 1899; Scott, ‘Studies in Fossil Botany,’ pp. 307 and 514, 1900. + G. Wild, “On Trigonocarpon oliveforme,’ ‘Manchester Geol. Soc. Trans.,’ vol. 26, p. 484, 1900. £ Scott, ‘Studies’ p. 334; Miss Benson, “ The Fructification of Lyginodendron Oldhamium,’ ‘ Aun. of Bot.,’ vol. 16, p. 575, 1902. § ‘Zeiller, ‘ Eléments de Paléobotanique,’ 1900, p. 370. VOL LXXI. 2 M 478 “Drs. F2W) Oliver*and™): He Scot: [ Mar. 19, described by Williamson ;* a third species, the subject of the present note, was left undescribed by him, though in his MS. catalogue he named it, after its discoverer, Lagenostoma Lomazi, a name which we here provisionally adopt. This seed occurs in calcareous nodules of the lower Coal-measures, and chiefly at Dulesgate, in Lancashire. In general structure the seed L. Lomaai agrees with L. ovoides. It is an orthotropous seed, circular in transverse section, and broadest midway between base and apex. The height of the seed slightly exceeds the diameter, and in general form it may be com- pared with a Jaffa orange. Its height in full-sized specimens is about 53 mm., the diameter at the equator 4, mm. Many of the specimens that have passed through our hands show signs of having become detached through the agency of a layer of separation and bear a low © conical papilla centrally placed at the chalazal end, beneath which the actual layer of abscission was situated. In the most general relations of its organisation the seed approaches the Gymnosperm type in that the integument and nucellus are distinct from one another in the apical region only, whilst the body of the seed, which contains the large single macrospore with traces of pro- thallial tissue, shows complete fusion of the integumental and nucellar tissues. But in other respects the seed is remarkable. The free portion of the nucellus which stands above the macrospore is conical in form; its base is about 0°75 mm. across, and its height some- what greater. The tapering apex reaches to the exterior, plugging the micropylar aperture like’ a cork. The whole of this structure, the ‘“lagenostome” of Williamson, constitutes a pollen-chamber, owing to the separation of the nucellar epidermis from the underlying parenchymatous body of the free part of the nucellus. The pollen- chamber thus has the form of a bell-shaped cleft situated between the persistent epidermis and the central cone of nucellar tissue. Access to the chamber is gained at the apex, which is open, and pollen- grains are found in its lower part. The integument, which is a simple shell where fused with the nucellus, becomes massive and complicated in its free part, which corresponds to the upper fifth of the seed. In this region it is usually composed of nine chambers radially disposed around the micropyle. The existence of these chambers is indicated on the outside surface of the seed by the presence of nine little ridges dis- posed like the rays of a star around the micropyle, but dying out almost at once. These ridges over-le the partitions of the chambered portion of the integument just as do the stigmatic bands the septa of a poppy capsule. The whole structure from within is like a fluted dome or canopy, the convexities of which correspond to the chambers, * “ Organisation,’ Part VIII, ‘ Phil. Trans.,’ vol. 167, p. 233, figs. 53—75 and 77—79, 1877; Pars X,‘ Phil. Trans.,’ Part II, 1880, p. 517, figs. 61—63. See Oliver, ‘New Phytologist,’ vol. 1, p. 145, 1902. 1903.] On Lagenostoma Lomaxi, the Seed of Lyginodendron. 479 and actually engage with broad low grooves on the surface of the wall of the pollen-chamber. The vascular system of the seed enters as a single supply-bundle at the chalazal papilla, and branches, a little below the base of the macro- spore, into nine radially-running bundles. ach of these bundles passes, without further branching, to the apex of the seed, running outside the macrospore and a little distance below the surface. At the canopy the bundles enter the chambers and end at the tips. ~ Lagenostoma Lomaaxt was thus a seed or seed-like structure, detached as a whole and containing pollen-grains in the remarkable cleft-like pollen-chamber ; the integument in its free part, when compared with that of Wilhamson’s Lagenostoma physoides, suggests a number of originally free arms or processes that have become laterally fused into a complex, chambered organ. | The seed, L. Lomax, is in some cases still attached to its pedicel ;* the great peculiarity of this seed, as compared with other members of the genus, is that when young, and sometimes even at maturity, it is found enclosed in an envelope or cupule, springing from the pedicel just below the base of the seed, and extending above the micropyle— at least in young specimens. The cupule appears to have been ribbed below, and deeply lobed in its upper gart; in form it may be roughly compared to the husk of a hazel-nut—of course on a very small scale. The pedicel and cupule bear numerous capitate glands, of which some are practically sessile, others shortly stalked, while in others again the stalk is of considerable length. The head, or secreting portion of the gland, which is spherical in form, is almost invariably empty, only the multicellular wall persisting. The tissue of the stalk of the gland, consisting of many layers of cells, is preserved, though in a somewhat disorganised state. These cupular glands present the closest agreement in size, form, and structure with the glands which occur on the vegetative organs of Lyginodendron Oldhaniwm,y and which are especially abundant on the particular form of that plant found in association with Lagenos- toma Lomaxi. Both on petiole and cupule the majority of the glands are short, those which are not sessile being commonly about 0-4 mm. in height. lLong-stalked glands, exceeding a millimetre in height, some- times occur both on the vegetative organs and on the cupule. The dimensions of the head of the gland agree exactly on cupule and petiole, the diameter averaging about 0°2 mm. in each case. In both the stalk is usually somewhat narrower than the head, except at the base, where it is oiten considerably enlarged. On the stem, as might * Of. Williamson, luc. cit., Part VIIL, fig. 68 (L. ovoides). + It has long been realised that the name Lyginodendron Oldhamium charac- terises a type rather than a species. It is probable that the very glandular form occurring at Dulesgate may deserve specific rank. 480 Lagenostoma Lomaxi, the Seed of Lyginodendron. [Mar. 19, be expected, the glands are usually somewhat larger than on petiole or cupule. . As a rule, the structure of the glands on the vegetative organs is well preserved, the secretory tissue in the head being perfect. But occasionally the vegetative glands are found in the same state of preservation as those on the cupule, with the head hollow, owing to disappearance of the secretory mass. Where we thus have the two organs in a corresponding state of preservation, the agreement between the vegetative glands of Lyginodendron and those on the cupule of Lagenostoma Lomazi is found to be exact. There is no other known plant from the Coal-measures with glands at all similar to those described, nor is it likely that any unknown Gymnosperm should so exactly resemble Lyginodendron in these charac- | ters. On the ground, then, of the glandular structure we are led to the conclusion that the seed Lagenostoma Lomazi can have belonged to no other plant than Lyginodendron Oldhamium, and more particularly to the glandular form of that type with which the seed is associated. The state of preservation of the glands and of the cupule as a whole, indicates clearly that this organ, as we find it, was in an effete condition, having, no doubt, already discharged its functions while the seed which it protected was still quite young. The vascular system of the cupule was well developed, and is very fairly preserved. A number cf bundles branched off from the main strand of the pedicel, and traversed the cupule throughout its whole extent. The structure of the large bundle, seen in the pedicel, agrees with that of a petiolar strand in Lyginodendron. The minute charac- ters of the tracheides are also in close agreement with those observed in the xylem of the foliar organs of the same plant. Hence, characters presented by the internal anatomical structure strengthen the conclusion drawn from a comparison of the glands, and thus further support the attribution of Lagenostoma Lomaai to Lygine- dendron. The evidence thus indicates that in a transitional type, such as Lyginodendron Oldhamium, with leaves wholly fern-like in structure and form, but with decided Cycadean as well as Filicinean characters in the -anatomy of stem and root, the seed habit had already been fully attained, as fully, at any rate, as in any known Paleozoic Gymnosperm. Lyginodendron retains, so far at least as its vegetative structure is concerned, the intermediate position already assigned to it, but, whereas the fern-like characters have hitherto seemed to preponderate, the discovery of the seed inclines the balance strongly on the Gymno- spermous side. It is not likely that Lyginodendron stood alone in this ; we must now be prepared to find, what has long been recognised as a possibility, that many of the plants grouped under Cycadofilices already possessed seeds, and thus that a considerable proportion of the 1903, ] On the Action of the Poison of the Hydrophide. 481 so-called ‘‘fern-fronds” of the Paleobotanist really belonged to Sper- mophyta. It is at present impossible to say at what stage in the evolution of the Fern-Cycad phylum, the great change in reproductive methods came, whether it followed in the wake of general anatomical advance, or vice versd. The discovery of further evidence as to the. reproductive processes of these ancient plants may be expected to yield interesting results. } The authors are much indebted to Miss Marie Stopes for her valuable aid in the examination of the numerous sections in the Williamson and various other Collections. Mr. James Lomax deserves high praise for his good judgment and skill in collecting and preparing the material for the investigation. A full account of the fossils dealt with in the present note is in preparation, and will shortly be submitted to the Royal Society. “On the Physiological Action of the Poison of the Hydrophide.” By Leonarp Rogers, M.D., B.S. (Lond.), M.R.C.P., F.B.C.S., lately officiating Professor of Pathology, Medical College, Caleutta. Communicated by Major A. Atcock, F.R.S. Re- ceived March 31,—Read May 7, 1903. It has long been known that the great group of the Hydrophide, or Sea-snakes, are poisonous, and cases of death produced by their bites have been recorded, for example, that in Sir Joseph Fayrer’s work on the Poisonous Snakes of India, of the ship’s captain bitten while bathing in the Bay of Bengal, with a fatal result. The fishermen on this coast are also well aware of the danger of the bites of these reptiles, and take such good care to avoid them, that deaths among them are quite uncommon as far as I can ascertain. Deaths, however, not very rarely occur among those employed in oyster fisheries in shallow water in some places on the Madras coast, owing to snakes being trodden on, so that a study of the nature of the poison of this class of snakes has a practical as well as a scientific side, and, as far as I can gather from the literature of the subject obtainable in Calcutta, it has not yet received much attention. During the last year I have been investigating the subject, and although the amount of poison I have been able to obtain has been very small, yet it has sufficed to allow of certain definite results being obtained, which will be summarised in the following paper. The Collection of the Potson. The Hydrophide are met with in large numbers all round the coasts of the Indian peninsula, and have been specially studied at Puri on the east coast in Orrisa. It was at this place that I obtained 482 Dr. L. Rogers. On the Physiological [ Mar. 31, my specimens, which are caught by the fishermen in their nets during the calm cold-weather months with a frequency which is in propor- tion to the number of fish taken. By small payments they were induced to bring them to a tank which I had constructed near the beach, in which they usually only lived a few days, although some survived several weeks. By making them bite on a watch-glass covered with a thin layer of guttapercha tissue stretched tightly across it, they eject the poison into the glass as clear drops free from all saliva. This is then dried over calcium chloride or strong sulphuric acid, and can then be kept indefinitely in dry well-corked glass tubes, without losing its potency. The snake which is met with in greatest abundance in Puri is the Hnhydrina Bengalensis, measuring from three to five feet in length, and it has a thick body and a large head. This species also furnishes the largest amount of poison, and from this alone have I yet been able to obtain a sufficient quantity to allow of a con- siderable number of experiments being performed with it. That ot four other species, belonging to three different genera, has also been obtained in small quantities, so that four out of the six genera of Indian Hydrophide have now been examined, and will be dealt with. Appearance and Quantity Ejected. When the clear watery drop of poison is dried, it forms white shining scales, freely soluble in water or normal salt solutions, and differimg from the poisons of Cobra and Daboia by the absence of the yellow tinge of the latter. The only exception I have met with was a faint yellow tinge in the dried poison of a Disieira cyanocincta, the others having all been colourless. The quantity of poison ejected at a single bite is of great import- ance in relationship to the deadliness of these snakes, and fortunately it is very small. In many of the smaller species it is often impossible to get a drop at all, but probably when free in the water they can eject more poison than when being held close behind the head, with conse- quent great limitations of their power of motion. The amount of dried poison obtained from a single bite of thirteen different fresh specimens of the Enhydrina was weighed, and the average quantity was found to be 0:0094 gramme, or almost one centigramme. This is very much less than that obtainable from a Cobra or a Daboia, for the average amount of poison (dried) obtained from a Cobra is, according to D. D. Cunningham, 0°:254 gramme, or twenty-five times as much as is obtained from an Enhydrina. In fact, so small is the amount, that at the end of a season I had only been able to obtain about one-third of a gramme of the latter poison, and for most of which I am greatly indebted to Dr. Reid of Puri. The poison also appears to be slowly formed, as a week after a snake had been made to bite, it is usually 1905.] Action of the Porson of the Hydrophide. 483 impossible to get any further poison from it, even if it bite vigorously. Yet if made to bite a small fish immediately after ejecting his poison, the bite is fatal in a short time, showing how fatal a trace of it is. The largest amount of poison obtained at a single bite was 0:023 gramme. ‘The other species mostly gave a smaller quantity than the Knhydrina. Effect of Heat on the Poison. On boiling a dilute solution of the poison, it becomes slightly opalescent. After being boiled for 15 seconds, two minimal lethal doses were recovered from, after slight symptoms had appeared, but four minimal lethal doses proved fatal in a somewhat longer time than with unheated poisons. After boiling for one minute, four minimal lethal doses were recovered from after only shght symptoms. Thus the poison is readily destroyed by boiling for a short time, but merely bringing it to the boiling point does not materially affect its strength. Some similar experiments with Cobra poison show that the latter is slightly more resistent to heat than is that of the Enhydrina. Symptoms Produced by the Poison. The following symptoms are common to all the species yet tested, no differences having been met with, except with regard to the exact amount of the minimal lethal doses in different animals, which will be dealt with presently. Briefly, the symptoms produced by the poison of the Hydrophidz may be said to be identical with those caused by cobra venom, with one very important exception, namely, that the former venom has no appreciable action on the blood, which is a marked feature of cobra toxin. In the case of warm-blooded animals, such as rabbits, rats or birds, the symptoms produced by sea-snake poisons are as follows. When minimal lethal, or slightly supra- minimal lethal doses are given subcutaneously, there is always a long period before any symptoms of poisoning occur, the time varying in accordance with the dose from half an hour to several hours, in which respect it resembles Cobra, and differs markedly from Daboia venom. It large doses are given, the symptoms set in much earlier, and in that case death rapidly results. The symptoms are best studied by the use of small doses, when the first thing noticed is that the animal remains quietly in one position, and soon begins to show signs of drowsiness, closing its eyes at intervals. Next it begins to nod its head, but every now and then appears to wake up again and opens its eyes. In the case of birds—in which the symptoms can be best seen— the subject of the experiment next sits down on the floor of the cage, and although it can be made to stand up if disturbed, yet there is now evident commencing muscular weakness, and it can only walk with an —— —— 484 Dr. L. Rogers. On the Physiological [Mar. 31, unsteady gait. By the time this stage is reached, it will be found that the animal is breathing more deeply than normal, while the number of respirations is also increased to a variable, but often considerable, degree. From this time the picture is one of progressive paralysis, affecting all the muscles of the body, and ending with respiratory con- vulsions. The animal nods more and more deeply, until the nose or beak touches the floor of the cage, only to be raised again with a jerky motion. It is now unable to stand upright, and the eyes remain closed. The respirations are now very deep and laboured, and in case of birds, the beak is half open, and gaping takes place with every in- spiration, while the head is more and more lowered until its vertex instead of the beak rests on the floor, and the animal is unable to raise its head. Very soon after this stage of paralysis is reached, convul- sions set in, and the respirations immediately fall very greatly in frequency, while they remain deep in character, although less regularly so than before, some being shallow, so that Cheyne-Stokes breathing is somewhat simulated. The convulsions recur, and soon respiration entirely ceases, but the heart continues beating for some time, usually two or three minutes in the case of warm-blooded animals, after the breathing has entirely ceased. When the convulsions com- mence, the anima] rolls over on its side in a state of nearly complete paralysis. Every word of the above description of the symptoms pro- duced by the poison of the Hydrophide is equally true of Cobra poisoning, so much so that if two animals are severally given minimal lethal doses of these two poisons, it is impossible to distinguish which animal has received which poison by the clinical symptoms produced, a fact which I have repeatedly demonstrated. Post mortem, after death from the poison of the Hydrophide, there is little or nothing noteworthy feund. The seat of injection is free from extravasation of blood, and presents little or no serous effu- sion. The blood is of a dark colour, no doubt due to the respiratory paralysis. It is fluid on opening the heart, but rapidly clots when placed in a small test-tube, doubtless owing to the large amount of CO, gas init. On standing it exudes serum, which is usually clear, but may be very slightly blood-stained, although very much less so than in the case of Cobra-poisoning under the same circumstances. There is no intravascular clotting to be found post mortem in the portal or other veins, as C. J. Martin first demonstrated in Pseudechis poisoning, and as occurs in acute Daboia poisoning as recently shown by Lamb. No other naked-eye changes have been found after death from sea-snake poisons. In the case of cold-blooded animals, such as fish, which have fre- quently been used in these experiments, the symptoms are essentially the same in kind as in warm-blooded animals, although less easy to observe. After small doses there is the same long latent period, often 1903. | Action of the Poison of the Hydrophide. 485 lasting for several hours. Sometimes temporary excitement with rapid motion may be observed for a short time, but more often the picture is simply one of slowly progressing paralysis. In most kinds of fish this is also very well shown by a gradually increasing difficulty in maintaining the upright position, the fish slowly turning over on one side and then swimming up into its upright position again, only to slowly sink on to its side once more. The respirations will now be found to be deeper than normal, although not as a rule quicker, but, on the contrary, they steadily slow down from the beginning of the symptoms to the end without any marked increase in the rate. This paralysis of all the muscles and of the respirations steadily progresses until convulsions set in, to be immediately followed by a very rapid failure of the respirations both in number and depth, so that they become difficult to detect, and death soon follows. The heart will be found beating some time after the breathing has ceased, and no extravasation of blood or other noteworthy change is found post mortem. Here again the symptoms are precisely similar in poisoning of fish by Cobra venom. The Potency of the Poison. By working out the smallest fatal doses of the poison per kilogramme of weight in different animals, and comparing them with those obtained by former workers for other snake venoms, we shall be able to esti- mate the potency of that now being dealt with. This has been done in the case of the poison of the Enhydrina by means of numerous experiments carried out with the mixed dried venom of a number of these snakes, with the following results, At the same time com- parative experiments were also carried out with fresh dried Cobra venom for comparative purposes. White rats were first tested, and 0-07 milligramme per kilo weight was found to prove fatal, but smaller doses were sometimes recovered from. In the case of Cobra poison 0°5 milligramme per kilo was necessary to produce death, while Lamb in Bombay found the fatal dose of this poison for rats to be 0°33 milligramme. It is evident then that the poison of the Enhydrina is several times as potent as is Cobra venom on rats. In the case of rabbits only a few experiments have been performed, but 0-04 milligramme per kilo proved fatal in under four hours in one case, while in another 0°01 milligramme per kilo produced no symptoms but loss of appetite; but on giving a second dose of 0:02 milligramme per kilo five days later (the animal having fully recovered from the first dose in one day), death resulted in a few hours. On the other hand, Elliot found the minimal lethal dose of Cobra venom for rabbits to be 0°7 milligramme per kilo weight, so that it is evident that these animals are many times as suscep- 486 Dr. L. Rogers. On the Physiological [Mar. 31, tible to the poison of Enhydrina as to that of Cobra, the former poison being some twenty times as potent for them as the latter— a remarkable difference. A larger number of experiments have been carried out with birds, pigeons and fowls. These also bear out the former ones in proving the far greater potency of the poison of the Enhydrina over that of the Cobra or other poisonous snake yet examined. In the case of pigeons the minimal lethal dose, 0°05 milligramme per kilogramme, always proves fatal, while in fowls the fatal dose is 0-04. These figures may be compared with those obtained by D. D. Cun- ningham in his numerous experiments with Cobra venom on fowls, for which he found the minimal lethal douse to be 0°5 milligramme per kilo, so that the poison of the Enhydrina for birds is at least ten times as potent as is Cobra venom, which goes far towards neutralising the effect of the much smaller dose of poison ejected by the Enhy- drina as compared withthe Cobra. Taking the minimal lethal dose of the Enhydrina for warm-blooded animals as 0°05 milligramme per kilo, the fatal dose for an average man of 70 kilogrammes would be 3°5 milligrammes, or about one-third of the average amount of venom ejected by afresh fuil-grown specimen of this, by far the most com- monly met with, kind of snake in the Bay of Bengal. There is good ground, then, for the belief in the deadliness of the Hydrophide. The Minimal Lethal Dose for Fish. It is well known that it is necessary to give many times as large a dose of Cobra venom, in proportion to the weight of the animal, in order to kill cold-blooded animals as is required for destroying the life of warm-blooded animals. Now there is no doubt that the Enhydrina live on fish, and I have been able to ascertain that they can swallow those of considerable size. One specimen of Enhydrina after being handled in the process of taking poison vomited a piece of hali- digested fish, which on comparison with complete fish of the same kind was found tc have certainly been a foot or more in length, while it was over 2 inches in depth. Such a fish could not have been swallowed if it had not first been killed, or at least paralysed to a marked degree. It is of interest, then, to ascertain the minimal lethal dose of these snakes against fish, and to compare it with that of the Cobra. As I have not been able to find accurate records of the effect of Cobra venom on fish, [ have also ascertained this by a series of experiments, using the hardy Mud-fish (Saccobranchus fosszlis), which lives for weeks in. a small’ vessel of water. It was found that 25 milligrammes per kilo of Cobra venom had to be given to be certain of causing death, although sometimes a slightly smaller dose was effective. Thus fifty times as much Cobra venom is required to 1903.] Action of the Poison of the Hydrophide. 487 kill a fish as is sufficient to kill a warm-blooded animal—a very marked difference. On testing the same species of fish with the poison of the Enhydrina, it was found that 0°5 milligramme per kilo of freshly dis- solved poison was always fatal, and sometimes a smaller dose caused death. Thus the dose of this sea-snake poison required to kill fish was but ten times as much as the minimal fatal dose for warm-blooded animals, that is, considerably less than we found to be the case with Cobra poison. In other words, the poison of the Enhydrina is much more deadly than is Cobra venom for fish, even allowing for the greater potency of the former for warm-blooded animals, so that it appears to be specially adapted for the needs of the Sea-snake, which lives on fish, being in all about fifty times as potent for fish as is Cobra venom. This great concentration of the poison may be of considerable advantage to the reptile when dealing with such active prey as fish in their own element. This special affinity of the poison for fish was even more marked in the case of some of the other species tested. Thus, that of a single species of the Disteira cyanocincta was tatal to pigeons in doses of 0°5 milligramme per kilo, being thus considerably weaker than that of the Enhydrina, but only 1 milligramme per kilo was required to kill fish, that is but twice as much as was needed to kill birds. Similarly with the Disteira viperina the minimal lethal dose for pigeons was 0°5 milligramme, and for fish only 0°75, or but very little more. Again, the poison of the Hydrophis cantoris for both pigeons and fish was just the same as the last-mentioned species. Lastly, the poison of the Hydrus platurus killed pigeons in doses of 0:075 and fish in one of 0°25 milligramme per kilo, being thus very deadly for both cold and warm blooded animals. The above include four out of the six genera of Hydrophide found in Indian waters, so that, although the poison obtained from the last four species was from single specimens, and therefore cannot be taken as more than approxi- mately accurate, yet they suffice to prove that the Hydrophide as a class secrete very virulent poisons, which are specially poisonous to fish. It is also worthy of note that the two genera which proved to be most deadly to warm-blooded animals, namely, the Enhydrina and the Hydrus platurus, are just the two which the fishermen at Puri said were the most dangerous ones, as the accuracy of their statement points to actual experience in the human subject of their deadliness having been handed down among them. Some of the smaller species, however, probably do not eject sufficient poison to prove fatal, to adults at any rate, and hence are not so much dreaded by the fishermen. It will also be observed that the poison of the Lnhydrina Lengalensis is the most potent of those so far tested, while it also yields the greatest amount of poison, with the exception, perhaps, of the Misteira cyanocincta. eae a ee 488 Dr. L. Rogers. On the Physiological [Mar. 31, Effect of the Poison on other Cold-blooded Animals. I have not yet been able to test any extensive series of other cold- blooded animals to see if they are equally susceptible to the poison of the Hydrophide as fish are, but in one instance a frog weighing 30 grammes was injected with a dose of 0:2 milligramme per kilo, with the result that it showed well-marked symptoms of paralysis, but eventually recovered, so that it would appear to have been about as susceptible as fish. Some harmless snakes were injected with note- worthy results. Thus, two specimens of the Coluber fasciolatus were injected with doses of 10 and 50 grammes per kilo respectively, with the poison of the Enhydrina, with no ill effect, and the former received a second dose of 50 milligrammes per kilo three days after the smaller dose, equally without effect. Here we have a harmless colubrine snake withstanding 100 times the fatal dose for a fish and 1000 times that for a warm-blooded animal. Further, two specimens of the harmless green Whip-snake (Dryophis mycterizans) were tested, but in this species 25 milligrammes per kilo in one instance, and 15 in the other, each produced death in less than two hours, so that a smaller dose would nearly certainly have been fatal. This opens up a large question which must await further investigation. The Physiological Action of the Poison on the Blood. The striking similarity of the symptoms produced by the poison of the Hydrophide and by Cobra venom leads one to expect a similarity of action on the blood. The researches of Cunningham have shown that Cobra poison has a very marked power of dissolving the red corpuscles of the blood and also in reducing its coagulability, and, contrary to the views of Lauder Brunton and Fayrer, he holds that these blood changes are the essential features of the action of the poison, and notits action on the nervous system, as held by the latter authors. Experiments have been carried out to test the effect of the poison of the Hydrophide on the blood, with unexpected and im- portant results. Taking first the poison of the Enhydrina, with which most of the observations have been made, and remembering that it is ten times as potent for warm-blooded animals as is Cobra venom, we may compare the action of the two poisons in dissolving the red cor- puscles of the warm-blooded animals, the blood of pigeons and of the human species having been used in the experiments. The method of mixing the poison in different degrees of dilution with a minute measured drop of blood, and counting the number of corpuscles with a hemocytometer before, and at varying periods after, the addition of the venom was adopted. The poisons were always dissolved in isotonic salt solutions, and equal quantities of blood in the same salt solution, 1903. ] Action of the Poison of the Hydrophide. 489 but without the addition of the venom, used as controls. These contro] solutions showed no dissolution of the red corpuscles after twenty- four hours. From 5 to 10 cubic centigrammes of blood were added to from 4 to 1 c.c. of the isotonic solution of the poison, varying strengths of the latter being tested in this way. Pigeon’s blood is specially well suited for these experiments, as the bodies of the corpuscles are dissolved while the nuclei remain visible. It was found that a l-in-1000 solution of Cobra venom (1 milligramme in 1 c.¢.) produced a very rapid solution of the red corpuscles, which had all dis- appeared in seven minutes. A 1-in-20,000 solution took a much longer time to produce complete dissolution, namely two and a half hours. In the case of human blood a 1-in-10,000 solution of cobra venom dis- solved the whole of the red corpuscles in from fifteen to thirty minutes, while one of a strength of 1 in 20,000 took about one hour todo so. A 1-in-100,000 solution had very much less effect, having produced only a slight diminution in the number of the red corpuscles within one hour’s time. The white corpuscles were not dissolved by the venom in the strengths used. Let us now compare these data with those obtained with the poison of the Enhydrina, bearing in mind the much greater potency of the latter as compared with Cobra venom. ‘The poison of the Enhydrina was mixed in the same way as above described with the blood of pigeons and with human blood, in strengths of 1 in 1000, with the result that at the end of one or two hours there had been no appreciable dissolution of the red corpuscles. On testing again several hours later, slight dissolution was found to have taken place, and by this time the solution also showed naked-eye evi- dence of commencing hemolysis. After having been kept at room temperature (from 70° to 80° I.) for twenty-four hours the dis- solution appeared to be complete, but, on examination with the microscope, a few red corpuscles were still found to be undissolved, showing that even after this lapse of time the hemolytic change was not quite complete. The poison of the Distevra cyanocincta and the Hydroplis cantoris were also tested in the same way with precisely similar results, namely, that a strength of 1 in 1000 had no appreciable hemolytic effect at the end of one hour, but caused nearly complete dissolution at the end of the course of twenty-four hours. This is about the same effect as is brought about by a solution of Cobra venom of a strength of 1 in 100,000, although Cobra venom has a potency of only one-tenth that of the poison of the Enhydrina. Thus we find that in proportion to its potency the poison of the Cobra has about 1000 times as great a hemolytic effect on the red corpuscles of warm- blooded animals as has that of the Enhydrina. We have already seen that the latter poison produces no blood-stained effusion at the site of the injection of a fatal dose, evidently on account of the strengths used. 490 Dr. L. Rogers. On the Physiological [ Mar. 31, having no hemolytic action, for the solutions employed for the small animals experimented on were | in 10,000 or less. If we work out the amount of poison required to dissolve a certain amount of the blood of a pigeon, for example, we find that it takes about 200 times a fatal dose to dissolve 1/2000th part of the bird’s blood in twenty- four hours, calculating this fluid to be one-thirteenth of its body weight. It is obvious, then, that ordinary fatal doses of the poison of the Hydrophide can have no appreciable hemolytic effect, and that death cannot be attributed, even in a partial degree, to its action on the blood of the animal killed by it. This can also be demonstrated by another method of experiment, namely, by counting the number of the red corpuscles before the administration of the fatal dose of the poison, and again immediately after death. This I have done several times, with the result of showing that no dissolution of the red corpuscles resulted from the action of a lethal dose of the Enhydrina. poison. For example, a fowl’s blood was counted, and 3,190,000 red corpuscles per cubic millimetre were found. A lethal dose of Enhy- drina poison was then injected subcutaneously, which proved fatal in just one hour, when the blood count showed 3,120,000 red corpuscles in the same quantity of blood. Next we have to deal with the action of the poison on the coagul- ability of the blood. In the case of Cobra venom marked changes are produced, as shown by D. D. Cunningham, and this point has recently been studied by Lamb. The virus has the action of reducing or totally destroying the clotting power of the blood when mixed with it in small quantities. JI have made a few observations on this point with the following results. Wright’s tubes were used, the solution of the poison being first drawn up into them, and then an equal quantity of the blood drawn up and quickly mixed with the venom solution in the mixing chamber, and blown down into the tube again, and the condi- tions as regards clotting examined in a series of such tubes at given intervals. The clotting time, when mixed with an equal quantity of the normal salt solution (in which the venom was also dissolved) of a rabbit, having first been found to be three minutes, those of different strengths of Cobra venom in normal salt solution were found to be as follows: when a 1-in-10,000 solution was added the coagulation time was seven and a half minutes; with 1-in-1000 solution it was twenty minutes, and with a 1-in-200 one the blood was still quite fluid after twenty-four hours, its coagulability having been completely destroyed. On testing the effect of the poison of the Enhydrina in a similar manner it was found that a 1-in-1000 solution had no effect in reducing the coagulability of the blood, which still clotted solid in three minutes ; when a 1-in-200 solution was added the blood still clotted in five minutes, showing only a slightly reduced time with the same strength, which in the case of Cobra venom had completely 1903.] Action of the Poison of the Hydrophide. 491 destroyed the clotting power, and this, too, it must be remembered, in spite of the Enhydrina poison being ten times as powerful as that of the Cobra. It is evident, therefore, that the poison of the HEnhy- drina has no appreciable effect in ordinary dilute minimal lethal doses on the coagulability of the blood, while, as a matter of fact, we have already seen that such doses do not produce any loss of the clotting power of the blood. This was also the case when fifty times a minimal lethal dose of the venom was injected into the vessels of rabbits with the result of causing death in about six minutes. The above experiments show that the poison of the Hydrophide has no appreciable action on the blood of animals, which can in any way account for the symptoms and fatality caused by it, yet it kills with precisely the same symptoms as are produced by Cobra venom, and, as we shall see presently, there are good reasons for believing that it has a special action on the nervous system. It will be evident at once that this furnishes a very strong argument in favour of the view that Cobra venom also kills through the nervous system, as held by Lauder Brunton and Fayrer, and not through the blood, as maintained by Cunningham. It is also of special interest to observe that although the action of the poison of the Hydrophidz on the blood is practically a negligible quantity in its lethal effects, yet it still persists to a slight, but easily demonstrable, degree ; for if it so persists in the Sea-snake, it may also persist in a still greater degree in the case of the Cobra without being a very active agent in the lethal effects produced by that poison, which kills through the nervous system as does that of the Hydrophidz. In this connection it is interesting to observe that all through the poisonous snakes we find evidence of an action on the blood and on the nervous system in different degrees. Thus, beginning with the Viperine snakes, we first have the Vipera Russell, which appears to be the purest blood poison of the known venomous snakes, killing by producing intravascular clotting in large doses, and the opposite effect of total loss of coagulability in repeated sub-minimal lethal ones. Then we come to the class of Pit-vipers, of which the rattlesnake of :America has been most closely investigated by Weir Mitchell and Reichert. They also found a very marked effect on the blood, apparently similar to that produced by the Daboia, but, combined with this, we have a marked paralytic effect on the nervous system, and especially on the respiratory centre, for the authors mentioned conclude that although death may occur through the effect on the blood, yet they add “There can be no question, however, that the respiratory centres are the parts of the nervous systems most vulnerable to the poison, and that death is commonly due to their paralysis.” Leaving the Viperine snakes and passing on to the poisonous Colubrines, we first come to the Australian species, so ably studied by C. J. Martin, namely, the Pseudechis, and we find again a combination of the two effects to +492 Dr. L. Rogers. On the Physiological [ Mar. 31, such a marked degree that, when the venom is administered intra- venously, death results from intravascular clotting, as In the Viperine snakes, while if minimal lethal doses are given subcutaneously death results through paralysis of the respiratory centres. Next we come to the Cobra, another Colubrine snake, and here we find the nerve symptoms quite predominate, although some considerable effect on the blood in the form of reduction of coagulability and dissolution. of the red corpuscles still survives, although it now takes quite a second- ary position to the effect on the nervous system. Lastly, we have the Hydrophide, which, morphologically considered, are but colubrines modified for an aquatic existence, and here we find a practically pure nervous poison, although there still persists a trace of action on the blood if strong solutions of the venom are employed, although it can have no actively poisonous effect. The very slight action found, however, may be of some value to the snake in the following way. We have seen that a 1-in-200 solution of the Enhydrina poison has a slight retarding effect on the clotting power of the blood, which would doubtless be more marked in still more concentrated solutions, so that it is highly probable that the pure poison would have the effect of preventing the clotting of the blood at the point of injection of the poison, and so allow of its more ready absorption into the circulatory system through the patent vessels severed by the fang. This will account for the extreme rapidity of the absorption of the poison of the Cobra, for Fayrer showed long ago that if immediately after a dog has been bitten by this snake the fold of skin punctured is raised and freely excised, still the animal dies of the poison. The survival of some degree of action on the blood in the case of the Cobra and the Hydrophide, although not in itself an important element in directly causing the death of the animal, may nevertheless be of service in causing the venom to be more rapidly absorbed in the way just pointed out. Action of the Poison on the Pulse and Respiration. We have already seen that in slow poisoning the respirations become more and more laboured until convulsions set in and they quickly cease, while the heart continues to beat for a short time. For the accurate study of the exact effects on the respiratory and circulatory systems, proper recording apparatus is necessary, but as these were not available, I had to content myself with a record of the rate of the pulse and respiration after the intravascular injection of a large and rapidly fatal dose of the poison into rabbits under the influence of chloroform, with the following results. A dose of 1 milligramme per kilo. weight, or at least twenty times a minimal lethal dose was used, and death resulted in from six to eight minutes, taking the time up to 1903. ] Action of the Poison of the Hydrophide. 493 the cessation of the heart’s beat. The effect on the respiration was simply a uniformly steady slowing down until convulsions set in, when the breathing finally ceased at once. For example, in a rabbit which had received a dose of 1 milligramme per kilo. directly into the carotid artery (the artery being clamped immediately afterwards to prevent hemorrhage), the respirations were 60 per minute imme- diately before the injection of the poison. During the four minutes immediately following the injection, the number of respirations were as follows :—first minute, 56; second minute, 51; third minute, 42, and the fourth minute 33. In the first quarter of the 5th minute they were 8, at which point convulsions set in and the breathing stopped. The respirations were written down every quarter of a minute, and the figures for the separate quarters show an equally steady diminution of the number of respirations as the minute periods just given. In the same experiment the pulse showed the following changes. Before the injection it was 105 per minute. During the second half of the first minute after the injection it was 47 (that of the first half minute was lost), during the second minute it was 106, showing no alteration up to this time. During the third minute it fell to 99, and during the fourth it further fell to 48, that for its first half having been 32, and for the second half 16. During this steady fall in the pulse rate, its volume and force became increased. During the fourth minute, as already mentioned, convulsions set in, and the pulse was lost for about a minute, only the first and third quarters of the fifth minute having been recorded as 8 and 11 beats respectively. During the last three quarters of the sixth minute the beats were 15, 15 and 17 respectively, being now very feeble instead of unusually full, as before the cessation of respiration and onset of convulsions. During the first and second quarters of the seventh minute, the beats were 26 and 20 respectively, at which point the heart finally ceased to beat, that is, three and a half minutes after the cessation of the breathing. Very similar results have been obtained in another experiment, in which the same dose was injected into the jugular vein, a steady fall in the respirations first occurring, and they ceased with the onset of convulsions, while an equally steady fall in the pulse rate occurred later than that of the respirations, accompanied with an increased volume of the artery, the tension rapidly falling when the respiratory convulsions set in, but the pulse at the same time became more rapid again until it finally declined once more and then ceased. ‘These experiments appear to show that the primary effect of the poison is a paralysing action on the respira- tory centre, and that the cardiac failure is secondary to that of the respirations. The exact explanation of the siowing of the pulse with mereased volume of the artery, | am not prepared to say without the aid of pressure tracings, which I have not yet been able to take. VOL. LXXI. 2N 494 Dr. L. Rogers. On the Physiological [ Mar. 31, -_* The Affinity of the Nervous System for the Poison. We have seen that the poison of the Enhydrina is much more potent than even that of Cobra, and it appears to be somewhere intermediate in virulence between Cobra and tetanus toxins. Further, we know that the repeated injection of gradually increasing doses of the latter two poisons into susceptible animals leads to the formation of an anti- toxin in the system. This marked similarity of the nerve poisons of the Colubrine Snakes and tetanus toxin leads one to inquire whether these snake venoms do not exert their noxious influences in the same way that tetanus toxin does, namely, by being taken up from the eir- culation and fixed in the nerve cells until a sufficient dose has been absorbed to paralyse the nervous matter. We know from the experi- ments of Wassermann that small amounts of tetanus toxin can be thus fixed by fresh nerve matter in a test-tube, and so rendered inert when subsequently injected into a susceptible animal. It seemed to be worth while to repeat these experiments with the poison of the Enhydrina, and although I have not had time to carry out a sufficiently exhaustive series of experiments to settle this point, yet the following data appear to me to have some value as being highly suggestive of the mode of action of these nerve-paralysing snake venoms. The experiments were carried out in the followmg manner. A weak solution of the venom, such as is used when giving minimal lethal doses, was placed in a small sterilised test-tube, and a given quantity of fresh brain matter from a pigeon was added to it, and the whole kept at blood temperature for a given time. Another solution of the same strength was kept at the same temperature for an equal period of time without the addition of any brain matter, for the purpose of injecting control animals, which were always used. Double and quadruple minimal lethal doses were used, and the brain matter was broken up so as to mix it with the poison as intimately as possible, and subsequently injected without filtering, so that most of the brain matter in a fine emulsion was injected with the poison. It was found that pigeons injected with these emulsions always lived longer than the control one, while they sometimes recovered from double, and in one instance from quadruple, minimal lethal doses of the poison aiter being mixed for from half an hour to eighteen hours with a small quantity (from 3 to 20 centigrammes) of fresh brain matter. The most marked effects were obtained by the use of the hemispheres of the cerebrum, the instances of complete recovery from lethal doses having occurred in these instances. The cerebellum had a less marked effect, only considerable prolongation of life having occurred, while in one experiment with the medulla and pons no very marked effect was observed. The grey matter, then, appears to have most effect in fixing the poison, as is also the case with tetanus toxin. These experi- 1903. ] Action of the Poison of the Hydrophide. 495 ments, then, point to the action of the toxins of the Enhydrina being very similar in nature to that produced by the tetanus bacillus. A few experiments were also done with Cobra poison in the same way, using the cerebrum only, but here the results were not so marked as in the case of the Enhydrina poison, only a retardation of the onset of symp- toms and of death having been observed. Antitoxins. Lastly, we have to deal with the question of the possibility of ob- taining an antitoxin against the poison of the Hydrophide. It has now been abundantly proved that Calmette’s antivenin is not a specific against all kinds of snake venom, as he claimed, although in large doses (40 c.c. according to Lamb) it is undoubtedly of great value against the poison of the Cobra. The very marked similarity of the symptoms of poisoning by the Hydrophide with that produced by Cobra, lead one to hope that the antitoxin, which is efficient against the latter, would also be of value against Sea-snake venom. This has been put to the test by adding minimal and slightly supraminimal lethal doses ef the poison of the Enhydrina to one half ¢.c. of fresh Calmette’s antivenin (which had only reached Calcutta a very short time before it was used), and after allowing the mixture to stand at blood heat for half an hour, injecting the whole subcutaneously. White rats were used in the experiments, and the amount of antivenin in proportion to the amount of poison was relatively enormous as compared with the dose recommended in the treatment of men bitten by venomous snakes. Yet the animals uniformly died in just about the same time as the controls, so that it is evident that Calmette’s serum is of no use against the poison of the Hydrophide. On the other hand, the similarity in the action of this poison to the ‘Cobra and tetanus toxins jeads one to expect that an antidote could be prepared against it in a similar way to those of the latter poisons. It is only during three months that I have been able to experiment on this point, fowls being used. It soon appeared that the doses had to be very slowly increased, or fatalities occurred, and in the limited time these experiments lasted, I was only able to immunise one fowl against the minimal lethal dose of this poison, and a slightly larger dose proved fatal with the usual symptoms. My intention was to immunise a series of animals against the Enhydrina poison, and then to test them with small doses of poisons from the other Hydrophidz, as owing to the large variety of this class of snakes, no antidote would be of any practical value unless it was equally potent against all the genera and species, or at least against the ones most commonly met with. This important and interesting question must await further investigation. One experiment, which was carried out in order to test if the serum 2°N 22 a! 496 Action of the Porson of the Hydrophide. [Mar. 31 or bile of the Enhydrina had any antidotal properties, deserves mention in this connection. Three puppies of the same litter were used, all very much of the same size. Hach received an equal quantity of Enhydrina poison, but in the first this was mixed with a four minims of the serum of the Enhydrina ; in the second it was mixed with four minims of the bile of the same snake ; and the third received the poison solution only as a control. The mixtures were injected ten minutes after being made. The result was that all three animals died in a little over an hour, the control surviving slightly longer than the others. It appears, then, that neither the serum nor the bile of this snake has any antidotal properties against the poison, and can not, therefore, be utilised in the treatment of their bites. Further re- search will be necessary to determine if a practically efficient antidote can be prepared, which I hope to undertake when sufficient venom for the purpose can be obtained. This concludes the most important experiments so far carried out by me with the poison of the Hydrophide. They have necessarily been strictly limited by the very small amount of poison which I have yet been able to obtain, and by the equipment of the laboratory at my disposal, for the use of which I am indebted to the kind permission. of the Committee of the Zoological Gardens of Calcutta. I am also indebted to the Bengal Government for a grant towards the expense of this investigation. 1903.) Experiments in. Hybridisation. 497 “ixperiments in Hybridisation, with Special Reference to the Effect of Conditions on Dominance.” By L. DoNcASTER, b.A., King’s College, Cambridge. Communicated by Dr. 8. F Harmer, F.R.S. Received March 19,—Read May 7, 1905. (Abstract.) ‘The paper describes experiments made at Naples with hybrid Echinoid larve. The object of the experiments was to determine whether the dominance of a character is influenced by the condition of the genital cells at the time of fertilisation. It had been suggested by Vernon* that the “prepotency” of the sexual cells varies with their maturity, and experiments were made to test this conclusion, and also to discover whether ‘“ prepotency ” could be influenced by other conditions acting on the eggs or spermatozoa before fertili- sation. | It was found that adverse conditions acting on the eggs did give rise to differences in the larve, but evidence is given to show that these differences are not due to a change in the dominance of characters, but are the result of differences in the vigour of the larve. It ig also shown that the seasonal changes observed by Vernon are probably due chiefly, if not entirely, to differences of temperature, and are not caused by a change of dominance accompanying difference of maturity. It is also shown that if an individual, A, shows greater dominance than B when each is crossed with a specimen X of the other sex, then A will also show greater dominance than B when both are crossed with a specimen Y. | It is shown that the different characters of one parent are inherited separately by the hybrid offspring, so that there is no pronounced correlation in the offspring between characters derived from the same parent. Further, a given character may appear in very different degrees on the two sides of the body of a hybrid larva, so that the hybrids are very frequently asymmetrical, although in the characters considered the two sides of the pure-bred larve are similar, . Experiments are described dealing with the causes which hinder cross-fertilisation between separate species, and it is shown that treat- ment of the eggs which tends to reduce their vitality usually renders their fertilisation by sperm of another species more easy. , } Seale Granss 0 8os8; 498 Miss D. M. A. Bate. On the Discovery of a [Apr. 23, “Preliminary Note on the Discovery of a Pigmy Elephant in the Pleistocene of Cyprus.” By Dorotuy M. A. Bare. Commu- - nicated by Henry Woopwanrp, LL.D., F.R.S., F.G.S., V.P.Z.S., late Keeper of Geology, British Museum, Natural History. Received April 23,—Read May 7, 1903. While still in Cyprus the receipt of a grant from the Royal Society in April, 1902, enabled me to devote a considerable amount of time not only to making more extensive excavations in some of the caves previously found, but also to a search for further cave deposits. I confined my attention chiefly to the Keryina range of limestone hills in the north of the island in the hope of finding bone caves containing other remains than those of the pigmy Hzppopotamus, of which Dr. Forsyth Major has already given a short description® from specimens discovered by myself. In this search I was at length successful, although it was not until a certain amount of tentative iene had been carried on in four out of five newly discovered deposits that work was started on what uppeared at first to be the most unpromising looking place which had been found, and was consequently the last to receive attention. However, during the first day one of the workmen found, not far from the surface, part of a tooth which was at once recognised as being that of an elephant. After this discovery every effort was made to procure a complete collection of the remains of this species, but at no time were either teeth or bones found to be so plentiful as those of Hippopotamus minutus, with which they were associated. Often not a single proboscidean tooth would be obtained during two or three days’ work, and only eleven molars and parts of molars were procured as the result of three weeks’ digging. It was then decided to continue excavations here for a short while longer, and this was done until the end of July, work being again resumed in the beginning of the following October. Altogether a good series was obtained of the teeth of this elephant, which is found to be a pigmy species. With the exception of the first milk molar (m.m. 2), specimens were procured of all the milk and permanent molars of both the upper and lower jaws; also a number of tusks of different sizes, though these included none of the tiny milk incisors. No teeth which could be referred to very aged individuals were obtained, for amongst the last true molars none have more than half their full number of plates in use. ‘The series of teeth consisting of specimens of very small size, it was natural in the first instance to compare them with the remains of the * © Proc, Zool. Sec.,’ June 3rd, 1902. 1903. | Pigmy Elephant in the Pleistocene of Cyprus. 499 dwartf species from the Pleistocene deposits of the caves and fissures of Malta and Sicily. It was thought probable that they would differ from these; the fact of the pigmy hippopotamus of Cyprus being distinct from those found in the other large Mediterranean islands, lending colour to the supposition ; this expectation was fulfilled, for the Cyprus fossils do not appear to be identical with any of the Maltese species, though they seem to come nearest to Hlephas melitensis both in size and in the number of plates in the molars. The number of these plates in any particular tooth is liable to vary to a certain extent, but on taking the average, as far as this can be judged from the amount of material available, the resulting ridge formula, exclusive of talons, is 3 = : - - which practically agrees with that of £. melitensis given by Dr. Falconer.* The teeth of the Cypriote elephant are considerably smaller than those of L. mnaidriensis, from both Sicily and Malta, this being the largest species from the last-named island. They also differ some- what in their ridge formula, which is that mentioned above, while Dr. Leith Adams? gives that of F. ERE Mas as 5 : = a = 12-13 Toa} The Cyprus form seems to have been also slightly inferior in size to L. melitensis, for its largest upper and lower molars do not equal, either in length or breadth, some of the specimens of the correspond- ing teeth of this Maltese species which are in the collection of the British Museum. Its tusks differ from all those from Malta in being compressed laterally, which character is especially noticeable in those of the female and young; further, they appear to be more strongly curved than those of F. melitensis. As a general feature it may be said that the molars from Cyprus are, on the whole, more simply constructed than those of /. melitensis. They show a still slighter tendency to “crimping” in the bands of enamel, and are less inclined to develop the mesial expansion of the plates of dentine which is not uncommonly found in the teeth of E. melitensis, and is so conspicuous in those of L. Africanus. It is well known that when the plates of an elephant’s tooth first come into use, the edging of enamel is in the form of a series of rings owing to the digitation of the plates. These are later worn into a single band surrounding the enclosed area of dentine. In the Maltese specimens it is not uncommon to find the encircling enamel persisting thus divided for a considerable time. Even four or five ridges may remain in this condition at one time in a single * “Pal. Mem.,’ vol. 2, p. 298. London, 1868. ft ‘Zool. Soc. Trans.,’ vol. 9, 1874, p. 112. 900 Ona Pigmy Elephant in the Pleistocene of Cyprus. [|Apr. 23, tooth, with perhaps an anteriorly decreasing number of rings. This is well shown in a tooth, now in the British Museum collection, doubt- fully ascribed by Mr. Busk* to the first upper true molar of L. Falconeri. This is not so much the case in the Cyprus specimens, in which the bands of enamel only remain thus separated into several annuli for a very short while after the plate comes into wear. The molars vary considerably, some specimens having very broad crowns while others are somewhat narrow. ‘The bands of cement are wide, in perhaps the majority of cases almost, or quite, equalling in width the plates of dentine; this seems to be the exception and not the rule in the molars of E. Hea enstes Taking into consideration the several characters in which the teeth of the Cyprus elephant differ from those of all the hitherto described dwarf species (putting on one side /. /amarnwrey from the Pleistocene of Sardinia, the teeth of which are unknown to science) as well as the distinct habitat of the animal, I have come to the conclusion that it is specifically distinct from these other small forms, though possibly they were derived from a common ancestor, and I, therefore, propose to name it Hlephas cypriotes. The discovery of the remains of this pigmy elephant, as well as of Tippopotamus nunutus, in Cyprus, is interesting in comparison with the dwarf species from Malta and Sicily, and because the presence of an extinct mammalian fauna in this locality had not previously been recorded. The occurrence of these different, though apparently closely related, races of small elephants in widely separated islands of the Mediterranean, lends probability to the theory that this is a case of independent development along similar lines, the result of similar circumstances and environments. Nevertheless, it would per- haps be wise not to take it for granted, without further evidence, that this diminutive size is wholly and entirely due to specialisa- tion, I hope shortly to be able to communicate a more detailed account, with figures and full descriptions, of this collection of elephant remains from Cyprus. =). Zool. Soc. ‘Vrans.,7 vol. 6.) Pl. oa, tig.\9) puedo: t Dr. Forsyth Major, “ Die Tyrrhenis,” ‘ Kosmos,’ vol. 7, 1883, p. 7. 1903.] Trypanosoma present im Cases of Sleeping Sickiress. 501 “On the Discovery of a Species of Trypanosoma in the Cerebro- Spinal Fluid of Cases of Sleeping Sickness.” By ALDO CASTELLANI, M.D. Communicated by the Malaria Committee of the Royal Society. Dated “Entebbe, Uganda, 5th April, 1905.” Received May 8,—Read May 14, 1905. On the 12th November, 1902, when examining a specimen of -cerebro-spinal fluid taken by lumbar puncture during life from a well- marked case of sleeping sickness, 1 was surprised to observe a living trypanosoma. Since that date I have made as many observations in this direction as possible, and the results are to my mind sufficiently ‘surprising to excuse me for presenting this preliminary note. . These trypanosomes are not in large numbers, so that to find them it is necessary to draw off at least 15 ¢.cs. of the cerebro-spinal fluid. It is better to reject the first few c.cs. as they are apt to contain blood. When the fluid comes away clear, 10 ¢.cs. are collected and centrifuged for 15 minutes. At the end of this time there is found at the bottom of the tube a slight deposit of whitish sediment, and in some cases also -a minute trace of blood. The liquid above the sediment is poured off and ane sediment examined under a moderately low power of the microscope. As the trypanosomes are at first fairly active they are easily detected. The following tables represent the results of this investigation :— [May 8, iscovery of Trypanosoma A. Castellani. > D Ke) “AjaAry otour ATyuorvdde O19M SJUOMeAOU ATeYyy ATUO ‘pmbry aangound avquimy ayy ut punoj asoyy 09 «Byturts A;yU1edde somosousdday Moy B punoy T Lep oes 9} AodUy BV WOA, UdyB4 pooyq jo suoyvavdord ysouy uy ‘guosoad *dAaT, “GUOA “4R] WoL; pInyZ ay *suoTyRorTdut0d ON "ZOGBT ‘RT Loquioo0¢gy UO porp JuorR | | "BYABULOY | | quasq ¥ quesq Vy qUO8Od quosqy JUOsSq (USI Us Vy qros¢ W ‘pug peurds -OLQ9L90 UL vutrosourd 414 JO ooudsoar | nm) a ee as ‘orga woz Aton | pus saydooone, owog | ‘Oa oulos pue sayfoooney = Moy | ‘Oras — || ON ‘sogh00on0T oul0g ‘ogy Aueut _ pur seyAo0one, outog ‘OMA MF Aaa. 80) Looney MOT -V HO)ra betel a) Kio A *sojKooone] MOF V ee Ouael (ON ‘Sogho0ono0], 9lWOg | Oey pur -soyhOOon0T MOF OUOG | soposnd.too poojq pot orva ATA aWog ‘avoponuodoUt aav yous Jo Aprcolvur | 044 ‘soyfooone, Moy YW | “yUaWILpos jo oounivodde dIdoosodor yy, pag pig PrE pug ple pag plg pig ‘osvestp jo a54C ‘sosuy) ssauyoig suldeo[G—T 9[qeJ, 0/1/24 ¢0/T/¢ ZO/Z1/3% 20/21/¢ Z0/Z1/ST Zo/z1/e1 zo/@t/2 Z0/11/4z 60/11/21 ‘Oye 06 OL | | ks ie ai vo yuepog, | ** ee weave ea MaRS hoe )410(-) 3 VOUeL . . ° - . . pot ease OZ “TOUTY eres 68 @8 oO | opunyy, ree 9 te) ———— le) =) 1D keness. UC g Si mM L Fluid in Sleep -spind the Cerebro ° Un 1903.] "SyIVULIY | | quesqy quesqy RON quosotg quesq Vy pus quosqy put quesq VW quesqy = pun quesqy pu qussolg pur quesetg = puv quosqy | pur quesqy | pur quosqy | pur qudsol | ON | ‘ping yeurds -O1Q9100 UT | eutosouvd say Jo oou0serg ‘panuiguoj—soseg ssouyorg sutdosjg— | o[qvy, bo ‘ya OUTOG “OEY ous “Oa ON ‘OMY Mey B pus seqgdooone] 20g OLS soyA000ne, == MOT Od soyfo0oneT §= MAT TOP aE tal OME | ONS soqdooone] vorvos {10 A Oana soyfo0one,T + awl0g VE soqyfo0oneT «= MOT “) gy Moz An0A seqydoooneyT alloy ‘Oa y Moz Atoa soqfoooney, + emlog ros ars i soyXooone] Moy VW OLA seyfd0on0, + euog ‘OT y woz A094 soqyho0ona] Mot VY Oh: esi ‘soq4o0one, awWlog “JWAUIIpes jo sousrvedde o1dooso.10 Ty { | | 381 981 4ST "aSBOSTP Jo 994 | £0/8/L2 £0/€/% 0/2/26 £0/8/8 €0/2/24¢ €0/2/01 20/2 /2 20/1/36 £0/1/¢z "OqU([ BE Oa Sie og | “a ZS | ‘W og | ‘W oe | W bl | CW eZ aT eee NC BLO, | Se ge CUOTU Tes, 010-0 OD oO OD TQUIO NT eee lee eG LUDLUCT pO BZIIVOTAL ee Suey issn soeaaae [os ouee "++ UssRARTAL at ST PI Miscovery of Trypanosoma {May 8, Dr. A. Castellani. 504 ‘s08¥0 TOY}O UL UY} sHoceuINU d1OUL TONU 910M SoMOsoURd 4.14 OILT, “SYABULOAT | JUISOL | qttasod quosolg quasar quasar quesq vy USA. quosad quosal quasar quesolg quosorg quasar quosq Y ess. quesq y i ee ‘ping yeurds “0149100 UT vutosoued 4.14 jo oouasortg pur DINE pue pue Oa soyXoo0oneT + autog deel ON: Oa ‘soqyMoooney, 9ULOg OL ‘saqho0onoy = MOT Oy ‘saqAooonoy ottog Ta ‘soyfo0oneT attog 6 ‘OE ON 5 ‘EW ON ‘Orgear ‘sayAo0oneT + etlog OLE saqfoooney, = Ma ‘Ona aT soyfo00n0T «= MOT Oa pue soy {o00ne] Moz amog Oad pure soz{o0on9] may vULOg ‘JUBUTTpes yo souvavodde d1OdsO.LOT TL “panwurqwogi—sasey ssouyoig surdesjg—'] o1qvy, | | | | ee } “OSUOSTP | OD jo I581C £0/8/8z 60/8/82 ¢o0/e/22 €0/2/12 ¢0/¢/8z ¢0/4/T €0/¢/2z £0/¢/9z €0/8/9% ¢0/¢/9z €0/¢/9z £0/¢/ 42 ¢0/¢/¢z ¢0/¢/62 £0/¢/8z 20/¢/¢z “Oye | 0c IN CZ WW Gs “iL OF | “W ek co CW 04 | un ec | ‘W ee | 71K eZ | “It o3 | ‘W ‘as | “xed ‘+ QULLOpLyy ** TUOqQoory 78 BIBIESqT “+ BZBqOROT "55+ Oqoote wWNnYVso0yf ‘+: ekosey ee Eee: 9° BIQUITSY rsss8 BIIAG a eZ, + crerelueg “OTLB NT ‘ONT ST S ~ r « L Fluid in Sleeping Sickness. *O-SPUna the Cerebi v7 1903.] ‘waptou-7s0d ayy 4U “QUOA [R1o4v] WOA, USye, pny ayy UL OsTe quosead vwosourdday, "peyoeqfoo pmmbiz jo a0 @ AjuUG SS ———— es — "SyAVULay {Used J pus quesqy pur Quorqy pue quosolg es quosal pue quesqy pur ques y pue quosqy put quosqw puv ‘pry jeutds -OQO100 UL eutosoued £14 JO eotlosotg 0 selerel soyMvoonoT = MOT So 0) 8h tel soyA000N0T = MOT rer: ire i sayA000N2T = MOT eb al ONE Od seqfooane, aulog OU soyfoooney, aod OL soyfooone,T oulog Ey soyAo0oneyT ouwlog Oe soj£000nNeT owog “JUOULL Pas jo souvavodde d1dooso.L0Ty, pig “ostastp jo 9984 | "PanUyVvoI—sosvy ssousyprg sutdosjg-——T| o[(v, 80/8/18 £0/0/z ¢0/¢/62 £0/¢/¢ €0/€/62 £0/F/2 e0/¢/8¢ ¢0/F/T “IN ssp’ OUNPOII ore SINT AC 0 DO DM CRA OUDS ee ee 86 8 @ BASIB AA see BULOLCIT YY, "898 VLOpIqRevayy “OTUUN && GE 1& O& 66 ‘ON [May 8, Discovery of Trypanosoma Dr. A. Castellani: 506 ‘SUITLIOUL OUTRS OY} poolq Ulasuaiquny nuos | -oundhsy, punojz s9oyeg “aq | quesq Vy | Te Ve | quosqy | quosq Vy quosqy FUSsq Vy HOE quosqy “‘SUIMAIOUT STUBS O17} poolq UL aswarquin,y Duos -oundhsy, pUnoj aoyeg “17 ques W quosqy ‘ose skep oul0s poolq UL aswarquin.y vwos -oundhw yz, punoj aoyeg 1q | quosq vy | quosqy | ‘eumosoued K.1} "‘SYIVULOY jo DOUISITT ‘O'a'y pue soyfooone, mez Ata A (0 ae ue sajooone, ~ moz Ato A Oa pue soyfooone, Moz AI0A4 Oa pus soyfooone,T maz AOA 3 jo celtal [OU SOJAOOONOT MOF ALIA Oe pues sayf00on0y «MOF A104. Oe pue soyfooonoy ~moz A104. OL ON ‘soqf00oneT Mo nid so}fo00neT ~=MoT One ‘soqyAooonoy =O pur ON 10) el tak |e sozkooonoyT «Mey =ATO A (Ors lieb [DERG sogfoooneT Moz AOA ‘quoUIIpos jo oourarvodde ordooso.or yy "LOAOT eutosouvd Any, oTPOVpVeFyT SI}UNo[ g qu See) IIAOT eutosourd ALT, Spurys [R1omo;z | JO juotmesaepuy TOAOF | wutosoued Ay, sty -Taqdeu o1moatyo ‘osvastcy ‘s[0.19UOQ— TT 91qvT, €0/¢/Ts €0/e/Te e0/¢/Te €0/€¢/08 €0/s/08 €0/¢/08 €0/¢/08 €0//08 ¢o/¢e/sz ¢0/e/FZ ¢0/8/Fz 20/1/1T **-uelanyA, WeTp.zo p eos @e@ eee II nqeusy i a ORATOR oe ee ae "> -os00g tee one OUNSTAB IL ose ee+ @e © oe npueT oSsuLing ouey RATUBO “OMe NT We) Coy iso oe |e 1903.] in the Cerebro-spinal Fliud in Sleeping Sickness. 507 Table I shows that in 34 cases of sleeping sickness, the trypano- somes were found in the cerebro-spinal fluid taken by lumbar puncture during life in 20 cases, giving a rate of 70 per cent. On two occasions I also examined in the same way fluid from the lateral ventricles and in both cases found the same parasite. In blood I found the trypanosoma once with certainty. It may be thought that the trypanosomes are found in the cerebro- spinal fluid on account of the trace of blood which sometimes forms part of the sediment. But it will be seen from the table that in several cases there was no trace of blood. Table II shows that in 12 cases of ordinary disease, the cerebro-spinal fluid taken during life by lumbar puncture, in no case contained trypanosoma, and it is important to note that 3 of these controls were eases of the usual trypanosoma fever, as described by Forde, Dutton, Manson, Daniels, &c. Here it may be remarked that trypanosoma fever is by no means uncommon among the natives in Uganda, 3 cases having been met with, by Dr. Baker, one of the colonial surgeons here (Entebbe), within the last 3 weeks. I understand that Dr. Baker is publishing this most interest- ing observation. It must be clearly understood that these cases of trypanosoma fever bear no resemblance in their clinical features to sleeping sickness. The trypanosoma found in the cerebro-spinal fluid of sleeping sick- ness does not as far as I have been able to make out differ materially in size and shape from the species one finds in the blood of trypano- soma fever, Trypanosoma Gambiense (Dutton), but possibly it is to be differentiated from this one, because in it, as a rule, the micro-nucleus lies nearer the extremity and the vacuole is apparently larger. Besides, its movements are not apparently so active, but this fact might be due to the effects of the centrifuge. In case it should prove to be a new species, the trypanosoma I have described might be called from the country where I have found it first-—TZ7rypanosoma Ugandense. Relation of the Trypanosoma te Sleeping Sickness. At the post-mortem examination of 80 per cent. of the cases where I found during life the trypanosoma, I grew from the blood of the heart and from the liquid of the lateral ventricles the variety of streptococcus I described many months ago in my first note. Up to that time I had never found the trypanosoma, but this is easily explained by the fact that I did not use the technique I have described in this note, viz., examination of a /arge quantity of liquid after long use of the centrifuge. Influenced by my last investigations I would suggest as a working hypothesis on which to base further investigation that sleeping sickness 508 Trypanosoma present in Cases of Sleeping Sickness. is due to the species of trypanosoma I have found in the cerebro-spinal fluid of the patients in this disease, and that at least in the last stages there is a concomitant streptococcus infection which plays a certain part in the course of the disease. Note by the Secretary of the Royal Society. As so far supporting the observations by Dr. Castellani recorded in the above communication, it may be desirable to state that Colonel Bruce, to whom in Uganda Dr. Castellani made known his discovery of the Trypanosoma, and who is now continuing the investigation begun by Dr. Castellani, has sent to the Royal Society a telegram, received May 4, stating that since Dr. Castellani left, in thirty-eight cases of sleeping sickness, he had found trypanosoma im every case in fluid obtained by lumbar puncture, and that he had found trypanosoma in the blood in twelve out of thirteen cases of sleeping sickness. MICHAEL FOSTER. On Skew Refraction through a Lens, ete. 509 “On Skew Refraction through a Lens; and on the Hollow Pencil given by an Annulus of a very Obliquely Placed Lens.” By J. D. Everett, F.RS. Received January 22,—Read January 29,1903. Received in revised form, April 20, 1903. [PLATES 9 AND 10. | Part 1.—Outline. 1. The investigation here described was undertaken with the view of explaining the curious curves obtained by receiving on a screen, at certain distances, the hollow pencil which emerges from an annulus of a lens placed at large obliquity (such as 30° or 45°) to the incident beam.* 2. The first requisite is a process for calculating the direction-cosines of a ray refracted at a given surface, when those of the incident ray and of the normal are given, with the relative index of refraction. In the original computations for this paper, the method of procedure was, to calculate first the direction-cosines of the tangent to the refracting surface inthe plane of incidence. A length unity along the refracted ray was then projected on this tangent and on the normal; and these two projections were themselves projected on the axes of co-ordinates, and added. It has subsequently been found better, instead of actually calculating the direction-cosines of the tangent, to eliminate them between two sets of equations, one of which is obtained by projecting the refracted ray in the manner above described, and the other by similar projection of the incident ray. 3. A single case of special simplicity is selected for the application of the process ; the case of a narrow and thin annulus of a plano-convex lens, with a parallel pencil incident at 45° on its plane face, the index being 1:5. The annulus is supposed to be graduated, the division 0° being at the point furthest from the source, and the graduations being from 0° to +180°. The plane containing the axis of the lens and the axis of the incident pencil passes through the divisions 0° and 180’, and is the only plane of symmetry of the system. For convenience of description, we shall suppose this plane to be ver tical, the point 0° being the highest point of the annulus. 4. Twelve points 30° apart, starting from 0°, aretaken onthe annulus, as points of-incidence ; and the direction-cosines of the corresponding emergent rays are calculated to three or four places of decimals. The- * See S. P. Thompson, “‘ Experiments on Zonal Aberration,” ‘ Archives Néer- landaises,’ 1901, and Fourth Traill Taylor Lecture, printed in the ‘ Almanae of the Brit. Journ. of Phot.’ for 1903; also Alice Everett, ‘‘ Photos. of Sections of Hollow Pencils,” ‘Jour. Brit. Astron. Assoc.,’ vol. 13, 1902, pp. 74-75 WOEs LXXI. 20 510 / eho “Prot. Jd Deiverert: ie an. 22, axis of the lens is taken as axis of 2, the axis of y being in and the axis of « perpendicular to the plane of symmetry. The equations of the twelve emergent rays thus found enable us to plot twelve points of the section made by a plane at any given distance z; and through these a curve can be drawn by hand. | 5. These sections are, however, inclined at about 45° to the emergent rays, and are about midway between the axes of z and y. To remedy this inconvenience, the equations of the rays are transformed to new axes (of 7 and ¢), the axis of ¢ being midway between the axes of z and y, and coinciding with the original direction of the incident beam. Sections perpendicular to this direction are found by assigning different constant values to ¢; and a ‘‘direction-curve” is drawn, which is a section of a cone whose generators are parallel to the emergent rays. 6. Harmonic reduction is applied to the direction-cosines; and harmonic expressions containing either two sines without cosines or two cosines without sines are found to give a remarkably close representa- tion of the facts. | 7. Each ray of the emergent pencil intersects two other rays. One set of intersections are in the plane of symmetry, and are the inter- sections of rays symmetrically placed with respect to this plane. These intersections constitute the secondary focal line, which is absolutely straight, and lies in the production of the straight line drawn through the centre of curvature of the convex face of the annulus, parallel to the rays in the glass. These may be called “left-and-right ” inter- sections. 8. The other set of intersections constitute the primary jocal line. They may be described as ‘“ up-and-down” intersections, inasmuch as the plane of any pair of intersecting rays is ata small inclination to the vertical—that is, to the plane of symmetry ; whereas the plane of a pair which intersect in the secondary line is perpendicular to the plane of symmetry. Each intersection involves an inversion of.relative position of the two rays concerned; and the combined effect of the ‘up-and-down ” inversion at the primary line, and the subsequent “ Jeft-and-right ” inversion at the secondary, is to cause a distant section to be an inverted image of the annulus. 9. In the “up-and-down” intersections, the pairing of the rays follows an unsymmetrical law. Each of the rays between 0° and 79° is paired with a ray between 180° and 79° ;-—a fact which I discovered empirically in my original calculations. The exact law of pairing has since been detected by Professor A. E. H. Love. It is, that the chord joining a pair passes through a fixed point, namely the point in which the plane of the annulus is cut by the secondary line. See §§ 23, 24, 26. 10. The first rays to intersect are those from 0° and 180°; and their intersection is the vertex of the primary focal line. This line is approximately a parabola, lying in a plane which recedes with a a. 1903. ] On Skew Refraction through a Lens, ete. 511 downward slope of about 1 in 6. Hach end of the primary line is the intersection of two ultimately coincident rays from near 79”. After a long interval, the intersections on the secondary line begin, the first being the intersection of rays from two points ultimately coincident at 180°, and the last the intersection of rays wire, coincident at 0°. 11. In the cross-sections, of which numerous specimens at gradually increasing distances are given in Plate 9, every intersection of two rays, or (what is the same thing) every intersection of a ray with one of the two focal lines, appears as a double point, which is generally a point of crossing of two branches, but is sometimes a cusp ; and in one instance the two double points coincide in a point of contact of two branches. For the purpose of identifying individual rays, the numbers 0, 30, 60, &c., are marked, indicating the positions, in each section, of the rays which came from the points 0°, 30°, 60°, &c., in the right hand half of the annulus. They facilitate the tracing of reversals of position. 12. It is by no means a general law for oblique refraction through annuli that the primary crossings are completed before the secondary begin. More usually there is a large region in which the two cross- ings overlap—a circumstance indicated by the presence of three double points in a section, the middle one of the three being in the secondary and the two outer in the primary focal line. 13. Sections of pencils from annuli of obliquely placed lenses Have been calculated by Steinheil* and by Finsterwalder,t the obliquity, however, being only 0° 48’ in Steinheil’s calculations, and not exceed- ing 6° in Finsterwalder’s. In both cases, the method of computation is that devised by Seidel{ based on spherical trigonometry ; and the calculations are only for the positions 0°, + 45°, + 90°, + 135°, 180°. Part II. —Detazls. 14. General Process for Skew Refraction. Let /;, 7, 2, be the direction-cosines of the normal, Io, 12, Ny those of the incident ray, then, calling the angle of incidence x, and the angle of refraction x , we have cos xX = Alot mymy + n4N2. Hence, knowing the index of refraction, we can deduce sin y’ from sin y. * “Munich, Akad. Sitz. Ber.,’ vol. 19 (1889). + ‘Munich, Akad. Abhandl.,’ vol. 17 (1892), p. 519. ~ ‘Munich, Akad. Sitz. Ber.,’ 1866, p. 263. 202 512 Prof. J. D. Everett. [Jan. 22, Let A » v denote the direction-cosines of the tangent to the refracting surface in the plane of incidence. Then /, mz 2 may be regarded as the projections of a unit incident ray on the axes. But this unit ray gives a projection sin x on the tangent, and a projection cos x on the normal. Adding the projections of these on the axes, we have Il, = Asin x +4, cos x Mo = psin x+M, COSY Po -erree eee eee eee GL: Nm. = vsinx +7 COS x And in the same way, by projecting a unit refracted ray, we have ie X sin x’ +/; cos y’ m = sin x’ +10, COS YX! Po vereeee serene ees (2), v sin x’ + COS x’ ~_ a) l’, m’, n’ denoting the required direction-cosines of the refracted ray. Substituting the values of A, u, v from (1), equations (2) become Rey (1s = l COS yY+h cos x’ m’ = k (m2 — My, COS xX) +71, COS YX’ Pee eee (3) n =k (nm2—% Cos x) +71 COS x’ k standing for sin x’/sin y the relative index from the second medium to the first. . Equations equivalent to (3) are given in Herman’s Geometrical Optics, S§ 18, 19 (Camb. Univ. Press, 1900) and have, I understand, been taught for many years by Mr. Webb at Cambridge. - 15. In the original calculations for this paper, a clumsier method was used, in which A » v were computed by means of deter- minants.* The original results have been tested, and in some instances made more exact, by the use of equations (3). * Since the tangent is coplanar with the normal and the incident ray, we have AA+ Bu+ Cy = 0, where A = My Ny | B ae Ny l, Mz Nz |’ ty ly o= | 22am Ll, mz Also, since it is perpendicular to the normal, we have /,A+my+mny = 0.- Hence A, », v are proportional to the three determinants BC a CA > Mm, 2 | ny AB | | bd Lm | L= M = , N=] and A, uw, v are the quotients of these by /(L?+M*+N?). J’, m’, n’ are then found by substituting the numerical values of A, u, v in equations (2). Tt can be shown, by expanding the determinants and making obvious reductions, that /(L?+ M?+ N?) is sin x, and that the final results reduce to equations (3). 1903. ] On Skew Refraction through a Lens, ete. 513 The Selected Case (see § 3). 16. The parallel rays meident at 45° on the first face, which is plane, are refracted into the lens so as to make with the normal (which is the axis of z) an angle of 28° 73’. The sine of this angle is 0:4714, and its cosine 0°8819. The direction-cosines of the rays incident on the second or convex face of the lens are therefore— pe. ime = 0-4714, no. = 0°8819 ; and are the same at all points. Our calculations relate to a single narrow annulus of the convex face,* the axis of this annulus being the same as that of the lens. The radius of the annulus will be taken as the unit of length, and the centre of the annulus as the origin of co-ordinates. The normals at all points are equally inclined to the axis of 2, and we assume the sine of this inclination to be 0°1000; in other words, the radius of curvature is taken as ten times the radius of the annulus.. This makes the cosine 0:9950, and the angle itself about 5° 44%’. Let @ denote the angular distance of any point of the annulus from the summit (which is the farthest point from the source). Then the co-ordinates of the point are— Zp. — sin 0, Ty = GOS Ay = (Op and the direction-cosines of the forward-drawn normal at the point are i = x5 sin 8, i = 5 cos 0, ni — 09950: From these we deduce, for the angle of incidence x, cos X = Gly + imym2+ mn. = 0:04714 cos 6+ 0°87749, and sin x’ is 2 siny; hence cos y’ is known, and we have all the data for calculating the direction-cosines /’ m’ m’ of the emergent ray by equations (3). The following values are thus found :— | | | | Opa 30% GONE Y\ SOPs 20%. t) HUSONI Riso ( bp A. 322. A. 323. A. 324, B. 207. B. 208. B. 209. B. 210 . 320. . 321. A Memoir on Integral Functions. By E. W. Banyzs, M.A., Fellow of Trinity College, Cambridge. Price 4s. 6d. Title, Contents, Index, &c., Vol. 199. Price 6d. Mathematical Contributions to the Theory of Evolution.—XI. On the © Influence of Natural Selection on the Variability and Correlation of Organs. By Kazu Pearson, F.R.S., University College, London. Price 3s. On the Equilibrium of Rotating Liquid Cylinders. By J.H. Juans, B.A., Isaac Newton Student and Fellow of Trinity College, Cambridge. Price 2s On the Measurement of Temperature. Part I.—On the Pressure Co- efficients of Hydrogen and Helium at Constant Volume and at different Initial Pressures. Part IJ.—On the Vapour Pressures Liquid Oxygen at Temperatures below its Boiling-Point on the Constant-Volume Hydrogen and Helium Scales. Part I1].—On the Vapour Pressures of Liquid Hydrogen at Temperatures below its Boiling-Point on the Con- stant- Volume Hydrogen and Helium Scales. By Morris W. Travers, D.Se., Fellow of University College, London, Grorcz SENTER, B.Sc., and ADRIEN JAQUEROD, D.Sc. Price 5s. On Some Definite Integrals and a New Method of Reducing a Function of Spherical Co-ordinates to a Series of Spherical Harmonics. By ARTHUR Scuuster, F.R.S. Price 2s. Title, Contents, Index, &c., Vol. 194. Price 6d. Thermal Adjustment and Respiratory Exchange in Monotremes and Marsupials.— A Study in the Development of Homeothermism. By C. J. Martin, M.B., D.Sc., Acting Professor of Physiology in the University of Melbourne. Price 2s. The Pharmacology of Pseudaconitine and Japaconitine considered iv Relation to that of Aconitine. By J. THEODORE CasH, M.D., F.RS. Regius Professor of Materia Medica in the University of Aberdeen, ana WrynpHAM R. Dunstan, M.A., F.R.S., Director of the Scientific ane Technical Department of the Imperial Institute. Price 3s. The Pharmacology of Pyraconitine and Methyl-Benzaconine considereé in Relation to their Chemical Constitution. By J. THEODORE Casu, M.D., F.R.S., Regius Professor of Materia Medica in the University of Aberdeen, and WynpHam R. Dunstan, M.A., F.R.S., Director of the Scientific and Technical Department of the Imperial Institute. Price 1s. The Structure and Development of the Stem in the Pteridophyta and Gymnosperms. By Epwarp OC. Jrerrrey, Ph.D. (Harvard). Price 3s. | CATALOGUE OF : SCIENTIFIC PAPERS COMPILED BY THE ROYAL SOCIETY OF LONDON. Volumes I—VI for the years 1800—1863, cloth (Vol. I in half morocco) £4 net, half morocco £5 5s. net. Vols. VII, VIII for the years 1864—1873, cloth £1 lls. 6d. net, half morocco £2 5s. net. Single volumes, cloth 20s., or half morocco 285. net. Vols. IX, X & XI for the years 18'74—1883, cloth £1 5s. net, half morocco, £1 12s. net, each. Now Ready. Demy Ato. Pp. xxxii and 807. 3 Vol. XII, being a Supplementary Volume for the Period 1800—1883. Price :—Cloth, 25s. net ; half morocco, 32s. net. (A reduction on the above prices to Fellows of the Royal Society.) Published and sold by C. J. Chay anp Sons, Cambridge University Press Warehouse, Ave Maria Lane. Published by Kr@an Pavx, TRENCH, TRUBNER, AND Co. In 1 vol., 4to. Pp. 500. With 6 Chromolithographs of the remarkable Sunsets — of 1883 and 40 Maps and Diagrams. THE ERUPTION OF KRAKATOA AND SUBSEQUENT PHENOMENA. Report of the Krakatoa Committee of the Royal Society. Edited by G. J. SYMONS, F.RB.S. Price 30s. To Fellows of the Royal Society and of the Royal Meteor Society, 20s. Just Published. Demy 8vo., pp. 160. Price 10s. REPORTS TO THE EVOLUTION COMMITTEE OF oe : : ROYAL SOCIETY. Reeort I.—Experiments undertaken by W. Batsson, F.R.S., and = E. R. SAUNDERS. REPORTS TO THE MALARIA COMMITTEE OF THE : ROYAL SOCIETY. First Series, 1900. Demy 8vo., pp. : ee Price 2s. 6d. Second ,, 1900. fe “ 2. sy nl Sem. ehird 3,69: 900: is i © i 2s. Rourth 3. 1001 i si 20. Oa ate Pigs ss 200 Os a A 79. eo DSO Sixth ,, 1902. 5 t 23) i ls. Seventh,, 1902. a 5 52. di: 35. Published and sold by HARRISON AND SONS, 45, 46, & 47, ST. MARTIN’S LANE, W.c, 9 R. FRIEDLANDER & SOHN, BERLIN, AND ALL BOOKSELLERS. Fo eet i © A Same ee Lap nig eR : JANUARY 16, 1903. apse re) % ty / ae PROCEEDINGS OF THE ROY AL Sf O0CTE TY. VOL. LXXI. No. 470. CONTENTS. Water with Temperature. By HE. G. Coxer, M.A. (Cantab.), D.Sc. (Edin.), Assistant Professor of Civil Engineering, and S. B. Crement, B.S8e., Demonstrator of Civil Engineering, both of McGill! University, Montreal. Communicated by Professor OsBorne ReynNoups, F.R.S. ° Isomeric Change in Benzene Derivatives.—The Interchange of Halogen and Hydroxyl in Benzenediazonium Hydroxides. By K. J. P. Onroy, Ph.D., M.A., St. John’s College, Cambridge, Demonstrator of Chemistry, . St. Bartholomew’s Hospital. Communicated by Professor H. E, Anm- STRONG, F.R.S. On certain Properties of the Alloys of the Gold-Silver Series. By the late Sir W. C. Roserts-Avsten, K.C.B., D.C.L., F.R.S., and T. Kirxr Rosz, D.Se. (Plate 3) ; : ; 5 Abnormal Changes in some Lines in the Specirum of Lithium. By Huan Ramaae, B.A., St. John’s College, Cambridge. Communicated by Pro- fessor G, D. LivxriIne, F.R.S. . : : ! : An Error in the Estimation of the Specific Gravity of the Blood by Ham- merschlag’s Method, when employed in connection with Hydrometers. By A. G. Levy, M.D. Gees Communicated by Sir Victor Horsizy, F.RS. . ; : 3 : : Quaternions and Projective Geometry. By Cuarizs J. Jory, F.T.C.D., Royal Astronomer of Ireland. Communicated by Sir Rozerr S. Le ERS. The Stability of the Pear-shaped Figure of Equilibrium of a Rotating Mass of Liquid. By G. H. Darwin, F.R.S., Plumian Professor and Fellow of ‘Trinity College, in the University of Cambridge _ Minutes of Meeting of December 11, 1902 Price One Shillz goad Ene FEB 9 1903 fies Lipeatls PAGE An Experimental Determination of the Variation of the Critical Velocity of 152 161 164 171 177 178 vi NOTICE TO AUTHORS. — Authors of Papers are urgently requested to send in all _ Drawings, Diagrams, or other Illustrations in a state suitable for direct photographie reproduction. They should be drawn on a large scale, in Indian ink, on a smooth white surface. Where illustrations are numerous, much time would be saved if authors would indicate in advance those considered essen- tial to the understanding of their papers. NOTICES TO FELLOWS OF THE ROYAL SOCIETY. The Council have directed that the Minutes of the Meetings of the Society shall in future be sent out as an inset in the ‘ Proceedings,’ separately paged, and shall afterwards be republished in the ‘ Year-Book.’ The ‘ Proceedings * are sent by post to every Fellow of the Society who resides within the limits of the Postal Union. The ‘Philosophical Transactions’ are delivered only to those Fellows who call for them, or who send a written applica- tion to the Assistant Secretary. Such an application may, if so desired, be filed as a standing order. The separate Papers forming the ‘ Philosophical Transactions’ will be sent post free, as issued, to those Fellows who prefer to have them in that form. And, on application to Messrs. Harrison and Sons, 46, St. Martin’s Lane, these will be bound in a cloth case for 2s. 6d., or the cloth cases for binding may be purchased, price ls. 6d. Demy 8vo, Cloth, 12s. 6d., with Plates and Woodcuts, 427 pp. THE RECORD OF THE ROYAL SOCIETY. SECOND EDITION, 1901. Containing an account of the Foundation and Early History of the Society, its Benefactors, Trust Funds, &c., with Chronological and Alphabetical LISTS OF THE FELLOWS FROM THE FOUNDATION. Demy 8vo, pp. 265. Price 5s. THE YEAR-BOOK OF THE ROYAL SOCIETY, 1902. Principal Contents :—List of Fellows, Committees, &c.; Statutes and Regulations; Business of the Society during 1901, including Anniversary Meeting with President’s Address, Council’s Report, and Treasurer’s Accounts; Arrangements for 1902; Obituary Notices of Fellows deceased. Demy 8vo, pp. 120. Price 4s. OBITUARY NOTICES OF FELLOWS OF THE ROYAL SOCIETY. Reprinted from the Year-Book of the Society, 1900, 1901. With an Index to the Obituaries published in the ‘ Proceedings’ from 1860 to 1899. Published and sold by HARRISON AND SONS, ST. MARTIN'S LANE. Sold by Harrison aud Sons, St. Martin’s Lane. PHILOSOPHICAL TRANSACTIONS FOR 1901-2. Szrizs A. Containing Papers of a Mathematical or Physical Character. Vol. 197. With 12 Plates. Price £1 15s. Od. Vol. 198. With 6 Plates. Price £1 8s. 6d. Srrizs B. Containing Papers of a Biological Character. Vol. 198. With 11 Plates. Price £1 5s. Od. Vol. 194. With 49 Plates. Price £3 Os. Od. Published by Dutav & Co., 37, Soho Square. RECENT PAPERS IN THE ‘PHILOSOPHICAL TRANSACTIONS.’ A. 323. On the Measurement of Temperature. Part I.—On- the Pressure Co- efficients of Hydrogen and Helium at Constant Volume and at different Initial Pressures. Part IJ.—On the Vapour Pressures of Liquid Oxygen at Temperatures below its Boiling-Point on the Coustant-Volume Hydrogen and Helium Seales. Part I11]—On the Vapour Pressures of Liquid Hydrogen at Temperatures below its Boiling-Point on the Con- stant- Volume Hydrogen and Helium Scales. By Morris W. TRAvERs, D.S8c., Fellow of University College, London, GzoreGr SrntrER, B.Se., and ADRIEN JAQUEROD, D.Sc. Price 5s. A. 324. On Some Definite Integrals and a New Method of Reducing a Function of Spherical Co-ordinates to a Series of Spherical Harmonics. By ARTHUR ScnuvusteR, F.R.S. Price 2s. A, 325. Effects of Strain on the Crystalline Structure of Lead. By J.C. W- Humerey, B.Sc. (Vict.), 1851 Exhibition Research Scholar (University . College, Liverpool), St. John’s College, Cambridge. Price 2s. i A. 326. The Fracture of Metals under Repeated Alternations of Stress. By J. A. Ewine, LL.D., F.R.S., Professor of Mechanism and Applied Mechanics in the University of Cambridge, and J.C. W. Humrrey, b.A., St. John’s College, Cambridge, 1851 Exhibition Research Scholar, University College, Liverpool. Price 1s. 6d. A. 327. The Stability of the Pear-shaped Figure of Equilibrium of a Rotating Mass of Liquid. By G. H. Darwin, F.R.S., Plumian Professor and i Fellow of Trinity College, in the University of Cambridge. Price 3s. { B. 206. ‘Title, Contents, Index, &c., Vol. 194. Price 6d. i B. 207. Thermal Adjustment and Respiratory Exchange in Monotremes and ik Marsupials.—A Study in the Development of Homeothermism. By | C. J. Martin, M.B., D.Sc:, Acting Professor of Physiology in the i University of Melbourne. Price 2s. B. 208. The Pharmacology of Pseudaconitine and Japaconitine considered in S i Relation to that of Aconitine. By J. THropors Casu, M.D., F.R.S. i . Regius Professor of Materia Medica in the University of Aberdeen, and i if WynpHam R. Dunstan, M.A., F.R.S., Director of the Scientific and ‘ Technical Department of the Imperial Institute. Price 3s. i _ 8B. 209. The Pharmacology of Pyraconitine and Methyl-Benzaconine considered ; . in Relation to their Chemical Constitution, By J. Tozoporr Cagn, . ; = M.D., F.R.S., Regius Professor of Materia Medica in the University of a | Aberdeen, and WynpHam R,. Dunstan, M.A., F.R.S., Director of the Scientific and Technical Department of the Imperial Institute. i Price Is, ie B. 210 The Structure and Development of the Stem in the Pteridophyta and Gymnosperms. By Epwarp C. JEFFREY, Ph.D. (Harvard). ; Price 3s. ” (aie aap > aoe Ser Pie CATALOGUE OF > SCIENTIFIC PAPERS COMPILED BY THE ROYAL SOCIETY OF LONDON. Volumes I—VI for the years 1800—1863, cloth (Vol. I in half morecco) £4 net, half morocco £5 5s. net. Vols. VII, VIII for the ysars 1864—1873, cloth £1 11s. 6d. net, half morocco £2 5s. net. er volumes, cloth 20s., or half morocco 28s. net. Vols. IX, X & XT for the years 1874—1883, ‘cloth £1 5s. net, half morocco, £1 12s. net, each. Now Ready. Demy 4to. Pp. xxxii and 807. Vol. XII, being a Supplementary Volume for the Period 1800—1883. Price :—Cloth, 25s. net ; half morocco, 82s. net. (A reduction on the above prices to Fellows of the Royal Society.) Published and sold by C. J. Cuay anp Sons, Cambridge University Press Warehouse, Ave Maris Lane. Published by Ke@an Patt, Trencu, TRUBNER, AND Co. Ia 1 vol., 4to. Pp. 500. With 6 Chromolithograpis of the remarkable Sunscts of 1883 and 40 Maps and Diagrams. THE ERUPTION OF KRAKATOA AND SUBSEQUENT PHENOMENA. Report of the Krakatoa Committee of the Royal Society. Edited by G. J. SYMONS, F-.R.S. Price 30s. To Fellows of the Royal Society and of the Royal Meteorological — Society, 20s. Just Published. Demy 8vo., pp. 160. Price 10s. REPORTS TO THE EVOLUTION COMMITTEE OF THE ROYAL SOCIETY. Reeort I.—Hxperiments undertaken by W. Bateson, F.R.S., and Miss EB, R. SAUNDERS. REPORTS TO THE MALARIA COMMITTEE OF THE ROYAL SOCIETY. First Series, 1900. Demy 8vo., pp. 76. Price 2s. 6d. Second ,, 1900. a Pe oe. so bate Fhird 535-1900. e Ss 45, A 2s. Fourth: 2 -190b os 3 * 20. 5). OSs, OG Fifth <9 34901; - os 19, 3p eS ae DIxth .s) 1902: = 2 23. 5 is. Seventh,, 1902. ao Re Hee Ht 3s. Published and sold by HARRISON AND SONS, 45, 46, & 47, ST. MARTIN’S LANH, W.C., R. FRIEDLANDER & SOHS, BERLIN, AND ALL BOOKEELLEES. ee) ee ee eee PROCEEDINGS OF THE ROYAL SOCIETY. VOL. LXXI. No. 471. CONTENTS. PAGH On the “ Blaze-currents” of the Incubated Hen’s Egg. By Aveustus D. Water, M.D.,F.RS. . ; ; ‘ : ; : ; : . 184 On the “Blaze-currents” of the Crystalline Lens. By Aveustts D. Water, M.D., F.RS., assisted by A. M. Water : : ; . 194 A Contribution to the Question of Blaze Currents. By Dr. ARNo~D Dotria, of Vienna, Communicated by Avetustus D. Watter, M.D.,F.RS. . 212 The Specific Heats of Metals and the Relation of Specific Heat to Atomic Weight. Part II. By W. A. Titpen, DSc., F.R.S., Professor of Chemistry in the Royal College of Science, London . : : . 220 Preliminary Note on the Relationships between Sun-spots and Terrestrial Magnetism. By C. Cures, 8c.D., LL.D., F.RS. . t : . 221 Characteristics of Electric Earth-current Disturbances, and their Origin. By J. H. Taytor. Communicated by Sir Oxiver Lopes, F.RS. - . 225 Solar Eclipse of 1900, May 28.—General Discussion of Spectroscopic Results. By J. EversHep, F.R.A.S. Communicated by the Joint Permanent Eclipse Committee . é ; : - ‘ : . : : . 228 a On the Electrodynamic and Thermal Relations of Energy of Magnetisation. & By J. Larmor, M.A., D.Sc., Sec. RS... : : 3 : : . 229 _ The Spectrum of y Cygni. By Sir Norman Lockyer, K.C.B., F.R.S., and i F. E. Baxanparr, A.R.CSe. . reeae : 0 246 E Minutes of Meetings of January 22 and January 29,1903. : : i NOTICE TO AUTHORS. Authors of Papers are urgently requested to send in all Drawings, Diagrams, or other Illustrations in a state suitable for direct photographic reproduction. They should be drawn on a large scale, in Indian ink, on a smooth white surface. Where illustrations are numerous, much time would be saved if authors would indicate in advance those considered essen- tial to the understanding of their papers. SEG NOTICES TO FELLOWS OF THE ROYAL SOCIETY. The Council have directed that the Minutes of the Meetings of the Society shall be sent out as an inset in the ‘ Proceedings,’ separately paged, and shall afterwards be republished in the ‘ Year-Book.’ The ‘ Proceedings * are sent by post to every Fellow of the Society who resides within the limits of the Postai Union. Tie ‘Philosophical Transactions’ are delivered only to those Fellows who call for them, or who send a written applica- tion to the Assistant Secretary. Such an application may, if so desired, be filed as a standing order. The separate Papers forming the ‘ Philosophical Transactions’ will be sent pest free, as issued, to those Fellows who prefer to have them in that form. And, on application to Messrs. Harrison and Sons, 46, St. Martin’s Lane, these will be bound in a cloth case for 2s. 6d., or the cloth cases for binding may be purchased, price Is. 6d. Demy 8vo, Cloth, 12s. 6d., with Plates and Woodcuts, 427 pp. THE RECORD OF THE ROYAL SOCIETY. SECOND EDITION, 1901. Containing an account of the Foundation and Early History of the Society, its Benefactors, Trust Funds, &c., with Chronological and Alphabetical LISTS OF THE FELLOWS FROM THE FOUNDATION. Demy 8vo, pp. 281. Price 5s. THE YEAR-BOOK OF THE ROYAL SOCIETY, 1903. Principal Contents :—List of Fellows, Committees, &c.; Statutes and Regulations; Business of the Society during 1902, including Anniversary Meeting with President’s Address, Council’s Report, and Treasurer’s Accounts; Arrangements for 1903; Obituary Notices of Fellows deceased. Demy 8vo, pp. 120. Price 4s. OBITUARY NOTICES OF FELLOWS OF THE ROYAL SOCIKTY. Reprinted from the Year-Book of the Society, 1900, 1901. With an Index to the Obituaries published in the ‘ Proceedings’ from 1860 to 1899. - Published and sold by HARRISON AND SONS, ST. MARTIN’S LANE. Oe No ni ean Sa ae Sap 3 Sold by Harrison and Sons, St. Martin’s Lane. ; “PHILOSOPHICAL TRANSACTIONS FoR 1901-2. Sprizs A. Contuining Papers of a Mathematical or Physical Character. Vol. 198. With 6 Plates. Price £1 8s. 6d. Vol. 199. Price £1 6s. Od. SzRizs B. Containing Papers of a Biological Character. Vol. 193. With 11 Plates. Price £1 5s. Od. Vol. 194. With 49 Plates. Price £3 Os. Od. ae, _ RECENT PAPERS IN THE ‘PHILOSOPHICAL TRANSACTIONS.’ Published by Duzav & Oo., 37, Soho Square. A. 323. On the Measurement of Temperature. Part I.—On the Pressure Co- efficients of Hydrogen and Helium at Constant Volume and at different Initial Pressures. Purt II.—On the Vapour Pressures of Liquid Oxygen at Temperatures below its Boiling-Point on the Constant-Volume Hydrogen and Helium Scales. Part IiI.—On the Vapour Pressures of Liquid Hydrogen at Temperatures below its Boiling-Point on the Con- stant- Volume Hydrogen and Helium Scales. By Morris W. Travers, D.Se., Fellow of University College, London, GEORGE SrNnTzER, B.Sc., and ADRIEN JAQUEROD, D.Sc. Price 5s. On Some Definite Integrals and a New Method of Reducing a Function of Spherical Co-ordinates to a Series of Spherical Harmonics. By ARTHUR SCHUSTER, F.R.S. Price 2s. Effects of Strain on the Crystalline Structure of Lead. By J.C. W. Houmrrey, B.Sc. (Vict.), 1851 Exhibition Research Scholar (Univer ary College, Liverpool), St. John’s College, Cambridge. Price 2s. The Fracture of Metals under Repeated Alternations of Stress. By in A. Ewine, LU.D., F.R.S., Professor of Mechanism and Applied Mechanics in the University of Cambridge, and J.C. W. Humrrey, v.A., St. John’s College, Cambridge, 1851 Exhibition Research Scholar, "Univer sity College, Liverpool. Price 1s. 6d. The Stability of the Pear-shaped Figure of Equilibrium of a Rotating Mass of Liquid. By G. H. Darwin, F.R.S., Plumian Professor and Fellow of Trinity Collen, § in the University of Cambridge. Price 3s. Thermal Adjustment and Respiratory Exchange in Monotremes and Marsupials.—A Study in the Development of Homeothermism. By C. J. Martin, M.B., D.Sc., Acting Professor of Physiology in the University of Melbourne. Pricé 2s. The Pharmacology of Pseudaconitine and Japaconitine considered in Relation to that of Aconitine. By J. THEoDoRE Casu, M.D., F.R.S. Regius Professor of Materia Medica in the University of Aberdeen, and WrynpHAm R. Dunstan, M.A., F.R.S8., Director of the Scientific and - Technical Department of the Imperial Institute. Price 3s. The Pharmacology of Pyraconitine and Methyl-Benzaconine considered in Relation to their Chemical Constitution. By J. THEoDoRE Casu, M.D., F.R.S., Regius Protessor of Materia Medica in the University of Aberdeen, and WynpHam R, Dunstan, M.A., F.R.S., Director of the Scientific and Technical Department of the Imperial Institute. Price 1s. The Structure and Development of the Stem in the Pteridophyta and Gymnosperms. By Epwarp (©. JErFFrey, Ph.D. (Harvard). Price 3s. On the Structure of the Gills of the Lamellibranchia. By W. G. Ripewoop, D.8c., F.L.S., Lecturer on Biology at the Medical School of St. Mary’s Hospital, London. Price 9s. 6d. Now Ready. Price 8s. | Royal 4to. 112 pages, with Map and Illustrations. MAGNETIC AND METEOROLOGICAL OBSERVATIONS MADE BY THE “SOUTHERN CROSS” ANTARCTIC EXPEDITION, — 1898-1900, Under the direction of C. E. BORCHGREVINK. Published by the Royat Society of London, and sold by Harrison” AND Sons, St. Martin’s Lane. Fellows of the Royal Society can obtain copies gratis, a : application. | Published by Knean Pavt, TraencH, TRUBNER, AND Co. In 1 vol., 4to. Pp. 500. With 6 Chromolithographs of the remarkable Sunsets of 1883 and 40 Maps and Diagrams. THE ERUPTION OF KRAKATOA AND SUBSEQUENT PHENOMENA. Report of the Krakatoa Committee of the Royal Society. Edited by G. J. SYMONS, E.R.S. Price 30s. To Fellows of the Royal Society and of the Royal Meteorological Society, 20s. Just Published. Demy 8vo., pp. 160. Price 10s. REPORTS TO THE EVOLUTION COMMITTEE OF THE ROYAL SOCIETY. Rerort I.—Experiments undertaken by W. Baruson, F.R.S., and Miss E. R. | SAUNDERS, | ‘REPORTS TO THE MALARIA COMMITTEE OF THE ROYAL SOCIETY. First Series, 1900. Demy 8vo., pp. 76. Price 2s. 6d. Second ,, 1900. ‘e Pe 22 Beane EES i) Third -s55%) 1900: x a AD. . 2s. Fourth ,, 1901. 5 vi 20. 1) OBO Pitth sce FAO 01; " 79. eee Ol, Sixth. @ 1909, 55 is 93) Bs Is. Seventh ,, 1902. i is 52. a) os; Published and sold by HARRISON AND SONS, 45, 46, & 47, ST. MARTEN’S LANE, W.C., R. FRIEDLANDER & SOHN , BERLIN, AND ALL BOOKSELLERS. PROCEEDINGS OF THE ROYAL SOCIETY. 3 ’ No. 472. 7 } q : o CONTENTS. i ) 7 PAGE | Some Dielectric Properties of Solid Glycerine. By Kxnest Witson, Pro- “4 fessor of Electrical Engineering, King’s College, London. Communi- cated by Sir WittiaM Preece, K.C.B., F.R.S. - : : : Daa ‘The Relation between Solar Prominences and Terrestrial Magnetism. By | Sir Norman Lockyer, K.C.B., F.R.S., and Witttam J. 8. Locxyer, Me Ae, RAS (Plates 455) 28 6 to " 'Lhe Bending of Electric Waves round a Conducting Obstacle. By H. M. oa Macponatp, F.B.S., Fellow of Clare College, Cambridge ; : Ag a59 | Studies in the Morphology of Spore-producing Members.—No. V. General Comparisons, and Conclusion. By F.O. Bower, Sc.D., F.R.S., Regius Professorjof Botany in the University of Glasgow. (Abstract) . . 258 On the Negative Variation in the Nerves of Warm-blooded Animals. a y NLE. es M.D. Communicated by A.D. Wattrr, M.D.,F.RS.. 264 _ On the Decline of the Injury Current in Mammalian Nerve, and its Modifi- i cation by Changes of Temperature. Preliminary Note. By S. C. M. Sowton and J. 8. Macponap (from the Thomson-Yates Laboratory of a unk University College, Meo Communicated Professor S. SHERRINGTON, F.R.S.. : Bish : : . 282 n the Formation of Definite Figures by the Deposition of Dust. By W. J Rossztn, Ph.D., F.R.S. (Abstract) 4 : i ‘ ; : . 285 M athematical Contributions to the Theory of Evolution. On Homotyposis in ' Homologous but Differentiated Organs. 2 Kart Pearson, F.R.S., University College, London. : : : : . 288 Primitive Knot and Early Gastrulation Cavity co-existing with Independent ' Primitive Streak in Ornithorhynchus. By Professor J. T. Wizsox, M.D., and J. P. Hitt, D.S8c., University of Sydney. Communicated by Professor C. B. Howes, LLD., DS8ce., F.RS. . : “ : . 314 “Whe Brain of the Archxoceti. By G. Exxior Surra, M.A., M.D., Fellow of St. Fohn’s College, Cambridge, Professor of Anatomy, Beeneiae Govern- \ ment School of Medicine, Cairo. Communicated by Professor G. B. Howes, LL.D., D.Sc., F.R.S. . : : ; j : d ‘ . 322 inutes of Meetings of February 12 and February 19, 1903 : : : iit Price Two Shillings and Sixpence. Tue £WO bt be e ll Istlgge aN \ i MAR 306 1903.) Mancn 14, 1903. ~ NOTICE TO AUTHORS. Authors of Papers are urgently requested to send in all Drawings, Diagrams, or other Illustrations in a state suitable for direct photographic reproduction. They should be drawn on a large seale, in Indian ink, on a smooth white surface. Where illustrations are numerous, much time would be saved if authors would indicate in advance those considered essen- tial to the understanding of their papers. NOTICES TO FELLOWS OF THE ROYAL SOCIETY. The Council have directed that the Minutes of the Meetings of the Society shall be sent out as an inset in the ‘ Proceedings,’ separately paged, and shall afterwards be republished in the ‘ Year-Book.’ The ‘ Proceedings * are sent by post to every Fellow of the Society who ae within the limits of the Postal Union. The ‘Philosophical Transactions’ are delivered only to those Fellows who call for them, or who send a written applica- tion to the Assistant Secretary. Such an application may, if so desired, be filed as a standing order. The separate Papers forming the ‘ Philosophical Transactions’ will be sent post free, as issued, to those Fellows who prefer to have them in that form. And, on application to Messrs. Harrison and Sons, 46, St. Martin’s Lane, these will be bound in a cloth case for 2s. 6d., or the cloth cases for binding may be purchased, price 1s. 6d. Demy 8vo, Cloth, 12s. 6d., with Plates and Woodcuts, 427 pp. THE RECORD OF THE ROYAL SOCIETY. SECOND EDITION, 1901. - Containing an account of the Foundation and Early History of the Society, its Benefactors, Trust Funds, &c., with Chronological and Alphabetical LISTS OF THE ee nes FROM THE FOUNDATION. Demy 8vo, pp. 281. Price ds. THE YEAR-BOOK OF THE ROYAL SOCIETY, 1903. ee Contents :—List of Fellows, Committees, &c. ; Statutes anda Regulations; Business of the Society during 1902, including Anniversary Meeting with President’s Address, Council’s Report, and Treasurer’s Accounts; Arrangements for 1903; Obituary Notices of Fellows deceased. Demy 8vo, pp. 120, Price 4s. ; OBITUARY NOTICES OF FELLOWS OF THE ROYAL SOCIETY. : Reprinted from the Year- Book of the Society, 1900, 1901. | With an Index to the Obituaries published i in-the ‘ Proceedings’ from 1860 to 1899, Published and sold by HARRISON AND SONS, ST. MARTIN’S LANE. ae ee a ee -- ae = 2 at re Pa Sold by Harrison and Sons, St. Martin’s Lane. PHILOSOPHICAL TRANSACTIONS FOR 1901-2. SErizs A. Containing Papers of a Mathematical or Physical Character. Vol. 198. With 6 Plates. Price £1 Ss. 6d. Vol. 199. Price £1 6s. Od. Series B. Containing Papers of a Biological Character. Vol. 198. With 11 Plates. Price £1 5s. Od. Vol. 194. With 49 Plates. Price £3 Os. Od. RECENT PAPERS IN THE ‘PHILOSOPHICAL TRANSACTIONS.’ . 324, . 326. . 326. . 327. . 328. . 329. . 208. . 209. Published by Dutav & Co., 37, Soho Square. On Some Definite Integrals and a New Method of Reducing a Function of Spherical Co-ordinates to a Series of Spherical Harmonics. By ARTHUR ScHusTER, F.R.S. Price 2s. Effects of Strain on the Crystalline Structure of Lead. By J.C. W. Humrrey, B.Sc. (Vict.), 1851 Exhibition Research Scholar (University College, Liverpool), St. John’s College, Cambridge. Price 2s. The Fracture of Metals under Repeated Alternations of Stress. By J. A. Ewine, LL.D., ¥.R.S., Professor of Mechanism and Applied Mechanics in the University of Cambridge, and J.C. W. Humrrey, b.A., St. John’s College, Cambridge, 1851 Exhibition Research Scholar, University College, Liverpool. Price ls. 6d. The Stability of the Pear-shaped Figure of Equilibrium of a Rotating Mass of Liquid. By G. H. Darwin, F.R.S., Plumian Professor and Fellow of Trinity College, in the University of Cambridge. Price 3s. On the Movements of the Flame in the Explosion of Gases. By Harotp Bb. Dixon, M.A., F.B.S., Professor of Chemistry in the Owens College, Manchester. Price 8s. 6d. } Report on the Eruptions of the Soufriére, in St. Vincent, in 1902, and on a visit to Montagne Pelée, in Martinique—Part I. By TEmpzsr AnpeErson, M.D., B.Sc., F.G.8., and Joun S. Frerr, M.A., D.Sc., E.R.S.E. Price 15s. 6d. The Pharmacology of Pseudaconitine and Japaconitine considered in | Relation to that of Aconitine. By J. THropoRE CasH, M.D., F.R.S. Regius Professor of Materia Medica in the University of Aberdeen, and Wrynpuam R. Dunstan, M.A., F.R.S., Director of the Scientific and Technical Department of the Imperial Institute. Price 3s. The Pharmacology of Pyraconitine and Methyl-Benzaconine considered in Relation to their Chemical Constitution. By J. THEODORE CasH, M.D., ¥.R.S., Regius Professor of Materia Medica in the University of Aberdeen, and WynpHam R. Dunstan, M.A., F.R.S., Director of the Scientific and Technical Department of the Imperial Institute. Price 1s. The Structure and Development of the Stem in the Pteridophyta and Gymnosperms. By Epwarp C. Jserrrey, Ph.D. (Harvard). Price 3s. On the Structure of the Gills of the Lamellibranchia. By W. G., RipEwoop, D.Sc., F.1.S., Lecturer on Biology at the Medical School of St. Mary’s Hospital, London. Price 9s. 6d. The Development of Echinus esculentus, together with some Points in the Development of H. miliaris and BE. acutus. By EK. W. MacBripz, M.A., D.Sc. (Lond.), sometime Fellow of St. John’s College, Cambridge, Strathcona Professor of Zoology in the McGill University, Montreal. Price 8s, 6d. Now Ready. Price 8. a Royal 4to. 112 pages, with Map and Illustrations. __ MAGNETIC AND METEOROLOGICAL OBSERVATIONS © MADE BY THE “SOUTHERN CROSS ” ANTARCTIC EXPEDITION, 1898-1900, if g Under the direction of C. E. BorCHGREVINK. i Published by the Roya Socrety of London, and sold by HARRISON y AND Sons, St. Martin’s Lane. » application. Published by Kegan Patt, TRENCH, TRUBNER, AND Co. | P In 1 vol., 4to. Pp.500. With 6 Chromolithographs of the remarkabie Sunsets q of 1883 and 40 Maps and Diagrams. oft THE ERUPTION OF KRAKATOA AND SUBSEQUENT PHENOMENA. Report of the Krakatoa Committee of the Royal Society. Edited by G. J. SYMONS, F-.R.S. - Price 30s. To Fellows of the Royal Society and of the Royal Meteorological a Society, 20s. P Just Published. Demy 8vo., pp. 160. Price 10s. REPORTS TO THE EVOLUTION COMMITTEE OF THE ROYAL SOCIETY. Report I.—Experiments undertaken by W. Barxson, Le R.S., and Miss E. R. 3 g SAUNDERS. os REPORTS TO THE MALARIA COMMITTER OF THE ROYAL SOCIETY. 4 First Series, 1900. Demy 8vo., pp. 76. Price 2s. 6d. ig Second ,, 1900. i 9 Be.) a Third ,, 1900. as i 45. ai 2s. a F ourth ue 1901. ”? ” 20. sx 38. 6d. a Fifth (4,190 Oo ae a. Sixth ,, 1902. 4 4 oS. 3 Lee Seventh,, 1902. a A 52. us 35. Published and sold by 4 HARRISON AND SONS, 45, 46, & 47, ST. MARTIN’S LANE, W.., — R. FRIEDLANDER & SOHN, BERLIN, AND ALL BOOKSRLUERS. PROCEEDINGS OF PEE ROYAL. SOCLETY. CONTENTS. The Differential Invariants of a Surface, and their Geometric Significance. By Professor A. R. Forsytu, F.R.S. (Abstract) . ; : The Hlectrical Conductivity of Solutions at the Freezing-point of Weaken By W. C. D. Wuertuam, F.R.S., Fellow of Trinity Gollese: Cambridge The Resistance of the Ions and the Mechanical Friction of the Solvent. HE Friepr. Kontravscu, Foreign Member B.S. 5 Upon the Immunising Effects of the Intracellular Contents of ae Typhoid Bacillus as obtained by the Disintegration of the Organism at the Tem- perature of Liquid Air. By ALLAN Macrapyen, M.D. Communicated by Lord Listsr, O.M., F.R.S. j Gn the Histology of Uredo dispersa, Erikss., and ie 6 AM soplaed® Hypo- me vou Uxxn, | Nor 473. PAGE thesis. By H. Marswatt Warp, Sc.D., F.R.S., Professor of ce in, the University of Cambridge. (itera) : The Cstrous Cycle and the Formation of the Corpus aie in the Shey! By Franois H. A. MarsHatz, B.A. Communicated by Professor J. C. Ewart, F.R.S. (Abstract) ; ; ; : : a ere On the Culture of the Nitroso-bacterium. By H. S. Fremurn, Lymph Laboratories, Chelsea Bridge. Communicated by Sir Micwarn Foster, K.C.B., Sec. RBS. The Statolith-theory of Geatropion By pees hee M. A,, MI. 8. E.RB.S. | : : : : : On the Laws governing cep Diuatane in ae at Low Pressures. By W. R. Carr, B.A., University of Toronto. Communicated by Professor J.J. THomsoy, FR.S. (Abstract) d : : : : : On the Optical Activity of Hemoglobin and Globin. By ArTHUR GAMGEE, M.D., F.R.S., Emeritus Professor of Physiology in the Owens. College, i ioria [aieeosity, and A. Crorr Hriu, M.A. , M.B., late George Henry Lewes Student in Physiology . : : : 5 4 On the Nucleoproteids of the Pancreas, aig wine, and Suprarenal Gland, with especial Reference to their Optical Activity. By ARTHUR GAMGEE, M.D., F.R.S., Emeritus Professor of Physiology in the Owens College, Victoria University, and WALTER JONES, Ph.D., Associate Professor of Physiological Chemistry in the Johns Hopkins University . Minutes of Meetings of February 26 and March 5, 12, 19, and 26, 1903 CD As u : Lp i ae NG Price Two Seg ova alg Aprit 7, 1903. {- ‘ aN | i o ni R: Ory 4 Poe, \ ra | r ie j i X, HY? q Wt fs 394: NOTICE TO AUTHORS. Authors of Papers are urgently requested to send in all ie Drawings, Diagrams, or other Illustrations in a state suitable for direct photographic reproduction. ‘ They should be drawn on a large scale, in Indian ink, on a smooth white surface. Where illustrations are numerous, much time would be saved : ] if authors would indicate in advance those considered essen- tial to the understanding of their papers. NOTICES TO FELLOWS OF THE ROYAL SOCIETY. The Offices and Library will be closed for the Easter Recess from the evening of Wednesday, April §, till noon on Monday, April 20. f _ The Council have directed that the Minutes of the Meetings of the Society shall be sent out as an inset in the ‘ Proceedings,’ separately paged, and shall afterwards — be republished in the ‘ Year-Book.’ . The ‘ Proceedings ' are sent by post to every Fellow of the Society who resides within the limits of the Postal Union. The ‘Philosophical Transactions’ are delivered only to those Fellows who call for them, or who send a written applica- tion to the Assistant Secretary. Such an application may, if so desired, be filed as a standing order. S The separate Papers forming the ‘ Philosophical Transactions’ will be sent post — free, as issued, to those Fellows who prefer to have them in that form. And, on application to Messrs. Harrison and Sons, 46, St. Martin’s Lane, these will be bound 04 in a cloth case for 2s. 6d., or the cloth cases for binding may be purchased, ~~ price ls. 6d. % ~ Demy 8vo, Cloth, 12s. 6d., with Piates and Woodcuts, 427 pp. THE RECORD OF THE ROYAL SOCIETY. SECOND EDITION, 1901. Containing an account of the Foundation and Early History of the Society, its ” | Benefactors, Trust Funds, &c., with Chronological and Alphabetical nt LISTS OF THE FELLOWS FROM THE FOUNDATION. Demy 8vo, pp. 281. Price 5s. THE YEAR-BOOK OF THE ROYAL SOCIETY, 1903. sg Principal Contents :—List of Fellows, Committees, &c.; Statutes and 8 Regulations; Business of the Society during 1902, including Anniversary Meeting with President’s Address, Council’s Report, 7 and Treasurer’s Accounts; Arrangements for 1903; Obituary ~~ Notices of Fellows deceased. Demy 8vo, pp. 120. Price 4s. i fe OBITUARY NOTICES OF FELLOWS OF THE ROYAL SOCIETY. ao Reprinted from the Year-Book of the Society, 1900, 1901. ia With an Index to the Obituaries published in the ‘ Proceedings’ from 1860 to 1899. Published and sold by HARRISON AND SONS, ST. MARTIN’S LANE. q A. 328. On the Movements of the Flame in the Explosion of Gases. By | a 2. ie. , f A. ae: es B A. A. Sold by Harrison and Sons, St. Martin’s Lane. PHILOSOPHICAL TRANSACTIONS FOR 1901-2. Serins A. Containing Papers of a Mathematical or Physical Character. Vol. 199. Price £1 6s. Od. Vol. 200 (1903). With 39 Plates. Price £2 3s. Od. Sgpizs B. Containing Papers of a Biological Character. Vol. 194. With 49 Plates. Price £3 Os. Od. Vol. 195. With 16 Plates. Price £1 7s. Od. RECENT PAPERS IN THE ‘PHILOSOPHICAL TRANSACTIONS.’ Published by Dunav & Co., 37, Soho Square. 327. The Stability of the Pear- eee Figure of Equilibrium of a Rotating Mass of Liquid. By G. H. Darwin, F.R.S., Plumian Professor and Fellow of Trinity College, in the University of Cambridge. Price 3s. Harorp B. Dixon, M.A., F.R.S., Professor of Chemistry in the Owens College, Manchester. Price 8s. 6d. 329. Report on the Hruptions of the Soufriére, in St. Vincent, in 1902, and on a visit to Montagne Pelée, in Martinique. —Part I. By TEMPEST AwnpeErson, M.D., BSc., F.G.S., and Joun S. Frerr, M.A., D.Sc., F.R.S.E. Price 15s. 6d, 330. Title, Contents, Index, &c., Vol. 200. Price 6d. 331. Experimental Researches on Drawn Steel. By J. ReGinaLD ASHWORTH, M.Sc. (Vict.). Price 3s. 332. The Specific Heats of Metalsand the Relation of Specific Heat to Atomic Weight.—Part II. By W. A. Titpry, D.Sc., F.R.S., Professor of a in the Royal ee of Science, London. Price 1s. the Critical Velocity of Flow of Water in ae By H. G. Coxzr; M.A. (Cantab.), D.Sc. (Hdin.), Assistant Professor of Civil Engineering McGill University, and 8. B. CimmeEnt, B.Sc. (McGill), Demonstrator in Civil Engineering, McGill University, Montreal. Price 1s. 209. The Pharmacology of Pyraconitine and Methyl-Benzaconine considered in Relation to their Chemical Constitution. By J. THEopoRE CasuH, M.D., F.R.S., Regius Protessor of Materia Medica in the University of Aberdeen, and Wrnpuam R. Dunstan, M.A., F.R.S., Director of the Scientific and Technical Department of the Imperial Institute. Price 1s. 210. QYhe Structure and Development of the! Stem in the Bisridophiytw and Gymnosperms. By Hpwarp C. Jurrrey, Ph.D. (Harvard). Price 3s. 211. On the Structure of the. Gills of the Lamellibranchia. By W. G. Ripewoop, D.8e., F.1.8., Lecturer on Biology at the Medical School of St. Mary’s Hospital, London. Price 9s. 6d. 212. The Development of Hchinus esculentus, together with some Points in the Development of #. miliaris and H. acutus. By E. W. MacBrips, M.A., D.Sc. (Lond.), sometime Fellow of St. John’s College, Cambridge, Strathcona. Professor of Zoology in the McGill University, Montreal. Price 8s. 6d. B. 213. Title, Contents, Index, &., Vol. 195. Price 6d. Now Ready. Price 8s. Royal 4to. 112 pages, with Map and Illustrations. MAGNETIC AND METEOROLOGICAL OBSERVATIONS. o MADE BY THE “SOUTHERN CROSS” ANTARCTIC EXPEDITION, 1898-1900, | Under the direction of C. E. BorCHGREVINK. Published by the Royat Society of London,,and sold by HARRISON AND Sons, St. Martin’s Lane. Fellows of the Royal Society can obtain copies gratis, on application. Published by Kraan Pat, TRENCH, TRUBNER, AND Co. i In 1 vol., 4to. Pp.500. With 6 Chromolithographs of the remarkabie Sunsets of 1883 and 40 Maps and Diagrams. THE ERUPTION OF KRAKATOA AND SUBSEQUENT PHENOMENA. Report of the Krakatoa Committee of the Royal Society. Edited by G. J. SYMONS, E.R.S. Price 30s. To Fellows of the Royal Society and of the Royal Meteoralogia a Society, 20s. x, ur 4 Just Published. Demy 8vo., pp. 160. Price 10s. | REPORTS TO THE EVOLUTION COMMITTEE OF THE ROYAL SOCIETY. | Revort I.—Experiments undertaken by W. Batsson, F.R.S., and Miss EB. By SAUNDERS. ‘ RAGS eA RR Ma tet GND a Bd Df REPORTS TO THE MALARIA COMMITTEE OF THE © ROYAL SOCIETY. i First Series, 1900. Demy 8vo., pp. ie Price 2s. 6d. ° Second ,, 1900. is - » Is. 6d. Third324)5 71900. 4 KA ip ¥ 9s. Fourth ,, 1901. Mh a 20. 9) ade he Bitth 7