eth: Cathe the th ads ee te werk 4 Wed NasbENed we scmacmeb iene LE Bs as Py Mey ~ d yA We Ook iiicta seni bara few. iat it th Y ae bane aes Ae Parca aera alie avheyvieral have Ande FP sith aad deiichcd qoute Vets by late 0a Gayot te re OMe ee ad tr tel Dl Gig Mod Hae’) rarer orip cath ‘ ER comet ot ah WN 4 Vet Mt td tether « as ads : miei Pa iy ‘ ah wiles arte w OP Oe em ae 4 4 ‘ idqud« “ey he oth Hooks et then . ei Wig ho eG, 8s vee ele ew eae Me ded: (aoa ieaapowrare hel Upeh deen cartes weaved bet, “ad ee i eee ” 4 CeCe ee te} ‘ Av Wottdanle jab PrP ee a oa ebay Heil Ms Deu ys be mr wn Se ee re ery wer i ee er ran) ee eee ee ie ys Natbetherh Otc 28 Hg a biledodelbea OU sn as ad at eB dey le sate eee +0 ‘ " 4a Wet al De ddr teehee bedi deg " f ‘ BH beri eek Wes tac as ah ol cue at ery 08D bd Pa eld dh ae Nd Ql Med be where te He He SOMA aR Herons O a it te Ma % Ecales sat la au-ad da ' (AD ee RAVE beh ke) eet ee eh ee att eri ie " Mpa bc bi alwys be iy C84 a deeded oe ed esrb Le Ream a ed ae » rarer) ; is ella “ 4 ue Ue eae Meee at err ee cd 4s PE a Wihade ee dolee Od " Oe the WE tes a Op aa 3 Wea Wei " 4a Leow det ee ee eat? elt d de duthaWh Ge Lp itieaed si tet ih ark. heey ia ivi 44 . Ue eo Phan Ho tk tH he Wald Wal Wy as [re Elegie we ) ‘ ‘ “Aiea De Wee Lt So ee en Wig de 4 Rid dons uth a 4 ‘ CA ee We ae a: tatiana agi ae Gi cic eneine r ‘ - wie Oe ee ee ee eC es Yee er aoe Et mua aeaase ie e 4 hou ata Me De Eh a hee oe Parhipnpsart ie Ne Geddy Dt rrp eu ea eek + fhe th Pei ements (ees bidet 404 ‘eo : We AeA RTE Goda eA dl edt de theMl ard! Hath (Abe ead shah ihe Hittbeibe dike be Gate ty sect Ley ee ‘ seo" oe 4 ‘ Hed og eel Oe dO Dd eth we Mie oh bebe del ds te the iain cient . a ‘ ‘ Mido edet @ that ey OU Mo eet bts 8-0: Vs OO Baobab ig ia ad Wer if ae “_ “+ t , Oh ele be teal dean ha data A i ya awae day hed Undedteees ide 4 ae chy "oe . * Ve hee bette A hel WD Dol De BAC W Gal sRuat | at aiits i ay * Get 1 He 4 ae | - pally Cy aa aris ” glen hay " Ln) A A 4 ieee Doms oe) a » ¥ ag “ . . fey ‘ Fd 4 Y “bedaas be . sua ’ Ht NWedadeded sete medics oh Ari anWod pi RN rt . see Ge > ‘ deh d & ee heO wide aed Und ae Wain "-* ny eM Oe Pde edo “ Uattoded te) arate 7) : aa ra EER tate AN A dit 10 ded hed bl eh ats OP De Oat wie c ‘ ‘ fe TeV My A et ihe Ne hag Whe the iba diy ie ade it ee ddl usd need i arnt A ae Nea r de Bah ed Agee ae at Healy i ha oil 4 tbh te a ate ° a he a a ean har iG on eye *" ei ° BAe wl da #0 ie é ii OD Mibsinde ott, " Baie ‘ rib ew nt y! ia MA 1 HN) Uy MORE ce Ue vue " ‘ et bebe ve PL ¢ ye ha ad heed wo dey im roy M En) ‘i K Ua LL sted 1” ir hade he Lah Geib gh dale tgitine » 4 ; ba Yeniet rartiaed cited Fae ie YI sa Ca or¢ if HA pe He ¥. +), todd fw Nowe na eda teDbdds. pte Ue fate ¢ i j : 14 Pe Abid Hod ay Ay Nplba Walp at Wet dheihsd deus i cy it Wen ‘ees «Vie (8\ US Mpadly a ; Cy aed hy mye Vailet eich hail Aa degen aie yeoe ade Me Be Beit a Asa SA Ch id i a tad Vette Ab itete 4 P Ge detested tity ‘ unin Ad ate ice ddan tied td Gadde ae eit HO ee Na wht hed : 4 Wok ean ( Puke, Cha tatk O8 ’ ‘ oy Pe oat Merk ett teas eat ty IV Ms bs dete heh dons dy yp : . Fue) aa tatls 8 Mie lhow aca t he naiahediesntted Tie Wty be tec etbolba the bets ‘ Pde ab iy iG af Cohg mike aA oS AM tha al) aids a pid bly ria ats ‘ , fe VP Wak dededune iat Oe NA gh eed oal sth teeta Wari vie itt * j ee bee ay ee iin bron arias «2% whetsioe Gatad (ie has . fie Ne pe of oem < S abak A ot a Ne in rast Heat Hebel a a elise a ¥ ant a, se ‘ ieee bed Hed 81a eed eu t NW S Gaisharwd P ’ Re Yo tap MoM Lk Nada Lana deat Seow i sts ease ° . «« 4 -D-ERobe ‘ Re ae he ed ai Hale ‘ie a0 ae ry ante PH, + ay sbivede a tales ah Th ed ryt Nite tei ¢ ¢ aes th Adhd dead dae la pA » SN aeH edb Qed aya da Seda Ts GeAd hall Pe LOOiva tC) + daa € ea et} het ’ we he id fed Og Oe tadethegh der i 7 . 4 +24 : Bad Ae odd the Je hates bod Date a Fin ih tide leds belt haa Pot toh OM thaibelty Ceres : ie ‘ boat “)O08-9 fap a NOR ne eT ni deta &. seas Crear . rf ‘ ‘ , 474 h Wen® NPM e ka Beal eh Weel ah Godt ad ge i uabe iat od 1M tethe th. kale i bal Web vain ) Wielew “hss icine reir "aa b Con) ant a Lai ded bed fide ee | nF WA AAG SS hm RN Mi ih) har Bada ak Bell Reda bat 40 Als dt di I 4 y site ise dal An , ane) Coie dale at f Oar dey Sena HMR heh be Didenit atone) a) ataeitenans a Sot ths + dt ood a file “ ) * ea Wded Wed WW waplouisqod te betel: my ‘he is Ma f fairey St , ¢ "Pg 4 ) ; wit ‘ 4) Bol oy 45Ba gy PEON Ho Dyan “tails ndeaened +e dade . idee rn) r { ta iad all \Y vi py! Wik Hiei ito F f ‘ sae tee Py *4 « - H tba Wi ath tthe ge dats gein dads * Le | whee wow d 4d eed ve) Cty are | Mea Wad irelber aie ee we dente’ Poy barn hed iy q ‘ PRM te Mast, Sele hs » HW Vi beastie ry ati eit WGeandsasanards ea ye pies om bay Ma - o @e i ‘ 4 ‘ hE 1 ae Wr By aut Wed 2 ‘ at Paint Hebe ISR ‘A Wales . 4 uw ‘ , We ot ” + belt (Poite Cea or by . abe > whut o De 4 ek te ; eu aa ai yt Fates Nays, i hdauaieey saest mei i) Be i ‘ aR Apa dal a bi ital Bis He We thn ti Set hed > a . Pa ON Ie nt i 10s tate ho ath be eng bee Devaney ye as oh , Hat Jadeded a) an ve. y ANKT ER a ie Wy Silky Heals a) as nee ba dedolts Bhedy Cabin Otehe Oe ed ia DAA a deed We ah ey ; . nd Webs dad , Wye ibe es ii bas ) ’ : 20 Cal be vdadea Laas i ‘: Ed behiaed: might AA aa At ' Hod fol ed Heed EDD Cade ark Lie aoa aetna ot Ua UNM ath Hea a f valet eMail ees : ‘ ‘4 At ete ht bed 0 TM dad! add at 4 i “4 = Pyt Mebed afdy dane eho wy 244 elite Uda tot ¥ aed UF de ey LW de Aaa bi Hen each agian ie atts | (et Vath eae Fe 4 , dftw i f] it ) t 1 yr , ae. at iki a ibed inace iigiatancl si uaa \ da? FU Odd ee dathy dod dow J be ths bas : loeb tay Wt 44% bed ha v.08.) wen ie at \ Pea ee i aw 1a ei? abel ihe debet ’ ay 4 ied be Ph om. De aa Nate be alget ; aed 4Uru-d Pe se by va Wy Peli Hans hag Wed ai dell Pd Mik Wade aa! * +" : : (at Dat OE deek ‘ Cade b a dadoll et - 4 ‘ ‘7 ae “aged Pitadd Wh eaedig daa ir ys au : hve ‘ a Pe LR ‘ daw 4 u Pane ad % 4 me eteiwne was ath, be O20 ide Diba Watie ee stad We Mb ibe | 4 aati a dete be he ude tee ke ella) 442 al ; S . Di tebe abe \ 4 ‘ ‘ i Waeded eb Wed bed we e 4 hea Thee is aa. y ike ¥ a4 ‘ ] ‘ , PEER Roe ten A, ions ‘ + O80 dn dea # tad at . ' wd vad Pia viet iii ee ‘ ry uy Edok delat hares he : ( 4 is ‘ Abe he eee thy mi . a ‘" : « Le we © . +d . ’ 4d Ok 8 ae eT a ia 44 7 te Ss aaa pier abd cn Vial sg Aer wares is ' i Erion . ‘oy 4 ‘ tas LAD Ce, eer cihe oo” “ imeay ae J “a * oz i 4 " + ihe waite x he try i , tet We da aug oted ea « ¢: ‘ 4; “ ¢ ‘ ae tard eed ‘ 4 | ty 4 a ] wok A) , r yi “i x aA sey Wier ie arn WEN aaa toaihsdh hel 7 ‘ ie acct wilt G tka ed DOs i aa i Hint us eal Bod Pd e sa a: 4 au oe Rey aa id ” Li Pineiro? Ve 408 Bie ) be ee ) alia deta aids em nt * 4 ‘ 4 #44 N aed wy) ey r Py device by 6 altoid ote de vb og ‘ ; ‘ #74 we ed ATA Regi ald ihe heM i fa) wand od itary wi eae ihe Bae hed Woda a ate i re . edn a } V pdeea a Uehd bded } Sade YT Be, Ped (had Cale i ; fabs We dode dah Hes Cima Cae o ) -) : A Koay sth Web theta ite) De bet dtr ewe He phere bi rH Cora, ada eae . PEVECK TEETER TMH Te MOB Te ViWtialiy dal de adads band aaa ae P) oh hea hesitant fa WF ha ea chat ty bs Heli g hy tee we 7 dee © ocd 2 PAL ROR DOPE Bi at bie or et Pst nee >ihe fags ooh ie Pa ae Width UE dosh ge Vidalia id WER oe Wow (haw de ivan rym Veo ode Wire sa ape 5 i {hdsdevads t Aya Mi “i? © detruddwt an é Ht (Fa Bt (Ot de died, i , ie: rived (eared eat ibe ay . fag atta hs hel HD whois ry ey Ui os sn . ‘ vs 4 7 de by ahi ak i ey F : te beled: : feta Hote . ae wee ‘ ‘ Sather ihe un be A 4 au : ve Murch te bps bid a4 © sea a 14, 4 a4, ee 14444 PW Wtaan ae ly eae Fat bode rt hagas es Weal aes . sas aan « 4 et ied Vath aiiie Gant . Wery ‘ we ey etd ails hee atalads & egu- vee “ 4 ne Fi +O And doi Oihed hadi de pital - rary Wail atl a Bow ‘ ‘ P ‘ a Wb iqilehnwedl Dee fa othe ie oh atte Vie Gaited he tan wi , wre ewee 7 FR hs. ine aN ateded it 7 . f at ak eR ee iv . eee oil ty *«* ban bs f ys week rr ‘f a alba etre i aie Ret en il - a4 ‘ “ he te ih Wine Ly i Me ny ie 1b ir RAD ee ald 4 on ew o> eee * pe Oe dab det ite ee] A hie naky %, renee ph tint eee ot a 4d tree ed 4 vod ard ede : ‘ 14 ate “Went “ MY AL bet basis f A tat wgighgtatrrtet WE a ail Haka ibe: visto i eek hians ait oe EG AU bg Hagin bait Rade Macon cS ret he ats » ae, Wide ene deel aig Wore eb da ae dof 6G) Alan ihed ef PC Kihara pir? Bowes 8M Ao iid: dedlett ded heal aitededods Gaded ted ale ded a Het Wide ba afew, Neth a ew oad itl “ Held iby dat i dda bot ee ee Be ke ty a Hh a fH Hepat ted 98h) edn Aad ge eC Wai aaiede) che a is ff wit u eh wba reugt ¥ , 4) the is Ad whee Ce res Uwe tow Aa aN ih Meteo lle “hohe Jed dil ged tnd A oe ee Aside oo Pate tl e : (pay Kei te Wed Lee ar vide wad tate tees ie Coe ate a ded HGS, Vemtbe abe Adee dai lho er cn Mie BUM M-Shettotl ate iby dor demi Sedo sel thot. sit hy Web Roadaak (oh heed eile PY te Hedi datiouee Cricy, oe went as Moe Baw tedi tad Hoth aw Vid tb te 6 a fl Pa edad p wo ees Meta Nirbdiifned by ead ath BU Ah hn m in ded, od CR coe aa meh deh A Mire eedla Ute Heth ed Godard < Whee Moke Het ee Set: they belle (od dey Saint HOt Mitte dy © he We et me Wend tng! Ga 4. " x Wad RiMatde a ee BR ea Wamen ae ie i aah ted Aid ed een Ae ek Reidy fue catenin . eat Was Qed they wa Reiié eet He Wt weg tat t vee We lege HAL ee te v dad arp eee dash aS . {Ve rt Ha det tet cabs ae . At shat iu eer “ tae at pid Add Me Gee hed ab tad . aso het ny ws A : We dat Pg ieee Vine hte J a shad 4 4 eet hele theft he ble Ged ark x) DW he Ret Wii Wed dat BM Bel aoe de Vk Ab usta le Gout pif het ee j Sao pany OH OT BH ded hed detailed sindbis Cray A Abe Hs shah eh Bette 7 BBL hoe eh dd de dae dH Oe icndl oe a Sones ¢ * i awe a Cee a are arian Citee oui ony eeosene a i Saas : 19 a Det ee reise abt eee a e Pee Oot een, Bee | ca ia i. ee ety ew weet meat See ne ae eH Hall oy ruin UNC 1 he etd gy abe Hered Dail i ae Witt HLA A Ae a he ae’ Me hd tb hd a hh e he ad Wedel ey alee HE sie bag 4 SE Ra OO eg thet) ay ry ry heii £ hay Ri PROCEEDINGS ROYAL SOCIETY OF LONDON. From November 30, 1899, to June 14, 1900. VOL. LXVL. LONDON: HARRISON AND SONS, ST. MARTIN’S LANE, Printers in Ordinary to Her Majesty. MDCCCC. LONDON HARRISON AND SONS, PRINTERS IN ORDINARY TO HER MAJESTY, ST. MARTIN’S LANE, CONTENTS. VOL, LXVI. — +g te No. 424. Anniversary Meeting, Prewom Ber SU, TEGO. iil detosgusesdnanasteese Meeting of December 7, 1899, with List of Vice-Presidents.... 0... List of Papers read On the Propagation of Earthquake Motion to Great Distances. By R. D. Oldham, Geological Surv ee of India. Communicated Lil Sir Robert 8. Ball, F. RS... R ae The Medusze of Stas 4 cane J. Pa ERS. . Vapour-density of Bromine at High Temperatures. By E. P. Perman, DSc, and G. A. §. . Atkinson, 1 B.Sc. Communicated by Professor Ramsay, E.R.S. e. Sua taaeiee ctraires ied socnesuaeen’ I(r) (mtr) a S, denoting 1+ = — ; [Sedan - q with the special case Sy=0, and ‘E _ 5 being log 2 + mit 2. When z isa real quantity, the function I,(~) increases from zero (or unity, when n = 0) to an infinitely large quantity, as x passes from zero to infinity, while K,(z) decreases numerically from infinity to zero under the same circumstances. The values of the functions K,(z), K;(~) have been tabulated by the present writer, and published in the ‘ Proceedings,’ for values of x at intervals of 0:1 from 0-1 to 12°0. The elements used in the calcula- tion of the earlier half of these results are available for computing the values of Ko(z) and Ki) in some cases when @ is a complex quantity. If x be a pure imaginary = 2i, z being a scalar, it is easily seen that eG RA desert, Oba seins aoa ae (4), where J,(z) is the ordinary Bessel’s function of the first kind and nth order. If also Y,,(z) denote Neumann’s function of the nth nalen and G,(z) be a function defined by the relation Cah = Es Ide Vike ea as (5), it can be shown without much difficulty that K,(@) = #G,(2) - nee A) aR en on en name (6). 3. The numerical calculation of the functions G)(7) and G,(z) can be made to depend on that of Ko(#) and K,(z) for any values of x for which the convergent series (1) and (3) are applicable. In doing this it is necessary to calculate the elements of Jo(a) and J;(z), and in- cidentally to compute these functions. With the notation used in the writer’s paper on the computation of K(x) and K,(2), it is easily seen that VOL. LXVIL. ) D yal Mr. W. Steadman Aldis. On the Numerical m= o m= © Jo(z} = By— By +B Pot sees. = 2 Byn-= Bamee ...- (7), m 1 = m = © nu = 2 oe a a B+ Bs - By +... = 2 Banya - = Bane m= 0 m= 0 Thus the elements 6, used in the computation of Ip(x) and I,(z), for any value of « can be easily used to derive the values of Jo(z) and J (7) 4, We have further (2) p Ge)? Get eae = Tyla) (Blog 2) — {rae ay ayer } Gey? @a)* , Gas also @) = Jo(v) —1+ T1(1)/2 T)P 0(3)P = wiche eo as } whence by addition Go(z) = Jo(a){H+1—-log a} + {ys—yetys—...-. a ee using the notation of the former paper | Again, Gi(0) = Ty(a)(B—log2) + - Lf Ga)X8,+8,) (= 1)°G2)4G,4 Se) +34 o— Tue) + TO “| aayTeee 4 T(r)(r + 1) bu 0 = Iya) — ae (Gee , aay tHe ee whence, adding, . 1 Hv G,(2) = J,(@){E+1-loga}+-- 7 Gai + 8s 2) 10 eee Srey) Sa 2T1(r) I(r + 1) But (ae (Si1+8_— 2) (37)°(81 — 1), Te)’ _ Bs ~ aT) ~~ TH(1)T() +45) ie. (Caen (oy + Spe _ 2) Se (30) (8, — 1) MEO 2+ 1) Wr) + 1) © 27+ 1)? @ o/2r af oF ae git gb | a Computation of the Functions G(x), Gy“), and ") Warf. 9D Hence Gi) = J,(@){H+1 -loga}+ - x ze) f ft 1 1 1 $e 4 by, —F1 t p79 — ant b {Ba Bo + Bs — -F, The last portion of this = a Ay Jo(z)-1+ ho} = - fh) Re : 2 % Mp 4 whence G(x) = Jy(o){E+1~logx} +=— a S 1 et Fe: += 2 Whe zve+ =78 - Fito po ot”) ee (10) 5. The quantities 6 and y, : 7? and the multiples of the different values of (E+1-log z) al been computed for the values of mb, Oro,...... ,6°0, m the process of calculation of K(x) and K,(z), given in the writer’s former paper. It has been, therefore, an easy matter to find by (7), (8), (9), and (10), the quantities Jo(x), Ji(~), Go(z), and G;(a) corresponding to the same values of x. The former two are of course well known, but the recalculation affords a valuable verification of the correctness of the quantities 6. The results are given in Table I, appended to this paper, negative values being indicated by the use of old numeral type. The formula used for verifying the values of I and K was ; 1 L(a) . K,(a) — 1,(#) . K,(2) = = Replacing « by 2, by means of (4) and (6), this gives a @ we Tee, tJ 1(2) { 6) - 59 o(2) } —J,(2) ft + 59 1(2) } = whence Ji(z) . Go(z) — J, (2) . Gx(e) = This formula has been applied throughout Table I to each set of four values, calculated to three places beyond those given. Where the last figure has been increased by unity, in consequence of the first omitted figure being equal to or greater than five, the fact is indicated _ by a dot after the last figure. The column Gp(x) has also been tested with satisfactory results by differencing. 6. The value of I,(~) can be readily expressed in terms of the quan- tities 6, when n is either zero or unity, in one or two other cases, beside those of #, being a pure imaginary or wholly real. 36. Mx. W. Steadman Aldis. On the Numerical For instance if = Ge — ep, then I,(22) = Po +Qoi, say, where ee Bo- By +Bs- seer, Qo = Bz - Be + Bro - ee Thus the values of Py and Qo are easily deduced, and, therefore, that of Ip(z?). The same process gives the value of Jo(z7*), for, since Jo(z) = By — Pot Ps-—Pot...... it is easily seen that Jo(ait) = By + Boi — Bit Be2 + Bs— --.-. =P) Oe...” (12). The values of Po and Qo are tabulated in the Report of the British Association for 1893, to nine places of decimals for intervals of 0°2 of aunit. Table II at the end of this paper gives them for the same number of places, and for the same intervals as have been used in the calculation of the K and G functions. — Py and Qo are denoted in the Table II by X and Y in accordance with the notation adopted by the Committee of the British Associa- tion, negative values being denoted by the use of old numeral type. 7. Assuming the accuracy of the values used for the quantities f, an accuracy guaranteed by the tests to which the Tables for I and K in the former paper have been subjected, the relation between I and J gives a very easy check for detecting and correcting any mistakes in addition or copying figures in finding the values of J. Thus nt = © L(oe= =. Bint 2Bam+2 m= 0 Jo(a) = ZBam— ZB yn+2. In finding Jo(x), =B4am and YBy,42 are separately computea by addition of alternate terms from I)(z), and the smaller sum written down below the larger. In all cases in Table I the sum of these has first been taken, and the agreement or disagreement of this sum with the known correct value of Ip(~) has shown either that there was no mistake, or has revealed where such mistake was Ss and secured its correction. A similar test of accuracy in finding J;(~) is derived from the known values of [,(2). In like manner, since Plein, Tae ~Psm + 2Bsm+4s 5 and Xx Po = LBs i >Pem+4s ti Computation of the Functions G,(a), Gy(#), and J,(a/t). 37 the known value of >8,, obtained in finding Jo(z), gives a check on mistakes in calculating X. The known value of =84n+2 does the same service in regard to the computation of Q, or Y. 8. By formula (8) J, (x) a Py ox Bs + Bs = By se ey Gs d Hence Ji(u Jt) = By — Bgit + Bylf— ...... Py cos <—Bs ee 7 +Bs cos — ee +i(Bs sin 7~Bs ones 7 +Bs ps es a ) / = 9 {Pit Ps—Bs—Br+ B+ Pa Ba-Bas. +7 (B; — Bs — Bs +B; + Bo— Pu — Bis + Bist... )s =—5 {Xi + Yyi}......, say «eee ot TS aes ae ee (13), where Xy ==) (Bsn 2S te Bsm+s) =2 (Psm+s ay Bsm-+ as Yi = (Bsm+1 3 Bsm+t) — (Bsm+s3 Bsm +5) the summation being in all cases from i = 0 to the largest value of m which gives sensible values for [. The values of Xj, Y;, computed by these formule from the known values of f, are given in Table II. : The computations evidently admit of a check to inaccuracy of the same nature as those given in the last article. Another form of the values of X, and Y; is given by : Xy oe (Bsm+1 ee Psin+s) - > (Bsm+3 a Psin+7)s Ny > (Psm+1 + Bsm+s) —2 (Bsm+s ra Bsm-+z)s which reduces the computation to that of the two quantities D2 (Bsm+1 a Psm-+s) and = (Bsm+3 = Psm+7)s so that if these be denoted by P, and Q, Xi = Pi +Q,, Yi = P, — Qi This form admits of somewhat different checks to mistakes. The values in Table II have been computed independently in the two ways, so that the writer has every confidence that they may be relied on as correct. The column for Y; has also been differenced with eatietaany results. 38° ~ Mr. W. Steadman Aldis. On the Numerical 9. The well-known sequence laws ' 2 In h@ “ Jala) dinate oh (14), AS y [da = Jy cele si (e020 16 re 8/louage 9: eile ee (15) can be utilised, the former to obtain the values of Ja(zi*), J3(z7) ..., and the latter to give a verification to some extent of the values of Ji(x*), by means of the formule given in the writer’s paper on | and K, which express dy/duz in terms of a series of equidistant values of y. Thus, since dJ,/dz = —di, replacing x by 27, and using the values already assumed for Jo(#7*) and J1(at*), it follows that d(X-Vi) og Mi+¥e 140 Mt+V dg ee Whence adX/dx = —4(X,-Yi), ae. (16). d¥/de = Y(X,+¥i) By means of the formule (18), (19), and (21) of Articles 17—19 in the paper above referred to, this formula gives a check to the series of values in Table II to a considerable number of decimal places, to thirteen places with the last approximation. 10. For determining the values of J2(ai*), J3(ai*), ... by the sequence law, it is convenient to denote these quantities by the symbol X, + Y,2 when 7 is even, and by = (Xp + Yni) when nis odd. This will be found to avoid irrational multipliers in the successive derivations. Equation (14), putting 22 for z, gives : 2n a ‘ Inia) = —.Jn(at*) — In_i(ar). al The cases of » odd and 1 even must be separately considered. First let » be odd. The equation gives, remembering that 7 = 1-2 ce Xn+1 ae Yn+10 ae ~ (L = ) (Xn ae Yn?) Fe (Xn=1 + Yn—11)- Whence, if ~ be odd, Xn41 a “(Xn T ¥;,) ne Xn nN t Yn41 = - (Yn Ai Xn) = Yn-1 Computation of the Functions G(x), G(x), and J,(a,/). 39 If, secondly, x be even, the equation gives 1 eee Sa we i) a il? Cnr 4 tYn+1) ea ay J2 (Xn os Ynt) a 12 (Xn-1 +¥ n—12); 2n whence Xn+1 =a a (Xn ah Yn) te Xn--1 2n ip Ncin == a igidde oa De sie be The most important special cases are when n = 1 and n = 2. In the first, remembering that Xo = X, Y) = — Y (Article 6), equations (17 ) give In the second case (18) gives X, = =(%+Y.)-X emi ce: Sere art Dive: (20). Y3 = fe (Yo A X2) = Yi In these derivations no labour is involved, except that of addition of known quantities, and division by z. At) |: Mr. W. Steadman Aldis. On the Numerical Table I. te Jo(a). J, (a). O 0°997 501 562 066 040 032 281 0°049 937 526 036 241 997 556 0 0°990 024 972 239 576 390 818° "099 500 832 639 235 995 398° 1) 0:977 626 246 588 296 087 570° 148 318 816 273 104 007 741 0) 0°960 398 226 659 563 450 344 "196 026 577 955 318 744 107° Q- 0°938 469 807 240 812 904 228 "242 268 457 674 873 886 384° 0 0°912 004 863 497 210 775 955° ‘286 700 988 063 915 739 746° 0 0°881 200 888 607 405 280 839° *328 995 741 540 058 947 849° 0) 0°846 287 352 750 480 266 089 *368 842 046 094 169 994 205 0 0°807 523 798 122 544 777 302 °405 949 546 078 805 674 605° ile 0°765 197 686 557 966 551 450° ‘440 050 585 744 933 515 960- 1 0:719 622 018 527 511 015 975° 470 902 394 866 292 936 849° 1 0°671 182 744 264 362 673 475 ‘498 289 057 567 215 480 211° 1 0°620 085 989 561 509 131 673 522 023 247 414 660 396 129° 1 0°566 855 120 374 288 721 361° 541 947 713 930 854 533 153° 1 0°511 827 671 735 918 128 749- 557 936 507 910 099 641 990 ‘5669 895 935 261 680 370 013 577 765 231 529 023 219 798 °581 516 951 731 165 183 470 °581 157 072 713 434 072 686° ‘576 724 807 756 873 387 202 ‘568 292 135 757 038 668 540° °5685 963 049 819 063 939 102 °589 872 532 604 313 665 317 °520 185 268 181 931 033 964 "497 094 102 464 274 038 011° ‘470 818 266 517 578 669 733° "441 601 379 118 253 106 422 “409 709 246 852 288 741 579 °375 427 481 813 095 896 391° °339 058 958 525 936 458 926° *300 921 133 101 057 626 662 ‘261 343 248 780 504 837 363° "220 663 452 985 241 082 698 "179 225 851 681 507 110 994 "137 377 527 362 327 185 716 ‘095 465 547 177 876 403 846° 053 833 987 745 461 864 015 °012 821 002 926 731 627 029 "027 244 039 620 779 926 253 "066 043 328 023 549 136 143 103 273 257 747 338 701 790° "138 646 942 126 046 167 310° "171 896 560 221 540 474 678° "202 775 521 923 086 594 695 °231 060 431 923 370 634 008 "256 552 836 097 444 561 708° "279 080 735 843 335 330 140 "298 499 858 099 557 876 149° "314 694 (671 (O15 gO. 002 2037 "327 579 137 591 465 222 038° *337 097 202 O18 gan e400 465° °343 223 (COS (871 eh 9Os 216 * °345 960 833 801 186 I99 542 345 344 790 779 586 326 575 "347 438 215 429 043 350 180° "334. 332 1836 290 Ge7 Bios 208 "324 147 680 222 656 214 217 "311 027 744 303 942 414 148 "295 142 444 729 016 123 857° "276 683 858 127 565 608 173° 0°455 402 167 639 380 713 311 *397 984 859 446 105 491 142° °339 986 411 042 558 350 093 "281 818 559 374 385 470 714° *223 890 779 141 285 668 052° "166 606 980 331 990 326 602 "110 362 266 922 173 950 988° "055 539 784 445 601 963 144 "002 507 683 297 242 813 015 "048 383 776 468 197 996 327 "096 804 954 397 038 249 909 "142 449 370 046 OII 821 820 "185 036 033 364 387 324 596 "224 311 545 791 968 114 187° "260 O51 954 901 933 437 624 "292 064 347 650 697 540 058 "320 188 169 657 122 907 289 "344 296 260 398 884 637 389 "364 295 596 762 000 469 831° *380 127 739 987 263 377 379° "391 768 983 700 797 768 519° "399 230 203 371 191 105 766 "402 556 410 178 564 169 319 "401 826 014 887 639 905 035° "397 149 809 863 847 372 287° *388 669 679 835 853 683 029 "376 557 054 367 567 663 516° ZOlImollm 7 2501525 102 tos "342 256 790 003 885 614 439° "320 542 508 985 T2T 424 355 °296 137 816 574 141 142 650° "269 330 789 419 752 826 396 "240 425 327 291 083 452 194° "209 738 327 585 326 314 755° "177 596 771 314 338 304 347 "144 334 747 060 500 516 529 "TIO 290 439 790 986 539 621 "075 803 I11- 585 584 160 063° "O41 ZO LOL) 244991 307 1084.7 "006 843 869 417 819 196 824° ‘026 970 884 685 114 476 356° "059 920 009 724 037 401 926° 0°091 702 567 574 816 188 248 0122 033 354 592 822 673 484° 0°150 645 257 250 996 931 662 (We PCDI Ee tee Ae ME Cal Ane. SS Se ore 2 2 DOE WOH OOHDIATIE WN HODHOTATEAMDNHKHOSOWAME SNHODOWACHA KE SEY SY OE OPV Os 1S) ) (Oe OTe) ©) 6) C10) (0). 0) (oe (0) Cl. 6). @)- C0) 10) OO GO OO) © ClCODADDDDADDDDDDDDDD DD DOD SOOO DSO OOOO SCOSD OOOO Oe COC OOO OOOO OCC OCCCCS QAKKAI AKT nA LA PL ADA Poo wwWwwowWnw nny nNOnnwnnwnNHeHene OXSDAIBDHLWNHOOAN | N.B.—Negative quantities are — Computation of the Functions G(x), G(x), and J,(a/t). 41 2°409 1°698 1°268 0°951 Go(z). 976 437 967 912 294, 196 269 260 531 005 062 370 733 913 360 941 166 032 609 089 248 393 783 854 194 606 170 757 539 963 495 770 651 788 694 348 702 042 021 281 840 923 388 656 204 633 715 204 053 999 725 363 498 849 106 Bye 729 071 792 761 088 686 532 541 263 764428 542 739 360 749 364-688 180 915 405 024 575 635 605 O42 351 497 427 739 947 984 061 056 004 402 985 970 970 798 696 231 883 694 215 133 899 087 413 666 046 042 540 OII 194 799 929 365 743 495 757 612 346 090 680 367 O91 369 019 468 072 323 668 009 246 357 283 363 643 701 735 229 033 948 656 746 308 772 709 085 954 611 480 711 143 944 758 310 761 413 318 117 438 413 367 688 717 401 572 140 678 928 264°659 951 914975 194 465 215 022 345 240 471 206 621 144 872 495 980 321 462 912 008 607 718 790 734 042 399 610 451 105 001 945 113 233 426 177 404 264 042 657 322 775 568 768 228 724 991 568 315 804 579 4389 841 912 363 794 369 010 072 657 030 554 737 200 240 423 447 733 800 104 815 454 758 283 862 071 400 618 352 492 666 714 171 945 773 571 858 327 751 662 450 130 045 273 081 540 584 334 453 689 368 500 254 868 317 059 586 914 487 744 863 636 468 031 370 191 891 114 925 038 474 697 38 096 884 841 038 443 695 151 000 080 566 867 denoted by old numeral type. VOL. LXVI. Table I. 552 «|: 10°145 616° | 5°221 785 3 602 045° | 2°797 778 2°311 705 ° 1°979 072° 1 °732 732° | 1°536 883° | 1°371 681 1 °227 693° | 1°096 119 0°975 114 0°861 172° | OF752 674 0°647 540 0 545 063° 0 *4.47 754° | O-°35l1 205° | 0°258 426° | 0:168 664 0-081 399° © *002 aoa - 0'082 037 O°157 035° 0°229 676 0°295 757 0°357 974 0°413 294 0 *464 de o.509 627 0°549 641° 0 °582 599 | 0609 O52 0 °629 873° 0 °644 252) | OPes2 327 0°654 gt. | ©7650 784. 0 °640 410 0°625 331 0 °604 155 0 °578 685 O°547 GIs erat 72° li pita 72 463 © "429 540° | 0°384 258 07335 448° | 07284 073 6\°232 198 - 0°178 166 OA 24. 629° 0 °069 921 O°OIs 380° | 0°037 272 0°089 880° | 0:°139 DOL B87 599 0 °232 0-274 696 052 OOo1 387 383 818 980 465 504: 126 603 678 612 642 652 974, 246 331 247 126 176 337 rig 847 207 880 564 976 861 997 196 312 237 913 322 495 510 489 602 060 118 073 255 033 805 998 O61 467 701 262, 656 391 975 907 319 230 366 292 599 905 @,(2). 654 082 128 265 429 098 846 279 028 230 640 743 777 307 876 258 939 953 983 150 574 O13 O15 655 675 763 209 136 550 393 706 008 959 347 460 354 574 $32 188 244 897 166 635 532 489 O19 738 380 638 882 785 899 236 873 304 050 297 507 047 605 5U5 235 335 631 386 470 329 810 549 143 617 748 028 119 756 657 888 309 282 312 108 951 702 Pie 130 567 833 265 729 867 485 724, 388 $32 Pia 105 o81 336 665 480 200 790 837 2,62, 45> 780 984 273 108 507 280 873 430 305 A83 440 368 445 277 978 820 445 994 180 204 115 515 311 450 038 382 571 960 170 331 771 467 506 742 999 347 430 327 404 804 986 978 315 291 745 216 205 298 774 677 572 513 516 292 477 408 34! 782 814 897 635 gio 139 o71 038 088 286 977 886 273 342 812 030 738 744 909 175 455 510 266 572 722 701 555 729 489 561 263 588 213 095 467 173 750 884: 935 549 941 981 366 077 R12 131 314 627 343 624 O45 935 686 G33 078 622 O94 742 903 671 300 775 4.04 279 344 875 857 295 221 BD: 134 453 552 230 414 402 860 357 743 883° 007 589 834 022 757° 299 782 243 767 436 ° 570° 489 ° 194: 494: 248 ° 701 849 - 228 ° 604; 779 444° 030 462 986° 344 572° 288 674° 493 6aT * 217 549° 523 102° 136 620 ost 495° 826° 183° 759° 617"; 056" 891° 442 255 409 237° 574 313 649 148° 517° 040° 285 142 025 452 * OurranGannns see PEE ER RPWOwWWWWawwowWwww 8 PNW NW NNER EB EP RIP RHE REREOSCOCOOCCCOO. SIE ONE OODTIAMEONHOCMTIAOA ONE OOOISHKSNWEOSHYIAGTKEADNHOSHATAHEWNE OCHIAI! We re lien ly ba J 42 Mr. W. Steadman Aldis. On the Numorgen Table II. Se) ee x ¥ 8 .e) NNN NNWNH EE EEE H HERO OOCOCOCCSC CIAWHO HE OSHABTKROWHOdOMAGMER Ow: "999 998 437 500 067 816 840° 0°002 499 999 565 972 229 004° °999 975 000 O17 361 109 182 0°009 999 972 222 229 166 666 "999 873 437 944 946 038 780 0°022 499 683 594 150 451 545 "999 600 004 444 436 543 214° | 0°039 998 222 229 333 326 883 "999 023 463 990 838 255 555° 0-062 493 218 382 199 458 650 "997 975 113 905 224 846 398° 0-089 979 750 410 060 617 063° "996 248 828 444 070 123 287 0°122 448 938 981 613 810 260 "993 601 137 745 414 585 178 0°159 886 229 503 894 323 928 "989 751 356 659 594 O19 089 0°202 269 363 489 470 399 618 ‘984 381 781 213 086 883 966° 0°249 566 040 036 659 721 419 ‘977 187 973 1638 994 306 095° | 0°301 731 269 206 265 863 908° ‘967 629 155 801 183 528 979° 0°358 704 419 873 150 681 448 "955 428 746 808 400 572 511 0°420 405 965 634 100 168 746 "940 075 056 652 724 712 846 0°486 733 933 588 908 060 448° "921 072 183 546 255 764 122 0°557 560 062 303 O86 694 894 0 0 0 S[Coo00ooqooeocoocooco °897 891 138 567 705 276 346 °632 725 677 031 398 154 882° ‘*869 971 236 987 757 520 821° ‘712 087 292 354 219 242 730° 836 721 794 210 160 854 515° ‘795 261 954 775 638 372 738° ‘797 524 166 991 521 789 701 "751 734 182 713 808 228 551° °698 685 001 425 6385 398 101°: "637 690 457 109 552 833 002 568 048 926 137 096 187 234 "489 047 772 101 826 069 086° °357 485 476 450 273 287 287 *399 968 417 129 531 339 957 ‘457 182 044 159 804 184 047° 0°882 122 340 574 509 297 036 O 1 I il 1 L °300 O42 090 306 787 850 787: 1°556 877 773 663 311 509 857 1 1 1 1 2 2 2 ‘972 291 627 306 661 206 104 "065 388 160 849 286 232 192° "160 969 943 770 221 785 831° °258 528 975 115 816 306 932° "188 706 303 992 608 423 524 ‘655 742 407 252 085 252 722 065 112 108 427 346 531 305 "752 850 563 814 438 038 253 1071-349 825 S3i sags Coz san °847 176 115 683 253 092 922° "221. 380 249 598-693 888 868 ‘937 586 785 266 042 766 897° 385° 531 454 077 28iea 13 e3ua "022 839 041 963 733 753 825 "564. 376 430 484 566 549 458° ‘101 573 388 135 250 371 321 "758 407 O12 072 785 084 982 °172 310 181 492 460 325 998° "968 038 995 314 976 506 884° "193 598 179 589 928 ‘060 082° "435 305 321 718 847 744 816° "693 259 984 269 599 885 400 "967 423 272 739 419 648 007 "257 599 466 142 987 708 599 "563 416 557 258 579 754 134° "884 305 732 008 850 753 468° "219 479 832 260 939 763 946° 567 GIO 1662 Soon 221 bona aya "928 306 621 502 089 326 988° "872 563 795 777 954 293 134 *299 086 551 599-756 238 427 686 O17 203 632 189 319 953° 2°233 445 750 279 040 972 132° 2 2 2 2 2 2 2 2 2 1 1 678 356 937 208 980 936 827° | 1°461 036 835 928 036 069 728 1 O 0 0 fo) fo) if 2, z 3 4 5 6 f ‘283 249 966 853 914 618 212 ‘319 863 654 812 663 506 793° 341 297 714 476 542 058 301° 345 433 061 385 529 680 393° 330 021 882 265 074 524 014°: ‘292 690 322 699 299 833 586° -230 942 780 326 965 102 027 142 167 986 657 022 °889 923° 023 647 069 440 171 807 909 1063 885 5X6 719 Sox 887 52m: "194 600 796 822 301 663 253° "453 076 174 855 458 180 119 ‘883 656 853 707 154 174 111 "842 942 441 915 628 551 218 525 146 810 908 826 889 589° 230 082478 1666 4578735 0085 "116 034 381 550 200 378 097 "610 653 357 304 £70 918 646 °346 663 217 591 247 641 801 "980 346 402 874 876 505 440° "865 839 727 484 430 267 303° "334 363 435 462 957 925 254° "444 260 150 604 921 519 731 "667 394 351 327 397 532 141 "084 516 693 093 664 203 000 "973 596 450 774 417 438 658° "788 980 154 734 066 597 920° "246 575 961 893 122 136 086 "559 746 593 355 732% 204 313° "479 37% 252 085 205 623 568 "398 579 IT1 649 335 813 378° "664 445 263 435 904 450 574 "306 844 640 335 221 439 301° 793 O06 G58) 132 1376 304 201 "285 445 622 573 310 185 248 "858 315 966 O45 036 088 551 "334 746 540 847 962 419 331° COO” COM mY vv NNANMrIMAMBBWWW Pp hd bP ew we “=O00000 000 CODeGCGCGOCOCO SCO Qaranradanands Abe PAAR Ea BHWwWwWwWWowwowWunwnDy OOHNIAGE ONO HOOHBIATLMONHOSMTAME AONHOSHOWS N.B.—Negative quantities are Computation of the Functions G(x), G(x), and J,(a/2). 43 Table IT. 0°050 062 473 952 908 664 100 499 165 972 569 "151 681 160 023 124 "203 973 244 622 459 "257 730 697 263 717 *313 295 988 104 834 °370 995 377 179 567 "431 135 380 394 708 "493 999 080 858 644 ‘009 842 263 647 128 *628 889 353 965 588 oooqoooqcoocoo xX, 1 Jy(uvi) = ve (X, + Yj) 0°701 329 140 828 180 0°777 310 270 858 051 ¢ 0°856 936 499 638 710 0°940 261 691 231 ty 212 *309 ‘410 *513 618 °725 833 "941 "049 "155 *258 NNNNNNNN NH BH HY BHP Re ee Re Se BESS 555 OO’ AMNPWNHH HO COOK HK ENNN NW Cae ee Oa . ee A : OU J ~ 943 109 581 080 242 608 625 631 852 394. 265 637 089 516 654. 089 267 514 046 997 440 417 975 801 740 538 704 502 88r 405 174 738 006 143 153 075 466 738 093 858 637 364 255 759 538 568 441 956 117 644 127 975 307 948 721 229 329 022 A492 054 747 359 664 650 542 917 215 O41 O37 144 823 669 346 188 263 733 400 362 "027 284 560 039 161 rage 872 093. 930 156 867 016 102 160 158 452 082 287 829 561 561 996 521 838 475 976 316 629 387 137 608 11] 541 328 265 866 560 019 033 955 119 178 350 866 287 933 143 149 654 790 065 231 480 661 393 259 276 817 962 149 253 742 032 838 629 100 345 464 751 066 822 862 798 894 174 046 612 966 207 728 009 804 775 090 254, 841 920 293 493 733 185 235 898 647 785 864. 907 953 209 262 508 230 ivy 354. 75d 876 343 673 580 801 879 723 IOI 006 780 958 862 820 143 636 870 172 983 477 352 956 228 196 608 221 416 285 089 387 864 207 802 983 598 259 630 365 398 482 341 609 083 880 648 750 803 395 874 denoted by old numeral type. = VOL. LXVI. IVE 710° 748° 206 626° goo 768 207° | | fe) IO ie) If EE WOO COCOVM TIT AAMNNABAAHRWWNRNRNH RM HOO OOOO coooecooooooecocecoeooesoece 148 °195 "242 286 °328 367 402 "434, 462 "485 *503 "515 520 518 *508 "489 “461 4.23 ‘375 314 °241 5 "054 "060 "192 “340 *5O5 °689 ‘891 "113 S55 617 "900 204 “529 "874 *240 "628 "029 “451 “889 342 "806 °280 °760 "242 724 “199 "663 Ilo °533 "926 279 "586 835 °o18 Ri 123 "139 My 049 937 473 963 '759 358 667 099 499 167 361 458 217 565 306 183 753 572 747 019 399 762 737 375° 968 702 488 384 010 431° 973 106 299 129 158 925 950 725 717 365 079 239 195 266 744. 892 946 120 608 587 225 685 868 261 298 754 364 806 693 B5° 805 762 59° 292 690 SOF 791 988 813 77° O16 315 023 038 480 145 479 541 471 759 215 945 323 976 764 771 393 422 544 025 312 134 975 759 771 856 933 851 656 283 739 826 433 488 584. 370 744 922 437 871 808 656 831 308 742 5 263 864. 374 952 106 z02 470 563 969 979 846 289 056 852 293 693 568 086 894 828 530 370 922 066 724 007 808 132 985 146 923 507 541 597 677 666 948 833 511 468 176 695 598 542 774 173 340 433 942 283 743 989 181 366 420 667 993 428 841 471 492 332 384 265 282 744 575 907 393 679 369 370 843 365 990 786° 045 769 443 366 536 o64 160 327 033 165 728 446 892 149 257 686 343 261 240 538 681 B29 931 723 109 849 196 255 309 224 119 O51 798 O15 082 265 a2 366 481 131 457 I15 231 446 796 879 219 115 219 446 731 273 678 753 ‘977 874° ‘621 018° 806 710 170 820° 860 550 244, 607 ° 781 030 818° 429 156 814 464 909 185 394° a11 948° ORE: 186° 956° Deo) 18 fe) 928° 438° 218 ney) 983 913° 800° 716 769° 240° O05 ° Bie 105 223 986 434° 035 423° 887 625° 829° 237° 508° OODTAGCE SVE OSHIACE ONE OSHDIAMK HONE OSOAAMKAWNYH OSHAWA KE wWNOHOO HTS co DOS Ounandanands pe ERE SE BRO wWHOWWwWwwWWWN NYDN NNN NNN HE ERE RHE ROOD DOSCOCOO Ad Mr. J. Lunt. On the Origin of certain Unknown December 14, 1899. Dr. G. J. STONEY, Vice-President, in the Chair. A List of the Presents received was laid on the table, and thanks ordered for them. The Right Hon. Lord Justice Romer, a member of Her Majesty’s Most Honourable Privy Council, was balloted for and elected a Fellow of the Society. The following Papers were read :—. I. “The Piscian Stars.” By Sir NormMAN LOCKYER, F.R.S. _ II. “On the Origin of certain Unknown Lines in the Spectra of Stars of the 6 Crucis Type, and on the Spectrum of Silicon.” By JOSEPH LuNT. Communicated by Dr. GILL, F.R.S. Il. ‘A Note on the Electrical Resistivity of Electrolytic Nickel.” By Professor J. A. FLEMING, F.R.S. IV. “ Investigations on Platinum Thermometry at Kew Observatory.” By Dr. C. CHREE, F.R.S. 7 V. “Observations on the Morphology of the Blastomycetes found in Carcinomata.” By Dr. K. W. Monsarrat. Communicated by Professor SHERRINGTON, F.R.S. The Society adjourned over the Christmas Recess to Thursday, January 18, 1900. “On the Origin of certain Unknown Lines in the Spectra of Stars of the 8 Crucis Type and onthe Spectrum of Silicon.” By JOSEPH Lunt, B.Sc. F.LC., Assistant, Royal Observatory Cape of Good Hope. Communicated by Davin Gi, C.B., F.R.S., Her Majesty’s Astronomer at the Cape. Received November 27—Read, December 14, 1899. In a recent paper “‘On the presence of Oxygen in the Atmospheres of certain Fixed Stars,”* Dr. Gill calls attention to three unknown lines * “Roy. Soe. Proc.,’ vol. 65, p. 205. Lines in the Spectra of Stars of the B Crucis Type. 45 in the spectra of 6 Crucis, « Canis Majoris, and stars of their type, viz., wave-lengths 4552°79, 4567-09, 4574°68. Mr. McClean had previously also recorded these lines in his measures of the spectrum of 6 Crucis* as wave lengths 4552-6, 4567°5, 4574°5, but beyond pointing out the approximate coincidence of the first of these with lines due to barium or titanium, he assigns no origin to them. Sir Norman Lockyer frequently records them as unknown lines. In his recent paper ‘‘ On the appearance of the Cleveite and other New Gas Lines in the Hottest Stars” (June, 1897),7 he records all three lines as unknown. ‘The first occurs in a map of the spectrum of Bellatrix as a line in a probable new series found by Dr. W. J.S. Lockyer. The second and third lines, given as 4566°8 and 4574:8, occur in a Table of Lines which Sir Norman regards as belonging, with high probability, to gaseous substances which have yet to be discovered. It will be noticed that the connection existing between the three lines is not there recognised. In none of the Tables of Wave-Lengths available for reference at the Cape could any satisfactory clue be obtained as to the origin of the lines. During some experiments made with a view to securing the best elementary line spectrum of oxygen as a comparison spectrum for stars of the 6 Crucis type, I found that a tube of carbon dioxide gave the best results, being freer from impurities, and giving stronger oxygen lines than any of the oxygen tubes at my disposal. By the use of a jar and air gap in the secondary circuit of the coil the gas was dis- sociated, and gave the spectra of carbon and oxygen. During use, the carbon Gaede tubes became more vacuous, and, with a view to obtaining a brighter discharge and shorter exposure, I passed the induced current from an 18-inch Apps’ coil, using four large jars and an air gap. | Whilst using these electrical conditions, | happened to expose an argon tube marked 2 mm. (pressure), and on developing the photo- graph was much surprised to find that it, too, gave the well-recognised lines of oxygen. Stronger than these I at once noticed two lines at the green end of the spectrum, which recalled the lines in 8 Crucis, which were unknown terrestrially, whilst the expected argon spectrum was almost entirely absent. On comparing the negative, film to film, with one of 8 Crucis, and allowing for the difference of temperature conditions under which the two negatives were taken, the identity of the three unknown lines in & Crucis with three lines on the argon negative was at once apparent, * ‘Spectra of Southern Stars’ (Stanford, 1898), p. 13. + ‘Roy. Soc. Proc.,’ vol. 62, p. 60. 46 Mr. J. Lunt. On the Origin of certain Unknown and a subsequent photograph of the spectrum of « Canis Majoris, in which the argon tube was used, as stated, as a comparison spectrum, established their absolute identity both as regards position and relative intensity. It was, therefore, evident that a terrestrial source of the three unknown lines had been discovered, and with the behaviour of the carbon dioxide tube fresh in mind, and the replacement of the argon spectrum by unknown lines and those of oxygen by use of a highly disruptive spark, it is not surprising that an obvious startling explana- tion as to the nature of the element thus found terrestrially should have suggested itself. It was at first assumed, erroneously as it afterwards proved, that the origin of the unknown lines lay in the gaseous contents of the argon tube. Four argon tubes in succession gave precisely the same results, viz., the argon spectrum with an ordimary discharge and the unknown lines and oxygen, together with the disappearance of the argon spectrum, as a result of using the jars and air gap. On communi- cating these results to Dr. Gill, he at once interested himself in the matter, and gave every facility for a further prosecution of the inquiry. He remembered that Professor Ramsay had furnished him with a specimen tube of pure argon, and this tube had not been examined. On trying this tube under the same conditions as the others, it was found to give the argon spectrum under all conditions. Neither the unknown lines nor oxygen made their appearance, even when the most intense disruptive spark available was employed. The first four tubes had aluminium electrodes, whilst Professor Ramsay’s tube had platinum electrodes, and was more vacuous and much shorter. A pair of aluminium electrodes was then taken from a vacuum tube, and a spark between the metal terminals in air was next examined, with the result that the unknown lines were not found. A line appeared very approximately in the same position as the strongest of the three lines, but this was only one of the numerous air lines, and was due to nitrogen (4552°6 Neovius). Therefore, the electrodes of the argon tubes did not account for the unknown lines. On a further examination of the negatives, the H and K lines of calcium were recognised in the spectra of the argon tubes subjected to the highly disruptive spark, pointing to the fact that the lime of the glass was being volatilised. This fact alone might account for the presence of the oxygen lines in the spectra, and the materials of the glass were then suspected as being the origin of the lines under consideration. _ A tube of pure helium, kindly furnished to Dr. Gill by Professor Ramsay, was next examined, and, with much surprise, this was found also to behave exactly as the first argon tubes had done. Lines in the Spectra of Stars of the B Crucis Type. 47 With an ordinary discharge it gave the pure helium spectrum, but with the highly disruptive discharge the helium spectrum vanished entirely, and was replaced by the unknown lines and the spectrum of oxygen. The helium spectrum could be obtained at will by reverting to the ordinary discharge. This helium tube had platinum electrodes, and these last observations finally banished any idea that the gaseous contents of the tubes or the metallic electrodes could be the origin of the substance searched for, and the conclusion that the glass of the tubes contained the substance sought was now irresistible. Yet in some of the spectra from the helium tube, the H and K lines of calcium were absent when those of oxygen were present, showing that the lime of the glass did not neces- sarily account for the presence of oxygen. After various fruitless experiments, sparks were taken between the platinum terminals of a broken up vacuum tube on which still adhered some of the blue fusible glass, commonly used in sealing in platinum wire in glass. The spectrum of this spark in air showed the unknown lines. { Beads of glass made from ordinary soda glass tubing, were then fused on platinum wires, and the spark from these was examined. The unknown lines again appeared. The substance sought was now strongly suspected to be the element silicon. The siliceous diatomace- ous earth “ kieselguhr ” was next used as the most convenient source of silica, and beads of sodium silicate were made by fusing this material with sodium carbonate on platinum wire. The result of the examina- tion of the spark was that the unknown lines were again found. The next step was to replace the kieselguhr by pure rock crystal obtained from the South African Museum by Dr. Gill. Sodium silicate made from the pure rock crystal, also furnished the unknown lines, whilst the sodium carbonate alone failed to give them. These experiments left little room for doubt that the element sought was silicon. Nevertheless, it was very desirable to confirm the result in another way, by examining the spectrum of a gaseous siliceous com- pound. Platinum wires were sealed into the ends of a piece of wide glass tubing, 2 inch internal diameter, the ends of the wires leaving a gap of only 2 inch for the passage of the spark. The tube was also fur- nished with an inlet and outlet tube for the gas. No capillary tube was used in order to avoid the hot spark coming into direct contact with glass. The tube was then filled with silicon tetrafluoride, and after the gas had been passing for some time, it was sealed off at atmospheric pressure. An ordinary discharge passed through the gas without jars or air gap gave a banded spectrum of the compound itself. The disruptive discharge obtained by using four jars and an air gap, A8 Mr. J. Lunt. On the Origin of certain Unknown at once gave the unknown lines, which were thus proved to be un- doubtedly due to silicon. This silicon spectrum was not accompanied by that of oxygen, thus proving that it could not be due to any dissociation of the silica of the glass, and that in this case, the gaseous contents of the tube and not the tube itself, furnished the lines under consideration. Sir Norman Lockyer’s papers were then consulted for any reference. to the presence of silicon in stars, and it is necessary to refer in some detail to his observations. It is evident that he has used similar powerful disruptive discharges with vacuum tubes, and obtained partial decomposition of the glass, for he says :* ‘The use of the spark with large jars In vacuum tubes results in the partial fusion of the glass, and the appearance of lines which have been traced to silicium.” Unfortunately he does not give the wave-lengths of the lines thus traced to silicon, and from his statement alone, one would surmise that the origin of the three lines was recognised by Lockyer. There is evidence, however, in the sume paper that he cannot have traced the lines in question to silicon notwithstanding the above state- ment, because, as previously pointed out, Sir Norman regards two of the lines as belonging to gases yet undiscovered, and includes them in a Table of Wave-Lengths of lines due to unknown gases. The other line he also includes as an unknown line in Bellatrix, and Dr. W. J. 8. Lockyer places this as a member of a probable rhythmic series due to an unknown substance. It is a curious fact that Hartley and Adeney, and Eder and Valenta, who alone give us any extended list of lines due to silicon, appear not to have examined the spectrum of this element in the region of the three lines here considered. Their’ published wavelengths show only lines in the extreme ultra-violet, and the majority of them are quite outside the region which can be examined by the McClean Star Spectroscope. Watts’s ‘ Index of Spectra’ (Appendix E, p. 21) records a line at 4566 (Salet), but no lines appear corresponding to 4552-79 and 4574°68. Sir Norman Lockyer? regards two lines at 4128°6 and 4131°4 as the most conspicuous enhanced lines of silicon, indeed these two lines are the only silicon lines he labels Si in his published photographs. Eder and Valenta give 4131°5 and 4126°5 as the least refrangible on their list, and although there is a rather excessive discrepancy in the wave-lengths of one of the lines, they are probably the same pair of lines. They are shown in Lockyer’s photographs of the eet of a Cygni and Siriust and also of « Cygni and Rigel.§ * “Roy. Soc. Proc.,’ vol. 62, 1897, p. 65. + ‘Roy. Soc. Proc., vol. 61, 1897, p. 443. t ‘ Roy. Soc. Proe.,’ vol: 65, p. 191. § ‘Phil. Trans.,’ A (1893), plate 2. Lines in the Spectra of Stars of the 8 Crucis Type. 49 It is a remarkable fact that these three stars, which may be con- sidered as amongst the best examples of silicon stars in the light of the spectrum of silicon hitherto known, do not show the three silicon lines which are so prominent in 6 Crucis, « Canis Majoris, &c. Scheiner has measured the spectra of all three stars* in this region, but does not record the lines in his Table of Wave-Lengths. Their absence from the spectra of these stars (as well as the presence of Lockyer’s enhanced silicon lines) is fully confirmed by photographs taken here with the special object of searching for the new silicon lines in the best known silicon stars. This can be readily understood in the light of the experiments with the tube of silicon tetrafluoride. With the highest disruptive spark, Lockyer’s silicon lines 4128-6 and 4131-4 are much enhanced as compared with the lines 4552-79, 4567-09, 4574°68, and it was found possible by suitable exposure to obtain the two enhanced lines without the presence of the other three lines becoming evident. The latter lines would be much more rapidly obliterated in the absorption spectra of stars, than in the bright line spectrum from the tube, and therefore their absence from certain stars in which the enhanced lines are strong need not occasion much surprise. In other stars, however, all five lines are present. Lockyer has recorded them in Bellatrix and their presence has been confirmed by photographs of the spectrum of this star taken here. Mr. McClean has measured all five lines in 8 Crucis where Lockyer’s enhanced silicon lines are certainly not so conspicuous as the lines 4552-79 and 4567-09. The same may be said of e Canis Majoris in which star the new silicon lines are very prominent, whilst the enhanced lines are very faint. In the silicon spectrum from the argon and helium vacuum tubes, the enhanced lines noted by Lockyer are by no means so prominent as they are in the silicon spectrum, obtained from silicon tetrafiuoride with the intense disruptive spark. It is evident, therefore, that great variations in the relative intensities of the silicon lines occur in stellar spectra, and that such variations can be produced to a certain extent in the laboratory, and these require further investigation. The behaviour of the silicon lines will give us valuable data for the elucidation of the problem of relative stellar temperatures. It is clear that if we regard, with Lockyer, the lines 4128°6 and 4131-4 to be the enhanced lines of silicon and their presence, enhanced, to be a criterion of a higher temperature than occurs in stars where these lines are not enhanced, it must follow that such stars as a Cygni, Rigel, and Sirius are hotter than Bellatrix, @ Crucis, and « Canis * Scheiner’s ‘ Astronomical Spectroscopy.’ 50 Prof. J. A. Fleming. Majoris. Whereas Lockyer* in his most recent paper ‘On the Chemical Classification of the Stars ” (April, 1899), regards the so-called “Crucian” stars, as at a higher temperature than the “ Rigelian ” and ‘‘Cygnian,” and indeed he regards Bellatrix “as a type of the hottest stars, exception being made of ¢ Puppis.” Of the other lines recorded by Eder and Valentat as due to silicon, 3905°4, 3862°5 and 3855-7 are present both in the spectra of the dissociated glass and in the high temperature spectrum of silicon obtained from the silicon tetrafluoride tube. They are enhanced lines in the latter case, occurring together with Lockyer’s enhanced lines in the absence of the three new silicon lines, but they lie outside the region measured by Scheiner in « Cygni, Sirius, and Rigel. In the Harvard “Spectra of Bright Stars”{ the two latter lines are however, specially noted in Rigel as 3863-2 and 38562 as “ con- spicuously strong in the ultra-violet,” whilst all three are recorded (3905°6, 3863°2, 3856-2) in stars of Groops VI to VIII (Harvard), comprising « Cygni, Sirius, and Rigel. They would thus appear in these stars to accompany the enhanced silicon lines, specially noted by Lockyer, viz. 4128-6 and 4131-4. The lines 3834°4 and 3836°7 recorded by Eder and Valenta are not present in any of the photographs of silicon spectra, and may possibly be due to impurities. The lines 3795:9 and 3791-1 recorded by Eder and Valenta are present in all the silicon photographs, but do not become enhanced at high temperatures. There is, however, a third line, approximately dX 3807, not recorded by them, but which appears in all the photo- graphs of silicon spectra. It is stronger than 3795-9 and 3791-1, and does not become enhanced with high temperature. All three lines accompany the three new silicon lines in « Canis Majoris. “A Note on the Electrical Resistivity of Electrolytic Nickel.” By J. A. FLemine, M.A., D.Sc., F.R.S., Professor of Electrical Engineering, University College, London. Received Novem- ber 21,—-Read December 14, 1899. The numerical values assigned by various experimentalists for the mass or volume electrical resistivity of certain metals differ very con- siderably. Some metals are without much difficulty prepared as often * © Roy. Soc. Proc.,’ vol. 65, No. 416, p. 189. t Waitts’s ‘ Index of Spectra.’ { ‘ Harvard Annals,’ vol. 28, Part I, Table 7, p. 23. On the Electrical Resistivity of Electrolytic Nickel. OL as required in a state of such chemical purity, and brought so easily into similar physical conditions as to annealing and density, that determinations made by different observers of their resistivity or specific electrical resistance are nearly identical. Matthiessen’s long-accepted value* for the mass resistivity of copper in the form of hard-drawn wire, viz., 0°14493 standard ohm per metre-gramme, was substantially confirmed by the more recent work of Mr. T. C. Fitzpatrick.— Even the purest electrolytic copper now obtainable in an annealed condition does not show an electric con- ductivity more than 3 per cent. greater than that of a similar character prepared thirty-five years ago by Matthiessen. In a research carried out in the years 1892 and 1893 by the author in conjunction. with Professor Dewar, careful redeterminations were made of the volume-resistivity at 0° C. of all ordinary metals, taken for the most part in an annealed condition and in a state of great chemical purity. The values so obtained for the electrical volume-resistivity of silver, copper, gold, aluminium, zinc, platinum, tin, lead, magnesium, and iron agreed fairly well with the values given by Matthiessen, and quoted in most of the treatises on electricity. The resistivity of cad- mium, as given by us was, however, considerably larger than that usually stated, although our sample of cadmium was very carefully prepared. In the case of nickel, the purest nickel we were able to obtain was that prepared from nickel-carbonyl. Dr. Ludwig Mond, F.R.S., was so kind as to furnish us with two tubes of this nickel. It was found, however, to be too brittle to draw into wire, and the operation of melting it would have most certainly introduced impurities. Accord- ingly, the nickel tube was cut up in the lathe into a spiral, and a resistance coil formed with it which could be used for taking the resistivity ratios at different temperatures, but which was not suffi- ciently uniform in dimensions to permit its volume-resistivity to be calculated. Hence, in our published results, we took the volume-re- sistivity of this nickel at 0° C. to be 12,320 C.G.S. units, which is the value given by Everett, said to be derived from Matthiessen’s experi- ments, and simply calculated the resistivity at other temperatures from the experiments given by our own observations with the Mond nickel. About a year ago, however, Mr. J. W. Swan, F.R.S., sent me a sample of nickel wire which he had prepared electrolytically from a hot solution of very carefully purified nickelous chloride. The electro- lytic metal was annealed by heating in an atmosphere of hydrogen, aiter having been drawn into wire through a die. The nickel wire so prepared and annealed is as soft as a silver wire. * See ‘B.A. Report,’ 1864, or ‘ Phil. Mag.,’ 1865. + See ‘B.A. Report,’ Leeds, 1890, or ‘ Electrician,’ vol. 25, p. 608, 1890. 52 Prof. J. A. Fleming. Tt had a fairly uniform diameter of about 0°01 inch, and a length of nearly 250 cm. The mean diameter of this wire was taken with the micrometer microscope at regular intervals of centimetres with the following results :— Diametral Measurements of Nickel Wire, as read by Microscope at _regular Intervals of about 5 cm. | Diameter Diameter | wes. (inches). ee (inches). | | on pie Aes) | | 1 0 -0097 26 00100 2 0 -0097 27 | ~~ 00096 | | 3 0 00975 23 00098 A 0 0107 joe Oa 0 -0093 0 -0100 30 | ~~ O-p090KN Bela 0 -0098 31 | 00102 7 0 -0100 Ee | 00098 8 0 -01015 33. | 000975 9 | 9 001015 | 34 1 oops ene 10 00100. «| ~~ 35 0:0098 11 0 -0098 36 0 00985 12 00100 37 0 -00965 13 001015 38 | | 00098 14. 0 -0099 39 0 -00985 15 0-0101 40 | @eonan 16 0 0099 41 | @-0096 17 0 -0099 42 | 00094 is |”, "0-01605 43 00098 19 0 -0097 AA 0 0098 20 0 -0098 A5 0:00995 21 00098 46 0 -0098 22 0 -00985 Zc 0 -0098 23 0 -0101 Agy on 0 -0096 24. 0 0100 49 00099 25 001005 50 00101 | The mean diameter of the wire, as obtained from the above fiity readings, was 0:00985 inch. This was checked by taking the density of the wire with all the usual precautions for obtaining a correct value. The length, weight in air, and weight in water were determined to be as follows :— ene thvot the waver ge oc. cena ee 246°98 cm. CLOG IAAL yg See a Al) cet aaa 1:1163 grammes. Weight in water at 18° C. +suspension... 100k was; Weleht OL Suspension Mace. peter 0°1000 :, Weight of wire at 18° C. in water ......... OOONi a eae. From the above observations the density was found to be 8°96 at On the Electrical Resistivity of Electrolytic Nickel. 53 18° C. The mean diameter calculated from the length and density ig then 0°00997 inch. Hence we have— Mean diameter from micrometer measurement ......... 000985 inch. Mean diameter by specific gravity and length measure- 2 eee LS GE SESE ee Ree ae ca ee ee 000997, 7% The mean of both means is 0°00991 inch = 0°02567 cm. =‘ This last number was taken as the value of the mean diameter. The nickel wire was then soldered to thick copper leading-in wires, and wound on a frame of a kind suitable for immersion in liquid air. A description of this particular kind of resistance coil, which has proved itself to be exceedingly suitable for low temperature work, was given in a paper describing the result of numerous observations on the resistivity of metals at low temperatures, published by Professor Dewar and the present author in 1893.* A coil having been thus constructed, its resistance was taken at various temperatures in a bath of paraffin oil, and the results are as shown in the table below. The temperature of the bath was measured Observations on the Resistance of Bie eo Nickel Wire. | | | Total | Platinum | Centigrade | Volume | Obs. | resistance of | temperature, temperature, | resistivity in | Md nickel wire. | pe. C. | C.G.S. units. - open Se Sy. | 1 3°4284 | 1-057 1 +232 6974 | I 3°7563 18 -489 18°29 | 7641 eh. . 30470, 28 676 28 °32 | 8029 | | deed 4°1109 | 36° 959 36°51 | 8363 | . ahs 4 4°3506 48 -740 48 °21 | 8850 | 6 4°5679 | 58 °773 58°23 9292 | 7 4°5778 | 59 °315 58°78 | 9312 | : Beh! 4°7493 | 67 “384 66°85 9661 | 9 | 50403 80 *550 80-11 | 10253 | 10 52018 88 -587 88 -25 | 10582 | eee | 5°3000 93 544 93 -29 | 10782 | ae 5 *3882 95 -731 94.88 10961 fete. | 5 +2379 8Y -289 88-97 | 10655 ie a 50094 78 482 78 “02 | 10190 Pinas | 4°7273 66-700 66-17 | 9616 eee. | 4 -5260 57 °151 56-61 | 9207 ore | 4. +3586 ' 48-704 48-18 | 8865 | eo an 41323) | 37 *906 37-44 84.06 ee | 3 9620 31 -226 30°83 | 8060 20 | 3°8112 21-550 21-30 / 7753 | ae | 3°4318 2°205 2°35 | 6981 | 22 2-090 —80-72 | 78-2 4251 | 23 0-710 — —196°89 —182°5 | 1444 | * See ‘ Phil. Mag.,’ September 1893, p. 279, “On the Resistance of Metals and Alloys at Temperatures approaching the Absolute Zero.” 54 Prof. J. A. Fleming. at the same time by means of a platinum thermometer (Pi) used in previous researches. The conversion of the platinum temperatures into contig tem- peratures was effected by means of a table drawn up Mr. J. Hamilton Dickson* for this thermometer. The measurements of the resistance at low temperatures (—78°2° and — 1825") was obtained by measuring the coil in liquid air and melting CO, at the Royal Institution Laboratories, by kind permission of Professor Dewar. The total resistance of the nickel calculated from the above observa- tions is set out in the form of a curve (fig. 1), having resistance as ordinates and temperature as abscisse. The curves are given both for temperature in centigrade degrees and temperature in platinum degrees. The curve shows that the resist- ance is falling steadily downwards to a zero value as the absolute zero of temperature is approached. In fig. 2 the portion of the curve between 0° C. and 100° C. is shown in an enlarged scale. In fig. 3 the volume-resistivity is set out in terms of temperature. The results show that the volume-resistivity or resistance per centi- metre-cube of this electrolytic nickel at 0° C. is 6935 C.G.S. units. The average temperature coefficient between 0° C. and 100° C. is 0-00618. The above observations indicate that this electrolytic nickel, as pre- pared by Mr. Swan, has an exceedingly different and much lower resistivity than that employed for test by Matthiessen thirty-five years ago. The value of the mean temperature-coefficient of the nickel used in the experiments of Fleming and Dewar in 1893, and prepared by Dr. Ludwig Mond, was 0:00622+ between 0° C. and 100°C. It is clear therefore that some extraordinary electrical difference exists between nickel as it can now be produced electrolytically and nickel - as it was produced by Matthiessen for his experiments. It would be interesting to ascertain if any specimen of nickel known to have been used by Matthiessen for his experiments still exists, and if so, to discover the nature of the impurity (if impurity was present), or at least the physical difference, which caused his nickel to have nearly double the electrical HU of that which can now be pro- duced. I desire to record my thanks to Mr. J. E. Petavel and to Mr. A. Blok for assistance in these experiments. * “Phil. Mag.,’ June, 1898. t ‘Phil. Mag.,’ September, 1893. (ohms) Resistance i) GD On the Electrical Resistivity of Electrolytic Nickel. 55 Fie. 1.—Temperature-resistance Curve for Electrolytic Nickel Wire. * [| | ble | : Ht ttt LY 45 Z TC PCEELELEELEE EL HHAVORIGDCAVAOTE | ci G a S fe) “260 240 220 200 180 160 140 120 100 80 Ca 0 +20 40 60 80 Toriperaticet Fo} Prof. J. A. Fleming. {oz a asyesadiuay vO! 00 96. 26. 88, ~8 09. 94 FL 9 09, 99." 29 oY Ov 9 Zo f2 p2 02 W. Zl 9. w+ 0, v= 8 e (}. a eee (Oe eats | = ieee Ls : ry | ad : - = tees = 0.98 rr ie 2) Oo o 10 20 50 40 lemperature Centigrade. Note added December 6, 1899.—Since writing the above short paper, I have discovered in a paper by Messrs. A. Matthiessen and C. Vogt,* a reference to the sample of nickel with which the present accepted figure for its resistivity was evidently obtained. This paper is entitled “On the influence of Temperature on the Electric Conducting Power of Iron and Thallium,” and its title would not lead a reader to look in it for a reference to the resistivity of nickel. Messrs. Matthiessen and Vogt therein state that samples of sup- posed chemically pure nickel and cobalt wires prepared by M. Deville. were given to them by M. Wohler. They measured the resistivity of these samples, but they state that their electrical behaviour gave them reason to believe that this nickel and cobalt were not pure. They give the electrical conductivity of the nickel as 13-11 at 0° C. taking hard drawn silver at 0° C. as 100. Hence if hard drawn silver has a volume resistivity of 1620 C.G.S. units at 0° C., it follows that Matthiessen and Vogt’s value for the resistivity of their sample of nickel would be 12,357 C.G.S. units at * “Phil, Trans. Roy. Soc.,’ 1863, p. 384. 58 Mr. K. W. Monsarrat. Observations on the 0° C., which is a number very close to that usually given in tables of electrical specific resistance.* Matthiessen and Vogt state in this paper, that they hope to be able to prepare pure nickel electrolytically, and obtain a value for its electrical resistivity. [I have been unable to discover, however, that they ever carried out their intention. At any rate, the number which they give for the electrical volume resistivity of this nickel of the purity of which they evidently had suspicions, has been accepted for the last thirty-six years as the true value. “Observations on the Morphology of the Blastomycetes found in Carcinomata.” By Kerth W. Monsakrat, M.B., F.R.C.S.E. Communicated by Professor SHERRINGTON, F.R.S. Received November 22,—Read December 14, 1899. (From the Pathological Laboratory of University College, Liverpool.) (Abstract. ) This research was undertaken in order to confirm if possible the observations of Sanfelice, Roncali, and others, on the presence of organisms of the order Blastomycetes in carcinomata, and to study the morphology of the same. The observations have been arranged under four headings :— | 1. Isolation by culture. 2. Staining reactions. 3. Histology. 4. Tissue reactions following inoculation. 1. Isolation by Culture.-—The tumours examined were carcinomata of the breast and uterus; incisions were made with a sterilised knife and scrapings from the edges of these inseminated on to media. Many kinds of media were tried, but a result was obtained only on glucose agar. Wort agar and wort bouillon were subsequently used for sub- cultures; on both the organism grows readily aérobically at 37° C. Sub-cultures on neutral gelatine appear as pale yellow slow-growing colonies without liquefaction of the medium. On neutral agar the colonies have a more marked yellow tinge; they do not appear until * The numerical values of the specific resistance of nickel given in various tables by different authors are not quite identical, and yet all so far found are stated to be derived from Matthiessen’s experiments. Thus, Everett (‘C.G.S. System of Units,’ 1891 ed.) gives 12,320 C.G.S. units at 0° C. as the value. Landolt and Bornstein give one value equivalent to 12,757 at 0° C. from the ratio of conduc- tivity of nickel to that of mercury, and another equivalent to 12,014 at 0° C., derived from the ratio of the conductivities of hard drawn silver to that of nickel. Morphology of the Blastomycetes found in Careinomata. 59 aiter four to five days’ incubation at 37°C. There is no growth on acidified gelatine, agar, or bouillon. Neutral bouillon yields a scanty growth after four days’ incubation at 37° C.; no scum is formed. On wort bouillon and wort agar the growth is plentiful after twenty-four hours’ incubation at 37° C. On potato at the same temperature there is a characteristic dark brown growth after forty-eight hours. The organism grows more readily anaérobically than aérobically ; the growth on potato under the former conditions is white, but turns brown when air is admitted, whereas the growth on agar is brownish-yellow, contrasting with the pure yellow colour of the aérobic growth. These appearances on culture media agree in the main with those described by Sanfelice and Plimmer in the case of the organisms which they have severally isolated. 2. The staining reactions of the organism in the tissues were specially studied in order to establish if possible the real characters of the “ cancer-bodies” described by many observers. After trial of several methods, the following was decided upon as giving character- istic and. distinguishing results. Carmine is first used as a nuclear stain, either in the form of lithium carmine, alcoholic borax carmine (when the pieces of tissue are stained in the mass), or acetic car- mine. The last is the only preparation which gives good results with tissue fixed in Flemming’s solution. The sections thus stained are placed in a | per cent. watery solution of methyl-violet for two minutes, then in a 0°25 per cent. solution of picric acid, washed, dried with filter-paper, and decolorised in clove oil. The methyl-violet is extracted from the plasma and nuclei, but remains in the organisms. The method was proved to give distinctive results in sections of the growths produced by experimental inoculation of the organism isolated. 3. The morphological characters of the organism are as follows :— Fresh specimens from cultures are spherical, from 4 to 10 microns in diameter, and in most cases take an aniline chromatin stain diffusely. There is, however, a great variety in the distribution of the chromatin, it is sometimes aggregated to one pole, sometimes divided up at different parts of the cell, and in other cases it is only represented by a few isolated granules. The capsule is delicate. Multiplication in cultures takes place by budding. In the primary growths produced by intraperitoneal inoculation of the organism, the latter is also in most cases spherical, possesses a delicate capsule, and multiplies by budding. Two peculiarities are, however, to be seen: firstly, in many cases, delicate processes connect adjacent specimens of the organisms; and, secondly, the capsule is often thickened and forms a kind of “ halo” round the central deeply staining body of the cell. : VOL. LXVI. G 60 Morphology of the Blastomycetes found in Caremmomata. In the nodules in lungs, liver, spleen, and kidneys, which are secondary to the growths on the peritoneum, in addition to the forms already described, spore-bearing forms are found. In these the capsule is much thickened, the chromatin of the cell breaks up irregularly, and portions are allowed to escape through dehiscences in the capsule. There is no regularity in the process, no simultaneous division of the cell-contents into a definite number of spores, and no simultaneous shedding of the same. The spores are without capsule when they escape, and are irregular in contour. ‘They stain deeply with chromatin stains, and are finely granular. This method of spore- formation is specially to be noted, as it is entirely unlike the methods described in the case of members of the Saccharomyces class. 4, Tissue Reactions—The animals used for inoculation were guinea- pigs, and the inoculations were made not so much for the purpose of estimating the capacity of the organism for producing cancer as for that of studying the morphological characters of the latter when in the tissues. The results following intraperitoneal injection of 1 ¢.c. of a culture forty-eight hours’ old are as follows:—The animals showed no symptoms of illness; they were killed at periods from two to six weeks after the injection. On opening the abdomen, the omentum and general peritoneal surface were found to be studded with nodules from the size of a pea to that of a pin’s head; of the other organs, nodules are visible to the naked eye in the lungs, liver, spleen, and kidneys. The primary growths on the peritoneum are composed of proliferated endothelial cells; the organisms are present in consider- able numbers, some within the cells, but most outside them. In many of the nodules there is some attempt at the formation of a connective tissue capsule, and in others the central parts are broken down. In the lungs, the nodules are made up of endothelial cells ; in each nodule there are organisms present and usually centrally situated. In the liver and spleen, the nodules are very similar in appearance to the primary omental growths; the origin of the cells composing them is doubtful. In the kidney, where again the nodules are of endothelial origin, the cells are derived from those lining the Malpighian corpuscles and tubules. In no case was there any alveolar arrangement of the cells or any appearance resembling the endotheliomata of man. Upon the Development of the Enamel in certaan Osseous Fish. 61 January 18, 1900. The LORD LISTER, F.R.C.S., D.C.L., President, in the Chair. The Right Hon. Lord Justice Romer was admitted into the Society. A List of the Presents received was laid on the table, and thanks ordered for them. | The following Papers were read :— I. Upon the Development of the Enamel in certain Osseous Fish.” By C. 8S. Tomes, F.R.S. IJ. “Further Observations on ‘ Nitragin’ and on the Nature and Functions of the Nodules of Leguminous Plants.” By Miss M. Dawson. Communicated by Professor MARSHALL WARD, oR ES. IIf. “ On the Innervation of Antagonistic Muscles. Sixth Note.” By Professor SHERRINGTON, F.R.S. IV. “On the Viscosity of Argon as affected by Temperature.” By Lorp RAYLEIGH, F.R.S. V. “On the Behaviour of the Becquerel and Réntgen Rays in a Mag- netic Field.” By the Hon. Rh. J. Strrutr. Communicated by Lorp RaAYLeEien, F.R.S. VI. “On an Experimental Investigation of the Thermo-dynamical Properties of Superheated Steam.” By J. H. GrinpLeEy. Communicated by Professor OSBORNE REYNOLDS, F.R.S. “ Upon the Development of the Enamel in certain Osseous Fish.” By Cuarues 8. Tomes, M.A., F\R.S. Received December 4, 1899—Read January 18, 1900. (Abstract.) The author has shown in previous communications to the Royal Society (which are to be found in its ‘Transactions’) that notwith- standing the fact that in all the vertebrata enamels present tolerably close resemblances in chemical, physical, and histological characters, differences far more considerable than might have been expected exist G 2 62 Upon the Development of the Enamel in certain Osseous Fish. in the formative processes. The present communication seeks to establish an additional method of enamel formation, essentially differing from any which has hitherto been described, and whilst the investiga- tion was undertaken in the hope of bridging over the gaps which exist between the methods previously known, it has only partly succeeded in doing so, as the process to be described stands somewhat alone. The principal point which is set forth in the paper is, that the forma- tive cells of enamel, known as ameloblasts, in all the Gadide and in Sargus and Labrus, undergo a preliminary transformation into a reticulated stroma, which is of the full dimensions of the ultimate enamel. During the calcification of enamel, the ameloblasts no longer - exist as such, nor do any other cells take their place, but the stroma itself seems able to segregate and properly apply the lime salts re- quired, which make their appearance at that side of the thick stroma which is most distant from the blood vessels. There have thus been demonstrated four varieties of the process by which enamel is formed, which, although there are gaps not at present bridged over, may perhaps be taken as representing certain stages in the evolution of enamel. These are— 1. Enamels not wholly epiblastic in origin, in which the stroma which is the seat of enamel calcification is furnished by a transforma- tion of the exterior of the mesoblastic dentine papilla, the ameloblasts apparently segregating the lime salts required for its hardening. This is found in the Elasmobranch fishes. 2. Enamels wholly epiblastic in origin, in which the ameloblasts undergo a prior transformation into a stroma of the dimensions of the finished enamel, and themselves disappear. This is the subject of the present communication, and is met with in the Gadide, in Sargus and in Labrus, and probably in many other fish. 3. Enamels wholly epiblastic in origin, in which the ameloblasts retain their integrity throughout the whole process. Their extremi- ties are, however, produced into long fibrillar processes, which are traceable far through the calcifying enamel, and these processes are prolongations of the plasm of the cells. This method is found in the enamel of Marsupials, and probably im all similar tubular enamels, such as are found in Hyrax and sporadically among other mammals. 4, Enamels wholly epiblastic in origin, in which the ameloblasts persist throughout the process of calcification. Their free ends are produced into short processes (Tomes’s processes), which penetrate but a short distance into the calcifymg enamel. This is the ordinary method found in placental mammals. It will be seen that the last two methods differ in degree rather than in kind, but that the first two stand markedly apart. Apparently in the Rays there is some sign of the approaching aban- donment of the share taken by the dentine papilla, as there is a less On ‘ Nitragin’ and the Nature, &c., of Nodules of Plants. 63 degree of specialisation of its outer portion, but no sign of any trans- formation of the ameloblasts themselves into any intermediate form of tissue has been observed. ) The sudden and entire transformation of the ameloblasts in the Gadidz may perhaps be correlated with the very early and rapid for- mation of the enamel, which is formed while there is yet: but little dentine calcified. In mammals enamel formation is a very much slower and more gradual process, and the dentine is always much further advanced towards completion than is the enamel. A comparison of the various processes now known as occurring in fish, in implacental mammals, and in placental mammals may, in the author’s opinion, be taken as finally disproving the idea, which is still entertained by some, that enamel is to be regarded as a sort of secre- tion shed out from the ends of the ameloblasts; for, imperfect though our knowledge remains in some respects, yet some form of conversion, direct or indirect, of a pre-existent organic matrix is common to all, though as in placental mammals it may be exceedingly small in amount, and the erroneous idea alluded to has proceeded from the study of the process exclusively in placental mammals, in whom its true nature is most difficult to decipher. > “Further Observations on ‘Nitragin’ and on the Nature and Functions of the Nodules of Leguminous Plants.” By Marta Dawson, B.Sc. (Lond. and Wales), 1851 Exhibition Science Research Scholar. Communicated by Professor H. MARSHALL Warp, F.R.S. Received December 5, 1899,—Read January £6, 1 900. | (Abstract.) In December, 1898, a paper by the author on “ Nitragin, and the Nodules of Leguminous Plants,”* was read to the Royal Society. Since that time the work on this subject has been considerably extended, and a brief summary of the additional results is given below. Investigations have been made in the following directions :— A. Microscopic Observations. A comparative study of various points of interest in the anatomy of nodules borne by several genera of different tribes of the order, with special reference to the mode of growth of the nodule organisms within the tissues of the host. | | * ‘Phil. Trans.,’ B, vol. 192, pp. 1—28. 64 Miss M. Dawson. B. Experimental Work. a. Pure cultures of the organisms from Piswm, Desmodium, and “‘ Nitragin,” upon various media, liquid and solid, organic and inorganic, employing the ordinary methods of bacteriology. B. Direct observations under the microscope of the various stages of growth of colonies, and the formation of bacteroids from straight rods, as seen in hanging drops. y. Experiments upon the effect of temperatures above the normal upon the direct infection of pea roots. 5. Cultures of various genera, representing different tribes of the order, to test the power of organisms proper to one genus to induce tubercle formation upon individuals of other genera or tribes. e. Crop cultures in the laboratory greenhouse of peas in sterilised media, with and without inoculation with “ nitragin,” also with and without a supply of nitrogenous food. ¢. Crop cultures of peas in ordinary garden soils and in subsoils, in the open, with and without inoculation with “nitragin,” also with and without an additional supply of nitrates. A further study of the morphology of nodules from various genera of the Leguminosex, leads to the conclusion that no definite line of dis- tinction can be drawn between genera in which filaments occur in the nodules and those in which they have not yet been observed. Several examples were found of fragmentary portions of filaments in the cells of very young nodules, whilst in older specimens these filaments were quite absent (¢.g., Phaseolus, Desmodium, Acacia, and others), suggesting an intermediate stage in the adaptation of the parasite to the special conditions existing in any given host. During the course of this study, some peculiar anatomical characters have been observed in certain nodules, ¢.g., the presence of a definite erystal layer in some genera, of apple-green nucleus-like bodies in Desmodium and Robinia, and of organisms of an. unusually large size in Desmodium, Coronilla, Psoralea, and others. A prolonged study was made of the organisms from Desmodium gyrans in particular. Pure cultures were obtained, and from these observations in hanging drops upon bacteroid formation showed that the X and Y forms arise by distinct lateral branching of the strarght rods. After twelve to fourteen days’ culture, the individual long rods tend to break up into small rodlets, and the branched forms become dis- jointed in a similar manner. A general study of these organisms and parallel cultures of “ nitragin,” compared with pure cultures of organisms obtained direct from Pzswm tubercles, shows that they all alike grow readily on gelatine or agar media containing an extract of pea stems, asparagine, and sugar, but very slowly on broth gelatine. They do not peptonise milk, but upon potato a watery streak is formed On ‘ Nitragin’ and the Nature, &c., of Nodules in Plants. 65 in about five days ; in a liquid medium—pea extract—a thick, zoogloea- like film forms in twelve to fourteen days. ‘The presence or absence of spores in these films is now under investigation. ‘The organisms are aérobic, and may pass through a short motile stage, but the presence of cilia has not been demonstrated. On a medium consisting of silica jelly and a mixture of salts, including ammonium sulphate, abun- dant growths of the organisms from Piswm and Desmodium have been obtained ; also in hanging drops of silica jelly, colonies of the latter type have grown to 30u diameter in seven days at 17° C. Further experi- ments are now in progress in order to test whether these organisms are per se capable of (a) fixing free nitrogen, or ()) converting nitrogen in the form of ammonium salts into nitrites or nitrates ; also to deter- mine whether or not the presence of nitrates in the culture medium is directly injurious to the organisms. At a temperature ranging from 24—35° C. (average 30°) a considerable increase in the percentage of direct infections of pea roots was obtained, but at temperatures above 35°, the host plants themselves succumbed after fourteen days. In water cultures only very early stages of infection were observed. Experiments to determine the action of the organisms proper to one genus upon plants of another tribe or genus suggest that there is probably only one organism capable of producing nodules on legu- minous plants, but that in each particular host special physiological conditions exist, to which the organisms become so specially adapted as to make it difficult for successful reciprocal action to take place between plants not nearly allied, though exceptions do occur. In connection with infection experiments conducted under, as far as possible, sterile conditions, it was determined that fifteen minutes’ treat- ment of seeds with a 0:1 per cent. solution of mercuric chloride before sowing, is without injurious effect upon the seeds, but that a longer action of the solution poisons the embryo. Crop cultures in sterilised media give best results when nitrates without organisms are supplied to the plants. The addition of “nitragin” under these conditions is of very little benefit, and if a sufficient supply of nitrogenous food be available, a reduction in the resulting crops ensues when this “ fertiliser” is employed. In unsterilised media a small increase in crop may result from the use of “nitragin.” The conclusion derived from the various experiments, however, is that the presence or absence of “nitragin” is but one factor in a complex problem, and that at the same time must be taken into account the complicated physical and biological conditions of the soil and atmospheric environments, as well as the symbiotic action of the host plants, in the removal of the products of metabolism from the field of action of the nodule organisms. 66 Dr. C. 8. Sherrington. “On the Innervation of Antagonistic Muscles. Sixth Note.” By C. S. SHERRINGTON, M.A., M.D., F.R.S. Received December 30, 1899,—Read January 18, 1900. . Machine-like regularity and fatality of reaction, although charac- teristic of spinal reflexes, is yet not exemplified by them to such extent that similar stimuli will always elicit from the spinal animal similar responses. This want of certainty as to response is an interesting difficulty attending the study of spinal reactions. The variation in the responses of the skeletal musculature manifests itself not only in regard to the extent of the movement but also in regard to the direction of the movement. Some of the factors determining the character of the reactions are factors contained within the stimulus. Important among these is the ** focus of the stimulus.” Thus it has long been known that the direc- tion and other characters of the reflex movement are influenced by the mere location of the stimulus. Nevertheless stimuli identical in all respects, including locality, may evoke reflex movements of widely different, even of absolutely opposite, character. Such differences of response must be referred to differences obtaining at the time in the spinal organ itself. One cause for such differences seems indicated by the following observations :— The most usual, indeed the almost invariable, primary reflex move- ment of the hind limb of the spinal dog (and cat), when spinal transec- tion has been performed in the cervical or upper thoracic region, is flexion at hip, knee, and ankle ; the limb is “drawn up.” This move- ment can be well obtained by, among other stimuli, the pressing of the pads of the digits upward so as to extend the toe-joints, a stimulus that in some measure imitates the effect upon those joints of the bearing of the foot upon the ground under the animal’s weight. ixtension as a reflex result from this stimulus is, in my experience, never met with in the homonymous limb in the early time after transec- tion. When a certain period has elapsed, three weeks or more after transection, and shock has largely subsided, it becomes possible to, at times, obtain extension at hip as the primary movement in the homonymous limb. The pressing of the toe-pads upwards, spreading and extending the digits, elicits a sharp movement of extension at the hip, if at the time the initial posture of hip and knee be flexion. If the initial posture of hip and knee be extension, the primary reflex movement excited is, in my experience, invariably flexion. The reflex movement is, it is true, not unfrequently flexion, even when the initial posture is one of flexion; but it is, on the other hand, very frequently, and especially preponderantly in certain individual animals, extension. The passive assumption of a flexed posture at hip and On the Innervation of Antagonistic Muscles. 67 knee seems to favour the reflex movement at those joints taking the form of extension. The influence of the posture of the ankle-joint upon the reflex movement at the hip seems negligible, for I have often remarked the reaction at the hip to be unaltered, whether the ankle were flexed or extended, at the time of excitation. In some dogs, when the spinal transection has been made at the hinder end of the thoracic region, stimulation of the skin of the limb evokes the usual primary flexion at hip and knee wherever the lacus of the stimulus, except it be in the upper three-fourths of the front of the thigh. Applied in this latter region the stimulus, if the limb be mid- way between extension and flexion, not unfrequently evokes reflex extension at hip and knee; it does not evoke extension if the initial posture of the limb be extension ; but if the limb be, at the time of application of the stimulus, well flexed at hip and knee, reflex exten- sion, instead of refiex flexion, becomes the rule. In the spinal frog, as in the spinal dog, flexion at hip and knee is the regular reflex response of the musculature of the homonymous hind limb to skin stimuli applied at any part of the surface of that limb. This being true when the initial posture of the limb is, as when pendent, one of extension at hip, knee, and ankle, a difference becomes evident when the initial posture is one of flexion at those joints. In the latter case excitation of the skin within a small gluteal and pubic area, lateral and somewhat ventral to the cloacal orifice, causes extremely frequently not flexion at hip, but extension at that joint. Stimuli , (mechanical and chemical) to that area which evoke flexion at the hip- joint when the initial posture of the limb involves extension at that jomt, evoke, when the initial posture is flexion, reflex extension at the joint. These instances seem to indicate distinctly that the direction which a spinal reflex movement elicited by stimuli similar in all respects, including “locality,” may take, is in part determined by the posture already obtaining in the limb at the time of the application of the stimulus. ; The reaction described above for the spinal frog holds good after previous removal of all the skin from both hind limbs, with the exception of the small gluteal piece necessary for application of the skin stimulus. It would appear, therefore, that the influence of the posture of the limb upon the spinal condition and reaction is not trace- able to the nerves of the cutaneous sense-organs of the limbs. There still remain the afferent nerves subserving muscular sense, and connected with the sense-organs in muscles, tendons, and joints. These, as is well known, are largely affected by the various postures of the limb, even by such postures as are passively induced. 68 : . Lord Rayleigh. “On the Viscosity of Argon as affected by Temperature.” By Lorp Ray ricH, F.R.S. Received January 12,—Read Jan- uary 18, 1900. According to the kinetic theory, as developed by Maxwell, the viscosity of a gas is independent of its density, whatever may be the character of the encounters taking place between the molecules. In the typical case of a gas subject to a uniform shearing motion, we may suppose that of the three component velocities ¢ and w vanish, while is a linear function of y, independent of z and z. Ii p be the viscosity, the force transmitted tangentially across unit of area perpendicular to y is measured by pdu/dy. This represents the relative momentum, parallel to ~, which in unit of time crosses the area in one direction, the area being supposed to move with the velocity of the fluid at the place in question. We may suppose, for the sake of simplicity, and without real loss of generality, that « is zero at the plane. The momentum, which may now be reckoned absolutely, does not vanish, as in the case of a gas at rest throughout, because the molecules come from a greater or less distance, where (e.g.) the value of w is positive. The distance from which (upon the average) the molecules may be supposed to have come depends upon circumstances. If, for example, the molecules, retaining their number and velocity, interfere less with each other’s motion, the distance in question will be creased. The same effect will be produced, without a change of quality, by a simple _ reduction in the number of molecules, 7.¢., in the density of the gas, and it is not difficnlt to recognise that the distance from which the molecules may be supposed to have come is inversely as the density. On this account the passage of tangential momentum per molecule is in- versely as the density, and since the number of molecules crossing is directly as the density, the two effects compensate, and upon the whole the tangential force and therefore the viscosity remain un- altered by a change of density. On the other hand, the manner in which this viscosity varies with temperature depends upon the nature of the encounters. If the molecules behave like Boscovich points, which exercise no force upon one another until the distance falls to a certain value, and which then repel one another infinitely (erroneously called the theory of elastic spheres), then, as Maxwell proved, the viscosity would be proportional to the square root of the absolute temperature. Or again, if the law of repulsion were as the inverse fifth power of the distance, viscosity would be as the absolute temperature. In the more general case where the repulsive force varies as 7~”, the dependence of » upon temperature may also be given. If v be the velocity of mean square, proportional to the square root of the tem- On the Viscosity of Argon «s affected by Temperature. 69 2+3 perature, p varies as v”—!, a formula which includes the cases (n=5, nm = 00) already specified. If we assume the law already discussed —that » is independent of density—this conclusion may be arrived at very simply by the method of “ dimensions.” In order to see this we note that the only quantities (besides the density) on which » can depend are m the mass of a particle, v the velocity of mean square, and / the repulsive force at unit distance. The dimensions of these quantities are as follows :— p = (mass)! (length)~? (time)"}, m = (mass)', (length)! (time)“}, (mass)! (length)*“? (time)~?. v k I Thus, if we assume FRCS RO CM SORE APE, SRR acon ee OPE B (1), wehave 1 = “+2, ~-1 = yt+(nt+l)z, -1 = —-y-2z, n n+ 3 2 whence % = wie a Sill eh a : om | n—-1 n-1 GE IN 0 ee ea Accordingly pe = 91 w—1 L ye-l : k JIA AAY boie\n sialnialece atel oles) wie: eiele (2), where « is a purely numerical coefficient. For a given kind of mole- cule, m and & are constant. Thus u+3 a+3 Dee ia Genoa (3). The case of sudden impacts (n = ©) gives, as already remarked, peava« @. Hence k disappears, and the consideration of dimensions shows that » « d~*, where d is the diameter of the particles. The best experiments on air show that, so far as a formula of this kind can represent the facts, p «a 6°. It may be observed that nm = 8 corresponds to p « 6°”. When we remember that the principal gases, such as oxygen, hydrogen, and nitrogen, are regarded as diatomic, we may be inclined to attribute the want of simplicity in the law connecting viscosity and temperature to the complication introduced by the want of symmetry in the molecules and consequent diversities of presentation in an encounter. It was with this idea that I thought it would be interest- ing to examine the influence of temperature upon the viscosity of argon, which in the matter of specific heat behaves as if composed of 70 . Lord Rayleigh. single atoms. From the fact that no appreciable part of the total energy is rotatory, we may infer that the forces called into play during our encounter are of a symmetrical character. It seemed, therefore, more likely that a simple relation between viscosity and temperature would obtain in the case of argon than in the case of the ‘diatomic ” gases. The best experimental arrangement for examining this question is probably that of Holman,* in which the same constant stream of gas passes in succession through two capillaries at different temperatures, the pressures being determined before the first and after the second passage, as well as between the two. But to a gas like argon, avail- able in small quantities only, the application of this method is difficult. And it seemed unnecessary to insist upon the use of constant pressures, seeing that it was not proposed to investigate experimentally the dependence of transpiration upon pressure. The theoretical formula for the volume of gas transpired, analogous to that first given by Stokes for an incompressible fluid, was developed by O. E. Meyer.t Although not quite rigorous, it probably suffices for the purpose in hand. If p;, V; denote the pressure and volume of the gas as it enters the capillary, po, V2 as it leaves the capillary, we have ) meee | mVi = pV » = Tea Oe (052 Haas so aac s SaaS (4). In this equation ¢ denotes the time of transpiration, R the radius of the tube, / its length, and p the viscosity measured in the usual . way. In order to understand the application of the formula for our pre- sent purpose, it will be simplest to consider first the passage of equal volumes of different gases through the capillary, the initial pressures, and the constant temperature being the same. In an apparatus, such as that about to be described, the pressures change as the gas flows, but if the pressures are definite functions of the amount of gas which at any moment has passed the capillary, this variation does not inter- fere with the proportionality between ¢ and ». For example, if the viscosity be doubled, the flow takes place precisely as before, except that the scale of time is doubled. It will take twice as long as before to pass the same quantity of gas. Although different gases have been employed in the present experi- ments, there has been no attempt to compare their viscosities, and indeed such a comparison would be difficult to carry out by this method. The question has been, how is the viscosity of a given gas affected by a change of temperature? In one set of experiments the * Phil. Mag.” vol. 3, p. 81, 1877. + ‘Pogg. Ann.,’ vol. 127, p. 269, 1866. — On the Viscosity of Argon as affected by Temperature. 71 capillary is at the temperature of the room; in a closely following set the capillary is bathed in saturated steam at a temperature that can be calculated from the height of the barometer. If the temperature were changed throughout the whole apparatus from one absolute temperature 6 to another absolute temperature 6’, we could make immediate application of (4); the viscosities (p, py’) at the two temperatures would be directly as the times of transpiration (é, ¢). The matter is not quite so simple when, as in these experiments, - the change of temperature takes place only in the capillary. A rise of temperature in the capillary now acts in two ways. Not only does it change the viscosity, but it increases the volwme of gas which has to pass. The ratio of volumes is 0’, ; and thus ies aes fag subject to a small correction for the effect of temperature upon the dimensions of the capillary. It is assumed that the temperature of the reservoirs is the same in both transpirations. . The apparatus is shown fig. 1. The gas flows to and fro between the bulbs A and B, the flow from A to B only being timed. It is confined by mercury, which can pass through U connections of blown glass from A to C and from B to D. The bulbs B, C, D are sup- ported upon their seats with a little plaster of Paris. The capillary is nearly 5 feet (150 cm.) in length and is connected with the bulbs by gas tubing of moderate diameter, all joints being blown. E represents the jacket through which steam can be passed ; its length exceeds that of the capillary by a few inches. In order to charge the apparatus, the first step is the exhaustion. This is effected through the tap, F, with the aid of a Toppler pump, and it is necessary to make a corresponding exhaustion in C and D, or the mercury would be drawn over. To this end the rubber terminal H is temporarily connected with G, while I leads to a common air pump. When the exhaustion is complete, the gas to be tried is admitted gradually at F, the atmosphere being allowed again to exert its pressure in Cand D. When the charge is sufficient, F is turned off, after which G remains open to the atmosphere, and H is connected to a manometer. When a measurement is commenced, the first step is to read the temperatures of the bulbs and of the capillary; I is then connected to a force pump, and pressure is applied until so much of the gas is driven over that the mercury below A and in B assumes the positions shown in the diagram. I is then suddenly released so that the atmospheric pressure asserts itself in D, and the gas begins to flow back into B. The bulb J allows the flow a short time in which to establish itself before the time measurement begins as the mercury passes the 72 Lord Rayleigh. connection passage K. When the mercury reaches L, the time measurement is closed. One of the points to be kept in view in designing the apparatus is to secure long enough time of transpiration without unduly lowering the driving pressure. At the beginning of the measured transpiration the pressure in A was about 30 cm. of mercury above atmosphere, and Ah (ee that in B about 2 cm. below atmosphere. At the end the pressure in A was 20 cm., and in B 3cm., both above atmosphere. Accordingly the driving pressure fell from 32 to 17 cm. Three, or, in the case of hydrogen, five, observations of the time were usually taken, and the agreement was such as to indicate that the mean would be correct to perhaps one-tenth of a second. ‘The time for air at the temperature of the room was about ninety seconds, and for hydrogen forty-four seconds, but these numbers are not strictly com- parable. On the Viscosity of Argon as affected by Temperature. 73 When the low temperature observations were finished, the gas was lighted under a small boiler placed upon a shelf above the apparatus, and steam was passed through the jacket. It was necessary to see that there was enough heat to maintain a steady issue of steam, yet not so much as to risk a sensible back pressure in the jacket. The time of transpiration for air was now about 139 seconds. Care was always taken to maintain the temperature of the bulbs at the same point as in the first observations. There are one or two matters as to which an apparatus on these lines is necessarily somewhat imperfect. In the high temperature measurements the whole of the gas in the capillary is assumed to be at the temperature of boiling water, and all that is not in the capil- lary to be at the temperature of the room, assumptions not strictly compatible. The compromise adopted was to enclose in the jacket the whole of the capillary and about 2 inches at each end of the approaches, and seems sufficient to exclude sensible error when we remember the rapidity with which heat is conducted in small spaces. A second weak point is the assumption that the instantaneous pressures are represented by the heights of the moving mercury columns. If the connecting U-tubes are too narrow, the resistance to the flow of mer- cury enters into the question in much the same way as the flow of gas in the capillary. In order to obtain a check upon this source of error the apparatus has been varied. In an earlier form the connecting U-tubes were comparatively narrow ; but the result for the ratio of viscosities of hot and cold air was substantially the same as that sub- sequently obtained with the improved apparatus, in which these tubes were much widened. [Even if there be a sensible residual error arising from this cause, it can hardly affect the comparison of temperature- coefficients of gases whose viscosity is nearly the same. I will now give an example in detail from the observations of December 21 with purified argon. The times of transpiration at the temperature of the room (15° C.) were in seconds 1042, 1044, 1043. Mean, 104-67. When the capillaries were bathed in steam, the corresponding times were 1674, 1674, 1673. Mean, 167-58. The barometer reading (corrected) being 767-4 mm., we deduce as the temperature of the jacket 100:°27°C. Thus 6 = 287°5, 0 = 372°8. The reduction was effected by assuming 74 On the Viscosity of Argon as affected by Temperature. With the above values we get es Ae As appears from (5), the integral part of x relates merely to the expansion of the gas by temperature. If we take we get i= aOrol 2s This number is, however, subject to a small correction for the ex- pansion of the glass of the capillary. As appears from (4), the ratio p, » as used above requires to be altered in the same ratio as that in which the glass expands by volume. The value of n must accordingly be increased by 0:010, making nm = 0°822 The followmg table embodies the results obtamed in a somewhat extended series of observations. The numbers given are the values of n in (7), corrected for the expansion of the glass. AaY (iby) <3. eee eee 0-754 Oxy cen So ee 0-782 Hivdrogen i tak iam case. 0-681 Argon (impure)............ 0-801 Argon (bestia ey aor 0-815 In the last trials, the argon was probably within 1 or 2 per cent. of absolute purity. The nitrogen lines could no longer be seen, and scarcely any further contraction could be effected on sparking with oxygen or hydrogen. It will be seen that the temperature change of viscosity in argon does not differ very greatly from the corresponding change in air and oxygen. At any rate the simpler conditions under which we may sup- pose the collisions to occur, do not lead to values of 7 such as 0°5, or 1:0, discussed by theoretical writers. I may recall that, on a former occasion,* | found the rates of argon to be 1:21 relatively to that of air, both being observed at the temperature of the room. * ‘Roy. Soe, Proc.,’ January, 1896. Beequerel and Rontgen Rays in a Magnetie Field. mo “On the Behaviour of the Becquerel and Rontgen Rays in a Mamene ield. By the Hon. hk. J. Strutt, “BAL Scholar of Trinity College, Cambridge. Communicated by Lorp Rayueien, F.RS. Received January 9,—Read January 18, 1900. In the current number of Wiedeman’s ‘ Annalen,’ an experiment is described by Giesel showing that the Becquerel rays are deflected in a magnetic field. This result is of great interest, on account of the light which it throws on the nature of the rays. Up to the present, the ‘evidence has tended to show that the Becquerel rays were of the same nature as the ‘Rontgen rays, both being capable of penetrating thin metal sheets, of affecting a photographic plate, and of producing ionisa- tion in the surrounding air. Neither could be refracted or reflected ; and so far as has yet appeared, neither could be polarised. These facts seemed to form a fairly strong body of evidence that the two kinds of radiation were essentially similar. But the announce- ment of the magnetic deflectibility of the Becquerel rays seems to throw doubt on this conclusion. The Rontgen rays, so far as is known, are quite unaffected by magnetic force. Under these circum- stances it seemed worth while to make a new attempt to discover such an effect on the Réntgen rays. This attempt I have carried out. It will be best to say at once that the result is negative. A focus tube was employed as the source of radiation. It was placed at a distance of about 35 cm. from a powerful electro-magnet, and in such a position that the cathode rays in the tube were parallel to the magnetic force due to the magnet. The line joining the oblique anti-cathode to the centre of the magnetic field lay in the plane of the anti-cathode. A short distance in front of the magnet a wire was placed at right angles to the direction of the rays, and in the plane of the anti-cathode, It was thus at an angle of about 50° to the magnetic force—the same angle as that between the axis of the cathode stream and the anti- cathode. This wire was used to cast a shadow on a photographic plate placed at a distance of 65 cm. on the other side of the magnet. An exposure was first made with the magnetic force in one direction. The exposure was then stopped, the field reversed, and another ex- posure given of course without shifting the plate. If then the rays had been appreciably deflected, the photograph should have shown two shadows, either overlapping, or altogether separated. The rays casting the shadow were those emitted at a grazing angle from the anti-cathode. The reason for using these very oblique rays was that owing to the foreshortening of the anti-cathode, the source was virtually narrower than it would have been, had rays VOL. LXVI. H 76 Hon. R. J. Strutt. On the Behaviour of the been used which left the anti-cathode at a greater angle. Thus sharper shadows were obtained, and a smaller magnetic deflection could have been detected. The tube was arranged with its cathode stream parallel to the eaenotic field, so as to avoid any shifting of the source of radiation when the magnet was reversed, owing to an effect of the magnet on the original cathode beam. Such a shifting would have given rise to a spurious effect. The only objection to this was that the shadow-casting wire had to be obliquely placed so as to be in the plane of the anti-cathode. Thus some sensitiveness was lost. I shall now give an estimate cf the smallest deflectibility which could have been detected. ‘The rays traversed a distance of 65 cm. after leaving the magnetic field. It was estimated that a lateral displacement of the shadow of the wire by 0:02 cm. could have been detected. But the wire was in- clined 50° to the resultant magnetic force. Thus the smallest real 0-02 in 50 _ The smallest angular deflection of the rays which could be detected would be, in circular measure, 0:02 65 sin 50 displacement that could with certainty be detected was cm = 0°000405. ‘The length through which the rays were exposed to the magnetic force was 8 cm. If in this distance they were bent through the above angle, the radius of curvature would be 8 = Cin, so y 0000405 em 9,800 cm The strength of the magnetic field was determined in the usual manner, by observing the throw of a galvanometer when a small coil of known dimensions connected up with it was suddenly withdrawn from the region between the pole pieces. To reduce the results +o absolute measure, the throw due to reversal of an earth-inductor in the same circuit was observed. In this way the strength of the field was found te be 3270 C.G.S. It is convenient to exhibit the result by givmg the maximum field which the experiments indicate as unable to produce a curvature of qagims lem, . Since a field of 3270 does not produce a curvature of radius less than 19,800 cm., we see that the field required to produce a curvature of radius 1 em. cannot be less than | | 6x 10°. Becquerei and Réintgen Rays in a Magnetic Freld. ME Owing to the fact that the magnetic field was reversed instead of being merely shut off, the experiment is really of double the sensitive- ness indicated above. But, in order to be well on the safe oe it has been thought best to leave this out of account. For the sake of comparison I have attempted a rough estimate of the amount of the magnetic deflection of the Becquerel ‘Tays. The method employed was as follows :— a pS Si a yl Se ” A photographic plate, shown in section at ad, was laid on the top of the square pole pieces of a magnet, the magnetic force being perpendi- cular to the plane of the diagram. The plate was covered with thin aluminium foil ; ¢ is a metal capsule filled with the substance d, which emitted the rays.* When no magnetic force was acting, the rays were emitted from the capsule as indicated in fig. 1, some of them striking obliquely on the plate. On development after one hour’s exposure, a shadow was obtained beginning at the edge of the capsule c, and extending a short distance. The effect gradually tailed off, and at a few cm. distance away from ¢ it was inappreciable. When the magnetic force was in such a direction as to bend the rays down into the plate (fig. 2), the Hie. uZs shadow extended further. When, on the other hand, the magnet was reversed so as to bend the rays away from the plate (fig. 3), the Ficus. 4 as *.The substance employed was a preparation from uranium residues, pees by de Haen, Hamburg. bit ne 2 78 Behaviour of Becquerel and Réntgen Rays in a Magnetic Field. shadow obtained on development was much shorter, the time of ex- posure being, of course, in each case the same. The numerical estimate of the curvature of the rays was obtained from an experiment of the latter kind. Wie. 4. Y nee OLS oo Let us suppose that cc (fig. 4) represents in section the front surface of the radiating substance, cf the surface of the photographie plate. Let f be the place furthest from c¢ at which the darkening on the photographic plate was perceptible. Now the rays which reach furthest are those which proceed from e¢, the highest point of the radiating surface, as may easily be seen from geometrical considera- tions. The rays which reach / must consequently proceed from e. Rays proceeding from any lower point of the surface will either be bent up so as never to reach the plate at all, or else they will strike it short of f. The ray which reaches / from ¢ will clearly just graze the surface of the plate at /. lf v be the radius of curvature, 5 the distance ec, and / the distance ef, then ‘[? (ar — ba ores h(Qr—b) = PB, or a7. +h) If then we measure 0, the height of the highest part of the radiating surface above the plate, and / the greatest distance to which the darkening of the plate extends, we have data for determining 7. It must be admitted that the measurement of / involves great un- certainty. The image gradually tails off, and any estimate of its length must to a great extent be arbitrary. The value of 7 deduced is more uncertain still, since /? is involved in calculating it. But, in spite of these objections, the method may, I think, be relied on to give the order of magnitude of 7, and that is all that is required, so far as the conclusions which it is here sought to draw are concerned. In one experiment, the length / was estimated at 2 cm.; } was 08cm. Thus7v = 3 cm. approximately. The strength af the magnetic field, measured as before, was 1680 C.G.S. Thus the field required to produce a curvature of radius 1 em. is about 5 x 10°. In another experiment, / was 1°8 cm., b was 0°8 em., and the field 2140. This gives practically the same result as the preceding. — The Thermo-dynamical Properties of Superheated Steam. 79 In an experiment described by Professor J. J. Thomson, a beam of cathode rays was bent to a radius of curvature of 9 cm. in a field of 35 units. Thus a field of 315 would have been required to bend it to a radius of 1 cm. Let us now collect the results obtained, and compare them with this. The field which would be required to produce a curvature of 1 cm. radius would be RIE P AGH ODS LAYS 22 scicni. sh vicvdeveyoedh vyee sane 3x 10? Paermecqtierel fays) (22.5.0 0.22.500. eee 5 x 108 » Rontgen rays not less than ............. 6 x 10°. If the Réntgen rays are magnetically deflected at all, it is by an amount less than a ten-thousandth part of that observed in the case of the Becquerel rays. The magnetic deflectibility of the Becquerel rays cannot but be con- sidered to be a most characteristic property. And the above result appears to make it tolerably certain that the Réntgen rays do not possess this property. It is to be concluded, therefore, that the Becquerel rays are, after all, essentially different in character from the Rontgen rays. “An Experimental Investigation of the Thermo-dynamical Pro- perties of Superheated Steam.” By Joun H. GRINDLEY, B.Se., Wh. Sch., Exhibition (1851) Scholar, late Fellow of the Victoria University. Communicated by Professor OSBORNE REYNOLDS, F.R.S. Received April 21, 1899—Read January fe 1900. (Abstract.) Part 1.—On the Law of Flow of Saturated Steam through Small Orifices. In making experiments on the thermal properties of superheated steam obtained by wiredrawing saturated steam, it is essential that certain laws assumed in theory to govern the flow through the orifice should obtain in practice. 7 Among these laws the only one on which a difference would be expected to exist between experiment and theory, is the law of adiabatic expansion assumed to hold during the flow. Since such adiabatical flow is not only assumed, but is indispensable in obtaining temperature results in the wiredrawn steam which will enabie deductions to be made by theory of the initial dryness of the steam or its thermal condition after wiredrawing, it was found im- 80 Mr. J. H. Grindley. An Experimental Investigation of portant to know the circumstances under: which adiabatic flow could be experimentally obtained. Many experimental results have already been given by various experimenters which indicate the laws governing the flow through various types of orifices, and to some extent bear out the theoretical conclusions on the subject, but so far as the author is aware, no experiments have yet been made with saturated steam, showing results which entirely agree with those deduced from theory by assuming adiabatic flow, and hence arose the necessity of making experiments - with this object in view. It appears from the theory, that when the ratio (the lower to the higher) of the two pressures causing the flow through the orifice is diminished below a certain value, the upper pressure being kept constant, the rate of discharge of the steam should be constant. This value of the pressure ratio depends entirely on the law of expansion assumed to hold during the flow. By assuming this law to be repre- sented by an equation of the form po = constant, p being the pressure, v the specific volume of the gas, and n a constant for any particular gas, we can deduce this value of the pressure ratio giving maximum flow in the form pe ie a — —_— = n—Il- Py ie se Putting x = 10/9 we get for saturated steam expanding adiabati- cally during its flow through the orifice. £2 0-584. di. If now the flow of steam be truly adiabatic in an experiment, this particular value of the pressure ratio giving the maximum flow should be actually found by the experiment, and if some other value than this be obtained the law of flow will not then be the true adiabatic one for saturated steam. Iience the attainment of this particular value of the pressure ratio’ giving the maximum discharge was made the object of the experi- mental inquiry here described, since it would follow that the law of expansion through the orifice was then truly the adiabatic law for saturated steam. To begin with, an orifice was drilled in a piece of thin sheet brass the nature of which should create, if possible, a large deviation from the adiabatic in the actual law of flow through the orifice. Experi- ments were then made with this orifice placed between a steam chest and a condenser, the weight of steam passing per minute being taken the Thermo-dynamical Properties of Superheated Steam. 81 at varlous pressures over a wide range of pressure ratio, the upper pressure being kept constant. It was found that the maximum rate of discharge did not occur until the ratio of pressures had fallen to 0°33, a value far below that given by the theory, and indicating a far different law of flow through the orifice than the adiabatic. As a contrast to this, since the main element in the question appeared to be the conductivity of the substance in which the orifice is made, the later experiments were made with an orifice drilled in a glass plate, the orifice being neither sharp lipped nor smoothly rounded, the lip in the best circumstances presenting a rough chipped edge. Now with such material for an orifice plate, it is evident that any passage of heat between the glass and the steam will be very small, and also that if adiabatic flow is not now obtained, then either there must be a passage of heat between various portions of the glass and the steam in contact with them, or heat must be conducted along the stream of vapour itself, the latter being considered negligible from considerations of gaseous conductivity. The experiments made with this orifice show a complete agreement between the results of experiment and theory, and that the law of flow through an orifice drilled in a plate of glass, no conditions being attached as to the roundness or otherwise of the lip of the orifice, is precisely the adiabatic law assumed in the theory. Part Il.—On the Cooling of Saturated Steam by Free Expansion. In Regnault’s experiments on the total heats of saturated steam under various pressures, the steam was withdrawn upwards from a boiler, allowing any entrained moisture in the steam to be separated by gravity. Saturated steam obtained in any other manner would not necessarily have the same total heat as that obtained by Regnault at the same pressure, and it is therefore of great importance to note that the dryness of the steam in Regnault’s experiments was obtained by the simple method of draining suspended moisture from it. Henee, since the foundation of most of the researches on het thermal properties of steam rests upon Regnauit’s results, it would be well to accept as a definition of dry saturated steam that condition of steam which is obtained by draining from wet steam any entangled moisture. In making experiments on the thermal condition of superheated steam obtained by wiredrawing saturated steam, a knowledge of the total heat of evaporation of the steam before wiredrawing is necessary, and as Regnault’s tables of the total heats of saturated steam only apply to steam obtained in the above manner, it must also be obtained in the same manner ior these tables to apply. The precise object of this paper is to describe a research on the thermal properties of superheated. steam, these properties being deduced - 82 Mr.J.H. Grindley. An Experimental Investigation of from a knowledge of those of saturated steam already obtained by Regnault. The temperature and pressure of saturated steam in a steam chest in which a constant supply of steam is kept is taken, the steam is then drawn upwards to an orifice, and, after wiredrawing, its pressure and temperature are again taken, using for the determination of the latter a thermo-electric junction immersed in the steam. Special precautions were found necessary, and special apparatus designed to prevent losses of heat by radiation from the channel con- taining the wiredrawn steam, a steam jacket of peculiar construction enveloping this channel completely, and by adjusting the temperature of the jacket to equality with that in the wiredrawn steam, all radia- tion was effectively prevented from this portion of the apparatus. Again, communication of heat from one side to the other of the orifice through the substance in which the orifice is made was pre- vented, and true adiabatic flow obtained through the orifice by drilling it in a piece of plate glass, such as that described in the research on the law of flow through orifices. | During an experiment, the pressure in the steam chest being kept constant, a series of temperature readings at various values of the lower pressure were observed in the wiredrawn steam. By this means a curve showing the cooling of the steam for any degree of wiredrawing from an initial constant pressure could be drawn on a pressure-tempera- ture diagram. Provided now that the total heat of steam before passing the orifice was known, it would be possible to deduce from these temperature and pressure results the values of the mean specific heat at constant pressure of superheated steam between the saturated condition and the tempera- ture of the wiredrawn steam at any given pressure, and further, the total heat of steam at any pressure and temperature obtained by such wiredrawing, would be known. Whether the steam was in the same condition before wiredrawing as that obtained in Regnault’s experiments was certainly not an easy point to decide. In both cases, however, the steam was obtained by draining any suspended moisture from steam initially wet, but whether this process of drainage always brought the steam into the same con-- dition as to dryness, whatever the degree of wetness originally in the steam, was as yet an open question, which could only be decided by experiment. Accordingly experiments were conducted with saturated steam at a known pressure and temperature in the steam chest, but at different degrees of wetness in different experiments. The results obtained are very important, as the maximum difference of temperature at any particular pressure in the wiredrawn steam which could be found to exist between experiments with different degrees of wetness in the steam in the steam chest was 0°35° F., and generally the differ- the Thermo-dynamical Properties of Superheated Steam. 83 ence could not be distinguished, it being remarked that if the dryness of the steam before passing the orifice had been altered by so little as 0:06 per cent., a difference of 1° F. should have been observed in the temperature of the wiredrawn steam. : It would, therefore, appear that saturated steam at any particular pressure obtained by relieving it of suspended moisture by gravitation has only one condition as to its dryness, and also that steam in this particular condition was obtained both in these experiments and in those of Regnault, and it is therefore taken that the steam before wire- drawing has a total heat given by Regnault’s tables of the total heats of saturated steam. Further experiments were also made to observe the effect of altering the position of the thermo-electric junction in the wiredrawn steam, of the effect of the steam jacket on the temperature of the wiredrawn steam, and of the effect of the velocity of the steam through the apparatus on these temperature readings. The amount of the corrections required for the conduction of heat between various portions of the apparatus and the steam was.also calculated, but on account of the precautions taken these were generally found to be negligible. The method of fixing the absolute temperature of the wiredrawn steam should be here men- tioned, as it is a point of great importance, on account of the diffi- culties attending the accurate measurement of the temperature. In — the experiments the thermometer was used merely as a scale to compare the temperature of the wiredrawn steam with that of saturated steam under a known pressure flowing through the same portions of the appa- ratus with about the same velocity, the fixing of the temperature being again dependent on Regnault’s tables of the pressure-temperature rela- tion of saturated steam. The final results obtained show clearly that within the limits of tem- perature obtained by wiredrawing saturated steam at temperatures varying from 240° to 380° F., the condition of the steam known as a pertect gas was not obtained, even when the wiredrawing was con- tinued to 3 lbs. or 4 Ibs. per square inch absolute pressure ; and further, that between the same temperatures and between pressures of 2°5 lbs. and 195 lbs. per square inch, there was not found any indication of a constant value of the specific heat at constant pressure in the super- heated steam. The specific heat at constant pressure was found to increase with temperature, the mean specific heat at atmospheric pressure between the temperatures 230°7° and 246°5° being 0°4317, and between temperatures 295° and 311°5° the mean specific heat was — -0°6482. 3 As regards the variation in the value of the specific heat at con- stant pressure of superheated steam with the pressure, it appears from an examination of the results obtained at about the same tem- perature but under different pressures, that if any such variation in the 84 Mr. J. H. Grindley. An Experimental Investigation of specific heat exists it will be very small compared with the variation with temperature, such examination indicating that the value of the specific heat is sensibly independent of the pressure. : The law of cooling followed by the wiredrawn steam is iHightly different from that obtaiming in many other gases, viz., that the fall of temperature varies directly as the difference of pressure. The rate of cooling was found to diminish with increase of initial temperature. The curves showing the pressure-temperature relations of the super- heated steam wiredrawn from definite initial pressures, seem to follow for a short distance the law of boiling points, and the experiments show that this coincidence always exists in saturated steam, and may well be mistaken for evidence of wetness in the steam. Tables showing the fall of temperature with pressure in the wire- drawn steam, of the total heat of the steam under certain pressures and temperatures, and of the mean value of the specific heat at con- stant pressure of superheated steam at definite pressures and between definite temperatures, accompany the paper. Parr Hi. In this portion of the paper the two properties of steam deduced directly from the experimental figures, viz., the specific heat K, and the cooling effect 66/dp or c, are more directly considered. In the first place, the cooling effect c is found to be inversely proportional to 7°", where 7 is the absolute temperature. . It is then shown that the pF aie formula 4, O 5K) = ~£ Ky) is capable of strict proof from Gti principles, the inter- pretation of the formula beimg that the variation of K, with the pressure at constant temperature is equal to the variation of the pro- duct cK, with the absolute temperature at constant pressure, but cf opposite sign. Applying this to steam when superheated, it has been shown in Part IL of the paper that the variation 9K, is zero to the degree Op of accuracy to which the experiments have been taken. It follows, therefore, from the above formula, that the variation * (cK,) should equal zero; hence, the values of the product cK, have been tabulated for different pressures and temperatures, and so far as the results go, it is clearly shown that the product cK, is an absolute constant, which means that the variations ¢ (cK,) and — ap ae are both zero. Cr the Thermo-dynamical Properties of Superheated Steam. 85 Since the variation es (cK,) = 0, it is possible to integrate at once for the case of superheated steam Thomson’s formula for the cooling effect c, which may be written dz i de the resulting equation being V+CKy 4 - — p] when A may be a function of the pressure. This equation has been used to find the specific volumes of superheated steam under various conditions of pressure and temperature, the value of A being deduced from known data in the saturated condition of the steam. The calculated specific volumes, the accuracy of which depends solely on the experimental results obtained in the research, are compared with those obtained experimentally by Hirn, the results in general “agreeing very well. It is also of interest to notice that in any gas in which K, does not vary with the pressure, the product cK, must also be independent of the temperature in that particular gas, since the equation O C 3, Bo) = 5 &) must be satisfied identically, and hence the equation v+cKy dv T ~~ dr must be immediately integrable for the gas in the form V+ Ky _ fo 86 My. S. A. Sworn. “ Researches in Absolute Mercurial Thermometry.” By the late S. A. Sworn, M.A. Communicated by H. B. Drxon, F.R.S. April 21,—Read June 15, 1899. (Abstract, prepared at the request of the Council by ARTHUR SCHUSTER, F'.R.S., December, 1899.) The experimental portion of this work consists of the careful com- parison of six thermometers, with the object of studying the effects of capillarity, and in the second place of obtaining a comparison between thermometers made of English flint glass with those of French “ verre dur” or Jena normal glass, and therefore indirectly with the hydrogen scale. The instruments employed consisted of a Tonnelot “verre dur ” thermometer, to be referred to as No. 4976, an English flint glass (No. 711,179) by J. J..Hicks, two normal thermometers (Nos. 2218 and 2219) of Jena 16"! by Gerhardt of Bonn, and two calorimetric thermo- meters Nos. 2220 and 2221 of Jena 16™ by Gerhardt, with a range from — 2° to 25° C. The Tonnelot instrument is divided on the transparent stem into tenth degrees, and is cylindrical in the bore. The other thermometers have enamelled backs, and are divided on the stem into half milli- metres. At the time the latter instruments were obtained elliptical bores were the only ones procurable, but care was taken that the bore was not unduly flattened, and was smooth in contour. The author considers the readings taken with these thermometers to be quite trust- worthy. The ratio of the major to the minor axis of the bore was about 2 for the Jena glass thermometers and 3 for 711,179. In each case the bulb (without enamel) was fused to the stem. Ampoules were avoided in all the instruments. The calibration corrections were obtained in the usual way, a micro- meter being used to measure the ends of the thread. The reduction was made by the Neumann-Thiessen method. All readings other than those for calibration were made with a telescope magnifying eighteen to twenty-four times, the eye-piece of which was provided with a micrometer scale by Zeiss. With the aid of this eye-piece, which serves to further subdivide the thermometer divisions, the readings agreed to 0:°005 mm. Several readings were always taken, generally three for zero readings, six for the indications in steam, twenty-one for coefficients of external pressure, fifty-four for coefficients of internal pressure ; twenty-seven of zero and fifty-four in steam for the fundamental internal correction, and ninety of comparison and eighteen of zero on each instrument during the comparisons. The Researches vn Absolute Mercurial Thermometry. 87 probable error of the separate results for the various constants is about 0-001° C. The Constant of Capillary Depression K and the Coefficient of External Pressuve.—It is usual to determine the pressure coefficient by suspending the thermometer in a tube, the pressure within which can be rapidly changed from atmospheric pressure to one of a few centimetres of mercury. A sudden diminution of pressure p causes a fall in the indications of the thermometer, and in the absence of capillary effects 5/p would measure the so-called coefficient of external pressure. But the fall of thermometer being accompanied by a change in the shape of the meniscus, the readings before and after the change of pressure are not directly comparable. If the meniscus is normal at high pressure (which can be realised in the experiment, by arranging for a slowly rising temperature), we must add to the reading at the low pressure a certain constant K, which represents the difference in the readings of a thermometer between a rising and falling thread. Hence (6+ K)/p will be the corrected coefficient of external pressure (6 ). If ten sets of observations are made, with different changes of pressure (1, 2) giving different falls of the thermometer (6,, 5,), we may put og ie Nl mao fy P2 and hence K Pe Po Be = eat P2- Pi P2- Pr In the actual observations the changes of temperature which take place between the readings must, of course, be allowed for, and the equations only hold if these changes are so slow that at the low pressure the actual temperature has not overtaken the apparent tem- perature at the moment the reading is taken. Previous observers not being interested directly in the quantity K, have arranged their experi- ment so that the readings of low pressure were only taken after a_ time sufficient to allow the rising temperature to have its effects, so that the thread was rising in all observations. Mr. Sworn, on the other hand, wishing to determine K and f. simultaneously, had to arrange the experiment so that at low pressures the hydrostatic pressure in the bulb was the same as with a falling thread. The following table gives an idea of the consistency of the results obtained :— 88 Mr. 8S. A. Sworn. Thermometer. Temperature. po. K, 2220 0° 654 p, = 525 mm. 655 561 42.4, J 654 © 0-1243 mm. | 574 414 2045? 669 i | | 664 0°1070 mm. | | ye: lal Mean K = 0°'111 mm. + 0°005. = 0:0065° + 0:0003. The value of K for the various thermometers was found to be as follows :— Thermometer. ie POZO) (i 2. a. eee 0:0065° + 0:0003 7, Re trea Behe 0:0098° t OP Bie eee 0:0087° + 0:0006 DEVO: cdc Re ae 0°0104° + 0°0013 BONG oh ls. eae: ee 0:0051° + 0:0004 CABRHOs | 0°0105° + 0:0004 Mr. Sworn concluded from his results that K is a constant not affected by a change in the rate of rise in temperature, and not appreci- ably different in different parts of the tube, if the average value of K over a space of several millimetres is always taken. The Fundamental Interval and the Coefficient of Internal Pressure —The zeros were determined by plunging the thermometer into a mixture of finely pounded ice and distilled water. Samples of ice were frequently prepared from distilled water, which had for some time been kept in a partial vacuum of 50—100 mm. Norwegian ice was also used, and within the limits of experimental error was always found to give the same results as the specially prepared ice. The purity of the ice was invariably controlled by testing for chlorides, by the Nessler test, and by evaporation to dryness in a platinum basin. In order to be sure Researches in Absolute Mercurial Thermometry. le that the two varieties of ice would give the same results, control determinations were made with Nos. 2220 and 2221, the indications of which could be relied upon to show differences exceeding 0:001° C. The apparatus for taking the zeros did not differ iP lange from that generally used and described by Guillaume. | The thermometer was plunged into the ice within one or two minutes after removing it from the hypsometer, whilst the bulb was still at 40—50°. The thermometer was held vertically in the hand until the mercury had fallen sufficiently for the bulb to be immersed with safety into the ice. 5-—10 mm. of the stem above 0° C. were exposed to the ice, adjustment to the vertical made, the thermometer raised so that the image of the meniscus was just clear of the ice, and the readings taken.. The stem was always well tapped. The indications of thermometers at the temperature of saturated steam were investigated in a form of rotary hypsometer which pre- sents some slight difference from that used at the Bureau Inter- national. The difference consists in an improvement of the position and construction of the manometer which measures the pressure excess ot the steam. In the Breteuil instrument the manometer keeps its vertical position while that part of the hypsometer which holds the thermometer may be placed in either a vertical or horizontal position. This construction renders it necessary for the manometer opening to be placed at some distance from the thermometer bulb, the two being separated by a narrow passage through which the steam has to pass. The manometer will, therefore, register too high a pressure. To correct for this the steam in passing into the condensers when it is at atmo- spheric pressure, is forced through an exactly similar passage, so that the hypsometer pressure may be assumed to be half way between the pressure indicated by the manometer and the atmospheric pressure. Mr. Sworn gave up the convenience of having the manometer in a fixed position and secured thereby greater certainty in measuring the actual steam pressure at the thermometer bulb. Arrangements had to be made, of course, for the manometer to turn so as to keep the water column vertical when the tube is placed in the horizontal posi- tion. Two manometers were used, one connected with the inner chamber holding the thermometer, and the other with the outer steam jacket, but no difference in pressure could ever be detected. In order to prevent the formation of troublesome water drops in the ma- nometers, short and wide glass chambers were interposed between them and the steam. It was thus ascertained that the pressure excess of the steam could be kept within 0-02 mm. of mercury. A distillation of mercury was avoided by leaving the ast two Fe tcas of the thread unexposed until it was thought that the depressed zero had attained a constant position. Tapping was always resorted to, but the author has been, unable to satisfy himself that it makes any 90 My. S. A. Sworn. difference for any of the instruments. By comparing the observations of the boiling point made with the vertical and horizontal thev- mometers, the coefficient of infernal pressure may be determined, as in the vertical position the hydrostatic pressure of the mercury acts on the bulb. The internal pressure correction (8;) is connected with the external pressure correction (Be) by the relation Bi = Be +k; where « is the compressibility of glass. The following table gives the comparison between the observed and calculated value of f;; i degrees per millimetre pressure of mercury. The column headed / shows the length of the mercury column between the boiling point and the centre of the bulb. | i Thermometer. | 8; (calculated).| Bi (observed). | h. k (assumed). pal Rape —— 4976 0 -0001229 0 -0001241 | 63°8 | 070000154. 711179 0 °001159 - 0°001172 | 57°3 0 -0000127 2218 0 000785 0 ‘000804 56°7 0 :000014:3 2219 0 °000825 0-000878 59 °2 0 00001438 / The observed values of 6; were obtained by dividing the differences in the observed reading (horizontal—vertical position) by the height of the mercury column above the centre of the bulb. No correction was made for the fact that when the column of mercury was raised from the horizontal to the vertical position, the thread descended, and the reading therefore corresponded to one taken with a falling temperature, while in the horizontal position the reading corresponded to one taken with rising temperature. Mr. Sworn concluded from the good agree- ment between the observed and calculated values of 6; that the effect of capillarity is somehow eliminated in these observations. Direct observations were made on this point. If the hypsometer is observed at an angle @ to the horizontal, the readings should differ by the quantity K according as the thermometer is brought into its position from the horizontal, or from the vertical. The observed differences are, however, for the most part x7, and in any case a mere fraction of K. If, further, the thermometer is gradually raised from the horizontal position, the observed differences in the readings should be expressible in the form 4; sin 6 — K, where / is the total height of the mercury column, but in reality they are well expressed by leaving K out of account. The author remarked on this point :— “The effect of capillarity on the advancing or receding columns is unquestionably liable to compensation, either by vibration or by Researches in Absolute Mercurial Thermometry. 91 momentary alterations in temperature or pressure not registered by the manometers. I am personally inclined to think that we are dealing in the hypsometer with steam under what I might term oscillatory conditions of temperature and pressure, the effect of which is to reduce ali the steam indications of the thermometer to what they would be with a receding menisous. Within narrow limits the mercury may advance along the tube, but of necessity there will subsequently be a capillary force erected which will, within the same narrow limits (viz., K), prevent its return. Comparison of Thermometers——The apparatus in which the compari- sons were conducted consisted of a cylindrical tank surrounded, except at the top, by a jacket kept at constant temperature, by a circulation of water heated in a thermo-regulator. The capacity of the tank was 5 litres, and it could be heated or cooled independently, and its temperature set a few degrees above or below that of the jacket. The contained water would then heat or cool at a definite and constant rate. The upper part of the tank has two plate glass sides let in parallel to one another and at right angles to the reading telescope. Two series of comparisons were made. In the first the normal thei- mometers were compared at 20°, 40°, and 60° C. At each tempera- ture the instruments were compared, two at a time in six pairs, the zeros being taken immediately after the second set of readings for each pair. In the second series the calorimetric thermometers, 2220 and 2221 were also utilised, the former in closed series with the normals. These comparisons were made at intervals of 5° from zero to 55° and at 80°. It is not necessary to refer further to the results obtained with the calorimetric thermometers, as it was found that the water in the comparison tank was slightly different at different levels according as the tank was at a temperature higher or lower than that of the atmosphere. The bulbs of the calorimetric thermo- meters being placed at different and varying levels as compared with the bulbs of the standard thermometer, the results were vitiated, but this source of error did not affect the comparison of the standards. In the reduction of observations Mr. Sworn reduced all readings to a falling meniscus. Assuming that the actual observations at the freezing and boiling points of water are those corresponding to a falling thread, he adds to such reading in the comparison the con- stant K previously determined by him. From the result of his investigations, Mr. Sworn drew the conclusion that there is no syste- matic difference between the indications of the verre dur and the Jena 16" thermometers, and that the flint glass thermometers give indica- tions which are practically identical with those of the hydrogen ther- mometer. [The details of the observations are deposited in the Archives of the Society. | VOL. LXVI. I 92 Researches in Absolute Mercurial Thermometry. Note on the above Paper. By ARTHUR SCHUSTER, F.R.S. Received January 4, 1900. Mr. Sworn’s investigations raise some questions of importance in the behaviour of mercury thermometers. The irregularities which are observed in the behaviour of the mercury thread of a thermometer while descending have led observers to take accurate measurements only in a slowly rising temperature. To avoid inconsistencies, the standard temperatures ought also to be measured under conditions which secure the normal formation of the mercury meniscus, which is that of a rising thread. At the temperature of boiling water it is supposed that this can be done by stopping momentarily the flow of steam, so as to lower the temperature betore bringing the mercury thread to its final position. At the freezing point a difficulty has always been felt about the influence of effects of capillarity, and there is no doubt that this is the weakest point at present in the accurate measurement of temperatures with mercury thermometers. Mr. Sworn’s investigations led him to conclude that if the fall of a thermometer is slow (7.¢., when the meniscus travels its own diameter in about one minute), the fall is regular, and not a series of disjointed steps. The difference in the readings of a falling and rising ther- mometer being, according to him, a constant (K), which can be deter- mined by the method described in his paper, it should be possible to reduce readings taken with a rising meniscus to readings with a falling thermometer by simply adding K to the reading. Mr. Sworn’s contention was that this should always.be done, because the freezing point is approached from above, and the boiling point also, according to him, corresponds to a measurement taken with a falling thread. His observations on the behaviour of thermometers in the hypsometer are of considerable importance, but some confirmation is required, because Guillaume describes an experiment which is not in agreement with Mr. Sworn’s conclusion, that the difference in the readings between a rising and falling thermometer disappears when the instrument is suspended in steam. On the contrary, Guillaume determines the amount of the difference by observations in the hypsometer, and states it to be between 0:002 and 0-003" with the standard Tonnelot thermometers. I am inclined to think that the truth lies between the two extremes, and that the effects of capillarity are still appreciable in the steam, but decidedly smaller than at lower temperatures. I am led to this conclusion through my observations at the freezing point with Tonnelot thermometers, which have always made me think that Guillaume must have underrated the effects otf capillarity. But even granting for a moment that there is no effect of capillarity in the hypsometer, I should not be inclined to accept Mr. Sworn’s explanation of the fact, which is that the temperature of Researches in Absolute Mercurial, Thermometry. 93 the steam is slightly fluctuating, and that when it is accidentally high the temperature rises, but when it is low, stiction prevents the thread from falling, so that the ultimate effect is to make the thermometer indicate too high by an amount equal to K. There is no evidence in support of such a fluctuation. It is at least equally probable that stiction is actually of smaller importance at the higher temperature, where the distillation of mercury, which is known to take place from the free surface, must assist the formation of the normal meniscus. Mr. Sworn’s method of reducing thermometric observations depending on the complete disappearance of K at the boiling point, cannot there fore be accepted without further evidence, but the matter is one well worthy of careful investigation. -Mr. Sworn was perfectly right in saying that the three readings, viz., freezing point, temperature, and boiling point, ought to be taken under like conditions of the meniscus, but the proper way of accomplishing this is to alter the usual practice of fixing the zero by substituting a method similar to that of deter- mining the freezing points of solutions. If the water is first slightly undercooled, and then brought to the proper temperature by the intro- duction of a few ice crystals, a great improvement in zero point deter- minations would be effected. Mr. Sworn’s comparisons between thermometers of different com- position were carried out with great care, and may be considered reliable, as far as the instruments used are concerned, but it is not quite certain in how far different thermometers purporting to be made of the same glass may differ. Thus the majority of the Jena glass thermometers carefully studied at Berlin showed a difference of over 001° at 50° when compared with the French hard glass,* but one instrument agreed throughout its range with the latter, while another differed in the other direction. Marek, at Vienna, did not find any systematic difference between the French and Jena glass, and Mr. Sworn’s thermometers also show a practical coincidence. Mr. Sworn’s evidence that his instruments were really made of the 16™! Jena glass rests on the assurance of the maker (Gerhardt, of Bonn), but the anomalous behaviour of two of the Berlin thermo- meters leaves a doubt as to how far blowers are careful to guard against accidental mixing up of different sorts of glass, It is not pcssible, moreover, to compare directly the result of Mr. Sworn’s com- parison with that of other observers, on account of the difference in the method of reduction, but, as far as I can see, the discrepancy would have been greater if Mr. Sworn had reduced his observations in the way adopted at Sevres and Charlottenburg. The same remark applies with greater force to Mr. Sworn’s reduction of the flint glass indica- tions to those of the hydrogen thermometer. He uses Chappuis’s . numbers for the relation between the French hard glass and hydro- * ‘ Zeits. f. Instrumentenkunde,’ vol. 15, p. 438 (1895). VOL. LXVI. K 94 Proceedings and Inst of Papers read. gen scales. But Chappuis’s numbers only apply when the glass ther- mometers are treated and read in the way in which he treated them in his comparisons. That is the great advantage in using the Tonnelot thermometer. It is not an absolute but an intermediate standard. It is immaterial whether Chappuis’s method of treating a glass thermo- meter can be improved upon by adopting a different way of obtaining the zero, or by making corrections for effects of capillarity. It is sufficient to know that consistent results can be obtained if the thermo- meters are always treated in the same way, and whether that way is good or bad, it is the only one which can be used, if we wish to refer the temperature measurement to Chappuis’s hydrogen thermometer: For this reason, and also because we have not at present any guarantee that flint glass thermometers agree sufficiently in composition to give identical results, Mr. Sworn’s conclusions cannot at present be accepted as final. Mr. Chree’s experience* tends in the direction of indicating differences in the behaviour of different thermometers nominally made of the same glass. January 25, 1900. The LORD LISTER, F.R.C.S., D.C.L., 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 :— I. ‘‘ Mathematical Contributions to the Theory of Evolution.—On the Law of Reversion.” By Professor Kari PEARSON, F. B.S. II. “On the Mechanism of Gelation in Reversible Colloidal Systems.” | By W. B. Harpy. Communicated by F. H. NEvILte, F.R.S. IJ. “A Preliminary Investigation of the Conditions which determine the Stability of Irreversible Hydrosols.” By W. B. Harpy. Communicated by F. H. Nevitie, F.R.S. IV. “On the Effects of Strain on the Thermo-electric Qualities of Metals. - Part II.” By Dr. Magnus MAcitEAN. Communi- cated by Lord KELvin, F.R.S. , V. “On the Periodicity in the Electric Touch of Chemical HKlements. Preliminary Notice.” By Professor JAGADIS CHUNDER Boss, Communicated by Lorp Ray.etcuH, F.R.S. * ‘Phil, Mag,’ vol. 45, p. 216 (1898). On the Mechanism of Gelation in Reversible Colioidal Systems. 95 “On the Mechanism of Gelation in Reversible Colloidal Systems.” By W. B. Harpy, Fellow of Gonville and Caius College, Cambridge. Communicated by F. H. NeEvILie,® F.RS. Received January 12,—Read January 25, 1900. Speaking generally, colloidal matter occurs in three conditions :— (1) As fluid mixture, colloidal solutions, or sols, as Graham called them ; (2) Solid mixtures of fluid and solid, the gels; and (3) Solids, such as dry silica or dry glass. The property of forming gels is not possessed by all those mix- tures which have been classed as colloids. Some only form slimes, which even to the point of actual drying retain the fluid property of flowing. Serum albumen and water is an instance. Those which form gels fall into two well-defined classes, according to whether the change from the sol to the gel is, or is not, reversible by a reversal of the conditions which produce it. Silica and water may be taken as the type of the latter, gelatine and water of the former. When a hydrosol of silica forms a hydrogel, the latter is “insoluble.” ‘To this class belong hydrosols of metallic hydrosulphides and oxides. A hydrosol of gelatine sets to a hydrogel by lowering the temperature ; the process is however reversed when the temperature is again raised. As the inner mechanism of the gelation of the hydrosols must differ in the two cases, since in the one irreversible, in the other reversible molecular aggregates are formed, I propose to distinguish the processes by different names. The production of an insoluble gel I will call “coagulation,” of a soluble gel “setting.” This nomen- clature is in accordance with general usage. Temperature is the most potent factor in determining whether a mixture which forms reversible gels is in the sol or gel state. There is also a limiting concentration of the solid below which the gel state is impossible at any temperature. “Setting,” as a rule, follows on a fall of temperature. Caseine, the chief proteid of milk, furnishes, I believe, the only known exception. ‘In the presence of a small quantity of free alkali it forms a hydrosol. When a small quantity of a solution of calcium chloride or nitrate Is added to this, a mixture is produced which forms a hydrogel on warming, and which reforms the hydrosol on again cooling.* Part I. Reverseble;Colloids. Systems containing two or three components occur, that is, binary. or ternary mixtures. The binary.system, agar-and-water, was studied * Sydney Ringer, ‘Journal of Physiology,’ vol. 11, p. 464, 1890. K 2 Ops 2 Mr. W. B. Hardy. On the Mechanism of at considerable length ; and initial experiments were made with the ternary systems, gelatine-water-alcohol, gelatine-water-mercuric chlor- ide, and agar-water-alcohol. These mixtures are homogeneous when heated, but, on cooling, there occurs a division into two fluid phases. In the binary systems and in the ternary system agar-water-alcohol, the conjugate phases have approximately the same refractive index. In the ternary system, gelatine-water-alcohol, the refractive index of the one phase differs so much from that of the other as to permit of - direct microscopical investigation of the form of the surface which separates the phases. For this reason I propose to treat this ternary system first. os The Ternary System—Gelatine-water-alcohol. When 13°5 grammes of dry gelatine are dissolved in 100 c.c. of a mixture of equal volumes of absolute alcohol and water, a system is produced which is clear and homogeneous at temperatures above 20°. As the temperature falls below this limit a clouding occurs, which I find to be due to the appearance of fluid droplets which gradually increase in size until they measure 3u. On cooling further these fluid droplets become solid, and they begin to adhere to one another. In this way a framework is built up, composed of spherical masses hanging together in linear rows which anastomose with one another. The framework, therefore, is an open structure, which holds the fluid phase in its interstices. The macroscopical result of the change is the con- version, with falling temperature, of the fluid into a loose gel. The droplets can be readily separated from the interstitial fluid by the help of a centrifugal machine. | The phenomena above described undoubtedly depend upon the separation of a homogeneous mixture into two phases owing to a fall of temperature. Each phase contains water, alcohol, and gelatine, and the system may be described as a conjugate composed of a fluid solu- tion of gelatine in a mixture of water and alcohol, and a solid solution of water with a trace of alcohol in gelatine. Both phases, however, are fluid within a small range of temperature. The surface of separa- tion of the two phases is curved, and at first discontinuous, and owing to the small size of the droplets it is very large. When the gel is heated the two phases again mix to form a clear fiuid. Owing however to the fact that the droplets adhere to one another, they tend to fuse as temperature rises, so as to form irregular masses of viscous fluid, which are separated from the other phase by a surface of no particular shape. The irregular form of this surface, and the ease with which it is modified by any chance slight currents, show that at this stage the surface tension between the two phases must be exceedingly slight. In order to simplify the description I shall call that phase which separates as small spheres the ‘‘ internal phase.” Gelation in Reversible Coilordal Systems. 97 The concentration of the gelatine in the mixture exerts a very remarkable influence upon the configuration of the hydrogel. When it is present in large quantity the internal phase is less viscous and of smaller gelatine content than the external phase, and on cooling it is the external phase which becomes a solid solution. The effect of increasing the proportion of gelatine above a certain amount is therefore very striking—it, so to speak, turns the system inside out, so that the gel is composed of a continuous framework of solid solution, out of which are hollowed spherical spaces filled with fluid. The general mechanical properties of the gel, built on this plan, naturally differ very much from those of a gel with a small proportion of gelatine, which consists of an open framework of solid holding fluid in its interstices. A mixture of gelatine, water, and alcohol is a ternary mixture which resembles a mixture of benzene, acetic acid, and water. In each there are two immiscible substances and a common solvent. The immiscible substances are gelatine and alcohol in the one case, and benzene and water in the other, while the common solvent is water in the former and acetic acid in the latter case. In both systems the solubility of the immiscible substances in the common solvent varies widely. Thus acetic acid and water, and water and alcohol, mix readily with rise of temperature; while acetic acid and benzene and water and gelatine mix freely only when the temperature is above 15° in the former case, and above 40° in the latter case. Duclaux’s researches* show that in ternary mixtures having this last charac- teristic the distribution of the constituents in the two phases varies widely with variations in the composition of the whole mass. These different characters are illustrated by the following figures, which give the amount of gelatine present in grammes per 100 c.c. They are, however, only approximate for the solid phase, owing to the difficulty in separating it completely from the fluid phase. Total mixture. Internal phase. External phase. ee 6°7 17-0 2°0 13°5 18:0 5D 36°D 8°5 40-0 The temperature at which the internal and external phases in this ternary system mix was found to be altered by altering the ratio of the masses of the components. Increasing the proportion of either of the two immiscible components, alcohol or gelatine, was found to raise the temperature, while an increase in the proportion of the common solvent water was found to lower it. The curvature of the surface which separates the phases was found * * ‘Ann, de Chim. et de Phys., série 5, tome 7, 1876, p. 264. 98 Mr. W. B. Hardy. . On the Mechanism of not to be constant for a given mixture. The internal phase formed droplets which were large or small according to whether the mixture was cooled slowly or rapidly. Thus with a mixture containing _ 13°5 grammes gelatine per 100 c.c. the droplets (of solid solution) were very regularly 3u in diameter when about 20 c.c. was allowed to cool slowly in air. Cooled rapidly, however, in an ether spray, the drop- lets were so minute as barely to be visible with a magnification of 400 diameters. The effect of the rate of cooling is the same when mixtures with a large gelatine content are used, and when, therefore, the internal phase is a fluid solution at ordinary temperatures. When cooling is very rapid the droplets are excessively minute ; when it is slow they may be as large as 10 in diameter (gelatine 36°5 per cent. of the mixture). One can therefore make the general statement that the more slowly the division into two phases occurs the smaller and less curved as the surface of separation. The effect upon the structure of the rate at which a fresh condition is imposed upon the system is manifested in a very striking way when an already formed gel is cooled. The experiments upon the effect of temperature on the composition of the two phases in the case of the hydrogel of agar show that when heat is added to or taken away from the system the balance of the phases is altered, water, and perhaps agar, passing from the one to the other. It might be ex- pected that this would take place solely by the passage of material across the surface which separates the two phases. The study of the ternary mixtures, however, makes it clear that a new approximate equilibrium may be reached in two distinct ways. When a portion of the hydrogel of gelatine-water-alcohol is cooled slowly from 16° to, say, 3° or 4°, one can see with the micro- scope no change beyond an alteration in the size of the droplets already present, that is to say, the fresh (approximate) equilibrium is attained by exchange across the surface which separates the phases. But if the cooling is rapid, say a fall of 10° in a few minutes, a second- ary system of small droplets appears. In all the mixtures which I examined these were formed in the external phase. Thus, when the concentration of gelatine in the whole mass was low, it was the fluid phase which underwent a division into secondary phases; when it was high, it was the solid phase. To put this fact in a general way, one can say that when the hydrogel 1s exposed toa rapid fall of temperature the phase which les on the convex side of the surface of separation undergoes division into two secondary phases.* When the temperature is again allowed to rise these secondary phases fuse before there is any obvious change in the relation of the primary phases. * The formation of the secondary phases therefore occurs in that one of the primary phases which is under the lower hydrostatic pressure. —Gelation in Reversible Colloidal Systems. H99 When once formed the phases have considerable stability. If the droplets are composed of a solid solution one may, by the addition of water, cause them to increase to relatively vast dimensions without their being destroyed, as they increase in size their refractive Index approximates more and more to that of the external phase until finally they are lost sight of. The addition of alcohol, however, once more brings them into view and causes them to shrink. Owing to this stability once a configuration is established one has to far overstep the conditions of its formation in order to destroy it. This would account for the remarkable hysteresis observed in reversible gels. Thus a 10 per cent. solution of gelatine in water sets at 21° and melts again at 29-6", and solutions of agar in water set at temperatures about 35° and melt at temperatures about 90°. Similarly with the ternary mixtures. In one holding about 35 per cent. gelatine, the internal and external phases separate at 20°, but they mix again only at 65°. When water is added to a ternary mixture so as considerably to swell the droplets the system is unstable, and the two phases mix at once when it is mechanically agitated. The properties of the ternary system: alcohol, gelatine, and water are the following :— i. Below a certain temperature it exists in two phases separated. by a well-defined surface. The temperature at which the separa- tion occurs depends upon the relative proportion of the com- ponents in the mixture. Increasing the proportion of gelatine raises it; as does also an increase in the proportion of alcohol. An increase in the proportion of the common solvent, water, however, lowers the temperature at which the biphasic cha- racter develops. | ii. Both phases are at first fluid; with further fall in temperature one becomes solid. ) iii. The surface of separation is curved and discontinuous. In some ~ cases, strictly as a secondary change, the discontinuous masses of the internal phase become continuous with one another. — iv. The more slowly the two phases are established the less is the surface which separates them both in extent and in curva- ture. : v. The solid solution phase is formed sometimes on the concave, sometimes on the convex side of the surface of separation. The former happens when the proportion of gelatine is small, the latter when it is large. It follows from the last (v) of these properties that a hydrogel may be built on two very different plans. It may consist of a solid mass ‘containing spherical fluid droplets, or of solid droplets which, by hanging one to the other, form a framework in the spaces of which 100 Mr. W, B. Hardy. On the Mechanism of fluid is held. These two types present important mechanical pecu- liarities. The former is firm and elastic, and it maintains its structural integrity even under high pressure. The latter is much more brittle, and manifests a tendency to spontaneous shrinking, which is due to a continuous increase in the surface of contact or possibly union between droplet and droplet. These gels with an open solid framework there- fore specially manifest that property of spontaneous shrinkage to which Graham applied the term “syneresis.” In the building of a hydrogel of the second type two distinct events occur. ‘The first is the separation of droplets, which rapidly become solid ; the second is the linking of these droplets together to a pattern - go that they build a framework throughout the fluid phase. The first is the separation of a homogeneous mass into two phases ; the second is a phenomenon akin to the grouping of particles which are suspended in a fluid. Itis probable that these two events are not directly connected with one another. Binary Mixtures (A gar-water). I know of no binary reversible system in which the optical character's of the two phases differ sufficiently to permit of direct microscopical investigation of the surface of separation. It is, however, easy to prove that in such a system as agar and water the property of gel building is dependent upon the appearance of two phases. The agar which was used was prepared from commercial agar as follows. The strips were suspended in a large volume of distilled water for twenty-four hours ; the water was then drained off, and a large part of the water absorbed by the strips was squeezed out by a powerful press. The strips were again suspended in distilled water, and again drained and squeezed after forty-eight hours. This washing with distilled water was continued for some weeks. The strips were then melted and the hydrosol filtered, and the filtrate allowed to set. The clear hydrogel so obtained was sliced and suspended for a further period of weeks in many changes.of distilled water. In this way a colourless gel was obtained free from all foreign diffusible bodies. It was not found necessary to take precautions against micro-organisms. With the removal of the salts the agar ceased to afford them a suitable nidus. Effects of Pressure upon the Hydrogel of Agar. When gels containing 1 to 3 per cent. solids* are broken up and slightly squeezed by hand a fluid exudes. In order to collect this fluid a screw press was made use of. The gel was cut into pieces, which were wrapped in fine cotton canvas which had been completely freed * By this is meant 1 to 3 grammes per 100 grammes. Gelation in Reversible Colloidal Systems. 101 from soluble substances by treatment for months with hot and cold distilled water. The packet was then pressed in a screw press, and the large yield of fluid collected. When the fluid ceased to flow, the solid which remained in the canvas was removed to a stoppered vessel. The fluid was found to be a solution of agar. This was proved by evaporating some after it had been thrice filtered to a small bulk, when it was found to set to a typical clear gel. The results of the study of the ternary systems give sufficient grounds for defining the expressed fluid as a solution of agar in water, and the solid which remains in the canvas a solution of water in agar. The effect of the composition of the gel and of temperature upon the distribution of the water and the agar in these two phases was determined. The percentage composition was arrived at by drying a known weight of each, and assuming that the residue was entirely composed of agar. The results which were obtained lie far outside of any error which could have been introduced by this assumption when one considers the pains which were taken to free the gels from foreign bodies. Further, in every case an examina- tion of the dry residue was made in order to prove that it was composed of matter capable of forming with water a typical agar gel. Kaperimental Difficulties. The method used to separate the two phases, though at first sight crude, was found to be the most effective. The great error to be avoided is the blocking of the canvas pores by a mass of the solid phase, so that, instead of the true fluid phase, one really expresses fluid which has been forced through a membrane (pressure filtered). Owing to this error, a press, which I had specially made, and in which the piston drove the gel directly down on to a disc of canvas, proved quite useless. Very great force was necessary to express a fluid which was found to be almost pure water. To succeed, it is necessary to avoid direct or great pressure. The masses of gel are loosely placed in a long canvas packet, which is then deformed by pressing the ends together. The pressure necessary to yield abundant fluid is now quite small, for the solid framework of the gel is destroyed by being rubbed against the canvas, and is reduced to fine particles, while the fluid easily makes its way through the coarse pores of the canvas. Raising the pressure always expresses a fluid poor in agar, while with slight to moderate pressure the concentration of the expressed fluid, as tested by determining the solids in the yield at different stages, remained fairly constant, but always with a slight decrease as time went on. The expressed fiuid was filtered before the solids were estimated ; this was found to lower the amount of solid to a very slight extent. 102 Mr. W. B. Hardy. On the Mechanism of The following figures illustrate the variation in the composition of the fluid as the gel becomes more completely expressed :— Successive equal: quantities of Repvennad Fluid contain Dry — in 100%¢:¢,° T= ee Experiment I. Experiment IT. 0°12 O-1l 0-14 0-12 0-1 0:09 The mechanical pressure used to separate the phases will modify their composition by deforming the surface of separation. This error cannot be estimated. The Influence of the Ratio of the Masses of the Two Components upon the Composition of the Phases. Two portions of a fairly concentrated gel were taken. To one part water was added to dilute it, and both were then heated to 100°, and equal portions of each were poured into two glass stoppered cylindrical vessels of identical shape, make, and size. The two vessels were then set aside to cool. After forty-eight hours samples were cut from different levels in each gel, and used to determine the percentage composition. Five hundred and eighty grammes of each of the gels were then expressed. The results were as follows :— ; Expressed fluid. Solid solution. Agar in 100 grams oe cn eieel: Volume. Agar. Volume. Agar. grammes. c.c. per cent. C.C. per cent. cas, Jk 440 Orl 140 A “7 3°3 230 0°14 390 5 °6 In another experiment— T. =.135° 1°6 a 0°12 — a 2°2 — 0°14 == — Thus an increase of the concentration of agar in the mixture pro- duces an increase in the concentration of the agar in both phases. An explanation of this relation is suggested, and discussed later. — Gelation in Reversible Colloidal Systems. 103 Effect of Temperature upon the Composition of the Phases. This was determined by running a large mass of the hydrosol into a number of glass vessels of the same shape and size. Each vessel held 600 ¢.c. of the hydrosol. They were close stoppered and allowed ‘to cool to the room temperature. After forty-eight hours they were placed in chambers of known temperature, where they were kept for five to seven days before the contents were subjected to pressure. In these experiments, as is obvious, the internal changes are those which follow on raising or lowering the temperature of the hydrogel from the air temperature. In other experiments the hydrosol was cooled down to, but not below, the temperature of observation. This dis- tinction is important, because it was found that the composition of the phases varied for a given temperature according to whether that tem- perature was the lowest of a descending series or the highest of an ascending series. ‘This is shown clearly in the two curves AB AB, fig. 1. The arrows indicate the direction, ascending or descending, of the changes of temperature. No. I.—Agar content of Mixture 1°6 per cent. Agar in Agar in HE CapeLaEate expressed fluid. solid. per cent. per cent. 14° 0°14 == Ascending series eee ee eee 33 8) *29 rae 50 0°80 rae ; ; 14 0°14 == Descending series ........ 1 33 1-10 Fas No. IJ.—Agar content of Mixture 2°23 per cent. a Agar in Agar in See expressed fluid. solid. 4 per cent. per cent. ; : 13 0°12 4°77 Ascending series ......... { 36 Q-25 5-0 ; 5 0:09 3°0 Descending series........ 18 0°12 4-7 36 0°47 3-2 Putting on one side for a moment the different effect of an ascending or a descending temperature change, these experiments show that (1) a hydrogel of agar is a structure form of a more solid part anda fluid, and (2) each of these two phases is a mixture of agar and water, 104 Mr. W. B. Hardy. On the Mechanism of (3) the composition of the phases is dependent to a lesser degree upon the ratio of agar to water in the entire mass, to a greater degree upon the temperature. | | While recognising as fully as possible that only an approximation to the actual composition of the two phases at different temperatures is obtained by these experiments, it is obvious that they afford reliable information on two points. These are, firstly, the marked “lagging” action or passive resistance to change offered by the system agar-water. The difference in composition of the phases according to whether any given temperature lies in an ascending or a descending series shows how slow the system is in reaching final equilibrium.* Secondly, the experimental results seem to me to indicate pretty clearly the general form of a part of the concentration temperature curve. I give the curve as it appears from the figures in Experiment I]. AB and CD are the curves for the system—solution of agar in water, solution of water in agar, and vapour. If they correctly represent the general form of the curve, then, by the theorem of Le Chatelier, it follows that the change from the system solution of water in agar and vapour to the system solution of water in agar, solution of agar in water, and vapour will be accompanied by a liberation of heat when the change takes place along the isotherms from 5° to + 20°, and by an absorption of heat when the change is along the isotherms + 20° to 35°, while the change from the system solution of agar in water and vapour to the system of two solutions and vapour will always be accompanied by absorption of heat. I have not established this deduction experimentally, but it finds a considerable amount of support in the following facts. When water is allowed to dissolve in pure dry agar at 14°, a considerable amount of heat is given off. 1 c¢.c. of dry agar in coarse powder added to 10 c.c. of water gives a rise of more than 6°, while control experiments with carefully dried finest graphite or sand gave a rise of temperature of 0-15° and 0:17° respectively. Wiedemann and Liidekingt also found * The systems salicylic acid and water, and thorium sulphate and water are perhaps comparable cases. The former readily yields two fluid phases which, however, are throughout in labile equilibrium (Bancroft, ‘The Phase Rule,’ p. 105). In the case of the latter system supersaturated solutions can be obtained over a wide range of temperatures, and even in presence of the stable hydrates it is often hours or days before equilibrium is reached (Bancroft, loc..cit., p. 54, or Roozebrom, ‘ Zeits. f. Phys. Chem,’ vol. 5, 1890, p. 198). The lagging action in colloids which is so markedly shown by van Bemmelen’s researches into the effect of time on the hydrogel of silicic acid, ceases to be extraordinary when one remembers that one of the phases is a solid. Gels reach equilibrium much more rapidly than does, for instance, a bar of metal in which the reaction velocity is so slow that final equilibrium miay never be reached, + The mercury in the Beckmann thermometer was driven beyond the scale into the upper reservoir, ' t ‘Ann. der Phys. u. Chem.,’ N.F., vol. 25, 1885, p. 145. Gelation in Reversible Colloidal Systems. 105 that when dried gelatine absorbs water heat is liberated, but when gelatine saturated with water is dissolved in water heat is absorbed. I have verified the general form of the curve AB in a way which eliminates all the errors due to the expression of the fluid phase from the gel. A cylindrical column of gel 15 cm. high was divided by two vertical cuts at right angles into four equal pieces. Four stoppered glass vessels were taken of the same size and shape, and in each one of the pieces was placed and just covered with water. Two of the bottles were kept at 14° for a week, and two at 44°; the water in both was found to have dissolved some of the agar, and to {140100 99 98 97 96 95 S54 contain per 100 grammes of the solution 0°50 gramme and 0°12 gramme of dry agar respectively. The curves AB, DC continued upwards will meet at some point which marks the consolute temperature for agar and water. I have attempted to fix this point by observing the changes in the intensity of the beam of polarised light scattered normal to the ray when parallel light is passed through a gradually cooling hydrosol. The observa- tions, which are still in an initial stage, have so far failed to fix the point. | The study of ternary systems under the microscope makes it probable that as the curves AB, DC are continued. upward they reach a point 106 Mr. W. B. Hardy. On the Mechanism of beyond which the equilibrium is no longer between a fluid solution and a solid solution, but between two fluid solutions. The first worker to regard gelation as being due to the formation of two phases, one fluid and the other solid, was van Bemmelen.* He has given a suggestive discussion of the formation and structure of gels, based chiefly upon the manner in which amorphous material is precipitated from a solution, and he is led to the conclusion that “ coagulation or the precipitation of a gel from a solution seems to be a similar phenomenon ; a desolution (Entmischung) which forms, not two layers completely separated from one another, but ©], A framework of a material which is in a more or less transitional state between fluid and solid, and which presents. those special properties to which the term colloid is applied. “9. A fluid which is enclosed within this framework.” Van Bemmelen, however, does not consider that these two parts can be considered as two phases in the sense of the phase rule, since there is no sharp line between them,t and he therefore concludes that the phase rule cannot be employed to elucidate the phenomena. ‘This opinion is based upon a study of the equilibrium between the water content of various gels and the vapour pressure, so patent and thorough as to give it very great weight. The curves of the equi- librium points are gradually bending lines if the dehydration of the gels is sufficiently slow, but if dehydration is relatively rapid there is a sudden change of direction (fig. 2) when the water content is very much diminished (1 to 2H gO to 18,02). It is possible that the tote of these curves does not necessarily depend upon the absence of a clear separation between the fluid and the solid portions of the gel. When one considers how small is the mutual solubility of silica and water and how slight therefore the in- fluence which a given mass of silica is likely to exert upon the vapour pressure of even a relatively small mass of water, it is probable that the form of the curve is determined more by the operation of seeondary influences, such as capillary tension, which depend on the structure of the gel, than upon the direct interaction of silica and water: Capillary tension would tend to lower the vapour pressure} with which the gel is in equilibrium to a greater and greater extent as the spaces in the solid framework of the gel became smaller and smaller with the decrease in the water content. The tendency'to reduce the surface * ‘Zeits. f. Anorg. Os vol, 18, 1898, p. 20. + « Zeits. f. Anorg. Chem.,’ vol. 18, 1898, p. 121. + The vapour pressure which van Bemmelen measured is that of the free surface of the gel. It is ameliee er to baci at the open ends of a number of capillary tubes filled with fluid. . Gelation in Reversible Colloidal Systems. 107 energy at the surface of separation of fluid and solid to a minimum, which manifests itself in the spontaneous shrinkage of some of these hydrogels, would act so as to raise the vapour pressure with which the gel is in equilibrium, but the operation of this factor would diminish as the surface was diminished by decrease of the water content. These two forces operating simultaneously would alone produce the character- istic gradual diminution in the vapour pressure of the gel as the fluid component is diminished. The break in the direction of the curve Fie. 2 (reproduced from van Bemmelen, ‘Zeits. f. Anorg. Chem.’ vol. 18, 1896, p- 233).—Equilibrium between a Hydrogel of Silica and Water Vapour. 8 re of walter. ® S 1) Vapour pressu N N EY a. Tee ey yA 2) 21 Mols.H,0 bo l3iO, | Curve - - - -- the rate of removal of water very slow. In curve ———— much more rapid. The arrows indicate whether the curve shows the removal or the reabsorption of water. when dehydration has been relatively rapid and is nearly complete, is what must occur when the capillary spaces in the framework become commensurate with the masses (small spheres for instance) of solid out of which the framework is built, and when, therefore, any further diminution in the capillary spaces involves deformation of those masses,, unless the removal of water is so slow that the very slow rate of readjustment in the solid phase is not exceeded. Lastly, the very. limited powers of reabsorption of fluid by «completely dried irrever- sible gels would, on this view, again not necessarily represent a reformation of the phases, that is to say, a real interaction between silica and water, but the refilling of capillary spaces by water due to the-excessive capillary tension of these very minute capillaries. The capacity for reabsorption would therefore be diminished by any agent which facilitates the annealing of the dried gel and so destroys the capillary interspaces. Such an agent is heat, and van Bemmelen found 108 Mr. W. B. Hardy.- On the Mechanism of that brief heating to red heat destroyed the reabsorptive powers of the gel of silica.* There is one binary system in which gelation is an irreversible process (7.¢., coagulation) which can be readily studied under the microscope. The hydrosol is a ternary system composed of water, a minute trace of free acid or alkali, and the modification of egg albumen which is produced by heating it to 100°. Coagulation occurs when the free acid or alkali is removed. As the coagulation point is neared the proteid particles in the hydrosol increase in size, so that spheres 0°75 to 1p in diameter are formed. These become arranged in rows which anastomose so that an open net with regular polygonal meshes is formed.t In this case the process of gel building is the same as that which can be followed so easily in ternary mixtures, and in both cases a definite surface separates the phases. It is probable that the hydrogel of silica is formed in the same way, since Picton and Linder have shown by optical tests that, as the point of coagulation is approached, larger and larger particles of silica form in the hydro- sol.t These particles may be solid solutions of water in silica, or they may-be large molecular aggregates of silica free from water. I incline to think that the latter is the more probable assumption, since, if they were solutions, it is difficult to see why the process should be irreversible. ? In the case of the reversible systems agar-water, or gelatine-water- alcohol, the particles seem to be of the nature of solid solutions. The system agar and water consists of two components, and, therefore, a nonvariant system should be defined by four coexistent phases. Since the gel stage consists of three phases, namely two solutions and a vapour phase, it should be a monovariant§ system. That is to say, the composition of the phases should be fixed by fixing either the temperature or the pressure of the vapour phase. The experiments show that this is not the case. The composition of the fluid and solid phases is not constant for a given temperature. This result might be regarded as being due to the passive resistance to change in the system which is introduced by the formation of a solid phase. On this view if the velocity of the reaction were known, the phases would be fixed if the element of time were introduced and accorded a definite finite value. This is the method which Bancroft sug- gests for dealing with such cases ||| it is, however, possible that there are * ‘ Zeits. f. Anorg. Chem.,’ vol. 13, 1896, p. 289. + The process is described in detail in an earlier paper by the author in the ‘Journal of Physiology,’ vol. 24, 1899, p. 182, and the information which the microscope affords as to the manner in which irreversible gels are built is discussed there. t ‘Journal of the Chemical Society,’ vol. 61, 1892, p. 148. § That is to say, a system having one degree of freedom. || ‘The Phase Rule,’ p. 234. — Gelation in Reversible Colloidal Systems. 109 really more than n+2 independent variables, so that the hydrogel is not a monovariant system. In an ordinary system the independent variables are the components (n), temperature and pressure. Agar- water, however, is a system with two components, temperature and two pressures. This follows from the fact that the surface which separates the fluid and solid phases is curved. In point of fact the system is most closely represented by a system of two solutions separated by a membrane which is permeable by only one of the com- ponents, for while water will readily pass the surface of separation, agar, having the heavy immobile molecule characteristic of such organic bodies, will be almost unable to do so. Hence, if time be con- sidered finite and small, the surface may practically be considered to be permeable by only one component. As Bancroft* points out, in a system of two solutions separated by semipermeable membrane, there are two pressures and there will be n+3 phases in a nonvariant system when » = the number of components. The hydrogel is a system of three phases and, therefore, on this view, to fix the com- position, it would be necessary to fix the temperature and one pres- sure. This relation would probably be true if the curvature of the surface of separation could be fixed. This, however, is not the case, and in order to fix the composition of the phases it would be necessary either to fix the temperature and both pressures, that of the internal _as well as of the external phase ; or to fix the temperature, one pressure, and the form of the surface. Practically we can only fix the tempera- ‘ture and the pressure of the external phase. I have succeeded in obtaining two phases separated by a plane surface by cooling a hydro- sol slowly in an electric field. This method may prove suitable if the ‘system is able to recover from the forces operating during its forma- tion. The method of taking known weights of dry agar and water and keeping them at constant pressure and temperature until equili- brium is obtained is simple, but unfortunately there is the fallacy that the dry agar is a preformed system. The structure of the hydrogel from which it is reproduced is not destroyed by drying, and the system tends to reform itself on the old lines by the filling of the original capillary spaces. To sum up these remarks, we may describe the hydrogel of agar as a system of three phases, a solid, a fluid, and a vapour phase, The equi- librium is determined by the chemical potential of the components in the various phases, by two pressures, and by temperature. Other operating variables are capillary tension and the energy of the surface between the fluid and solid phases. I have made no measurements to determine how soon the system reaches equilibrium, but the analogous system, gelatine and water, attains to a constant melting point twenty- four hours after the formation of the hydrogel.t * Loc. cit. + ‘Gela‘intése Lésungen,’ van der Heide, Miinchen, 1897. VOl« LXVL P 110 Mr. W. B. Hardy. On the Conditions which “A Preliminary Investigation of the Conditions which determine the Stability of Irreversible Hydrosols.” By W. B. Harpy, Fellow of Gonville and Caius College, Cambridge. Com- municated by F. H. NEVILLE, F.R.S. Received January 12,— Read January 25, 1900. It has long been held that a large number of colloidal solutions are related to or identical with suspensions of solid matter in a fluid in which the particles of solid are so small as to settle at an infinitely slow rate. Such solutions are the colloidal solutions of metals and of sulphides such as those of antimony, arsenic, and cadmium. Such solutions belong to the class of irreversible colloidal mixtures. A rise of temperature assists the process of coagulation or precipitation ;* but neither a further rise nor a fall of temperature will cause the reformation of the hydrosol. On this ground they may provisionally be classed with such colloidal solutions as those of silica, ferric hydrate, alumina, &c., and with the modification of the albumen of white of egg which is produced by heating an aqueous solution to the boilmg point. I also add to the class, for reasons to be developed in the following pages, the suspension of mastic in water which is produced by adding a dilute alcoholic solution of the gum to water. Looked at from the point of view of the phase rule, the equilibrium in these hydrosols, if they really consist of minute solid particles dis- persed in a fluid, is not necessarily between the solid particle and water, but between the solid particle and a solution of the particular solid in water. The hydrosol of gum mastic gives off a vapour of the gum of a density sufficient to affect the olfactory organs, and, therefore, the water must contain a definite quantity in solution. Similarly, as it is probable that no substance is completely insoluble, we may assume that in all the examples a portion of the solid is in true solution in the fluid. As the solid which is not in true solution is dispersed in particles whose diameter is, as a rule, very much smaller than the mean wave length of light, it follows that the surface of contact between solid and fluid is very great for unit mass of the former. The opportunity for evaporation and condensation of the solid matter of the particles afforded by the immense surface of contact is so very great that, although only an immeasurably minute quantity of the solid may be in true solution at any one time, this quantity, minute though it be, is probably an important factor in determining the equilibrium between solid and fluid. * Elsewhere (‘Journal of Physiology,’ vol. 24, 1899, p. 172) I have shown that precipitation and coagulation are not discontinuous processes. Coagulation gives way to precipitation when the concentration of the solid phase falls below a certain amount. deternvine the Stability of Irreversible Hydrosols. Pt It is necessary to keep such considerations as these in mind in view of the readiness with which these mixtures have been regarded as simple suspensions* in which the only relation between solid and fluid is a mechanical one. These hydrosols are, as a matter of fact, singularly stable when pure. They can, for instance, be concentrated by boiling to a remarkable extent, and their stability depends upon complex relations between fluid and solid, which gives the former, so to speak, a definite hold over the latter. ! Mode of Preparation of the Different Solutions —The hydrosol of gold was prepared by adding a couple of drops of a solution of phosphorus in ether to about a litre of a very dilute solution of gold chloride. The fine ruby-coloured fluid which was formed was dialysed against distilled watert for fourteen days, and then concentrated by pe: The hydrosol of silicic acid was prepared by acting on soluble glass with excess of hydrochloric acid, and dialysing the product. A hydrosol of ferric hydrate was prepared by prolonged dialysis of the solution in ferric chloride. The hydrosol of gum mastic was prepared by adding a very dilute solution of the gum in alcohol to distilled water. It was dialysed for fourteen days against distilled water. The hydrosol of heat-modified egg-white was prepared by dissolving white of egg in nine times its volume of distilled water, filtering and boiling. The result should be a brilliant fluid which scatters blue light. Surface action, however, plays an extraordinary part. If the solution is boiled in a test-tube a milky fluid is formed and a film of proteid is left on the glass; a second quantity boiled in the same test-tube comes out less milky, until, when the proteid film is sufficiently thick to eliminate all action by the glass, the solution after boiling contains the proteid dispersed as particles so small that they scatter pure blue light. After pre- paration the hydrosol was ae against distilled water for some days. Behamour of the Hydrosols in an Electric Field.—It has long been known that the particles in these colloidal solutions move in an electric field. Zgismondyt found that the gold in colloidal solutions moves against the current. Picton and Linder§ established the important fact that the direction of movement of the particles, as compared with the direction of the current, depends upon their chemical nature. I have shown that the heat-modified proteid is remarkable in that its direction of movement is determined by the reaction acid, or alkaline of the fluid * Cf., for instance, Stoeck! and Vanino, ‘Zeits. f. phys. Chem.,’ vol. 30, 1899, p: 98; also Ostwald, ‘ Lehrbuch.’ + In working with these colloidal solutions it is very necessary to use distilled water freed from dissolved carbonic acid. t ‘ Lieb. Ann.,’ vol. 301, p. 29. § ‘Journal of Chem. Soc.,’ vol. 70, 1897, p. 568. 112 Mr. W. B. Hardy. On the Conditions which in which it is suspended.* An immeasurably minute amount of free alkali causes the proteid particles to move against the stream, while in presence of an equally minute amount of free acid the particles move with the stream. In the one case, therefore, the particles are electro- negative, in the other they are electro-positive. Since one can take a hydrosol in which the particles are electro- negative and, by the addition of free acid, decrease their negativity, and ultimately make them electro-positive, it is clear that there exists some point at which the particles and the fluid in which they are immersed are iso-electric. | This iso-electric point is found to be one of great importance. As it is neared, the stability of the hydrosol diminishes until, at the iso- electric point, it vanishes, and coagulation or precipitation occurs, the one or the other according to whether the concentration of the proteid is high or low, and whether the iso-electric point is reached slowly or quickly, and without or with mechanical agitation. This conclusion can be verified experimentally in many ways. If a coagulum or precipitate of the proteid particles made either by the addition of a neutral salt, or by the addition of acid or alkalis, be thoroughly washed, made into a fine mud in an agate mortar, and suspended in water in a U-tube, it rapidly subsides. The establishment of an electric field having a potential gradient of 100 volts in 10 cm. has no influence on the level of water or precipitate in forty-eight hours. If, now, the smallest possible amount of caustic soda or acetic acid be added, the proteid will commence to move, so that in twenty hours the precipitate will rise in one or other limb until it nearly touches the platinum electrode. Speaking generally, the hydrosol of ferric hydrate is stable only in the absence of free acids or alkalis or neutral salts. The hydrosol of heat-modified proteid is stable only in presence of free acid or alkali. The hydrosol of gum mastic is readily. precipitated by acids, but is stable in presence of any concentration of monovalent alkalis. The general conditions of stability of these various hydrosols, therefore, are very different, yet they agree in manifesting the same important rela- tion between the isoelectric point and the point of precipitation as is shown by the hydrosol of proteid. In the hydrosol of ferric hydrate the particles are markedly electro- positive. A dilute hydrosol is coagulated by citric acid when the concentration of the latter reaches 1 gramme-molecule in 4,000,000 c.c. No matter how small the concentration of the ferric hydrate, the hydrosol becomes cloudy and settles. The rate of settling is, however, slow, being about 1 cm. an hour. In an electric field, having the form of a U-tube, the particles always settle slightly faster from the negative * “The Coagulation of Proteid by Electricity,’ W. B. Hardy, reece of Physiology, vol. 24, 1899, p. 288. determine the Stability of Irreversible Hydrosols. 113 electrode—the acceleration due to the electric field being about 5 mm. an hour. The suspended particles of ferric hydrate show, therefore, an exceedingly slight movement in a direction opposite to that which they manifest when in colloidal solution. In the latter condition they are markedly electro-positive ; in the former they are exceedingly faintly electronegative. An exceedingly faint electro-negative character is also conferred upon the ferric hydrate when the hydrosol is coagulated by ammonia, 1 gramme-molecule of the latter being present in 100,000 c.c. If a fresh gel of silica is broken up in distilled water and carefully washed to free it from still uncoagulated silica, and from impurities, it is completely iso-electric with the water. It becomes markedly electro- negative, however, on the addition of the minutest trace of free alkali. Gum mastic precipitated from a dilute hydrosol by adding barium chloride until the concentration is 1 gramme-molecule in 600,000 cm. is found to be iso-electric with the fluid. It is markedly electro-negative when in colloidal: solution. Picton and Linder have shown that the particles in these hydrosols gradually grow in size as the coagulation or precipitation point is neared.* It might, therefore, be urged that, as the movement of the particles in the electric field is, on Quincke’s theory of electric endosmose, due to surface action, the fact that they do not move when in simple suspension as opposed to colloidal solution may be due to the diminution of the impelling force acting on a given volume.t This is, however, negatived by the character of the experiments. The addition of a minute amount of free alkali to a mass of particles of coagulated silica which have settled to form a “mud” cannot alter the size of these relatively very large masses to any appreciable extent. And since in the case of ferric hydrate and proteid, the sign _ of the charge which the particles carry in the electric field is different on each side of the actual point of precipitation, that point must of necessity be an iso-electric point. If the stability of the hydrosol is dependent upon a difference in electrical potential between the solid particles and the fluid, then one would expect that for, at any rate a short distance from the iso-electric point, the stability would vary simultaneously with the variation in the difference of potential. The experimental investigation of this question is beset by many difficulties. At present | know of no way of approaching the iso-electric point other than by the addition of salts, * ‘Journ. of Chem. Soc.,’ vol. 61, 1892, p. 148. + As a matter of fact, Lamb finds that the velocity of a partic’e is independent of its size or shape, provided that its dimensions are large compared with the slip, so perhaps the objection scarcely needs discussion. Lamb, ‘ Brit. Assoc. Report,’ 1887, p. 502. 114 Mr. W. B. Hardy. On the Conditions which acids, or alkalis. One may, therefore, approach the point by the addi- tion of, say, acid or alkali, and use a salt to measure the stability of the system, as in the experiment described later. In such experiments, however, the colloid particles are immersed in a complicated system of three components, the conditions of equilibrium of which cannot be arrived at from existing data. The conditions could be simplified by using, say, KHO or H»SO, to approach the iso-electric point, and K,SO, as a measure of the change of stability. A series of determi- nations with different systems of this kind may afford the requisite measurements. A direct and conclusive proof that stability does decrease as the iso-electric point is approached was however obtained in two ways. The iso-electric point can be approached in the case of the hydrosol of proteid _ by the withdrawal of either the free acid or the free alkali, as the case may be. As it is neared, the proteid particles increase in size, so that, instead of scattering blue light, they scatter white light; thus the sur- face of contact of fluid and solid gradually diminishes as the point is neared. The second experiment, though not a quantitative one, is very convincing. A hydrosol of gum mastic dialysed as pure as possible is not destroyed by mechanical agitation even when long continued. UH, however, a salt is added in an amount so small that it just fails to coagulate the hydrosol, the latter is rendered so unstable that it is destroyed by shaking. Experiments were made to determine whether the particles aie carry a charge. An electric field which was practically uniform was made by using flat electrodes of the same size, which were placed parallel to one another at the ends of a straight tube. The particles were found to move in all parts of the field; they therefore carry a definite charge which, according to Quincke’s theory of the movement of particles in an electric field, would be a surface charge, each particle being surrounded by a double layer of electricity. : Action of Salts——The power possessed by salts of destroying colloidal - solutions was noticed by Graham. The subject was, however, first accurately investigated by H. Schulze.* He showed that the power which various salts possess of precipitating a hydrosol of sulphide of arsenic is related to the valency of the metal, while the valency of the acid has little influence. The increase in the precipitating or coagu- lating power produced by increase in valency is very great. If coagulative power be defined as the inverse of the concentration in gramme-molecules per litre necessary to convert a given hydrosol into a hydrogel, then from Schulze’s measurements the coagulative power of metals of different valency is :— R 2 Roo RR” 21:30" 650: * ‘Journ. f. prakt. Chemie,’ vol. 25, 1882, p. 431. determine the Stability of Irreversible Hydrosols. 115 Schulze’s conclusions were verified by Prost,* who used sulphide of cadmium, and Picton and Linder, who used the sulphide of antimony.t The last-named wofkers added the important fact that a small portion of the coagulating salt is decomposed, the metal being entangled in the coagulum. The measurements which I have made with various colloidal solu- tions both confirm Schulze’s results, and bring out a new relation, which may be stated as follows :— The coagulative power of a salt is determined by the valency of one of its ions. This prepotent ion is erther the negative or the positive ion, according to whether the colloidal particles move down or up the potential gradient. The coagulating ion is always of the opposite electrical sign to the particle. The salts employed to determine this point were the sulphates of aluminium, copper, magnesium, potassium, and sodium; the chlorides of copper, barium, calcium and sodium, and the nitrate of cadmium. Solutions containing 1 gramme-molecule in 2000 c.c. were prepared. The experiments may be’summarised as follows :— Silica dialysed free from Chlorides, Electro-negative. Concentration of Coagulating Salt 1 gramme-mol. in 120,000 c.c. Temperature 16°. Coagulated at once. InlOmins. In2hours, I1n24hours. Still fluid. Al,(SO,)s CusO, MgSO, K,SO, NaCl CuCl, Na,SO4 Control. Cd(NO,)2 BaCl, This illustrates many experiments. Proteid in presence of trace of Alkali, Electro-negative. Temperature 16°. Coagulating Salt 1 gramme-mol. in 80,000 c.c. Coagulated at once. On slightly warming. Did not coagulate. Al,(SO,)3 MgSO, Na SO, Cd(NOs), BaCly K,SO, CuSO, CaCl, NaCl CuCl, Proteid in presence of trace of Acetic Acid, Electro-positive. Coagulated instantly. No effect. Al.SOx CuCl, CuSO, Cd(NOs)s K,SO,4 BaCl, Na SO. | NaCl MgSO, * ‘Bull. de l’Acad. Roy. de Sci. de Belg.,’ ser. 3, vol. 14, 1887, p. 312. _ + ‘Journ. Chem. Soc.,’ vol. 67, 1895, p. 63. 116. - Mr. W. B. Hardy. On the Conditions which Mastic, Dialysed, Neutral, Electro-negative. Temperature 16°, Concentration of Coagulating Salts, 1 gramme-mol. in 50,000. Coagulates at once, No coagulation. Alx(SOx4)3 K»SO. CuSO, Na SO, CuCl, NaCl Cd(NOs)2 MgSO, BaCl, Ferric Hydrate, Dialysed, Neutral, Electro-positive. Temperature 16°. Coagulating Salt 1 gramme-mol. in 100,000. Coagulates at once, Does not coagulate. Alx(SOx)3 CuCl, CuSO, Cd(NO,)s MgSO, NaCl K.SO4 BaCl, Na SO, Gold, Dialysed for fourteen days against Distilled Water, very faintly Acid. Hlectro-negative. Temperature 16°. Coagulating Salt 1 gramme-mol. in 200,000. Red changes to blue* instantly in— No change. Al,(SO,), NaCl CuSO, Na.SO, CuCl, K,S0O, Cd(NOs)2 MgSO, BaCl, Only one comment on these experiments is needed. Solutions of Al,(SO4)s, Cd(NO3)s, CuCls, and CuSOu,, are acid to litmus, while MgSO,, and BaCl, are neutral to litmus, but acid to phenol phthalein. This acidity has a disturbing action in some cases—the system acts not only as a neutral salt, but also as a free acid. Thus the hydrosol of proteid when brought very near to the point of precipitation by dialysis is more sensitive to the more acid than to the less acid salts of the bivalent metals. The effect of the acid or basic reaction of the * The relation of the colours of hydrosols of gold to the size of the particles has been investigated by Stoeckl and Vanino (‘Zeits. f. phys. Chem.,’ vol. 30, 1899, p. 98). The change from red to blue indicates an increase in the size of the particles. determine the Stability of Irreversible Hydrosols. 117 salt on the hydrosol is as a rule small compared with the effect of the metal ion. Thus the stability of a hydrosol of electro-positive proteid is increased by free acid, yet the acid salts find their proper place in the scale of valency. Again, ferric hydrate is coagulated by nitric acid when the concentration reaches 1 gramme-mol. in 2500 ¢.c.; yet the cadmium salt of this acid is not much more potent than the “ none is salts MgSO,, BaCl.,. Temperature 16°, Concentration necessary to Coagulate Ferric Hydrate. Salt. a. gun cas 1 gramme-mol. in 4,000,000 c.c. BigsO,° ...... 2 8 4,000,000 ,, Bas «5.... 0 = a 10,000 ,, RD sca. oi # i 30,000 *,, Cd(NOs3)2 ... Z a3 50,000 ,, The extraordinary rise in coagulative power with an increase in valency, which was observed by Schulze, Prost, and Picton and Linder, holds in all cases. In order to measure it for ferric hydrate, I used Schulze’s method, in which a drop of the hydrosol is allowed to fall into a large volume of the solution of the salt. A number of experiments were made until the concentration of the salt was found which just sufficed to coagulate the drop. In the case of gold and mastic the process was reversed, the salt solution being added drop by drop to a measured quantity of the hydrosol. J append the results :--- Gum Mastic, Neutral. Temperature 40°. MEMO), win cine von 5 03 1 gramme-mol. in 86,000 c.c. MBO cel. - S000: Gs Oe Fe zt 68,000 ,, |: Ce, 0 eee +3 53 8,000 ,, Gold, very faintly acid. Temperature 16 ONAC so. sk .es-- 1 gramme-mol. in 72,000 e:c. Ae bee a , 500,000, SS) Orne i ar (0,000)... The figures for ferric hydrate have already been given. It has been pointed out that if specific molecular coagulative power be defined as the inverse of the volume occupied by one gramme-molecule of a sub- stance when it just suffices to bring about coagulation, then this value (K) varies with the valency of the active ion approximately according to the square and cube :-— (Bigsby, celts =e 1 12s Ke 118 Mr. W.B. Hardy. On the Conditions ivhich The relation really is not as simple as this ; it is complicated by the change which the specific molecular conductivity of a salt undergoes with change in concentration. The theoretical considerations have been dealt with elsewhere.* For convenience of description, how- ever, I will call this relation the relation of the square and cube. Action of Acids and Alkalis—The values for K furnished by these substances show relations to valency even more interesting than that found with salts. As in the case of salts, their action is entirely dependent upon the electric properties of the colloid particles. When the colloid particles are electro-negative, alkalis either do not cause precipitation at any concentration, or if they do cause pre- cipitation the value of K does not vary in any simple way with varia- tions in valency. When the particles are electro-positive, K increases with valency, but the relation of the square and cube does not hold. Instead, one finds that K varies directly with the chemical activity of the solution. Acids have the reverse relations. When the particles are electro- negative, the value of K varies directly with the chemical activity of the solution ; while if these particles are electro-positive, acids either ~ have no precipitating power, or if K has any value, then (in the par- ticular case measured) the value varies with valency according to the square and cube. The various measurements are brought together in the following table. The specific conductivities were calculated from the British Association tables. * Hardy and Whetham, ‘Journal of Physiology,’ vol. 24, 1899, p. 288, and Whetham, ‘ Phil. Mag.,’ November, 1899. + “The Electro-chemical Properties of Aqueous Solutions.” TT. C. Fitzpatrick ‘Brit. Assoc. Report,’ 1893. 119 sols. ality of Irreversible Hydro determine the Stab “U0T{OB OU SVT LT IB pozyVanges WoTyNTOR "UOT}VIAT [BIOMED 9Y4 MOS 07 oAros LOYY nq ‘UOMVAIASqO JO 4vyy UtOoAT YuUdLO_IP ornyesodure, 8 LO} UsaTd ore AYTATJONpUOD OYIOods TOF sonTRA OT, OOoT LT Lt OOT—ZL£T a OP | oF } OF L OF 41 LT AT LT LT 41 9T 00I—9T 00I—9T oJ0IT—9T ‘aangerodure 7, "eOT X (8ST 4@ MOI4NTOS Surye[nsvoo fo Aqratyonp -uo) oywodg “UOTNTOY SuyepMseog oO “OTF zod ‘atnbo OULUTRL) 000°6T 008‘OT 008‘IT 0 000‘8 000°8 00089 00098 000‘02% 000‘09F 000‘092 000092 098'T OOF FET 008‘0F ) 0 ‘O'O UT *[OUL -OULULVAD T "MOTYBINSBOD OONpoad 04 AIVISINIM WOT}VIPMIIMOD “°° S(O) *& ee ee **HOX + HOUR ** “BrUOULUL VY eee “Foey Hunt TOBIN Vee QOS THT sooo ee ORE * “pyoe o1[exQ canecrggery eevee fONH tees Tae * plo’ o1J00 V7 eee ee 'oa'H "**5(HO)*8@ 1 ees OR “ee TORY ** "BITOULUL YW ‘OAT yesSou -0.1400]9 ‘POD "OATZB90TL -01999]9 ‘O14st WW -_———————— — .. —— — | ————. ‘josorp ATT On the Conditions which Mr. W. B. Hardy. 120 9T 8% GS-0 | 00002 2 TOEN 9T GGZ 8-0 000‘9 were: STO 91 G.0 <000-0 | 000.000 j|°****’OSsIT 91 LL-0 9000-0 | oo0‘00es |°''*' OSH “UDAIS ON[LA OY} ULI[Y 1o9vo0IS oq You TTT splow avprunis yyim Adopeus Aq Ayrayyonpuoo oproodg 9T [4-0] 4000-0 | 000‘000‘F O1a4I9 9T as 200: 0 000°000'T | * ProB O11exO 9T 8.9 Z00-0 | O00‘000'T |**'** 'OS*H 91 68ST G0 | 0008 sees ONT OT oso |- ¢0O | 008‘T "ereeee TH | ‘eaTzTsod (J ¢ ° ee 3 eB -O1409T[9 91 g.2 100-0 | 000‘000°% HO)*E | ‘ogeapcy 9T 3% 100-0 | O00‘000'T |**'***OHM | AMOR LT 83 920-0 | 000‘S4 OS OL m3 PD F00-0 | o00;00g § j*"''** TOV 41 1 €10-0 | 000° oe fe UN Lt 9 ps00-0 | PPS‘ses |**'** 'ostH LT 6G 800. 0 000‘EZT die poms eiONS! *poyBinjges UIA WOTIOV ON ait Oe sie oe ** 8(HO)80 "e0T * 081 pal 7 qe WOTJNT[OS aod ‘ainbo | *o'o ur ‘Tout ‘eanqeredmay| sutyepnseoo owe) | -oULUTeIs T ‘josorp AFT go Aytayonp -100 oytoedg ‘MOTJB[NSROO sonpoad 09 A vessovou WOT}B1yMDTO(/) ‘panijwuog—uorynjog surye[aseog determine the Stability of Irreversible Hydrosols. 121 The figures in the fourth column are very remarkable. When the particles are electro-negative, equicoagulative solutions of acids agree in their electric conductivity within the limits of experimental error. The same relation is clearly shown if one takes the measurements which Picton and Linder made of the power possessed by acids of coagulating the hydrosol of arsenious sulphide. Value of K referred Sp. mol. conductivity when Acid. to Al,Cl, as unity. 1 gr. equiv. = 1000 c.c. HBr HI HCl 0-001 2950 AN O3 H3SO, 0:0006 1935 Oxalic 0:0005 578 When, however, the particles are electro-positive, the conductivity of equicoagulative strengths of acids varies to a remarkable extent. Acids. H’ : | a : BY Mastic, electro-negative ......... 12°6 14-4 13°9 Ferric hydrate, electro-positive 1650 6°8 )°7 Now specific conductivity (C) has the relation C = na(ut+v) where ~ is the fraction of the total number of molecules (n) which are dissociated at any one moment, and w + vis the sum of the velocities of the two ions. The factor u + v plays an important part, as will be seen by comparing the values for na in equicoagulative solutions of acids with slowly moving ions with those with rapidly moving ions :— 12a H3POx,4 at ate = wefel d xlatal atoll aieatta ate 0-01 iAgeticihs.csc Ape 0:07 BOG es aah ke. i das EEO k «ok monte ee } 0:004 BENNO. 55:08, 08) L220. 5. This, however, is probably partly due to the fact that owing to the manner in which the coagulative power was measured, time has practically a constant small value. The values for n might, perhaps, be different if the duration of the experiments were prolonged indefi- nitely. 7 122 Mr. W. B: Hardy. On the Conditions which The important point, however, which calls for notice is that the function a(w + v) is a numerical measure of the chemical activity of the substance at a given concentration, so that we reach the important conclusion that the concentration of acids necessary to coagulate electro- negative colloid particles, and of alkalis necessary to coagulate electro-positive particles, is determined by the laws which govern ordinary chemical equilibrium. In the case of the action of salts on these hydrosols, the relation is not so simple. K does not vary directly with a(u + %), but contains a factor which is approximately squared or cubed by a change from a mono-valent to di- or tri-valent ions. The relation can therefore be best expressed as thie K = na(u+v) A® where “is positive and increases rapidly with an increase in the valency of the ion whose electric charge is of the opposite sign to that on the particles. | I should interpret these relations by the suggestion that in the former the acid or alkali alters the difference of potential at the surface of the particles by altering the character of the fluid, and in that way modifies the stability of the hydrosol ; in the latter the active ions of the salt act directly upon the solid particles, or, perhaps, on the charge which these carry, and thus play a part which is, perhaps, generally similar to the action of ions when they furnish nuclei for the condensa- tion of vapour. Picton and Linder have shown that the active ions are actually entangled in, and form part of, the coagulum.* The former relation may profitably be placed beside Brihl’s con- clusions that the action exerted by a fluid upon the substance dis- solved in it is determined by the chemical characters of the former, as well as of the latter. He has shown that the molecular refraction, the dielectric coefficient, and the power possessed by the fluid of disso- ciating or chemically changing the molecules of the substance dissolved in it are measured by the unsatisfied valency, or, to use another phrase, the residual energy of its molecules. The action of acids or alkalis ona hydrosol, the particles of which are of the opposite electrical sign, seems to be compounded of these two actions. The acid or alkali may act as a salt, and exhibit the characteristic relation between K and the valency of the ion of the opposite electrical sign. An instance is furnished by the action of various acids on ferric hydrate, or the acid or alkali by increasing the difference of potential between the fluid and the solid particles may increase the stability of the hydrosol. ‘This is markedly manifested by the increased stability given to the hydrosol of gum mastic by the addition of univalent alkalis. In the action of barium hydrate on | this hydrosol, the segregating: action of the metal ion overcomes the * © Journ. of the Chem. Soc.,’ vol. 67, 1895, p. 63. determine the Stability of Irreversible Hydrosols. 123 action exerted by the reagent in virtue of its alkalinity, the result is that the coagulative concentration of the alkali Ba(OH), gives'a value for K which is less than that given by salts of bivalent metals, and the specific conductivity of the solution is of the same order as that of the coagulating concentration of salts of univalent metals. Against these suggestions, however, must be set the anomalous relations of be: various alkalis to the hydrosol of gold. | Action of a Salt in presence of Varying Amounts of Acid or Alkali. This was measured for one salt only, potassium sulphate, the col- loidal solution being gold. The figures are as follows :— Temperature 16° Concentration of the salt ne- Concentration cessary to produce blue tint. 1 gr.-mol. in— 1 gr.-mol. K,SO, in— C.C. c.C. Acetic acid......... 1,087 0 16,000 324,000 * 66,000 64,000 330,000 50,000 (neutral*) 28,500 PImmonia ......... 113,333 10,000 22,666 9,000 4,900 20,000 2,450 24,000 980 20,000 200 Large amount of salt needed. 100 Salt unable to act when saturated at 16° or at. 100°. These results are shown in the curve. Ammonia alone will not aggregate the particles of gold. Up to a certain point, however, it decreases the stability of the system. The conclusions can be summarised as follows :—The irreversible hydrosols which have been investigated are systems composed of solid particles dispersed through a solution of the substance of phe = in the water. The stability of the se okaes is related tothe contact difference of potential which exists between the solid and the fluid phases, and which forms round each solid particle a double electrical layer. Such double electric layers round particles of any kind immersed in a fluid * Except for e faint acid reaction of the gold solution, due HEoeeeny to a trace of phosphoric acid. 124 Mr. W.B. Hardy. ° On the Dolitatens which would resist any movement of the particles through the fluid, because, as Dorn’s experiments show, electric work is done in displacing the particles.* The effect would be the same as if the viscosity of the fluid was increased.T The stability of the system may be destroyed by altering the differ- ence of potential. Free acid, added to a hydrosol in which the particles are negative to pure water, will diminish the relative differ- Fia. 1. Acetic acid tooh% © Line of neutratiiy. KH, 50, 100% Ammonia 100% Action of potassium sulphate in presence of varying amounts of acetic acid or ammonia upon the hydrosol of gold. The abscisse represent the volume of water which holds 1 gramme-molecule of the salt. The positive ordinates represent the reciprocals of the volume which holds 1 gramme-molecule of the acid, and the negative ordinates the reciprocals of the volume which holds '- 1 gramme-molecule of the alkali.. Hach division = 50,000 c.c. ence of potential of the water. In this case the reagent acts directly on the water, and the coagulative activity of unit mass of the sub- stance varies directly with its chemical activity when dissolved in water. The same relation seems to hold when free alkali is added to a hydrosol in which the particles are electro-positive. 2 The stability of the system may also be destroyed by induction, the active agents being free ions carrying a static charge.{ In this case .. Wied. Ann.,’ vol. 10, 1880, p. 70. ‘+ This mode of stating the result I owe to Professor J. J. Thomson, and gladly acknowledge his kindness in discussing this and kindred points with me. Whetham, loc. cit. determine the Stability of Irreversible Hydrosols. 125 the action may be said to be on the particles, or rather on the electric layers immediately around them, and the active ions are those whose electric sign is the opposite of that of the charge on the surface of the particles. In this case coagulative power does not vary directly with variations in chemical activity. It rises exceedingly rapidly with a rise in the valency of the active ion, so that the relation l’:I’:1” =n:n?:n8 is approximately satisfied. _ Picton and Linder have shown that when the concentration of the salt is insufficient to completely destroy the system, it is not wholly without action. A fresh point of equilibrium between solid and fluid is reached by an increase in the size of the particles and therefore a diminution in the extent and curvature of the surface of contact. The fact is of importance, because it introduces us to the possibility that the reagent may affect the size of the particles by altering the equi- librium between the part of the solid in solution and the part in suspension. Double electric layers round each particle are, according to Thomson, separated by a region of ‘uncompleted chemical combina- tion” between the components.* The density of the field round the particles in hydrosols will therefore be a measure of the velocity of the solution and condensation between the particle and the liquid, and therefore the factor which determines whether the particles will on the whole grow, diminish, or remain stationary in size. When acids or alkalis are added to hydrosols holding particles of the opposite electric sign to themselves, the simplest relation seems to be that univalent acids or alkalis increase the stability ; bivalent acids or alkalis decrease it. The view advanced in this paper implies that each particle in a hydrosol is surrounded by a zone in which the components are in a condition of chemical instability. According to Rayleigh,+ such a zone is of finite thickness, and deep enough to contain several molecules. We therefore have in these hydrosols two phases, separated by a layer of extraordinarily large extent, which possesses considerable chemical energy. This, it seems to me, suggests an explanation of the catalytic powers so markedly manifested by hydrosols. * “Discharge of Electricity through Gases.’ Scribner, 1898, p. 24. + Thomson, loc. cit., p. 26; Rayleigh, ‘Phil. Mag.,’ vol. 33, 1892, p. 468. VOL. LXVI. M 126 | Sir Norman Lockyer. “The Piscian Stars.” By Sir Norman Lockyer, K.C.B., F.R.S. Received November 20,—Read December 14, 1899. Introductory. In the Bakerian Lecture for 1888* I briefly considered the relation of the stars having spectra with predominant carbon absorption (for which I have recently suggested. the name Pisciant) to the various other types of celestial bodies. Shortly afterwards I began the dis- cussion of the observations which had been made by Dunér in the case of these stars. Those relating to Group II (now named Antarian), another group studied by Dunér, were fully dealt with in the Bakerian Lecture, but after all the available information as to the former had been brought together, I found that notwithstanding the admirable data which Dunér had put on record, there were some points on which further information was desirable. In the Antarian stars we had evidence as to the lines present in company with the various flutings, but in the Piscian stars even the presence or absence of lines was somewhat uncertain. The publication of this investigation was therefore postponed to see what light could be thrown upon the subject by further observations. At various times, as the work permitted, such observations have been attempted, and the results, so far as they went, did not disturb the classification at which I had already arrived. Still, the information thus gained was less complete than desired. The photographic work which has quite recently been done on these stars by Dr. McClean and Professor Hale has now furnished the additional information required, and it is therefore unnecessary to delay the publication of the memoir, some ten years old, which in the main points stands as it was written. Historical Statement. Secchi was the first to recognise stars with spectra of the type under discussion during his spectroscopic survey made in 1866 and 1867. They constituted his fourth type. All the stars of the group are of small magnitude, and of a deep red colour. He was inclined to believe that a radiation spectrum was in question, but pointed out that there was a relation between the flutings of carbon and the dark bands seen in the spectra of the stars. He says :—‘ Quelques-unes des raies noires, et les plus importantes, coincident a tres-peu prés avec celles du troisieme type; cependant le spectre, dans son ensemble, se présente * ©Roy. Soc. Proc.,’ vol. 44, p. 26. + ‘Roy. Soc. Proc.’ (1899), vol. 65, p. 191. The Piscian Stars. TOT comme un spectre direct appartenant 4 un corps gazeux, plutot que comme un spectre d’absorption. Si on le considére comme un spectre dabsorption on trouve quill présente le caractére des composés du carbone, tels qu’on les obtient en produisant une série d’étincelles électriques dans un mélange de vapeur de. benzine et d’air atmo- sphérique et dans l’are voltaique entres les charbons.”* Secchi also statest that bright lines are occasionally seen in the spectra of stars of this type, and in a diagram of the spectrum of 152 Schjellerup he indicates no less than six bright lines. | Dunér in 1884 catalogued fifty-five stars of this group, and recorded the details of their spectra, so far as he was able to observe them. t Referring to Secchi’s work, Dunér says:—‘“ Secchi s’est beaucoup occupé de ces spectres, mais il y a de trés graves erreurs dans ce quil dit sur ’apparence qu'ils offrent et sur leur nature. D’abord il prononce a plusieurs occasions dans sa ‘ Memoria seconda’ et dans son ouvrage ‘Le Soleil’ que le rouge leur manque presque absolument. Pour ma part, je l’ai trouvé tres vif dans tous, seulement un peu pale en com- paraison avec la sous-zone jaune excessivement brillante. Puis Secchi parle des raies vives qui termineraient, du cété du violet, les zones brillantes. Mais ni dans la zone verte ni dans la bleu je n’ai vu la moindre chose qui pit expliquer un tel énoncé, et je sais que M. Vogel na pas été plus heureux. Quant aux deux raies brillantes que Secchi dit avoir vu dans le jaune, elles se rapportent selon toute probabilité a la sous-zone jaune, laquelle, comme je viens de le dire, est divisée en deux par une bande étroite. Secchi s’est plus tard persuadé, par des mesures, que les deux raies jaunes n’ont pas la méme position que celles du sodium ; mais il est néanmoins difficile de comprendre comment il a pu croire que cette zone, quarante fois plus large que la distance entre D, et D,, fit les raies du sodium. Au reste il parait disposé 4 admettre que tout le spectre est un spectre direct émis par un gaz incandescent. Pour moi, il est tout a fait incontestable que c’est un spectre d’absorp- tion, tout aussi bien que celui des étoiles de la Classe IIIa, et M. Vogel a déja, il y a quelques années, emis une pareille opinion, et il l’a répétée tout récemment.”§ In 1865 Zoéllner pointed out that spectra might enable us to deter- mine the relative ages of celestial bodies, and suggested that the yellow and red light of certain stars were indications of a reduction of tem- perature.|| There is now no doubt that stars with fluted spectra, whether of Group II (Antarian) or Group VI (Piscian), are cooler than stars like the sun and « Lyre, which have line spectra. But there is an * ‘Le Soleil,’ vol. 2, p. 458. + Ibid., p. 457. t ‘Sur les Etoiles 4 Spectre de la Troisime Classe,’ Stockholm, 1884. § Ibid., p. 10. || ‘Phot. Unters.,’ p. 248. 128 Sir Norman Lockyer. important difference between the two groups, as I pointed out in the Bakerian Lecture. In the Antarian stars we have to deal with con- densing swarms of meteorites in which the temperature is increasing, whereas in the Piscian stars we have an advanced stage in the cooling of masses of meteoritic-_vapours. In the case of Antarian stars we have ‘mixed radiation and absorption flutings, and I suggested a way in which the stars of the group might be divided into fifteen distinct species, representing successive stages in condensation. We have now to consider a similar classification of the Piscian stars, of which I stated in the Bakerian Lecture (p. 26):—‘‘ The species of which it will ultimately be composed are already apparently shadowed forth in the map which accompanies Dunér’s volume, and they will evidently be subsequently differentiated by the gradual addition of other absorptions to that of carbon, while at the same time the absorption of carbon gets less and less distinct.” _ Inconsidering the stars of this group, it is most important to bear in ‘mind that there are indications of carbon absorption in the spectrum of the sun. I first obtained evidence of the existence of carbon vapour in the solar atmosphere in 1874, and in 1878 I communicated a paper to the Royal Society on that subject.* Angstrém had already shown that the true carbon lines were not reversed in the solar spectrum, but I demonstrated by photographic comparisons that there was a perfect correspondence between the individual members of the brightest part of the fluting in the ultra violet (commencing at 3883°55), and a series of fine dark lines in the solar spectrum. I pointed out that this carbon vapour existed in a more complicated molecular condition (as is evidenced by the flutings) than the metallic vapours in the sun’s _atmosphere.t There can therefore be no doubt that in the sun carbon absorption is Just commencing, and that, as I stated in a former paper,{ “the ' indications of carbon will go on increasing in intensity slowly, until a stage is reached when, owing to reduction of temperature of the most effective absorbing layer, the chief absorption will be that of carbon— a stage in which we now find the stars of Class IIIb of Vogel’s classification.” * ‘Roy. Soc. Proc.,’ vol. 27, p. 308. + Note (added 1899).—Professor Rowland has since identified a considerable number of faint lines in the solar spectrum, on the more refrangible side of 6, with the constituents of the green carbon fluting (‘Prelim. Table of Solar Spectrum Wave-lengths,’ p. 90), and we thus have direct evidence of the presence in the solar spectrum of the band which is perhaps the most characteristic feature of the Piscian stars. ft ‘Roy. Soc. Proc.,’ vol. 43 (1887), p. 155. pie Pistiin Stars. 129 General Characteristics of the Spectra. The main features of the spectra are three broad dark fiutings, which fade off to the violet end of the spectrum. These, as is now well known, coincide with the three principal bands in the spectrum of high temperature* carbon. The wave-lengths of the bands, and Dunér’s numbers, are as follows :— No. of band. | Duner’s mean A. | Vogel’s mean A.f | ik eee fee 6 663 °3 563 ‘1 563 °3 9 516 °3 515 °9 516 °4 10 472 °7 472°9 473 °6 The greatest discrepancy is in the case of band 10, and this is easily explained when we consider the variation of the position of the maxi- mum intensity of the band to which I have previously drawn attention. t At different temperatures the position of. the brightest part of the band changes, and in Angstrém’s measure of the carbon fluting this was not taken into account. Tn addition to these principal bands, Dunér mapped seven secondary bands. In the Bakerian Lecture, I stated that “ there is evidence that some of the absorption is produced by substances which remain in the atmosphere during the next stage, that of Group VII (dark bodies). This probability is based upon the fact that some of the bands are apparently coincident with bands in the telluric spectrum as mapped by Brewster, Angstrém, Smyth, and others.” * As in former papers, the term “high temperature,’ as used here, is only relative, and refers to the spectrum of carbon which is seen in the electric arc, Bunsen burner, or vacuum tube under certain conditions. The spectrum of carbon at a still higher temperature consists of lines. + ‘ Potsdam Observations,’ No. 14, 1884, p. 31. ‘Roy. Soc. Proc.,’ vol. 45 (1889), p. 167. § Note (added 1899).—Although my subdivision of the group into species, made ten years ago, is based in part upon the secondary bands, it is not materially affected by the new information asto the nature of these bands, since, as cooling goes on, low temperature metallic lines would become more prominent relatively, just as we might have expected the secondary bands to become stronger on the supposition that they had the same origin as the telluric bands. 130 Sir Norman Lockyer. Specific differences in the Spectra. In considering the question of variations of spectra with temperature in these stars, the importance of taking differences of magnitude into account must not be lost sight of. A general examination of Dunér’s observations indicates that there are two marked differences in the spectra of the different stars. (1) Some of them give secondary bands, whilst in others they are absent. (2) Some of them have longer continuous spectra than others, as indicated by the number of “ zones ” visible. If the continuous spectrum extends far enough towards the violet, the three dark flutings of carbon will divide the spectrum into four bright zones. If it does not extend beyond the most refrangible of the flutings (A473) only three zones will be visible, and the continuous spectrum will appear to end sharply at wave-length 473. In one case it does not extend beyond the fluting at A517, and then only two zones are visible. These differences might evidently depend upon differences of mag- nitude of the stars concerned, but a detailed examination of the observations shows that some of the differences do not depend upon brightness. If we consider the visibility of the secondary bands according to Dunér, we have the following result :— Me GF Wand Warelourtn | Magnitude of stars Magnitude of stars in | in which it is seen. | which it is not seen. 1 . 656 -O 5 °*4—6°2 6 -0—9°5 2 621°0 5 °4—8°1 6°0—9°*5 3 604 °8 5°4—8 °°] 6 °0—9°5 4 589 *8 & 4—8°5 6-6—9°5 . 5) 576 °0 5 °4—8°1 7°5—9°5 7 551 °0 5 4—6°5 6 -O—9 °*5 8 528 °3 5°4—7°0 6 “C—9°5 This table shows that the visibility of the secondary bands is not altogether dependent upon the magnitudes of the stars observed. Thus the bands 2 and 3 are seen in some of the stars of the group as low as magnitude 8:1, whilst they are absent from some of the stars of the sixth magnitude. If we consider the question of the length of the continuous spectrum, we have the following result, the maximum number of zones referring to the longest continuous spectrum :— The Piscian Stars. 131 Magnitudes of stars in Number of zones. which they are seen. bw > lop) co or | octal Here, again, the visibility of the zones does not depend altogether upon the magnitude. There are stars as bright as the sixth magni- tude which only give three zones, whilst some as low as 8:2 give four. Dunér refers to this difference as follows :— “Puis Vintensité de la lumiére des zones brillantes peut varier considérablement chez les étoiles de la méme grandeur. Dans les étoiles d’un rouge foncé, la zone ultra-bleue est extrémement faible en comparaison avec la méme zone dans les étoiles rouge jaune; et chez les étoiles faibles, cette zone est tout-a-fait invisible, et méme la zone bleue est tres difficile 4 voir si elles sont trés rouge.”* Another important difference is the variation in the intensity of the citron band of carbon (band 6) as compared with the other bands. Dunér also refers to this point (p. 10) as follows :—“ Mais aussi la bande principale a la longueur d’onde 563 est d’une opacité trés variée. Chez certaines étoiles, elle est presque aussi foncée que les deux autres bandes principales ; mais dans certains spectres elle est assez faible, et semble, probablement a cause de cela, étre beaucoup moins large que les bandes aux longueurs d’onde 5:6 et 473. Celles-ci, et surtout la premiere d’entre elles, sont toujours trés fortes et tres larges, et forment le caractére le plus prononcé de ces spectres.” glimpsed. - No. 21, 89 Schj., Mag. 7°5 Feri Thee zones. ‘Band 6 rather dark, .i.% No. 24, 124 Schj., Maes 65 frj.—Three zones. Gaxition bands very wide and-dark; 4 and 5-well seen. No. 27, 136 Sely. “ ee 6: 0 ae ——ihree zones. Banas 4 and 5 well seen. ' No. 28, 145 Schj. ., Mag. 81 Bij <-Three’ zones. eens 4 and 5 well seen ; 2 and 3 possibly present. Rn Oe Se ae een 134 Sir Norman Lockyer. No. 30, 1556 Schj., Mag. 7:3 Rrj.—Three zones. Carbon bands very dark; bands 4 and 5 well seen; 1, 2, 3 very feebly visible. No. 33, 202 Schj., Mag. 8:5 Rrrj.—Three zones. Band 6 dark; 4band 4. No. 35, D.M. + 36° 3168, Mag. 8:5 Rrj.—Three zones. Carbon bands wide and dark ; trace of band 4. No. 37, 219 Schj., Mag. 8°0 Rrj—Three zones. Carbon bands very wide and dark; bands 4 and 5 seen; band 2 present, but very weak. No. 38, 222 Schj., Mag. 9:0 Rrrj—Three zones ; blue very weak. Carbon bands strong. No secondary bands were recorded, probably because of the faintness of the spectrum. No. 39, 222e Schj., Mag. 7°3 Frrj.—Three zones ; blue very weak ; 4 and 5 distinctly visible. No. 41, 228 Schj., Mag. 7:0 Rj—Three zones. Bands 2, 3, 4, and 5 visible. No. 42, D.M. + 32° 3522, Mag. 8:0 Frj.—Three zones; blue rather bright. Carbon bands very wide and dark ; 4 and 5 seen. No. 45, D.M. + 35° 4002, Mag. 9°5 krj—Three zones. Carbon bands very wide and dark. (The absence of secondary bands probably due to faintness of star.) No. 49, V Cygni (Var.) Rrrj—Three zones. Band 6 wide and dark. No. 51, S Cephei (Var.) Rrrj—Three zones. Band 6 wide and dark. No. 53, 251 Schj.. Mag. 7-8 Rrrj—Three zones. Bands 4 and 5 doubtful. Species 7. Characteristics. —Three zones. Band 6 pale. No. 9, 43 Schj., Mag. 8-1 Frrj—Three zones. Band 9 strong ; band 6 pale. No. 12, 99 Birm., Mag. 8:0 Rrj.—Three zones. Band 9 very strong; band 6 weak. No. 16, 72 Schj.. Mag. 7:4 Rrj—Three zones. Bands 9 and 10 wide and dark; band 6 weaker; possibly bands 4 and 5 are visible. No. 31, V Corone (Var.) Rrj—Three zones. Band 9 strong; 6 rather weak. | No. 48, U Cygni (Var.) Rrrj.—Three feeble zones. Band 6 weak. The remaining stars of the group observed by Dunér are not described with sufficient detail to enable them to be included in the foregoing classification. The Piscian Stars. $39 The variations from one species to another are shown in the accom- panying map (p. 136), which also indicates the connection with stars of the group which precedes, it, and suggests a later Species 8, not yet identified by observation. Owing to the difficulties attending the observations, the actual appearances of the secondary bands are not quite so regular as is shown in the map. | The Colours of the Piscoan Stars. The classification arrived at was next tested by reference to the colour phenomena. Dunér employs two methods of indicating the colours of the stars, which he has specially observed—first, by means of initial letters, and second, by means of numbers, such that 10 = blood red. ‘Thus : Rj = rouge jaune. Rrj = rouge jaune foncé. _Rrrj = presque rouge absolue. The numbers show a fair agreement with the letters employed, if we omit one Rrj star, which is given the number 9°3. The following table compares the colour numbers corresponding to the initial letters :-—— Initial letters. Range of numbers Mean No, Remarks. corresponding. | Rj ne 8:2 One member only. Rrj 7°8—8°6 8°3 One star, 9°3, omitted. Rrrj 8°8—9°5 9-04 If we consider the colour numbers corresponding to the various species into which I have divided the group, we find the average numbers to be as follows :— Species ree 8-1 GES eee any 8°6 SS aE A ce eet aera 8°2 ee ey 8:1 Oa deca ote ae 8:2 Ts. tee is cate 8°8 i iceatirte este 8°6 On the whole, therefore, considering the difficulty of the observa- tions, there is an increase of redness as we pass successively through ckyer Norman Lo Sir 136 a WH i nt iNT tii IBIG WeIosty Jo sowodg oy} Sutrmoys dey Qe = SF 2 RG Hl < Be ga sa Ge La Sw 8 a TT | pang 2 SB Lm anc aii Be a : a ba ca si allt Te in ui J a E:T HN A ao t dl nS HH iit HT LE The Piscian Stars. AST successive species, which is exactly what we should expect if these species truly represent the effect of gradual cooling. Variability of Piscian Stars. Of the fifty-five stars in Dunér’s list, ten exhibit fluctuations in brilliancy. ) On the whole the light changes are not so great as .in the stars of Group II, and the periods tend to greater length. As to the cause of variability, the increase of light at maximum may be due, as I suggested in 1890,* to the light added by bodies of a cometary character when they reach periastron, the increase of luminosity being produced by tidal action, as in the case of comets in our own solar system. If there be any truth in this idea, it seems probable that the added light of the comet at maximum, which would give a spectrum consisting of bright carbon flutings, would produce a paling of the carbon absorption flutings. _ As in the variables of the Antarian group, which are uncondensed swarms, and where, on the meteoritic hypothesis, the increased light at maximum is produced by the collision of a revolving swarm at ‘periastron, irregularity is a natural consequence of the revolution of more than one secondary body. ADDENDUM. Recent Observations. The Kensington observations were made chiefly during 1894 and 1895, with special reference to the lines involved. The stars selected for observation were 132 Schjellerup, 152 Schjellerup, 115 Schjellerup, and 19 Pisctum. The 3-foot reflector was used. In addition to the carbon bands, numerous lines were seen without much difficulty, but only the more prominent ones could be satisfactorily measured. Among the lines recorded in 132 Schj. were Hf, the E line of iron at 5269, and a group of lines near A 5380. In 115 Schjellerup addi- tional lines were measured near 5005, 5762, and 5429, and the presence of H@ was again determined by comparison with a hydrogen vacuum tube. In 19 Piscium numerous lines were observed, among them being D and F. No suspicion of bright lines was entertained during these observations. Attempts to photograph the spectra were not sufficiently ‘successful to help matters. In 1898, Dr. McClean published photographs of the spectra of 19 Piscium and 152 Schjellerup,+ showing that these stars have a line spectrum similar to « Tauri, in addition to the well-marked bands * © Nature,’ vol. 42, pp. 419, 548. + ‘Phil. Trans.,’ vol. 191, A, p. 181, plate 14. 138 Sir Norman Lockyer. of carbon. This was the information wanted, but more recently Professor Hale has published photographs of the spectra of 280 Schj., 273 Schj. (19 Piscium), 132 Schj. (U Hydre), and 152 Schj., taken with the aid of the Yerke’s telescope at Chicago, and showing a wealth of fine detail.* The dark line spectrum is very marked, and the details of the carbon bands themselves are clearly revealed. Besides these, there are certain bright places in the spectrum which Professor Hale has been led to believe are true bright lines, and he mentions that Professor Keeler has arrived at the same conclusion as a result of his observations with the Lick refractor. Dunér appears to have continued his observations of this group of stars after his removal from Lund to Upsala, and he states that with the Upsala refractor he was able to see more detail, and could detect without difficulty bright lines in the spectra of various stars of the group.T The Question of Bright Lines. As I have already pointed out, Professors Hale, Keeler, and Dunér consider that there are bright lines in some of these spectra, but I must confess that the published photographs do not convince me upon this point. In the plate which accompanies Professor Hale’s paper of April, 1899,{ the spectra of four stars are shown, namely, 280 Schj., 273 Schj., 182 Schj., and 152 Schj. A study of these photographs shows that the supposed bright lines are involved in the carbon ab- sorption bands in the yellow green, and occur where there is reduced absorption, on the less refrangible sides of the dark flutings. This at once led me to suppose that they could not be real bright lines, but simply places in the continuous spectrum where there is least absorption. These supposed bright lines are most marked in 152 Schj., and there is no suggestion of them in 280 Schj., while I think few © would be disposed to suggest their presence in 273 Schj. and 132 Schj. without having 152 Schj. as a guide. Nevertheless, in these inter- mediate stars there are certainly bright places corresponding in poai- tion with the “ bright lines” in 152 Schj., the principal one being at 5592. So far as appearances go, the greater apparent intensity of the bright line in 152 Schj. appears to be due to the introduction of a - strong absorption line on the less refrangible side. In another paper§ Professor Hale reproduces photographs of 152 Schj. in which the contrast has been increased by photographic means, so that the whole spectrum appears to consist of bright lines, rather than dark ones. * ¢ Astrophys. Journ.,’ vol. &, pp. 288-9; vol. 9, p. 271; vol. 10, p. 110. + ‘ Astrophys. Journ.,’ vol. 9, p. 121. t ‘Astrophys. Journ.,’ vol. 9, p. 271. § ‘ Astrophys. Journ.,’ vol. 10, August, 1899, p. 108. The Piscian Stars. 139 In favour of the real existence of bright lines, Professor Hale points out that the contrast between the line and the continuous spectrum increased rather than diminished when dispersion was increased, and that there was no decrease in contrast as the slit was widened. The question, however, is so complicated by the presence of the carbon fluting and other absorptions, that I shall not follow Professor Hale in his definite conclusions as to bright lines upon these grounds. Before we can admit the certain presence of bright lines in 152 Schj., we must consider whether similar appearances occur in other stars where bright lines have not been previously suspected. As a matter of fact, in the photographic spectra of a Tauri, 6 Andromede, and @ Orionis, which I published in 1893,* the spectra might, so far as mere appearance goes, be regarded as containing both bright and dark lines, some of the bright spaces between obvious dark lines being very conspicuous ; the same remark applies in a less degree to the spectrum of Arcturus which I published at the same time. But we find a com- plete explanation of these spectra if we regard them as consisting of dark lines, whereas if we take the bright spaces we cannot match them at all. We do not hesitate in these cases to treat.the spectra as consisting of dark lines only, the apparent bright lines being simply spaces between dark ones. I find that practically in all dark line spectra where the lines are from some cause or other thick, the inter- vals between them are apt to appear as bright lines, and this bright- ness can readily be intensified by purely photographic processes. I have accordingly thought it unnecessary to modify the division into species on account of the supposed presence in some of them of bright lines. If the presence of bright lines be eventually established, may they not indicate that we are observing the effects of volcanic gases floating over a “ photosphere ” which has attained the consistency of lava ? Bearing on the Meteoritic Hypothesis. The photographs taken by McClean and Hale have now sufficiently shown that there is much in common between the line spectra of the Antarian and Piscian stars. This indicates that there is a practical equality of mean temperature in the reversing layers of the two groups, but we find a very great difference in the conditions as to carbon; while carbon is undoubtedly absorbing in the Piscian stars, it is cer- tainly not absorbing in the Antarian, and there is in fact strong evidence that it is radiating.T We cannot imagine different kinds of stars of the same temperature as representing the same stage in any evolutionary scheme, so that the * ¢Phil. Trans.,’ A, vol. 186 (1893), plate 23. + ‘Roy. Soc. Proc.,’ vol. 44 (1888), p. 52; ‘ Phi!. Trans.,’ A, vol. 186 (1898), p. 704. (140 Prof. Karl Pearson. separation of the two groups which I suggested in 1887 is fully justified by the recent work to which I have referred. By putting the two groups on the same level of temperature, but on opposite sides of the temperature curve, as in the evolutionary order forming part of the meteoritic hypothesis, the differences are fully explained. It will be seen that this work carries us a step beyond that with which I have recently been engaged in connection with the hotter “stars. General Conclusions. (1) The undoubted presence of dark carbon flutings in the solar spectrum, including that near ), and of solar lines in the Piscian stars, indicates that the Piscian stars are next in order of development to the Arcturian stars. (2) The stars observed by Dunér may be divided into seven species, beginning with the hottest and ending with the coolest stars. (3) The reported presence of bright lines in the Piscian stars must be received with caution, as similar evidence of bright lines might be adduced in the case of other classes of stars in which the spectrum is fully explained by dark lines alone. _ (4) The redness of the stars increases as we pass from the earlier to the later species of the group. iil _ (5) The variability in this group is less marked than in the Anta- rian stars, and may perhaps be accounted for by the revolution of secondary bodies of the nature of comets round the stars themselves. (6) The place on the temperature curve assigned to these stars on the meteoritic hypothesis is fully confirmed by the more detailed ‘inquiry, and the hypothesis is thereby strengthened. I am indebted to Mr. Fowler for assistance in the determination of the species and the construction of the map ten years ago, and for additional assistance in discussing the recent work. I have also to express my thanks to Mr. Shackleton for a detailed examination of the recent photographs. “ Mathematical Contributions to the Theory of Evolution.— On the Law of Reversion.” By Kari Pearson, F.RS. (A New Year’s Greeting to Francis Galton.—January 1, 1900.) Received December 28, 1899,—Read January 25, 1900. (1) Introductory—In a memoir recently presented to the Royal Society, I have endeavoured to emphasise the importance of distin- Mathematical Contributions to the Theory of Evolution. 141 guishing between three diverse types of heredity, namely (i), Blended Inheritance, (ii) Exclusive Inheritance, and (iii) Particulate Inherit- . ance. In a memoir printed in vol. 62, pp. 386412 of the ‘ Proceedings,’ . I have dealt at length with the theory of blended inheritance, general- ising for this purpose Mr. Galton’s Law of Ancestral Heredity. Allowing for a certain degree of variation in the constant y, or. “ coefficient of heredity,” there discussed, I consider that this theory gives a fairly good first approximation to the facts hitherto observed in| this field. But blended inheritance certainly does not cover the whole field of heredity. When a character blends, then this law of ancestral heredity tells us the most probable blend for the offspring of given ancestry. It shows us the offspring of exceptional parents regressing towards mediocrity, owing to the fact that without stringent selec- tion the great bulk of their ancestry must be mediocre and not exceptional.* Thus the main feature of the law of ancestral heredity is regression. Such regression is not what most biologists would un-. derstand by reversion. In fact, when the inheritance from a variety of ancestry is blended, the idea of reversion becomes very obscure ; I venture to think meaningless. Let us suppose stature a blended character, then the array of off-. — spring of a definite short statured ancestry will have a mean regressing (here progressing) towards the population mean and a definite vari- ability. Hence the theory of chance enables us at once to determine, the frequency of a very tall man born of such short ancestry. The frequency may be small, but sooner or later the tall man will appear. Now let us suppose onc distant ancestor in the otherwise short ancestry to have been tall. Clearly his existence will hardly affect. at all the mean of the array of offspring. . He will not materially influence the chance of a very tall man appearing among the offspring ; yet a superficial observer might easily. describe the appearance of the very tall man as a case of reversion to the distant tall ancestor. The absurdity of this attribution is mani-, fest when we remember that persons like him would have had sensibly equal frequency with or without the distant tall ancestor. In fact, it seems to me that in the case of characters which continuously vary, and which blend their inheritance, it is hopeless to look for any evidence whatever of reversion. The term is, then, meaningless. To find reversion we must investigate cases in which characters do not, blend, 7.¢., the individual takes exclusively after some one member of the ancestry. In this case the appreciation of reversion becomes possible and its meaning intelligible. Cases of this kind are by no means un- * An individual has 1024 1(th great parents, and these can hardly be anything else but a fair sample of the population of their generation, if there has not been an excessive amount of in-and-in breeding or much selection. — Lia Ls VOE.LXV1. N 142 Prof. Karl Pearson. common. Thus, Mr. Galton writes in his ‘ Natural Inheritance ’(p. 139) : “Parents of different statures usually transmit a blended heritage to their children, but parents of different eye-colours usually transmit an aiternative heritage . . . . . . if one parent has a light eye- colour and the other a dark eye-colour, some of the children will, as a rule, be light and the rest dark; they will seldom be medium eye- coloured like the children of medium eye-coloured parents.” Again, in his paper on ‘‘ Basset Hounds,”* Mr. Galton classifies these hounds as tricolour (T) and non-tricolour (N), remarking, “ I am assured that transitional cases between T and N are very rare, and that experts would hardly ever disagree about the class to which any particular hound should be assigned.” In other words, Mr. Galton appears to assume exclusive inheritance.t Roughly, in such exclusive inherit- ance, the offspring takes after one or other parent, or reverts to more distant ancestry. It becomes accordingly somewhat difficult to see how the law of ancestral heredity, which applies to blended inherit- ance, can be transferred to this different field. Yet Mr. Galton in his ‘Natural Inheritance’ (p. 153) writes: “The broad conclusion to which the present results irresistibly lead, is that the same peculiar hereditary relation that was shown to subsist between a man and each of his ancestors in respect to the quality of stature, also subsists in respect to that of eye-colour.” Further, in the paper on Basset Hounds, he actually endeavours to demonstrate the truth of the law on the exclusive colour of these hounds. Now I think we must keep these two matters quite apart. The average stature of an individual is a blend of all his progenitors’ characters ; even in a single individual we find contributions from many ancestors; this is not the case with an exclusive inheritance, and it does not accordingly seem to me possible that “the same peculiar hereditary relation that was shown. to subsist between a man and each of his ancestors” for a blended character can also hold for an exclusive character. ) It is no longer of the proportions of a character in one individual that we speak, but of the frequency of various types of individuals among the total offspring of a given ancestry. The one statement is a law of blending characters, and the other is a law of distributing the exclusive characters among a group of individuals. In the first case we deal with regression, in the second with reversion. What Mr. Galton really asserts is, that the proportions of reversion in an array of offspring are identical with the proportions of blend in the average * “Roy. Soc. Proc.,’ vol. 61, p. 403. + A remark in the ‘ Natural Inheritance’ (p. 139) that “Stature is due to its - being the aggregate of the quasi-independent inheritances of many separate parts, while eye-colour appears to be much less various in its origin,’ would seem to indicate that Mr. Galton considers that blended inheritance is ultimately based upon exclusive inheritance of parts—a suggestion well worth investigation. Mathematical Contributions to the Theory of Evolution. 1438 individual. If this be true, then his law, or possibly some generalisa- tion of it, is very comprehensive ; it embraces the two distinct types of heritage, blended and exclusive. But I think it most desirable to keep the two ideas quite separate, and speak of the one dealing with blended inheritance as the Law of Ancestral Heredity ; the second, dealing with exclusive inheritance, as the Law of Reversion. If this be done, we shall, I venture to think, keep not only our minds, but our points for observation, clearer ; and further, the failure of Mr. Galton’s statement in the one case will not in the least affect its validity in the other. (2) The Law of Reversion—Let us examine first what I take to be Mr. Galton’s view of this law. Out of an array of N offspring, 1/4 N will follow each parent, 1/16 N follow each of the four grandparents, nth great parents. In this manner the total offspring N is distributed by reversion among the ancestry. Now I want to draw attention to one or two points here. 1/4.N will not be all the children like, say, their father; for out of the 1/4.N who are like members of his ancestry, those who are like ancestors like him—and these ancestors will occur in certain proportions—will thus also be like him. This holds for each individual ancestor; the number like any ancestor will be considerably greater than the number who “follow” that special ancestor. Now let piN, psN, p3N, piN, : pnN... be the number of the offspring like a cand a grand- parent, a great grandparent, and mth parent, &c. This brings me to my second point. A special meaning is here given to the word like. p,N is not in the usual sense of the word all the number like the father. If the offspring had the same distribution of character as we find in the general population, then undoubtedly some would have the same quantity or quality of the character as he has—some, for instance, would be blue-eyed if he were blue-eyed—but this is a random likeness and not like in the special sense in which we are using the word. p,N are like the father owing to the laws of heredity, the remainder have a random distribution so far as he is concerned, and we exclude any random likeness from our considera- tion. How then are we in actual observation to distinguish hereditary from random likeness?* ‘The answer is simple; p,N out of N pairs of parent and offspring will be absolutely correlated, 7.¢., have a correlation equal to unity, but the remaining (1-—p,)xN pairs will have zero correlation, although there may be random likenesses. Hence, by the theorem given by me in the ‘ Phil. Trans.,’ vol. 192, p. 276, the actual correlation will be perfect correlation reduced in the ratio of the * I exclude for the present the influence of assortative mating. A likeness to the mother, otherwise random so far as the father is concerned, may thus become a real likeness to the father. N 2 144 Prof. Karl Pearson. number of correlated pairs to the total number of pairs.* Thus the correlation of parent and offspring = 1 x pi:N/N = py. It thus follows that pi, po, p3..-... 5 PAbieeie are the correlation coefficients to be expected between offspring and parent, grandparent, ....mth great parent, &c. Here we have assumed equal potency for both sexes and all lines of descent, otherwise these coefficients must be looked upon as mean values of the correlations for different genera- tions of ancestry. Lastly, it seems to me that reversion may not be the proper word to apply to those who directly follow their parents, and that these may be fairly considered direct inheritors and distinguished from reverters. I shall accordingly assume no @ priori relation between these two classes, certainly not that direct inheritors and reverters are equally numerous, 2.¢., 4 and 4, as in Mr. Galton’s Law. As for reversion itself I will only suppose it to diminish in geometrical progression as we step backward to more and more distant ancestry. I shall accordingly take BN offspring to follow either parent and yaN, ya2N, ye@N, &c., to follow grandparent, great grandparent, great great grandparent, &c. With these preliminaries arranged we can now pro- ceed with the analysis. (3) The Generalised Law of Reversion—The total number of off- spring N is clearly the sum of all those that follow all the successive ancestors, 2.¢., N = 26N + 4yaN + 8ya?N + 16ya3N +...... 26+ 4ya/(l — 2a)... i ies .scccae sheen eee (i) Now consider how the number of offspring “like” or absolutely corre- lated with one parent are made up: they are p,N in number; they consist first of BN, the number directly inheriting from this parent ; also there will be yaN like each of the parent’s parents, and the parent will be like one or other of the parent’s parents in p; proportion of cases ; similarly there will be ya?N like each of the parents’ grandparents, and the parent is like each of the parents’ grandparents in pz cases ; and soon. Thus we have I or 1 piN = BN + 2yapiN + 4ya2p.N + 8ya?pgN +...... jenn Py = B+2yalp, + 2aps+ 4a7p3t ...... ) ese (ii) Now note how the p2N like any one grandparent is made up. We have directly yaN reverting to this grandparent, ya?N to each of the grand- parents’ parents, and in each case pyya?N like the grandparent ; similarly out of those ya#N reverting to any grandparents’ grand- * In this case there is every reason for supposing o, = 0, = 6)’ = og, and mM, = Mg, m = m/. Thus 2 = o,/ 2’ = o;’, and since r = 1, R = 1)/N. _ Mathematical Contributions to the Theory of Evolution. 145 parent, there will be poya®N like the grandparent, and so on. But beyond these contributions, certain of the BN who follow the parent will be like the grandparent, for the parent is like the grandparent in p; fraction of cases. Hence we have finally : poN = pi PN +yaN + 2p,ya"N + 4poya®N 4+...... or po = piP+yat 2ya?(p; + 2ap.+ 4a%3+...... Nee Sr accseaugae (iii) Proceeding in the same manner we find ps = po +yapit ya? + 2ya8(p; + 2ape+ 4a3p3+...... Veen See (iv) ps = pa + yap, + ya%pr +22 + 2yaX(py + Laps + 4a ps +...) ...(¥) and so on. Hence we deduce from (iii) and (11) p2 = piB+ya+a(pi — £) or Pa eee Ray E(B) ooo cee cacecs ces wees ieee (v1) Similarly from (iv) and (iii) P3 = Po + yap + &(p2 — pif) or Ps = (4+ P)pota(y—B)p1 ...sececeeee ceeeees (vii) Again from (v) and (iv) ps = ps8 + yap2+ a(p3 — p2/) or Pa = (a+ P)pg toy = B) po vavessiee crear danse (viii) Generally Pe (OEP par + (YB) paige es ine een ace ene sos (ix) with p, = 1, by (vi). To solve equation (ix) assume as usual py, = Am”, and we find m?—(a+B)m + a(y—f) = 0. Thus m = a+ B+ see eon FI, _ But by (i) 1-2(a+) = 4a(y—-). Hence m= 5 or m= a+p—, mee cc ha (x) We have then Aaa A,(5) 4 Aga ee 5) oe (xi) where A, and Aj are constants. 146 Prof. Karl Pearson. But p= 1, .*. A,+Ae = 1. Thus we may put = (1-—c)(4)"+ce(a+B—4)"............ (+p ome EAL) where ¢ is a constant. Let us substitute this in equation (ii); we find (1 -0)5+a+B-5) = B+2ya4 (1-054 2a(l -2) a, gt 4a%(1 — A Byer } te(a+B—5)+da+B 5) 2a+a+B-5) 47+... a 3(1-c) ca+B-$) | 2s or = B+2ya4 20-9 epee eo — Bala Bi da ear (xiii) Write 6 = a+ —-43, then by (i) we have = B(a HOPE 28) ih, cesses acces ees (xiv) Hence from (xiil) | 2yad ya oS De ya o(3- 7 sty aaa) 1- and by (xiv) (1 — 2a6)(8 — «) oe ~ (1—28){ (= Ol = 2a) tl on 200d) oS es Suppose the parental and grandparental correlations observed, then pi = 3(1—¢) +68 » +0 catelp nas ree (xvi) po = Wc) +08? will both be known. These will give c and 6; then (xv) will give e and (xiv) y, while will determine 2, and the whole law of inheritance and reversion will have its constants fully determined. We have, indeed, from (xvi) <—e (xviii) = (ge 2p (xix) Mathematical Contributions to the Theory of Evolution. 147 From (xv) 2+¢+46+4 2262 lice a OS aa = a O(1 +8408) oy se ae (xx) where é= ns : Lastly from (xiv) = gelaerts Fh RRR 26 (xxi) a Thus (xviii) and (xix) give 6 andc. (xx) then gives a, taking the root less than unity. Finally (xvii) and (xxi) give 6 and y, completing the solution. (4) Comparison with Law of Ancestral Heredity.—Now let us compare these results with those I have obtained from the law of ancestral heredity.* On p. 390 of the memoir on that subject we have for the nth midparental correlation with the offspring pn = 2?”7n, where rp 1s the correlation of the offspring with the individual nth great grand- parent. By p. 394 p, = ca” Hence ai CONS) pcomne a iceniemes tee: (xxii) or the correlations of the offspring with the ancestry follow a simple geometrical progression. Comparing this with the result (xii) of this paper, or tm = (L—c)(h)*+c(at+B—$)™ oc eee (xii) where ¢ is now a different constant, we see that the two cannot possibly be in agreement, unless one of the terms of the latter result vanishes. Thus there is in general a fundamental difference between the law of ancestral heredity and the law of reversion; they give expressions differing in character for the ‘correlations between the offspring and individual ancestors. Let us see when the two laws will agree. There is unfortunately a bad slip in my memoir of 1898. The series at the top of p. 403 leads to | yP'/1-f') = 1 and not as there given yf'(1—yf’) = 1. Thus we have P(1+y) =1 or 4 = 1/(1+y). Hence B = 98 Soa) ch kt Y) panna ans cae (xxiii) and therefore by (xii) & == 1) 2. Thus by (xxii) of this paper * ‘Roy. Soc. Proc.,’ vol. 62, pp. 386—412, 148 i ~ Prof. Karl Pearson. This shows us that the correlation with the individual ancestor is halved at each backward step in the pedigree. We see at once that (xii) can only be in agreement with (xxiv)—the letter ¢ being different in the two, and merely standing for a constant—provided a+ = i, or by (i), provided-y = 8. Thus the condition that blended and exclusive inheritance should lead to the same values for the corre- lations with the ancestry is : that reversion should form a series starting with the actual parent. If this condition should hold, then, for example, the grandparental correlation must always be one-half the parental either for blended or for exclusive inheritance. (5) Correction of an Error in Memoir on Ancestral Heredity.—lt may be of value to insert here the modifications required in my memoir on blended inheritance, owing to the slip just referred to: they apply to the results deduced from (xvii) on p. 403; these are the table on p. 403, and the result immediately under (xvii) on p. 406. In the first place the law of ancestral heredity may now be written AS Sop, Ao See eae Dia) git cede oa? eat mm ae ) a very simple form. In the second place we may replace the equation for con p. 394 by 2y+1 ~ Pa Oy 41 cee peise Odean ee (xxvi) whence we find for the parental correlation ca/ ,/2 or $c | Di 2y+1 ie i ares | J ee (xxvil) Each succeeding ancestral correlation will be obtained by repeated halving of this value. Lastly, the result on p. 406 for the fraternal correlation becomes*— eeeeeeeseoeeeocsreeseteosn The following table indicates the effect of varying y on the intensity of heredity, and should replace that on p. 403 of the memoir on the ancestral law :— * Equation (xviii) on p. 406 of the memoir is correct, but the value of r in terms of y below it, since it depends on the erroneous Equation (xvii) of p. 403, as well as the limit given for y in the foot-note, must be cancelled. Mathematical Contributions to the Theory of Evolution. 149 Table of Heredity for divers Values of y. (6) Difficulties arising when we apply these Results for Blended Inheritance. —Now the above table shows us that by varying y sufficiently we can obtain a considerable range of values for the correlation of characters in kindred. But these values are limited by two serious considerations, namely :— (i) The ancestral correlation is halved at each stage. (i) The fraternal correlation appears to become perfect as we approach the upper limit of parental correlation, 7.¢., 0°5. Now actual determinations of grandparental correlation in the cases of eye-colour in man, of coat-colour in horses, and of coat-colour in hounds, which I have recently made, do not as a rule seem to justify the statement that the grandparental is half the parental cor- relation. Further, in two of these cases, the average parental correlation is quite 0-5, but the fraternal correlation is, while larger than 0:4, still a good deal short of perfect. Hence I am bound to conclude that :— (i) These characters do not obey the laws of blended inheritance as deduced from the law of ancestral heredity ; or, (i) The laws of blended inheritance, as deduced from the law of ancestral heredity, would be largely modified if we considered the influence of assortative mating, or (ii) The fundamental assumption that if all the midparents right away back had the same amount of the character, the average offspring would have also the same amount, is not justified. Thus the result 8 = 1/(,/2(1+y) ) in Equation (xxiii), per- haps, is unnecessary, or there may be ¢wo independent con- stants of inheritance. ‘ Value of y. 0'7 0°9 1, 12. 2°35. con Parental corre- UIT ae 0°2485 0 °2851 0 °3000 0 °3248 0°4000 0°5C00 Grandparental correlation .. 0°1243 0°1425 0°1500 0°1624 0 °2000 0°2500 Great grand- parental cor- relation ......... 0°0621 0:0713 0°0750 0°0812 0:1000 0 °1750 Fraternal corre- lation .........00. 0 -2899 0 °3665 0 -4000 0 °4586 0 °6596 1-0000 Regression on nth midparent 0°4970 0°5701 0°6000 0°6497 0°7999 170000 Correlation with nth midparent | 0°4970(0°7071)" | 0°5701(0°7071)" | 0°6000(0°7071)”| 0°6497(0°7071)" | 0°7999(0°7071) | (0°7071)" 150 Prof. Karl Pearson. It is quite possible that eye-colour in man and coat-colour in hounds are exclusive and not blended inheritances, so that (i) would cover these cases. On the other hand, I have found parental correlations as high as 0°5 for a new and large series of stature data in man, without fraternal correlation approaching unity. Here (i) can hardly apply, although (ii) may, for the coefficient of assortative mating in this case is remarkably high, nearly 0°3. But I think that, even if (i) or (ii) might help us over our difficulties in certain cases, we ought to carefully reconsider the assumption referred to in (iii). It would surely only be justifiable in the case of an absolutely stable popula- tion, each generation of which has existed under an identical en- vironment. In itself it seems to exclude any secular change due to natural selection, or to improved physical or organic environ- ment. In fact, we must proceed with caution when applying the statement that the average of all the offspring of an absolutely ~ same system of midparents would be like those midparents; for a portion of such offspring have very probably been removed by selection, and our average is not really that of all the offspring, but of the fitter. In the like manner, we must treat with some caution the principle on which Equation (i) of the present paper is based. It assumes that all the ancestral contributions are to be found in the present progeny ; but what if the contributions of certain ancestors by selection, artificial or natural, have been eliminated before reaching the existing generation? What if the coat-colours of certain ancestors were unfashionable, and only their unlike descendants have been put to the stud? Our theory may be quite correct, but it may appear erroneous when tested by facts observed in the case of horse or dog breeding. Let us investigate whether independent y and ( in our expressions for parental and fraternal correlations would enable us in the case of blended inheritance to reach a value of the former as high as 0°5 without the latter becoming perfect. I find if 7; be the parental corre- lation, = ca/ ,/2, from Equation (xi) of my memoir on the ancestral law (p. 394), and if r+ be the fraternal correlation obtained from Equation (xviii) of the same memoir : a Se Ea (xxix). /2 eyez hl +27) wich £2 RE . ie XXX pos 1a BAL ay) ee (xxx). Whence, eliminating yB, we have > _ 2(r—4r?) Hs r(2—7r)- BU (EC) CORAM OE Oe Rt MM) oe ae 4. BOS (xxxi). Mathematical Contributions to the Theory of Evolution. 151 These give y and @ when the parental and fraternal correlations are known. Now, since r is <1, 6? will be imaginary, if 7 be not >4r,?.. Hence we should again need perfect fraternal correlation for 7; to be as large as 0°5. Thus with blended inheritance and little or no assortative mating we cannot get a parental correlation as high as the value 0°5, which actually does occur in my data for both men and horses. We must now consider how the problem will be affected, if we suppose exclusive and not blended inheritance. (7) Illustrations of the Law of Reversion in Exclusive Inheritance.— (i) Let us first consider what happens if we take the chief feature of Mr. Galton’s view, 7.¢., that the likeness to the parent is the beginning, so to speak, of the reversion series. Then y = £ (in the notation of the present memoir). It follows from Equation (i) that: 1-2(2+6)=0, ord=0. Thus by (xviii) : Pl = 2Po and generally Pn = 2Pn-1- _ Equation (xx) to find « now becomes —(l+te)a+}4 =0, while y =4-«4. Thus as soon as we know p, we can find all the ancestral correlations and the whole series of reversions. For example: if p, = 0°4 we should have p2 = 0°2, e = 5, and a?-35a4+05=0. ..a=0°149 and y = 0°351 = 8. ‘Thus in this case 35 per cent. of the offspring take after each parent, and 30 per cent. revert to higher ancestry. Of this 30 per cent. 100 x 0°35 x 0°149, or 5:23 per cent., revert to each of the four grandparents, leaving 9 per cent., about, to revert to great grandparents and higher ancestry still. (ii) Next suppose Mr. Gaiton’s full view to be correct, and that 1/4 of the offspring follow each parent, 1/16 each grandparent, 1/64 each great grandparent, and soon. Then we have— (ef == ¥ — B == 4. Hence from a?—-(1+4e)a+4 = 0, we find G=420 and, prti=. 073. Thus we should have: p; = 0°3, pe = 0°15, p3 = 0°075, &c., or, pre- cisely the same ancestral correlations in the case of exclusive that we have in the case of blended inheritance by the law of ancestral heredity for the special case of y = 1 (see table, p. 149). 152 | # Prof. Karl Pearson. Thus the law of reversion fits no better than the law of blended inheritance the data to which I have referred (in § 6) when we adopt the 1/4, 1/16, 1/64 hypothesis, @.¢., the original form of Mr. Galton’s statement.* 7 (iii) Let us suppose the parental correlation to be 0°5, a value not very far from what I have found for eye-colour in man aad coat-colour in horses. ‘Then by (xvi) Of) al? aaa Putting ¢ = o in (xx) we deduce: i 1 + 26? ee which gives us a = 1 or 4. But remembering the value of 6 we have, using (xxi), ay = —(a-4)? and a+P=1. The first equation shows us that a = 1 is impossible, for # gives y negative. Accordingly we conclude that « = 4 and 6 = 3, while y= 0. Thus reversion is totally eacluded and one-half the offspring take after each parent. In this case the grandparental correlation, p2, is 0:25, the great grandparental 0°125, and so on. ‘The ancestry beyond the parents have no direct influence on the offspring, beyond the fact that they have determined the parents. We are dealing indeed with a case like that investigated in my memoir on “ Regression, Heredity, and Panmixia.”t So far our theory of exclusive inheritance with parental correlation = 0°5 agrees with that of blended inheritance with the same value of the parental correlation. But we have seen that the latter leads to an impossible value for fraternal correlation, 1... one which does not fit the facts. Does perfect fraternal correla- tion necessarily flow from exclusive inheritance without reversion ? Certainly not, for this would connote that all the offspring of a given set of parents would be alike, or one parent in each family be abso- lutely prepotent. This is of course not the fact. Supposing all families to consist of members, and that both 2\2 parents were equipotent in the family, there would be 253 7 1) pairs of brethren alike, out of a total of acid) pairs, or the fraternal correlation would be (4n—1)/(n—1). The average size of a human * See ‘ Natural Inheritance,’ chapter viii, p 149, &c. Mr. Galton there uses the correlation coefficients corresponding to blended inheritance for eye-colour, an exclusive inheritance. But, directly investigated, such values are far from holding for eye-colour. ¢ ‘Phil. Trans.,’ vol. 187, p. 303. Mathematical Contributions to the Theory of Evolution. 153 family is about 4:5; but if we confine ourselves to one sex, we must exclude all sterile marriages and all not leading to two brothers, or two sisters. We might then very well take n = 6, which gives 0°4 for the fraternal correlation. Thus we might expect in the case of exclu- sive inheritance that the fraternal correlation would lie between 0°4 and 1, according as to the degree of prepotency of one or other parent in the individual marriage.* Thus our theory of exclusive inheritance is not, like that of blended inheritance, incompatible with observed facts, 7.¢., high values of parental correlation and values substantially less than unity of the fraternal correlation. But for such cases we must deny the existence of any regular and continuous law of reversion. We should have to look upon reversion, if it occurred at all, as merely an irregular and infrequent phenomenon. On the other hand, if we differentiate the taking after parents from the reversion to ancestry as phenomena of a quite distinct nature, our theory will enable us to surmount, for some cases at least, those difficulties in ancestral correlation, which arise when we take Mr. Galton’s Law in its original form to cover both blended and exclusive inheritance. [I illustrate this from data for the coat-colour of Basset Hounds in the following section. (8) Application to «Basset Hownds.—Understanding that I was desirous of testing my theory on a character which was definitely exclusive, Mr. Galton, with his invariable kindness, at once placed at my disposal his material on Basset Hounds. The reader will remember from the statements in Mr. Galton’s own memoir (‘ Roy. Soc. Proc.,’ vol. 61, p. 403) that these dwarf bloodhounds are either lemon and white, or black, lemon, and white; and here, as in Mr. Galton’s work, they will be classified as non-tricolour and tricolour, or by the symbols Nand T. In dealing with the offspring I was in many cases unable to determine the sex of the dog, as that information is not in the stud book,t+ and all individuals are not again recorded as sires or dams, nor do they possess obviously male or female names. Thus in my | principal tables all the offspring of both sexes are clubbed together. To measure the legitimacy of this, I have formed separate tables of the two sexes in the case of sires and dams. Further, in dealing with great grandparents there were so few of each of the eight individual types alone, that I have formed merely one table, that of great grand- parent and offspring, disregarding the line of descent. * The mother and father may be equipotent on the average, but in the individual family one or other be markedly prepotent. It is to this prepotency of the individual, regardless of sex, that the increase of fraternal correlation beyond 0°4 is very probably due. + Sir Everett Millais, ‘The Basset Hounds Club Rules and Stud Book,’ 1874— 1896. Sr ee ee ee ee n ee a iC er on eT eee ee ee i TT Sassen teh ore 921 | O1Z | STRs0,, Vag cel egz | s}vqo7, £68 C8Z seo | s]8q0], | ect 2 oso 4 or oS ost || 26 Bo. 2 Ne |e ose || tet | oot | “wn a eee ge Se a re, 0 , ee eee = Se a ee oo ee. Sno Se oS a pe te yt || ow | ONT eae 12 Op C61 1 ae EOF 16 G98 ' "18907, || “LN oT ‘sTeqoy, || “LN oT ‘s[BjOg, | LN “L 6 ‘sulidsyQ (9) ‘suladsyQ P (g) — ‘Bultds#O [[V (F) é ‘suldsyQ pur weg ow — ies Cc as Lee Tel 912 | sprog, OSh cat 612 | STSIOL, 628 682 Ors | s[Rio¥g, eae ae ie a 2 | fee Eee 2 (S) ay ee | he, See a a ry | eae ae ae = ne a ra SI PS IN | » 6S 8Z Te iE Nita ee 9IT gg 29 ‘de Nl = SSR AS = | ely | a a ae Se ag ee = ee ene ee | en | a C63 OT Z6L “if @ 19e ell 8hZ 6 fa eI 982 LL L ‘s]8407, | “LN c ‘s[@]07, ||. “LN ‘i ‘s1RI0, || LN ai ‘sudsyQ $ (g) “sutadsyOQ P (Z) ‘SuldszO TLV (1) ‘SutdsyoQ pur oi1g ‘SpunOP] Josseg Ul IMO][00-}v0_ JO oouLzIsOYyUyT Jo soiqey, 154 “SLUU(T 46 Ven) 1 Mathematical Contributions to the Theory of Evolution. 6E9 “818407, SLT | €66 s]Bq0T, iN iL (6) e1¢ IS9T | *IRIOL | Y 919 SST FEE | SIRO, TLS 1 19% ‘UN | 93 &9Z 101 TOs |b ibe Nl eas 177 BG 6F ee fw) —. | | ——S=Ss ————_ |_————_- ————_ _ | ———- @ ———— | | | HOF 11 Came |i |e rae 18 eee. | Snail cos || gst | te S| eee a ae ae Eve 2 TE || eee Nie pean Sper eee : ea: ES) eee c ia ‘VN A eS ‘S180 | ‘L'N L ‘s[e}0L || “LN ‘L | ‘surtdsyO TV (11) “BurzdeZO TLV (01) ‘suytds¥O TLV ‘BulldsyoQ puv quoredpuris 4voiy ‘SulldsyQ) pue weg s,weq ‘SutldsyOQ pue stg suey 91g Z81 T6E | SPIO, PLS 181 6g | si07, 861 OL Q20 | iN Pa Bg 1% 8 | IN | 3, ee — — ee | | Cr Ge ie Bile 89S AD gaa. | Jet | -d9e_! ‘1 2 || — —|—__-—_| 2 sye07, | “IN ‘L S1@30L || “LN L ‘suladsyO IV (8) ‘suladsyQ ITV (1) ‘SutsdsyQ pue we s,o11g ‘SuLldsyQ pur WIG so1Ig *SOITS 8 BCT Prof. Karl Pearson. 156 06062 VOR 9BLET "S[BIOT, 628 96L8T | SI"4OL 99PE 828P LN 88h 8688 \L ‘TN | ‘L “BUILIG IAT (91) ‘(opis soatg) sBurqig eH B[BIOT, 896 O9T ‘LN Surqqig seat (s1094T[ JUOLOYIP) SAuTTqIg OTOU, cEP 8[84O,L. s0L \L'N PCE iL \L “Sullgig puosag ‘Bur[qig puovag LIP HOST “B[BI0 7, POST 809% | s[RI0,, CSL, 208 “LN rags: YOLT i lo ht si "BUILGIS ali (FT) VEO6 OL 69ST | ST*I0%, Shy LI LN 118 206 ‘lL “eke 7 ‘BUIGIS FANT (21) ‘(10941] OUNES) SHUTTGIG OTOYM ‘SUITGIg Pucoag ‘Su[qig puoseg Mathematical Contributions to the Theory of Evolution. 157 The extraction of eleven tables from Mr. Galton’s data papers I owe to Miss Alice Lee, D.Sc. ) Offspring take more after the dam than the sire, but ¢ offspring more than 92 after the sire, and ? offspring more than ¢ after the dam. In other words, the parent hands down its characteristics more strongly to its own than to the opposite sex. (c) Curiously enough, the sire’s parents seem to have more influence than the dam’s. In particular the dam’s sire has, within the probable error of our determinations, no influence at all. In the unchanging line of descent, the dam’s dam has more influence than the sire’s sire, which is what we should expect from (a) ; but (a) also makes the male element of much less importance than the female, and so the dam’s sire insig- nificant as compared with the sire’s dam. The final result is thus to give a slight preponderance to the sire’s over the dam’s parents. (2) Collateral Inheritance. (a) The degree of resemblance between puppies of the same parents is not greater when they are of the same than when they are of differ- ent litters. Itis clear, however, that we have only been able to find comparatively few pairs of whole siblings from different litters, and the difference between 0:5084 and 0°5257 is of the order of the probable error of the differences. With greater numbers, possibly a more sensible difference might be found for the correlation of siblings from the same and different litters. At present there seems nothing to warrant the idea that puppies from the same litter have the high degree of resemblance which we find between twins in the case of mankind. (6) A comparison of the correlations for half siblings on the dam’s side and on the sire’s side again emphasises, if the breeding be straight- forward, the great prepotency of the dam in the matter of coat-colour. The fact that we have upwards of five times as many pairs of half siblings on the sire’s side as on the dam’s side shows how large a fashion there is in selecting sires. It is possible that largely used and # © Roy. Soc. Proc.,’ vol. 61, p. 404. The Table II, p. 410, requires interchange of headings, as already pointed out by Mr. Galton. Mathematical Contributions to the Theory of Evolution. 159 possibly overworked sires lose some of their hereditary influence, while not losing their power of fertilising the dam.* (c) The great reduction in the degeee of fraternal correlation when we turn from whole to half siblings is very remarkable, and is, at any rate for half siblings on the dam’s side, not very explicable. Had we assumed the parental correlation to be 0°3507, and found y from (xxvii), 7.c. = 1:4722, we should have deduced from (xxviii) for the fraternal correlation the value 0°5236, which is in fair accord- ance with the observed result for whole siblings. But, as we have seen, (xxvii) and (xxviii) belong to a theory which gives very poor values for the grandparental and great-grandparental correlations, 7.¢c., 0°1753 and 0:0877, instead of 0°1326 and 0:0402. Further, we should on that theory have expected the average correlation for half siblings to be half the value above, since one-half of the common ancestry 1s cut off, z.¢., 0°2618, and not 0°1646, as it actually is. Thus the fra- ternal correlation does not appear to be in accord with the theory of blended inheritance. Its determination in the general case of exclu- sive inheritance with reversion seems a problem of considerable diffi- culty, which in this case is rendered much greater by the immense prepotency of the dam, so that it would seem very desirable to differ- entiate the sexes when dealing with the resemblance of siblings revert- ing to ancestral types. (9) Application of the Theory of Reversion to Busset Hounds.—We have for mean values p, = 0°3507, pe = 0°1326, p, = 0:0404. Now these correlations certainly do not obey the relation p, = 2p., pz = 2p, required, when we take (xxiii) to govern the law of ancestral heredity (cf. § 6 (iii) ). A glance at the table on p. 149 will show that such a series of p’s as the above cannot fit into it. Still less do they appear consonant (except to the first roughest approximation) with Mr. Galton’s form of that law,7.c., y = 1. Nor do they satisfy for the same reasons the law of reversion when we start the reversion series from the parents, 7.¢., put B = y as in § 7 (ii). Accordingly, I have tried to find what would be the value of ps, if p, and pz had the values given above, and our generalised law of reversion were correct. Turning to § 3, and substituting in (xviii) and (xix) for py, and pz we have : 6 = 0'286336, 6.3, 0-698 761; * There is also to be considered the possibility of error of the record in the sire’s case. Given a large stud of hounds and servants of average carelessness, and a bitch may easily go astray even after she is lined by the dog required. Sir Everett Millais, in a lecture on Telegony, delivered at St. Thomas’ Hospital, in 1895, stated that he knew of “quite two dozen such examples resulting in supposed telegony.” “The master is the last person to whom such little lapses of duty are confided.” But if two dozen mésalliances can be palmed off as telegony, how many alliances within the blood may be conveniently overlooked ? 1G Prof. Karl Pearson. Hquation (xx) becomes a2 ~ 156851240 = 0 or, a = 0°445057, Hence y = 0:019594, 8 =, Or34d2 7. This effectually shows us that for this case y cannot be taken equal to £, or the reversion series started from the parents, Further, we reach from Equation (xi) : pn = 0°30124(4)" + 0°698761(0-286336)". Such a value again demonstrates that in this case the ancestral correla- tion differs totally in form from what might be deduced from the theory of blended inheritance, 7.c., it shows us how we must distinguish between a law of regression and a law of reversion. Putting n = 3, we have p3 = 0°0541 for the calculated value of the great grandparental correlation. This may be, I think, considered in satisfactory agreement with the observed value 0:0404. Had we deter- mined our 6 and c by the method of least squares, so as to satisfy the three relations p= 1-408, pp = EL - 40, pp = ML -) as Closely as possible, we should have got, of course, still more accord- ant results. We can now put down our general conclusions : 63'256 per cent. of Basset Hounds take after their parents .......... (50) 3'488 e 3 revert to grandparents ............ (25) 3°105 5 ‘3 By great grandparents....... (12°25) 2°764 3 és 5 great? : 5 $,olatenecca et (dare) 2°460 3 “ e great® % Mee ace teil (ial B43) 2-190 m es i great? ie ose wee el ob2p) 1-949 5 b, f great? 53 OE eee (0°78125) 1-735 ea se » great? ni weeeees (0390625) 14:053 per cent. of Basset Hounds revert to still higher ancestry (0°390625). Now the divergence here from Mr. Galton’s original statement of the law is most significant; I have put in brackets the percentages deduced from that statement. In our case we have a comparatively small reversion to each generation of ancestry, but the percentage, 1°7,is still sensible in the case of the 8th ascending genera- tion. In Mr. Galton’s case we have very substantial reversions to grandparents and great grandparents, but the rate of diminution in- stead of being the loss of about 1/9 at each stage is 1/2! As a result, the reversion to the 8th ascending generation is less than 0-4. It can- Mathematical Contributions to the Theory of Evolution. 161 not be denied that the difference is of extreme biological interest. In the former case we have a comparatively small total reversion widely spread; in the latter a much larger total reversion concentrated on the first stages of ancestry. Which system is more accordant with facts? It needs far wider observation and experiment than are yet available to settle this. So far we can only say that the former case covers Mr. Galton’s as a special sub-case, and that the data for Basset Hounds appear capable of treatment under the wider rule, but can only be fitted with some straining to the special case. Are there or are there not physiological reasons for supposing that resemblance to a parent arises from a different source from reversion to an ancestor ? Here reversion to an ancestor must not be measured by cases of resemblance to an ancestor, for a portion of this resemblance is due to common likeness to the parent; we must approach the matter from the standpoint of cases in which the offspring inherits a character like that of the grandparent or higher ancestor, and wnlike that of the parent.* Current use of the term “reversion” at least justifies us for the time being in not speaking of all inheritance, including likeness to parents as reversion, and in our theory we may be permitted to dif- ferentiate parental and reversional inheritance of exclusive characters, if we find it needed by our numerical data. Here, as elsewhere, we sadly need a widely extended range of observation and experiment. (10) On the Variability of Basset Hounds having regard to Sex and Pedigree.—We have already indicated that our justification in applying the methods of normal correlation to coat-colour is considered in another memoir. We merely suppose at present that there is some variable, following approximately the law of normal variation, on which the coat-colour can be thrown back. This being so, let / in terms of this variable be the distance of its mean value from the division line between tricolour and non-tricolour, and let o be the standard devia- tion of this variable for the same group. Then if x = h/c, we easily deduce by aid of tables of the probability integral from the correla- tion tables in § 8 the following results :— * Of C and D brown-eyed parents, A and B, the offspring, have respectively brown and blue eyes. Is the process by which B gets the blue eyes of his grand- father or great grandfather qualitatively different from that by which A gets the brown eyes of his parents? The problem can hardly be answered at present, but I see no @ priori reason for a negative. 162 Prof. Karl Pearson. Table of Values of y. Xe DIGS. siete ae 1:0806 with all his offspring...... 0°3891 ah eeeeei gers. eke O82 3%, hiss TAG coe 0:42.42 MC Steed ic oa kU ee IQ i ete fee ee 0°3611 Mama bodt. fen. ae 01575 ,, all her offspring... 0°3954 Be aver Subir ear bs! 01315 *,, her ig) Ga 0°3521 Ba ae mapas 01499 1, 4 «02 ee 0:3186 DINE SSIs Hattie hace 1'3372.,, : all his oftspring wee 0°4808 seth (ORIN ieemieac!. & 0°4023)) 4,°:. 5; her ae ee 0:4790 DD ain's: Sine wereees, +a LAD SGueeyy Gs whiis sf ae 0°4909 Dani's damn wre alae 01134... 5, herd eee 0°4790 Great grandparents .. 05824 ,, all their offspring... 0°6592 Whole siblings, same litters)...::0 1). 1.1: See ee 0°3108 io » | different litters :.2..:).00. eee 0°2980 Halt siblings, sire’s side ..00.07.0. s.02:. 0. er 0°3143 és ve damis side aeite. Haas ath « ages eee: ae 0°3190 Now, if there were no secular change due to artificial or other selec- tion, 4 would remain the same in each generation, and therefore o = h/x would give a proper measure of the variability due to sex or to relative position in ancestry. The table at once suggests a number of interesting points, which I proceed to note. (a) Turning first to the offspring, as given in the second column, we observe that the dogs with the longest pedigree have the largest x. ‘We have, in fact, the values of 1/y for pedigrees stretching to great grandparents, grandparents, parents, and merely to brethren, respec- tively, as 1517 : 2:064 : 2°550 : 3°221, or, roughly, we have a geometrical series with the ratio of about 1:3. Now this result may be reached in more than one way, either (i) decreasing h, or (ii) increasing o. Increase of / would signify that, for the longer pedigree, the mean of the quantity on which the coat-colour depends is being thrust further into the tricolour section ; decrease of c would signify greater concentration in the tricolour section within which the mean lies. Whether the longer pedigree signifies the more modern hounds, or the more careful preservation of ancestors’ names in the more fashionable hounds, we reach practically the same conclusion : the process of breeding is emphasising melanism. Further, by com- paring the g and ¢ offspring in the case of both sire and dam, we conclude that this process is sensibly more significant for the male than the female offspring. As very often only one or two puppies out of a litter are recorded in the ‘Stud Book,’ this apparently artificial selec- Mathematical Contributions.to the Theory of Evolution. 163 tion, which is more stringent for the males than the females, appears to be, theoretically at least, a mistake when we remember that the potency of the female is thrice that of the male in coat-colour. Within the groups of grandoffspring and siblings the differences are hardly significant enough for special conclusions to be drawn. (b) Turning next to the parental groups, we see that (i) the sires and dams, neither of which can be considered to form a more modern group than the other, have yet remarkably different values of x, that for the sire being about seven times as great as that for the dam. The sires are thus far more stringently selected than the dams, and a great deal of this difference must undoubtedly be due to the lesser variability of the sires. Here again the breeders, if they are selecting at all for coat-colour, would appear, at least theoretically, to be in error. (ii) Grandsires, ¢ and 2, appear to be less variable than the sires, and granddams, ¢ and 2, less variable than the dams. This may be due to the original paucity of the breed, or be an instance of the general rule to which I have elsewhere referred, 7.c., that parents are a selection out of the general population, and so less variable than their offspring. (c) But this rule meets with a remarkable exception in the case of 2 parentage; both granddams and dams are more variable than their off- spring, and very significantly so. An examination of dam and female offspring shows that the ? offspring have a value of x double as great as that of their dams. With few original dams, it is difficult to under- stand how they could be more variable than their offspring. Consider- ing the great prepotency of the dam, it is difficult to attribute this increase of x entirely to the action of the less variable sire; one is more or less forced to believe that there is a process of stringent selection of the offspring which are entered on the record going on, and that thus a group of dams possibly fairly variable, and with a not very marked tendency to melanism, is represented in the next generation by offspring of a more stringently selected character ; the stringent selection of sires may have contributed, but can hardly be the sole source of this change. A further conclusion is worth noting: Parents, whether male or female, when they have male are apparently more variable than when they have female offspring. (10) General Results. (a) The laws hitherto propounded for blended inheritance do not appear to cover the cases of exclusive inheritance, ¢.g., such cases as eye-colour in man, coat-colour in horses or hounds, &c. (b) The law of ancestral heredity must be distinguished from a law of reversion. Neither seem to fit the facts if we adopt the amounts of heritage, 1, go, sb, &e., from parent, grandparent, great grandparent, &c., originally taken as a first approximation by Mr. Galton. 164 Mathematical Contributions to the Theory of Evolution. (c) That the mean correlation of an nth parent with the offspring is one-half that of an (nm —1)th parent also appears doubtful. (This would follow if reversion were started from the parent.) (d) Testing theory by the case of Basset Hounds, we find much difficulty, owing partly to the great prepotency of the dam, and partly to the large amount of artificial selection which is evidenced at every turn, and obscures what may be termed the natural laws of inheritance. (¢) There is an urgent need to widely extend our knowledge of heredity by new experiments and observations on other organs in different races. Facts are of the first necessity at the present time, and facts collected on a large scale for a wide range.* * It may be of service to indicate to would-be investigators what has already been done or is now in hand :— In men :— (2) Stature (direct to first degree and collateral, Frater, (b) Head index ( s ” (c) Span and forearm (direct to first degree and ooMatedall fraternal). (d) Bye-colour (direct to second degree, collateral, fraternal and avuncular). (e) Shape of head, physique, intellectual capacities, tastes (collateral only). (f) Fertility (direct to second degree). (g) Longevity (direct and collateral, fraternal). In horses :— (2) Coat-colour (direct to second degree and collateral). (<) Fecundity (direct to second degree and collateral, fraternal and ayun- cular). In hounds :— (j) Coat-colour (direct to third degree and collateral). In moths :— (Z) Wing-markings (direct and collateral). In daphnia :— (2) Shape of spine (direct and collateral). In all these cases the coefficients of correlation have already been worked out, or material is being collected to determine them, by Mr. Francis Galton, Professor W. FE. RB. Weldon, Dr. Warren, or by my collaborators and myself at University College. Hence I would impress upon others to take as far as possible widely different characters in widely different races. Above all, cases in which artificial selection plays a great part, z.e., dogs, fancy pigeons, &c., ought to be avoided. Effects of Strain on the Thermo-electrie Qualities of Metals. 165 — February 1, 1900. | The LORD LISTER, F.R.C.S., D.C.L., 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 :— I. “A Case of Monochromatic Vision.” By Sir W. bE W. ABNEY, GB. ERS. Il. “Thermal Radiation in Absolute Measure.” By Dr. J. T. BotroMLeEY, F.R.S., and Dr. J. C. BEATTIE. III. “Electrical Conductivity in Gases traversed by Cathode Rays. By J. OC. McLennan. Communicated by Professor J. J. THOMSON, F.R.S. IV. “Researches on Modern Explosives. Second Communication.” By W. Macnas and E. Ristori. Communicated by Professor Ramsay, F.R.S. V. “On the Influence of the Temperature of Liquid Air on Bacteria.” By Dr. ALLAN MACFADYEN. Communicated by Lorp LIsTER, P-R.S. 99 “On the Effects of Strain on the Thermo-electric Qualities of Metals. Part Il.’ By Macnus Macniean, M.A., D.Se. Com- municated by Lorp KeEtvin, G.C.V.O., F.R.S. Received _November 22, 1899,—Read January 25, 1900. A.—Thermo-electric difference between free wires and wires previously subjected to longitudinal extension and lateral compression, by drawing them through the holes of a draw-plate (§§ 1—7). §1. In Part I of this paper, read to the Society on 2nd February, 1899, the object of the experiments was stated to be the determina- tion of the magnitude of the thermo-electric effects, obtained from any one metal strained and unstrained. The results then given were obtained from two wires of the same material, one wire being pre- viously drawn through a draw-plate, so as to reduce it in size from No. 18 standard gauge (0°122 cm. diameter) to about No. 24 stan- dard gauge (0:0559 cm. diameter). The arrangement of the experi- VOL. LXVI. P 166 Dr. M. Maclean. On the Effects of ments to measure the thermo-electric effect is shown diagrammatically in fig. 1. One junction of the wires was kept in a glycerine bath which could be heated by a Bunsen burner. This junction was tied by a fine copper wire to the bulb of a thermometer T. The other ends of the wires were joined to short copper wires, which served as terminals of the low resistance galvanometer used in the experiments, These junctions were wrapped in paraffin paper or cotton wool, which contained the bulb of a thermometer T’ reading half degrees from 0° C. to 25° C. A paper screen S was hanging vertically between the Bunsen burner and the thermometer T’ and the galvanometer, to pre- vent any heat from the flame reaching the rest of the circuit by radiation. ‘These precautions were taken to make certain that all junctions, except the hot junction, would be at the same temperature. The sensitiveness of the galvanometer was 0:09 mikroampere per division, and as its resistance was 1°5 ohms, the electromotive force at its terminals was 0°135 mikrovolt per scale division. §2. The metals for which results were given in Part I were copper (six specimens), lead (two specimens),-platinoid, german silver, reostene, and manganin.* The present paper gives the results of similar experiments made on specimens of commercialf and pure lead, obtained from Messrs. Johnson and Matthey ; and specimens of annealed steel, of aluminium and of nickel. § 3. The method of experimenting was to take a piece of the wire and draw it through a few holes of a draw-plate, so as to reduce its cross sectional area to about a quarter. Then two pieces of the wire, one drawn and one undrawn, each 60 cm. long, were joined as in fig. 1. The glycerine bath was very slowly heated by the Bunsen burner. When there was a rise of temperature of 5° C. the Bunsen burner was drawn slightly aside, so as to give as much heat to the glycerine bath as it lost by radiation. When both the thermometer reading and the spot of light on the scale were seen to be steady, the readings were noted. The circuit was then broken and readings taken of the gal- vanometer zero and the thermometer T’. ‘The circuit was again com- * Dr. Anderson, Chemical Laboratory, the University, Glasgow, gave me the following analyses for reostene and for manganin :— Reostene. Manganin. | Si ecg psismeaie Oil per cent, DI) es «steele 0-073 per cent. eins sls SSE OOD sh a5 He ieee cscs, OD : NW seis ee rele Oe Tl, CU weccceesee S662 Fs Min! Givees'se ce M2 SY 4, Min wie se o'e'ealne as Commies Ni eeeoaeeeoee@ 3:°261 33 Total.... 98°30 Total.... 98°585 + Dr. Anderson analysed the commercial lead and found it contained 99°12 per cent. of lead. Strain on the Thermo-electric Qualities of Metals. 167 pleted, and readings taken when there was 5° C. further rise of temperature, and so on up to a difference of about 100° C., the greatest difference tried in these experiments. A curve was plotted with Mirror Galvanometer SSS ess es "” 4 Screen Mv AG WaVS Cu Cee weeewe a OS 8 res 8 differences of temperature between the hot and the cold junction as abscissee, and currents through the galvanometer as ordinates. The mean current per degree difference of temperature as found from each curve is given in Table I. Table I. * Condition of conductor. Current in mikro- Conductor. Current from 1 to 2 ampere per degree through the hot-junction. up to 100° C. | | 1. Draws dhceseee:. . | é 7 Annealed steel.........! 2. ipa cig” aa ts 0 -0567 | j | | Aluminium............! A pee ia 0 -0065 Sr y rics a be 0-213 Lead, commercial .... 7 ; amish = aN S i 0 °0124 Vie. | Undeawits. os cc se, ; Lead, pure eseeeeeeensese ; ran Drawn .. neh aes eae } 0 :0036 ! 4 4 | § 4. The steel wire was annealed by coiling it round a large cast-iron ball, which was heated in a bright coal fire for about an hour. After ‘being taken out of the fire, the ball, with the iron wire round it, was P 2 168 Dr. M. Maclean. On the Effects of allowed to cool slowly for about an hour and a half in the ashes below the fire. It was previously found that heating the wire to red heat by an electric current and allowing it to cool slowly did not anneal it. § 5. The resistances of all the undrawn wires, and the specific gravities of both the undrawn and drawn wires, were carefully deter- mined by the usual laboratory methods. The results obtained are given in Table II. Table II. Resistance of the undrawn : wires in C.G.S. units Cross section Specific at 14° C. of undrawn EN) Conductor. oa ies undrawn seus and drawn Per cm. wires in sq. cm. : 1 wires. Poe ae long, ; weighing a gramme. Sieelmmesled 2. { ve sy \ 13,900 108,100 Seat 0 °04594. 2°8 | Aluminium........ { 0 “01697 2-796 } 3,546 9,931 easeeeie ue, ee { Sous cee \ 9,430 83,920 | Lead, commereial .. { Oe ai ne \ 20,560 233,100 | Lead, pure .....e.. { Scgaaee s ee \ 20,300 230,200 awe $6. By multiplying the current per division given in Table I, by the total resistance in the circuit, caleulated from the results in Table II, the thermo-electric difference per degree between drawn and un- drawn wires is found. The numbers are given in Table III. Table IIT. Thermo- Resistance: in ohms electric of 60 cms. of wire. | Total difference | | resistance | Total in mikro- | Conductor. ‘external | resistance | volt per | | to galvano-| in circuit. | degree C. | | meter. difference | | Undrawn. | ‘Drawn. of tem- | perature. | Steel, annealed....| 0°1111 0°2015 0 °3126 1°813 01028 | Aluminium.......| 0°0047 0°0125 0 ‘0172 1°517 0 -0099 Nickel...........| 0°0480 0 '2287 0 °2767 ae br 0 °3784, Lead, commercial .| 0°1077 0 °4820 0 5897 2°09 0 :026 Lead; pure.......| 0°1059 0 °5162 0°6221 2°12 00076 Strain on the Thermo-electric Qualities of Metals. 169 §7. The thermo-electric difference between glass hard tempered steel, annealed steel, and unannealed steel, was _~ by similar ex- periments to be :— . Glass hard steel . Unannealed steel . Unannealed steel ‘| 0-18 . Annealed steel.. . Glass hard steel beer . Annealed steel... hr ‘5 mikrovolts per degree C. 39 79 bo = bo ee be 23 99 The direction of the current through the hot junction was in every case from hard steel to soft steel. B.—Thermo-electric difference between free wires and wires previously permanently elongated by longitudinal stresses (§§ 8—10). § 8. Attempts were now made to determine the thermo-electric difference between free. wires and wires previously permanently elon- gated by a longitudinal stress. It was found difficult to elongate the hard wires permanently to any appreciable extent before they broke. Several methods for stretching the wires were tried, and the method finally adopted, was to take two pieces of ‘stout copper rod, bent into the shapes shown at A and B in fig. 2, and to wind the wire to be Fie. 2. eee ae as eS ES stretched several times round A and B. The end A was clamped in a fixed vice and the end B fixed in the clamp of a screw arrangement. By turning the screw the wire was stretched tight. Two ink marks were then put on the wire at C and D 60 cm. apart. The screw was very slowly turned, and the distance between C and D measured until the necessary elongation was produced or until the wire broke. The wire generally broke where it lay tangentially to either rod A or B. §9. The greatest percentage permanent elongation that could by this method be got in hard drawn copper, manganin, nickel, and German silver, was 0°7, 0°5, 0:7, and 0°5 respectively. The thermo- electric difference between the stretched and the unstretched wires was then determined, as described in § 3, and the results are given in . Table IV. 170 Dr. M. Maclean. On the Effects of § 10. It will be noticed that the current is from unstretched to stretched, through the hot junction for three specimens of copper, and from stretched to unstretched through the hot junction for other three specimens. The probable explanation of these results is suggested in §§ 13, 14, 15. ‘ ; Table IV. | Current; Thermo- Condition of Per- in Total electric conductor. centage | mikro- noaate difference in Bho dnctor Current from | perma- | ampere ance jn |. mikrovolt : 1 to 2 nent per - + | per degree through the | elonga- | degree Ba See centigrade hot junction. | tion. up to "| difference of 100°C. temperature. Messrs. Johnson and Matthey :— (a) Copper, elec- { | 1 stretched \ i ‘ ; 2 trotype [| 2 unstretched 10.1) 0 Olt ae ee : (b) Copper, for { 1 stretched \ 1:0 | 0-003 1°519 0 -0046 alloy 2 unstretched c) Copper, com- f | 1 stretched : -mercial . . 2 unstretched 22 Messrs. Glover :— 1 unstretched : ; ; ‘ (a) Copper, hard { opr 2 y 0-7. | 0-0018 | 1°519| 0-0027 1 unstretched (2) Copper, soft 2 stretched Copper, labora- he 1 unstretched tory | | 2 stretched 00051 | 1°548| 0-0079 1°5 | 0°005 1°531 0 -0077 | 1°5 | 0°0024 |} 1°518 0 :0036 1 unstretched ” 2 stretched i unstretched 2 stretched i i i 1 hed f stretche \ j i i i 20°0 | 0°0174| 1°52 0 '0264. Reostene ... 2°0 | 0:009 2°312 0 :0208 Platinoid . 1:0 | 0°047 1 °937 0 °0910 : unstretched 1 stretched 4 unstretched 1 stretched 2 unstretched © 1 stretched 2 unstretched 1 unstretched 2 stretched 0°5 | 0°056 1°835 0°1027 Manganin.... 0°5 | 0:036 1-924 00693 Aluminium .. 1°0 | 0°01385 | 1°51 0 :0204 Nickel .... 0°7° | 0084 |: 1°596 0°1341 ai ah German silver .. { | a Hel C.—Thermo-electric difference between free wires and wires under stress, : producing (1) temporary elongation, (2) per manent elongation (§§ 11—18). § 1, The arrangements shown diagrammatically i in fig. 3 were now - made to determine the thermo-electric difference between free. wires : and wires while (a) under stress, stretching them within their limits of elasticity, and (b) under stress, stretching them beyond their limits of elas-. Strain on the Thermo-eléctric Qualities of Metals. 171 ticity. AB is a brass tube through which steam from a kettle is allowed to pass. The wire under test is wound round this brass tube three times, and then round a small brass tube in the triangular frame below, and then to one of the terminals of the galvanometer. The wire is thus quite continuous, from one terminal of the galvanometer to the other terminal. Some cotton wool is loosely packed at the hot junction of the wire to ensure that the temperature of the wire is at steam tem- perature. The weight of the triangular frame (two sides wood and lo Galvanometer. one side brass tube) with its hook for hanging weights on was 220 grammes. ‘This is the smallest weight used for each wire, every one of which was about No. 30 8.W.G. (diameter, 0°0315 cm.). The. object of the small brass tube in the triangular frame was to keep the temperature of the cold junction at any determined temperature by allowing water or other fluid to flow through it. In all the experi- ments hitherto made, the temperature of the cold junction was taken as the temperature of the air at the time of each observation. § 12. The experiments were performed as follows :—The wire was put into the circuit, as shown in fig. 3. After steam was allowed to pass through the tube for some time, the galvanometer reading and the air temperature were taken. The circuit was then broken, and the metallic zero of the galvanometer was noted. The circuit was made, and a weight was added on to the hook of the triangular frame. Three readings of the galvanometer were now taken : (1) with the weight on, (2) with the weight off, and (3) with the circuit broken. A heavier weight was hung on, and other three readings taken, and so on to the heaviest weight used in the experiments. -§ 13. The readings of the galvanometer were in the same direction 172 ~~ Dr. M: Maclean. * On the Effects of for all the wires tried with weights on and off, except for soft copper and iron. The greatest permanent elongation produced in any of the hard copper wires experimented on was 0°17 per cent., and for this permanent elongation the reading on the galvanometer was in the same direction for weights off and on, though always greater for the latter. § 14. For the soft copper wire (‘Table [X below) the readings were in the same direction for weights on and off, up to a permanent elongation of 1 per cent. After a permanent elongation of 4°72 per cent., the current with weight on was 0:00103 mikroampere per degree from stretched to unstretched through the hot junction, while with the weight off, the current was 000075 mikroampere per degree from wnstreiched to stretched through the hot junction. For iron wire the current was in the same direction for weights on and off, up to a permanent elongation of 0°35 per cent. ; but after a permanent elongation of 3°41 per cent. the current with weight on was 0-:00461 mikroampere per degree from unstretched to stretched through the, hot junction, and with weight off, 0°0069 mikroampere per degree from stretched to unstretched through the hot junction. § 15. In ‘Mathematical and Physical Papers,’ vol. 2, p. 270, § 109, Kelvin says :—“I have thus arrived at the remarkable conclusion that when a permanent elongation is left after the withdrawal of a longi- tudinal force which has been applied to an iron or copper wire, the residual thermo-electric effect is the reverse of the thermo-electric effect which is induced by the force, and which subsists as long as the force acts.” It seems (1) that for small longitudinal strain in copper or in iron the direction of the current through the hot junction is the same, whether the force which produced the permanent strain is on or off, (2) that as the permanent elongation is increased by increased longitudinal forces, a stage is reached which gives zero current when the forces are removed, and (3) that for greater longitudinal forces and permanent elongations the direction of the current is opposite, with the pulling forces off andon. It seems, in fact, that the permanent elongation must exceed a definite limit, to produce reverse thermo-electric effects with the longitudinal force on and removed. I hope to further investigate this point and to report the results to the Society. § 16. The galvanometer used for the investigation of these temporary and permanent strains was one of the Kelvin recorder pattern, namely, a movable coil between the poles of a strong permanent magnet of circular form. ‘The coil had 81 turns and a resistance 14°94 ohms at 17°-5 C. Its constant was determined in the usual way, and found to be 0:029 mikroampere per division of the scale. $17. To find the stress-strain diagram, experiments were performed on a specimen of each wire, in the following manner. ‘Two pieces of the wire were passed through two small holes in a metal plate, and Strain on the Thermo-electric Qualities of Metals. 173 soldered at the back of the plate. This plate was fixed to a horizontal support at a convenient height. One of the wires had a half millimetre i eee Yr N Manganin. 4 2 0 Percentage permanent set x io~*. Li TORY | lron Wire. 2 N rey 9) v Ny oy ~ QoQ ‘P4BAMPOLYIU 5_Of ¥ JUBMND scale near its lower end, and a weight hanging on it to keep the wire straight. The other wire had a scale pan and a pointer in the manner generally used in laboratories for finding the Young’s modulus of 174 Dr. M. Maclean. On the Effects of materials. By putting weights into the scale pan, and taking them out, and noting the readings of the pointer on the scale, the temporary and permanent elongation i in the wire for different weights were found. The numbers are given in the second and third columns of Tables V to XIV. The currents in the fourth columns are calculated from the deflections of the galvanometer when the stated weights are on the wire, and the thermo-electric differences in the fifth column are found by multiplying the currents in the fourth column by the total resistance in the circuit in each case. § 18. The numbers in the tables are plotted in curves with percentage permanent elongation as abscisse, and thermo-electric differences as ordinates. The thermo-electric difference between free platinoid and strained platinoid, rises rapidly with the permanent elongation of the strained wire. For manganin and lead, both commercial and pure, it is very nearly constant up to a permanent elongation of + per cent. Mr. Alexander Wood, Thomson Experimental Scholar in the Physical Laboratory, The University, Glasgow, has rendered valuable help in the experimental work and in the calculations. Table V.—Electrotype Copper. Current from Stretched to Unstretched. Temperature difference = 87° C. Total resistance in circuit = 15°56 ohms. Current in Thermo-electric Percentage Percentage Total weight ye ne eran mikroampere difference in in grammes. ae Pee ce Se per degree mikrovolt 5 i 8 with weight on. per degree. 250 0-06 ys 0-00020 0°00311 | 500 0-1 es 0 :000267 0°00415 | 750 0°137 of 0 :000567 000882 : 1000 0-18 0-036 0 -000934: 0 01452 1250 0°21 0°057 0 :001200 0°01867 | 1500 0 °264. 0°072 0 °00137 0 -02127 1750 0°318 0-104 0 001866 0 °02905 2000 0 °348 0 °144 0 002 0:031138 Strain on the Thermo-electric Qualities of Metals. Ls Table VI.—Commercial Copper Wire. Current from Stretched to Unstretched. Temperature difference = 88° C. Total resistance in circuit = 16°66 ohms. . Current in | Thermo-electric Total weight | pee eee mikroampere | difference in : emporary permanent d esc ate pu eee elongation. elongation. pa ia ie er aa with weight on. per degree. 250 | 0-05 : 0°00119 0°02 500 0-1 ee 0 °00142 0 °0236 750 0°14 ve 0 -00138 0°0231 1000 0°19 ay 0:00148 0 :0247 1250 0°23 0 °024 0 °00152 0 -0253 1500 | 0°28 0°058 0 °00161 0 0269 1750 | 0-32 0-083 0°00175 0°0291 2000 | 0°37 0°112 0 -00191 0°0318 Table VII.—Copper for Alloy. Current from Stretched to Unstretched. Temperature difference = 87°C. Total resistance in circuit = 15°62 ohms. Current in | Thermo-electric Total weight oe pee mikroampere difference in . emporary permanent i mk 1 ea elongation. elongation. Serene’ ee | oF with weight on. | per degree. 250 0:08 se 0 000333 .0:00521 . 500 0°145 0°041 0 000533 0 00833 750 0-176 0:059 0 -000600 0 :00937 1000 0°215 0 °076 0 -000833 0°01302 1250 0 °256 0 °105 0 001167 0 °01823 1500 0°285 0°118 0 °001324 0:02031 _ 1750 0 340 0°142 0°001433 0 02239 0°170 0 °001600 0 02500 2000 0 °402 ps a 176 “Dr. M. Maclean. Table VIII.—Glover’s Hard Copper. Current from Stretched to Unstretched. On the Effects of Temperature difference = 86°°5 C. Total resistance in circuit = 15:59 ohms. Percentage Percentage Cee Total weight ‘ 8 = mikroampere TAI emporary permanent a. dewiee an grammes: | elongation. elongation. 0 eas with weight on. 250 Se pe 0 000335 500 0 006 oe 0 -000469 750 0°014 Py 0 -001106 1000 0:019 0-006 0001241 1250 0°0285 0°0135 0 -001375 Table [X.—Glover’s Soit Copper. Current from Stretched to Unstretched. ‘Thermo-electric difference in mikrovolt per degree. 0 005226 0 -007316 0 ‘01725 0 °01934. 0 °02143 Temperature difference = 87° C. Total resistance in circuit = 15°56 ohms. . Percentage Percentage fee Total weight S 8 mikroampere in grammes. posporary permanew: per degree elongation. elongation. with weight on. 300 oe oe 0°00020 500 0°14 0°03 0 -00083 750 0°28 0°105 0 -00123 1000 1°18 1:00 9°00143 1200 4:°84 4°72 0 -00103 Table X.—Reostene Wire. Current from Unstretched to Stretched. Thermo-electric difference in mikrovolt per degree. 0 °003113 0°01296 0 °01919 0 °02225 0 -01608 Temperature difference = 86°5 C. Total resistance in circuit = 43°39 ohms. Current in Total weight peace makes: mikroampere ce ae He tion ais ation per degre: Oe Ewa 8 * |with weight on. 250 0-037 oe 0 ‘002682 500 0°085 ieee 0 °003521 750 0°152 0 °037 0°003521 1000 0 ‘220 0:073 0 003588 1250 0°257 0-104: 0003721 1500 0318 0°134 0 ‘003957 1750 0 °417 0°201 0 -004526 Thermo-electric difference in mikrovolt per degree. 0°1163 0°1528 0 °1528 0°1557 0 ‘1614 0°1717 0°1963 Strain on the Thermo-electric Qualities of Metals. | 177 Table XI.—Platinoid Wire. - Current from Stretched to Unstretched. Temperature difference = 86°C. ‘Total resistance in circuit = 29°89 ohms. C t i | Th lectri ‘ Current in ermo-electric Total weight oye ir ansg 5 Eee mikroampere difference in in grammes PeTAOPArY porns per degree mikrovolt Br *| elongation. elongation. with weight on. per degree. 250 | 0 °035 ee 0°003541 0 °1058 500 0-071 RS 0 -003979 01190 750 0°101 a 0 -004451 0°1330 1000 0°130 0 :025 0 -005125 0°1532 1250 0°16 0-03 0 ‘005631 0°1683 1500 0°19 0:035 | 0:°007285 0°2177 Table XII.—German Silver Wire. Current from Stretched to Unstretched. Temperature difference = 86°C. Total resistance in circuit = 26°85 ohms. é Current in | Thermo-electric Total weight open ee mikroampere difference in or Bovor. elongation. elongation. ee ea Satad | gree. 250 | Se Se 0 °002529 | 0 -0679 500 0-08 as 0 :002360 0 :0634 i, «| 07105 sd 0002474 | 0-0661 1000. 0-13- | : 0°003020 0-0813 1250 | 0°18 0°051 0 *003710 | 0 :0996 1500 0°35 0 162 0 °003946 0:1059 1750 | 0°80 0-588 0°004552 0 1222 178 = Lffects of Strain on the Thermo-electric Qualities of Metals. Table XIII.—Manganin Wire. Current from Stretched to Unstretched, Temperature difference: = 86°. Total resistance in circuit = 30 ohms. ; Current in Thermo-electric Total weight i ee: Betoun age mikroampere difference in ile porary permanent : | 72 grammes. | longation Elasicou oa per degree mikrovolt j * | with weight on. per degree. 250 le oye 0 -001888 0 05665 500 0°145 0 °012 0 °002124 . 0 :06372 750 0 260 0°072 0 002090 0 °06272 1000 0°327 0°120 0 °002259 0 06778 1250 0 °382 0°151 0 002158 0 06474 1500 0°427 0°166 0 002158 0 °06474 1750 0°484: 0°193 0 :002192 0 °06575 Table XIV.—Iron Wire. Current from Unstretched to Stretched. Temperature difference = 85°. ‘Total resistance in circuit = 19:20 ohms. Current in Thermo-electric: Total weight eee IEMEEAE Be mikroampere difference in © é emporary permanent Fi mik It bn erates: | elongation. elongation. per eae ape ee 5 with weight on. per degree. 250 36 36 0 :0000727 0 °000912 500 35 | ie 0 000131 0 006216 750 0 023 ate 0 :000131 0 -006216 1000 0°03 0-02 0 °003813 0 -07192 1250 0°06 0-027 0007816 0 09485 1500 Omg 0:08 0 007736 0 -09430 1750 0°39 0°23 0 006602 0 08696 2000 sie 3°41 0 °00461 0 07576 A Case of Monochromatic Vision. 179 “A Case of Monochromatic Vision.” By Sir W. DE W. ABNEY, K.C.B., F.R.S. Received January 17,—Read February 1, 1900, | Cases of monochromatic vision are rare, and I have thought it right to put on record one which was kindly brought me some time ago by Mr. Parker. The patient, whom we will call K. B., was aged twenty-five at the time when I examined him for colour vision. The notes of his case are as follows :—Vision always defective ; has always been colour blind. Has quick horizontal nystagmus; probably an absolute central scotoma. He is always ‘day blind.” His vision for right and left eyes is 6/60. He is not night blind. His fields are nearly, but not quite, full for white. He shows no definite changes in his eyes, I took his luminosity curve, and all colours he matched with white with the same facility as if they were white. The following table ae i B.'s tog oS K. B.'s Pis ee ciatic). luminosity. bere riaicsiey luminosity.| luminosity. 56 2°5 — 32 61°5 65 0 54 9°0 aa 30 43 °O 50 °0 52 16°0 a0 28 37 ‘0 36 °0 50 27 °5 19°0 26 30°0 26 °5 48 42 °5 39 °0 24 24°0 19°5 | 46 61°0 65 °0 22 18°5 14°0 | 44 82 °5 85-0 20 14°5 10-0 42 96 °0 98-0 18 11°5 — 40 100 ‘0 99:0 16 9°0 5°5 38 95 °5 97 °5 14 7°0 — 36 87 °5 90-0 12 5°0 _ 34 75 °0 80 °0 10 3°0 2°5 i neneee cn Wee aM dl gives the luminosity of the spectrum to him, and for the convenience of reference a previous case, which has already appeared in the ‘ Pro- ceedings,’ is given for comparison, In the accompanying diagram both these curves are shown, together with the curve of luminosity for the normaleye. As regards the first two, it will be seen that the maximum of each curve is about scale number 40, or close to E. On the right- hand side of the maximum the curves do not absolutely agree. K.B.’s observations were first made in the red and green, and his readings at first were not very close, and a mean had to be taken. As the colours he had measured went towards the blue his measures were much more accordant, as he had become accustomed to the methods 180 Dr. A. Macfadyen. On the Injluence of employed. The slight divergence on the left-hand side of the curve from that of P is probably due to his colouring matter in the yellow spot. Attention must be again called to the fact that these curves are practically identical with those obtained by the normal eye when it measures a spectrum of very feeble luminosity, and also agree with Pra a oa Bite eae OMABi 6 4 0124618 20 £224 sEeRSEESESERERITL =e) the results obtained by measuring the diminution of each ray when it first becomes invisible, and making a curve of the reciprocals of the numbers, taking the highest point of it as 100. This is clearly shown in Part III, “Colour Photometry.”* It may be mentioned the scale of the prismatic spectrum employed is the same in this communica- tion as in that paper, the wave-lengths of each scale number being given in it. “On the Influence of the Temperature of Liquid Air on Bacteria.” By ALLAN MacrapyEN, M.D. Communicated by Lorp LISTER, Pres. R.S. Received December 15, 1899,—Read February 1, 1900. The experiments of Dr. Horace T. Brown and Mr. Escombet have shown that no appreciable infiuence is exerted upon the germinative power of seeds, when exposed for 110 hours to the temperature of * «Phil. Trans.,’ A, vol. 1&3, 1892. + ‘Roy. Soc. Proc.,’ vol. 62, 1898, p. 160. the Temperature of Inquid Air on Bacteria. 181 liquid air (— 183° C. to — 192° C.). The results were equally nega- tive in the recent experiments of Sir W. Thiselton-Dyer,* in which seeds survived exposure for upwards of six hours to the temperature of liquid hydrogen (— 250° C. to — 252° C.). The following investigation on the influence of the temperature of liquid air on bacteria, was carried out at the suggestion of Sir James Crichton Browne and Professor Dewar. The necessary facilities were most kindly given at the Royal Institution. The experiments were conducted under the personal supervision of Professor Dewar, and he has asked me to put the results on record, although it must be acknowledged that the essential features of the investigation are due to him. | The bacteria employed were selected from the stock of the Jenner Institute of Preventive Medicine, where the results were also con- trolled. Pure cultures of the several micro-organisms were employed, and the series included typical representatives of saprophytic and parasitic bacteria. The organisms chosen possessed varying degrees of resistance to external agents—the extremes in this respect being repre- sented by the very sensitive spirillum of Cholera Asiatica and the highly resistant spores of B. anthracis. Ten organisms were used for the experiments, viz. :—B. typhosus, B. cola communis, B. diphtherie, Spirillum cholere Asiatice, B. proteus vulgaris, B. acidi lactici, B. anthracis (sporing culture), Staphylococcus pyogenes aureus, B. phosphorescens and Photobacterwwm balticum. The cultures of the organisms were young and vigorous, and were tested both on solid and in fluid media, viz. :—Nutrient gelatin, agar- agar, potato and peptone broth. The cultures on these media were simultaneously exposed to the temperature of liquid air for twenty hours (- 182° C. to — 190°C.). They were then carefully thawed and examined. The results may be briefly stated. In no instance, whether on solid or in liquid media, could any impairment of the vitality of the micro-organisms be detected. The fresh growths obtained from the exposed tubes were normal in every respect, and the functional activities of the bacteria were equally unaffected. The colon bacillus produced its typical effects—such as the curdling of milk, the fermentation of sugar and the production of indol; the Staphylococcus pyogenes aureus retained its pigment producing properties and the anthrax spores their pathogenic action on animals. The photogenic bacteria preserved their normal luminous properties. These photogenic properties are intimately connected with the functional activities of the cells. The cells emit light which is apparently produced by a chemical process of intra- cellular oxidation and the phenomenon ceases with the cessation of their activity. These organisms therefore furnished a very happy * “Roy. Soc. Proc.,’ vol. 65, 1899, p. 361. VOL. LXVI. Q 182 Influence of Temperature of Liquid Air on Bacteria. test of the influence of low temperatures on vital phenomena. Their cultures, when cooled down in the liquid air for twenty hours, became non-luminous, but on re-thawing the luminosity returned with unim- paired vigour as the cells renewed their activity. Watery emulsions of the photogenic bacteria, on immersion in liquid air for a few minutes, ceased to emit light, but on withdrawal the luminosity reappeared in avery short time. Strips of filter paper soaked in the watery emul- sions and brightly luminous were immersed directly in the liquid air with similar results. The sudden cessation and rapid renewal of. the photo- genic properties of the cells, despite the extreme changes:of tempera- ture, was remarkable and striking. The following experiment was made :—Fifty litres of the laboratory air about six feet from the ground were liquefied at atmospheric pressure in a glass bulb by means of boiling liquid air am vacuo. The tempera- ture reached was about — 210°C. The bulb was then sealed off, the contents being still at a temperature below zero, and was subsequently opened and washed out with sterile broth. A series of plate cultures were made from the broth on nutrient gelatin, agar-agar and sugar agar, and were incubated under aérobic and anaérobic. conditions at 22° and 37° C. for a period of ten days. The anaérobic plate cultures remained sterile. The aérobic plates yielded forty-four organisms which had survived an exposure to — 210°C. The organisms were representative types of those to be usually met with in the air, viz., moulds, bacilli, cocci, torule and sarcine. It may also be mentioned that a sample of | yeast cll plasma (Buchner’s zymase) subjected to — 182° C. to — 190° C. for twenty hours, retained its peculiar properties ae Vi1zZ., as ices the production of CQ, and alcohol. The above experiments show that baci may be. scold down to — 190°C. fora period of twenty hours without losing any of their vital properties. Further experiments are in progress with ihe above-mentioned and with other micro-organisms, exposed to the temperature of liquid. air for still longer periods of time, as well as to that of liquid hydrogen. These experiments will form the subject of a future communication. Electrical Effects due to Evaporation of Sodiwm in Air, &c. 183 February 8, 1900. The LORD LISTER,. F.R.C.S., D.C.L., 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 :— I. “The Spectrum of « Aquile.” By Sir NorMANn Lockyer, K.C.B., F.R.S., and A. Fow.er. Tl. *“*On the Production af Artificial Colour-blindness by Moonlight.” By G. J. BurcH. Communicated by Professor Gotcu, F.R.S. III. “On the Relation of Artificial Colour-blindness to Successive Con- _ trast.” By G. J. BurcH. Communicated by Professor GoTcu, E.RS. IV. “On Electrical Effects due to Evaporation of Sodium in Air and other Gases.” By W. CRAIG HENDERSON. Communicated by Lorp KELVIN, F.R.S. V. “On Electric Touch and the Molecular Changes produced in Matter by Electric Waves.” By Professor J. CHUNDER BOSE. Communicated by LorpD RAYLEIGH, F.R.S. “On Electrical Effects due to Evaporation of Sodium in Air and other Gases.” By W. Craic HENDERSON, M.A., B.Sc., late 1851 Exhibition Science Scholar. Communicated by Lorp KeELvIN, F.R.S. Received November 30, 1899,—Read Feb- ruary 8, 1900. The experiments described below form part of a research which I began in the summer of 1897 in the Cavendish Laboratory, Cam- bridge, but was unable to complete before leaving Cambridge that same year. The object of this part of the research, which was sug- gested by Professor J. J. Thomson, was to determine whether evapora- tion of an unelectrified liquid produces any electrification or not. The liquid used was fused sodium and the arrangement of apparatus is shown in the accompanying drawing. The sodium to be fused was held in a vertical iron cylinder, A, Q 2 184 Mr. W. Craig Henderson. On Electrical Effects closed at the bottom and having a tightly fitting asbestos plug at the mouth. Through a small hole in the asbestos plug there passed a tightly fitting glass tube, G, only a little longer than the thickness of the plug. A stout bare copper wire, C, passed through this tube G without touching the sides, and had a copper disc at the end inside Bunsem Burners. the iron cylinder. Outside the cylinder, this wire passed direct to the insulated quadrants of an electrometer E, and was surrounded throughout this portion of its length by a metal guard-tube D, which screened it from outside electrostatic influences. Besides the fastening to the electrometer, the sole supports of the wire C were two parafiin plugs fixed into the ends of this tube D. The iron cylinder A, the tube D, and the uninsulated quadrants of the electrometer EH, were connected by a wire with one another and with the sheath of the electrometer, denoted by S in the diagram. The heat to fuse the sodium was supplied by two Bunsen burners placed below the cylinder A; and in order to protect the insulated. wire from the hot gases rising from the burners, a metal screen M was fixed on the cylinder and bent up on the side remote from the electro- meter to serve as a funnel, due to Evaporation of Sodiwm in Avr and other Gases, 185 With these arrangements it was found that very soon after heat was applied to the sodium, a negative electrification of 2 to 3 volts was indicated by the electrometer. This electrification persisted for a considerable time, but eventually the insulation broke down, owing to the sodium vapour condensing in the glass tube G at the mouth of the cylinder, sufficiently to cause a solid connection between the tube and wire. When the same experiment was repeated without the sodium, no electrification was indicated by the electrometer. These results were confirmed by repeated experiments, and the question then arose whether this negative electrification is due to the evaporation of the sodium, or to oxidation of the sodium going on in the iron cylinder. The latter seemed probable, as the electrometer showed electrifica- tion almost from the moment when heat was applied to the sodium, whereas sodium does not fuse till at temperature of 96° C. and boils at about 400° C. To determine this point, the same experiment was repeated with this difference, that the air in the cylinder was replaced by an atmo- sphere in which the sodium could not oxidise. Carbonic acid gas was first tried, being kept flowing into the cylinder after passing through a drying apparatus; but in this gas the sodium became coated with a white encrustation, and showed no signs of evaporation or of boiling even at a red heat. | Coal gas was next tried, and no difficulty was found in boiling the sodium in this gas. To prevent accident by explosion, the apparatus was set up in the fire-place, so that the escaping coal gas might pass up the chimney, and not mix with the air of the room. Great care was taken to ensure the complete removal of air from the iron cylinder before heat was applied. With this atmosphere of coal gas no electrification was obtained while heat was applied to the tube for over an hour. The insulation was tested from time to time during this period by giving a charge to the wire from an electrified vulcanite rod, watching the rate of leak indicated on the electrometer scale, and then discharging. It was found to be excellent; but eventually, as before, it broke down when some of the sodium vapour condensed in the mouth of the cylinder between the wire and the glass tube. Repetition of this experiment confirmed this result. This problem of the possible generation of electricity by evapora- tion of a liquid, has been recently investigated for the case of water by Pellat,* who found no trace of electrification. A similar result was found in the earlier experiments of Blake,t who used water and solutions * Pellat, ‘Séances de la Société Francaise de Physique, 1899,’ ler Fascicule, p- 21. + Blake, ‘Wiedemann’s Annalen,’ vol. 19, 1883, p. 518. 186 Mr. W. Gardiner. The Genesis and Development of copper sulphate and of sodium chloride, but found no electrification. The present. experiments lead to the conclusion that evaporation of fused sodium does not give electrification, such as could be detected by the method used, unless oxidation is going on. February 15, 1900. The LORD LISTER, F.R.C.S., D.C.L., 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 yall I. “The Genesis and Development of the Wall and Connecting Threads in the Plant Cell. Preliminary Communication.” By WALTER GARDINER, F.R.S. II. “ Photography of Sound-waves and the Kinematographic Demon- stration of the Evolutions of Reflected Wave-fronts.” By R. W. Woop. Communicated by C. V. Boys, F.R.S. “The Genesis and Development of the Wall and Connecting Threads in the Plant Cell. Preliminary Communication.” By WALTER GARDINER, M.A., F-B.S., Fellow and Bursar of Clare College, Cambridge. Received February 1,—Read February 15, 1900. In the course of my investigations in connection with the forth- coming paper on “The Histology of the Cell Wall with special Reference to the Mode of Connection of Cells,”* certain observations and conclusions concerning the origin and development of the wall- threads and cell-wall have come to light which seem to be of suffi- cient interest to warrant my bringing them to the notice of the Society without delay. 1. Origin and Development of the ‘Wall Connecting Threads.” The “ connecting threads” are found to arise from the median nodes: of the fibres of the achromatic spindle. The nodes are either (@) all continued as connecting threads, ¢e.g., the endosperm cells of Tamus: * ‘Roy. Soc. Proc.,’ vol. 62, 1897. (Preliminary Communication.) of the Wall and Connecting Threads in the Plant Cell. 187 communis ;'(b) in part continued, and in part overlaid by superposed lamellze of cellulose membrane, e.g., the endosperm cells of Liliwm Martagon ; or, (c) all overlaid, e.g., the pollen mother-cells and pollen grains of Helleborus fatidus. 2. Origin and Development of the Cell Wall. (a) Origin.—Seeing that spindle fibre nodes (apparently intact) can be recognised in a mature wall of considerable thickness, there would seem little doubt that the existing views with regard to the genesis of the cell plate and first formed cell wall cannot be entirely correct. I am inclined to believe that the cell plate arises not directly from the spindle fibres, in the manner described by Strasburger and others, but rather indirectly; that is to say, that although it is possibly pro- vided by or even proceeds from the fibres in question, yet it exhibits a certain structural distinction, in that it is pierced by the persistent nodes of the spindle fibres, and is not merged into their substance. The cell plate would appear to consist of cytoplasm, and cytoplasm, moreover, practically identical with the ordinary cytoplasm of the cell, and from it is secreted the first formed cell wall as an equatorial mem- brane traversed by the nodes of the achromatin spindle fibres. (b) Development.—There are grounds for regarding the primary cell wall as different in genesis and character from the secondary formations which succeed it and arise from the general cytoplasm. In any case, the wall rapidly grows in thickness as layer after layer of cellulose is deposited. In the course of my work certain observations were made, which appeared to throw some light on the structure and genesis of the wall thus produced. It was found that many walls, and especially mucilaginous walls, when strongly swollen and stained, after passing through the stage of stratification, became resolved into numberless and often well-defined spherical droplets or spherules which not unfrequently exhibit a markedly high refraction, and are embedded in a hyaline and possibly mucilaginous ground-substance or matrix. I am of opinion that these spheroidal droplets represent swollen granules or spherules, which are practically homologous with the droplets or the drops (and I am disposed to think with the droplets) described by myself and Ito in our paper “ On the Structure of the Mucilage-secreting Cells of Blechnum occidentale, L., and Osmunda regalis, L.,” published in 1887 in the August number of the ‘Annals of Botany’; and I believe that the phenomena in the two cases of internal mucilage there described, were in essence, instances of internal wall formation, or, in other words, that the formation of the cell wall takes place in a similar way. Moreover, the “ mucilage ” described by us, both gave the reactions of cellulose, and also exhibited the forma- tion of a firm, clear, and stratified membrane. 188 | Proceedings and List of Papers read. In the above paper we compared the droplets of Blechnum occidentale with the granules or spherules described by Langley as occurring in certain gland cells, ¢.g., the mucous cells of the sub-maxillary gland of the dog; and I am still of opinion that such a comparison was a pertinent one, and not entirely without significance in the case of the plant cell wall also. I am disposed to regard the cell wall as fundamentally of the nature of a mucous or, rather, mucilage secretion ; the droplet or spherules (shall I call them provisionally “‘tezchosomes” ?) being composed of a substance which, when more hydrated, passes as “a mucilage,” and when less hydrated functions as ‘‘a cellulose.” The spherules are embedded in the “‘ ground substance,” and possibly the remains of even a proto- plasmic framework (which may undergo mucilaginous change) is also present. hy, I regard stratification as the necessary accompaniment of the rhythmic periods of activity and rest of the secreting protoplasm ; and as to the method of secretion, it is eaternal and not internal, as in the mucilage cells described by Gardiner and Ito. ! The changes incident upon lignification and the like I have always regarded as induced by secondary secretion or post-formation chemical change. I may add that I see little in the above view of the structure of the cell wall which militates against the facts which we have at our disposal, either with regard to the properties of the cell wall or to the phenomena associated with growth in thickness or in surface. I am aware that much remains to be done before the above views are placed on a proper basis, but I have great hopes that this is only a matter of time and of further detailed research. February 22, 1900. The LORD LISTER, F.R.C.S., D.C.L., 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 :— I. “Preliminary Note on the Spectrum of the Corona. Part 2.” By Sir Norman Lockyer, K.C.B., F.R.S. II. “On the Structure of Coccospheres and the Origin of Coccoliths.” By Dr. H. H. Dixon. Communicated by Professor J. Joy, F.RS. Preliminary Note on the Spectrum of the Corona. 189 III. “The Ionisation of Dilute Solutions at the Freezing Point.” By W. C. D. WHETHAM. Communicated by E. H. GRIFFITHS, F.RS. “Preliminary Note on the Spectrum of the Corona. Part 2.” By Sir Norman Lockyer, K.C.B., F.R.S. Received February 8,—Read February 22, 1900. One of the chief results which, in my opinion, would be secured by the use of the prismatic camera in eclipse work was the differentiation between chromospheric and coronal phenomena. The photographs taken during the eclipses of 1893, 1896, and 1898 all enabled this distinction to be made very clearly, and various radiations formerly attributed to the corona have been shown to belong to the chromo- sphere alone. The photographs taken in Africa in 1893 showed eight rings in the spectrum of the corona ; in Novaya Zemlaya, in 1896, with a less powerful instrument, a smaller number was secured ; but those taken with increased dispersion in India, in 1898, show a much greater number. I have already given the results of an inquiry into the wave-lengths of two of the chief coronal rings (5303-7 and 4231°3) as determined from photographs taken in 1898 with the 6-inch prismatic camera ;* and as the results of the continued investigations may be of service to intending observers of the eclipse of next May, I give a short abstract of them in the present note. Eight photographs were obtained during the time the spectrum of the corona was least admixed with that of chromosphere ; of these, three taken with instantaneous exposures show only two or three of the brighter rings, so that five, showing many coronal rings, are suitable ‘for measurement. The exposures of the photographs used for measure- ment were as follows, the ratio of focal length to aperture being 15 :-— ig 2 eee as acura 50 seconds. Ve oP Or aces ss eke ete, WR ens «sea hae ® 2) Ee Sa, During the earlier part of the exposure of plate 2a, the upper regions of the chromosphere were visible in the north-east, and some of the stronger chromospheric arcs appear, together with the coronal rings, in the corresponding parts of the images. In plates 3¢ and 3d chromospheric arcs appear in the south-west quadrant, together with * © Roy. Soc. Proc.,’ vol. 64, p. 168. 190 mee v0) “asia Norman Loekyer, coronal rings ; but, as I’ pointed out in the preliminary report on the observations, at. Viziadrug, the arcs and rings. are readily distin- guished.* The coronal rings which have been noted on the photographs may be divided into three groups, defined by the position-angles in which they have their greatest brightness. The typical rings are (1) the green ring at 4 5303°7 ; (2) a violet ring at 4 3987, near He; and (3) a blue ring at A 4359- 5, near Hy. The structure and brightness of these are shown in the -cispapenr is diagram, but it may be remarked that the fainter members of the three groups do not exhibit the differ- ences of structure so clearly. Fie. 1.—Diagram showing the Forms of Three Typical Coronal Rings, and the Positions of the Prominences photographed at the same time. The tables which follow show the wave-lengths of the rings which are believed to belong to each of the three groups, and indicate also the average brightness of each ring. : * ‘Roy. Soc. Proc.,’ vol. 64, p. 38. Preliminary Note on the Spectrum.of the Corona, LOE Table of Coronal Rings. Group I. Typical ring, 1 5303-7. ovine. | RET Wave-length. Meena Aen ners Wave-length. Tamia 3952 °5 2 4536 1 4007 1 4588 *5 1 4022 1 4657 1 4056 2 | 4685 °5 2 4068 1 | 4714 1 4085 1 | A727 1 4121 1 4737 1 4168 1 | 4768 1 4220 2g 4808 1 4231-3 5 | 4922 2 4248 °5 2 1 5125 1 4262 1 | 5137 1 | 4400 I | 5303 °7 10 4430 1 | | ee 4518 1 | | | Group II. Typical ring, 4 3987-0. Group III. Typical ring, \ 4359-5. Brightness. Brightness. Max. = 10. Wave-length. rightness { | | Wave-length. Max. = 10. 3800 3987°0 4275 4568 °5 4030 4192 4204 © 4302 4323 4359 ° 4485 4648 4662 4788 4890 5001 5285 We ow Or ceil pearl soe ceri meee el need SNS sel cell sell I have already suggested that the different forms of the coronal rings indicate that they are not all due to the same substance, and the foregoing tables suggest that at least three substances are in question. The attempts which have so far been made to trace the origins of the rings, however, have led to no very definite results, and the coincidences with lines in the spectra of stars and nebule which were formerly suspected have not yet been completely established. Special interest is attached to the question of the presence or absence of carbon flutings. There is a possible trace of the fluting, 992 Mr. W. C. D. Whetham. commencing at X 4736°18, which so far has not been observed in the chromosphere. The other flutings of carbon which are present in the chromosphere do not appear in the coronal spectrum. The reductions indicate that there may be feeble indications of the presence of some of the chromospheric gases in the inner corona. Thus in photograph 3d, on the north-eastern edge, fragments of rings corresponding to lines of helium at AA 4472, 4714, and 4922 have been recorded ; these occur also on the south-western limb, where the chromosphere itself is coming into view, but as the chromosphere was completely eclipsed in the north-east at this stage, the radiations mentioned as occurring there perhaps belong to the inner corona. A very interesting result of this detailed examination of the photo- graphs is that the chief coronal ring in the green is very closely associated with the form of the inner, and appears to have no distinct connection with the outer, corona. This suggests that the green line of the coronal spectrum is not produced in the outer corona, and that the indications of its presence there on previous occasions, as obtained by slit spectroscopes, were simply due to glare, as in the case of hydro- gen and calcium. So far as the photographs taken with the prismatic cameras are concerned, the spectrum of the outer corona gives no indi- cations of bright rings. The measurements of the coronal rings and the diagram which accompanies this paper have been made by Mr. Fowler. Dr. Lockyer has investigated the coronal spectrum in relation to carbon, and Mr. Baxandall has made comparisons with the spectra of stars and nebule. “The Ionisation of Dilute Solutions at the Freezing Point.” By W. C. D. WHETHAM, M.A., Fellow of Trinity College, Cam- bridge. Communicated by E. H. Grirrirus, F.R.S. Received February 14,—Read February 22, 1900. (Abstract.) It is known that the depression of the freezing point of water, pro- duced by dissolving molecularly equivalent amounts of different acids and salts in a given quantity of it, is approximately proportional to the number of ions which these substances must be supposed to yield in order to explain their electrical conductivities. Again, as the con- centration of a solution of one such substance is gradually increased, the molecular depression of the freezing point, and the equivalent electrical conductivity, both vary, and vary by amounts which seem in some cases to correspond, but in others to differ considerably. There appeared reason to suppose that it was desirable to increase The Ionisation of Dilute Solutions at the Freezing Point. 193 both the extent and the accuracy of our experimental knowledge of these relations. Freezing-point determinations for very dilute solu- tions are extremely difficult, owing to the minute differences of tem- perature to be measured, and the results given by various observers showed great discrepancies. On the other hand, the most satisfactory experiments on the electrical ionisation of corresponding solutions had been made at higher temperatures, instead of at the freezing point, at which they should be obtained for purposes of comparison. The fact that the temperature coefficient of conductivity differs for solutions of different concentration, showed that the values of the ionisation would vary if the temperature was changed. Mr. E. H. Griffiths therefore undertook the examination of the freezing points by the method of platinum thermometry, and the present paper contains an account of corresponding measurements of the electrical conductivities at 0° C. In order to avoid any possible action of glass on the solvent used, it was determined that the water should be obtained from a platinum still and collected in platinum bottles, and that both the freezing point and the electrical measurements should be made in platinum vessels. The structure of the resistance cell is represented in fig. 1. The walls of the vessel itself are used as one electrode, and an insu- lated platinum cage, suspended inside, forms the other. Within the cage is a platinum screw, mounted on a shaft, which can be turned by means of a hand wheel and cord. This screw is used to insure tem- perature equality throughout the liquid, and to mix the solutions when made. The shaft of the screw is a hollow tube, closed at the bottom, which contains a thermometer. Instead of beginning with a strong solution and gradually diluting, it was thought better to begin with a definite quantity of the pure solvent, and, when its resistance had been observed, to add weighed amounts of stock solution of known strength by means of the platinum vessel shown in fig. 2. This vessel will obviously empty itself if a flow of liquid is started by slightly increasing the air pressure at the neck. In order to obtain a definite quantity of solvent, slightly more than the volume needed was placed in the cell, and the level of the liquid was then adjusted by sucking water through a capillary platinum tube into the glass vessel shown in fig. 3. The bottom of the capillary always comes to the same position relatively to the cell, and, if the sucking pressure is kept constant and equal to that of a water column of about a foot in height, it is found that the amount of water left in the cell is constant to within about one-tenth of a gramme. ‘Thus three independent withdrawals left 219-60, 219°63, and 219-59 grammes. Whenever the cell was dismounted and set up again, this measurement was repeated. The platinum vessel was surrounded by a brass case, coils of metal te Mr. W. C. D. Whetham. ra. i. tubing being placed in the narrow air space between them. LEvapo- rated ether vapour could be drawn through these coils by an air-pump, and thus the whole vessel cooled. The apparatus was fixed i in a large copper tank, which was filled with melting ice. The Ionisation of Dilute Solutions at the Freezing Point. 195 _ The electrical resistance measurements were made by the method of alternating currents, but the usual telephone indicator was replaced by a D’Arsonval galvanometer. This was done by using a revolving commutator, which, turned by a hand-wheel and cord, alternated the connections of the bridge with the battery and with the galvanometer simultaneously. The usual Wheatstone-bridge method could then be used, and measurements obtained in the same cell of resistances varying from 10 to 50,000 ohms, the accuracy throughout: being at Fie. 3. least 1 in 1000. The method eliminates several troublesome periodic disturbances, and, in this form, seems entirely satisfactory. The surface of the electrodes was platinised in the usual manner, but was afterwards heated to redness. This process gives a roughened plati- num surface of large area, which is less liable to absorb matter from the solution than is the unheated platinum black. The water used was thrice distilled, twice with alkaline perman- ganate and once in a platinum still with a trace of acid potassium sulphate. It had an ahaa conductivity at 18° of about.0°9 x rs bent C.G.S. units. 196 Mr. W. C. D. Whethai. Some of the stock solutions for the early part of the work were prepared by Miss D. Marshall, and most of those used in the later measurements were made up at the Cambridge University Chemical Laboratory by Mr. G. Hall, under the advice of Mr. H. J. H. Fenton. Others were prepared by the writer from recrystallised salts obtained from Kahlbaum, of Berlin. From a knowledge of the weight of solvent used and the weight of stock solution added it was easy to calculate the concentration (m) of the resulting solution in terms of gramme-equivalents of solute per thousand grammes of solution. Allthe experiments were made on solu- tions so dilute that this way of defining m leads to practically the same results as though the gramme-equivalents of solute were referred to 1000 grammes of solvent, or to one litre of solution. The differences only become visible on the curves in the cases of two or three of the strongest solutions of some of the substances used. The observed resistance is corrected for any slight difference in temperature from zero, and for the increased volume of liquid in the cell due to the volume of stock solution added. The reciprocal of this corrected resistance is the conductivity in arbitrary cell units, and from this the corresponding arbitrary con- ductivity of the solvent is subtracted. The resultant conductivity, 4, due to the added solute alone, is divided by m, and k/m, the equivalent conductivity, plotted on a diagram as ordinate, the value of m', a number proportional to the average nearness of the molecules, being used as abscissa. From these curves the maximum value of £/m is estimated, and taken to represent complete ionisation, the ionisation for the solutions measured being calculated as the ratio between the actual value of k/m and its maximum. The values obtained for these ionisations were arranged as shown in the following table, which is given as an example, and are plotted as curves on the diagrams appended. Sulphuric Acid.—Prepared at the Chemical Laboratory by adding the calculated amount of SO, to distilled acid. Successive crystallisation brought the melting point of the resultant H,SO, to + 10°5° Cent. The crystals were dissolved in water and the concentration of the solution estimated by the barium sulphate method. Similar measurements were made on Potassium Chloride, Barium Chloride, Copper Sulphate, Potassium Permanganate, Potassium Ferri- cyanide, and Potassium Bichromate. In discussing the results, we may first notice that, in cases where it has been possible to obtain values for the ionisation at 18° from Kohlrausch’s work, the ionisation curve at 0° is appreciably different from that at 18°, the ionisation falling off more rapidly with increasing concentration at the higher temperature. In the diagrams the values for 0° are given by dots inside circles, and those for 18° by crosses. The Ionisation of Dilute Solutions at the Freezing Point. 197 In the case of copper sulphate, measurements by the present method were made at 18° as well as at 0°, and are indicated by crosses inside circles ; giving a curve which agrees with Kohlrausch’s observations at moderate concentrations, but differs from them at extreme dilution. The normal type of curve is given by potassium chloride, barium chioride, &c. The curve for sulphuric acid departs from this form, as other observers, using glass vessels and working at higher tempera- tures, have previously found. The drop in this curve at extreme dilution is seen also in solutions of other acids and alkalies, and it has been usual to explain it by supposing that the effective amount of acid is reduced at extreme dilution by interaction with the residual im- purities of the solvent. The phenomenon seems too constant for this Table I. 4H.SO, = 49-04, Solvent. Weight = 219-42, R = 40420. Mm. mi, R. K/m. a. Peeaeacsseen.) 12 204x10—* | 0-0319 | 3106 9-122 - }-0°809 Peeesinsscecas| 9°628_ ,, 0°0459 | 954°3 | 10°62 0 ‘941 Re So gidicie ne sin «+ 2°001 x 10-4 | 0°0585 | 444°7 11°10 0°984 MRitestaes sass.) oa" 5, 0 ‘0709 247°8 =| 11°26 0°999 Pies ee aaee ss) G40 ,, 0 *0859 139 11°26 0-998 We eeeeseevecee ss] 1°426x10-* | 071125 63-11 | 11°09 0 °984, Me... oss. 2411 , |0-1341| 37°75 | 10-90 | 0-967 eee, ..| 3-423) ., | 071507} 27-25 | 10:73. | 0-951 eat ree ee ns | 4° T2R-Y yy 0°1678 20°05 | 10°53 0 °934. In Glass Cell. I . "660 x 10-3 | 0°1541 1789 0 °1552 | 0:948 TEL ¢ "158 x 10-2 | 0°2262 | 615°7 0-1426 | 0°871 i) ae ? ) eeee @eeserereeenee 3 | Wee coe teen! BSB. \ 0-1752 | 1246 0°1516 | 0-926 | Ps es 1 | 1-747 ,, | 0°2595 |) 424°4 | 0°1847 | 0°837 | | explanation to be satisfactory, and the cause of it may perhaps be con- nected in some way with the fact that it occurs only in solutions the solute of which gives ions either of hydrogen or hydroxyl, which are ions (1) present in the solvent, (2) possessing greater velocities than any other ions. The drop in the curve for potassium permanganate is, on the other hand, probably due to interaction between the salt and the solvent impurities. The effect is completed by the first addition of salt, for if a correction be made in the case of the first solution for the salt thus put out of action, it is found that the curve reverts to the normal type. This is clearly shown by the diagram. Again, both in this case and in that of sulphuric acid, it was found that for solutions of great dilution the resistance showed a gradual rise for some time after the VOL. LXVI. R 198 . Mr. W. C. D. Whetham, ‘50 28 *26 24 "20 «6-22 eZ fi id (z H, SQ). +6 ." : 14 ‘2 10 ~~ _ ~ee 02 mémw—>.: :08 0:3 The LIonisation of Dilute Solutions at the Freezing Point. 199 ~O-4-+-o- he Sichrameate (J Oa ui 200 Mr. W. C. D. Whetham. The Ionisation of Dilute Solutions at the Freezing Point. 201 stock solution was added to the solvent, as though the action took time for its completion. While, however, in the case of permanganate this phenomenon was only observed on adding the first lot of stock solution, in the case of acid it appeared in the second solution also. This confirms the idea that the action is not completed by the first addition of acid, though the quantity of acid present must be large compared with the amount of residual impurity in the solvent. The permanganate measurements also show that the slant of the curve is that of a salt like potassium chloride, with a monovalent acid radicle, rather than that of a salt such as copper sulphate, with a divalent acid. The chemical structure of permanganate in water solution is therefore probably represented by the formula KMnQy. The curve for potassium bichromate appears to consist of two parts, an indication, perhaps, that the ions are different at different concen- trations. In order to collect the results, smoothed values have been obtained from the curves and are appended in Tables IX, X, XI, and XII. The first three tables contain ionisation coefficients at 0°, the concentration being tabulated in different ways. Table XII shows approximate values for the equivalent conductivities at 0°. These were not neces- sary for the determination of ionisation, so a single value of the cell constant, obtained by comparison of the copper sulphate measurements at 18° with Kohlrausch’s absolute values, was used, except for the potassium chloride solutions, which were reduced by a figure given by Kohlrausch for this salt at 0°. The errors will be small, for the amount of solvent left in the cell in each case was very nearly con- stant, and this is a measure of the accuracy with which the cell is re- adjusted after being taken to pieces. The results, however, are not supposed to be as trustworthy as those of the ionisation coefficients. Mr. W. C. D. Whetham. GOS. 0 644-0 188-0 086-0 a 298-0 ¥2 Zr6-0 183-0 L8-0 ¢18-0 TS6.0 +2 T98- 0 0&8: 0 296: 0 999-0 106-0 806-0 896.0 ie £98: 0 . 688-0 L16-0 SrL-0 Pr6-0 P86: 0 616-0 sal a 998-0 026-0 186-0 608: 0 16-0 S6:0 886-0 ae 048: 0 PPG: O £66- 0 C98: 0 266-0 696-0 266-0 =oet 088-0 296-0 966: 0 606-0 000: T O86: 0 G66. 0 Sree 668-0 PL6- 0 866: 0 0P6- 0 166-0 886: 0 866: 0 asoe 126-0 €86-0 000- T $96: 0 896-0 66-0 666: 0 ests. L¥6-0 066: 0 000: T 846.0 106-0 166-0 000- T sags sca 696: 0 66: 0 000: T 886- 0 G64 -0 666-0 000: T sie aaa G86: 0 166-0 000: T 966-0 ae 000-T 000: T oat coma 66: O 666: 0 000: T 666- 0 as 000: T 000: T fare “oro mE Kooga t *OUNM ‘Fogno? ros Ht *oeqs ‘TOM "ult "MOTINIOS JO souUIVId puvsnoyy od oynyos Jo syueTeAtInbe-sutmess Jo equnU = wW "0 98 SPUSIOLYO-0F) UOIQVSIUOT—X 9]4e, 603-0 PSL. 0 £e8.0 BEG. 0 £0-0 L¥8.0 2 26-0 193.0 6Z8. 0 098: 0 PEG: O Z0-0 £98. 0 gi PPG. O 168-0 TS8- 0 918-0 296.0 S10. 0 ggg: 0 PS8-0 GGG. 0 889.0 €88- 0 968: 0 296. 0 10-0 298.0 918: 0 116-0 LIL-0 126-0 CZ6- 0 916-0 c00- 0 —98-0 616-0 986-0 408: 0 PL6-0 ECG. 0 186-0 Z00- 0 048-0 PPG. 0 £66. 0 £98. 0 Z66- 0 696: 0 266-0 T00- 0 088-0 196-0 866: 0 806-0 000- T 086: 0 966- 0 £000: 0 Z06-0 116-0 666-0 L¥6. 0 86> 0 066: 0 866: 0 2000-0 626. 0 S86-0 000: T 196.0 PPG. O C66: 0 666. 0 1000.0 296-0 166-0 000. T 186-0 088: 0 866: 0 000: T £0000: 0 086-0 966-0 000-T 266-0 bs 000- T 000:T 20000: 0 166-0 866-0 000- T 866- 0 a 000: T 000: T T0000: 0 “orp at ooq* at YOUN “rosnoe *osH? “19 vg t ‘TOM “ue 202 ‘UOTINTOS Jo soMUTeIS puvsnoyy Tod oqnjos Jo syuepeamMbo-ouuivis Jo toqunU = wW O38 SJUILOYJO-O9 UOT}SIUOT—' YX] 9G, 203 Point. ung of Dilute Solutions at the Freez 10N 0 t onisa The It ‘aqnjos JO o[noopour-emuvsis rod queAaos Jo soTMoo[OW-oWUIIS Jo JequINU = U "0. W S]UOLOYJo-oF) UOIYUSIUOT— TX fe, 6S L902 829 BSL, €0-0 TIL vie 904, S68 COIZ 8s9 T92 Z0:0 91 x PIL LIP SES 119 89, CT0.0 ZL, cog Zel OcP L182 ¢g9 911 10-0 CSL 9F8 ceZ 90¢ ShPS 804 182 c00-0 96L . 188 OF. 69S 969% 622 961 Z00- 0 TeZ 116 T¢Z 609 £09Z TPL 008 100-0 6eL 826 Gey TP9 P29 OcL FOS S000: 0 8g) SEG Gol, 899 raster Lcd G08 Z000- 0 08d 1S6 9cL 289 LL¥G 192 908 1000: 0 008 16 9S), 269 60EZ E91 108 60000-0 £28 296 9cL TOL os POL 108 Z0000: 0 £e8 796 9S2 FOL a GOL 108 T0000- 0 “0% Me oan Ht TOUNM Fognos Fos*Hs “19eqt ‘10M “ ‘uoINjos Jo soutureis puvsnoy} red oynjos jo sqyuoyeAmMbo-oummess Jo Joquinu = w oOL X Sqtun "9°9'9 Ul ‘0 4B SoTyTATJONPUOD yuoTeaInbyY oyeurxosddy— TX 214", £26-0 a oe ge ce6-.0 0002 9F8: 0 - 26-0 GES. 0 P18: 0 LS8.0 096-0 000 8c8-0 ee 696- 0 929- 0 G18: 0 268-0 ~L6-0 0000T 298-0 958-0 Z86- 0 COL. 0 #26: 0 226-0 P86- 0 00002 998: 0 168-0 £66-0 L6L-0 016-0 0S6: 0 Z66- 0 0000s 048-0 926-0 166-0 968-0 686: 0 996-0 566-0 00000T 818-0 6P6- 0 666: 0 206-0 000- T 616: 0 166-0 000002 106-0 696-0 000: T PEG: 0 686-0 686-0 666-0 00000 CZ6- 0 616: 0 000-T 696-0 0&6: 0 P66- O 000: T QO0000T 6h6- 0 L86- 0 000: T 616: 0 £68: 0 866- 0 000- T 0000002 916-0 F66- 0 000- T 266-0 ie 000: T 000- T 0000008 066: 0 166-0 166-0 ys 000: T i. 0000000T ee 866: (8) ee ee ee ee ee 00000006 “op HE Koo qe YOUN rosnot *os*tHt 10% t ‘OM ou y 47: a Mr. G. J. Burch. On the Relation of “On the Relation of Artificial Colour-blindness to Successive Contrast.” By Grorce J. Burcu, M.A. Oxon., Reading College, Reading. Communicated by Professor Gorton, F.R.S. Received January 30—Read February 8, 1900. I have elsewhere pointed out that my observations on artificial colour-blindness seem unfavourable to the theory of Hering, and favourable to that of Young. The experiments on successive contrast described in the following pages tend also to confirm in a remarkable manner opinions held before the time of Young, and which must be considered as incorporated in his theory. The method I have pursued throughout this investigation consists essentially in the use of the spectroscope to analyse sensations of contrast, and I have accordingly been able to make certain experiments which would have been beyond the resources of those earlier writers; but the following account of their opinions, expressed as far as possible in the words of the authors themselves, is intended to show in some detail how closely their views agree with my own results. | During the 18th century the phenomena of after-images and suc- cessive contrast attracted a good deal of attention, and although in most cases the physical conditions of the experiments were too com- plex to afford much information as to their true nature, there were some remarkable exceptions, to which I desire to direct attention. The opinions enunciated during the 18th century may be divided broadly into two groups. The one school held that after the stimulus of a strong light of any given colour, a species of reaction sets in, by. which a sensation of the complementary colour is produced. This view may be regarded as belonging to the same category as the theory of pe It may be doubted whether it was definitely adopted by Jurin,* who says only that this “contrary sensation is apt to arise in us sometimes of itself, and sometimes from such causes as at another time would not produce the sensation at all, or at least not to the same degree,” and preserves a like caution throughout his description of the phenomena. Other experimenters, however, advocated this theory, and it was strongly upheld in 1801 by Venturi,t who maintained that the changing tints of after-images excited by the pure colours of the spectrum proved the existence of a multiple function for each nerve- fibre, as opposed to the theory of one nerve, one function, taught by Bonnet.{ #* “ Essay on Distinct and Indistinct Vision.” In Smith’s ‘ Opticks,’ 1738. + “Dei Colori Immaginarii.” ‘ aa scelti sulle Scienze,’ da Carlo Amoretti. Soave, vol 21, p. 274. { “Essai Anslytique sur ’Ame,” and “Essai de Psychologie.” Works. 1785. Artificial Colour-blindness to Successive Contrast. 205 The other school, from which the theory of Young may be considered to have developed, seems to have been founded by Scherffer, who, in 1761, published a long series of experiments on contrast based on Buffon’s work, but criticising his conclusions. Scherffer’s standpoint is briefly expressed in the following passage :— “‘ Perhaps the Creator has so constructed the entire organ of vision that each kind of ray can only act upon such of the parts of which the eye is composed as are particularly appropriated to it. But I pre- suppose that the whole action of light consists in attraction and repul- sion. . . . . It may be that a continuous action of, for instance, red light, may so, change the order and arrangement of the parts of the back of the eye . . . . that those rays may be no longer strong enough to communicate to these parts the necessary vibratory movement, until a little rest shall have restored them to their condi- tion . . . . and during this time the other aye of different kinds will not cease to act 1 8 He points out that if this ehiitictiond of eéSociBiatital colours” is the true one, it must follow that the after-image of a coloured object viewed upon a ground of the same colour must. be black, just as a white spot upon a dark ground gives a black after-image upon white paper.* Two of his experiments may be specially noted :— In order to determine the complementaries of the primary colours by experiment, he projected the solar spectrum on a white surface, and observed the colours of the after-image produced by it.t He then compared these with the corresponding colours, as calculated by Newton’s method. He gives the following list of colours observed :— 2 er Blue, verging on green. .. TEC ie ee eines Blue, almost indigo. PN esses Sane en ono wes A more violet-blue. TST SONS RS aaa ey Purple earns cde e neti ee os sles Red Oe ee ee ... Orange, but rather pale. RG ee ee crass cok ion A very yellow green. He also made drawings of flowers, and painted them with colours complementary to those they naturally possessed. These, when steadily looked at in a bright light, gave after-images in their true colours. He even went so far as to copy a picture, painting it with a green face shaded with yellow, white hair and eyebrows, black eye-balls. with white pupils, and green lips, so that the accidental image of it had the colours of the original.t * Compare this paper, Section II (2), p. 208. + Compare this paper, Section IT (1), p. 208. { Compare this paper, Section III, p. 213. 206 Mr. G. J. Burch. On the Relation of He does not, however, deal with the positive after-image and after- effects. This was done by Robert Waring Darwin,* in his paper on the ‘Ocular Spectra of Light and Colours,” in which he treats of the ‘direct and reverse spectra” of brightly illuminated pieces of silk of various colours. He describes very clearly the series of changes of these after-images from negative to positive and back again, observable under certain conditions, and points out that in order to see the direct spectrum (positive after-image) all extraneous light must be excluded, whereas “it is difficult to gain the reverse spectrum (negative after- image) where there is no lateral light to contribute to its formation.” “The reverse spectrum is instantaneously converted into the direct spectrum by excluding lateral light, and the direct into the reverse by admitting it. . . . .” “The green spectrum which is perceived on removing the eye from a piece of red silk to a sheet of white paper, may either be called the reverse spectrum of the red silk, or the direct spectrum of all the rays from the white paper except the red, for in truth it is both.” Thus the “direct spectrum” is the sensation of each colour persisting after the cause that produced it has ceased to act, and the “reverse spectrum” is the effect of compound colours upon the retina, which still remains liable to be excited “by any other colours except the colour with which it has been fatigued.” He proves this by showing that the colour of the “reverse spectrum” depends upon that of the “lateral light,” 7.¢., the light which reaches the eye after the retina has been fatigued; He compares these phenomena with those of taste, touch, and hearing, and shows that each of these senses undergoes a partial temporary paralysis after being strongly excited. Although, therefore, the unaided evidence of the senses might have suggested that each pair of complementary colour-sensa- tions, such as red and green, or blue and yellow, were conjugate iunctions of some nerve structure, it is plain that he desired to empha- sise the fact that they must be regarded as due to separate nerve structures. One nerve, one sensation. The sensation might be weak or strong, according to the physiological condition of the organ at the time, but its character could nut be changed. Darwin, following Newton, refers to seven colours as primaries. In 1792 Wiinsch,t whose method consisted in the superposition of spectra projected upon a screen, stated that a mixture of three colours, namely, red, green, and violet-blue, could be made to match any given tint. For the basis of Young’s theory there existed, therefore, experimental evidence of the small number of the primary colour-sensations, and of their being functions each of some nerve structure specially * ©Phil. Trans.,’ vol. 76 (1786), p. 318. + Compare this paper, Section II (1), p. 208, and Section III, pp. 212, 216, t+ ‘Ueber die Farben des Lichtes.’ Leipzig, 1792. Artificial Colour-blindness to Successive Contrast. 207 appropriated to it. The phenomena of negative and positive after- images were known—it was known that certain colours are altered in hue to an eye previously exposed to coloured light, and that experi- ments on this subject should be made with the spectrum rather than with pigments. The following papers also are of special interest in connection with the present communication. Brewster,* by looking at the spectrum through a coloured medium, was able to trace the green as far as C. His mode of explaining the phenomenon led to a controversy, in which the true merit of the observation was lost sight of. A momentary colour-blindness is, in. reality, as was shown by Hunt, produced by the contrast of adjacent parts of the spectrum differing greatly in brightness. Piazzi Smyth,+ working with a very long spectrum, observed the boundaries of the colours to change when any alteration was made in the intensity of the illumination. The earliest account of a systematic investigation of the effect of retinal fatigue on the colours of the spectrum is in a paper of John. Aitken.{ Similar observations were made by Edmund Hunt,§ who also describes the appearance of the spectrum when observed through certain coloured media, after the manner of Brewster. Coloured figures of the results are given. Both these authors used light much less intense than that employed by me, and did not obtain the full effect. I was not aware of their work until after my own paper had. been read, and therefore take this opportunity of calling attention to it. To these may be added a paper by Hess|| on the Alterations of the: spectral colours by retinal fatigue. | I].—Experimental Investigation of the Phenomena of Successive Contrast. Successive contrast is an effect of two stimuli—a primary stimulus by which the retina is fatigued, and a secondary stimulus, the effect of which is modified in consequence of the first. The colour-sensations excited may be reduced to four at the most. To arrive at the funda- mental laws of contrast, we may vary the conditions in the following manner :— Let the first stimulus excite a single colour-sensation, taking each in turn, or separated from the rest as in the spectrum. Four cases: arise : (1.) The second stimulus may excite all the sensations, 7.¢., it may consist of white light. * “Kdin. Trans.,’ vol. 12 (1834), p. 182. + ‘ Roy. Soc. Edin. Trans.,’ vol 28 (1879), p. 792. ~ “ Colour and Colour-Sensation,” ‘ Roy. Scot. Soc. of Arts Proc.,’ 1871-72.. § Hunt, ‘ Colour Vision,’ Glasgow, 1892. | Graefe’s ‘ Archiv fir Ophthalmologie,’ 36, abth. 1, pp. 1—32. 208 Mr. G. J. Burch. On the Relation of This gives the complementary colour, i.c., the direct spectrum, as Darwin calls it, of all the colours save that excited by the first stimulus. (2.) The second stimulus may excite the same sensation as the first, but less strongly. This gives a black after-image. (3.) The second stimulus may excite two or more colour-sensations, including that of the first stimulus. The colour of the resulting image is that of the second stimulus minus the first. (4.) The second stimulus may excite one or more sensations, none of which were included in the first stimulus. The colour of the resulting image is that of the second stimulus plus an admixture, usually small, of the first. This scheme is covered by the experiments described in the following pages. They are so arranged as to require but little special apparatus, and to employ spectral colours by direct observation. Some little care must be taken to adjust the relative intensity of the two stimuli correctly, and to effect the change from one to the other as suddenly as possible. fl 1. After fatiguing the retina by the spectrum, to observe a uniform white light. This is most easily done by means of a low-power spectroscope with a reflected-scale tube. Unscrew the cap containing the scale, and place a mirror so as to reflect white light from the sky into the tube. Cover the open end of the scale tube by a black card held in the right hand, and have a similar card in the left hand in readiness to cover the slit of the spectroscope. Let the slit be fairly wide so as to give a rather bright spectrum. Look steadily at, the spectrum for half a minute, keeping the eye fixed on the intersection of the cross wires, and then suddenly cover the slit and uncover the scale tube. A complementary spectrum will be seen, brilliantly defined, for a fraction of a second. To myself, by daylight the spectral red is replaced not by green, but by blue, and the complementary of green is a pinkish purple, but by lamp- light the complementary of red is green, and that of green is red. The advantage of this mode of experimenting is that it utilises existing apparatus. 2. After fatiguing the retina by the spectrum to observe a less intense spectrum. The phenomenon of successive contrast is shown by the preceding method in its least simple form. To analyse it, the effect of retinal fatigue by each spectral colour on the perception of that same colour must be determined. This may be done by focussing with a lens A | (fig. 1) a glow lamp B on the slit C of an ordinary spectroscope, and at the same time illuminating it by a second glow lamp D placed between Artificial Colour-blindness to Successive Contrast. 209 the lens and the slit. The effect produced is that of a broad, continuous. spectrum, with a narrow but much brighter spectrum in the middle of it. After a few seconds a black card is suddenly brought behind the. lens, so as to screen off the light of the focussed lamp B. A dark band. like a shadow instantly appears in place of the narrow bright spectrum— that is to say, the effect upon the retina of light of any wave-length is to blind the eye temporarily for light of that same wave-length. This may be illustrated in another way. Place near the slit of the spectroscope a. Bunsen burner, and behind it, a few inches farther off, a lamp, and hold. between the lamp and the Bunsen flame a black card. Burn some calcium or strontium chloride, or common salt, or anything that gives a good bright-line spectrum, in the Bunsen flame, keeping the eye fixed on one of the lines. On snatching away the card and the Bunsen flame a dark-line spectrum will be seen momentarily against the con- tinuous spectrum of the lamp, so sharply defined that it is difficult to realise that it is merely an illusion. These results are of cardinal importance. They mean that the green or blue subjective impression produced by a white surface when the eye has been fatigued for red does not indicate that red excites an after-sensation of green or blue, or renders the eye more sensitive to green or blue, but that the eye has become less sensitive to: red. And similarly with the other colours. This point is clearly brought out by Darwin. An “accidental” colour has therefore this. in common with an absorption spectrum—that it involves a diminution | of the intensity of a certain portion or portions of the spectrum. The line of proof is completed by the third disposition of the vari- ables. 3. After fatiguing the retina by any one colour, to observe the entire spectrum. This is in effect a mere variant of the method described in my paper,* and depends on the production. of a very transient colour-blindness. It is necessary to make special arrangements for suddenly substituting a complete spectrum for a field of view illuminated by monochromatic light. Among the methods I have tried, the following may be mentioned :— * © Phil. Trans.’ B, vol. 191 (1899), p. 4. 210 Mr. G. J. Burch. On the Relation of (1.) Place a coloured screen over the end of the scale tube of an ordinary spectroscope, the scale being removed, and cover the slit with a card. After looking at the coloured light for some seconds, cover the scale tube, and simultaneously uncover the slit. This ex- periment is easily tried, but is open to the objection that the first stimulus is not perfectly monochromatic. (2.) Illuminate the scale tube with monochromatic light from a prism, and proceed as before This plan obviates the difficulty referred to, but does not afford sufficient light to produce the ful effect. (3.) A single prism A, fig. 2, with collimator B, and slit C, is fixed near the back surface of the last prism D of the large spectroscope in such a position that the rays from it are reflected into the telescope E of the large spectroscope. This second spectrum is of course much fainter than the one observed directly through the instrument, but that is an advantage rather than otherwise. It is only necessary to arrange two black cards with slits in them in sucha way that when light passes through the first spectroscope the second is obscured, and on touching a spring the conditions are reversed. In order to make the effects more marked, a short slit diould be used for the large spectroscope, so that it may give a band of mono- chromatic light across the middle of the field, fairly bright but rather narrow. The eye should be fixed on the centre of this band. After afew moments, by the action of the spring referred to, a black card is suddenly brought over the slit of the spectroscope, shutting off the light, while at the same moment the screen is removed from the other spectroscope, and the complete spectrum appears, filling the entire field of view. For an instant, a dark shadow is seen, not extending Artificial Colour-blindness to Successive Contrast. 211 across the entire spectrum, but only that part of it corresponding to the colour-sensation excited by the monochromatic light. The effect is very striking after red light. An intensely black band cuts through the spectrum from the ultra-red, as far as C, where it begins to fade away into a pure green. After violet light, a similar black band cuts through the spectrum from the ultra-violet, and if care has been taken not to implicate the blue in the fatigue, the black band fades away into blue. After green light, most frequently the red and blue are seen to stretch across and meet in the middle of the 0 lines; but sometimes, if the exposure is exactly right, a well-marked darkening of that part of the spectrum is seen. Blue light is the most difficult to manage, unless a wide dispersion is used, the blue being otherwise not sufficiently separated from the green and the violet. After getting the adjustments right, it is better either to wait ten minutes, or use the other eye. With these precautions, it is easy to see the green and violet meet in the place of the blue, and to note that the after-image of the blue casts no shadow on the violet near H. Sometimes a momentary shadow may be seen in the blue. This experiment is of interest as affording additional evidence of the existence of a separate sensation for blue. It should be noted that the two spectra must have the correspond- ing colours on the same side. If the prism of the second spectro- scope is reversed so as to bring the red of one spectrum towards the violet of the other, a black shade, very well defined, can be produced in any part of the second spectrum. For if the two spectra are so arranged that any given portion of the one corresponds with the same wave-length on the other, then no part of the one spectrum on either side of that one part will be of equal wave-length with the portion of the other spectrum which coincides with it. Accordingly, a negative after-image will be produced only of the short space within which the wave-lengths are approximately the same. But it is clear that such an experiment is more curious than useful. III. Contrast Phenomena by Intermittent Stimulation. In 1868 Sigmund Exner* made a series of experiments on the fol- lowing plan. After an interval of darkness, he presented to the eye, for a fraction of a second, the image of a small white disc. This was succeeded by a disc of considerably larger diameter, which in turn was followed by darkness. The illumination of either disc, and the period during which it was visible, could be independently varied. Thus the portion of the retina on which fell the image of the small disc received * Exner,‘ Sitzungsberichte d. Wiener Akad.’ Abth. 2, vol, 58 (1868), pp. 601 —632. 212 Mr. G. J. Burch. On the Relation of not merely the light of the larger disc, but in addition the light of _the smaller disc. Yet in spite of this, with certain relations of in- tensity and duration between the two images, the total sensation evoked by the larger quantity of light was less, so that the small disc appeared as a black spot on the larger disc. I was curious to ascer- tain to what extent this principle could be carried. By the following experiment it may be demonstrated in a striking manner. A card- board disc, A, figs. 3, 4, 200 mm. diameter, of which 180° is black and. Picea. Fie. 4. re the rest white, has a slit about half a millimetre wide cut at the junction of black with white. It is made to revolve so that black precedes and white follows the slit. An incandescent lamp B is placed behind the disc, and another C in front of it, so as to throw a strong light upon its surface.. While the disc is moving slowly the incandescent filament of the lamp B, seen through the slit as it passés across, looks bright against the white card, but when a certain speed is reached it appears as a black thread against a brighter background. In spite of the very short duration of the intense light of the fila- ment, the fatigue induced by it is out of all proportion to the sensa- tion it excites, and in consequence the less fatiguing illumination of the white card produces a greater effect on the senses than the sum of the sensations due to the filament and the subsequent light. With a single flash the filament looks, to the rested eye, black all over, but with a succession of flashes there is generally an appearance as though the luminous filament were partly covered by an opaque black thread, but could be seen in places behind it. This I think is due to the shifting of the images on the retina, which after the first flash is no longer in a uniform condition. The retina is, as it were, scarred with after-images, and the ratio of illumination to length of flash which suits one part is incorrect for another. When a red glass is placed over the lamp the reversed image of the filament is green if the card is not quite white, or if the light falling on it is yellowish, but blue-green or blue if it is illuminated by sun- light or the arc lamp. The use of coloured glasses, however, so far diminishes the light that in most cases a disc with a different flash tatio has to be employed, and the lamp C placed at a greater | distance. 4 , Artificial Colour-blindness to Suécessive Contrast. 213 I next attempted to get the reversed image of the sun. This was attended with difficulty, owing to the lack of a disc witha slit sufficiently fine to reduce the sensation evoked by the direct light of the sun within the limits required. I succeeded at last by increasing the intensity of the illumination of the white card. This was effected by holding in front of it a large lens by which the sun’s rays could be concentrated, the degree of concentration being regulated by adjusting the distance of the lens from the card. In this way I was able to see the sun’s disc black upon a white ground. The experiment tended to confirm the explanation already given of the appearance of the incandescent filament under similar conditions. If the visual axis was not fixed, three or four black discs would appear, and on looking directly at one of the more central ones, the sun’s disc seemed to be — partly visible behind it. A very curious effect was produced by “sweeping” with the eye along a faint circle marked on the revolving eard. A whole series of black discs started into view one after the other without a glimpse of the luminous disc that produced them. The principle underlying Sigmund Exner’s method is illustrated in an even more striking manner by the remarkable experiment of Shelford Bidwell. In this a coloured object is placed behind a disc half black and half white, with a sector 30° wide cut out of the white portion. As the disc revolves, the eye is kept in darkness for a space, then sees the coloured object for a short time, and immediately after- wards a white surface for a considerably longer time. The retinal fatigue induced by the colours of the object causes a negative after- effect so strong that the object is seen in its complementary colours. From the point of view of my own investigations it was necessary to repeat these experiments with the pure colours of the spectrum. There are several positions in which a disc, such as Shelford Bidwell employs, can be used in conjunction with a spectroscope. It may be placed between the prism and the telescope, the latter being set back an inch or two to make room for it, or it may work in a gap cut in the body of the telescope, being illuminated by front light through a side tube. But either arrangement involves some alteration of the spectro- scope. The following method is free from this objection and has a certain interest of its own :— The disc is placed in front of the eye-piece of the spectroscope, and the spectrum viewed through a second telescope fixed in the optic axis an inch or two from the eye-piece. But the second telescope magnifies the spectrum and consequently renders it less bright. The definition is, however, much better than would be expected, and is so little affected by slight displacement of the second telescope from the optic axis that it occurred to me to try the arrangement shown in figs. 5, 6. A telescope A, magnifying ten times, is placed with its eye-piece close to the eye-piece E of the spectroscope. The disc B revolves VOL. LXVI. $ 214 Mr. G. J. Burch. On the Relation of between the objective of A and the objective of a second telescope C, magnifying five times. The first telescope A being inverted, diminishes the image to one-tenth of its size, and the second tele- scope only magnifies it five times, so that it appears to the eye half the size it would without the telescopes, and correspondingly brighter. Although the two telescopes were merely supported by Fie. 5. retort stands and roughly adjusted, the definition was quite good enough and the light strong enough to show the complementary spectrum extremely well. But the white light used was merely that reflected from card, and was in consequence weak in the extreme violet rays. By the following arrangement white light reflected from a mirror may be employed :—- The dise A, figs. 7, 8, which is 25 cms. in diameter, has two sectors of 30° aperture, and reaching within 2 cms. of the centre, cut away at opposite ends of a diameter. The disc B, 15 cms. in diameter, has two narrow slits of about 1° or 2° aperture and 180° apart. Both discs are blacked and monnted upon the same shaft, which is furnished with a nut and broad washer, so that they can be clamped together. The Artificial Colour-blindness to Successive Contrast. 215 shaft is so fixed that the slits of the smaller disc may revolve close in front of the slit C of the spectroscope, an ordinary single-prism instru- ment, furnished with a reflected-scale tube, the scale being removed, leaving the tube open. A mirror placed at D in front of the sectors of the larger disc reflects light from the sky on to a second mirror E, by which it is reflected into the scale tube, causing the field of view to be filled with a soft white light. Fig. 8. The mirrors and discs must be adjusted until on rotating the shaft slowly the flashes occur in the following order :—First a sharp flash through the small disc, giving a momentary view of the spectrum. As soon as possible after this is over, but not before, there is a rather long soft flash of pure white light, followed by a much longer period of perfect darkness. The reason for having two slits and two sectors on the discs is simply that they may be well balanced on the shaft, and therefore rotate more steadily. There should be no overlapping 216 Mr. G. J. Burch. On the Production of of the spectral flash by the white flash, a short interval of darkness between them being preferable to the smallest overlap. For this reason the shaft is fitted with a screw nut, which being slackened, the angular position of the slits with respect to the sectors can be accu- rately adjusted. On rotating the discs steadily, but not too quickly, a spectrum of complementary colours is seen with the greatest distinct- ness. By placing a narrow strip of black card across the mouth of the scale tube, a portion of the white flash may be stopped out, allow- ing the normal spectrum to be seen in that part of the field. It is necessary, however, to shade the corresponding part of the slit some- what, so that the normal spectrum may not overpower the comple- mentary spectrum. ‘The colours as I see them are as follows :—Red is replaced by Prussian blue, green by purple (a red shade of Hoffmann’s violet), blue by orange, and violet by yellow. ‘To show the comple- mentary of violet it is necessary to use sunlight, or, better still, the arc light. I have never been able to see it properly by any of the methods involving the use of white card or paper surfaces as reflectors. The experiments of Section 3, for which a wide dispersion was required, were made with a large direct-vision spectroscope belonging to the Marlborough Collection, for the use of which I am indebted to the Aldrichian Demonstrator of Chemistry, Mr. W. W. Fisher. I have also to thank Professor Gotch for the use of the electric light in the physiological laboratory. The remainder of the work was done at Reading College, and the expenses have been defrayed by a portion of the sum of £10 allotted to me by the hee Society out of the Government Grant. “On the Production of Artificial Colour-blindness by Moonlight.” By Grorce J. Burcu, M.A. Oxon., Reading College, Reading. Communicated by Professor Gorcu, F.R.S. Received January 30,—Read February 8, 1900. Since the publication of my paper on “ Artificial Colour-blindness ”* I have found a very general and not unnatural tendency to regard the results described therein as phenomena of a pathological condition induced by the severe strain to which the structures of the eye had been subjected. In my paper I indicated, perhaps too briefly, that this could not be the case, since “the same general phenomena are observable alike with strong sunlight and with the faintest light the eye is capable of perceiving.” The purpose therefore of the present communication is to describe some of the experiments on which that statement was based. * ‘Phil, Trans.,’ B, vol. 191 (1899), p. 1. Artificial Colowr-blindness by Moonlight. 217 When green-blindness is induced by exposure of the eye to intense green light, not only is the observer unable to perceive the colour of green objects, but the sensation of green is no longer excited by the intense green light that caused the blindness. And the same may be said of blue-blindness. On the other hand, with artificial red-blindness the exciting light still looks reddish, though greatly dulled and much paler in hue, but all objects less brightly illuminated fail to excite the red sensation. Probably in the case of green-blindness the green sensation is not entirely destroyed, but reduced so much that the red and blue sensations, which are also excited by that same part of the spectrum, completely overpower it. In producing red-blindness, as there are no colours to the left of red, I have generally used a part of the spectrum which excites only the red sensation, and which therefore must con- tinue to appear red if visible at all. But I have not thought it desir able to push the fatigue of the retina far enough to destroy the sensation of light. For the mere demonstration of the phenomena of colour-blindness, light of quite moderate intensity is amply sufficient if the precaution is taken of shielding the eye from all other light during the experiment, and of giving it time to recover from the effects of previous illumina- tion. The colour-blindness so produced is, however, not absolute, but merely relative, the sensation which has been fatigued, whether red, green, blue, or violet, being still excited by a stronger stimulus. The following is perhaps the most striking and suggestive way of making the experiment :— 1. IT exposed my left eye to direct moonlight in the focus of a lens behind a screen of ruby glass combined with a gelatine film stained with magenta. After three minutes I looked through a spectroscope directed to the moon. The red had entirely disappeared, and only the green, blue, and violet were visible. With the right eye I could see the red as well as the other colours. 2. I exposed my right eye in the same manner to moonlight, using a screen of green glass instead of the red. On looking through the spec- troscope I found the green sensation had entirely vanished, the red meeting the blue in the same part of the spectrum, viz., between E and b, as in my experiments with sunlight.* The violet was easily distin- guished from the blue in this case also. The left eye was still partially red-blind, and the contrast between the spectrum as seen by it and by the right eye was very marked. I was unable to use spectral colours for fatiguing the eye because the full moon is not visible at this season of the year from the laboratory in which the large spectroscope is mounted, and the intensity of the light was too much reduced by reflection from the two mirrors of the * ‘Phil. Trans.,’ B, vol. 191, Plate I, figs. 4 and 5. 218 Mr. G. J. Burch. On the Production of : heliostat. I did not therefore make any observation with blue or violet light. As may be imagined, it is necessary to use a larger lens with moon- light than with sunlight. In practice I have found an ordinary reading lens, of 4 inches diameter, sufficient. To obtain the full intensity of illu- mination, the focal length should be such that the moon’s image may be not smaller than the pupil of the eye. There are two points of interest in connection with this experiment. The first is that the illumination of surrounding objects is on the same scale, as regards contrast of light and shade, in moonlight as in sun- light—that is to say, in each case the source of light is an object sub- tending an angle of about 30’ at a distance which is practically infinite. Whatever difference may seem to exist must be of physio- logical or psychical origin. The deeper shadows in moonlight probably afford too little stimulus to fully excite the sensation of vision even in an eye accustomed to darkness ; but it must not be forgotten that we accentuate this difference by a habit of looking at the moon itself and at the bright sky near it, thus blinding ourselves to the faintly illumi- nated details of the shadows. If we were to do the same with sunlight the shadows would seem equally lacking in detail. In a room arti- ficially lighted there is seldom so much contrast between lights- and shadows. Light-coloured objects are usually to be found in close proximity to the lamps, even where white shades or globes are not used to diffuse the light. It is less easy to demonstrate the phenomena of temporary colour-blindness under these circumstances, owing to the greater relative intensity of the dazzle-tints* resulting from the action of the diffused light before the experiment began. Until these are gone the retinal fatigue is not confined to one colour. If in experimenting with moonlight the observer accidentally looks at the moon’s disc before his eye is protected by the coloured screen, a well-defined after- image is produced, and the subsequent phenomena of colour-blindness are only locally modified, whereas if an after-effect even of less intensity, due to diffused light, is present, the colour-blindness may be to a great extent masked. The other point of interest in connection with this experiment is that colour-blindness has been produced by light no stronger than that reflected by ordinary pigments in sunshine. That this is so is evident if we look at the moon’s disc in the daytime through the same red glass and lens and compare it with a piece of coloured paper. It can therefore be hardly maintained that the condition of temporary colour- blindness should be regarded as a pathological result of excessive stimulation of the colour sensations. Merely to look for a few seconds * It will be convenient, in describing my own experiments, to retain this word, which I have used to signify the “elementary component sensations of the positive after-effect.” ‘Phil. Trans.,’ B, vol. 191 (1899), p. 6. Production of Artificial Colowr-blindness by Moonlight. 219 at a scarlet poppy in a cornfield causes a measurable degree of red- blindness for the next two or three minutes. In applying the spectro- scopic method of measuring the colour sensations described in my previous paper* it is necessary to guard against this source of error. Last summer I had a case of a man who seemed at first to be very nearly green-blind. His green sensation did not reach more than half way between 0) and F, even after fatiguing for thirty seconds with blue light, and it was correspondingly shortened on the red side. After conversing with him for some time in the subdued light of the labora- tory I repeated the measurements, and found that his green sensation then extended considerably beyond F. It appeared that he had been strolling about the Parks on the grass in the bright sunshine. I have myself frequently experienced a temporary green-blindness from a similar cause. The effect seems to be intensified by looking at a white surface, as, for instance, in reading a book while sitting on the grass. After a time the green leaves seem to lose their colour and become greyish. This effect may be often noticed during a long walk through the fields. If during this condition the eyes are directed to a small red spot on a black surface, as, for instance, a single geranium petal on the black cover of a book, and the observer walks with it quickly into a dark shed or barn, the colour of the geranium petal will seem to change from red to orange and then to yellow, and finally almost whitish, owing to the subjective admixture with the red of the green dazzle-tint. On coming out into the light again the red colour will reappear. These changes are similar to those observed in the red end of the spectrum during green-blindness on opening and closing the slit ;— and as the experiment requires no apparatus, I have recom- mended it in my lectures for the last two years. The retinal fatigue induced by white light under various conditiens forms the subject of a recent paper by Beck.t * ‘Phil. Trans.,’ B, vol. 191 (1899), p. 19. ¢ ‘Phil. Trans.,’ B, vol. 191 (1899), p. 8, and Plate I, figs. 4 and 5. ~ Archiv fiir die ges. Physiologie,’ vol. 76, p. 634. YOL. LXVI. T Proceedings and List of Candidates. March 1, 1900. The LORD LISTER, F.R.C.S., D.C.L., President, in the Chair. A List of the Presents received was laid on the table, and thanks ordered for them. | In pursuance of the Statutes, the names of Candidates for election into the Society were read as follows :— Adeney, Walter Ernest, D.Sc. Alcock, Alfred William, Major, I.M.S. Allen, Alfred Henry, F.C.S. Ardagh, Sir John, Major-General, R.E. | Ballance, Charles Alfred, F.R.C.S. Bourne, Gilbert C., M.A. Bovey, Professor Henry T., M.A. Boyce, Professor Rubert. Bridge, Professor Thomas William, M.A. Brown, Adrian John, F.C.S. Brown, John. Bruce, John Mitchell, ML D. Budge, Ernest A. Wallis, D.Litt. Burch, George James, M.A. Callaway, Charles, D.Sc. Cardew, Philip, Major, R.E. Clowes, Frank, D.Sc. Copeman, Sydney Monckton, M.D. Crookshank, Professor Edgar March, M.B. Darwin, Horace, M.A. David, Professor T. W. Edgeworth, B.A. Dixon, Professor Alfred Cardew, MA Dixon, Professor Augustus Kd- ward, F.C.S. Dyson, Frank Watson, M.A. Farmer, John Bretland, M.A. Feilden, Colonel Henry Wemyss. Gray, Professor Thomas, B.Sc. | Hamilton, Professor David James, M.D, Hardy, William Bate, M.A. Harmer, Frederic William, F.G.S. Hiern, William Philip, M.A. Hill, Leonard, M.B. wit Hills, Edmond Herbert, Captain, R.E. Hopkinson, Edward, M. A. Horne, John, F.G.S. Jackson, Henry mregirentitie Captain, R.N. Knott, Cargill Gilston, D. Sc. Letts, Edmund Albert, D.Sc. Lewis, Sir William Thomas, Bart., M.Inst.C.E. Lister, Joseph Jackson, M.A. MacArthur, John Stewart, F.C.S. Macdonald, Hector Munro, M.A. MacGregor, Professor James Gordon, D.Sc. Maclean, Magnus, D.Sc. Mallock, Henry Reginald Arnulph. Mance, Sir Henry C., C.L.E. Manson, Patrick, M.D. Marsh, James Ernest, M.A. Martin, Charles James, M.B. Mather, Thomas. Matthey, Edward, F.C.8. Meyrick, Edward, B.A. Researches on Modern Explosives. ‘221 Mill, Hugh Robert, D.Sc. ‘Swinton, Alan Archibald Camp- Muir, Sivinas, M.A. bell, Assoc. M.Inst.C.E: — -:. ‘Oliver, John Ryder, Major-General | Symington, Johnson, M.D... (late R.A.). | Tatham, John F. W., F.R.C.P. Payne, Joseph Frank, M.D. _ Thomas, Michael Rogers Oldfield, F.ZS. Ulrich, Professor okie Hae Perkin, Arthur George. Rambaut, Professor Arthur A., M.A. | Frederic, F.G.S. Russell, James Samuel Risien, | Walker, James, M.A. M.D. Walker, Professor James, D:Sc: Salomons, Sir David, M.A. | Waterhouse, James, Colonel... . Saunders, Edward. | Watkin, Colonel, R.A. Schlich, Professor William, C.I.E. | Watson, William, B.Sc. Sell, William James, M.A. _ Watts, Philip. : Sidgreaves, Rev. Walter, S.J. | Whetham, William C. D., M.A. Smith, James Lorrain, M.D. White, William Hale, M.D. | Smith, Professor William Robert, Wilson, Charles T. R., M.A. M.D. - Woodhead,Professor German Sims, Smithells, Bidet Arthur, B.Sc. | M.D: Spencer, — W. Bildiwiky Woodward, Arthor Smith, F. G. s. B.A: Wright, Professor Bdwalia Per- Swinburne, James. | ak M.A. ranrye The following Papers were read :— I. “An Experimental Inquiry into. Scurvy.” By F. G. Jackson and VAUGHAN HARLEY. Communicated by Lorp LIsTEr, P.R.S. Bite: Il. “The Velocity of the lons produced in Gases by Rontgen Rays.” By Professor J. ZELENY. Communicated by Professor J. J. THOMSON, F.R.S. | III. “ Mathematical Contributions to the Theory of Evolution. VIII. — —On the Correlation of Characters not Quantitatively Measur- able.” By Professor KARL PEARSON, F.R.S. “Researches on Modern Explosives. Second Communication.” By W. Macnas, F.I.C., and E. Risrori, Associate M. Inst., C.E., F.R.A.S. Communicated by Professor Ramsay, F.RS. Received January 29,—Read February 1, 1900. In our communication published in the ‘ Proceedings of the Rasa ‘Society,’ vol. 56 (in which we gave results of the determination of, the calories evolved and analysis of the products of combustion of various PZ 229 Messrs. W. Macnab and E. Ristori. explosives), reference was made to certain experiments we had then begun for the purpose of determining the actual maximum tempera- ture reached during explosion. We have now made a long series of experiments in this direction, and propose to communicate some of the results so far obtained, although the research is not yet complete. Some experiments have already been made by others for the pur- pose of determining the temperature during explosion by placing strips of metal of different melting point in a closed bomb and observing the result after firing. Noble and Abel in their well-known communication on explosives found in this way that the temperature produced by the explosion of black gunpowder was slightly above the melting point of platinum. There have been also several communications, on the same subject, made by others, who have deduced the temperature during explosion from theoretical considerations, but these calculations involved assump- tions which as yet do not rest on an experimental basis. It appeared to us desirable, therefore, to endeavour to determine experimentally, and with greater accuracy than has hitherto been done, the actual temperature developed when an explosive is fired in a closed vessel. The practical solution of this problem is, however, beset by several difficulties; amongst others, the intensity of the temperature, the extreme shortness of duration of the maximum temperature, and the necessity of conducting the explosive reactions in a closed vessel. We were led to try a modification of the pyrometric method - developed by Sir W. C. Roberts-Austen, by observing that a thin platinum wire used for firing the explosive in the vessel by electricity was often melted by the heat produced by the explosion, while thicker platinum wires, which served to support the jaar containing the explosive, were unaffected. This showed that the temperature reached was above the melting point of platinum, and also that the duration of the maximum tem- perature was very short. In the case of the thin wire, the small mass of the metal allows the heat to penetrate it with sufficient rapidity to raise it to the melting point before the period of maximum temperature s past, while with the thick wire the time does not suffice for the larger mass to be heated to the same extent. These considerations led us to argue that if rhodium-platinum couples of wires of different diameters, sufficiently thick not to be melted during explosion, were used in a bomb, the deflections of the galvanometer indicated would vary inversely with the sizes of the wires forming the couples; that in this way we might get data which would enable us to calculate the deflection of an infinitely thin couple which could be capable of taking up the heat in an infinitely short time, and that this deflection, expressed in degrees, would repre- sent the actual maximum temperature reached. We also expected Researches on Modern Explosives. 223 that the indication of the galvanometer would show the rapidity with which the temperature rose at the moment of the explosion, as moll as the rate at which the cooling took place. Through the kindness of Sir W. C. Roberts-Austen we were enabled to make some preliminary experiments in his laboratory, connecting the wires from a couple in our bomb with the galvanometer in his photographic-recording apparatus. The results of these experiments were so encouraging that a similar photographic-recording apparatus was procured, only introducing such slight modifications as were required to make it more suitable for our purpose. We also had a special lid made for the calorimetric bomb previously described in our former communication, this lid being similar to the other, with the exception of two insulated conical pins, one made of pure platinum and the other of platinum alloyed with 10 per cent. of rhodium. These pins were used in the inside of the lid as points of attachment for the thermo-couple, their ends outside the lid being con- nected with the galvanometer. The couples were made of platinum and rhodium-platinum wires in contact. Several couples of different thicknesses were prepared for us by Johnson and Matthey, fusing the ends of the platinum and rhodium-platinum wires together and then drawing the junction through a die until it had the same diameter as the rest of the wire. As we shall have to refer to these couples by number in future, we give in Table I the diameters and areas, and the number by which we distinguish each couple. As will be seen, ten different couples were made, the diameters varying from 0-044 to 0:01 of 1 inch :— Table I. Couple No. Diameter in inches. Area in sq. inches, 1 - 0°044 0-00152 2 0-040 0:00125 3 0-035 0-00099 + 0-028 0-00061 5) 0-026 000053. 6 0-022 p 0:00037 7 0-018 0:00025 8 0-015 0:00017 9 0-012 000011 10 0-010 0°00008 Fig. 1 shows the explosive bomb with the lid separate, and the arrangement of connections which are made through the lid are clearly seen. ‘Three insulated pins pass through the lid ; two of them, as above indicated, are those which connect with the thermo-couple, the other one is used as one of the terminals for the firing wire, the © 224 Messrs. W. Macnab and E. Ristori. other terminal being attached to the body of the bomb outside. Thus there are two electric circuits absolutely insulated from each other. The position of the cup which contains the explosive can be seen in the figure (fig. 1), and the position of the couple in respect to the cup BG. 1. \ aaa mn! iN ow a has been varied in a way that will be explained later on. The con- nections from the firing wire were led to an electric battery, which is set in action when it is required to ignite the explosive. | The connections from the thermo-couple are led to the galvanometer, which is inside the recording apparatus shown in fig. 2. The general details of the apparatus are well known, and have been described by Sir W. C. Roberts-Austen. ‘ Lime-light is used to throw a spot of light on the mirror of the galvanometer, from which it is thrown on to a photographic plate. through a horizontal slit, equal in length to the breadth of the plate. This photographic plate is held in a weighted frame, which falls past the slit with a rate of descent uniformly regulated by clock-work, con- Researches on Modern Explosives. ay trolled by adjustable vanes. The rate of descent of the plate in the experiments was one inch in 24 seconds, or equal to four-tenths of an inch for every second of time. Fig. 2. In using the apparatus, a datum line is first traced, and this is done by letting the plate fall with the galvanometer at rest, the circuit being closed, when the spot of light traces a continuous vertical line at one edge of the plate. The plate is then brought back to its original posi-. tion and re-started on its descent. After it has fallen a short distance, in order to make sure it has acquired a steady rate of motion, the explosive is fired and the thermo-couple in the bomb is in consequence heated, and the current generated deflects the mirror of the galvano- meter, and therefore the spot of light, horizontally and proportionately to the temperature attained by the couple ; and then, as cooling sets in, the original position of the spot of light is gradually resumed. The result is recorded as a curve, which shows the combined move- ments of the spot of light and of the photographic plate. Fig. 3 shows one of the plates on which three records were succes- sively taken. In these three separate experiments, couples of different thicknesses, but with the same charge and kind of explosive, were used. — | | | i 226 Messrs. W. Macnab and E. Ristor1. AA is the datum line; the point of departure a of the spot of light from it shows when the first charge was fired; the spot of light, owing to the deflection of the galvanometer, travelled rapidly to the Fie. 3: MEDIUM COUPLE THIN COUPLE SSS, left, until the maximum point P, was reached, then more slowly returned, as cooling set in, towards its normal position. 0, ¢, Ps, Ps, are the corresponding points in the other two experiments. The maximum deflections, as shown by Pj, Ps, Ps, are seen to be in inverse order to the thickness of the couples used, the greatest deflection being obtained with the thinnest couple. Researches on Modern Hauplosiwes. 227 It is also noticeable that the thinner the couple the sooner is the maximum point reached, 7.¢., the steeper is the curve and the sharper its point, while with a thicker couple the point is more rounded, and the maximum more slowly reached. Fig. 4 shows photographs of five results, two with a thick, and three with a very thin couple. Fie. 4, Fig. 4. THIN COUPLE THICK COUPLE Many difficulties were encountered in carrying out these researches in order to secure reasonable accuracy in the work. Several hundred 228 Messrs. W. Macnab and E. Ristori. experiments have been made on the lines above indicated, but a unige number of the results have had to be discarded. To begin with, one point which had to be studied was peice the size of the grain of the explosive would make any difference in the results ; but after numerous experiments we have come to the conclt.- sion a within reasonable limits, the size of the grain exercises no influence under the conditions of these experiments. In all the experiments about tv be described, the explosives used were all gelatinised preparations of gun-cotton alone, or mixed with nitro- -glycerine, all in the form of small grains, and the charges were fir ed in the bomb full of air. _A difficulty observed was that the deflections of the galvanometer were different, everything else being equal, when the position of the couples varied in relation to the position of the explosive. It became necessary, therefore, to carry out a series of experiments in order to determine which position of the couple gave greatest and most uniform results. | _ The following Table II gives the results of the experiments made to find the hottest place in the bomb. The same charge of the same explosive and the same couple were used in all the experiments, the only difference being that the position of the couple in relation to the explosive which was held in a platinum capsule. Three experiments were made in each position. : Table II. { ' Couple No. 5. Deflection of light | ee | Mean. Position of couple bent into capsule in centre of charge..... «ea ne ele alias sae glee | ULM gu lcs een 126 °3 Couple 1” above explosion .. ae cceeee we ne ee | LOO. (i gence 1546 gute: ‘, s chielo- afueta fatal tale alfa etices 156, 163, 153 167 °5 Seer aN Oona ae Foon 168, 165, 163 165 °5 Boer ‘ (furthest practicable)| 160°5, 168° 5, 155°4 | 161-4 These results are shown graphically in Diagram 1, where the vertical distance represents the deflection of the galvanometer, and the hori- zontal the position of the junction of the couple in relation to the charge. The deflection is least when the junction is embedded in the charge, and greatest when about 3 inches above it. As will be seen from the diagram, there is very little variation in the results between 24 inches and 3} inches (the maximum distance allowed by the size of the bomb), and as the actual maximum was shown at 3 inches, we have in all] future experiments placed the thermo-couple at 3 inches above the surface of the charge. Researches on Modern Explosives. 229 Another point which we had to consider was the different deflection caused by the firing of different quantities of the same explosive in the same bomb. DraGrRamM 1. 2 3 3S/nches above charge. Diagram 2 shows the result of firing 4, 5, 6, and 7 grammes of the same explosive in the same volume with couples 2 and 5. DraaRam 2, COMPOSITION OF 70% Guncotton, 30% Nitro- glycerine. CHARGE a =4Grammes. = 5 “ Cc = 6 “ d=7 o 150 S CS) O ie) Deflection in Gq. V? of Couple. 125 53 Area of Wires OLN : ; ; (urease of an inch. The points of deflection have been connected by a straight line simply for the purpose of showing the general parallelism of these observations. ay With charges of less than 4 grammes the results were very irregular, and we have not fired charges larger than 7 grammes, the bomb not being constructed to withstand high pressures. Having thus ascertained the conditions of working which gave fairly concordant results, a series of experiments was made with couples of different diameters, and, by means of ‘introducing some suitable resist- ance into the circuit of the galvanometer, it was arranged that the 230 Messrs. W. Macnab and E. Ristori. deflection when the thinnest wire was used should be about the maxi- mum that the photographic plate could record. | The explosive used for this series of experiments was Ardeer ballistite, composed of 70 per cent. gun-cotton and 30 per cent. nitro- glycerine, and was in the form of thin square flakes. In each case three experiments were made, and the results are shown in Table III. Table ITI. Charge, 4 grammes Ardeer Ballistite, 70 per cent. Gun-cotton, 30 per cent. Nitro-glycerine. } No. of couple.) | Defiection in mm. Mean. Maximum. | j | SS Wise =: mE i 1 85, 81°5, 83:5 83 85 | Z 102, 90:5, 9875 97 102 3 109, 115°5, 1125 115°5 4 131, 128°5, 1388°5 132 °d 138°5 5 99) 140: coe ae 148 149 | 6 | 152-5. 158°5, 151 154. 158-5 | 7 = 6b ee (eee 170 | 8 185 5* 192% 189 192 * After the first experiment, the wire was partially fused, but not broken. It will be seen that the results are fairly uniform, the average variation from the maximum to mean being only between 2 and 23 per cent. In the case of couple No. 8, which is only 0-015 inch in diameter, we found that after the first experiment it was partially fused but not broken, thus showing that we had reached the practical limit of fineness of wire for this particular explosive. The character of the increase of the defiection in inverse proportion to the diameter of the couple is clearly shown in Diagram 3, where two similar curves are shown for two different explosives. The curve A shows the results given in Table III, and the curve B the results of a similar series of experiments made with gun-cotton fired with couples 2, 5, 8, 9, and 10. In the case of gun-cotton, the temperature being so much lower, as clearly shown in the diagram, we have been able to use thinner couples, and there was no fusion up to No. 10, which is.0-01 inch diameter, but a thinner couple (0°005 inch) was fused. The curves have been drawn as nearly as possible following the actual measurements, which are indicated by dots surrounded by circles ; as it will be seen, the curves are very regular, and it is particularly moeteaiie that the curves of the two explosives are very similar in character. ; In order to get comparative data for several explosives, we made bo Se) ey Researches on Modern Explosives. DIAGRAM 8. a = Balistite ;(30% Nitroglycerine, 70% Guncotton). 6 = Guncotton. Deflectio oO re) i Oo / 2 3 4 5 6 7 8 310N®%f Coupe. 152 125 33 66 55 ao. Zo 7 WE Oo i rea of Wires 4 100,000 of an inch. another series of experiments, the results of which are indicated in Diagram 4. DIAGRAM 4. NATURE (2 = 30% Soluble Nitrocotton, 50 2 Nitroglycerine. OF ie = Se 7] “ 350% “ Cc = Cordiée. EXPLOSIVE \ gq = Guncotton. 200; nN a oO Deflection i 5 WN°of Couple. 53 t) Area of Couple in (aa000% aN inch. The vertical lines indicate, as before, the maximum deflection of the galvanometer obtained with charges of 4 grammes with couples 2 and5) for the following explosives : Gun-cotton, cordite, ballistite of 70 per cent. soluble nitro-cotton, and 30 per cent. nitro-glycerine, and ballistite containing 50 per cent. soluble nitro-cotton and 50 per cent. nitro- glycerine. In the same way as in Diagram 2, each of the points corresponding have been connected by straight lines to show the parallelism of the observations. It will be seen from the diagram that the larger the proportion of nitro-glycerine in the ballistite the higher is the temperature during bo: SX) bo Sir Norman Lockyer and Mr. A. Fowler. explosion ; but, on the other hand, cordite, although it contains as much as 58 per cent. nitro-glycerine, owing to the fact that it contains also vaseline, gives a temperature lower than that of ballistite containing only 30 per cent. nitro-glycerine and no vaseline. Of course, the minimum deflection is the one due to gun-cotton, which contains no nitro-glycerine. Similar experiments have also been repeated with other explosives and with different charges, and, in every case, the same comparative results have been obtained. The above refers only to a part of the experiments which have been carried out so far. Another series is now in progress for determining the other necessary elements which will be required before we can accurately express the value of these deflections of the galvanometer in degrees of temperature. One important element which comes into play is the inertia of the galvanometer itself in connection with the short- ness of the time during which the maximum temperature exists, and there are also other points which are being investigated, and these will form the subject of a further communication. We have, however, thought it advisable not to delay communicating the above results, as already the described method shows the possibility first of all of obtaining approximately an idea of the temperature during explosion, and, secondly, it shows a clear way by which the comparative temperatures for various explosives can be determined. These, taken in connection with the results shown in our former com- munication, will serve, we hope, to give a better knowledge of the different modern explosives which are now commonly used. “The Spectrum of « Aquile.” By Str Norman Lockyer, K.C.B., F.R.S., and A. FOWLER. Received January 18,—Read Feb- ruary 8, 1900. [Prate 1.] The study of enhanced lines throws considerable light on the spectrum of « Aquilz, the peculiarities of which were first described by Professor Pickering* and Dr. Scheiner? in 1889. In this spectrum the lines of hydrogen are strong and broad, but the additional lines, instead of being faint and sharp as in most other stars of this class, are faint and diffuse. Dr. Scheiner stated that these apparent bands were identical with the most conspicuous groups of lines in the solar spectrum, and further that this appearance of the spectrum can be * Third Annual Rep. Henry Draper Memorial, p. 5. ¢ ‘Ast. Nach.,’ 2924. The Spectvum of « Aquile. 233 imitated by holding a rather faint drawing. of the solar spectrum at such a distance that the individual lines are no longer visible. From a consideration of the photographs taken at Harvard College Observatory, Professor Pickering suggested in 1891 that the diffuse- ness of the lines in the spectrum of « Aquile and certain other stars was perhaps due to a rapid rotation of the star.* That rotation might be capable of producing such effects had already been suggested by Abney in 1877.7 Photographs taken at Kensington with large dispersion, during 1892, led to the adoption of Professor Pickering’s view, and the spectrum of a Aquile was classed with that of @ Arietis, in which we apparently got the same lines quite sharp.j 3 In 1895 Dr. Scheiner again referred to this spectrum, and sug- gested that it represents a transition stage from the first to the second type; as an alternative explanation, he mentions the view that the spectrum may be a composite one, in which a spectrum of the first type is superposed upon one of the second. In a recent paper|| Dr. Vogel has discussed the spectrum of « sdb chiefly with reference to its motion in the line of sight, but he also considers the question of the haziness of the lines. He refers to some experimental photographs which depict the solar spectrum with its lines broadened by a cylindrical lens, or by a photograph taken out of focus, and states that spectra of this kind have been obtained in which the close lines run together so as to produce a spectrum resembling that of « Aquile. He adds that the exact comparison of the two spectra shows that the agreement is not perfect, in particular that the G group is hardly indicated in the spectrum of « Aquile, while it comes out strongly in the solar spectrum when thrown out of focus. He accordingly places the spectrum of « Aquile in his Class Ia 3, of which « Cygni, 6 Cassiopeie, and Procyon are members, and further concludes that the lines are broadened in consequence of rapid rotation, without, however, referring to previous suggestions to the same effect. The general result is that while Vogel classes « Aquile with a Cygni, which, on the meteoritic hypothesis, is a star of increasing tempera- ture, the work at Kensington indicates that it should be classed with stars like 8 Arietis, which there is every reason to believe to be cool- ing. This difference as to facts is so important that the whole question has been re-investigated. * ¢ Annals Harv. Coll. Obs.,’ vol. 26 (1891), Pt. I, p. 21. + ‘Monthly Notices R.A.S.,’ vol. 37, p. 278. - ¥ ‘Phil. Trans.,’ A, vol. 184 (1898), p. 697. § ‘Pub. Ast. Obs. zu Potsdam,’ vol. 7, Part IT, p. 232. | ‘Sitzber. Akad. Berlin,’ Nov. 1898; translated in Re cere Journ.,’ Jan. 1899. ‘] ‘Astrophys. Journ.,’ vol. 2 (1895), p. 346. 234 Sir Norman Lockyer and Mr. A. Fowler. The Kensington Photographs. An investigation of the spectrum of « Aquile was commenced at Kensington in 1890, and, with the various instruments employed up to 1892, fifteen negatives were obtained.* In all these the lines were ill defined, and it was decided to take a special series of photo- graphs “in order to determine whether the haziness of its spectrum lines is invariable.’+ Since then a considerable number of photo- graphs has been obtained, but although variations have been suspected it is found difficult to establish their reality. One thing seems quite certain, namely, that the lines are always ill defined. At the Royal Society Conversazione in 1894, enlarged copies of photographs of the spectra of a Aquile and f Arietis were exhibited which indicated that Pickering’s view that the haziness of the lines is due to rotation is probably correct. Dr. Scheiner’s experiment of photographing the solar spectrum out of focus has since been repeated; but while it was found possible to produce bands in this way, only a few of them agree with those in a Aquile. Among these coincident bands, are 4031—4036 (Mn), 4046 (Fe), 4064 (Fe), 4132—4135 (chiefly Fe), 4143-6—4144 (Fe), 4226°9 (Ca), 4250°3—4251 (Fe), 4260-2—4260°6 (Fe), 4271-3—4271-9 Fe). a the other hand, by taking an out-of-focus enlargement of a negative of the spectrum of @ Arietis, the violet being put more out of focus than the blue, the spectrum of « Aquile is almost perfectly reproduced (see Plate). The difference in width of the bands appears to be sufficiently explained by the gradually increasing dispersion in prismatic spectra as the violet end is approached. With the instru- ment employed at Kensington a tenth-metre near » 4046 is repre- sented by a distance on the photographs about 1-4 times as great as that corresponding to the same difference of wave-length near A 4384 ; and since the velocity which would produce a displacement of one- tenth metre at » 4384 would produce a displacement of 0°92 tenth- metre at A 4046, the displacements on the photographs for the same velocity, with the particular instrument employed, will be in the pro- 3 portion of 1 to 1:29 at A 4384 and A 4046 respectively. Classification of the Star. This experiment appears to be a sufficient demonstration of the essential similarity of the spectra of a Aquile and f Arietis, so that the former conclusion that the two stars should be classed together is perfectly justified. * ¢ Phil. Trans.,’ A, vol. 184 (1893), pp. 685-688. + Ibid., p. 696. The Spectrum of «a Aquile. 235 According to the earlier work at Kensington, stars like 6 Arietis were classed in Group Va,* that is, between stars like Sirius and those like Procyon. The work on enhanced lines which has been done since then enables us to carry on the work of classification with much greater precision, since we have now a means of estimating relative temperatures with considerable accuracy. In this way we learn that stars at each stage of temperature fall into two groups, one of which represents stars of increasing temperature, and the other including stars of decreasing temperature. «a Aquile and £ Arietis fall in the latter group, and are to be regarded therefore as stars in which photo- spheres have formed. The later work on the classification of spectra has shown that it is sufficient for all practical purposes to include both in the Sirian group of stars. This question of classification is further elucidated by a more detailed examination of the spectrun of a Aquile in relation to a Cygni and the Sirian stars. The foregoing demonstration of the likeness between a Aquile and 6 Arietis leads us to expect that the origins of the lines in the spectrum of a Aquilz will be the same in the main as those of 6 Arietis and Sirius. In these stars the temperature of the absorbing vapours is intermediate between that of the arc and that at. which enhanced lines appear alone, so that the spectra show both are and enhanced lines. The origins of the chief enhanced lines in the spectrum of Sirius have already been investigated,t and practically the same lines occur in f Arietis. Besides these enhanced lines — there are several well-known arc lines, such as the iron triplets and the blue line of calcium, which can be readily identified. The origins of some of the lines of both classes are shown in the plate which accompanies this paper, enhanced lines being shown at the bottom and are lines at the top. It will be seen that enhanced lines of iron appear in a Aquila, but have not the same relative intensity as in « Cygni; the most enhanced. line of iron (A 4233°3), for example, which in « Cygni is represented by a very strong and well-defined line, is in a Aquile very weak and hazy. On the other hand, some of the enhanced lines of iron less. refrangible than Hy are fairly prominent. The principal enhanced lines of magnesium, strontium, and titanium are also certainly present, as. shown in the plate. The enhanced double line of silicium at AA 4128-1, 4131°1, if present, is very weak, a moderately strong hazy line, rather less refrangible than the silicium double, making it rather difficult to. determine whether the latter is certainly present. Among the arc lines present are those of the iron triplet in the: violet (AA 4045-90, 4063-76, 4071:°79), which are clearly seen, but the iron triplet in the blue (Ad 4383°70—4415:27) cannot be identified. * «Phil. Trans.,’ A, vol. 184 (1893), p. 726. fT ‘Roy. Soc. Proe.,’ vol. 65 (1899). Plate 7. VOL. LXVI. U 286 Sir Norman Lockyer and Mr. A. Fowler. with certainty. The place occupied by the manganese quartet (AA 4030°88—4035°88) is covered in the spectrum of the star by what appears to be a broad hazy line, which is probably composed of the individual components of the quartet merged together. The arc line of calcium at » 4226-90 is one of the most prominent lines in the spectrum. The classification of the spectrum of « Aquile may therefore be considered as settled; it does not sufficiently resemble a Cygni to justify Vogel’s view that it should be classed with that type of star, while, on the other hand, apart from the haziness of the lines, it does bear a very strong resemblance to 6 Arietis and other Sirian stars, and should therefore be classed with them. [Note, February 8.—In a later publication* Vogel places a Cygni in his Class Ia 2, with Sirius, 6 Arietis, &c., but this does not materially modify the conclusions arrived at. | There are other points on which this demonstration of the similarity of « Aquile and 6 Arietis may be brought to bear, among them being the determination of the lines most suitable for the measurement of the velocity of the star in the line of sight, and the approximate deter- mination of the velocity of rotation necessary to produce the observed haziness of the lines. Lines suitable for the Determination of the Velocity of the Star in the Line of Sight. ; For the measurement of the velecity of « Aquilz, Deslandres has employed comparison spectra of hydrogen, iron, and calecium.t Vogel, however, questions the advantage of using the spectra of iron and calcium as comparisons for this purpose, on the ground that “the lines in the spectrum of a Aquile are so diffuse . . . . that between He and Hs; no lines except those of hydrogen and the magnesium line at d 4481 can be identified with known lines.” He himself has used the Hy, line alone as a term of comparison, and concludes that there are no indications of a periodic change in the velocity of the star in the line of sight, such as was supposed by Deslandres. We now know with certainty the origins of a considerable number of the lines in a Aquilz, so that measurements of the velocity of the star are placed on a surer basis. Since the spectrum of a Aquile is simpler than that of the sun, some of the broad lines do not represent confused groups of lines, but are broadened individual lines. The latter class of lines, when of known origin, seems to be well adapted for the * “Pub. Ast. Obs. Potsdam,’ 1899, vol. 12, Part I, p..49. + ‘Comptes Rendus,’ vol. 121 (1895), p. 629. BSeteel 706. col 60 Flale «0. ~ Roi Le wl 66, Plate Pree, See. Rov. Lochver & Fowler. Ca ENHANCED LINES The Spectrum of « Aquile. 237 measurement of the velocity of the star, for though they may be still somewhat wide, they are much less so than the lines of hydrogen. Among these broadened individual lines are :— Tron (enhanced lines).—4178°95, 4233°25, 4385°55, 4549-64. (arc lines).—4045°90. 7 Deaaium (enhanced lines).— 4417-98, 4443-98. Strontiwm.—4215°66. The enhanced line of magnesium at 4481°3 is usually sharply defined _ in stellar spectra, but the fact that it is generally fluffy in the compari- son spark disqualifies it for accurate measurements. The Velocity of Rotation. Assuming that 6 Arietis represents the spectrum of a Aquile as it would appear if the axis were directed towards the earth, we can get a general idea of the velocity of rotation necessary to produce the observed broadening of the lines. For this purpose lines which occur in groups are obviously unsuitable, but we can utilise the lines to which attention has just been drawn. ‘Taking the enhanced line of iron at \ 4178-95, we find that‘its thickness is increased from about two to four tenth-metres, and this corresponds to a surface velocity of the star at the equator of about 45 miles per second, supposing that the axis is perpendicular to the line of sight. Similar measurements of the broad- ening of the magnesium line 4481-3 yield a velocity of about 40 miles per second. Since only a small portion of the surface of the star could exhibit the effects of the maximum velocity, it is probable that these values are too low, really representing the equatorial velocity of rotation resolved along the line of sight with reference to a point some- where between the limb and centre of the star. Dr. Vogel gives reasons in his paper for supposing the velocity of rotation to be possibly 27 kilometres (16°8 miles) per second, but this determination does not depend upon measurements of individual lines. General Conclusions. The investigation of the Kensington photographs of the spectrum of a Aquilz has thus led to the following conclusions :— (1) Apart from the general haziness of the lines, the spectrum pre- sents no unfamiliar features. (2) The spectrum is of the Sirian type, showing enhanced lines of various metals, and a smaller number of arc lines. (3) A rapid rotation of the star, as first suggested by Pickering, appears to be a simple and sufficient explanation of the peculiarities of the spectrum. tc 238 Prof. J. Zeleny. The Velocity of DESCRIPTION OF PLATE. A. Solar spectrum, purposely out of focus. B. a Aquile. C. 6 Arietis, purposely out of focus. D. 8 Arietis, in focus. The photographs of the spectrum of a Aquile which have been obtained at Kensington since 1890 were nearly all taken by Messrs. Fowler, Baxandall, Shackleton, and North. Mr. Baxandall has assisted in the determination of origins. The photographic plate has been prepared from the original negatives by Sapper Wilkie, R.E. “The Velocity of the Ions produced in Gases by Rontgen Rays.” By JoHN ZELENY, B.Sc., B.A., Assistant Professor of Physics, University of Minnesota. Communicated by Professor J. J. THOMSON, F.R.S. Received February 15,—Read March 1, 1900. (Abstract.) The sum of the velocities with which the positive and the negative ions that are produced in gases by the Rontgen rays move when in a unit electric field has already been determined by an indirect method by E. Rutherford.* In the experiments here described the velocity was determined in a number of gases for the positive and negative ions separately, by comparing the ionic velocity directly with that of a stream of gas. The stream of gas was made to flow between two concentric cylinders, which were maintained at different potentials. By passing a narrow beam of Rontgen rays through the cylinders at right angles to their length, a narrow layer of ionised gas was pro- duced. Due to the electric field between the two cylinders, the ions of this layer tended to move radially towards, or away from, the axis of the cylinders, but at the same time they were carried along by the stream of gas. Of the ions of this layer which travelled inwards, those that started from the inner surface of the outer cylinder were carried a distance X by the gas stream before they reached the surtace of the inner cylinder. This distance is dependent directly upon the mean velocity of the gas stream, and inversely upon the difference of potential between the two cylinders. For obtaining the difference of potential which must be used to allow the ions to be carried a certain distance along the tubes by the gas stream the inner cylinder was divided at some distance from the beam of rays into two parts, insulated from each * H. Rutherford, ‘ Phil. Mag.,’ November, 1897. the Ions produced in Gases by Rontgen Rays. 239 other. That one of these parts which was not traversed by the rays was connected to a pair of quadrants of an electrometer, so that it was possible to tell when any ions reached it. A series of readings was taken for the charge reaching the electrometer in a given time for different values of the potential of the outer cylinder. From this was determined the value of this potential for which the ions starting from the outer edge of the ionised layer were just able to reach the juncture in the inner cylinder. The ionic velocity in a unit electric field is given by the equation— eG? — a") CL — log : 2AX at where U is the mean velocity of the gas stream between the cylinders, 6 is the inner radius of the outer cylinder, a is the outer radius of the inner cylinder, A is the potential of the outer cylinder, corresponding to X the distance defined above. To avoid the presence of vortices in the gas at the place where it was exposed to the rays, a sufficiently small velocity was used, and the gas was previously passed through a long portion of the cylinder to allow the motion to assume a steady state. The disturbing influence upon the electric field between the cylinders of the free charges formed in the gas during the conduction was diminished by using weak rays. The fall of potential at the electrodes* was also reduced by this means. For diminishing the amount of ionisation due to the secondary radiation produced at the metal sur- face,t the cylinders were made of aluminium, for which metal the effect is the least. The spreading of the ions due to diffusion produces an error, the amount of which increases with the time required for the ions to travel between the two cylinders. The value of this time is found from the equation T = X/U, where X and U have the same significance as above. The experimental values obtained for the velocity decreased as T increased, and from a series of results with different values of T the velocity could be obtained corresponding to T = O. Since in that case the effects of diffusion and similar causes disappear, this result was taken as the desired value of the ionic velocity. For testing the accuracy of the method, in addition to using different values of U and X, changes were also made in the intensity of the rays, in the diameter of the internal cylinder, and in the metal which formed the inner surface of the outer cylinder. Determinations were made with the gases when dry and when satu. * J. Zeleny, ‘Camb. Phil. Soc. Proc.,’ vol. 10, Part I, p. 17. + J. Perrin. ‘Comptes Rendus,’ vol. 124, p. 455. 240 Velocity of the Ions produced in Gases by Rontgen Rays. rated with aqueous vapour, as the results were found to be different in the two cases. This is in agreement with the effect of moisture upon the coefficients of diffusion of the ions, as observed by J. S. Townsend.* A summary of the results obtained is given in the follow- ing table. The results are reduced to a pressure of 76 cm. of mercury, but are not corrected for temperature, the effect of which is not known. Ionic Velocities. Velocitvn . | Velocity in centi- | elocity in centi- d tres per secoud’| 1°’? Ret eee . | aa f 4 £ 1 volt in a field of Ratio of | Feet br ie im Gut oe 1EH.S.U. per | negative \Tempera- AG p : centimetre. to ture. | i esis Positive.|Negative.| Positive. Negative. Min dirys cial? sutea «bes 1°36 .| 1°87 408 561 | 1-375 | 13°5°C PAE SINOIR aia acalnitaielie liu eel 1°51 411 | 453 1°100 | 14 Oxygen, dry'..5... .. 1°36 1-80 408 _ 540 1°320 | 17 Oxygen, moist ...... 1°29 1°52 387 456 | 1:180 | 16 | Carbonic acid, dry ..| 0°76 0°81 228 243 1-070 | 17°5 Carbonic acid, moist | 0°82 0°75 246 | 225 0-915 |. 17 | Hydrogen, dry......; 6°70 | 7°95 | 2010 | 2385 | 1-190 | 20 | Hydrogen, moist....| 5°30 5 60 1590 | 1680 | 1-050 | 20 | It is believed that in no case is the error greater than 5 per cent. while most of the observations indicate a considerably greater accu- racy. It is observed that the presence of moisture always diminishes the velocity of the negative ions, and that in carbonic acid the velocity of the positive ions is at the same time markedly increased. The velocity of the negative ions is the greater in all of the cases except for moist carbonic acid. The ratios of the velocities of the ions pre- viously determined for these gases by the writert were between those given above for the dry and for the moist. gases, as the influence of moisture was unknown at that time, and the gases had not been dried. K. Rutherfordt does not state whether he used dry gases in deter- mining the sum of the velocities of the two ions produced by Réntgen rays; but his result for air (3:2 em. per second) agrees with the sum of the values separately obtained above for the two ions in dry air, while his values for oxygen (2°8 cm. per second) and hydrogen (10:4 cm. per second) correspond to those for the moist gases. His value for carbonic acid (2°15 cm. per second) is higher than those * J.S. Townsend, ‘ Phil. Trans.,’ A, vol. 193, 1899. t J. Zeieny, ‘ Phil. Mag.,’ July, 1898. { E. Rutherford, ‘Phil. Mag.,’ November, 1897. Mathematical Contributions to the Theory of Evolution. 241 here obtained. The value obtained by EH. Rutherford* for the velocity of the negative ions produced in dry carbonic acid (0°78 em. per second) by the action of ultra-violet light, is quite near to that here obtained (0°81 em. per second) for the ions produced by Réntgen rays, but his values for dry air (1:4 cm. per second) and dry hydrogen (3°9 cm. per second) are considerably smaller. In discharge from points, A. P. Chattockt has obtained for the velocities of the positive and negative ions in dry air 413 and 540 cm. per second respectively for a field of 1 H.S.U. per em., which values are quite close to those obtained here for the ions produced by Réntgen rays. J. S. Townsend{ has shown that from the coefficients of diffusion of the ions and from their velocities it is possible to compare the charges carried by the different ions, and also to compare them with those carried by the ions in the electrolysis of liquids. By using the velocities given above with the coefficients of diffusion determined by J. 8S. Townsend, the values of Ne are obtained, N being the number of molecules in 1 c.c. of the gas and e the charge carried by each ion. The results thus obtained for the moist gases, air, oxygen, and hydrogen, perhaps justify the statement that the charges carried by the positive and the negative ions are equal, and that the charge is the same for the different gases, and is equal to the charge carried by the hydrogen ion in the electrolysis of liquids. The values of Ne obtained for the positive ions in these gases when dry are considerably larger than the above, while in carbonio acid ail of the results are over 20 per cent. smaller. “ Mathematical Contributions to the Theory of Evolution. VIII. On the Correlation of Characters not Quantitatively Mea- surable.” By Karu Parson, F.R.S. Received February 7, —Read March 1, 1900. (From the Department of Applied Mathematics, University College, London.) | (Abstract.) 1. In August last I presented to the Society a memoir on the inheri- tance of coat-colour in thoroughbred horses, and of eye-colour in man. This memoir, which was read in November of last year, presented the novel feature of determining correlation between characters which were not capable & priori of being quantitatively measured. The theoretical * EH. Rutherford, ‘Camb. Phil. Soc. Proc.,’ vol. 9, Part VIII. + A. P. Chattock, ‘ Phil. Mag.,’ November, 1899. t J. 8. Townsend, ‘ Phil. Trans.,’ A, vol. 198, 1899. 242 Prof. Karl Pearson. part of that memoir was somewhat brief, but I showed by illustrations that the method could be extended to deal with problems like the effectiveness of vaccination and of the antitoxin treatment in diphtheria. More recently, in studying the phenomena of reversion in Basset Hounds, Mr. Bramley-Moore indicated to me how my method, although correct in theory, differed sensibly in the numerical results with the pro- cesses of interpolating employed. I then proposed a new method, and the analytical discussion of its details was worked out in part by Mr. Bramley-Moore himself, by Mr. L. N. G. Filon, M.A., and by myself. Dr. Alice Lee also came to our assistance, and the result is the present joint paper. On the basis of the new methods, we have already worked out upwards of sixty coefficients of correlation, principally of heredity. Thus the thirty-six coefficients of heredity for coat-colour in horses and eye-colour in man have been re-calculated, as well as twelve coefficients for heredity in coat-colour of Basset Hounds given in a paper on the Law of Reversion presented on December 28th, and about to appear in the ‘ Proceedings.’ The great growth of the theoretical investigations has, however, compelled me to break up the old memoir* of last August into two parts, the one (the present) dealing only with theory, and the other with its application to inheritance in the horse and man. 2. The theory of the present memoir depends upon a very simple feature of normal correlation. If z6z,dx%2 . . . . 5% be the frequency of a complex of characters lying between 2 and a + 6x, 2 and dg + 0X2... ») Im and % + S%n, where x» is the deviation of the pth character from its mean, then dz dz U'pq — ik ditg. where 7p, is the correlation of the pth and qth organs. This simple differential relation enables us to expand z for any number of characters in powers of the correlation coefficients (neces- sarily less than unity) by Maclaurin’s theorem. But since we may replace a differential with regard to a coefficient of correlation by a double differential with regard to the corresponding organs, the coefi- cients of correlation may be put zero before instead of after the differentiation. In other words, we obtain a symbolic operator which, applied to a normal surface of frequency for n-uncorrelated organs, converts it into a correlated surface of frequency with $n(n-1) coefficients of correlation of arbitrary values. This operator gives us by aid of certain symbolic equations the expansion of the n-fold integral | | | cane | 2 da dig dig ....0- AXn Hie By he eae * That memoir was at my own request returned for revision after being accepted for the ‘ Philosophical Transactions.’ Mathematical Contributions to the Theory of Evolution. 2438 in terms of the 4n(n—1) coefficients of correlation, and a series of new functions which we term the 7-functions. These satisfy the difference equation : Un = LWp—1— (N—-1)vn—2 and the differential equation Wn = NUy»—1- dz In fact n(n —1 (n—1)(n-—2)(n-8 Uy = x- a gn? + “e ae M\ ) La Mth Ed oo. 5: n— Drs = ( — py adn oe) gre OF oe oe es The calculation of these functions is shown to be easy, and their properties are investigated. In this manner the volume of a frequency surface of the nth order cut off by m planes parallel to the co-ordi- nate planes is shown to be capable of calculation, and its value is determined in the numerical illustrations given for example of 1, 2, 3 up to 6-fold correlation. It may be noted that by putting 2 = 2 = 2 4= ee = 2n, we have really obtained a result which enables us to find the “area” of a “spherical triangle ” in n-fold hyperspace in terms of a series ascending by powers and products of the cosines of the angles between its faces. The application of these results to the correlation of characters not quantitatively measurable, arises from the fact that the n-fold integral above given, and which we have shown how to evalute, measures the total frequency beyond certain boundaries. We can observe, for example, whether horses’ coats are bay or darker (or chestnut or lighter), whether eyes are grey or lighter (or, dark grey or darker). Thus by forming mass frequencies instead of frequency distributions for small changes of character, we can find equations to determine the correla- tion. The probable error of such correlation, the convergency of the series, and other points are investigated. 3. Some discusssion is given to the problem of association, and coefficients allied to Mr. Yule’s coefficient of association but somewhat closer in value to the coefficient of correlation are considered, and their relative closeness measured. 4. A number of illustrations of the new method are given from heredity in horses, dogs, and man, and it is shown how normality of frequency must even for such a character as stature* only be looked upon as a first approximation. An investigation is also made into the influence of superior stock * Cited by so many as an example of “ normality.” 24-4 Prof. W. A. Tilden. On the Specific Heat of Metals in producing superior offspring. It is shown, for example, that if an individual who possesses a degree of character only found in one in twenty be considered “exceptional,” then eighteen times as many exceptional men will be born of non-exceptional parents as of excep- tional parents; but on the other hand, exceptional parents produce exceptional offspring at a rate ten times as great as non-exceptional parents, the greater gross product of the latter being due to their much greater numbers. In other words, distinguished parents are more likely to have distinguished offspring than undistinguished—ten times as likely—and yet only one distinguished man in nineteen will be born of distinguished parents. ‘The importance of such conceptions for both natural and artificial breeding can hardly be over-estimated. March 8, 1900. The LORD LISTER, F.R.CS., D.C.L., President, in the Chair. — A List of the Presents received was laid on the table, and thanks ordered for them. The Bakerian Lecture, “The Specific Heat of Metals and the Relation of Specific Heat to Atomic Weight,” was delivered by Pro- fessor W. A. TILDEN, F.R.S. BaKERIAN LecTuRE.—“On the Specific Heat of Metals and the Relation of Specific Heat to Atomic Weight.” By W. A. TILDEN, D.Sc., F.R.S., Professor of Chemistry in the Royal College of Science, London. With an Appendix by Professor JOHN Perry, F.R.S. Received February 9—Read March 8, 1900. (Abstract.) The experiments described in this paper were begun with the object of assisting in the determination of the relative values of the atomic weights of cobalt and nickel, but were continued with the further purpose of testing the validity of the law of Dulong and Petit. The metals cobalt and nickel closely resemble each other in density, melting point, and other physical properties, as well as in atomic weight. The pure metals were prepared for the purposes of these experiments with the most scrupulous care, the cobalt by taking and the Relation of Specific Heat to Atomic Weight. 245 advantage of the slight solubility of purpureo-cobaltamine hydrochloride in strongly acid solutions, and the nickel by deposition from the car- bonyl compound and subsequent solution of the metal and electro deposition. Both were fused by means of an oxyhydrogen flame, and afterwards shaped into bars. For the estimations of specific heat between 15° C. and 100° C. the differential steam calorimeter oi Professor Joly was employed. The mean specific heat of cobalt within these limits of temperature was found to be 010303. The mean specific heat of nickel was found to be 0710842. In order further to test the method and the conclusion from the case of cobalt and nickel, gold was compared with platinum, and copper with iron. The following are the mean specific heats for these pure metals after fusion, and within the same limits of temperature :— Cranes 20. NES 0:03035 Blatt 6! 0:03147 Copper ..2..)4. 0:09232 Rome 282 6 BA 0:10983 When these values for the specific heats are multiplied by the respective atomic weights of the several metals, the products are not constant, as the law of Dulong and Petit would seem to require if applicable at all temperatures. The influence of impurities on the specific heat of several metals was then investigated, and a number of results are given, from which it appears that small quantities of carbon or other non-metallic element tend to increase the specific heat appreciably, while the presence of a small quantity of a foreign metal seems to produce little effect. A series of calorimetric experiments were next made by the method of mixtures on the two pure metals cobalt and nickel, at the temperature of solid carbon dioxide, — 78°4°, and at that of boiling oxygen, — 182-5". The results, which are given below, show that as the temperature is reduced the value for nickel declines more rapidly than that for cobalt, and hence that, when the mean results are plotted out, the curves steadily approach each other. The mean specific heats of cobalt and nickel now stand as follows :— Temperature. Cobalt. Nickel. Promo toil 8... aus 0°10303 0:10842 ey LOK. spect LO’ lease seu OL O9 0:0975 be vel Vortec LBQ:4r dan: 070822 0:0838 and by calculation from the last two results from — 78-4 to — 182-4’, 00712, 0-0719. oo On the Specific Heat of Metals, &e. It seems probable that at absolute zero the values of the products of specific heat, multiplied by atomic weight, would be identical, or differ only by the very small amount due to experimental error.* Appendix. The following calculations have been made by E. R. Verity, Assoc. R.C.S8., and H. L. Mann, Assoc. R.C.S. :— If cobalt and nickel were in the states of perfect gases, their specific heats at constant volume would be ky = 0°04123 for cobalt. 0°04145 for nickel. Ig This is on the assumption that ko multiplied by the atomic weight is the same as for hydrogen. It is not unreasonable to assume that in no state can the substance have a smaller specific heat than k,. It is easy to show that for any of the metals the specific heat at atmospheric pressure K is not more than 2 per cent. different from the specific heat under any other condition, such as great hydrostatic pressure, to keep its volume constant. Our ignorance of the molecular state of a solid is so great, that we cannot even speculate on how it is that when 1 gramme of cobalt (we have much the same figures for other metals) at 50° C. rises in temperature to 51° C., whether it is allowed to expand freely or is subjected to great hydrostatic pressure which prevents expansion, the energy 0-041 enters it as what may be called the real sensible heat, and the energy 0-062 enters it as some kind of energy of-: disgregation, necessary because of change of tempera- ture, and having nothing to do with change of volume or pressure. The facts are not explainable by assuming that the atomic weights are wrong, because, as we see from cobalt and nickel, K approaches the value iy at low temperatures. Indeed, if the following formula is correctly deduced from the above measurements, we may say that in the solid state the product of the atomic weight and specific heat may be anything between what it is for hydrogen and 2-7 times this amount. The formula is K = i, ae : : 1+ where ¢ is the absolute temperature. * Note added March 3.—Further experiments made, since the date of communi- tion, upon the metals silver, copper, iron, and aluminium show, however, that this suggestion will not be realised. The mean specific heat of silver between 15° and 182'4°, for example, is 0°0519, while from 100° to 15° it is 00558. The decrease of specific heat at the lower temperature is therefore much less than in the case of cobalt and nickel. Total Eclipse of the Sun, January 22, 1898. 247 Nickel. | 2°35 x 10-7 | | Cobalt. | | | | Ky. | 0°0412 | 0°0415 | b | 1-764 x 10-§ | 1°764 x 10-8 | e | 2°55 x 10-7 | This formula must give fairly correct values of K from — 180° C. to 100° C. | March 15, 1900. The LORD LISTER, F.R.C.S., D.C.L., 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 :— I. “Total Eclipse of the Sun, January 22, 1898. Observations at Viziadrug.” By Sir Norman Lockyer, K.C.B., F.RS., Captain CHISHOLM-BaTTEN, R.N., and Professor PEDLER, F.RS. II. “A Comparative Crystallographical Study of the Double Sele- nates of the Series R2M(SeO.)2,6H2O. Part I.—Salts in which M is Zinc.” By A. E. Turron, F.RS. . Ii. “The Theory of the Double Gamma Function.” -By E. W. BARNES. Communicated by Professor Forsytu, F.R.S. “Total Eclipse of the Sun, January 22, 1898. Observations at Viziadrug.” By Sir Norman Lockyer, K.O0.B., FERS. Captain CHISHOLM-BATTEN, R.N., and Professor A. PEDLER, F.R.S. Received January 25,—Read February 22, 1900. (Abstract.) The paper is in three parts. The first, by Sir Norman Lockyer, gives an account of the arrangements made for the observations and the general conditions of the eclipse ; but the results obtained with the 248 Mr. A: E. Tutton. Crystallographical Study of the prismatic cameras are not included, for the reason that the reduction of the photographs is not yet completed. The second part, by Captain Chisholm-Batten, R.N., gives a full account of the observations made by the officers and men of H.MS. ‘“‘Melpomene ”; four photographs of the corona, taken with different exposures, accompany the paper. In the third part, Professor Pedler gives details of the spectroscopic observations which he made by eye with a 6-inch short focus lens and a grating spectroscope. A general idea of the work at Viziadrug has already been given in a preliminary report.* “ A Comparative Crystallographical Study of the Double Selenates of the Series R,M(SeO,),,6H,0.—Part I. Salts in which M is Zinc.” By A. E. Turron, B.Se., F.R.S. Received March 5, —Read March 15, 1900. — (Abstract.) In this communication are presented the results of the investigation of the group of salts of the above series in which M is represented by zinc, R being represented by potassium, rubidium, and cesium. The investigation is similar to that which has previously been-carried out for the double sulphates of the analogous series.t |The work consists of very large numbers of measurements of the exterior angles of the crystals, determinations of density, refractive index, optic axial angle, orientation of optical ellipsoid, and effect of change of temperature on the optical properties, together with the calculation of all the morpho- logical and physical constants derivable from the measurements. The main results are as follows :-— The morphological axial angle of the rubidium zinc salt is apprexi- mately the mean of the axial angles of potassium zinc and cesium zine selenate. , In the cases of thirty-three out of thirty-six angles between the exterior faces, the value for the rubidium zinc salt is intermediate between the values for the other two salts, and the exceptions are only apparent, being due to changes of opposite sign in adjacent angles following the rule. The morphological axial ratios for rubidium zinc selenate are inter- mediate between the ratios of potassium zinc and cesium zine selenate. The common habit of the crystals of the rubidium zine salt is of an * ©Roy. Soc. Proc.,’ vol. 64, p. 27. + ‘Journ. Chem. Soc., Trans.,’ 1893, 337, and 1896, 344. Double Selenates of the Series R5M(SeO,),,6H,0. 249 intermediate character to the habits of the crystals of the potassium zine and cesium zinc salts. An increase of density accompanies a rise in the atomic weight of the alkali metal, and it is greater for the replacement of potassium by rubidium than for that of the latter by cesium in the proportion of 5:4. The molecular volumes show a similar progression in the order of the atomic weights of the alkali metals, but the replacement of rubi- dium by cesium is marked by the greater change. The distance ratios (topic axes) indicate an extension of the distance separating the structural units in all three axial directions, the maxi- mum being along the symmetry axis. All these rules relating to the exterior merraala ae of the crystals are precisely analogous to those previously shown to apply to the double sulphates. The optical ellipsoid (indicatrix) is found to rotate about the sym- metry axis, when the atomic weight of the alkali metal is raised. The amount of rotation is twice as great when cesium replaces rubidium as when the latter replaces potassium. The mean refractive index (mean of all three indices) of the rubi- dium zinc salt is intermediate between the mean indices of the other two salts. The greatest change accompanies the replacement of rubidium by cesium. The double refraction diminishes at an increasing rate as the atomic weight of the alkali metal increases. It follows from the above that the axial ratios of the optical indi- catrix for the rubidium zine salt are intermediate between those of the other two salts, the second replacement being accompanied by the greater change. The optic axial angle of the rubidium zine salt is almost exactly the mean of the widely different optic axial angles of the potassium zinc and cesium zinc salts. The optic axes have a common plane, and the bisectrices are similarly situated, subject to the rotation of the whole ellipsoid already specified. The optic axial angles show a progressive change on heating the section plates—namely, a slight increase in the case of the potassium zine salt, a slight decrease in the case of the cesium zinc salt, and an almost complete indifference to change on the part of the rubidium salt. The whole of the specific and molecular optical constants of rubidium zinc selenate are intermediate between those of the potassium and cesium zine salts. The molecular refraction and dispersion increase at an accelerating rate with the rise of atomic weight of the alkali metal. 250 My. F. G. Jackson and Dr. V. Harley. On instituting a comparison between the results now communicated and those formerly published for the triplet of zine double sulphates, it is found that the replacement of sulphur by selenium is generally accompanied by a change in the morphological and physical constants similar to that which accompanies the replacement of one alkali metal by another of higher atomic weight. The changes due to the latter chemical change are often smaller in the selenate series than in the sulphate series, the greater weight of the initial molecule appearing to offer greater resistance to change. The intermediate character of the constants of the rubidium salt is, however, the invariable rule in both cases. “An Experimental Inquiry into Scurvy.”* By FREDERICK G. JACKSON and VAUGHAN Har.tey, M.D. Communicated by Lorp Lister, P.R.S. Received February 15,—Read March 1, 1900. (From the Department of Pathological Chemistry, University College, London.) The view that scurvy is caused by the want of fresh vegetables or lime juice, which has been the teaching of physicians and scientists in past years, would appear to require modification. In the early part of this century, through the efforts largely of Lind, the better feeding of sailors led to the gradual disappearance of scurvy in the naval service, and from this and other observed facts it was conceived that the disease developed whenever individuals did not receive a sufficient quantity of fresh vegetables, or some substitute, such as lime juice, in the diet. Garrod held that the cause of scurvy was a deficiency of potassium salts, while others believed the essential factor to be the absence of organic salts, which are present in fruits and vegetables. Ralfe believed the absence from the food of malates, citrates, and lactates reduced the alkalinity of the blood, and thus was the cause of scurvy. It was proved, however, by analysis that the alkalinity of the blood was not diminished, and the majority of evidence showed no diminution in the quantity of potash salts in the scorbutic blood, so that these explanations had to be abandoned. Neale, in an article on “Scurvy in the Arctic Regions,” published in the ‘ Practitioner,’ 1896, stated that ‘scurvy is a disease due to want of proper ventilation and want of proper blood nourishment ; in fact, scurvy begins with anemia, and its great antidote is fresh blood.” He consequently did not consider that fresh vegetables were of such * Towards the expenses of this research a grant was received from the Royal Lociety. An Experimental Inquiry into Scurvy. 201 prominent importance in the warding off of scurvy as the general teaching up to the present has led us to believe. These remarks of his were drawn from practical experience in the Arctic regions, as will be later mentioned. It is to the sojourners in the Arctic and Antarctic climes that scurvy nowadays is of such overwhelming importance, although in some lands nearer home it is still rife, and occasional cases even now occur in our marine services. Credit, as has been already stated, has been given to the use of lime juice in the Royal Navy and the Merchant Service for the great reduction of scurvy on board ship, but, as this research will presently show, this pomekistauis is probably without justi- fication. In the Nares Polar Expedition the crews of both the “ Alert” eae the “Discovery” suffered greatly from this affliction, although lime juice was taken daily by all hands when on board. When on the sledging expeditions, in consequence of the necessities of such a con- dition of travel, only a small quantity of lime juice was carried, yet an outbreak of scurvy occurred, not only amongst. the sledging parties, but also amongst the men that remained on board ship and continued to take the prescribed allowance of lime juice daily. On the other hand Mr. Leigh Smith’s party, with its medical ofiicer. Dr. Neale, after the loss of their ship the “ Eira,” spent nine months, including a winter, upon Franz Josef Land, in the severest, and, neces- sarily, the most unsanitary, conditions imaginable. They had no lime juice whatever; however, they almost entirely lived upon freshly lulled meat and frozen blood, and no case of scurvy occurred amongst them. On comparing these two examples we see that in the case of a body of picked men, well housed, well cared for, and with all possible means _ for procuring health adopted by those in command, taking the pre- scribed quantity of lime juice daily (except in the case of the sledge crews when absent from the ship for a few weeks) but who lived almost — entirely on preserved meat, we have universal scurvy ; while, on the other hand, we have a party who had not been selected on account of physical fitness, who were cast upon the desolate shores of Franz Josef Land, and with only the bare necessaries of existence, passed through nine months of their life there under conditions of considerable priva- tion and hardship, and in circumstances which would hardly meet with the approval of any sanitary inspector. They had no lime juice, but lived on freshly killed bear and walrus meat, and no symptoms of scurvy appeared amongst them. No less striking results were obtained by Dr. Nansen and Lieutenant Johansen. These two individuals, after they left the ‘“‘ Fram,” had to spend nine months, including the winter of 1895 and 1896, on Frederick Jackson Island. They were forced to live in a rudely constructed hut, VOU. LXVI. x 252 Mr. F. G. Jackson and Dr. V. Harley. without any changes of clothing, and no possibilities of washing, so that their sanitary conditions were not only of the roughest description, but the most unsuitable possible for health, and during the whole of this time they had no fresh vegetables whatsoever, and not even any lime juice. The only food they were able to use was fresh walrus or bear’s meat, which had been preserved simply by the cold. During this entire period they ate no salted or tinned meats, and we can pre- sume that the bear and walrus meat, as it would freeze almost imme- diately on being killed, would remain perfectly fresh. In consequence apparently of the purely fresh meat, and in spite of the most unsanitary conditions, Nansen and Johansen passed the whole winter almost constantly in the dark, without exercise, and yet showed no symptoms of scurvy whatsoever. In fact, the results with these two individuals show that Neale’s view, who had laid special stress on the want of proper ventilation as one of the causes of scurvy, was not realised, for the ventilation of their hut must have been extremely bad, as they describe the soot from their blubber lamps being deposited everywhere. Before commenting on this a few more examples in reference to this important subject may well be given. | Orie of us (F. G. J.), when living amongst the Samoyads on Waigatz Island, and the Bolshaia Zemclskija Tundra in 1893 and 1894, observed some striking facts as to the cause of scurvy. Amongst those of the Samoyads who invariably winter upon Waigatz, who never take vegetables nor know of lime juice, scurvy is unknown. They, however, live entirely upon fresh reindeer meat. On the other hand amongst those Samoyads, who in the autumn migrate south with the Russian peasant traders from the neighbour- hood of Yugor Straits, and live in common with them in the districts adjoining the large rivers in North-East Russia upon salted fish—the chief winter food there until the following May—scurvy is prevalent. That this fish is invariably tainted can be testified to from personal experience. In 1893, when at Kharborova, a shang settlement on the Yugor Straits, a remarkable case pointing to the cause of scurvy came under his notice. Six Russian priests, whose religion forbade them to eat reindeer or other such meats, but allowed salted fish, were left in a hut by Siberiakoff, the wealthy mine-owner, to pass the winter, a year or two prior to F. G. J.’s. visit. A small Russian peasant boy—whom he conversed with—was left to wait upon them. - The priests lived almost exclusively on tea, bread, and salted fish; the boy lived upon similar food, except that instead of the salted fish he ate fresh reindeer meat. None’ of them had any vegetables. In the following May, when the Samoyads and. peasant traders returned, they found that all the six priests had died of scurvy, whereas the little boy, who had lived upon An Experimental Inquiry into Scurvy. 253 fresh meat and had not eaten salted fish, was alive and well, and had buried all his late masters in the snow, he being the only living being in Kharborova in the spring. In the experiences collected by one of us—F. G. J., during his late expedition to Franz Josef Land in 1894, 1895, 1896, and 1897—we have two parties to consider: that of the crew of the ‘ Windward,” who spent the winter of 1894 and 1895 there, no individual of which ever failed to take his prescribed ounce of lime juice daily, and yet scurvy broke out, causing’ at least one death ; and on the other hand the land party on shore who took no lime juice, except two or three of them, who used it as a refreshing drink during the first few months, after which none was used. During the three years that they passed in Franz Josef Land none of them suffered from any symptoms of scurvy. The difference between these parties was principally, if not entirely, due to the meat. The “ Windward” party used largely tinned and salted meat; while, on the other hand, the land party principally lived on bear’s meat, and when tinned meat was employed, it passed a severe scrutiny in order that as far as possible it might be not even tainted. From these and other facts it would appear that neither lime juice nor fresh vegetables either prevent scurvy or cure it, and it is not the absence of this which is the catise of the disease, but that scurvy is a disease produced through the eating of tainted food. The view that scurvy is essentially due to poisoning by the ptomaines of tainted animal food was first propounded by Professor Torup, of Christiania, and it would appear from the foregoing evidence that such is the case, for in all the cases above mentioned where any scurvy occurred the men had lived on tinned meats or salted foods. Confirming this view, Dr. George M. Robertson, of the Perth District Asylum, relates a case of a woman who had become an inmate chiefly owing to her malady having taken the form of eating filth from pigs troughs. On arrival in the asylum she was found to a suffering . from spongy ulcerated gums—in fact, from “land scurvy,’ "and ulti- mately all the teeth, except the canines, fell out. . In the many instances of scurvy that we have investigated, in no single case have the circumstances rendered inadmissible, or even improbable, the theory that this disease is due to ptomaine poisoning. Before giving to lime juice the credit of having practically swept away scurvy from the naval and marine services, it is necessary to remember _that other causes have at the same time been at work to promote health, such as improved sanitation, better quarters for the men, shorter voyages through the enormous increase in the use of Ee, and above all better food. | | : The evidence so far shown, in which men have unwillingly experi- mented on the effects of ptomaine poisoning, proves that scurvy is x 2 254 Mr. F. G. Jackson and Dr. V. Harley. produced by the eating of tainted meat, and not by the want of fresh vegetables. In order to confirm or negative this view, it was decided to carry out some experiments in this country. After careful con- sideration, it was concluded that the most suitable animals for such experiments would be monkeys, since they are mostly nearly allied to man. Monkeys are not naturally carnivorous, and therefore it would be necessary to give the meat mixed with food that would not possess any alleged anti-scorbutic properties; and for this purpose it was decided to feed the monkeys on boiled rice and maize. In order to keep the standard of meat as nearly as possible aiways the same, a certain brand of tinned Australian beef was employed. Daily the rice was well boiled, and, after becoming thoroughly softened, 50 grammes of meat was added to each portion of rice for the various monkeys. It was then well stirred and gently heated ; by this means the meat got well mixed with the rice, and, although the monkeys might reject some of the larger lumps of meat, a con- siderable portion of it was eaten. At the same time any soluble ptomaines would be absorbed by the rice, and thus eaten by the monkeys. To this mixture daily was added a certain amount of maize. The results can be best described by dividing the experiments into three groups. First Group —The monkeys in this group were given daily, together with their boiled rice, 50 grammes of meat from a freshly opened tin, together with maize. Second Group—The monkeys in this group were given the same quantities of meat as in the previous group, but from tins which had been opened for a few days, and had stood in the laboratory. The meat in these was not what one would call bad, although it had a dis- tinctly sour smell. Rice and maize as before. Third Group—The monkeys in this series were given exactly the same diet as was employed in the second group, except that each monkey received daily either an apple or a banana. We have found in these three groups three conditions, | so far as diet is concerned, which ought to yield definite results in reference to the subject of ptomaine poisoning, In order that the general surroundings of the monkeys should be as nearly as possible the same, and that each should be properly observed, every monkey was kept in a separate cage of similar con- struction, so that the feces could be easily examined. The cages were kept in a room warmed by hot-water pipes, so that. they were under as nearly as possible similar conditions as regards light and heat. The excreta of the monkeys were examined daily, and the general appearance of the animals noted, more especially as regards the condition of the gums. Every few days they were weighed. We can now proceed to describe the results of the experiments. In An Experimental Inquiry into Scurvy. 255 erder to study the results of each group more easily they are put together in a tabular form. First Group —The monkeys fed on boiled rice with 50 grammes of fresh meat and maize daily. Six monkeys in this group (Table I) were kept under observation. Table I—Group 1. Monkeys fed on Boiled Rice, with 50 grammes of Fresh Meat and Maize daily. | | | | | | Weight in kilos. Blood | No Duration of | Diarrhea and | Gums _~ “*| observation. ) commenced.| mucus | spongy. Original. | Final. | Loss. | in stools. | | — - ss a! ~—|—___— ——-—— days. days. days. | days. 1 | 70 2-000 1°450 0-550 62 Othe atintg 2 31 2°250' | 1°700 | 0°550 23 0 0 3 28 2°200 | 1°850 | 0°350 3 0 | 8) 4 45 2°650 1°950 | 0-700 | Pep OO ) 73 ~~ 2°750~ | 1°650-|-t -100 13 0 | 0 9 | 39 1°250 | 1°050 | 0°200 5 0 | 0 | | | In the case of No. 5, which lost no less than 1-100 kilos. in seventy- three days, this loss of weight principally occurred during the last seven days, for up to that period this monkey had only lost 300 grammes. During the last seven days, however, severe diarrhcea set in, followed by rapid wasting. The same holds good as regards the other cases. So that although the monkeys on this diet lost weight, possibly from the food not being sufficiently nutritious, the principal loss of weight was apparently due to the diarrheea. When we come to consider the diarrhcea, we see that all six monkeys developed this condition sooner or later. As a rule it commenced by being very intermittent, and becoming more severe towards the end, when it speedily proved fatal by the general loss of strength, &c., which it occasioned. The diarrhoea in monkey No. 1 did not commence until it had been sixty-two days on the diet, when it proved fatal after seven days. In the case of monkey No. 5, on the other hand, it began after thirteen days, but was only slight, and at intervals of three or four days, after which it became very severe ; during the last seven days there was rapid wasting, and death occurred on the seventy-third day. In the case of monkey No. 3, the diarrhcea commenced, however, on the third day, when it was very severe, and afterwards, although the diarrhoea was not so acute and occurred only at intervals, the animal became very feeble, and on the twenty-eighth day was killed by chloroform. At the autopsy no cause for the feeble condition could be 256 Mr. F. G. Jackson and Dr. V. Harley. found, although the large intestine was somewhat congested in this case. In spite of the appearance of this diarrhoea, none of the monkeys in this group showed any signs of either blood or mucus in the motions, the liquid stools being merely of a pale yellowish colour. And in all these cases the gums, although frequently examined, showed no sponginess nor signs of bleeding. ‘The monkeys of this group, as they became emaciated, sat hunched up in their cages, the most usual attitude being with their heads between their knees, as if they were trying to keep themselves warm, although the room was, as already stated, heated by hot-water pipes. They also showed signs of being out of condition by the general roughened condition of their coats. At the autopsy all these monkeys exhibited more or less marked emaciation, but with the exception of No. 2, which died from pneumonia, in no case was any direct cause of death discoverable. In the bowels were found liquid, light-coloured contents, and only in the case of No. 3 were there any signs of congestion to be noted in the large intestine. Table II.—Group 2. Monkeys fed on Boiled Rice, with 50 grammes of Tainted Meat and Maize daily. Weight in kilos. Blood | Gums N ‘Duration of Diarrhea .| and spongy ‘| observation. commenced.| mucus and | Original, | Final. | Loss. in stools.| bleeding. days. days. days. days. 6 18 2°000 | 1°400 | 0°600 5 | 0 0 a 14 1°350 | 1°050 | 0°300 5 8 13 8 55 1°500 | 0°950 | 0°550 | 17 23 27 10 65 1:600 | 1°050 | 0°500 7 28 28 11 54: 1°650 | 1°250 | 0°400 ay 26 27 12 11 2°425 | 0°150 | 0°275 6 9 0 13 80 2:050 | 1-900 | 0°180 | 0 ) 0 20 62 2°450 | 1°400 | 1:050 | 40) 40 40 In the above table (II) the results of eight observations under these conditions are recorded. The monkeys of this group lived from eleven to eighty days, although in the cases of Nos. 8, 10, and 20 we cannot call this the limit of their life, as they were killed in order to examine the influence scurvy, thus artificially pro- duced, would have on their blood. We see in this Table II that we have the same loss of weight in these monkeys as we had in the six monkeys fed on the fresh meat. In this group the diarrhcea commenced earlier than in those pre- An Experimental Ingwry into Scurvy. 257 viously described. In only one case was it delayed forty days; in the other cases it commenced between the fourth and seventh day, except in monkey No. 8, in which it did not commence until the seventeenth day. In these, as in the previous monkeys, diarrhea, although always occurring, was somewhat intermittent. Out of those eight monkeys, in no less than six was it seen that the motions were not of the simple diarrhceaic character of the former group, but contained blood and mucus. In monkeys Nos. 7 and 12 the blood and mucus appeared on the eighth and ninth day respectively, while in the other cases it was more delayed, the diarrhea having continued for some time previous to its appearance. In some of the cases the motions just before death consisted principally of blood and mucus. When we turn to the appearance of the gums, we find that in five out of the eight monkeys included in this group they showed spongi- ness, and in some cases even small ulcers forming. The sponginess of the gums was most marked around the incisors and bicuspids, and, as a rule, did not occur around the molars at all. ‘The monkeys belonging to this group sat in the same cramped posi- tion, with roughened coats, as already described in the previous group. They showed a more marked disinclination to move, or to take interest jn objects around them; but in no cases did they show any signs of definite tenderness of their limbs when handled. Only in one case (No. 7) was there any indication of bruising. In this monkey a few days betore death two bruises developed on his left knee, about $ cm. in diameter, of a dirty red-brown colour, and also sores showed on the sole of the right foot and at the root of the tail. In all these cases, as in previous monkeys, an autopsy was carried out. In no single instance was there found any signs of hemorrhage or hemorrhages into the pleura, pericardium, or peritoneum. The gums in Nos. 8 and 10 were not only spongy, but had a tendency to the formation of ulcers at the root of the incisors. The stomach and small intestine showed little or no change, while, on the other hand, the large intestine was, in the majority of cases (except Nos. 6 and 13), markedly congested, the congestion being, as a rule, most noticeable at the sigmoid flexure and cecum. The contents of the small intestine were light yellow, while the large intestine con- tained more or less bloody mucus. The only thing else abnormal to be noticed was that in No. 7 the liver was enlarged and fatty, showing markedly the line of the ribs. | | We now come to consider the Third Group (Table III). In this the monkeys were fed on boiled rice with 50 grammes of tainted meat and maize daily, but each monkey received, in addition, an apple or a banana. : | 258 My. F. G, Jackson and Dr. V. Harley. Table I1I.—Group 3. Monkeys fed on Boiled Rice, with 50 grammes of Tainted Meat and Maize daily. Each Monkey received a Banana or Apple, these being given on alternate days. Gar" : | | | spiobal4|* Hanis | Weight in kilos. No Duration of Diarrhea and | spongy | observation. | commenced.| mucus | and | Original. | Final. | Loss. in ets bleeding. days. | days days days 14: 22 1°600 | 1°350 | 0°250 16 18 0 15 180 2°200 | 1°300 | 0°900 150 0 0 15 13 1°750 | 1 °500 | 0°250 8 11 11 17 31 2°476 | 1°500 | 0°975 9 20 present* | 21 | 123 2°000 | 1-650 | 0350 90 90 0 | In this group the animals, therefore, in addition to the rice and maize and tainted meat, received fresh vegetables daily, and may be considered to have been well fed. Five monkeys only were used, and, in spite of the extra food, they all lost weight. In fact, Nos. 15 and 17 lost no less than 900 and 975 grammes of weight in 180 and 31 days. respectively. The others, however, did not lose so much weight. .In three of the monkeys, Nos. 14, 16, and 17, diarrhea com- menced from. eight to sixteen days after the monkeys had been on this diet. But in No. 15 diarrhcea did not commence until the 150th day of observation, and in No. 21 not until the 90th day. In four of the cases blood and mucus appeared in the motions in from eight to eighteen days. When it first occurred in No. 21 was unfortunately not noticed. The other monkey, No. 15, did not show any signs of it. In two of the monkeys, Nos. 16 and 17, spongy gums occurred; in the former, on the eleventh day, while in the latter it was only noted at the autopsy. The other three monkeys showed no signs of bleeding gums. The autopsy of those monkeys which showed scurvy exhibited, as in those of the second group, marked congestion of the large intestine, with bloody mucus in the contents. In all the cases there was marked emaciation but no hemorrhage, either into the pleura, pericardium, or peritoneum. After this general description of the results obtained in the three groups of experiments, and before discussing their significance, we can consider the changes produced in the blood of animals suffering from the results of scurvy—scurvy being defined by Bristow as “a peculiar form of anemia arising from a deficiency of vegetable diet, and attended with a tendency to the occurrence of hemorrhages, profound impairment of nutrition, and great mental and bodily prostration.” * Only noted at the autopsy. An Experimental Inquiry into Scurvy. 259 Blood in the Scurvy of Monkeys—In order to get monkeys with as well-developed symptoms of scurvy as possible, the animals were kept until not only were they passing bloody mucus by the bowels, but the gums were spongy and easily bled. Unfortunately, the blood of only two monkeys could be examined, as the others died too speedily. The monkeys Nos. 8 and 10 in the second group, however, both showed very well marked symptoms of scurvy, as found in monkeys—diarrheea, wasting, the motions containing blood and mucus, and the gums spongy and easily bleeding. Table [V.—Comparing Analysis of Blood of a Normal Monkey with that of two suffering from Scurvy. — — Normal. Scurvy. | | | Pemehe im kilos... 6.2 e en cec ce sece 2-000 0950 1-050 Number of corpuscles ............| 4,730,000 4,220,000 4,500,000 — re Be PEUCOBVIES Fe ccc c sc cc 8125 40,000 — PEEIROPIOUI Bovis Sete sce te eae 75 48 45 | OS er 1046 1035 1034: | Water ..... i eee 83 °37 85°18 84°59 ES GEESE ees a Co 14 *82 ES"oL e's | Proteids..... Mipmalale ye rah lo jamicre 5 18°27 12°37 15°69 | PRRREUMI oy ais wn es ners neces sees 0°52 — OF 7G | Time of coagulation..............| 3 minutes 2 minutes 1 minute | Discs cwseesees| 2°72 re a eee | Ash....... 0 Ee eee ee 0°75 0°79 | — — In the above Table IV the results of the analyses of the blood in the case of these two monkeys are compared with that of a normal monkey, so that we can more readily see the changes produced. The blood to be examined was collected from the carotid artery while the animal was kept under the influence of ether, and when sufficient blood had been collected for the various analyses, the animal was killed by an overdose of anesthetic before it returned to con- sciousness. The Number of Red Blood Corpuscles.—These were estimated by the Thoma-Zeiss’ Counter, and in each of these cases the average of two blood counts is given, the result being the number of red blood cor- puscles contained in a cubic millimetre of blood. It is seen that while in the normal monkeys the number of red blood corpuscles is 4,730,000 per cubic millimetre, in the two monkeys suffering from scurvy it is respectively 4,220,000 and 4,500,000. Other observers have drawn attention to the diminution in the number of red corpuscles in cases of scurvy as quoted by Philip Wales in Ashurst’s ‘ International Encyclopedia of Surgery.’ 260 Mr. F. G. Jackson and Dr. V. Harley. From these results one may conclude that there is a slight reduction in the number of red blood corpuscles in monkeys fed on tainted meat, although this reduction is nothing like the reduction one finds in the human subject in many of the more severe forms of anzmia. The Leucocytes—The white blood corpuscles were counted in the same way as the red blood corpuscles. This was carried out on one of the monkeys affected wtih scurvy. It is seen in the case of the normal monkey that there were 8125 leucocytes per cubic millimetre, while in the case of the monkey suffering from scurvy there were no less than 40,000. Laboulbéne notes the occurrence of an unusual number of white globules in scurvy. We certainly can conclude that there is a very marked leucocytosis produced by the diet of tainted meat. The Hemoglotin.—The quantity of colouring matter in the blood was estimated by Fleischl’s hemometer. In each case precautions were taken to use the same illumination power, and at least two calculations were made, the average being taken. It is seen in the above table (IV) that the hemoglobin present repre- sent 75, while in both the monkeys suffering from scurvy it is very considerably reduced, being 48 and 45 respectively. When we compare this with the small reduction in the number of red blood corpuscles, we see that in the monkeys fed on tainted meat there is produced a very marked hemoglobinemia, while at the same time there is probably oligocytheemia. In fact this condition corresponds with that in ie human being in cases of chlorosis, as against those forms of pernicious anemia or secondary anemia, where either the red blood corpuscles are reduced out of proportion to the hemoglobin, or the hemoglobin and corpuscles are reduced in equal proportions. The Specific Gravity of the Blood—The estimation of the specific gravity of the blood in these monkeys was carried out by means of the picnometer. In the normal monkey we have a specific gravity of 1046, while in the case of the two monkeys suffering from scurvy we have a specific gravity of 1035 and 1034; so that we ean conclude that there is a slight decrease in the specific gravity produced by the diet of panied meat. The Wa A given quantity of blood was collected in a penne: crucible ai acd to constant weight at 70° C. It is seen that in the normal blood there was 83°37 per cent. of water and 16°63 per cent. solids, while in the case of the monkeys suffermg from scurvy the percentage of water was 85°18, with only 14:12 per cent. of solids, in monkey No. 8, while 84:99 per cent. of water, with 15-01 per cent. of solids, occurred in No. 10. It would appear that all observers who have noted the blood in scurvy speak of the marked diminution in the total solids which occurs in this disease. So that we An Experimental Inquiry into Scurvy. 261 have with the decrease in the specific gravity of the blood a diminution in the total solids, as one naturally would expect. The Proteids of the Blood.—The total quantity of proteids in the blood was estimated by precipitating a given quantity of blood in 15 volumes of absolute alcohol. After allowing it to stand some days, with frequent stirring, the precipitate was collected on a weighed filter paper, and dried, &c., in the usual manner. By this method it was found that the normal blood contained 18:27 per cent. of proteids, while in the case of the two monkeys fed on tainted meat the quantity of proteids was only 12°37 and 15-69, so that there is a very marked decrease in the quantity of proteids in the blood. The Quantity of Fibrine.—This was estimated. by the method of Hoppe-Seyler in one case. It is seen that in the normal monkey there is 0°52 per cent. of fibrine in the blood, while in the case of the monkey suffering from scurvy the fibrine was no less than 0°76, so that we see there was a very marked increase in the quantity of fibrine. Chalvet, in his, analysis of a case of scurvy, comparing it with healthy blood in the normal individual, found the fibrine 0-216, while in the case of scurvy it was 0°434. Busk, in three well-marked cases of scurvy which occurred in the *“* Dreadnought ” hospital ship, found the fibrine in excess of the normal amount, the least being 0-45 and the greatest 0°65 per cent. The Time of Coagulation.—The time that it took for the blood to coagulate was estimated by Professor Wright’s tubes, and it is seen that in the normal monkeys this is three minutes, while in the case ot the monkeys suffering from scurvy it was found to be one and two minutes respectively. The increase in the quantity of fibrine (hyperinosis) with the shorten- ing of the time of coagulation is what one commonly finds in hydrzemia in the human subject, and may therefore in all probability be put down to the same cause. The Quantity of Nitrogen.—The total quantity of nitrogen in the blood was estimated by the method of Kjeldahl, and the average of two analyses is given as before. In the normal monkey it is seen that the total nitrogen was 2°72, while in the case of the monkey suffering from scurvy, 1n which it was analysed, it was 2°31 per cent. There is, therefore, a small decrease in the quantity of nitrogen, this decrease corresponding to the decrease in the quantity of proteids, and as this was only analysed in one monkey, it is as well perhaps not to discuss any theories as to.its significance. The Ash of the Blood—The estimation of the ash was ¢arried out in the ordinary way, but only in the case of one monkey suffering from the effects of tainted meat. It is seen that in the case of the normal monkey the ash was 0°75, while in the .case of the monkey fed with tainted meat it was 0°79 per cent. 262 Mr. F. G. Jackson and Dr. V. Harley. The average amount in Becquerel’s, Rodies’s, and Busk’s cases of scurvy in the human subject was 0°81, the smallest being 0°55 and the largest 1:15 per cent. Garrod, in the analysis of the blood in one case of scurvy, found a deficiency of the potassium salts, upon which he formed his well-known theory that scurvy was due to their want. We therefore see, in reference to this monkey, that the ash does not tend to be decreased with the low specific gravity and diminution of the total solids. It is apparent that the diminution in the total solids is principally due to a lessening in the quantity of proteids. Conclusions. The descriptions brought forward in the first part, of several cases of scurvy which occurred in the Arctic regions when the individuals were under the influence of preserved or salted meats, in spite of their taking at the same time either vegetables or lime juice, of the Nares Polar Expedition, in which scurvy occurred, as well as the very striking case of the six priests already mentioned, can,be compared on the other hand with conditions of the greatest hardship and privation in the Leigh Smith Expedition, Nansen and Johansen, and the Frederick Jackson land party, as well as in the instance given of the Samoyads who winter on Waigatz and who live on fresh meat ; in all of which cases, in spite of the entire absence of vegetables or even lime juice, no scurvy occurred. If we look at this evidence alone we could almost say we have con- clusive experimental evidence that the eating of salted or improperly preserved meat, or tainted meat in any form, can produce scurvy, even when lime juice or vegetables are being taken at the same time. We have also the support of the fact that bad ventilation, believed by Dr. Neale to be one of the causes of scurvy, was not the cause of scurvy with Dr. Nansen, living on fresh meat and blood—probably owing to the fact that he introduced no ptomaine, and therefore no scurvy occurred. We now come to consider the experiments on monkeys, and how much or otherwise these experiments confirm the results already given in man. It is necessary to consider what are the symptoms of scurvy. In the present paper it is impossible to go through all the symptoms of scurvy described by the various observers, since different epidemics have shown more markedly various symptoms. We, however, can compare the symptoms in our monkeys with those generally described as accompanying scurvy. The pallor and yellowish colour of the face, which is described as distinctive in scurvy, is of course impossible to be observed in monkeys, although in those monkeys which we con sider to suffer from scurvy there was generally a good deal of blue- An Experimental Inquiry into Scurvy. 263 ness about the lips and gums. There was a very marked disinclina- tion to bodily movement and a general tendency to mental prostration, for the monkeys took little interest in the things surrounding them, in those cases which showed what we might consider the more definite symptoms of scurvy, such as the bloody mucus and bleeding gums. At the same time in none of these monkeys did we find any definite tenderness of the limbs, no swelling of the legs, or any purpura. Only in one case do we get the formation of bruise-like sores in an animal which apparently was suffering from scurvy. In the monkeys included in the first group, which were fed on fresh meat as well as maize and rice, the only symptom we note beyond the wasting is diarrheea, and none of these monkeys showed anything like the muscular feebleness or general ill-health which was noted in the scorbutic monkeys. When we compare the second group, Table II, in which the monkeys received the same diet, except that the meat instead of being fresh was tainted, we find a very different state of affairs. These monkeys showed a very much greater prostration, and although it is difficult to judge by the eye they certainly seemed paler and generally out of con- dition. No less than six out of the eight monkeys thus fed passed blood and mucus in their motions. _ The question whether blood and mucus in the motions is to be regarded as one of of the symptoms of scurvy, can be easily answered by the fact that, first, Bristow states that “patients suffer from looseness of the bowels, the motions frequently being highly offensive and containing blood.” It is also stated by other observers, such as Hilton-Fagge and Osler, as well as in an able article on scurvy in Ashurst’s ‘International Encyclopedia of Surgery,’ by Philip Wales, that the occurrence of hemorrhages in the mucous membrane of the stomach and bowels is of frequent occurrence. None of our monkeys vomited, so that whether they suffered from hemorrhages in tne stomach cannot be noted. In the post mortem there was no evidence of any such thing. If this bloody diarrhoea is an evidence of scurvy, we find that no less than six out of the eight monkeys which were given the tainted meat showed this symptom. We now compare the third group, Table III, in which the monkeys were given fresh fruit, apples or bananas, every day. One can say that the monkeys in this case were well fed. Five monkeys were observed, and out of these five monkeys no less than four developed the symptoms of bloody mucus in their stools, so that in spite of good feeding this symptom of scurvy developed in four out of the five monkeys on full diet. We come to the next symptom, undoubtedly the most definite sign of scurvy that occurs in man—in fact, it is about the only condition 264 An Keperimental Inquiry into Scurvy. which is universal in scurvy, and is always found except in those cases in which the teeth are absent—that is, the spongy condition of the gums. When we consider the first group of monkeys which were merely fed on rice, maize, and fresh meat, out of the six monkeys not one showed any appearance of spongy gums, so that we can conclude that in these monkeys none of them showed any scurvy whatever. We now come to the eight monkeys fed on exactly the same diet, in which the meat was tainted instead of fresh, and here no less than five of the eight monkeys showed sponginess of the gums, and some, not only the sponginess but the gums even ulcerated, so that five out of the eight monkeys showed this sign, which is considered by all who describe scurvy as the most significant. We now come to those monkeys which were well fed, getting the fresh vegetables every day. Four of these monkeys showed bloody mucus in the stools, and two of them spongy and bleeding gums. We have further to consider the condition of the blood in the monkeys fed on tainted meat. Scurvy is considered by all authorities to be a peculiar form of anemia. In two monkeys we had the oppor- tunity of analysing the blood, and it is seen by the analyses that we have marked changes in the blood, a very great diminution in the quantity of hemoglobin, with a slight diminution in the number of red blood corpuscles—in fact, a condition corresponding to chlorosis, and that is accompanied by leucocytosis. The specific gravity of the blood is reduced, and this is due to the reduction in the quantity of proteids and not to a marked reduction in the quantity of salines in the blood. The fibrine of the blood is increased, together with an increase in the coagulability of the blood. When we consider what has been found in the blood of man suffering from scurvy, we find it is universally accepted that there is this condition of anemia with low specific gravity, the blood being distinctly watery and with a marked excess in the quantity of fibrine. So the condition of the blood of the two monkeys which we have analysed corresponds with that found in the majority of analyses of scorbutic blood. Considering therefore the occurrence, in the monkeys fed on tainted meat, of bloody mucus in the stools, spongy gums, and characteristic anemia, we assume that, although the symptom of hemorrhages into the tissues was not observed, we may fairly conclude that they were really scorbutic. This conclusion is further justified when we wont on looking over the description of scurvy given by authors, that all the symptoms except the spongy gums are very often absent in different epidemics. The fact that the five monkeys fed on tainted meat, in which fresh vegetables were given, showed in a smaller proportion the symptoms of scurvy than the monkeys in the second group, can be sufficiently The Theory of the Double Gamma Function. 265 explained by the fact that when the monkeys received a banana or an apple every day, they would be less likely to eat as much of the meat as they would otherwise do, and would thus daily receive a smaller dose of ptomaine. In spite of this fact, in no less than four cases out of the five did we - get bloody motions, and in two of the cases spongy gums. In these eases tainted meat alone seems to have produced scorbutic symptoms, although the animals in this group took longer to develop the symptoms, and seemed not to suffer in such a severe form. Looking at the results of our experiments on monkeys, as a whole, we venture to think that they afford important confirmation of the conclusion derived from Arctic experience, as referred to in the early part of this paper, that the absence or presence of fresh vegetables or lime juice is not alone sufficient for the prevention or the cure of seurvy, and that we must regard the condition of the food in general, and especially the state of preservation of the meat, as the essential factor in the etiology of the disease. We have to express our thanks to Dr. Francis Goodbody for his untiring assistance in the numerous observations that had to be made during the eighteen months the research was continued. “The Theory of the Double Gamma Function.”. By. HE. W. Barnes, B.A., Fellow of Trinity College, Cambridge. Com=- municated by Professor A. R. Forsytu, 8e.D., F.R.S.. Re- ceived February 26,—Read March 15, 1900. (Abstract.) The memoir deals with a function which is substantially one-quarter of the o function of Weierstrass, just as the ordinary (simple) gamma function is substantially one-half of the infinite sine product. The analogy between the two functions determines the nomenclature. In any development of the simple gamma function from the function- theory point of view, it is necessary to use Euler’s theorem Li [1+4+ ececees +2 = log | —_ U=on n, to obtain the product I’) (z/w) given by Dy? @/e) = a Géyaee I | 1+— ) m7 | els \ , as a solution of the difference equation /(7+ 0) = 2f(2). Similarly for the elementary theory of the double gamma function 266 My. E. W. Barnes. investigated in Part II of the memoir, two forms, y2:(o;, #2) and ¥22(@1, 2), are considered which arise substantially as finite terms in the approximations for T nv 1 nN nT 1 4 DD A) Sees anne Se oe tase eae ow! thy == 10 diy Sg where 2 = 71; +M2w2, when n is large. These approximations are shown to involve logarithms of ©, w2, and (w;+ 2); and the relative distribution of the points in the Argand diagram representing these quantities causes the introduction of two numbers m and m’ of fundamental importance in the theory. The double gamma function Tr, (z/o, 2) given by -] = yar Wy os) +2Yn (069) iy TI Il’ 1+) aa PG) eyo) =¢ re Co is shown to satisfy the two difference relations oad (2 + 1) ii T; (z/@2) e—2mri8,'(z/w2) ; Pooh) ie pi (2) T'.71 (4 + w) ee Ty (2/1) g- 2m! ni8y'(z/o) . Py? (@) pi (1) The functions y21(1, 2) and yo2(@;, w2) are called the first and second double gamma modular forms respectively. The double gamma function can .be expressed in two ways as an infinite product of simple gamma functions ; it can be connected with an unsymmetrical function G (z/r) first considered by Alexeiewsky ; and in terms of it, Weierstrass’ elliptic functions can be expressed. By means of the values of the numbers m and m’ the well-known relation | M2 — Nowy = + ¥7t can be obtained, as well as the fundamental formule of the o function. Fundamental in the theory of double gamma functions is the double Riemann ¢ function, ¢2(s, @/, #2), which is considered in Part III, and is the simplest solution of the difference equation | 1 JG + oy + 09) — f (G+ oon) — FG 42) Fa where 8, @, w, ,and w. have any complex values such that w/w; 1s not real and negative, and a~* has its principal value with respect to the axis of —(@,+ 2). This function is expressible as a contour-integral by means of the relation The Theory of the Double Gamma Function. 267 - US) re i iad (a) a (2 (8,4) = migen st oa la — eer) (1 —e-#8)’ and the determination of the axis of the contour and the number M depends upon the distribution of the points , »,, —1, and (w; + .). By means of this function we obtain asymptotic approximations for summations of the type pn gn on) te Viewed at Scare my, =0 my = 0 (4+ 0) + Mz@2)5 when ” is a very large number and s has any complex value. We can also obtain an asymptotic approximation for the product pn qn TIL IL (@+mo, + mw2). m, =0 m,=0 Since Stirling’s theorem gives the asymptotic evaluation of n!, we obtain, on putting a = 0, an extension of Stirling’s theorem to two parometers. We find, as the absolute term, the double Stirling func- tion p2(;,w2), which is the analogue of the simple Stirling form pi() = /27/v. All the double asymptotic expansions involve as their coefficients double Bernoullian functions and numbers. The double Bernoullian function 2S,(a/o;,2) is an algebraic polynomial which satisfies the two difference equations f (a+) - — f (a) =S, (gin y ep? ete’ S'n41 ad f (a+) —f(a) = Sp (a/or) + Sen(0), and possesses properties exactly analogous to the corresponding simple forms. The theory of this function forms Part I of the memoir. From the contour-integral expression for the double Riemann ¢ func- — tion it is possible to obtain similar expressions for the logarithm of the double gamma function and its derivatives, for the first and second double gamma modular forms, and for the logarithm of the double Stirling function. Under certain restrictions these contour-integrals ean be transformed into line-integrals. The double gamma function admits of transformation and multipli- cation theories developed in Part IV. By means of the latter theory we may express the double Stirling form as a product of double gamma functions of arguments 40;, $w2, and 4$(w;+:2) respectively. There is also a transformation theory for the double gamma modular forms and the double Stirling function. The extension of Raabe’s formula for the simple gamma function leads to certain “integral formule.” ‘The integral . VOL. LXVI. Y 268 Proceedings and Lnst of Papers read. 5 MLE | log Ty (2) dz 0 can be expressed in terms of double gamma functions of equal para- meters, and the two integrals | log Pe (2) dz, | | log T's (2) dz 0 0 can be substantially expressed in terms of the double Stirling function of w; and ay. It 1s shown in Harb X, that it is ie to obtain an asymptotic log Ty (e+), . which is valid over that part of the region at infinity in which there are no zeros of I',"1(z). The coefficients in this expansion are double Bernoullian functions of a. Finally it is proved that the double gamma function does not satisfy any differential equation whose coefficients are not substantially that function or its derivatives. There exist n-ple gamma functions whose properties can be obtained by an easy generalisation of the previous theory. March 22, 1900. The LORD LISTER, F.R.C.S., D.C.L., President, in the Chair. A List of the Presents received was laid on the table, and thanks ordered for them. - The Croonian Lecture—“ On Immunity, with Special Reference to Cell Life ”—was delivered by Professor PAUL EnR.IcH, of Frankfort- on-the-Main. March 29, 1900. _ The LORD LISTER, F.R.C.S., D.C.L., President, in the Chair. A List of the Presents received was laid on the table, and ao ordered for them, The following Papers were read :— Thermal Radiation in Absolute Measure. 269 I. “On the Retinal Currents of ithe Frog’s Eye, excited by Light and excited Electrically.” By A. D. WALLER, M.D., F.R.S. » | II. “Observations on the Electromotive Phenomena of Non-medul- lated Nerve.” By Miss 8. C. M. Sowron. Communicated by Dr. WALLER, F.RS. Hl. “Variation.” By Professor J. C. Ewart, F.R.S. TV. “Certain Laws of Variation.” By Dr. H. M. VERNon. Communi- eated by Professor LANKESTER, F.R.S. ! i ey “Data for the Problem of Evolution in Man. IV. Note ca the Effect of Fertility depending on Homogamy.” By Professor Karu Pearson, F.R.S. | VI. “ Mathematical Contributions to the Theory of Evolution. VII. On the Inheritance of Characters not capable of Exact Quanti- tative Measurement.” (Revised.) By Kari PEARSON, F.R.S, “Thermal Radiation in Absolute Measure.” By J. T. BorroM.ey, M.A., D.Sc., F.R.S., and J. C. BEATTIE, D.Se., F.RS.E. Re- ved Retanber 28, 1899,—Read February 1, 1900. The experiments* described in the following paper form a continua- tion of researches on thermal radiation by one of the present authors, the results of which have been communicated to the Royal Society from time to time since 1884. The main object of the present experi- ments was to push forward the inquiry as to the amount, and the relative quality, of the radiation from surinces of various kinds 1 in high vacuum. When a body is maintained at a high temperature the total radiation from its surface depends, other things being the same, on the tempera- — ture and on the character of the radiating surface. With a given temperature the total radiation, consisting of thermal, luminous, and actinic rays, seems to depend on the nature and on the ultimate texture of the radiating surface; and the proportion in which vibrations of longer and shorter conrad are present seems to be governed by the * The experimental results of the paper were obtained two years ago. Various circumstances have prevented earlier publication; and it was originally intended to carry the investigation further before publishing. Want of opportunity, how- ever, makes this difficult for the present; and we therefore deem it advisable to put our results on record just now, as they stand. The present investigation, as well as the former work referred to in the text above, has eee assisted By aaa EeOra the Government Grant Fund. + “On Thermal Radiation in Absolute Measure,” J. T. Dothaaley: “Roy. Soe. Proc. and ‘ Phil. Trans.,’ 1884 —1893. Y 2 270 Drs. J. T. Bottomley and J. C. Beattie. coarseness or fineness of the structure of the surface at which the rays take their origin. Very little progress has yet been made towards an investigation of the question just referred to; and the results of our experiments are intended to be a contribution in this direction. In a former paper* by one of us the loss of heat in vacuum from the metallic surface of platinum wires was determined ; and Schleiermachert has compared the loss from bright platinum wires and from platinum wires whose surface was coated thinly with black oxide of copper. Further experiments on this part of the subject seemed highly desirable, and were, therefore, undertaken by us. : The radiating body was a platinum wire. The way in which it was mounted is shown in figs. 1 and 2. The platinum wire, ad, is held, = == = Baitery B heoste ¢ F A _——— ea wens rr = = aa Ses Pg Standard . ‘ /) Resistance ee in Oit Vesset ee = (Oma VENSTTTVI STS TIN —p Sab v. Or Potential Gabv? Fre. 2. stretched between two spiral springs, in a glass tube. The outer ends of the spiral springs terminate in loops ; and two pieces of glass rod, which are passed into tubes, cc, c’c’ (see figure), pass through the loops, so that the springs pull on these glass rods. After the rods have been passed into their places, the ends of the tubes cc, cc’ are closed up, except one, which is used for exhausting. Flexible electrodes are soldered to the loops of the spiral springs, and are silver soldered to * Bottomley, ‘ Roy. Soc. Proc ,’ 1887. + Schleiermacher, ‘ Wiedemann Annalen,’ vol. 26, 1885. t The arrangement has been already described, J.T. Bottomley, ‘ Phil. Trans,,’ A, 1887, vol. 178, p. 448 Thermal Radiation in Absolute. Measure. Del stout multiple platinum terminals; and by means of these, which are fused, with the help of some white enamel, into the glass at d,d, the current is passed through the platinum wire. At ¢,e, platinum wires are brought through the sides of the tube, and serve as potential electrodes ; and it is to keep the platinum wire ab in the middle of the length of the tube, and to avoid pulling unduly on the potential electrodes, that the two spiral springs, one at either end of the tube, are introduced. Two exactly similar tubes were employed, as shown in fig. 2. They were connected together by a side tube, as shown ; and by means of a branch tube, attached. to this side tube and connected to a Sprengel pump, the air was withdrawn from both tubes at the same time ; and by this arrangement it was provided that the vacuum in the two tubes should be at all times precisely the same. In one of the tubes the platinum wire was brightly polished and perfectly smooth, just as it came from the maker’s hands. The other tube contained a platinum wire cut from the same hank, but with the surface covered with an excessively fine coating of soot. The soot was put on by passing the wire carefully through the upper part of a ae paraffin flame.* The usual arrangements were made fee drying the vacuum of the tube, and of the pump, by means of phosphorus pentoxide; and the vacuum was measured by means of the Gimmingham modification of the McLeod gauge. The wires were heated, as in the former experiments, by means of an electric current. Fig. 2 shows the electric connections. A battery consisting of a sufficient number of secondary cells was em- ployed ; and ,the current was controlled by means of suitable resist- ances, including a rheostat. In the experiments here described the platinum wires of the two tubes, the resistances, and the battery were all connected in series, so that the same current passed through all.t The current in the circuit was measured by means of a Kelvin ampere gauge, and the difference of potentials at the two ends of each * The texture of the soot depends greatly on the source from which it is obtained and on the way in which it is applied to the wire. Some preliminary experiments have been made with various coatings of soot, and comparisons have been attemped between surfaces finely coated with soot, and surfaces prepared with platinum black and with a fine coating of black oxide of copper chemically applied to the wire (cf. J. T. Bottomley, ‘ Phil. Trans.,’ A, 1887, p. 449). + In another set of experiments the platinum wires were joined in parallel, and, by means of two rheostats, one connected in series with each platinum wire, an attempt was made to regulate the current in each wire so that the temperatures in the two should be the same. This was found very difficult to carry out; but it is intended to renew the attempt, and determine simultaneously the radiation from two wires with different surfaces, in the same vacuum, and at the same tem- perature. 272. Drs. J. T. Bottomley and J. C. Beattie. of the platinum wires was measured by means of a high resistance ' reflecting galvanometer. This potential galvanometer had a resistance of about 5000 ohms, and-it was possible to insert in the galvanometer circuit an additional : resistance of 10,000 ohms. In order to ascertain the absolute value of the readings of the potential galvanometer, a standard coil of platinoid wire, whose resist- ance was known very accurately, was joined into the circuit, as shown . in fig. 2. This resistance was of considerable length, and it was kept cool by being immersed in a bath of oil. The following was the order of experimenting. The pressure in the tubes was first reduced as much as possible by means of the Sprengel pump ; then a very small current, practically unable to heat any part of the circuit, was sent through the two platinum wires and the standard coil, and the potential difference between the two ends of each was determined. This gave the ratio of the resistance of each of the platinum wires to that of the standard coil, all ‘being cold, and at the same temperature. The current from the battery of storage cells was now suitably increased, and readings were taken in the following order :—The current passing was first read. Then the zero of the potential galvano- meter was noted, and the deflection of the potential galvanometer when connected to the two ends of the standard coil was observed. The electrodes of the potential galvanometer were next applied to one of the platinum wires, and the deflection noted ; then the deflection due to the second wire was observed. A second reading was taken from the first wire and also from the second wire. Usually these pairs of readings were identical, or nearly so, as no reading was taken until after the strong current had been passing through the circuit: for sufficient length of time to allow the temperature of the whole to. become perfectly steady. Generally speaking, five minutes or more was allowed for this purpose. Lastly the current was again read, and the zero of the potential galvanometer noted. The readings detailed above enabled us to calculate the current: passing through each wire and the resistance in that wire. The length and cross section of each of the platinum wires (practically identical) were also known. Thus the energy lost by radiation per square centimetre per second, C2R/JS, could be calculated; C being the current, R the resistance and S the surface of the ER wire, and J being the dynamical equivalent of heat, all in absolute measure. The measurement of the electric resistance of the wires also enabled us to calculate the temperatures of the wires by means of the results. of a separate determination of the electric resistances, at different. temperatures, of the wires themselves. | In a former paper,* the precautions and difficulties connected with * J.T. Bottomley, ‘ Phil. Trans.,’ A, 1887. Thermal Radiation in Absolute Measure. 273 the determination of change of resistances of platinum wires with temperature have been fully discussed. In the present case each platinum wire, after having been used in the radiation experiment, was wrapped round the bulb of an air thermometer* of special construc- tion; the bulb and wire were then packed in asbestos wool, and placed in the laminated copper heating jacket described and figured in the paper just referred to. The jacket was heated by means of one of Fletcher’s powerful ‘solid flame” burners, by means of which it could be kept for any length of time almost absolutely steady, at any temperature below the softening point of glass. By means of stout copper electrodes the platinum wire was made one of the branches of a Wheatstone balance, and the electric resist- ance and temperature were simultaneously determined. A consider- able number of readings between 15° C. and 350° C. were taken, and from these an empirical formula was constructed, or a curve drawn to represent the relation between temperature and pressure at all inter- mediate points. In one respect, the determinations, an account of which is given in the present: paper, are not perfectly satisfactory. We have not been able to take account in a proper way of the temperature of the enclosing envelope. In order to be able to see the condition of the wires, and in particular to observe their appearance when they became luminous, glass envelopes were used in these experiments; and owing to the nature of the arrangements and the method of experimenting, it was not found possible to immerse the glass envelopes in a cooling bath. Consequently the glass became more or less heated during the experi- ments, and the heating was unequal in the cases of the bright wire and the sooted wire. It has already been pointed outt that the proportions in which the radiations of longer period and shorter period are present in the total radiation depends on the radiating surface, other things being the same. In the case of the sooted wire, the quantity of long-period radiation is, in proportion, far in excess of that proceeding from a bright metallic polished surface. Consequently, with the same total electric energy supplied to both wires, the glass tube containing the sooted wire became much hotter than the tube containing the bright wire. It has also been pointed out{ that with a substance like glass, con- ducting badly and somewhat diathermanous, it is impossible to tell how _ * J.T. Bottomley, “On a Practical Constant-volume Air Thermometer, ‘ Proc. Roy. Soc. Edin.,’ December 19, 1887, and ‘Phil. Mag.” August, 1888. This thermometer has proved perfectly satisfactory; and the separation of the volume gauge and pressure gauge make it extremely convenient for applications of the kind referred to in the text. t ‘Phil. Trans.,’ A, 1887, p. sel { Loe. cit., p. 444. 274 Drs. J. T. Bottomley and J. C. Beattie. much heat is returned to the radiating wire from the interior skin of the tube, which no doubt rises to a high temperature during the experi- ment. To a certain extent, therefore, the results which we have obtained must be considered as not affording results strictly com- parable with those formerly obtained in which a metallic envelope cooled with water was used. The absolute value of the radiation observed ought to be somewhat lower in amount than would have been found had the enclosing envelope been of metal and properly kept cool, and the disturbance from this cause must have been relatively greater in the case of the dull, than in the case of the bright, wire. Experiments were made with platinum wires from three separate hanks. -cceee ‘A ARS er ea on e werlita K 650 700 750 ade in ps oes eae 280 Drs. J. T. Bottomley and J. C. Beattie. Fie. 5.—Curves showing emission of heat from Pt, (bright) and Pt, (sooted), from the same hank of wire, diameter 0°015 cm., at various temperatures. 85 eames Sate Chl LE Le a “CLUE a 75 | sal | | | |) 70 ee + | | | | Lo) ee 65\= eet | coves be NO. ) a ee & [| feel [| | bd) ee 50 ais I L RE BRE So UU ao-& Sa Pe, Uulby ho /| i ot Bee || Lb ee 40-28 HMEERGEER aa Pies “TRS 8 bs | 350} my wT Ne mEVAbee iy DS £5,583 — 32° TAY [a 8 4 oA a A ett: | tts) a FERREPZEEeeE ss | ee 28a aake oO 30 450 200 250. 300 450 500 550 600 650 700 450 800 . Tempera aoe in ae Centigrade. Thermal Radiation ir Absolute Measure. 281 Fie. 6.—Emission of heat from three bright platinum wires, Pt,, Pt, Pt;, of different diameters as indicated. Se ee a a ES aS SSS a 7 Meee Tedd ba ¢ oe «eee iv 23 INTIEPOIOS (000. .s- pia emi di 2 RN 5: 169 175 PAT CiCLAN MM OLIZONS 2-25. 508 oo: ek see umes 9 9 Barometers, Marine 0000.2. 000.0 122 92 os StanGand acne ssceet cece 58 85 i. SbARTOM oro. i. hence 55 : 15 IMO CULATS “Ices eae eee ee 314 404 (Womipasses Gs ee a ee 44 43 Wetlectors cise etek vis. oe ne ee 3 6 lyGrOmMeversie Per. esse rene. ts sec see 463 241 Inclinometers’\00.000.2/0°: PREM NP i 5 9 Photographie tenses) 7... er) seen 13 160 IE Vee NSS ah Mra aus mR ATA, CF 2 3 Naw TelesCOpes. 2 ciccn +5: arlene orks eee 681 561 RVATT GAUL ES. a cc oes som ytea + -aeRce ane 12 19 Rain-measuring Glasses .................. 10 44 Scalea tie oan | een: ach ah (eae ee 2 — SYD BEHIND. TA Mle 3 750 876 SUishimen lve cOGUeELs a... se eee ee 15 6 AM COM OMCs ect ies. san eeu ace eee 26 24 Thermometers, Avitreous or Immisch’s 10 5 55 Clinical rca a ieaee 17,962 16,020 ‘ Deep isaac eee 79 19 % Eieh Tange) 20. ..20. 1 56 62 is Hypsometric ............ 38 39 a Low Range See. 94 103 i Meteorological ......... 3,296 2,892 Ks Solar radiation ......-- 2 — yi Spancard: jy.t eee eee 66 104 VOR OTT 2 Ga Ne RS SIR eI Th 8 6 5 Vertical Force Instruments ............ == 1 Declinometbers cei) acted. teers — — Geel ih wh cake 94,434 22,051 Duplicate copies of corrections have been supplied in 97 cases. Report of the Kew Observatory Commitiec. 347 The number of instruments rejected in 1898 and 1899 on account of excessive error, or for other reasons, was as follows :— 18938 1899 Whermometers, clinteal...:.0 0.200.600.0000, 173 149 A ordinary meteorological... 92 78 immer es) ire 0. WY, 106 151 0 RE ee 60 49 eeresecor hi. Wee eR ALON 30 21 Es esl haalle ARs ee 26 14 Two Standard Thermometers have been constructed during the year. There were at the end of the year in the Observatory, undergoing verification, 6 Barometers, 450 Thermometers, 24 Sextants, 150 Tele- scopes, 75 Binoculars, 6 Hydrometers, 2 Rain Measures, 2 Rain Gauges, and 2 Unifilar Magnetometers. VI. RATING OF WATCHES AND CHRONOMETERS. The number of watches sent for trial this year is slightly less than in 1898, the total entries being 469, as compared with 483 in the pre- ceding year. The “especially good” class A certificate was obtained by 78 watches. The highest number of marks obtained is a fraction lower than the highest obtained in 1898, but the average performance shows no falling off, as appears from the following figures showing the per- centage number of watches obtaining the distinction ‘“ especially good,” as compared to the total number obtaining class A certifi- cates :— Year ............ 1894 1895. 1896. 1897. 1898. 1899. Percentage “especially good” 16:1 166 30:5 280 22:1 266 The 469 watches received were entered for trial as below :— For class A, 362; class B, 86; and 21 for the subsidiary trial. Of these 19 passed the subsidiary test, 62 were awarded class B, and 293 — class A certificates, while 95 failed from various causes to gain any certificate. In Appendix IV will be found a table giving the results of trial of the 50 watches which gained the highest number of marks during the year. The highest place was taken by Messrs. S. Smith & Son, 9, Strand, London, with a keyless fusee tourbillon lever watch, No. 238-99, which obtained 88-7 marks out of a maximum of 100. Marine Chronometers.—During the year, 56 chronometers have been entered for the Kew A trials; of these 34 gained certificates, and 22 failed. No movements were sent in for the class B trials, and as the demand 348 Report of the Kew Observatory Committee. for the B certificate has been very small indeed for some years past, the question of the retention of the class B trial seems to require con- sideration. The electrical contact-piece of the mean-time clock “ French ” failed in its action frequently in the early part of the year. This was found to be mainly due to the unequal wearing of the teeth of the old escape wheel. The clock was sent to Messrs. Dent, who fitted a new escape wheel, &c., and its general performance since has been much more satisfactory. VII. MISCELLANEOUS. Commissions—The work under this heading has been of a very varied character during the year. The following instruments have been procured, examined, and forwarded to the various Observatories on whose behalf they were purchased :— 3 1 dip circle and 4 extra needles for St. Petersburg. 1 * yg pene is Toronto. 2 pairs dip needles for Upsala. 1 pair nt » Mauritius. : 1 Kew pattern self-recording Robinson anemometer and sheets, and 1 pocket aneroid for St. Petersburg. 2 Kew standard thermometers and a barograph tabulator for Colaba (Bombay). A standard Fortin barometer, an astronomical globe, maximum and minimum thermometers, and an ozone cage for Mauritius. Anemograph sheets, sunshine cards, and rain-gauge forms have been sent to Hong Kong and Mauritius ; prepared photographic paper to Batavia, Aberdeen, Fort William, and Valencia, for the Meteorological Office ; and to Hong Kong, Mauritius, Toronto, and Lisbon. Gas Thermometer—The instrument referred to in last year’s ‘ Report’ arrived at the Observatory in February. Prior to its receipt, Dr. J. A. Harker went over to Germany and was shown the methods of using the gas thermometer adopted at the Reichsanstalt, Charlottenburg. The Committee are much indebted to Dr. Kohlrausch and other authorities of the Reichsanstalt for the courtesy shown by them on this occasion. The cost of the instrument, including its carriage and Dr. Harker’s expenses at Berlin, was borne by Sir A. Noble, who also kindly expressed his willingness to pay for the auxiliary appliances required in gas thermometry. Owing to the want of a suitable building in which to erect the gas thermometer, the Committee were unable to take full advantage of Sir Andrew’s generous offer for the immediate present, and they have been obliged to leave it to their successors, the Executive Committee of the National Physical Labora- Report of the Kew Observatory Committee. 349 tory, to carry out the final arrangements for the installation of the gas thermometer. Collimator Magnets—A critical and experimental paper dealing with the data obtained in the verification of collimator magnets at the Observatory during the last forty years was prepared by the Superin- tendent, and has been published in the Royal Society’s ‘ Proceedings.’ Discussion on Platinum Thermometry—A discussion on platinum thermometry having been arranged for the British Association meeting at Dover, Dr. Harker attended, with the Committee’s approval, and in concert with Dr. Chappuis gave a summary of their joint work at Sevres. Professor Carey Foster and Mr. Shaw also took part in the debate as well as the Superintendent, who had been instructed by the Com- mittee to attend. 3 Compass-testing Regulations—In consequence of representations by Mr. J. White, of Glasgow, the regulations for the testing of ships’ com- passes have been revised. In this process the Committee had the advantage of the advice of Lord Kelvin and Captain Creak, whose views were laid before a sub-committee appointed for the purpose. At the request of the Danish Legation, the methods employed at the Observatory for the verification of compasses, sextants, and naval tele- scopes were shown to Commander Clausen, of the Royal Danish Navy, who has charge of the verification of naval instruments at Copenhagen. National Physical Laboratory.—Parliament having, on the motion of Her Majesty’s Ministers, voted a sum of money for the establishment of a National Physical Laboratory, to be under the management of a committee nominated by the Council of the Royal Society, the Royal Society have drawn up, and the Government have approved, a scheme for the organisation of the Laboratory. In accordance with this scheme, the Kew Observatory is incorporated with the National Physical Laboratory, and becomes part of the organisation thereof as from the Ist January, 1900. The Kew Observatory Committee as hitherto constituted ceases to exist at the same date, and its property © is to be transferred to the Royal Society. The work of the Observa- tory will, however, proceed as heretofore, and will be carried on by the existing staff. The scheme of organisation already mentioned constitutes an Execu- tive Committee as the authority having the immediate management of the National Physical Laboratory, and this Committee includes at present six members of the Kew Observatory Committee. The scheme also provides for the appointment of a Director, who, subject to the authority of the Executive Committee, is to have sole control and direction of the officials of the National Physical Laboratory and of the work done within it. Mr. R. T. Glazebrook, F.R.S., has been appointed to this office. : 300 Report of the Kew Observatory Committee. The Kew Observatory Committee having been incorporated under the Companies Act, 1867, certain legal forms have to be complied with in order to wind it up, transfer its property to the Royal Society, and. put an end to its liabilities. ‘The steps required for these purposes are: being taken. Inspection of the Observatory.—An inspection by the General Board of the National Physical Laboratory was arranged for October 16th and 18th, when the Chairman and some other members of the Kew Com- mittee attended at the Observatory to assist in showing it to their visitors. On the second occasion the Observatory was visited by fully twenty members of the General Board, including the Vice-Chairman of the Executive Committee and the Director of the National Physical Laboratory. By the courtesy of the Mid-Surrey Golf Club, arrange- ments were made for examining the most likely sites for building afforded by the Old Deer Park. Library—During the year the library has received publications. from— | 21 Scientific Societies and Institutions of Great Britain and Ireland, 103 Foreign and Colonial Scientific Establishments, as well as from. several private individuals. The card catalogue has been proceeded with. Audit, éc.—The accounts for 1899 have been audited by Messrs. W.. B. Keen & Co., Chartered Accountants. The balance sheet, with a comparison of the expenditure for the: two years 1898 and 1899, is appended. PERSONAL ESTABLISHMENT. The staff employed is as follows :— C. Chree, Sc.D., F.R.S., Superintendent. T. W. Baker, Chief Assistant. E. G. Constable, Observations and Rating. W. Hugo, Verification Department. J. Foster be es T. Gunter "i Se Mi... Boxall 23 if G. E. Bailey, Accounts and Library. E. Boxall, Observations and Rating. G. Badderly, Verification Department, and six other Assistants. A Caretaker and a Housekeeper are also employed. In addition to the above, Dr. J. A. Harker has been employed in the: capacity of special assistant to the Superintendent. (Signed) G. CAREY FOSTER, Interim Chairman. Report of the Kew Ubservatory Committee. List of Instruments, Apparatus, &c., Jol the Property of the Kew Observatory Committee, at the present date out of the custody of the Superintendent, on Loan. To whom lent. G. J. Symons, F.R.S. The Science and Art Department, South Kensington. Professor W. Grylls Adams, F.R.S. Lord Rayleigh, F.R.S. Radcliffe Observa- tory, Oxford. The Borchgrevink- Newnes Antarctic Expedition. The New Zealand Government. _C.T. R. Wilson, Esq., Cambridge. Articles. Portable Transit Instrument.......... 2000 Articles specified in the list in the Annual Report for 1893. Unifilar Magnetometer, by Jones, No. 101, complete. . Pair 9-inch Dip ‘Needles with ‘Bar Magnets . os Ne Standard Barometer (Adie, No. 655) .... eerte Black Bulb Thermometer in vacuo .....ee00. Dip Circle, by Barrow, No. 24, with four Needles and Bar Magnets...... Unifilar gis marked N.A.B.C., complete. . Dip Cir cle, by Barrow, ‘with one air ‘of Needles and Bar Magnets. atare diane ‘Evipad Stand: ei iahe «cys by Jones, eeeeoeeeervneeaeaeeve Electrograms for 1897 .... Date of loan. 1869 1876 1883 1887 1885 1897 1898 1899 1899 1899 1899 Report of the Kew Observatory Committee. 352 9 €& P6EhF 9 & P6ErF ae 6 O18 Lal ols t% ° seers (Used AW10g) puBY UL 9 ¢ 61 Pence meee e een neee sae nee see censeeeceree Suryueg Sunteay 6 PFI 9OL “ooveretetteeeerereress Sue AJUNOD pus UopuoT—eourleg 0 G 98 ‘ULOT JOJ SJLOTUIMAYSUT OTYOUSU, LOJ 90}{LUIWOD JUeLH JUOUTUTOADD 0 ¢ 98 ULOT LO} SJUBUINAISU] OIUSEAI—I9.AOG “WM “VW “AIA 0 0 GL SMeeeees es ToMNGuqUOD $,a[GON MalpUY IIS—AojouOUTLOYL, sey OR aoe 0 el 6 seeeeeoeee TIO 07 QISTA § 1dyLBH "Id jo sosued xy 8 6I er POE O HOO e OHO OOO EEE OEE OOOO SOHO OEE SEDO EEOH SEC OSE EEE EHy HHO ees yo0qg BIpUy uo SPpUOPIAIC 0 OLt9 CCT 8 UlLeg WO osvILieD pue snjeieddy — caJUIOWLaUL, SBD 0 aL seeessenessesssestseasesvenseneatonnesvnrsaneceserene sesteeeeereenenensensenseeentetes SaUIQUT FOOL -=——==— 0 0 Off “CC aIngIpuedxy UOTZeMstUIUIpY Jo uoTAodo1g 2 & OLF °* ‘Om ‘SUOIINITSU] USIaL0y PUL [VIUO[OD IO}Z payndaxe sUOTSSIMIUIOD 9 OL OPE “"""* “Os ‘SUOTINITISUY USIO10g PUB [BIUOTOH 10}F sodeg dIYdvisojoug pus SJUOUWUNISUT jo esvyoing 0 OL POO O ere Coen e eS OOOO EOOs sue eeeeeses es SESE ODEs rH OOSOOHnesens wel-[joq usyorq Oy — SUOISSIULWIOD uoyesusdu0p—Auvdu1og ABATIEY UL9}SOA\ YINOG pue uopuoTy ’ Tt WI— — : sotpo1easey b IL tg """"""' OINgIpuedxy UOTeTstUlUIpY Jo UoT0do1g 6. oui —— 0 a OCI Pee eeoreeessscnee ee ererewosce “oO” ‘sosuodxq [ejueprouy OL 8 CP OOO OOOO SCOOTER ser SSS SEC LET: 08084 OOH 1 2005S CEH 08s OHS SOC OSeEseeES sosueT 0 SI 696 PURE raetrmacoer nay oot SSS CSRS Ste Neel ne Sek Reese ee Eee ast LORS 0 Il 9IG COOOL EOS HH OO OER EO OOD EOED! SESH HOO EHAHS! COO OETOETEOT OEE OESESe SUIICH —:8189], GQ OL OIL reeteeerttesseteteste ee estes etsevesee enero neees UOT ROTLG A, GeGaGc) == —:8}sa.L 0 0 68g “""""* aINGIpUsdxY UOTZeAYs!UTUApY JO UoTytodo1g | o 8 Ser == LOL BZ tetettttesteteteeteeeeeeeeteeeeeenny (soguad xg [eJU@PIOUT 0 8 GJ “Wreeeersreteserseseserssececeee SHAM HIBPUBIS S,YIBIO JO 0g OL Il 0% Ce ee ee eee ee ee SOLIBIeS 9 Or c OC C80 OCS POLS CHOSE OOOeeeseeeeosed “oO” ‘sodBIS0g —: sayo.vasoyy PT) a Gag CaO sae —‘[L0UNOD [eolsofO.LNAIE TT 0 OT 61 '* OInjIpuedxA UOleAysIUIUpY Jo uoTJodo1g i ORS == IL SI FL CCE meee rerereronsererssearsroes sence Bo a er wooo Ot 5) 6 6I tI Soseeeesocereresear® poumjos XB, amo0uyT oe ce ‘redeg o1yde.sojoug ‘snyereddy ‘sosuedxq [euepiouy OL OL OFF CT ree quautded [enUUY “4S, JOISSey € Gl 6pe cot’ “om ‘suoelnqgey, ‘suOIyeALosqQ—sorte|eg ee Ser —: Aqa1o0g [elo ——* ALOJVEALISAO [BULLION 4g ¥ 6 0c&9 SOCKS OEE OTE THe FCHHHE SHEE GEEEEH FSH EEF SSS OEEDE S681 (vox woud QOUB[GG OL “p Uy F “Pp ks} F “p Urs) ¥ “SINGWAVd . “SLdIHOeN "6681 “OS “aquiaray Huapua woah ay sof (Qs saquasag aous wopopmbryT ayy fo asoyg buapryour) squauhvg pun spdvooary fo ywnosopr “f.sognasasgg nay 303 Report of the Kew Observatory Committee. *2Uapuajpwrladingy ‘TUUHO SUTUVHO = (peusts) €& 8 0908F * ¢ $862 PTUUUERUUT TUPI ooUBleg [e10uex) PTT TTURETT TIRE ee ee ide 10J9UL -OULLOYJ, SV) OF UOLINGIAJUOD 8,0[GON MOIPUY ALS ‘ Hh Gc v sereeeneonseaerossoncserecsesesoerst1 TBI GOUSIIG 10} que.ty jo aouvleq quodsuy, Pee eeeeeererseeeesaseeeseee®e PUUUUEUE TE ee eee) SIYOLBIS9tT PTTTTTTTTET TTT eee SUOISSIULOLO/) ceecenssrooeensenecouseansensassaeaearenseusosesrusesenssaesssarsersreerss TINODIB S489, [, srerennvanssseseseoseorers tony Cody a1yderdojoyg—sjunooov ALOPVEALOSAO Fabieceieecen sien vouseneens eine ‘saredoyy ‘109CM ‘sei —8]UN0008 UOLBAISIULWUI pV OL ‘SHILITIAVIT CHLVWILSH b LT 8621F 6 be os ‘sayedoy ‘toxeja18D Y 96 wee eerewereseee ny, ‘Tong ‘que L LGp ‘op UVeiquy GUeISISSY SIT] 0 *s a b I 862lF 0 0 O&T y IL 8c 0 0 ese “" ** SQYDIBISOY 0 Ol 61 eee eoeeeerereee ALOYCALOSYC) "p *$ F ft “yuauuon.woddy eee ree eee eee ene reeeee SUOISSIUIULO(K) id ssa, | + 6 0 00¢ Pere eee quopusqulsodng F *SL0ININMD EEE SS Sn eee a: Se oe ne dee ee OE Oe "AAOLIGNAdXY NOILVULSININGV $ 8 09087 | 0 8 6L G 91 &¢ G¢ Fy &ér >) VU 985) = 0% 2 G LI §9S b GI OOL p's 3 D'S F Vunjunoop paiavoy) ‘NAM “A “AA ‘O06T “2761 Auwnuns OORSe COEF HOSTESS SEE OHT EH THEH EHH HEH Hee SLOJOULOWLLOT, J, parepueys Sees teeseseeeecoeeesesasseserare SOJBOYIFLID pues SUTIOW UV —: yoo Saboue cBanea ses vend stews rcreceuecnss ces se cizcsles S248 4)700 ‘SUOISSTUIUIOD sseaneeserosaseosneseneecaesennerenarensecessensasennersrerers BOQ q GRO T reseee ecm ‘godeqeog ‘AOUBMO]LY —[IOUNDD [BoIdo[O.LOojoI —‘onp sjuow ded Messi seas Oe. 2 CM Ee eer Unb r caenige SY cee gee owe ges ean (CES PI ‘; Atenuve uo onjwa ‘yoo}g “quod aod Fg VIPUL OOSTF SOO PORT EHH SET OSES EE EEE BE EEHHeEE EHH EEE Eee queuleyeys ted gv DIUG[VET AST ‘SLUSSV CHLVWILSH (pousig) ‘0061 ‘Scenuee UIST ‘qo04100 punoy pus AJoI00g [vAOY oY JO JeYoq UO pozIpNV fieport of the Kew Observatory Committee. Comparison of Expenditure during the Years 1898 and 1899. Expenditure. Administration :— ae oa ee jajeieieus First Assistant. . Office. . Rent, Fuel, ‘Lighting, & &e. Caretaker. . Incidental Expenses .. Normal Observatory :— Salaries—Observations, &e.. : cece Incidental Expenses soilale Prop. Adm. Expenditure Researches :— RSANATACS <5 5/0. s)0 (nie aiateians siete Incidental Expenses . Prop. Adm. Expenditure Tests :-— Salaries.. ae Incidental “Expenses .. Geis Prop. Adm. Expenditure Commissions :— Purchases for Colonial Institutions, &c. ... Prop. Adm. Expenditure Seismograph.. Shia Gas Thermometer ....... Magnetic Instruments for HiGan ss. “epee eeceoeesceeeeee Gross Expenditure.... (showing an increase of £8 2s. 1d.). Extraordinary Expenditure. Normal Observatory :— Incidental Expenses .... Researches :— Salaries) sic \ejeyel sic sie'e 6 Purchase of ee BEC vs diets Commissions : Purchases for Colonial Institutions, &c..... Seismograph...........- Gas Thermometer....... Magnetic Instruments for GB 206 vere ia ae eeecte #60 ee Leaving for Ordinary Nett PR PEMOIDUINE Falele ee! eas (showing an increase of £59 18s. 5d.). 805 1 11 1898. 1899. BS) PS) 2 Gs GE srimnae 500 O O 500 O O 335 8. 0 331 18 0O 121 10 O 125.9. 9 87 16 6 96 4 4 68 18 0]. 6818 O Ve 12... } oat 3 1249 4° 77) 1268 te 38386 15 6 349 15 8 Al a Cer 74,18 11 187 10 O 19410 O 158 8 O 204 11 10 64 9 2 28 10 27 375 0 O 389 0 O 918 6 9O 969 18 O 222 9 5 156 120 0 499 4 7 584 11 4 520 aa 340 10 6 187 10 O 130 O O 5515 O [Ad aw 86 5 O 3575 12 4 | 3588 14 5 | > Sm 158 8 O 204 11 10 61 15 10 2 Nias 529 3 1 340 10 6 bo 15 © 7411 O 86750 753 5 7 2770 19 5 | 2880 8 10 Increase. Decrease. CBr ies ss. a, © 107 6 319 9 8 7 10 3719 2 50 6 9 E10 O 1219 9 a8 17 VA i (Oreo 46 3 10 35 18 7 14 0 0 51 12 40 65.17 5 85 6 9 188 12 7 57 10 O 55 5) 0. 7411 0O 86 5 O 411 15 8 | 4038 13 7 23. 80 46 310 Bt 26, 7 188 12 7 bm 1d 0 74 dh, WG. 86 5 0 230 710| 282 4 2 eee L8L 7 10 Wai 9). Report of the Kew Observatory Commiatice. 300 APPENDIX I. MAGNETICAL OBSERVATIONS, 1899. Made at the Kew Observatory, Old Deer Park, Rich- mond, Lat. 51° 28’ 6” N. and Long. 0° 1" 151 W. The results given in the following tables are deduced from the magnetograph curves which have been standardised by observations of deflection and vibration. These were made with the Collimator Magnet K.C.I. and the Declinometer Magnet marked K.O. 90 in the 9-inch Unifilar Magnetometer by Jones. The Inclination was observed with the Inclinometer by Basie No. 33, and needles 33 inches in length. The Declination and Force values given in Tables I to VIII are prepared in accordance with the suggestions made in the fifth report of the Committee of the British Association on comparing and reducing Magnetic Observations. The following is a list of the days during the year 1899 which were selected by the Astronomer Royal, as suitable for the deter- mination of the magnetic diurnal inequalities, and which have been employed in the preparation of the magnetic tables :— DUAL. oes acre ce ses ihe eelOe Pas 21, PE WPUREV 52 cia ese ve A EGE Ss PO as cciste cance Bn a Sie s 14 ee 13, 15, 16; 29, 22. IE ie Tis. ii Ses 13,°14, 24, 25, 29. A sie ls viiastans asks Gris UC, 20,526) Sse cau page 15, 17, 22, 28, 29. POU USE: Yes. tec 1 ilgeh Geel id ia DCPLCMADET s.0...,+.--- 20) On ty VA 20). Cletomen. f.sk:...-esnen- 2 oA, 20, 29: INevermber so fois DAO TO 20. Bi. December... o.is:..+ Goll pals, 24. to >) VOr, LAVI. 306 Report of the Kew Observatory Commuttee. Table I.— Hourly Means of the Declination, as determined from the Preceding | 4,- | | | | HeOWES | 7) Mid: | 1. | 2. | Bo) a. | 5. | | 7 Be ae | i, '| 11. (16° +) West Winter. 1899. | | Months. U / , / , | / / | if / / / / / Jan. .. 61°3 57 *9| 58°3 58-2) 58°38) 58-2) 57-9 58:0) 57 9) 58°1) 58°4' 59-0) 60°0 : Hebs =. 61°4 58°0\ 58°1) 58°1| 58°0, 57°9| 58:0 58:0) 58:1) 58-5) 59-1) 60°1|60°6 : March. 62 °1 56°9} 56°8 56°6) 56 7) 56°8| 56°9 56°7| 56°7) 56-0 55°6 56°81 59°5 Oct... 59 °7 55°3| 55°38) 55°3) 55 | 55°1| 55°O 54°8) 54°2) 53°5) 53°8 55 °2) 57°3 NOV: 5’. 58°3 54°5| 55°1 55 °2) 55°4) 55°38) 55°O 54°9) 54°5) 54°1) 54°4) 55°7/ 57-1 MCC. Ta. 57 °5 55.°0}. 55°1, 55°1) 55 2 55°38) 55°38 55°1) 54°9) 54°9 550, 55 °5| 56°3 : | SS SS |S ee S| — = SSS -_ —— = | : Means 60 °1 56°3) 56°5| 56°4) 56 5) 56 °4 56 is 56°3| 56°1) 55 2 56 * 57°1) 58°5 ! | | Summer. , / | Uf / | / | / / / | , / | / / / April.. 61°5 56°9) 57 °0| 57 "2| 57 ‘1 56°7| 56°6) 55°9) 54-8] 53-6) 54-2] 56°3/ 59-1 May .. 62°3 57 °0| 57:0} 56°8' 56°6) 56-1] 54°9) 53°9) 53-0} 53 "1) 54°77) 57°2/59°8 June.. 61°6 56 8] 56°7| 56°5) 56°5! 55 °7| 54°3) 53-0) 52°6) 52-6) 58°3) 55-5) 58 °1 July .. 60 °2 56 °4) 55°8) 55°7, 55°4 55:0) 53°9| 53°6) 58 °6| 53°1) 54-1) 55°7/ 57-9 JE SG 61 °5 55 5] 55 °4) 55 °5) 55°38 54°8 54°0) 53°2) 52°5) 52°8 54°1) 57-0) 59-7 Sept... 61°6 56°1! 56°0) 56°2) 55°9, 55°7| 55°3) 54°7) 53°6 53 °4 54°4) 56°4) 59-4 Means 61°5 56 °4; 56°3 oe 56°1 55°7| 04°8) 54-O)bsce 53-1) 54°1| 56°3) 59-0 | | | Table II.—Diurnal Inequality of the l Hours | Mid. ‘| 2, | 3 ba | ee a | 7. | eee | 10s 4. ia Summer Means. , ’ , | , , ? , | , t 7 | , , —0°7 |—0°8 |—0°8 +1°8 ee 8 =1°5 |—2°3 |—3-1 |=3°8 | =4r0 eae | Winter Means. | / , | / / | / / , , / / / / —0'8 |—0°6 |—0°6 |—0°6 0-6 —0°7 |-0°8 |—1°0 |—-1°'2 |-1'0 | 0°0 |+1°4 | | Annual Means. , ’ ‘ D , , , | , , tA , / 4 = 0-7 |= 027 eae a a ape —2-°0 -2-4 —2°6 coe +16 Norre.—When the sign is + the magnet 9 A dee »” Report of the Kew Observatory Convmittee. 307 selected quiet Days in 1899. (The Mean for the Year = 16° 57''1 West.) Succeeding | Noon. noon Je 10. | 1 oa | 2 Winter. , ’ , , / t 7 | , t , , ? / 7] 8:9} 59:3) 59°0| 58°8 58°7) 58°4) 58°2| 58°0)| 58:1) 58°0} 61°1 61-1 | 61:0 | 60°7/59°7| 58-8} 59-3) 59-1) 58-9; 58°5) 58°0| 58-0) 58-0) 57°8| 61-1 1:0} 59:2) 58°4|-57°9| 57°5| 56°9) 57°2) 57°0) 56°9] 57°1) 62°2 58°6 | 59°5 |59°3| 58-2! 57°1) 56°5) 56-2) 56°1| 55°8) 55.6) 55°5| 55-1] 55-2 58°9 67°8 | 57°8 |57°1|55°9| 55°6) 55°6) 55-5) 55°3] 54°9) 54°6] 54°7) 54°7| 54-7) 957-9 57°O | 57°3 |56°6| 56°1| 55°7| 55°5| 55°38) 55°0] 54°9) 54°7| 54°8) 54°9] 54°9) 57°3 59°5 | 59-9 |59°3|58-3| 57-6] 57-4) 57-1) 56°9| 56-6) 56°4/ 56:3) 56°3| 56:3; 59-8 Summer. / / / / / / 4 | / / / / / / / 62°1 | 68°9 |63°8|62°5| 60°5|) 59:1) 57°9) 56°9) 57-3) 57°3) 57-2) 57°2) 57-1 61 °7 62°2 | 62°7 |}62°1/60°5| 58:9) 58°1| 57°7| 57°7| 57°7| 57°5| 57-4) 57°38) 57-0 63 °2 61-0 | 62°2 |62°7/61°9) 60°6| 59:0) 58°5| 57°8) 57°5| 57°4| 57-2) 57°38! 56°9 61 °7 60°S | 61°4 |61°4/60°2) 59°0) 57°9| 57°4) 57:0} 57°1) 56°9) 56°9| 56°6) 56°5 61°0 61°0 | 61°9 60 °8 | 59°3) 58:0) 56°4| 55°9, 56°1| 56°2) 56°2) 55°9; 55°6) 55°6} 62°5 - 61°9 | 62°9 | 62-3) 60°1| 58°3| 57-0) 56°6 56°4) 56°5) 56°3) 56°3) 56°2) 56-1) 62°7 | | | | EE 61°4 | 62°5 |62°2|60°7| 59:2) 57°9| 57°3) 57°0) 57-1) 56°9| 56°8) 56°7/ 56°35) = =62°1 | | Declination as deduced from Table I. Noon 1. 2. oe 4. F 6 | 7 | 8 | 9 | 10 11. | Mid Summer Means. / / / / 7 / / i , | , / | , v4 +4°3 |+5°4 /+5°0 |+3°6 |+2°1 |+0°8 |+0°2 |—0°2 |—0°1 0-2 —0°3 —0°5 —0°6 | | Winter Means. / js / / / / / / / / / / / +2°S |+2°8 |+2°2 |+1°2 |+0°6 |+0°3 |+0°1 |—0°2 |—0°5 |-0°7 |—0°7 |—0°7 |-0'8 Annual Means. / / / / / / +0°5 }+0°1 |—0°2 |—0°3 |—0°5 |—0°5 ‘ / : 7 / / +3°4 |+4°1 14+3°6 ll ae —0°6 |-O°7 points to the west of its mean position, ” east ” ” {5 ie] bo 358 Feport of the Kew Observatory Committee. Table III.—Hourly Means of the Horizontal Force in C.G.S. units (corrected Hours | Preceding | wria. noon. 0°18000 + 1899 Months. Jane .'. 381 381 Hebi: 380 385 March. 374: 383 Oct. .. 387 405 Nov. .. 400 402 Dec 404, 407 Means 388 394 April 363 386 May 370 394 June 376 399 July .. 381 397 ATES. « 388 401 Sept. 381 404. — —=—— | ————— | Means SENG 397 Hours| Mid. | Ie | 2. + 00009 + 20004 + °00004 - ooo ~ ‘00001 — 00001 *00000 + 00002 + nae -00002|+« “00002 EOP ON NER IOM AI FOS ALLL (The Mean for the if 2. ae. 4., 1) 1, a Winter. 338 388 386 406 410 411 388 389 387 405 407 411 381 381 368 387 394 406 382 384 383 4.06 403 406 388° 386 387 407 410 411 380 383 373 392 398 408 381 380 367 387 392 405 382 383 382 405 4.04 407 ee | 394 | 394 | 395 | 396 | 398 | 398 | 398 | 394 | 389 “386° 385 385 383 405 409 410 : fal? Summer. 372 362 377 381 380 | 375 | 376 | 379 383 | 374 | 371 | 376 386 384 391 391 383 376 387 388 392 | 386 397 | 390 378 367 381 383 366 361 376 379 362 363 377 380 386 390 394 397 399 | 398 403 | 402 388 387 395 395 395 401 387 390 395 397 386 393 396 396 4.00 404, 396 | 395 | 394 393 390 | 385 | 379 | 374 | 372 | 373 Table 1V.—Diurnal Inequality of the ee | hee s | a | pa | 4 | + *00003 + -00002 Summer Means. + :00001 — oa 00 *000i3/— 0018 — *00020 |- 00019 Winter Means. | + *00002 + 00004 se -o0004 + °00003 — -00008 - “00009 \ 0000 — 00005 Annual Means. | | + *00002) + °00003) + 03 — “00002 2 00007 00012) — ot | ~ 00014 Note.—When the signis + the Report of the Kew Observatory Committee. 309 for Temperature) as determined from the selected quiet Days in 1899. Year = 0°18393.) | ., | Succeeding Meon.| t. | 2. | 3. | Bee BE Gs | lS | a: ae 11. | Mid. nee i | | Winter. 384 388 | 388 | 384 | 380 | 384 | 385 | 386 | 385 | 384 | 383 | 383 | 384 388 382 387 | 389 | 387 | 385 | 385 | 386 | 387 | 389 | 389 | 388 | 388 | 389 382 372 | 878 | 382 | 383 | 383 382 | 383 | 385 | 384 | 388 | 384 | 385 | 384 373 389 393 | 399 | 402 | 402 | 405 | 406 | 407 | 408 | 408 | 408 | 407 | 408 395 397 403 | 406 | 407 | 409 | 410 | 411 | 410 | 408 | 408 | 407 | 407 | 407 397 406 408 | 406 | 408 | 409 | 411 | 412 | 412 | 411 | 411 | 410 | 410 | 411 410 388 893. | 395 | 395 | 395 | 396 | 397 | 398 | 397 | 397 | 397 | 397 | 397 391— Summer. 365 371 | 378 | 386 | 387 | 388 | 394 | 391 | 391 | 392 | 391 | 390 | 391 363 376 385 | 390 | 391 | 391 | 392 | 395 | 399 | 400 | 397 | 396 | 396 | 395 378 381 886 | 391 | 396 | 399 | 400 | 406 | 409 | 408 | 406 | 404 | 402 | 400 380 385 888 | 394 | 399 | 399 | 401 | 403 | 407 | 408 | 406 | 405 | 403 | 402 386 384 390 | 396 | 398 | 399 | 402 | 404 | 410 | 411 | 410 | 407 | 407.| 406 385 386 394 | 401 | 404 | 405 | 406 | 406 | 408 | 406 | 405 | 406 | 406 | 405 381 380 386 | 392 | 396 | 897 | 398 | 401 | 404 | 404 | 408 | 402 | 401 | 400 379 | Horizontal Force as deduced from Table III. l Noon m4 | 4, | 5 | 6. 7 | g | 9 | 10 | 11 | Mid 1 | 2, Summer Means. | an Rao + *C0009 — °00012 + 00012 + “oon oon + 0010+ 009+ 00008 | || = a6 00000 ae “00004 + °00005 Winter Means. - 0008 — 0002 + 00001 + “00001 0000+ “00002 + -00003 a 0005 + oon + 00008 + ogo + -00002| + "00002 | Annual Means. — 00009) — -00004 0000+ non + oe °00004/ + °00006)+ *00008 + °00006 a5 ona 00007 a 5+ "00005 reading is above the mean. | 360 Leport of the Kew Observatory Committee. Table V.—Hourly Means of une Vertical Force in C.G.S. units (corrected (The Mean for the Eee: | eee eee | eos 2 | 6. | 7) gag eer eto, an: 0:43000 + Winter. 1899. : | Months. Daye... 840 846 | 845 | 844 | 844 | 844 | 843 | 843 | 842 | 841 | 841 | 841 | 841 Helo 7... 841 844 | 843 | 843 | 842 | 843 | 842 | 842 | 841 | 840 | 841 | 841 | 841 March.. 848 866 | 865 | 864 | 864 | 868 | 862 | 861 | 862 | 861 | 859 | 855 | 850 Oct... 863 871 | 870 | 869 | 869 | 868 | 867 | 868 | 869 | 869 | 867 | 864 | 861 Now. cs 820 825 | 826 | 827 | 826 | 826 | 826 | 824 | 825 | 825 | 823 | 822 | 822 Deer. « « 827 830 | 880 | 830 | 830 | 829 | 830 | 830 | 830 | 829 | 828 | 827 | 827 Means 840 847 | 847 | 846 | 846 | 846 | 845 | 845 | 845 | 844 | 848 | 842 | 840 ~ Summer. April... 843 865 | 863 | 863 | 863 | 863 | 863 | 864 | 864 | 864 | 859 | 854 | 846 May)... 843 861 | 861 | 860 | 860 | 860 | 860 | 862 | 860 | 858 | 851 | 843 | 840 June ... 840 851 | 850 | 849 | 849 | 847 | 847 | 846 | 846 | 844 | 840 | 837 | 831 July . 852 866 | 864 | 863 | 862 | 862 | 863 | 861 | 861 | 859 | 857 | 854 | 847 PATI 6 ave 841 856 | 855 | 855 | 854 | 854 | 854 | 854 | 854 | 851 | 847 | 841 | 838 Sept. . 855 868 | 868 | 867 | 866 | 866 | 866 | 866 | 867 | 865 | 859 | 857 | 855 Means 846 861 | 860 | 860 | 859 | 859 | 859 | 859 | 859 | 857 | 852 | 848 | 843 |- Table VI.—Diurnal Inequality of the Hour Mid. | a: 2 | 3. | A 5 | 6 | 7 | 8 | 9 | 10 1. Summer Means. | + "00003| + 2 oc 00002 + 0001 + *00001 + °00001 - *00001 - 00006) 0} ~ 00015 + °00001 | + °00001 Winter Means. + 00001 — °00001 |— °00001 |— °00002 - 0005 “00004 - *00006 Se 000 00000 | “00000 | 00000 - 00001 Annual Means. c 00002 + 001+ no “00000 | *00000 | *00000 | 00000 | 00000 - 00001 - “00004 - *00007 |— -00010 Note.— When the sign is + the e Report of the Kew Observatory Commuttee. 361 for Temperature), as determined from the selected quiet Days in 1899. Year = 0°43852.) | | | | | ny Succeeding Noon. | 1. | 2. 3. | 4 | 5. | Get. ee | SF. pe | 10. Li. aa, noid | | / ; | 841 | 842 | 845 | 847 | 847 | 848 | 841 | 839 | 840 | 841 | 842 | 842 | 842 | 843 | 841 | 842 | 841 | 842 | 842 851 | 853 | 861 | 864 | 871 | 874 872 | 872 | 871 | 869 | 868 | 866 | 864 861 | 862 | 864 871 | 873 | 872 | 825 , 828 | 831 | 832 | 831 | 831 | 830 | 829 | 828 | 828 | 827 | 827 | 827 828 | 830 | 831 | 833 | 834 | 834 | 834 | 834 | 833 833 | 833 | 833 | 832 @ =] = @ ~I He 2) sJ je (e.2) ~J i ios) ~I jo) (0.9) “I j=) (@ ) ~I (=) 841 842 845 | 848 | 850 | 850 _ 849 | 849 | 848 | 848 | 847 | 847 | 847 839 Summer. | | | | ere. 843 846 | 853 | 859 | 863 | 867 | 869 | 868 867 | 866 | 866 | 865 864 | 842 841 846 | 851 | 859 | 862 | 864 | 865 | 864 | 864 | 862 | 860 | 859 | 858 835 836 841 | 849 | 854 | 858 | 859 | 862 | 862 | 860 | 857 | 852 | 849 | 849 829 847 852 | 857 | 860 | 866 | 869 871 | 873 | 872 | 870 | 868 | 867 | 865 | 850 838 843 | 851 | 858 | 862 | 863 | 863 | 861 | 859 | 857 | 855 | 854 | 854 | 839 854 | 857 | 863 | 865 | 869 | 870 | 869 870 | 870 | 869 | 869 | 867 | 866 | 849 843 847 | 854 | 859 | 863 865 | 867 | 866 | 865 | 864 | 862 860 | 859 | 841 | \ 4 | | | | | Vertical Force as deduced from Table V. Noon 1. A Vin a | 4, | Bee Vp | 7. | 8. | 9. | 10, | ul. | Mid. | { | Summer Means. i | i i | — °00015 — caeial -00004| + 0001 == *00005| + °00007) + “00009 + 00008 + 00007) a= aa + nc + 00002) + ‘00001 { | eI : | Winter Means. | | | | | = 005 - ‘eae “00000 + 0002 a= hasty + “00004! + °00003 + 00003! + *00002|+ °00002 + °00001) i | ai -00002 + °00001 \ Annual Means. | | | | | = 01 — 00007 = — + 00002 + Aa eg "00006| + 00006) + aca + 00005) + 00004 + be: na} 00001), | | ( i reading is above the mean. 362 Report of the Kew Observatory Committee. Table VII.—Hourly Means of the Inclination, calculated from the Horizontal | | | Hours | Freeeding | aria} 1. | 2 | Be | s | arn & ewe | 1 | noon. | | | ‘ 67° + Winter. i i | 1899. Months. , , , / Ae bale | , | , / / ’ , f | Jan....., 15°2 |15°4/15-3|15-2/15-1|15-0)14°8| 14°8|14°8| 14-8] 15 2| 15 -2/ 15-2 Feb 15°3 ESO Dae 15°1/) 15-0; 15°0| 14°9|14°8 14°7 | 14-7] 15:1) 15°2| 15°73 March.. 15°9 17 S97 11548 15°8/15°7 | 15-4 / 15 °4. | 15°4)15°7|16°2/16°4|16°4 Oct ‘ 15 °4 14°4 | 14°3 | 14:4) 14°4.| 14°4| 14-2) 14°3 | 14°4) 14°91} 15°2 15°4|15°4 Nov. .. 13 °4 13°41 913-31 13s te 12-9 | 12 AZ 8 | 13 °0| 13 °4| 18°6)138°8|13°S Dec... 13 °3 1352 ovo eso ele 012-9 12°9 12 9112-91 12-9) 13-0).43°1) 13°2 = ae SESE) eee Sr ee ee eee ——— |__| ——- | ——-|-———_ ss SSS ee | 14°2|14°2)14°4)14°7|14°9|14°9 Means..] 14°7 | 14°5/14°5/14-5/14-4/14°3/ 14-2 | | Summer / / / / , ri / / / 7 / / / April 16°5 15°5|15°5|15°5 | 15°4:)15.-5 | 15-4) 1575) 15-7) te eee 16°6 May.. 16°0 1459115 0 | 14591 1521415 1|15°3 15°6|16°1|16°6|16°7/|16°6 | 16°4 June. 1555 14°3|14°4/14°4/14°5]14°5)14°4| 14°7|15°0/] 15°3|15°4/|15°4| 15° July.. Les 14°8 | 14°8|14°8|14°71/14°7114°9115-1/15°3|15°6|15°7|15°7| 15 °4 Aug. aie 14°7 14°3/14°3/14°3]14°4) 14°4/14°6] 14°8| 15-2) 15-6| 15°8/15°6| 15°3 Sept... 15:6 14°4| 14°4)14°4/14°4]14°5|14°6|14°8|15°3 | 15°7| 16°2|16°3| 15°9 we ee ee | a ——— |—__- | _—— a Pees |S es Means.. 15 °6 14°7|14°7(14°7|14-°8/14°8|14°9|15°1|15- 4) 15-8] 16-0) 16°0 158 Table VIII.—Diurnal Inequality of the Hours Ma. z, | 2. 3. 4. | Boge | "7. 8. : Goede. 14s Summer Means. | 4 , / / / | 7 | / / / / 7 7 —0°3 |-0°2 |—0°2 |-0-2 |—0°2 fee bs +0°5 |+0°S |+1°1 |+1°1 |+0°9 Winter Means. | ) ‘ | / / : / / | / / | / / / | , f +0O°1 |+0°1 |}+0°1! 0°0 |—0'1 |—0°3 |—0°3 |—0'2 | 0°0 |+0°3 |+0°4 |+0°5 | | | | | | Annual Means. / / / / —0O'1 |—O'l |-—O°1 |-0°1 / | / | / , / , , —O°l |-—0°2 -o1 +0°1 |+0°4 |4+0°7 |+0°7 | | { +0°7 | Note.—When the sign is + ? ’ Report of the Kew Observatory Committee. 363 and Vertical Forces (Tables III and V). (The Mean for the Year = 67° 14’-7.) | Wome) 1.) 2 | 3. | 4 ls. | 6) 2 ts. | 9} to.) i, (wee poem . | noon. Winter. / ? , , / , , , / / 7 / U / mee 4 149) 15.°2 | 15°4)15°2) 15°12! 15°0/15°2) 15°11 15°2)15°2)15°1 14°7 15°1 |14°8/14°6/14°8)15°0/15°0/14°9 | 14°9)14°7 | 14-7 | 14°7 | 14°8 | 14°7 tok Seems 7) bo 7 | £5°71)15°9|16°1|15°9/ 15°8 | 15°8 | 15°9) 15:8 | 15°6)15°6 16°0 15°2 |15°0/14°6(14°6) 14:7 | 14°5 | 14°4) 14°3 | 14°2|14°2|14°2/14°3/14°2 14°7 Pepe ts 41) $o°3) 13°2) 13-71 | 13°0)12°9 | 12-9) 13°01 13-0) 138°1 | 138°1 | 138°1 1376 Peewee bia o> | 13°2)13°1/13°0)} 13:0; 12-9/ 18°0/ 18°0/ 18-1) 18°11) 13°0 12°8 14°7 | 14°5/ 14-4) 14°5 | 14°5|14°5) 14-4) 14°3 | 14°38 | 14°38 | 14:4.) 14°3 | 14°3 14°5 | Summer. , / , vi / / / / , / , / / , 16°3 | 16°0/15°7|15°4|15°4| 15°5 | 15°1)15°3 | 15°38) 15°2| 15°2 | 15°3 | 15-2 16°4 15°6 |15°1)14°9/}15°0/15°1 | 15°1)15°0)14°7 | 14°6|14°7| 14°7 | 14°7 | 14°7 15 °2 15°1 | 14°9| 14°8 | 14°8 | 14°5 | 14°4/14°1/13°9|138°9| 14-01 14°0| 14-0 | 14°2 14°9 15°1 | 15-0} 14°8) 14°56 | 14°7 | 14°6 | 14-6 | 14-4 | 14°3) 14°38 | 14-4 | 14°5 | 14°5 15°1 14°9 | 14°7 | 14°5 | 14°5 | 14°6 | 14-4.) 14-3 |} 13°8|13°7 | 1377 | 13°91 18°8)13°9 14:°9 15°2 | 14°8|14°5 | 14:°4 | 14°4 | 14°4) 14°83 | 142) 14°3 | 14°4)14°3 | 14-3 | 14:°3 15 °4 15°4 | 15°1| 14°9 |-14°8 14°8|14°7 14.°6 | 14°4| 14:°3 144)| 14-4 14°4] 14°5 15°3 Inclination as deduced from Table VII. Noon Summer Means. / / , / , / / , / / , / / +0°41+0°1 |—0-1 |—0-2 |-0-2 |-0-2 |-0-4 |—0-6 |—0°6 |—0°6 |—0°5 |—0'5 |—0°5 Winter Means. , / , , , / / / , / / / U +0°3 |} 0°0} 0°70} 0°0/+0°1 | 0:0 /—0°1 |—0°1 |—O°1 |—O°'1 |—O°1 |—G°l |—0°2 Annual Means. , , , / , jas / / / r) , 1 , +0°4 |+0°1 |—0°1 |—0'1 | 0°0 |—0°1 |—0°2 |—0°3 |—0°4 |—0°3 |—0°3 |—0°3 |—0°8 the reading is above the mean. _| Falmouth 5364 Report of the Kew Observatory Comnuttee. APPENDIX AS Mean VAuues, for the years specified, of the Magnetic Elements at Observatories whose Publications are received at Kew Observatory. Place. ee Pawlowsk . Katharinenburg Kasam .... .% Copenhagen ... Stonyhurst .... Pambure, «6. Wilhelmshaven Potsdam ... ss Trkutsk., ... os Winecht .. <7. WRC tows «'«. ole oi= Greenwich*.... Uccle (Brussels) PPA 2. = aie St. Helier (Jer er- sey) - Pare St. “Maur Ghats) 5... WACK 55246 .. O’ Gyalla(Pesth) Odessaic. cs..0 0: Pols ie os DAC ET, oo aijoiaie'e e's Roronto. .. -*'.. Perpignan..... HOME... 36 e+e. UGH Pore es oo: 5 6 Capodimonte (Naples) .... WETTIG ee ss os WOUND. soso os * Observations taken on site of new magnetic pavilion. Latitude. TZN needles alone employed. + In last year’s table the Declination at Nice should be 12° 12°8’ (not 12° 18'8’). 5 0 0 4 21 E. 5 4 25 E. 2 2 29 K. 21 EH. 12 EH. 46 E. 51 EK. 7 164. 79 2 53 EH. 12 27 EK. 44 48 B. 14,15 E. Longitude. 5 W. 5 W. 30 W. 3 40W. 8 25W. Year. 1897 1897 1892 1898 .| 1898 1896 1898| 1899 1898 1897 1897 .| 1899 1898 1898 1898 1898 1899 1897 1898 1897 {1308 1899 1897 1898 1898 1899 1897 1896 1891 1896 1897 1896 1897 1897 1895 | Declination. TH HWOONMOSDSDHHA ~T BNNBNWENACOUYUSBDONA {4245 PASSeeaaeeaes = SAARSAP AAAS SAREE 16d 15 56° if vs2- U7 20 6 09 6 7 Inclination. DOOVWOKRMWORDAONON that Sh So a i al thf a oy) bo ew) Plea led D ro) a EOOo SIH © on On tN Or fap) (Je) he is 59 3 59 33° Hori- zontal Force. C.G.S. Units. 16514 17812 *18551 “17467 17260 *18061 "18045 "18072 18794 °20145 °18511 *18393 18387 "18930 *18627 *19906 19717 20797 *21114 °21114 *21129 -22039 oA "22349 *22390 *16650 "223898 "2324 °25670 "24075 22658 22691 | Vertical Force C.G.S8. Units. "42372 “39065 * *39087 *38948 "3730 387775 36406 “38628 38613 In case of Inclination 3-inch Report of the Kew Observatory Committee. 365 APPENDIX JA—continued. | Hori Vertical zontal F | Place. Latitude. | Longitude.| Year. | Declination. | Inclination. | Force. penis | . ; C.G.8. Oe | Units. ae Washington ..| 3855N.| 77 4W1894| 3 39-9 W.| 70 34-3 N. | -19979 | -56646 Wasoen........| 38 43 N. 9 9W.) 1899 | 17 22°6 W.| 57 58°4 N. | °23451 | °37484 Zi-ka-wei .....| 31 12 N. | 121 26E.| 1896 2 18°1 W.| 45 52°7 N. | °32676 | +33693 Barnna..5....) 238 8 N. 82 25 W.| 1898 38 10°8 E. | 52 30°7.N. | °381166 | °40634 | Hong Kong....| 22 18 N. | 114 10E.| 1898 | 0 22-6 E. | 31 33°3 N. | *36607 | :22481 ) Tacubaya...... / 19 24 N. 99 12EH.| 1895 7 45°6 E. | 44 22-2 N. | °33428 | -32764 | Colaba(Bombay) 18 54N. 72 49 E.| 1896 0 33°8 E. | 20 55°6 N. | °37463 | °14326 Manila........| 14 35 N. | 120 58E.| 1897 0 51°4E. | 16 33°2 N. | °37910 | °11268 Batavia . 6 11S. | 106 49E.| 1897 | 118-6. | 29 37-8 S. | °86767 | +20913 Macias Vaden | 20, 6.8. 57 33 E.| 1897 9 43°6 W.| 54 27°4S. *23900 | *383452 Melbourne. ....|-37 50S. | 144 58 E.| 1898 | 8 20'1E. | 67 22°48. | °233641] °56050 ‘G68 10} ,, suvopT ATMO ,, Jo oWMNIOA oY} UT NOTQRoITQnd tox popuozUr sonTeA UMOIT OOTYO [BS0TOLODjo py O44 4v poptdutoo useq ser oTqQu} sITTT, “TSW 1 .ZE 0} poonpay » Report of the Kew Observatory Committee. 366 ee ee eeee ee @eee ee e@oee ee SUBOTA, 68z 000-08 4-09 | ¥-8h| 6-19] 9.09 1 reo Solo | wae 6c | ObPee) © OL & | 829-081 01662) ~ © 9 PI) S12 | “ 6 9 | 0OF8 1'8.98 |-P.28| LTP) O48 |" 09 082- | ‘WV 8 068-62 | ‘WV OTL ZT | 622-08! 86T-0& ‘ L O€| 8.22 See ep 8:09 | O.2b | F-IP| $-2S| ep) °° “aon C66. | Way T TPI-62 | “WIG TZ | Shh-08| FL0-08 “ L 1 0.18 come ae 6-29 | L-6P | S-IP| 9-99] 9.87) °°°*"999 PVE. | WV OF | G9B-6z L I | 662-08 | 798.62 = G62 z.ce Se" te o-P8 | 0-85 | 8-6F| T-99] 8.Z¢| °° *-9deg pep | McIN 18 | OF4.62) "NVS.OT | €9F-08| 960-08) “ ¢ 82) 28h | “ € ST | 8-48 | 1-99 | 99] 8-94) 4.99 | "°° “Sny 61> | ‘WVILT 82¢-62| “WA OL TE | FLF-0€! 8210-08 ‘ ’ G!| .0S Hig. iz | 7-98 | Z-99 | |_0-2¢ 17.62 | 1-99) =e Amie poe. |‘mw'ag oc | tee.62| “ ots | 2FF-08! 290.08 { 7 a | ogy wap ¢ | 9.08 | 6.09 | 0.1¢| 404] 6.09|**:eung (a3 Oz. | « oe \etr.cz “ TL 82 | ssr.-0g! 9z0.0¢| “ ¢ F| gre lwacypte | etg | 9.19 | e-er| 66¢| ere|°:: Sem Tag. | “_T PL | ¢88-8c] “ T €% | 8ee.0€| pe8-62| “ 9 81] ace | “ & T | 889 | 4Lh | ZIP] 2-69) F-2h) °° Indy c6Il. | KVP 6 | 610-62 OL T | 2489.08} 460.08] “ 4 12) && “ € T&| 869 | %0F | $88] T-8P| 6.68 | ° He] Olga a J et | TAL6a| MVE -8Z | 6-08) 606-62 SS eeecGic 2 — BOL | GO 56.1 8-98) OE | Onl eee can Gcé. | “NY 8 6 18.86 | “Wd6 GS | 669.08 | 3h8-62| “NVZ 9 8-86 | WaT 1 Poo | 9.2b | 6.28 e-Lb| 8.aF | caer “ul SPS | - “SUL UP | ‘sur } ‘sur 7 Pp T 'P "66ST “UIT ia te " eT @ fl “UIT aqeql “xe *04bq UIT aD eT @ | “XUT | pue | uy | xe | g } -uorsu9} “UBOTT “XVI B = -inodea a UGS ‘soUOT]XA OyNTOsg VW “soULoA4XH OyNlOsqy —Jo suveqy x LOJOULOL’ | *I9}9TLOULLOT, J, “6681 ‘AIOJVATOSYO) MOY ‘“oInsserg puv singerodmoey, Jo syqnsey A[qQuoyy, wees ‘T "LI XIGNHddV ~ Ne) ae) Report of the Kew Observatory Committee. ‘MOY UB SeTTUL G popoddxe Jou sey stnoy «noj-AqU9My OG} LOZ A7LIOTAA PUIM UeaUT OT} ,, WI[BO,, B UT || “Inoj-44U0M} OY} JO INOY OUO 4svOT 4B UL IMoY Ue soTtUL Ge popsd0xe sey AZLI0TAA PUIM UBIUT OY} ,,9[vd ,, B UT § “‘popiodat SVM MOUS PozTOT IO ULeA YOU [9.0 Yor uo osoy} ore sup Aurea Jo coquinu oy, ff ‘ydearsourour oy} Aq poroqstsor sy 4 | 69 | 92 | 9F oz | 2¢ ‘punois oaoqe oof GZ.T osnes Aq ATIep ‘WV OT FB pomnsvoy x COnt Ps re | 68 |, FL Get 99 L if 8 C&T 8 G € G € Polk € i Zameen 0G v es ee T GT Fl Lehane ease 0 | T Glee shh elseere |) on nas dM tk (ER Seale GL ‘G G 9 P sera ONE Vv § ae 6 6 ra Be wiaasl le P § GL Z € sa tara] T € a 8 S G poe SEN NAL I § G G : G GL € I aa Vv Z we 2 paral fe G G 8 c v peat T G wy 6 9 g 28 SG v € € if G ee V AE he as, 6 8 it ps aig — ae G 8 ¢ T i 9 G if 6 oF I mee Tee V G 8 6 (f T T ie ait aes oe 9T I ss oe G 0G 9 c 9 L I I 9 1 i iE 8 L = Se § OL Z G I 4 9 i 8 G T g OT OT ae Ef I 6 Ve Va G Or | 9 5 ae L T G val L ra T T 61 2 ae - “kys |, fxg | SUEtOI8S if B | AACN] “AA MAS) “CS [OS | “A | ON | ON ey (|, 1880 maerS “top | [IBA MoUs) Urey at | cor | AO UN], O10ISTS0L OOM TOM SUM JI WOIyM UO skep joaoquinyy L'purAA | = aie ie "yequin yy autee At “£IOJVAIOSYQ MOY | ‘IT 91942 T—'swoyyeasesqg [BOLso[O.10940TT GP8.0Z 092-0 | SSz.1 OSZ-T ae Ge ore GGO.T | SL6-T 006-0 | G9T-Z O6T-0 | SPP-0 OFT-O | O19-0 0ZS.0 | OT9-T CPP.O | SOV-T 028-0 | G48. OFT-0 | 099-0 088.0 | 020.2 OZF-.0 | S8E-Z “SUT “SUL ‘Tan anal AXCY [240], x [[BFULe YY 0-9 g.2 0-8 6.7 0-9 GP 9. 6G 6-9 G4 PG . VG v9 “(aseo -L9AO=(0T ‘1e9[0=0) pnoyo jo quUNOUB UvOTA, “suBoIm 8[®I0 7, ee "** aquis00q pue ** * Taq UIOAO NT * 99+ gaqo,00 ** -raquieydog eeeee qsnsny eee sere At reeeees oon reece ee err eee eevee [udy ress ee TOdeTty ree Kengo cores KTBNUB LP “668T *sTUOTL ‘ep oures 09 podoqud pus “N'Y OT 9% pwoy T -Aep snoradid 09 poroyue pur “HV OL ye peoty 4 ‘pasn Surleq ¢ 10,0Vz [VUISITO oy ‘puNodd OYY Jo oovjANS [Boles OYy EAOGE 4003 OL ‘qdeisomour 8, uosuIqoy v fq pozworpul SV » *{( \ o - OGRE Paes Ks BGaanl# % A RG sual aaa Lg Of GOLT [°° *' ttt + suvoy_ pus sTeqoy, 66 OF 0-6 at 9 ec ek %8 gc. | rs 96 9 £1 0€ OV nee ve eee PES OO CE 8 & _ bP 6-6 8T 1Z rE | & 96 oF) 8T 8 61 9¢ TS ‘eee ee geen # ee GUIOAO NT T S ST VE c.L PL GG cé | & 6IL 68 €T ZV 6 1g Sh SOT teen nee seis OF0G) 3S ra 9% F6 | 63_| 8 pp |S S61 SIL-| SI ZI aig @ LOT [*c°ttteteeeees goquiegdeg a is Qe econ eee ol ared ere am giecere Ts | et et | Sey) ap aoe jose Oe renee gen 3 i 9% age eli lamer pe Uh hese Jet ten |-0ee4|.08-pT gg BpspOe. a Mecsas ene sees ee = a VG Gop aigi || 6 ep a Geto.) Her St | 9 ET Tg Q 2Ge |Oe Seat cesta tee oun 8 LT Ge 8-01 | > 0% Ge (leer \ Cele snl) te ||. 79 PL OF ZL GIS [tires tsetse se tes Leg = OFS ia oe 6 IL 61 0G ce a SIL 90T | 6L VS GL Tg (eg kt Megoer regu ee) Sty S rs GGr | «7 LE T-6 GG 6 6% 64 OTT 06 Go | «FS 6 && CIOL Tee See oe eae = €I OF Fo eb0e | et 62 Ore ecole | 1678 | ce a ey 2 63 PG-OG —[turreteree esse Comagay = 16 8P 6-VI | $2 8T T& | G6 06 69 | 6 8st 9 LG 8p OL siege tgesseiae ie ® seals Arama Ke "sOTIUL | “SO[TUL ‘sop | ‘dap ‘Dp | “sop ‘um ‘Y Ur ery ‘6681 | ——-— Bee eal ky ul ree ae ee We pee |e eel ee a = | QUITEDUE| yaprogas ~~ “Aqwoyea | Agto0poa | 4 ‘paiooar | ofqtssod smoy = ‘ayeq Ayanoy | Ajanoy |ozeq|4semorT | ‘avoyy 99% ‘qsoysIpT| Uvoyy |oyvq| Ayrep jo ee 70 zoquunu S | qsoqtety | ose.10A VW qsoywoty | -wedre OR S, Uv RS “STFU JT F ‘onona UL G{nq yoVTq) | x IIB OY} JO punors otf} UO eAny ( ‘sea § UNS UT 9In4 ‘QUIYSUNG JYsIAg | JUSWOAOUL [VIUOZMOFT | -CAoduIe, WNUIUTL | yroduroy wn UIXE -£1098A1084(0 MOY 368 ‘ITI °19%,,—'stonearosq(¢ [VoLdoTo100}07q Report of the Kew Observatory Comnuittee. 369 APPENDIX III.—Table I. Register of principal Seismograph Disturbances. 1899. No. in Commence- ha | Spal ie Ke Dat eaten Duration | Ist ee ia in duration oe = ePT >, % | of P.T.’s.*| maximum. | maximum.| seconds of | of disturb- register. id aew ee | te gees sa. | m. ae) ha. hy ta, he 7a 49 Jan. 14 2 58 ‘2 27 °2 3 26°5 3 28°2 1-03 pe 50 oe 22 Suszse 7). ) 5G 8 29°1 — 0°77 O 27°5 52 » 24-25 | 28 47°7 | 43°4 0 35°6 0 42°6 2°44, 2 59-6. 89 meets 17 47-2. | 296 18 35:1 — 0°46 1 44°6 112 dune 5/]-1515°4 |; 30:0 15 46°2 | 15 55°2 0°58 1 LEG 114 » 14 11 28 °6 22 °5 EE 52 °Q) | 11. 59-2 E90 2 9°4 124 July 12 1 55°3 10°8 2 12°3 | 2 13°3 1°12 1 4:0 125 pen la |. 15. 31°4 21-9 13 54°0 | 13 56°8 1°60 3 35°6 142 Sept. 4 0 33 °6 8°3 E's 6 y's 7°49 2 49 °2 144. eee) 17 15 °3 7°3 LG 50°78 | 0F 53°5 2°18 1 39:0 exceeded 145 (es es a 58 ‘9 22 20-21 | 22 25°6 10°80 3 00 149 ee 0) 2 16°7 4°6 2 21°8 2 27°5 3°20 1 22°8 150 5 22] 112873 |. 20°2 11 46°7 | 11 49°8 0:70 1 19°2 151 mere) 14 8 |: | 190 14 24°3 | 14 26°3 0-70 1 12°4 152 eee We LT 23 °2 5°3 V7 aoe | VT 4a] 0°51 2 13°8 168 Nov, 23) 10 1:0 9°8 1O410°S | 10 TEs 1040 ob Se Ae 169 ” 24 19 5°5 ED <1 19 41 °4 19 43 °4 0°69 | 0 59°5 179 Dee ai 1059-2 . S-o cp ls — 0°88 0 53°3 180 Pe are 20 87-Le 18-7 20) a2) ob 27 O 57:7 0°50 | _* PTs = preliminary tremors. The times recorded are G.M.T.; midnight = 0 or 24 hours. The figures given above are obtained from the photographic records of a Milne Horizontal Pendulum; they represent E—W displacements. Report of the Kew Observatory Committee. 82+ 9-8+ F-O+ 9-I+ ¥-0+ sececgeccaecnagnoore TOSI ob borg “ors ereg 20716 srevoeocsovovoevee KTATIAOD ‘SMoU}IBI L-$8 | PLL | Ge) 18; $F | 70-0] F-0 F-F8 @.GI Ta 9¢ 0. gg 1.9 10-0 ¥-0 0-G+ 6-T+ Beot Pe ot 9-1+ sen eneeeccaeceereosoevenscvccnccorseese ered are erg LOPTI Soeoearerevecce AIqUeA0g ‘JopuBIpLi 9.79 | LSE | 9-78 | &-1e| 0-9 | 20-0 | 0 | O-G+| SI+] LO+| SOF] BO-+[ tr fesmEEy ,, 0's “a's “xg | S8TOST jets Arquaa0D ‘TTH “f “SeuD 8 | HBL) 8-2 | ¢-82] 2-8 | 20-0] 9:0] HEt| 8 F+) SFT) OFT! Det” LASNATEY 5, “ow eqs “ag | LOOTET |" Aaqgueaog “OD a ouOsMaNT Perl POE | Gc 9Gck @ GE | G8 1280-0 | G20 | G-O—) 0-8) 6.9] GG—|6-F-y ne TOSME NL ys Sos Qs org | LLeGs [eet ttesss Arguaaoy ‘repuslpiig 6-%8 | 6ST | 0-92 | 0-e8 | 2 | 90-0} €-0] 3+) 6-O+| GO-| T-I—-} G-0+ reosereeeseoerecoueeo de souolya ‘0's “q°scag | 6601 “= BAQUOY “4.1OqOY-MOpuLIuOTY Mee GOT | 0.80 2.06 | Gap G0-0.| 6:01. 18C= (Spa |G ral GO| Oba) TSNS “ogegs org | LOIGG [ttt AIQMoAOD ‘UesyTUUOg 0-.S8 6-81 | 6-98 | 3- HE 1-9 | 60-0 @.0 | G-O-| &-IT+] S-T+| G-3+| 6-04 reeeessseeersavenererstecesssonerseee song GQeg Greg GP99e "+ £WUOAOD ‘WOY * aT A ‘fC Z- G8 L. 9T Se Le €§-1€ 0. G0. 0 F-.0 8.[- L-0+ ¥-1I—| I-I—]| 8 i= ceesercceresoeeoeerse _, OSNAIey e borg “a3 Gre LELLE seececeeccee £14U2A09 ‘SMOUIICTL “M 7.98 | ¢6r | e-9¢ | 9-62 | ¢-¢/] 10-0] #0] Le+| ZL+| &F+| I-94] G+]. [sMMIeY,, oeuseUM-TOU “o's "Q"B “4's een “terre TOPUOTT “WOS I YTS *§ c. gs I- LI y- 98 0. os 0-2 v0-0 P-0 TZt 8. O= G.0-— 8-0- 8.0o+ PeRTIUE LE) ‘o's easorg ZLOPE seroeeree KTITIOAOD ‘aog 2” oqtU M ‘p 9.48 | 2-91 | G92 17-08 | &F | $0.0] &-0 | 20+] €-0—| G0—| 0-0—| €-O— |" TSSMtLey “ors “qrd “ig | 99696 "* AIGUOAOD ‘SMOTIFEHL “MA 3.68 | 1-91 | 6-Ge | 2-@ | &F | 90-0} 8-0 | 20+] F-0—-| G:3—-| ¥-0F] G.T—]"""" TeQOUTOMOIYS ,, WOLIGINOL >, “ors “q'y | I-668T |" UOpuOT ‘uOg wp YWTUIG “gs 8-¢8 6-91 | GS-98 | F-2E &:F | $0-0 ¥-0 | 8 St] Get] 6 I+] 2-04) 1-64 setaeeeneenneeeeererasense seeremeeren tong Garg Gye 0901 "=" eAguexy “4.1eqoy-UOpuL}uOy, 0- 98 6- 61 8. 98 g- 66 cP 10-0 c.0 g. O+ T- o+ G.0— 0-2- 0- I- CO ower reer HF OSErOF OSE EH SFR OOS SOHDE1OCD Wonrs} “q's “US PLESLZ eee WVY SUT “nary ‘apaleylyy 0-98 | [Ot | S281 9.2e | 0-¢ | 90-0 | F-0 | 8:0—) G0) O- T=) T<8—) ot revererrereers Touvadar aynurm “o's “qa “ig | G60T |'**'* Vaauey “410qoy-copuRyuoN G.98 9-91 9.1¢ | ¢.2e z-% | G0-0 $0 | #E—| 6 I-| 2I-—| ¢-1I-| 81I- seasnaaneceesensecm== TOSMILGSL y5 borg os crs LG99) ve seereessoes KINTQAOD ‘suemoa x “sg @yeer elects) G.7e" | G.e| Zak | POs0' | GO | LS) See) esOa ee) OOS eshte, “ors‘q'd rg | pee, |" eas “ as 9. 98 ¥-9L }. cg Q. +29 L-@ c0.0 °.0 9.2+ ZG €.0— Z-0- 70+ see cewererarccse eos eenererescssscneseves seg “q's cNTaG OFSFI se cteeesenscoen AIGUdAOD ‘JepuelpIT 2-98 | GLI | &-22 | 6-18 | ZG | 70-0 | 0 | 0-E+| GOF| OT+] T-T+] 8 I+]/"" LOA0T , WOTMTAIMOL 5, “ors ‘gasng “aq | L980 | Uopuo’y ‘OD = UvYspory ‘seyD 6-98 @.61 | 6-LE | L-6z 2: | 10-0 Gl | Geek kOnea | Laer | OnGetele Gs Sma oe, ce LESTE,» bors “qed “aig BGIO) |rrcttttttesss ATgUaAOD ‘suUBULOa “Ss Parcel apc aseerGueen| 2.64) 90-0 |) GeO) PaG=—|) Qe = |e@nuliee| Ou Dee(sOsliee| oe GIPRLMIB DL Sos eqs ad | ZIIL | vaouey 4toqoy-copueyuoyy Perec TeGenicGeTG)| O.p |H0h 0) | uO Olt GsOmr | Calc aiGseleG.Qce| 18 lOSThte ys, Sos *qrs “arg | gGggL [ttt * *ATjeAOD ‘suBUIOeX *S ee 88 c. OL c. ge Ze ge C.F GQ. n) e.0 9. T = 0: Z- 6-2 6: I- c. I- se deeseneosene weaceesrsocesesecceeness #1) 0K “q:a “rg SLOT eevee BADUOX) ‘1 1oqOY-UOpuULIUO}T 1.88 | 620 | Zee | 9-28 | 0-G | £0-0 | FO | BO-| T-I=| F-0--) 0-0—| GOH] LOAST ,, UOTTIGMOL, “o's ‘gasny “aq | 66-8es [TT UOpuOT ‘Mog P YYTUIG “Ss "so0S | “SOOS | “Sods | “SOAS | “SOVS | “SOAS “soos | ‘SOOS or—0/0e—o|or—0/0V—-0) | Bl ae] Bie) 2 | |S 3 a go | BR] op | ee = = = = = Se@|ene|ts | BS] yb | PB a S © © iS) o) ley" 2 =. aren See Sesame Big. S 4 ° 3 B B 3 as 5B D eq m5 2 = ct ct ot Beles ji a|e.e| §) te) & Bis Sola = Ca") ~ a 2 es (re) lo} 5 ‘sre | & Soleo eh eS aa ct. & 2, : yoyem SeIBT : 5 oe Sol o 8 st ‘om ‘Sulids soured ‘ueutedvosy jo fq poqisodap qove g)g8| fifi | 2) & ie B| Ba ee lateo o 5. sees sa alt et Sa 5 “ 5 ‘ayer Ajrep ULdTT IO} POpIVAL SPLEP De Ne en ed eS -ceok oyy SULAMp SYTVUL FO LOG UNU FSOYSIY OY} POULeIGO TOIT SoToF AA OG OY} JOOOUVULAOJAeg “STVIU, HOLV AA 40 SLINSAY ‘T 9% L— AT XTC NWddV a a rn t= 9 DD OWMWWOMwrAMmMNHNOOWIOM iO IDODAMDHDIOM NAH GO OO ee | 1d _-—_ es Report of the Kew Observatory Committee. “motyesued -W109 aimjeiodua J, ¥-1E 8-€& I-Gé L-GE 9. Gs ¢- FS 6: 8& L. Gg F- GE 8- GE G- LE 6: FE 1. 8§ 6: 9E 1. G§ €- 88 0- LE 6- 8E §- €& I-86 8. CE 6- LE 6+ GE ¥- 68 9-9€ *moTjIsod Jo asueyo UIA 9VL Jo osueyO L- 82 4-08 G-1& 9-T& 8-62 9-1¢ 0. 62 ¢. 62 L-82 6-82 &- 12 1-38 &- 82 §- 0& 6- LZ T- 62 &- 0€ 8-18 ¥-&& €- 18 §- 66 L.0& &- 82 L-1€ 0Z—0 | 0F—0 | 0F—0 “oyel jo uoTjeLIeA S[Ieq 10} POpAVMe SHIE]y Cia Ml ee eae S Oy em Lert cee Nene ie wD HID O16 1D OOO 1019 We O . DMDWONnNOoOrWoOAIIDOH 1900 —_ HH 19 WOO HO GmorNdnoonrannr be [77 o *$0481I SUISO] pues SuluIes OULAIJX9 WIDMJOG BOUALOYI 90-0 £0-0 10-0 80-0 ¢0- 0 90-0 L0-0 40-0 60-0 £0-0 £0-0 90-0 - 90-0 ¢0-0 10-0 90-0 90-0 II-0 ¢0-0 60-0 60:0 90-0 G0-0 10-0 10-0 “S008 “A of IO} 9JVI JO oSuLYO UIT ° . . ry OWO OM O OH OHI ID © MONDO ONMO DWN WHIDc BROOSCSOSSSOSS 0 SGOSCOSSGOOCSDOSCC0SO o °o4yel + Ajrep JO WoNeLea ueeTT 1-0- ¢-0- e.6+ T-o+ 6-G+ 8-I+ Sor vPt 8-64 9-T— G- I+ E+ 8-ot 6-e+ 0-Gt 9-64 9-0—- Lit 8-E+ 8-0+ FeI+ 9-0—- 8-T+ 9-0— 9-o+ “soos “MMOD [BI a Oe) eu et lee alel lara ON HOOD Nn OSS FHttt+e+ee it ONO OHHH ON HO *gJo| JUBpUEg “4311 JUepUdg ‘ayer Alrep vey ‘oj SUiso, I0f — I-0— see eereccccseerveaseseaenesecsveeseoees eran vad “rg #1G9 eeocecoce uopuoT *00 ” uesuuLyo Lr 20+ Creer aeerooovacces ydvisouo1yo ’org “qa aan Or Gf 881FSZ SOOO oor 1 uopuo’7y “09 » euNnvg 6-G+ seeeceeseerenecceets*/ TONMRLE + Sorg ‘qa LN OS trace ecvcceves 66 6b 66 i g.Z+ ceeeaeoencvcrreneseerecsseccereeesssees 01) ec vas “rg sgsce weococsss 4x4usa0p ‘TOS ” O71 AA am 0b SI a et OC SECON ORCS O YF “q's rg E8Ch8Z sp wey surwlg “pry “preyIUuy g. I- to eveeereeeneveveeaseeesecnsssesnsenses ereg Wo 3 Sore a eglee *reeeeeee KAT OAOD ‘tog mY) OqTU AA L Te THpceerereeeereres rasnnieyy ,, “o's “q's “rg | ELTIQG fut: Arquaaog ‘491MoHT UYOLr 8. Gt[ errs odersouoiyo ‘o's “q's “rq | OFEG9T |" UopuoT ‘0D wp ‘uog ‘1aynTIS [2 (9 Ja bP EOISOECOOO ECD O00 GRICE a ALCOLOC ‘o's &q: Bg ber g 809ez cee erevesose Arquaao0g ‘sureyroujoy 0= © Seevereeeecererewoussvcaesrecneoseses eye igi: 3 Gere g 92898 sreceeeeecoecesoes KIATQAOD “SMOTIIBIA 0. a+ |eeerseees weoreeerees a Jesnarey ,, &org Sg (| ire ZV 261 seeereoeeeeTTODUO'] ‘u0g AY) YjIWS ‘S Te I- SPP T COE Deer se rereceooaseressescens See rec ‘aq: 8 bere a ZL01 eeeece BAQUNXD) “19qoy -uopueyUO, a e+ Ug ailel seb Cy [OSULIv i so: Ss acl 3 Coys g 981ZZ% OOOO A14U9A09 ‘u0JSUIPpB A “ay Ze Te HESeReRseic eae ses'esd [OSNAIe yy aa bo: s Sq: 3 Rens S 88908 eooeeseencee AIqgUaA0g ‘SMOUIICT “M 7.¢+|oneusvu-uou ,,‘jesnaiey,, ‘0's “q's “0g ee strewesesore TOPUO'T ‘UO Zp YIUIS “S 0-Z+ Toe eeeweroesesevevaesererssvegrereseres ere “a's rg Z866% soeneereroeseer tO DUO], bf) (0@) ~ I9YIsp T-o+ [Sem Pere eeeeeserecveserseessereesseeenes sey “a3 org OZFER sereeeees KTQUBAOD ‘0g mY) Ou MA a Sit». WOIqinoy,, Iunuus ‘o's “qa “ad | gGgL frewesttereeees UOpUOT “L0TABT, “V ¢.o+ Bare sainecier sos esos eer OST MLE NT 5 borg “qd erg 08ccz Coerercseresves AMUdADD ‘LOpuUvpPIAT, Qa Ql t eeeereeerseeres TMM LIVI 5 borg “qs “Ig OTM Z6L | wopuo'T ‘U0. ap YIIUIS *g GU Efe been cereee rereeeoee 1 RTIT UE ues OnSeeaCa dg aetna | Mea G CC om pee neaearnc Asquoa0y ‘LOpuUvTpPLLA 9.2— PR OO CCR [OSNIIVy, ‘a Sorg “as Lies z2eece eveescere £IGUIAOD © dos WY) OT A “cf 0.F+ steeeeeecoseesoeooees TORNLLIGY a borg “a3 ‘1g ZPLOe eer eeeoresesces Atos syeeHsUkeneseeseosees THEN IIEY 55 borg “qd PTS 8eL9E seeceeveneesoervoe KTAUDAOD 'SMOUICIT TO [teers rosnaay,, o's “q's wag | ggceg. [reese £uguaaog ‘topuElpLiT *soas a] oO B 5 =] cr i=] Ss ‘YOICA ‘om ‘Butids oouvjeq ‘juowodvosy jo fq poqisodop yore Ay 1aquinN $aqe1 SUlUTES Io} + { (10010A0 O[SuI9 oJ ‘o's $po1IVq Buro9 10} °q'S $19T[O1 9[QNOp log *a°p $A9[[OI O[FuIs Oy *s—: *ZIA ‘pasn ore suonetaarqqe SUIMOTLOJ O49 4ST] DAOGR at] UT “panuyuor—T 2[QeBy, 372 Report of the Kew Observatory Committee. (O0T=s0 “syareut T®}O], ee So en ee | e eae noe on eee mano 119 1g He | BOND teak imal fo oe ~ 4 0c—0 “oIny -elod ura J, 6-98 1: 62 @. ge 6. LZ 1- G8 1. 8Z g. ee F- 08 8.22 @. 8 P-ce $- 63 G. ge P- O& 1-G2 G. 63 0.98 0.82 T- L£2 CG. FZ T- 8 Gige 0. Fe G. GZ 2.08 9. LT 9. 6Z 6-&Z OF—O OF—0 "MOTJISO Eee eed -VLIB A LOF POpIEMe SyV]y ee See er a ce ce ‘+ uopuoy “g “9 Laeyr p Lamy | 999 Ce, u “ « ** AtqUeAOD ‘uog pu rokeTg "fF | SZ66Z Henig Sie Sere rs af is eee veee KT9UOAOD ‘suvyLoyyOY E9TZ e222 60808 Be oe 73 6¢ 66 "****s uopuoTy “og pus ouneg | ggTPez seeeceescees udegisouoryo spuodes puv oynUITY, ee ArquUaArog ‘MOG pue O9It AA “f zZGegg eee eceoee 4c 9 (43 UOPUOT “YOoT[NHITT PUe SoTBAA | EFE pag f “ “ ee ee ee uopuo'T “O*d pue oumneg 609ZT eoeevececee (<3 ¢ (74 uopuo'T “O09 pure ‘OG *1OIINVIG OFE69L eoereeeee (79 66 (9 aus BADUIK) ‘q1aqoy-WOpuByuoy 660L Pabee ye cess ese ydeascuoxzyo Spuooes quyds pue oynuUrypy eeoee (73 6¢ G-€CT eseeree “cc ce (73 eevee wuopuoryT Og pue yy zs) P-kCl eeereee (19 (73 (73 Se a cae uopuo'y ‘uosuog. “MM. ‘f TGEGS peo en Iayeodar oFNUIU put ydeadouo.1yo oynuyy pele - 2 6ST8I (oye yooyo pur) 66 (73 ¢¢ (a4 eooees ee 08 080 08 OF cE ee dec tT CY Se OE KOLO NG Dt oy} jo saseyd YIM ‘aepueyeo pengedaod pue ‘royvodaa ‘Yydeasouoayo spuoses 41]ds pue oqnuIpy "*** gopuoTy ‘nog pue YANG “S| GE-SFT £q poytsodeg “1aquin NT “TPOVEM jo uorydrxosoqy ‘rvoX oy} SulINp soyoze AA pozeorjdurog Aq pourezqo syrepy ysoyorpyy TT S98 L or) I~ oO Report of the Kew Observatory. Committee. €- 08 9. ST €-4& F- 08 6-08 L- PL 6: TE 0-¢& T- 8 c: OF L.G€ 6-16 P- C8 ¢- 61 €. 9& 9°66 v9L 0-61 9-T& 8- 66 9-LL 6: GL 6.86 Gg. 86 0:98 T- 91 £16 9.36 ‘OOL-0 06-0 Ov—O Cv—O0 *OINy *u014 “SyaAvua 7 “HONISOg 12I0T, vaso uo J, “CI A 1OJ POpATMs SYACVPAL "**-TOpuOrT ‘UOg puv YUL ‘g reeeesos KIQUDAOZD “LOPUL[PIL : TLSSS = *'TOpuoy GOR Pua UIUC Se, **uopuory “GQ Gg Lavy H AWAY uOpUOTT ‘POOT[NDII PUB SoTR AA "* BAMA) “YLOQOY-UOpUByUOTY | £q poysodacy ee “panurjuoo—]JT o[Gey, 66) c6L VIS&s - SiGbte-|* 808-666 OVEG PTS C6OL “LOquUN Ny 608-626 | icy La | eutp ek we eb) Sle) 6 Behe 66 fe iC Le dY cbt Cerh wd) ot cre CieCmai $ 6s CUS) (uNWore aieltenre) aay varenne eine 16) 66 ee abe eS ewer re OT tis Ue O Naay ee ee se (oqnUTW G) 6¢ (79 CRS ero ta a REN SN ct «¢ *ee@eweeeeeoe TSE co fal zfsols) ix OFQUT TAL "T]OywM JO uoTdrLOSsog ; z b> * r4 be a] 3 enone beet. lake, a ‘Me CBT4 Proceedings and List of Papers read. May 10, 1900. The LORD LISTER, F.R.C.S., _D.C.L.; President, in the Chair. A List 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 :— Burch, George James, M.A. Manson, Patrick, M.D. David, Professor T. W. Edgeworth, | Muir, Thomas, M.A. B.A. | Rambaut, Professor Arthur A., Farmer, Professor John Bretland, M.A. M.A. Sell, William James, M.A. Hill, Leonard, M.B. Spencer, Professor W. Baldwin, Horne, John, F.G.S. B.A. Lister, Joseph Jackson, M.A. _ Walker, Professor James, D.Sc.. MacGregor, Professor James | Watts, Philip. Gordon, D.Se. Wilson, Charles T. R., M.A.* The following Papers were read :— I. “On the Diffusion of Gold in Solid Lead at the Ordinary Tempera- ture.” By Sir W. C. RoBerts-AvusTEN, K.C.B., F.RS. II. “On Certain Properties of the Alloys of the Copper and Gold Series.” By Sir W. C. RoBerts-AUSTEN, K.C.B., F.R.S., and T. KirKE Rose, D.Sc. III. ‘“‘ Experiments on the Value of Vascular and Visceral Factors for the Genesis of Emotion.” By Professor C. 5. SHERRINGTON, F.R.S. IV. “On the Brightness of the Corona of April 16, 1893. Preliminary Note.” By Professor H. H. TuRNER, F.R.S. V. “Radio-activity of Uranium.” By Sir W. Crooxss, F.R.S. May 11, 1900. The LORD LISTER, F.B.C.S., D.C.L., President, in the Chair. A List of the Presents received was late on ae table, and thanks ordered for them. Electrical Conductivity in Gases traversed by Cathode Rays, 375 The following Papers were read :— I. “The Circulation of the Surface Waters of the North Atlantic Ocean.” By H. N. Dickson. Communicated by Sir JoHn Murray, F.RS. Il. “On Cerebral Anzemia and the Effects which follow Ligation of the Cerebral Arteries.” By Dr. LEoNARD Hinn. Communi- cated by Dr. Mort, F.R.S. Ilf. “The Influence of Increased Atmospheric Pressure on the Cireula- tion of the Blood. Preliminary Note.” By Dr. LEonarp Hitt. Communicated by Dr. Mort, F.R.S. IV. “Contributions to the Comparative Anatomy of the Mammalian Lye, chiefly based on Ophthalmoscopic Examination.” By Dr. G.L. JoHNsonN. Communicated by Dr. Gapow, F-.R.S. The Society adjourned over Ascension Day to Thursday, May 31. “Electrical Conductivity in Gases traversed by Cathode. Rays.” By J. C. McLennan, Demonstrator in Physics, University of Toronto. Communicated by Professor J. J. THomson, F.R.S. Received December 7, 1899,—Read February 1, 1900. (Abstract. ) The object of the experiments which are described in this paper was to investigate the nature of the conductivity produced in different gases when cathode rays of definite strength passed through them. In a series of papers,* Professors J. J. Thomson and Rutherford have recently shown that gases become conductors, when traversed — either by Rontgen or by uranium rays, owing to the production of positive and negative ions throughout their volume. In the present investigation cathode rays were found to impress a condition of the same kind upon a gas, and laws have been de- veloped which connect the absorption of these rays with the number of ions produced by them in the absorbing gases. The investigation is described under the following subdivisions :— (1) Form of tube adopted for the production of cathode rays. (2) Ionisation by cathode rays. (3) Discharging action of cathode rays. (4) Ionisation not due to Réntgen rays. = Phil) Mac,’ November, 1896, p. 393; 2bid., January, 1899, p. 109.. 2F2 376 Mr. J. C. McLennan. (5) Discussion of methods for measuring the ionisations produced in different gases. (6) Description of apparatus used. (7) Explanation of the method adopted for comparing ionisations. (8) Ionisation in different gases at the same pressure. (9) Ionisation in air at different pressures. - (10) Ionisation in a gas independent of its chemical composition. (11) Comparison of ionisations produced by cathode and by Rontgen rays. | (12) Summary of results. The tube used for the production of cathode rays was similar in form to that devised by Lenard,* but, as the brass plate carrying the aluminium window was found to act very well as an anode, the ordinary positive electrode in his apparatus was dispensed with. The paper commences with a series of experiments illustrating the conductivity produced by cathode rays, and the various phenomena met with are shown to be fully explained on the supposition that pesitive and negative ions are produced in a gas by the radiation, and that the conductivity arises from the motion of these ions under the action of an electric force. This view of the conductivity is also shown to explain the loss of charge sustained by a conductor upon which the rays fall. Lenard’s experimentst in this connection were repeated, and, contrary to his observations, negative charges were not in any case found to be com- pletely dissipated by the rays, but were reduced, at atmospheric pres- sure, to small limiting values of the order of 0°25 volt. These values were found to be slightly increased when a blast was directed so as to remove the air close to the conductor, and when the latter was placed in a vacuum, the limiting charge rapidly assumed a very high value. The value of the limiting charge was found to be affected also by the proximity of conductors whose potentials were different to that of the one upon which the rays fell. | Conductors initially unelectrified gained the limiting negative charge under the action of the rays, and positive electrifications upon con- ductors, surrounded by air at normal pressure, were completely dis- charged. The explanation offered regarding this limiting or steady state is that 1t represents a condition of equilibrium in which the electric con- vection by the rays to the conductor is just equal to the conduction by the ionised gas away from it. It has been thought by some that the ionisation under consideration may be due to Rontgen rays sent out from the aluminium window at * ‘Wied. Ann.,’ vol. 51, 1894, p. 225. t+ ‘ Wied. Ann.,’ vol. 68, 1897, p. 253. = Electrical Conductivity in Gases traversed by Cathode Rays. 377. the same time as the cathode rays. The results of experiment, how- ever, are entirely opposed to this view, and lead to the conclusion that, if any Rontgen rays are present in the cathode pencil, they must be of so weak a character that their ionising action can be neglected. A direct comparison showed the ionisation by cathode rays to be about 300 times that due to an intense Réntgen radiation. In the conductivity produced by cathode rays, the current of lew tricity does not increase in proportion to the electromotive force applied. The current, after reaching a certain critical value, becomes practically stationary and increases but little when very large increases are made in the electric field. With Rontgen or uranium radiation fields of 400 or 500 volts a centimetre have sufficed to give saturation in the case of most simple gases, but in the present investigation it was necessary to go as high as 1000 volts a centimetre before ie maximum current was reached. : In order to compare the ionisaticns in two different gases, or in the same gas under different conditions, recourse was had to the use of two ionising chambers. ‘lhe discharge tube was provided with a double cathode, and carried two aluminium windows. Two pencils of rays were obtained in this way, whose intensities were found to maintain a constant ratio, and these were used to produce the ionisations in the two chambers. | The ionisation in air, kept at a constant pressure in one of the chambers, was taken as the standard. The gases, whose ionisations were to be compared, were placed in turn in the other chamber, and their conductivities, as measured by saturation currents, were found in terms of the standard. An important result, obtained by this method with cathode rays of constant intensity, was the agreement found to exist between the ionisation in hydrogen at atmospheric pressure, and that in air at a pressure of 53mm. At these pressures the two gases had the same density, and in both cases, therefore, according to Lenard’s absorption law, the disposition of the rays, their actual intensities, and the amount of them absorbed from point to point in the ionising chamber, were precisely the same. Under these conditions the equal ionisations obtained not only form a strong confirmation of Lenard’s absorption law, but they also show that, where equal absorption of cathode rays occurs, equal ionisation is produced. In order to test the conclusion more closely experiments were made with air, hydrogen, carbon dioxide, oxygen, nitrogen, and nitrous oxide, and in all cases it was found that, when these gases were re- duced to the same density, the same ionisation was produced in them by rays of constant intensity. It follows, therefore, that an ionisation law exists exactly analogous to that of absorption, namely, that when cathode rays‘ of a given 378 Electrical Conductivity in Gases traversed by Cathode Rays. strength pass through a gas, the number of ions produced per second in 1 c.c. depends only upon the density of the gas, and is independent of its chemical composition. From the results thus obtained, the conclusion is drawn that, when cathode rays are absorbed to any extent, the positive and negative ions produced by these absorbed rays are of a definite amount, which bears a constant ratio to the quantity of the rays absorbed ; that is to say, in order to ascertain the relative ionisations produced in any two gases by cathode rays of the same intensity, it is sufficient to determine the absorbing powers of the two gases for the same rays. In other words, the coefficients of ionisation are determined when the coeffi- cients of absorption for the same gases are known. The paper then deals with the ionisation in any particular gas under varying pressures. ‘The inference is drawn that, under rays of con- stant intensity, the ionisation in a particular gas varies directly with the pressure. The very great absorption of the rays by gases at ordinary pressures prevented a direct verification of this relation ; but, as Lenard has shown, the coefficients of absorption for any particular gas to vary directly with the pressure, the conclusion seems quite justifiable in the light of the connection established between ionisation and absorption. Assuming this relation to be true, it follows at once that, if rays of constant. intensity are allowed to traverse different gases at the same pressure, the ionisations produced would be directly proportional to the densities of these gases. These numbers for the gases examined are given in column | of the appended table, while in column 2 are given the values found by Professor J. J. Thomson for the relative ionisation produced in these same gases by Réntgen rays of constant intensity. Column I. - Column II. Cee eee Tonisation by Tonisation by ; cathode rays Kontgen rays (calculated). (observed). WAIT Meitiney'e ga othe clketss ehasohgeger ee 1:00 1°00 ORViEOW veins se seen: 1°106 1°10 INGGrO mem haters ale ate Ong 0°89 Carbon dioxide ....... 1°53 1-40 ETOP OM. erie cnalats sel 0 069 0°33 Nitrous oxide.......... 1°52 1°47 _ The numbers, with the exception of those for hydrogen, present a fair agreement, and they show that although the two forms of radia- tion are so very different in many respects, still the products of their actions on the gases examined are practically the same. Electromotive Phenomena of Non-medullated Nerve. 319: “Observations on the Electromotive Phenomena of Non-medul- lated* Nerve.” By Miss 8. C. M. Sowron. Communicated by Dr. WALLER, F.R.S. Received March 8,—Read March 29, 1900. [PLATE 4,] Kiihne and Steiner in their work on the olfactory nerve of the pike,T published in 1880, showed that the laws of electromotive action in medullated nerve, as formulated by Du-Bois Reymond, held good in the case of non-medullated fibres. Having demonstrated the resting current in a suitably prepared olfactory nerve, they proceeded to com- pare the value of this current with that obtained from medullated nerves of the same fish as well as from the sciatic nerve of the frog. The result showed a much higher E.M.F. for non-medullated than for medullated nerves, their diameter being about equal; a result con- firmed by Biedermann and others working on the non-medullated nerves of Anodonta. Kiihne and Steiner found that the resting current diminished rapidly, but a new transverse section restored it to, and in some cases even augmented, its original E.M.F. i | On stimulation with induction currents, a negative variation of high value was obtained. The response was unfailing while the nerve was fresh, provided that the stimulating electrodes were not applied too near the peripheral end of the nerve where its character becomes modified. . While working at Leipzig during the early part of 1899, Professor Hering was good enough to suggest that I should use the nerve which had given such good results in the hands of Kiihne and Steiner for a further study of electromotive phenomena in non-medullated nerve, and especialty with reference to the occurrence in such fibres of the *‘ positive after-variation” which his own researchest and those of Head§ had made familiar in the case of medullated nerve. Hering had himself occasionally observed this effect in 1885, while using the olfactory nerve for experiments on electrotonus. The positive after- effect occurred most frequently on mechanical stimulation of the nerve by cutting, more rarely it followed when induction currents were used. In the present experiments, the pike used ranged from 14 to 23 kilos. * “Non-medullated” is used in its ordinary sense as applied to grey nerve, and without prejudice to the conclusions of Gad and Heymans that such nerves may be slightly myelinated, conclusions which Ambronn and Held have confirmed by means of their “ optical method.” ‘ Archiv fir Anat. und Physiol.,’ 1896, p. 210. + K.und SS. ‘ Untersuchungen des Physiol. Instituts der Universitat Heidelberg,’ Ba, 3, p. 149. tT ‘W.S. B.,’ vol. 89, 3 Abth., p. 187. § ‘ Pfltiger’s Arch.,’ vol. 40, p. 207. 380. Miss S.C. M. Sowton. Odservations on the in weight. As described by Kiihne and Steiner the E.M.F. of the resting current of the olfactory nerve was found to be very high, and. a few experiments, made by the “ opposition method ” of those authors, gave results similar to theirs; that is to say, the resting current of the non-medullated nerve overpowered that of a medullated nerve of approximately similar thickness, whether taken from the fish itself or from a frog. The galvanomecer used was a sensitive and quickly reacting instru- ment by Siemens and Halske, of Berlin, a modified form of the D’Arsonval pattern. It was used with a telescope in the usual way. Method.—To prepare the olfactory nerves: the head of the pike was cut off, the lower jaw removed, and the head fixed toa small board. The bones of the upper surface of the skull, from the brain down to the nostrils, were removed with bone forceps, care being taken that only bone was cut away, the cartilage below being left uninjured. The cartilage thus exposed forms a thin capsule over the lobes of the brain, becoming thicker towards the olfactory lobes, and forming, where these are prolonged into the olfactory nerves, a substantial sheath which encloses the nerves entirely, and through the semitransparent walls of which the nerves are just visible. To expose the nerves a fine sharp scalpel was used, with which the upper surface of the cartilaginous sheath was sliced away, great care being taken not to cut too deeply and so injure the nerves themselves. Having removed in this way their upper covering, the two grey nerves are found lying side by side in their canal: they run parallel for the greater part of their length, then fork, right and left, to enter either nostril. Before attempting to remove the nerves, it must be carefully ascertained that they are freely exposed in their entire length, with no overhanging shreds of cartilage to catch and injure them. The way being clear, the end-organ of each nerve is separated from its nostril and serves as a handle by which the nerve is lifted from its canal; a clean scissor cut then severs its central end from the olfactory lobe, and the nerve is ready for experiment* (see Plate). The central end of the nerve was led off from transverse and longi tudinal sections by brush electrodes ; their distance apart was usually 5 mm. ‘The stimulating electrodes were of platinum wire, with Hering’st extra loop to cut off unipolar effects. The induction coil was supplied by a single Daniell cell, the distance between primary and secondary coils varying from 6 to 0 ¢.m. A good olfactory nerve, freshly prepared, gave in response to single stimuli, electrical or mechanical, a negative variation that was per- fectly legible on the galvanometer scale, but it must be noted that * Tf, as sometimes happened, the two nerves were united near their central end they were used together as one nerve. t Described by Pereles and Sachs, ‘ Pfliiger’s Arch.,’ vol. 52, p. 529. Miss Sowton, Roy. Soc. Proc., Vol. 66, Plate 4, OLFACTORY NERVES OF PIKE PREPARED FOR EXPERIMENT. hie OY li ee OR ee OR Te Me | FOE et READ ae, . ee [PRT ee Tey yt ee ee” ' RR Ly) a Re ied. | PEs A TaN ee oy Electromotive Phenomena of Non-medullated Nerve. 381 only while the nerve was quite fresh could such responses be obtained by sengle stimuli. The momentary mechanical stimulus consisted in a clean cut through the nerve with fine sharp scissors, the blades of which had been moistened with normal saline ; the nerve was supported, so that there should be no pull upon the led-off portion. In the tables of experiments given below, a few cases will be noticed in which there is a back swing + beyond the position of rest. This is partly instrumental swing, but partly also due to a slight positive after-effect. The galvanometer used, although highly damped, was not perfectly dead-beat, and the small after-effects could not be accu- rately estimated. All that can be said is that there is a slight ten- dency towards positive after-effects on good nerves when quite fresh ; later the back swing tends to fall short of the position of rest. With tetanising induction currents the positive after-effect was only once observed with any certainty. Olfactory Nerve of Pike.-—Nerve current compensated. The nerve current is +, the Repatize variation —, on the scale throughout the experiments. al 1.—March 8, 1899. Pike, weighing 53 lbs. Nerve I. Stimu- lation by single break Defictinn shocks. | Coil Position | Negative Back Value of Back swing + or —*. | distance.| ofrest. | variation. | swing. neg. var. cae : PELE URS 0 cm. fal 69°5 el —1°5 +0°7 a 63 61 5 63 °2 ed face G 5 |b 52°7 54 °4 eee +0 °4 0 cm 47 45°6 46 °7 —1°4 ies a AL 39°7 40°7 —1°4 =0F3 z 45 43°7 | 44.°5 —1°3 —0°5 Nerve moistened. Stimulation by single make induction shocks. O em. 49 4S 49 ‘6 a 47 45°9 46 +2 —l —1il +0°6 ee —0°'8 Same nerve. Ten stimuli make and break alternately at half- second intervals. is 49 38 ‘6 47 —10°4 a 29 O em. 38°5 277 36 °8 —10°8 “ee -—1°7 * In this column the + sign signifies a back swing beyond the position of rest, the — sign signifies that the back swing fell short of that position. 382 Miss S. C. M. Sowton. Observations on the Exp. 2.—Nerve 2. sag aie te by single break induction shocks. } Coil | Position | N egative | Back Value of distance. | of rest. variation. | swing. neg. var. a tke | Ieee.) eavat | Bem. | 52:5 |) 50-7 : af 52-5 so H E 49 | 4772. | 48-8 i -Sea -p:2 0 cm. AB gh ABA | ye —1°9 | —0'1 Stimulation by ten shocks, alternate make and break, at half- ‘second intervals. |. [ 41'5 | —14-3 eee | a7ea ot le Qe] i @ Oo Ww Iw He bo Back swing + or — : { | | xp. 3.—February 21, 1899. Pike, 44 Ibs. Nerve I]. Momentary mechanical stimulation by cutting. | Position | Negative Back | Value of Bick 2 ee of rest. | variation.| swing. |. neg. var. | ee | | | [46 26-5,t 25-5t | 43 ee be | 26 43 | 24°5, 25, 23+ | 40 | — 206.57 ieee at , ak AD an 1 Pee et ee | —19, s) cae hee. | 37 | 20,19,18 | 35 | | -19 | — | Le | 35 | 19, 19 5. 16 | 33 [ =39 79 Paes we ; | 09488 ;0: 416-5, 17, 14-8 | 305/30. 18 ¥ ‘ ai. | 15,155,138 |) — | = 3 CO, 2 mins. : : | : , || 28 24 35 — 4 7 [ f4 nee 30 31°5 ae Wee ' » | 48 44°5 50°53 | — 3° + 2°5 | | 34* 31 51 — 3 14aF | i aie 26 45 he ee 33 | 29 23 34°5 — 6 (+15°5 H ee 19 27 ee . er 16, 20-S5}+ | 25 -8 +1 | , [a-5 | 18, 1615-5 4 as [10°51 = eed ae ie 11°5, 744 | :20°5 | -10°5 | “<= ee 7 > | 19°5 | 9°5,11,10 | 18 | -10.) | S23 hein co, 1 5 sy _o- Fie EEIVANN | wt UL ALL ANN IAIN ot 1 VV YY I LY WA | Leese) ot LL Ee EE ot tf UE tt | Misutes.o 9 WW JE oe bs pe pao s* Curve of first part of Experiment 6. * The + after-effect began before the end of stimulation. + A pause of a second or so took place here. Electromotive Phenomena of Non-medullated Nerve. 387 r _ April 14, 1899. Olfactory Nerve. Bet Valud Back swing. Stim. | Coil. | Zero.| Neg. var. eee of 7 Peri: OF Incomplete : fal | ! — oo 5” 6 em. | 53 18°5 47 —34°5 — Less 6 a (40 17 a) 30 | | 4 CO, 1 min. (slow stream). | tt BO 25 50 | 25 i sr a 50 30 44, — 20 — » 6 | 44, 25°5 ieee 22a ete CO, 2 mins. el 4, «| 40°5 | 28 47-5 | -12°5)+ 7 mee) ae 38°5 48 16-5 [os aa 35°5 Ad Bias \+ 3 | 415 37 50 — 45 |+ 8:5 | ae O75 52 es +17 | A 28 19 °5 45°5 | — 8:5 [+175 (2405 14 87°5 | —10°5 |+13 ee eee 8°5 27 295-5 4 3 yh RET 5 23 Sie ine’. 1 SL 225 1 20-5 | —21:5| = , 2 Spot adjusted. | wo 0 23 ik a Al Ua al i Wl | 4675 Baveiul ehonie-t $= 24 fo | CO, 2 mins 50 44 60 | 2A BPE TO er Ok MYSeled) 48 48 — 25|+ 0% | ee. | 43 °5 52°5 | — 2°53 |+ 6°5 ee es 40 58 fois eed iia Ly ania | Be | 30 55 is Gal) eto min Leta 65 |4+24°5 28 °5 59 °5 | — Primi 2 Sat Ree eeee TUIAVI ANA SF eieHSReAD ST CCIE Minutes.O 1 ee Coit be since once gan min. Be 5 ote Curve of first part of Experiment 7. 388 Miss 8S. C. M. Sowton. Observations on the— Fig. 6 is a photographic record of a similar experiment. og Fie. 6. | Coit ab 5 | | | Effect of CO, on the negative variation of non-medullated nerve. Stimulation at 1-minute intervals. Another condition in which the olfactory nerve gave positive after- effects was after being kept for some time in normal saline. Isolated grey nerve appears to be far less resistant than white nerve, but the olfactory sometimes retained its excitability for as long as four hours after excision, and the electrical response of such a kept nerve was usually a negative effect followed by a large positive after-effect. In frog nerve there is also a development of positive after-effect in stale nerve ; but whereas such effects are in this case markedly reduced if not abolished by a new transverse section, in the case of the olfactory nerve the positive after-variation persists after fresh section. A few experiments were made to test the effect of ether and chloro- form vapour on the negative variation of non-medullated nerve. The galvanometric effect was promptly abolished by brief administration of ether or chloroform vapour. Recovery after anesthesia occurred to some extent—more markedly in the case of ether than in that of chloroform. - I wish to take this opportunity of thanking Professor Hering for the means of study which I enjoyed at Leipzig, and for his kindness in sparing me much of his own valuable time. I would also thank Dr. Waller for kind help and suggestions. Electromotive Phenomena of Non-medullated Nerve. 389 Fie. 7. Ether Effect of ether on negative variation of non-medullated nerve (Pike). Fie. 8. S ~|8 Coil at o Effect of chloroform on the negative variation of non-medullated nerve (Pike). VOL. LXVI. ee G 390 Dr.C.S. Sherrington. Hxperiments on the Value of “ Experiments on the Value of Vascular and Visceral Factors for the Genesis of Emotion.” ByC. 8. SHerrinaton, M.A., M.D., F.R.S. Received April 5,—Read May 10, 1900. (From the Physiological Laboratory, University College, Liverpool.) That marked reactions of those portions of the nervous system which regulate the activity of the thoracic and abdominal organs and the skin do contribute characteristically to the phenomena of emotion has long been common knowledge. In descriptions of emotion furnished in recent years by certain leading psychologists these purely physio- logical processes have been given a place more important than was attributed to them formerly. To changes induced in the condition of the heart and blood vessels, lungs, abdominal and pelvic viscera and skin has been assigned a large causal réle in the genesis of affective psychological states. Whereas the cardiac, vascular, respiratory, and visceral phenomena accompanying emotion were wont to be regarded as secondary to the cerebral and psychological, we find their position reversed in the writings of Professor W. James,* Professor C. Lange,t and Professor Sergi. It is true that it is claimed that this more recent position has been foreseen and partly preoccupied by older writers, by Descartes§ and Malebranche,||; but as Professor Ribot, who with some reservation, endorses the new theory,{l writes: “ La supe- HUEe de James et de Lange, est de avoir posée clairement et de ’étre efforcés de l’appuyer sur des preuves experimentales.”** It is Pay fitting here to enter on a full statement of the doctrine. I may, however, be allowed some brief quotations from the authorities indicating their teaching. After having, in a previous chapter, given an account of the influence that a shock of feeling exerts on the nerve centres con- trolling circulation, respiration, skin glands, abdominal and pelvic viscera, Professor James writes: ‘“ Our natural way of thinking about these coarser emotions (¢.g., ‘ grief, fear, rage, love’) is that the mental perception of some fact excites the mental affection called the emotion, and that this latter state of mind gives rise to the bodily expression. * ‘Mind,’ London, 1884, PEnneaples of Psychology,’ London, 1890, vol. 2, pp. 448, Ke. + Om Sindsbevigelser, Copenhagen, 1885; German by Kurella, Leipzig, 1887 ; French by Georges Dumas, Paris, 1895. t Dolore e Piacere, Milano, 1894, 398 pp. ‘Zeitschft. f. Psychologie u. d. ehysiol. der Sinnesorgane,’ Hamburg and Leipzig, April, 1897, p. 96, &e. § ‘Passions de l’Ame,’ Paris, 1648-9; ‘Passiones sive Affectus Animae,’ Amstelod. 1677, || ‘ Recherche de la Verité,’ 1672. { ‘ La Psychologie des Sentiments,’ p. 92—113. Paris, 1896. ae Thid., p. 112, > Vascular and Visceral Factors for the Genesis of Emotion. 391 My theory on the contrary is that the bodily changes follow directly the perception of the exciting fact, and that. our feeling of the same changes as they occur 1S the emotion.”* ‘“ Every ene of the bodily changes, whatsoever it be, is FELT, acutely or obscurely, the moment it occurs. If the reader has never paid attention to this matter, he will be both interested and astonished to learn how many different local bodily feelings he can detect in himself as characteristic of his various emotional moods.”T “Tf we fancy some strong emotion and then try to abstract from our consciousness of it all the feelings of its bodily symptoms we find we have nothing left behind, no ‘ mindstuff’ out of which the emotion can be constituted, and that a cold and neutral state of intellectual per- ception is all that remains.”{ “If I were to become corporeally anesthetic, I should be excluded from the life of the affections, harsh and tender alike, and drag out an existence of merely cognitive or intellectual form.”§ This view is the extreme antithesis to the spiritualistic conception of emotion. On it the “coarser emotions” come to consist In essence merely of sensations which arise in consequence of the effect of an idea upon the internal organs. M. Jules Soury says,|| ‘ Pour James Vemotion n’est que la conscience que nous avons des réactions organ- iques, vasculaires, glandulaires, oe &c., provoqués par certaines perceptions ou certains souvenirs.’ Professor Lange traces the whole psycho-physiology of emotion to certain excitations of the vasomotor centre. He conceives all the other of the organic reflexes occurrent in emotion to be attributable mediately to the vasomotor. For him, as for Professor James, the emotion is the outcome and not the cause or the concomitant of the organic reaction ; but for him the foundation and corner-stone of the organic reaction is as to physiological quality vascular, namely, vaso- motor. Emotion is an outcome of vasomotor reaction to stimuli of a particular kind. The stimulus is some sensation acting often by inter- mediation through some memorial-idea linked to it by association. This stimulus induces a vasomotor action in viscera, skin, and brain. The change thus induced in the circulatory condition of these organs induces changes in the actions of the organs themselves, and these latter changes evoke sensations which constitute the essential part of emotion. It is by excitation of the vasomotor centre therefore that the exciting cause, whatever it may chance to be, of emotion produces the organic phe- nomena which as felt constitute for Lange the whole essence of emotion. It is noteworthy that in Lange’s view the action of the * The italics and emphasis stand as in the original. + ‘Principles of Psychology,’ vol. 2, p. 450, London, 1890. + Ibid., vol. 2, p. 451. § Ibid., vol. 2, p. 452. || ‘Du Systéme Nerveux,’ Paris, 1899, vol, 2, p. 1338. : 24 2 392 Dr. ©.8. Sherrington. Zxperiments on the Value of vasomotor centre upon the blood-vessels of the brain, as well as on those of the viscera and skin, plays an important part. The teaching of Professor Sergi closely approximates to that of Lange. He argues that the exciting stimulus acts on nervous centres in the bulb (medulla oblongata) producing cardiac vascular and respi- ratory effects as well as effects upon the abdominal and pelvic viscera. He writes recently: ‘‘ Lange hat gemeint, die Affekte hingen von dem vasomotorischen Zentrum ab; doch ist dieses Zentrum zu eng, um die Mannigfaltigkeit der visceralen Erscheinungen des Ernahrungs lebens erklaren zu konnen. Dagegen hat mich die Analyse zu der Erkenntniss gebracht dass der Bulbus rachidicus, wo die reflex und automatischen Zentren der Nerven, die das ganze Hrnahrungsleben regulieren, zusammenlaufen, das Zentrum der Affekte und im allge- meinen das der Gefiihle ist.”* The views of James, Lange, and Sergi have common to them this, that according to them the psychological process of emotion is secondary to a discharge of nervous impulses into the vascular and visceral organs of the body suddenly excited by certain peculiar stimuli, and depends upon the reaction of those organs. Professor James’s position in the matter is, however, not wholly like that of Professor Lange. In the first place, he does not consider vasomotor reaction to be primary to all the other organic and visceral disturbances that carry in their train the psychological appanage of emotion ; and to a certain extent Professor Sergi, though more nearly in harmony with Lange, agrees with James in this. In the second place, Professor James seems to distinctly include other ‘‘ motor” sensations and centripetal im- pulses from musculature other than visceral and vascular, among those which causally contribute to emotion. Thirdly, he urges his theory as one which is completely competent only for the ‘coarser ” emotions, among which he instances “fear, anger, love, grief.” For Lange and Sergi the basis of apparition of all feeling and emotion is physio- logical, visceral, and organic, and has seat for the former authority exclusively, and for the latter eminently, in the vasomotor system. This view, which some may conceivably tax with “ materialism,” T has a merit that materialism does oftentimes possess, namely, relative accessibility to experimental test.t Such test it is attempted in certain measure to apply in the observations which I herewith report. They have been obtained from five young dogs. In these * ‘Ztschf. f. Psychologie u. Physiologie d. Sinnesorgane,’ Hamburg and Leipzig, 1897, vol. 14, p. 93. . + James, ‘ Principles of Psychology,’ vol. 2, p. 453. t Sollier, ‘Revue philosophique,’: Paris, March, 1894. Dr. Sollier records experiments made on subjects in the condition of deep hypnosis; their sensation, both cutaneous and deep, was believed to be abolished; the conclusions he draws from the experiments are in support of the theory of James and Lange. Vascular and Visceral Factors for the Genesis of Emotion. 393 the spinal cord has been severed in the lower cervical region. Such a severance lies headward of the exit and entrance of all that system of nerves usually embraced under the term ‘“ sympathetic system.” It therefore sunders from the brain all nexus with the thoracic, abdominal and pelvic viscera, except that existent through certain cranial nerves. It also cuts off all the blood vessels from the bulbar vasomotor centre, except for certain scanty communications through the cranial nerves. The skin and motor organs are, as far as the shoulder, likewise cut off from all communication with the brain. Therefore behind that level they are precluded from contributing to nervous processes of emotion, either in their centripetal or their centri- fugal phases. In each of these dogs the observations have been prolonged for several months subsequent to the operation of transection ; in none has any impairment whatever of emotional character, so far as demonstrable, been detected. To study emotion in a lower animal is not altogether easy—even ina dog. But if reliance be placed on the signs that are usually taken to signify pleasure, anger, fear, disgust, then these animals showed them as unmistakably after as prior to the transection of the cervical spinal cord. The sight of, or the sound of, the attendant who kept them evoked from them the same joyous activity and animated caressful pose of head and feature as formerly. Towards friends and enemies among their fellow-inmates of the animal house they displayed as markedly as ever their liking or their rage.. To © give an instance, I saw fear notably displayed by one of the dogs, a young animal, approached and threatened by a powerful old Macaque monkey. The lowering of the head, the dejected half-averted face, and the drooped ears contributed to indicate existence of an emotion as lively as the animal had ever shown us before the spinal operation had been made. An observation of confirmatory kind I once obtained in the labora- tory of my friend Professor Mosso of Turin. Ina young dog under deep chloroform narcosis, I had performed a spinal transection close behind the origin of the phrenic nerves. Six weeks later, the trauma having completely healed and the condition of spinal shock having largely subsided, I placed the animal once more under chloroform, but this time not profoundly. I connected the femoral artery with the mercurial kymograph and proceeded to record the arterial pres- sure, allowing the chloroformisation gradually to pass off. As the depth of the narcosis waned, the breathing became quicker and less regular. The waking of the animal was accompanied by no pain, because the whole body was insentient behind the cervical region, and the kymograph attachment was in the femoral region. I was intend- ing to faradise a branch of one of the nerves of the right hind limb. Inductorium, electrodes, galvanic cells, and whole electric circuit stood 394 Dr.C.S. Sherrington. xpervments on the Value of on a table near, but not on that on which the kymograph observation was in process. In order to be sure that all was ready, I closed the electric key and touched the vibrator of the inductorium. The harsh rattling noise of the vibrator lasted a few seconds, and I then stopped it by re-opening the key. Turning thereupon toward the arterial record, I was a little disappointed to see that a marked oscillation had suddenly upset the already somewhat undesirably irregular line that had to serve as starting level for the vasomotor reflexes I was wishful to study. It was clear that one would have to wait for greater quietude to re-establish itself again. I waited; the disturbance of the arterial pressure subsided; the previous fairly equable cardiac beat, despite somewhat disquiet respiration, returned. A few minutes later I again started, by force of habit, trying the inductorium for a couple of seconds preparatory to proceeding to excite and observe the vasomotor reflexes. Again, on turning toward the trace running on the kymograph, I was met by a sudden disturbance that had altered it. This time it occurred to me that the sudden whirring noise of the magnetic interruptor might have caused the reaction. This supposition I proceeded to test, and soon found that each time the noise was repeated the disturbance of the circulation followed. If the reaction had become less, as it frequently did after a number of repeti- tions, it was only necessary to wait for ten minutes or a quarter of an hour in order to re-obtain it in its original extent. I then remembered that in examining the limits of the cutaneous anes- thesia in this animal from week to week, I had at several times employed the inductorium ; sometimes the electrodes had in making the delimi- tation been applied to points of skin still sentient, and no doubt had there caused sensations of unpleasant quality. The recurrence of the sound to the awakening animal occasioned now emotional anxiety. But in this animal the vasomotor centre cut off by the spinal section from practically the whole of the rest of the vasomotor mechanism was quite unable to affect the arterial pressure.* Hence that rise of pressure observed by Couty and Charpentiert to occur under emotion of fear was impossible in this case. All the more obvious and un- complicated for that reason appeared the inhibitory action exerted on the heart. ‘The heart that had been beating at the rate of 180 per minute, suddenly fell for twenty seconds to a rate of 54 per minute. The respiratory rhythm was easily seen to be also altered, but no graphic record of the respiratory movement was being employed. A slight elevation of the mean arterial tension immediately preceding the * Some description of the spinal reflex and other vasomotor reactions obtained. from these animals I hove to give in the ‘Journal of Physiology’; they are not necessary to the argument here. . + ‘‘ Effets cardio-vasculaires des excitations des sens,” fig. 4, ‘Archives de Physiologie normale et pathologique,’ 1877, p. 560. —_— Vascular and Visceral Factors for the Genesis of E'motin, 395 vagus action on the heart I incline to attribute to mechanical effect on the circulation, secondary to alteration in respiratory movement. The _ Artertal ' PY€SSUTE. Time tn Seconas. Fic. 1.—Record of the arterial pressure in a dog forty-one days after: spinal transection at the 7th cervical segment. The arterial pressure is high and good in spite of the transection, the period of vasomotor shock having passed by. For the short period marked by the signal the noise of the vibrator of an inductorium sounded and was heard by the animal. The point of the signal marked nearly 8 mm. further to the right than did the kymograph pen. The inhibition of the heart is shown by the oscillations on the kymograph trace. The kymograph paper moved from right to left, so that the tracing reads from left to right. The line marked “‘ Zero of B.P.” signifies the height of the zero of the manometer recording the arterial pressure. interest of the observation here is that it gives an objective illustra- tion of a disturbance emotional in character occurring in an animal 396 - Dr. C.S. Sherrington. Lxperiments on the Value of after the possibility of vasomotor reaction had been set aside, and after the vastly larger portion of all visceral reaction had also been removed. All the evidence I obtained from all the dogs went absolutely con- cordantly to show that in spite of exclusion of such a huge field of vascular, visceral, cutaneous, and motor reaction the emotional states of anger, delight at being caressed by the master or at approach of a friend, fear, and disgust were developed with as far as could be seen unlessened strength. The horripilation of the coat along the crest of the back between the shoulders, so usual an accompaniment of anger in the dog, was of course absent in these dogs, the spinal pilomotor* nerve-fibres having had all connection with the brain ruptured. But absence of this reaction could not for a moment mask emotional dis- turbance so vividly indicated by other features of expression. Regard- ing emotions of fear and disgust, the former was evoked by threatening with the voice or gesturing with a whip, the latter by the expedient of substituting dog’s-flesh for horse- and ox-flesh in the dog’s food-pan. Few dogs, even when hungry, can be prevailed on even to touch dog’s- flesh as food; almost all turn away from it at once with obvious signs of repugnance and dislike. Fear and disgust in answer to these tests seemed as indubitable in these dogs as in normal. Great care was taken throughout not to establish in the dogs under observation by too frequent repetition or encouragement a habit or trick of response by emotional signs that might thus become so to say pseudo-emotional with the artificiality of an acted performance. No employment of the special tests for eliciting the emotions was made in them until after performance of the spinal section. Even then the tests were never frequently rehearsed; nor were the animals ever encouraged or incited to respond unduly or in a particular manner. It was sought to as far as possible approximate the tests to natural incidents, and to as far as possible collect the observations from natural incidents. The above was the condition found to obtain in the animals after the cervical spinal transection. I then proceeded in two animals to carry the test further by additional severance of both the vagi nerves in the neck. ‘The vagus may be regarded as the great visceral unit of the cranial series of nerves. Its section subsequent to prethoracic spinal transection relegates to the field of insentience the stomach, the lungs, and the heart, in addition to the other viscera previously rendered apzsthetic.j It also limits still more narrowly the number * Langley and Sherrington, ‘Journal of Physiology,’ Cambridge and London, 1891. + By apesthetic is meant not only devoid of sensitivity but deprived of all con- nection with the nervous centres necessary to conscious reaction, a meaning for which the word apesthesia was suggested by Dr. Mott and myself, these ‘ Pro- ceedings,’ vol. 56, 1895. Vascular and Visceral Factors for the Genesis of Emotion. 397 of efferent and afferent channels by which the vascular system can be possibly affected. Of the animals chosen for these further observations, one was selected because we soon noted marked emotional characteristics in her beha- viour even on her first arrival in the laboratory. She was a mongrel- bred fox-terrier with rather wiry coat, white in colour. She was older than the other dogs; her exact age we did not know. She quickly showed herself affectionate toward the laboratory attendants, one of whom had her in charge ; but toward some persons and toward several © inmates of the animal house, she frequently exhibited violent dis- plays of anger. Her ebullitions of rage were sudden. Their expres- sion accorded well with a description of the symptoms of rage in the dog furnished by Darwin.* Besides the utterance of the growl, ‘the ears are pressed closely backwards, and the upper lip is retracted out of the way of the teeth, especially of the canines.” The mouth was slightly opened and lifted; the eyelids widely parted; the pupils dilated. The hair along the mid-dorsum, from close behind the head to a point more than half way down the trunk, became rough and bristling. A particularly violent outburst of anger was once suddenly, without warning, exhibited against a visitor who happened to enter with me, and had not before visited the room. Spinal tran- section in the cervical region was performed on this animal (under deep anzsthesia). Subsequent examination months later at the autopsy proved the section to have been through the 6th cervical segment, where it trenches on the 7th. The severance was complete, as was confirmed by microscopic examination. Rapid recovery from the trauma followed. An interval of depression of the spinal functions behind the site of lesion was succeeded by gradual restoration of reflex activity, surface temperature, &c. Sensation, superficial and deep, was found to be lost behind the limit shown by the skin line indicated in the accompanying figure (fig. 2, p. 400, the lower diagram). The flexors of the elbow were not paralysed, but the extensors were com- pletely so. I have showny that the sensory nerve supply and motor nerve supply to any muscle have both of them the same segmental position in the spinal cord. Therefore the only muscle still sentient behind the shoulder region must have been the diaphragm. No alteration whatever was detected in consequence of this lesion in the occurrence of emotion, as judged by anger, by delight, or, when provocation arose, by fear. Her joy at the approach or notice of the attendant, her rage at the intrusion of a cat with which she was unfriendly, appeared as active and thorough as before. But among the signs expressive of rage the bristling of the coat along the back no longer occurred. On the other hand, the eyes were * “ Expression of the Emotions,” Darwin, London, 1872, p. 117. + ‘Phil. Trans.,’ London, 1897. 398 Dr. C©.8. Sherrington. Hxperiments on the Value of well opened, and the pupil distinctly dilated in the paroxysm of anger. Since the brain had been by the transection shut out from discharging impulses wa the cervical sympathetic, the dilatation of pupil must have occurred by inhibition of the action of the oculomotorius centre. That the cervical sympathetic had been cut off from its normal bulbar and cerebral excitation was shown by the semi-paralysis of the membrana nictitans in this dog, as in all the others, after cervical spinal transection. This partial closure of the eye, due to impaired tonus in the third eyelid, was little if at all diminished during the outburst of rage ; but we sometimes thought that during the fit of anger the third eyelid was a little more retracted than in its usual paretic condition. As in the other dogs after spinal transection, so in this, the spinal transection markedly enfeebled the voice. This we thought traceable entirely to the enfeeblement of the respiratory muscles, the only respi- ratory muscle left unparalysed after the transection being the diaphragm. The viciousness of the enfeebled and short-winded growl and bark remained as unmistakable and virulent as ever. Apart from this change in the ocular and vocal factors of the facial and respiratory expression of anger, we detected no departure from their previous normal in any direction whatsoever. The heart-beat could be felt to be altered, sometimes becoming quick and sometimes slow, but rarely remaining unchanged during the exhibition of wrath. I thought I could feel “ vagus beats,” but the upset of respiratory movement made the judgment difficult. One hundred and eighty days after the spinal transection, I degaded under deep chloroform anesthesia the right vagus nerve in the neck, about the level of the cricoid cartilage, and therefore well below the superior laryngeal branch, but above the recurrent laryngeal. The cervical sympathetic trunk in the dog is contained in the same sheath as the vagus, so also is the depressor branch of the superior laryngeal, and all three were divided by the same section. This operation pro- duced curiously small obvious result. There ensued little or no difference between the pupils; the right was generally a little the smaller. Absolutely no difference was discoverable between the degree of protrusion of the third eyelids right and left. The palpebral open- ings on the two sides appeared the same. The pose of the pinna of the ear of each side, both right and left, seemed quite similar. The voice, after the first day succeeding the operation, when it seemed altered in some quality difficult of description, reassumed the char- acter it had had since the spinal transection. The exhibition of emotion, as tested by delight, anger, and fear, indicated emotional states as marked and violent as ever. The trauma was rapidly recovered from as regards the healing of the small wound necessary. Twenty-eight days later the left vagus nerve was similarly divided under deep anesthesia, and at the same level as the right. The left Vascular and Visceral Factors for the Genesis of Emotion. 3899 depressor nerve and the left sympathetic trunk were severed at the same time. Subsequent examination at the autopsy proved that: both nerves had been completely cut. An additional guarantee: was given by the absolute absence of all effect on stimulating the distal end of each of the four trunks before proceeding to the autopsy —that is, twenty-one days after the second vagotomy, when complete nerve degeneration had been allowed time to occur. In this animal, the superior laryngeal was the lowest branch of the vagus remaining intact and connected with the brain. The recurrent laryngeals proceeded from the vagus trunks, below the level of the sections. The results of the operation were dyspneea tending to occur in short-lasting attacks, but often passing off entirely ; some loss of appetite, which was in the course of seven days recovered from ; con- siderable enfeeblement of, and alteration of, the growl and bark, both these however still remaining, although modified. The attacks of dyspnoea diminished, and in the course of ten days disturbed the animal rarely and but little. We began to think they might be avoided altogether. The animal seemed quickly to learn what. postures were least hampering to respiratory movement, and this had as result a marked improvement in breathing and general condition. In this animal the capacity of the nervous system differed from that obtaining in those subjected solely to the spinal transection, in that to the body regions and organs already cut off from the brain and rendered anesthetic and put beyond power of contributing to con- scious reaction, there were added in this case the stomach and lower half (%) of cesophagus, the lungs and lower half (?) of the trachea, and finally the heart itself. (Compare diagrams, fig. 2.) Of any diminution or change in the emotional character of the animal we could detect no trace. The following illustrates her condi- tion in that regard. The approach of the visitor whose advent months previously elicited such violent anger, again provoked an exhibition of wrath as significant as before. The expression was indubitably that of aggressive rage. The animal propped itself against its kennel, and followed each movement of the stranger as though of an opponent, growling viciously, and barking in spite of increasing dyspnoea under the excitement. On other occasions a cat with which she was never friendly, and a monkey new to the laboratory, approaching too near the kennel, excited similar ebullitions. No doubt was left in our minds that, sudden attacks of violent anger were still easily excited in this animal. She also gave evidence daily that she experienced the acces- sion of joyous pleasure and delight she had always shown at the approach of the attendant the first thing of a morning, or at feeding time, or when caressed by him, or encouraged by his voice. I had carefully refrained from testing this animal previously with 400 Dr. C.S. Sherrington. Zxperiments on the Value of regard to disgust at dog’s-flesh if it were offered in her food. After her recovery from the last operation, that is to say from the eighth x \ * ‘ WS Me, Ww, ‘ iy ‘A 1 SS r] Ww ie t poe | 4 sen) 4 U t ( 4 \ \ 4 ’ t ' ' t . NG oN . ' Wor bP J Sees 4 “Sa ’ U D ? a i aay ‘ A“ Hs LN ‘ \\ 4 BNET ‘ i 1 ; 4 . es H ‘ \ mad ‘ ‘ ayaa so ! ‘4 / ; os ne “ eu -- 4 Be ‘ ry ; i Dh as i ae Fie. 2.—Diagram to indicate the extent of the parts still retaining sensitivity after the spinal (upper figure) and combined spinal and vagosympathetic (lower figure) nerve sections described in the text, pp. 398, 396. The extent of skin surface left sentient is delimited by the continuous (not dotted) lines in the figures. The limit of “deep,” 7.e., muscular, articular, &c., sensitivity also corresponds with this line. But the limit to which the respiratory and ali- mentary tracts still retained sensation is shown by dotted outlines of the lungs, heart, and stomach in the upper figure, of the larynx and upper part of cesophagus in the lower figure. From anatomical data it is presumed that the trachea and csophagus had been deprived of all sensitivity somewhere about those levels. The curved line behind the chest indicates the diaphragm as the only muscle behind the shoulder still retaining afferent nerves. day after it, we proceeded to this observation. Flesh was given her daily in a bowl of milk, and this (after return of her appetite a week subsequent to the second vagotomy) she took with relish. The meat Vascular and Visceral Factors for the Genesis of Emotion. 401 was always already cut into pieces of a size rather larger than the lumps of sugar usual for the breakfast table. It was generally horse-flesh, sometimes ox-flesh. On the tenth day after the final operation the bowl was placed by the attendant, as usual, in the corner of the stall, with milk and meat in it in every way as usual; but the meat was flesh from a large dog killed in the laboratory on the previous day. Our animal eagerly drew itself toward the food ; it had seen the other dogs fed and evidently itself was hungry. Its muzzle had almost dipped into the milk before it suddenly seemed to find something amiss there. It hesitated, moved its muzzle about above the milk, made a venture to take a piece of the meat, but before actually seizing it stopped short and withdrew again from it. Finally, after some further examination of the contents of the bowl (it was usual for it to begin the attack of its food by taking out and eating the pieces of meat), without touching them, the creature turned away from the bowl and withdrew itself to the opposite side of the cage. Some minutes later, in result it seemed of encouragement from us to try the food again, it returned to the bowl. The same hesitant display of conflicting desire and dislike was once more gone through. The bowl was then removed by the attendant, emptied, washed, and horse-flesh similarly prepared and placed in a fresh quantity of milk was offered in it to the animal. The animal once more drew itself toward the bowl and this time began to eat the meat, soon emptying the dish. This test was similarly applied afterwards on various occa- sions ; always with the above result, except that twice the animal did, after much hesitancy, lap some of the milk out of the bowl, although dog’s-flesh was immersed in it. We have occasionally seen a normal dog do likewise when hungry. To press the flesh upon our animal was of no real avail on any occasion; the coaxing only succeeded in _ getting her to, as it were, re-examine but not to touch the morsels. The impression made on all of us by the dog’s behaviour has been that. there existed in the dog’s-flesh something which was repulsive to — the animal and excited in it disgust unconquerable by ordinary hunger. Some odour attaching to the flesh seemed the mark for its recognition. Fear seemed to be clearly elicitable in this animal. While I held her in my arms the attendant, approaching from another room the door from which was open, chid the dog in high scolding tones. The creature's head sank, her gaze turned away from her advancing master, and her face seemed to betray dejection and anxiety. The respiration altered and became unquiet, but the pulse was never altered in rate although perhaps slightly in volume. Twenty days after the last vagotomy the animal suddenly de- veloped a serious attack of dyspnoea; this was recovered from, but in the course of the following day a similar attack occurred. Fearing 402 Vascular and Visceral Factors for the Genesis of Emotion. lest in our absence such an attack should recur and prove fatal, pre- cluding the possibility of testing the excitability of the vagi and sympathetic and of thus examining the completeness of their functional removal, I therefore on the 21st day after the latter vagotomy and the 229th after the cervical transection killed the animal under deep chloroform narcosis. A second dog, a quite young puppy, was made the subject of a course of similar experiments. The animal was hardly so suitable, because it at no time, either before commencement of or in the course of the experiment, exhibited to similar degree the signs of anger. Its exhibitions of joy and pleasure, as also, on occasion, of fear, were well marked. The spinal transection in the cervical region, and the double section of the vagi and the cervical sympathetic nerves in the neck, seemed not to dull at all its emotional character as tested by these tests. The spinal transection was through the seventh cervical segment, the section of the vagi above both recurrent laryngeal branches and of the sympathetic trunks at the same level. The animal was preserved under observation 174 days from the time of the first operation, namely, from the spinal transection. The autopsy proved all the sections to have been complete. The results have substantiated those noted in the previously related observations. The only point of difference worth remarking seems that in this last case the reactions were given by a very young animal practically reared in the laboratory, whereas in the just mentioned older bitch there was a past history, with habit and education, of the details of which we in the laboratory knew nothing and have failed to obtain information. These experimental observations yield no support to the theories of the production of emotion quoted at the opening of this communica- tion. On the contrary, I cannot but think that they go some way toward negativing them.* A vasomotor theory of the production of emotion seems at any rate rendered quite untenable. I am not sure if I understand Professor Sergi aright when I think that he suggests that in absence of vascular and visceral reactions, the very quality of affective tone must be lacking to sensations. Such a suggestion is opposed by the ease with which evidence of unpleasant, ¢.g., painful quality of sensation could be evoked by appropriate excitation of the still sentient regions of skin in the animals we had before us, the subject of this note. It need hardly be added that the importance of the concurrence, together with the other reactions in emotion, of marked vascular and * All who have visited and seen the animals, the subject of this communica- tion, have fully concurred in the opinion of myself and others in the laboratory as to the possession by them of ample and lively emotions. I would especially mention and thank for their attention to the matter, Dr. Abram, Professor Paul, Dr. Warrington, Sir James Russell, and Dr. James Mackenzie, On the Brightness of the Corona of April 16, 1893. 403 visceral and cutaneous, seems to me as to others hardly over-estimable for the study of the subject. Wundt,* A. Mosso,t Alf. Lehmann, t Head,§ and Wright|| are among those who have in recent years laid stress on that aspect of the phenomenon. I would not be thought to impugn the importance of the study of such organic phenomena in connection with emotional mental states. The only respect in which the here given observations affect the position of affairs is, that they, I think, render it necessary to attribute to these elements of emotion another significance than that imputed by the authorities quoted in my opening paragraph. The picturesque incisiveness of all that comes from Professor James’s pen, renders the more persuasive any argument that it pursues. His suggestive chapters led to the above attempt at examination of his theory, an examination the incom- pleteness of which I wish to unreservedly acknowledge. The expenses of this investigation have been in part met by a grant given by the Research Committee of the British Medical Association. I would take this opportunity of tendering to them my best thanks for their generous support. “On the Brightness of the Corona of April 16, 1893. Preliminary Note.” By H. H. Turner, M.A, F.B.S., Savilian Professor. Received March 29,—Read May 10, 1900. The visual brightness of the corona was measured at the total eclipses of August 29, 1886, and April 16, 1893, by Professor T. E. Thorpe, using a method arranged by Sir W. Abney (‘Phil. Trans.,’ A, 1889, p. 363, and 1896, p. 433). Soon after the first of these eclipses, Sir W. Abney devised a method of measuring the brightness photographically, by exposing a portion of the plate, which was not exposed to the sky, to a standard light passed through a row of small square screens of varying and known thickness. The result is a series of “standard squares” on the plate, which show the density of deposit due to standard lights of known values ; and by comparing the density of the coronal image we find the brightness of the corona in terms of these standard lights. These squares were first put on the coronal photographs by English observers at the eclipse of December, 1889, and have been syste- matically used by them since. The 1889 photographs have not yet been measured. Some measures of the 1893 photographs were made by me in Sir W. Abney’s laboratory at South Kensington, in July, * “Grundriss d. Psychologie,’ vol. 1, 3te Auflage, Leipzig, 1893. + ‘La Paura,’ Milano, 1885. = ‘Das Gefiithlsleben,’ Leipzig, 1892. § “Visceral and referred Pains,” Parts I, II, and III, Brain, 1893-7, London. (| Zbid., ‘The Physiological Element in Emotion,” 1894, London. 404 Prof Hu. Durer: 1894; but the standard squares had not received a sufficiently long exposure and additional experiments were required. These were carried out by Sir W. Abney during the year; but causes which need not here be dwelt on (chiefly the desire to make further measures which other work has hitherto prevented but which are now being made) have delayed the publication of the results far too long: and, while still reserving the details for a more complete account, I publish one or two general results which may be useful to others in preparing for the forthcoming eclipse. Three plates were measured, taken by Serg. Kearney, at Fundium. He was provided with a “double-tube,” which took photographs of two sizes, the diameters of the moon’s image being 0°6 inch and 1°5 inches respectively. The following table shews the details of ex- posure :— Table I. Plate numbers. | Exposures. tad Time in seconds cee 13 inch. Duration. from commencement é of totality. . seconds. 1 Bare! GN 20 Stores 2 2A les 120 41 ,, 161 3 3A 50 169 5. 219 A, 4A 5 227 ,, 232 5 5A 2 240 ,, 242 6 | 6A 1 250 ,, 251 Totality ends... ae ue “sf e. 250 The plates measured were those numbered 3a (large scale), 3 and 5 (small scale), of which the first was measured rather elaborately along four radii extending due N., 8., E., W, from the limb respectively, and the reference table or curve of standard lights was constructed for this plate. The following table shows the actual measures, each number being deduced from the mean of three sector readings by application of the curve of standard lights. The unit of the table is the effect caused by an amyl acetate lamp shining on the plate from a distance of 9 feet for 1 second. The numbers represent powers of 2, giving the intensities in terms of this unit. Thus 7:7 means 27-7, or 208 times the unit. The distances from the limb are in terms of a cardboard scale, and it was found by measurement of the moon’s diameter that 100 div. = 156 = solar radius nearly. Hence 10 div. may be regarded as 0:1 solar radius. — On the Brightness of the Corona of April 16,1893. 405 - Table IT. Measures of Plate 3A (large scale). Exposure 50 seconds along four radii (N., S., E., W.). Powers of 2 representing intensity of light. Distance from moon’s limb. | ta N. S. i. AW ; } | | Ae meetet : Diy. f | : | H 0 0-0 Pies. 858 78 rie. ae 10 | 1°6 i fis i em De ocr PT yoly 7-9 eo : 30 | Gr ue Lia) 7G iF Mey Gare 40 | 6°2 6-9 7-4 7-8 7:0 50 / 7-8 el) i 68 at 6°6 60 | 9 +4 ae oe. OBS 70 10°9 4°8 5 °8 5A 80 11°5 4:4 5-1 5-5 47 90 13-0 3:8 46 42 100 15°6 scares We a) 53 110 el 870. ls Arb "i 3:1 120 | 18-7 2:8 3 °2 3°3 a tiie! 130 Bese se PB 28 a Pe es, 140 ONS eee ane ee ’ Kae a: : 150 | 23-4 22 | 26 rs ee ite 160 | 25-0 | 2s ho) Legpeg 2°5 a 170 2 26:5 E 2°3 ED Lo ae 180 / 9871 1°9 2-0 ms c 190 29°6 ss 1°8 6 1°7 200 31:2 enh Meg 1°9 1°68 210 328 a | a mS 1°7 220 24.°3 ey a 2, abi | 230 35-9 ay | on Lee 240 | 37 -4 £2 | 1°5 | 1°4 | This table shows (1) The accuracy of the method. The ae of the light is clearly determinable within an error of 0-1 or 0°2, 2.¢., of 2°7 or 2°, which are ratios of 1:07 and 1°15 respectively. (2) The intensity falls off in nearly the same manner in all four directions. [It may be remarked that 1893 was near a sun-spot maximum, and the corona of the kind approximating to symmetry all round the limb.| Such differences as there are do not arise from the excentric position of the sun behind the moon; for the photograph was taken after the middle of totality, when the moon would have advanced towards the east, and hence the western radius should be brighter than the eastern at the same distance from moon’s limb; whereas the contrary is the case. Also there is a marked difference between the N. and §. radii. : VOL, LXVI. 2H 406 fe -- Prof, H. H. ‘Turmer, (3) The falling off in intensity is very rapid at first. At one radius from the limb it has fallen to one-twentieth, at two radii to one- hundredth. | (4) As regards the absolute intensity, it will be more convenient to refer the brightness to the more familiar unit of the moon’s brightness. It is a fair. assumption to take the total brightness of the moon as 0-02 candle at one foot ; or (since an amyl acetate lamp = 0°8 candle) to 0°02 x 81 + 0°8 of the units (amyl acetate lamp at 9 feet) adopted in the intensity scale: that is, to 2°025 of such units. Now if the moon were to shine for 50 seconds (the exposure of the photograph) instead of 1, we must multiply this number by 50; and if further it shines on the plate through a 4-inch object glass, so that the light falling on this 4-inch circle is collected into an image of 1:5 inches diameter, the brightness of any point of this patch would represent 2°025 x 50 x (4/1°5)? = 720 units = 29° units. Thus if we subtract 9°5 from-all the numbers in Table I, we shall get numbers fairly expressing the coronal brightness in terms of that of the moon, in powers of 2 as before. (5) The results from the other photographs need not be given (in this preliminary note) in detail: they confirm those already given remarkably well; and show that the diminution of light is very gradual indeed after 45 minutes from the limb. They also seem to show that the readings near the edges of the plate in 3A are too low by about 1:0; which is due to the fact that the magnifier is too small ‘to take in the whole object glass at the edges of the field. (6) We can now give the comparison of visual and photographic observations. The results from plates 3 and 5 are preferred to those of 3A far from the limb, for the reason stated in the last paragraph, and for measures near the limb, results obtained from special measures with a plain glass reflector. In ‘Phil. Trans.,’ A, vol. 180 (1889), pp. 380—1, Abney and Thorpe give their visual readings for the 1886 eclipse in terms of a Siemens’ unit, and remark that the moon would have given an image equal to 1-2 candles or 1:4 Siemens’ units on the same scale. Hence we must divide their figures by 1:4 to get results comparable with the above ; and the same must be done for the 1893 eclipse, the numbers for which are given in ‘ Phil. Trans.,’ 1896 (A), p. 433. We thus get the following table of comparative results, replacmg now the powers of 2 by ordinary decimals :— On the Brightness of the Corona of April 16,1893. ..407 Table ITI. | Brightness compared with moon. d = distance pa a hes Observed Caleulated — from limd in Observed visual. photographic. |0-05/(d-+0°182). solar radii. | 1886. | 1893. 1893. 0-0 | (0°40) (1°54) | O-1 | 0°49 0°64. 0:2 0°33 0°35 | 0-4 me ber os O-18 «°° 0-15 0°6 0047 | 07048 0-090 0-083 1-0 0-038 0-034 0-026 | 0:036 : 1°4 0-031 | 0-027 0017 0-020 | 18 0-024. @-o22 1 @-012 0-013 |. 2-2 e019 4 O01 ~ t -0T04 7 } O-G088 2-6 | 0-015 | 0-013 0 0092 0 -0064: 1 No measures of brightness were made visually within 0°6 of a radius of the sun from the limb. It would be interesting to have this comparison made. For the region where we have a comparison, it appears that the light falls off photographically more rapidly than visually. This is in accordance with experience, the faint extensions having been seen more easily than photographed. As regards the central portions, we have some indirect information ; for the total brightness of the 1893 corona was found to be piareiest to 0:026 Siemens’ unit at one foot or 0:022 candle, 7.¢., rather more than the value assumed above for the moon. Integrating numerically for the part in the annulus extending from 0-6 radii to 2-6 radii, we find that this portion is equal to 0: 20 moon photographically and 0°25 moon visually. This leaves about 0°75 moon (visual) for the part within this distance (i.c., from the limb to 0°6 radius), while photographically the value got from the curve is only 0-44 moon. It seems as though the corona were altogether brighter visually than photographically, in the ratio of about 3 to 2; but this conclusion needs confirmation. An attempt has been made (in the column “calculated” of the above table) to represent the brightness by a formula. The American photographs of 1878 suggested that the coronal light varied inversely as the square of the distance from the sun’s limb. Abney and Thorpe find that this law does not hold; but the photographic observations can be made to obey the law approximately. The calculated numbers are obtained from the formula 0-05/(d + 0-18), “408 ‘On the Brightness of the Corona of April 16, 1893. where d is the distance from moon’s limb given in first column. The distance is thus measured from a point 0°18 = 2'"7 within the circle on the photograph taken as the moon’s limb. This is well within the sun’s limb, though near it. It need,not cause much surprise that the calculated numbers close to the limb exceed the observed ; for the corona close to the limb was obscured during part of the exposure by the advancing moon. Further examination of these points is required. The following diagrams exhibit graphically the figures of - Table III. é 29) Limb Soler Radii. a : Radio-activity of Uraniwm. 409. sy— Square Root of Brightness. a 2-0 20 : Infinite - Reciprocat of Distance ~ Listance. from assumed timb. “ Radio-activity of Uranium.” By Sir WitttaAm Crookes, F.R.S. Received May 3,—Read May 10, 1900. [Pate 5.1} 1. The researches of M. Henri Becquerel have shown that com- pounds of uranium possess the property now called “radio-activity ” ; that jis, rays emitted by them affect a sensitive photographic plate through bodies usually considered opaque to light; they discharge an electrometer when brought near it; and they are deflected by a magnet. These rays are now called “Becquerel rays,” or “uranic rays.” mee 2. On the discovery by M. and Mdme. Curie of polonium and 410 Sir W. Crookes. radium, bodies of enormous radio-active powers, it was suggested that: uranium might possibly owe its power to the presence of a small quantity of one of these bodies. But in a paper published in the ‘Revue Générale des Sciences’ for January, 1899, Mdme. Curie says :— “This does not appear probable, for if such were the case different samples of uranium compounds would have very. different radio- activities, but in the course of a number of experiments made with various samples of metallic uranium, as well as with oxides and salts from various sources, I have never found any marked difference between the relative activities of the same compound.” In another paper* the same author says that ‘the property of emitting rays . . . which act on photographic plates is a specific property of wranium and thorium.” “The physical condition of the metal seems to be of an altogether secondary importance.” “ Uranium and thorium alone are practically active.” 3. When the discovery of radium was announced, and it was said to have ‘to all appearance the properties of almost pure barium,’7 it occurred to me that radium might be found in detectable quantities in some barium minerals were search made among them from different localities. Accordingly specimens of the following minerals were put on sensitive plates, a sheet of black paper separating them from the sensitive surface. As it was probable that the radio-active substance would be present, if at all, in very minute quantities, the sensitive plate was exposed to their influence for forty- “Hep hours. Pe (Heavy Spar). | , from Hungary. (Three specimens.) ud » Cumberland. (Hight.) 5 , Westmorland. (One.) > em -y Cumberland” (A fine crystal.)ay =e a ,, Derbyshire. (Three.) as , Arkendale. (One.) a ~ tlartz, ) (One) a , Scotland. (One.) S) Weds treland.., {a wa.) 33 » Northumberland. (Two.) as > Arran. (One) FA ;, Cherbourg. (One.) Several unnamed, but finely crystallised, peer! Witherite from Lancashire: (One.) $d ‘Cumberland. (Four.) is » Northumberland. ius ) _ * M. and. Mdme Curie, Eanes ie csacdian vol. 127, ils 195; ‘Chem. Dette vol. 78, p. 49, July 29, 1898. . t M. and Mdme. Curie and M. Bémont, oe Ba , vol. a, P. 1215; ‘Clie: News,’ vol. 79, p. 1, January 8, ee mR —- Radio-activity of Uranium. ATd Not one of these minerals showed the slightest action on the sensi- tive plate. 4. Having obtained negative results with barium herpeiceata eh J went through every mineral in my cabinet—a somewhat extensive collection, numbering many fine specimens. Large photographic plates were covered with black paper, and the minerals were laid on them as close as they could conveniently be placed, accurate note of their names and positions being recorded. They were exposed in total darkness for forty-eight hours. By this means a list of radio- active minerals was ultimately obtained. They were then tested for order of intensity of action. The following is a list of active minerals arranged in order, the most active heading the list :— 1. Pitchblende. 9. Broggerite. 2. Uranite. 10. Monazite. 3. Autunite. 11. Xenotime. 4.. Orangite. 12. Arrhenite. 5. Thorite. 13. Sipilite. 6. Euxenite. 14. Fergusonite. 7. Samarskite. 15. Chalcolite. 8. Alvite. 16. Hielmite. It will be observed that these minerals all contain either uranium or thorium. She 5. Pitchblende was the most radio-active mineral, but it varied much in different parts. A slice was cut from a piece of pitchblende from Cornwall and the surface was polished. A sensitive photo- graphic plate was pressed against it, and after twenty-four hours the plate was developed. The impression showed the structure of the mineral in a remarkable manner, every little piece of pitchblende showing black, those portions in which the radio-active substance was not so operative showing in half tint, while the felspar, quartz, pyrites, | &¢., having no radio-activity, left the plate transparent (see Plate 5). ° Pitchblende from different localities differed greatly in action. 6. A large crystal of orangite from Arendal was ground flat and polished at one end and side, and a piece of sensitive celluloid film, cut half through so as to allow it to bend sharply, was put on the polished surfaces, one half pressing against the end and the other half against the side. The exposure was continued for seventy-two hours. On developing, no difference could be seen in the intensity of the impression, whether made by the end or the side of the crystal. The impression also was uniform over the surface, the pias in the surfaces not having impressed themselves. cn These experiments were repeated with the interposition of a thin sheet of celluloid between the mineral and the sensitive plate.” ae results were practically the same as before. ee 4t2 Sir W. Crookes. . 7. Roughly speaking, the action of pitchblende is in proportion to the percentage of uranium in the mineral. Finely powdered pitch- blende of different degrees of richness were experimented with. Cells were made of thick lead pipes half an inch internal diameter and one inch long, closed at the lower ends with card. They were filled with powdered pitchblende, one containing 43 per cent. Ur3Og and the other 12 per cent. Ur3;03. A sensitive plate being covered with black paper, the lead..cells were laid on it and kept in total darkness for 120 hours. The intensity of the spot under the 43 per cent. ore on development was top to be at least three times that of the one under the 12 per cent. ore. Two lead cells were Bish one being a quarter of an inch long and the other two inches long. They were filled completely with the 43 per cent. ore, and a sensitive plate exposed to their action for forty- eight hours. On developing it was doubtful whether any difference existed in the intensity of the two spots, proving that the action does not pass through much thickness of active material, a quarter of an inch being equal in effect to two inches. No difference in the action was noticed when the bottom of the cell was made of thin glass cemented on, instead of card. Four cells were filled with pitchblende and placed side by side on a sensitive plate. . After having acted twenty-four hours the first was removed, the second after forty-eight hours, the third after seventy- two hours, and the last was kept on for ninety-six hours. On develop- ing the plate the spots had intensities varying with the lengths of exposure, and in about the right proportion, on the assumption that double the time of action gives double the intensity of blackening. ‘8. For convenience of comparison I had a number of glass cells made, three-quarters of an inch wide and deep, so that they could either be-sealed up or closed with a cork. A piece of, apparatus was made so.as to take radiographs of samples with more ease and cer- tainty. A lead plate, 2 mm. thick, 64 inches long, and 22 inches wide, has circular holes punched in it, one inch in diameter. Under the thick plate of lead is another thinner plate, made of pure assay foil, and. having holes in it, concentric with the others, but barely ? inch in diameter, so that one of the small cells will not pass through the lower hole, but will pass easily through the upper hole, and thus be kept in place, To prevent contact between the lead plate and the sensitive surface a thin sheet of celluloid is fixed beneath, with holes punched. in it concentric with those in the lead plates. In the top left corner of the lead plates is a short steel pin, which can be pressed on the sensitive plate and so register its position in respect to the cells experimented with. The lead and celluloid plates are then bound together and the whole is auhee into a shallow wooden tray with a light-tight cover. Radio-activity of Uraniwn. 413 ‘A sensitive film is laid face upwards at the bottom of the wooden tray; on this are put the lead screens, and then the experimental cells of radio-active bodies in order, — note being taken of their rela- tive positions. 9. Wishing to prepare compounds of radium and polonium, the very curious bodies discovered by M. and Mdme. Curie, I arranged with my friend Mr. Tyrer, Stirling Chemical Works, for the systematic working up of half a ton of pitchblende. It was necessary to examine every precipitate and filtrate in each stage of the operation, and for convenience of registration they had to be compared with a standard cell filled with a substance unvarying in action. After many trials, I selected crystallised uranium nitrate as being strongly radio-active and easily prepared pure. This led me to the observation which forms the subject of the present paper. 10. The following compounds of uranium were tested simultaneously, being put into glass cells and arranged on a lead screen with seven holes :— Metallic uranium, from M. Moissan. Uranium nitrate, UO,(NO3)2.6H20. Uranium acetate. Uranium persulphate, Uranium protosulphate. Uranium oxide (green), UQ)2U0O3. Uranium oxide (black), UO2UOs. For twenty-four hours the sensitive plate was exposed to the influence of these bodies. With the exception of metallic uranium, which showed least action, there was not much difference between the effect produced by any of the others. 11. In order to prepare uranium nitrate of great cae for a standard, I took some pounds of the commercial salt and purified it, first by solution in ether (13), and then by repeated. crystallisation. After many operations a cell was filled with the crystals, and it was used as a standard. ‘To my surprise, after having acted on a sensitive plate for twenty-four hours, on development not a trace of ee could be seen. 12. Thereupon I tried the following, experiments to ascertain if és great radio-activity of some uranium compounds and the absence of it in others might be caused by some variation of physical, crystalline, or chemical condition. Commercial uranium nitrate was taken, and—_ ew A portion was heated with excess of nitric acid to dryness. 0 on the water-bath, It was powdered and put.in a cell. (2) A similar portion of the salt was dissolved in alcohol, and the salgeion evaporated to dryness over the water-bath. The resulting orange-coloured pasty material was ground and put inacell. |. 4: Posey. © 414 Sir W. Crookes. (3) The salt was treated in the same way as No. 1, except that the excess of water and acid were not driven off. (4) A small quantity was heated for some time on a water-bath till it was thoroughly dry. It was then powdered and put in a cell. (5) 50 grains were put in a glass cell and heated on a water-bath to about 75° C. till it had dissolved in its water of crystallisation. (6) 50 grains in a glass cell were heated on a sand-bath to about 230°C. The water ais crystallisation having been driven off the salt fused, and on cooling remained a hard, yellow, glassy mass. (7) A similar quantity of the same salt was heated to a little above 230° C. till it commenced to decompose. (8) A similar quantity was heated more strongly, till about half the nitrate had decomposed. (9) The same quantity was heated till decomposition was complete. (10) As a standard, some of the same lot of commercial uranium nitrate from which these lots were taken was put ina cell. These ten cells were placed in a lead screen apparatus, and a sensi- tive plate was exposed to their influence for twenty-four hours. On development there was not much difference between any of the impres- sions, that under No. 9 being a little the strongest. Thus it appears that no modification of physical or chemical con- dition materially affects the radio-active property of a uranium com- pound when, to begin with, the salt experimented on possesses it ; other imilar experiments show that, starting with an inactive uranium salt, nothing that can be done to it will cause it to acquire this property. It is therefore evident that, as I had suspected, the radio-active pro- perty ascribed to uranium and _ its compounds is not an inherent property of the element, but resides in some outside body which can be separated from it. Having by repeated crystallisation succeeded in preparing a photo- graphically inactive uranium nitrate, I started experiments with several pounds of the commercial salt to ascertain the readiest means of separating from it the active body. 13. Into a stoppered cylinder I put 1 lb. of erystdliaeee nitrate and poured on it a pound of methylated ether, sp. gr. 0°72. The salt easily dissolved on shaking, and after a few hours the whole of the crystals had disappeared, leaving at the bottom of the cylinder 1000 fluid grains of a heavy aqueous solution. I separated the aqueous solution from the ethereal solution, and evaporated it to dryness to remove traces of ether. The ethereal solution was allowed to evapo- rate spontaneously. Equal quantities of the soluble and the insoluble in ether I put into glass cells and added sufficient dilute nitric acid to dissolvé the salt, and then evaporated each lot to dryness on the water-bath. When dry the two cells, and a third containing some of the original nitrate, were put on a sensitive plate for twenty-four hours: Radto-activity of Uranium. 415 On development it was found that the action of the part undis- solved by ether was very strong, that of the original nitrate not more than half as strong, while no action whatever could be detected on the part of the plate covered:by the salt soluble in ether. The portion insoluble in ether, after evaporation to dryness with nitric acid, and then crystallisation in water, in no way differed in appearance from ordinary uranium nitrate. The portion soluble in ether, when dried, heated with dilute nitric acid and crystallised, also had the same appearance as the initial salt. | 14. The crystallised nitrate from the portion insoluble in ether I again extracted with ether. Most of it dissolved, and a small portion of heavy aqueous liquid settled at the bottom. As before, the nitrate which dissolved in ether had scarcely any radio-active power, while the residue from this second extraction possessed it in a strong degree. The residue after the second extraction was about double the activity of the residue after the first extraction, showing that ether, while dissolving uranium nitrate itself with facility, does not dissolve the body to which it owes its radio-activity. 15. The uranium nitrate from the portion insoluble in ether was submitted to fractional crystallisation in the following manner :— The solution was evaporated until on cooling about three-fourths would crystallise out. The beaker in which this operation was per- formed was called No. 1. When crystallisation had finished, the mother-liquor was poured into a beaker called No. 2. A little water was added to No. 1 and it was warmed to dissolve the crystals ; No. 2 was evaporated a little, and both were set aside to crystallise. When cold the mother-liquor from 2 was poured into a beaker No. 3, the mother-liquor from 1 was poured into 2, a little water being added to dissolve the crystals in 1, and the contents of the three beakers were warmed and allowed again to crystallise separately. This operation was continued as long as the uranium salt would hold out, or till the : tests showed that the operations had gone far enough. 16. Tests were made to see how the operations were proceeding. A portion of the crystals from No. 1 beaker was dried and put into a cell. Some mother-liquor from the last beaker was also evaporated and crystallised, and the erystals were put in a cell. The two were placed side by side on a sensitive plate and the action was allowed to proceed for twenty-four hours. On development there was no visible spot beneath No. 1 nitrate, while that beneath the nitrate from the other end of the fractionation was strong and black. 17. The crystals were removed from their cells, ignited to the green oxide, and replaced. Tested again on a sensitive plate the results were similar to those given by the unignited nitrates. . The active ‘substance, therefore, is seen to réside in the mother-liquor. FiO 18. In making photographic tests it is not necessary to take much 416. Sir W. Crookes. of the substance. On an ordinary microscopic slide I put a small drop of liquid from each of the highest five fractionations, Nos. 10, 9, 8,7, and 6. The drops were allowed to crystallise and the slide was aid on a sensitive plate. In twenty-four hours a good impression of No. 10, the highest fraction, was obtained ; a less strong impression of the next, No. 9; a fainter one of 8; a scarcely perceptible one of 7; and no impression at all from No. 6. The slide containing the five crystalline spots was then covered with another glass, and the whole cemented together with Canada balsam and mounted in the manner usual with microscopic slides. When the balsam was dry the slide was put on a sensitive plate. In twenty-four hours a good graduated image was developed. , 19. There are in commerce two kinds of uranium nies one, the eB mercial variety, and another called ‘“‘ purissimum.” I am informed that the “purissimum” is prepared from the former by repeated crystallisation. I purchased some of each of these nitrates and tested them on a photographic plate. The commercial variety proved to be at least twice as radio-active as the “ purissimum ” salt. ,20. Experiments were now instituted with 4 view of obtaining a wholly inactive uranium nitrate. About two pounds of the’salt that had been obtained from the solution in ether was repeatedly crystal- lised, pouring off the mother-liquor each time. This and the next succeeding two lots of crystals (Nos. 1, 2, and 3) were put into cells, and kept on a sensitive plate for seven days. On developing the plate no image could be detected where No. 1 had been; a scarcely perceptible impression could be just detected at No. 2, aa a little stronger impression at No. 3 21. Other methods were perpied for the separation of the adabod substance from uranium. Uranium nitrate fuses at a moderate temperature, and after some time it becomes darker, and nitrous fumes come off. Finally, the mass becomes semi-fluid, and will not run. The operation is then stopped, and the mass transferred to water; the undecomposed nitrate is dissolved out, leaving an insoluble basic nitrate. The basic nitrate is of an orange-yellow colour, easily dissolved in nitric acid to again form the normal nitrate. By using this..method of fractionation the active body gradually accumulates towards the basic end. But the method is neither so complete nor so easily effected as the crystallisation method, and therefore I have not pushed it very far. I have, however, proved that the radio-activity of nitrate of uranium can be concentrated by fractionation to the basic nitrate end, the nitrate at the other end being diminished in radio- activity. 22. A highly active uranium nitrate, prepared by fractionation from the part insoluble in ether, was dissolved in water, and ammonia in. excess was added. Yellow ammonium uranate was precipitated. Radio-activity of Uranium. ‘ALT The filtrate was evaporated to dryness, and heated with nitric acid. The yellow precipitate and the residue of the filtrate were put into cells, and laid on a sensitive plate. After twenty-four hours’ action the plate was developed, when it was seen that the whole of the radio-activity resided in the ammonium uranate, the other substance showing nothing. This experiment proves that the active body is precipitated by ammonia, and is insoluble in excess. 23. Another portion of active uranium nitrate was dissolved in water, with an excess of ammonium carbonate. ‘The first formed precipitate almost entirely re-dissolved, leaving a small quantity of insoluble light brown flocculent precipitate. This collected on warming, like alumina. It was filtered off, well washed and dried, and put ina glass cell. The filtrate from the above precipitate was evaporated to drive the ammonium carbonate, when a yellow precipitate came down This was filtered, washed, dried, and put into a cell. These precipitates were exposed for twenty-four hours in a lead screen apparatus. On developing, it was seen that the residue insoluble in ammonium carbonate instantly flashed out black and dense, while the salt precipitated from the ammonium carbonate solution gave a ‘scarcely discernible image. | 24. The action of the precipitate insoluble in ammonium carbonate was so strong that another experiment was tried, exposing the sensitive plate to its action for one hour. On development, the disc of action came out strong and black, although not so black as in the twenty- four hours’ experiment. It was now laid for five minutes on a sensitive plate. Here the action was distinct—about as strong as that given by ordinary uranium nitrate in twenty-four hours. These experiments prove that the active body can exist apart from uranium. If a sheet of thin glass or celluloid is laid on a sensitive plate, and the dried filter-paper with its contents laid on that, and kept down by a weight, an impression is given in as short a space of time as if a glass cell had been used. 25. The radio-active body is not entirely insoluble in ammonium carbonate. A portion of the very active precipitate left after separa- tion from the uranium was dissolved in dilute hydrochloric acid, and an excess of ammonium carbonate added. The precipitate was very brown due to the presence of iron; it was dried and tested. The filtrate was well boiled, and as the ammonium carbonate evaporated a slight precipitate came down. ‘This was collected on a filter and ‘tested by the side of the first precipitate. On developing the plate. the images produced by each precipitate were of about equal intensity. The brown precipitate was digested in very dilute hydrochloric acid in the cold. The iron partially dissolved -before the rest of the sub- stance, leaving the residue decidedly paler in colour.. This pale body 418 a Sir W. Crookes. was just as radio-active as before the partial removal of. the iron. Therefore the presence of iron does not interfere with the activity of the substance. 26. Having thus definitely proved that the supposed He oe of uranium ae its salts is not an inherent property of the element, but is due to the presence of a foreign body,* it is necessary patiently to determine the nature of the foreign body. Several radio-active bodies claimed to be new have already been extracted from pitch- blende, and experiments have been instituted to see if the newly found body UrX had similar chemical properties to those of the older active ae | . Polonium was first mea A photographic plate had a thin ne of celluloid laid on it, and over this a sheet of aluminium foil, 0-05 mm. thick. On this double layer were put two cells, one containing basic polonium nitrate, the other active UrX. Action was allowed to proceed for twenty-four hours, and the plate was then ‘developed. A disc of blackening was seen under where the UrX stood, the action having passed through the glass, celluloid, and aluminium. Under the polonium nitrate no trace of action could be detected. The experiment was repeated, minus the aluminium foil, and the action continued only two and a quarter hours. On development, the UrX. was found to have acted well, while the polonium showed no trace of action. 28. This behaviour of polonium being excentric or contrary to published accounts,j I put some polonium nitrate in a very thin gelatine capsule, and laid it for eight hours on a sensitive plate. No trace of an image could be seen on development. The same polonium nitrate was put in a watch-glass, and the sensi- tive plate put over it face downwards, so that it might be exposed * For the sake of lucidity the new body must have a name. Until it is more tractable I will call it provisionally UrX—the unknown substance in uranium. + “The rays emitted by compounds of polonium render barium platinocyanide fluorescent . . . To make the experiment, place on the active substance a very thin sheet of aluminium, and on this a thin layer of barium platinocyanide; in the dark the barium platinocyanide appears feebly luminous over the active substance” (M. and Mdme. Curie and M. Bémont, ‘Comptes Rendus,’ vol. 127, p. 1215; ‘Chem. News,’ vol. 79, p. 1). ‘Polonic rays act on sensitive plates. The substance we call sulphide of polonium gives a good impression after only three minutes, and there is a decided action noticed after even half a minute” (Mdme. Curie, ‘Revue Générale des Sci.,’ January 30, 1899; ‘Chem. News,’ vol. 79, p. 77). In the same paper the authoress, after describing the power pos- sessed by a polonium compound to excite phosphorescence, says: “the rays emitted by this latter body have traversed the aluminium and excited the fluore- scence of the platinocyanide above it.” It is evident from the above extracts that I was justified in thinking that polonium rays were not entirely as by thin aluminium, glass, or celluloid. Radw-activity of Uranium. 419 to the direct emanations from the polonium nitrate. On the top of the plate was laid a sheet of lead to press it tight to the edges of the watch-glass. | 1 The exposure was continued for twenty-eight hours. On develop- ing, a strong action was seen, strongest in the middle where opposed to the thickest part of the heap of polonium nitrate, and weaker towards the edge. A well-marked action took place all over the plate exposed to the interior of the watch-glass, but it was sharply cut off at the edges. This confirms the previous results—that the emanations from polonium are of a different character to those from radium or UrX, both of which pass through glass, aluminium, and lead. 29. Another property of polonium sharply distinguishing it from UrX is volatility. The discoverers first obtained it by subliming pitchblende in vacuo. Afterwards they used this property to separate it from bismuth, the polonium and the bismuth sulphides depositing at different parts of the hot tube. A strongly radio-active compound of UrX was ignited in a blowpipe flame with the addition of a drop of sulphuric acid. Its radio-activity, on a sensitive plate, was not diminished by this treat- ment. This experiment was tried several times at increasingly higher temperatures, and always with the same result. 30. Polonium is precipitated by sulphuretted hydrogen, in an acid solution. An acid or neutral solution of UrX is not precipitated by this reagent. Therefore I am justified in saying my UrX is not polonium. 31. But it is not so easy to settle whether UrX is distinct from radium, although many arguments point to its not being radium. The discoverers of radium give several of its chemical properties, and in most of these UrX and Ra are entirely different. Thus, radium sulphate is said to be insoluble in water and acids, while UrX dissolves easily to a clear solution in dilute sulphuric acid. Radium salts are said not to be precipitated by ammonium sulphide or by ammonia, while UrX is precipitated by both. 32. It was hoped that doubtful points might be settled conclusively by the spectrum, as both radium and polonium give well-defined and characteristic lines, especially in the ultra-violet part of the spectrum where I have chiefly worked. M. Demarcay* has given a list of some of the principal lines in the radium spectrum between the wave-lengths 3649°6 to 4826°3, the one at 3814-7 being very strong, and those at 4683-0, 4340°6, and 3649-6 being next in intensity. He draws special attention to the line at 3814-7 as the line showing first in a compound poor in radium. In none of my UrX compounds have I been able to detect a trace of this line; on the other hand I have * “Comptes Rendus,’ vol. 124, p. 716; ‘Chem. News,’ vol. 80, p. 259. 420 7 Tee Sir W. Crookes. failed to photograph this line in products which I know contain radium. The reason is my radium compound is too weak. M: Demar¢ay says the line is scarcely visible with a radium compound only sixty times as active as uranium. My substance containing radium was still weaker, judging from its action on a photographic plate. 33. The same reasoning applies to polonium. With polonium I have obtained strong lines in the ultra-violet, but I can detect none of them in the spectrum of my compound of UrX. sos that I can see are lines belonging to— Platinum (from the poles), Uranium, Calcium, Aluminium, and afew of the strongest air lines, besides a large number of faint lines difficult to identify. 34. Spectrum experiments having failed to show a difference —— radium and UrX, it was thought that possibly some information might be gained by submitting them to the radiant matter test, which has proved so fruitful in its application to the yttrium earths. Some of the most active UrX was put in a tube furnished with a pair of terminals, and it was exhausted to a high point, heat being appiied during exhaustion. Simultaneously a self-luminous radium compound was sealed in a vacuum tube and exhausted, heat being likewise applied. When fully exhausted a strong induction spark was passed through each tube. The UrX compound phosphoresced of a fine blue colour. In the spectroscope no discontinuity could be seen in the spectrum of the phosphorescent light. Under the influence of the induction spark, the radium compound phosphoresced of a luminous rose-colour, showing in the spectroscope a concentration of light in the red-orange, and a very faint citron band, due to a trace of yttrium, probably an impurity. 35. A powerful radium compound and one of UrX, each in a glass cell, and a paper tray full of polonium sub-nitrate, were placed side by side, and a strip of white card was put as a reflector at the back. In front a photographic camera was arranged so as to throw full-sized images of the polonium, UrX, and radium compounds on a sensitive plate, and the whole was kept in total darkness for five days. On development the image of the radium with the containing bottle was visible, but not a trace of image from the polonium could be seen- This confirms previous observations that the radiations from polonium will not pass through glass. Those from radium and UrX easily penetrate glass and other media (28). 36. Recently claims have’ been put forward for the existence of a Radio-activity of Uranium. i eaog third radio-active body in pitchblende. In the ‘Comptes Rendus’ for October 16, 1899, and April 2, 1900, M. A. Debierne describes a radio- active body, to which he gives the name of “Actinium.” At first he said “actinium” showed the principal analytical. properties of tita- nium, but later he describes it as not resembling titanium in all its reac- tions. M.Debierne gives many reactions of the new substance, and in some instances they are like those of radium. But he qualifies them by the statement that they cannot yet be considered as belonging definitely to the new radio-active substance, because up to the present it has not been obtained sufficiently concentrated. He believes rather that these reactions should be looked upon as the result of retention, analogous to that of iron oxide by barium sulphate. He says that the chemical reactions of the most active substance which he obtained, together with its spectroscopic examination, showed that it chiefly consists of thorium. He cannot, however, be sure that it resembles thorium in all its reactions. 37. Experiments have been commenced to see if it is possible to separate thorium compounds into an active and an inactive body. A strong solution of thorium sulphate was slightly acidulated with sul- phurie acid, and gradually raised to the boiling point. A copious precipitate of sulphate came down, and was filtered hot. The pre- cipitate was dissolved in cold water, and the solution re-heated, when a precipitation of the sulphate again occurred. The mother- liquor from one crystallisation was added to the crystals from another in the systematic manner adopted in fractionation, and when the opera- tions had proceeded some time a test was made on the “head” and “tail.” A small quantity of solution from each was evaporated to dryness and strongly ignited before the blowpipe. The two lots of earth were put in cells and a sensitive plate exposed to their action for seventy-two hours. On development not the slightest difference could be detected between the impressions produced by either of the fractions. I next tried partial crystallisation of thorium nitrate, fractionating it in the way already described in the case of uranium (15). Great difficulties were here encountered, owing to the tendency of a strong solution of thorium nitrate to remain supersaturated for several days, when it would suddenly crystallise to a solid mass. After some weeks, however, six fractionations were effected, and tests were made on the first and last of the series. The sensitive plate was exposed to their action for 120 hours. On development, the fraction at the first end (crystals) gave a very feeble action, while that at the other end (mother-liquors) gave an impres- sion about three times as intense. This points to the possibility of separating from thorium its radio-active substance. By the kindness of Dr. Knofler, of Berlin, who makes thorium VOL. LXVI. Zt 422 hadio-activity of Uranium. nitrate by the ton, I have at my disposal some specially prepared thorium nitrate, which is chemically pure. Thoria prepared from this, tested on a sensitive plate, gave a feeble impression in 120 hours. 38. In the present state of our knowledge of these radio-active bodies it is safest to retain an open, or even a slightly sceptical, mind. We recognise them mainly by the photographic and the electrical tests—reactions which are so sensitive that they give strong results, even when the active body is present in too small a quantity to be detected by its spectrum—one of the most delicate of tests. Knowing the tendency of ordinary chemical bodies to be carried down when a precipitate is formed in their presence, even when no question of sparse solubility is involved, it is not surprising that radium and actinium, to say nothing of UrX, appear to simulate elements which may ultimately prove to be very different from them in chemical characters. For instance, UrX dissolves easily in dilute sulphuric acid, and, I have reason to believe, forms a soluble sulphate; still, when chloride of barium is mixed with it and precipitated as a sulphate, I invariably find strong radio-activity in the precipitated sulphate as well as in the filtrate from the barium sulphate. To adduce a simile from my previous researches, the first surmises as to the chemical characteristics of the bodies now known to be yttrium and samarium, were widely different from reality. The differences were entirely due to the perturbing cause which is active in the present case—the tendency of the bodies to be carried down and entangled in precipitates, where, according to ordinary chemical Jaws, they ought not to occur; and to the extreme delicacy of the radiant matter test, which in the case of samarium detects one part in 24 million parts of calcium, and in the case of yttrium detects one part in the presence of a million parts of extraneous matter. 39. The radiographic test for these active bodies presents another point to be borne in mind. Other tests for the presence of an element either act quickly, or do not act at all, with a comparatively narrow - margin of debateable land where the indications of the test may be doubtful. Here, however, the testiscumulative. Like an astronomer photographing stars too faint for his telescope to disclose, he has only to expose the plate for a sufficiently long time and the star reveals itself on development. So, in the case of radio-active minerals or precipitates, if no action is apparent at the end of an hour, one may be shown after twenty-four hours. If a day’s exposure will show nothing, try a week’s. Considering my most active UrX does not contain sufficient of the real material to show in the spectrograph, yet is powerful enough to give a good impression on a photographic plate in five minutes, what must be its dilution in compounds which require an hour, a day, or a week to give an action ? Sir Wm. Crookes. Roy. Soc. Proc. Vol. 66, PI. 5. FIG. 2. FIG. 1. i iaelia Seae Proceedings and List of Papers read. — 423 DESCRIPTION OF PLATE 5. Fie. 1.—Photograph taken by daylight of a cut and polished surface of pitch- blende. Fie. 2.—Radiograph impressed in the dark by the same surface, showing the portions (white) emitting radiant energy. The luminous parts are pitch- blende, the dark parts are felspar, quartz, pyrites, &c. (see para. 5). May 31, 1900. | The LORD LISTER, F.R.C.S., D.C.L., President, in the Chair. Dr. Robert Bell (elected 1897) was admitted into the Society. A List of the Presents received was laid on the table, and thanks ordered for them. The President announced that the subscription portrait of Sir John Evans, painted by Mr. A. S. Cope, A.R.A., had been handed over for the acceptance of the Society by the Portrait Committee. The following Papers were read :— I, “ Paleolithic Man in Africa.” By Sir Jonn Evans, K.C.B., F.R.S. II. “On the Estimation of the Luminosity of Coloured Surfaces used for Colour Discs.” By Sir W. DE W. ABNEY, K.C.B., F.B.S. Ill. “ The Sensitiveness of Silver and of some other Metals to Light.” By Major-General J. WATERHOUSE, I.S.C. Communicated by Sir W. DE W. ABNEY, K.C.B., F.R.S. IV. “ The Crystalline Structure of Metals. Second Paper.” By Pro- fessor EwInG, F.R.S., and W. RoSENHAIN. VY. “ Vapour-density of Bromine at High Temperatures. Supple- mentary Note.” By Dr. E. P. PERMAN and G. A. 8. ATKINSON, Communicated by Professor RAMSAY, F.R.S. VI. “ Influence of the Temperature of Liquid Hydrogen on Bacteria.” By Dr. A. MACFADYEN and 8. RowLAND. Communicated by Lorp LIsTER, P.R.S. The Society adjourned over the Whitsuntide Recess to Thursday, June 14. VOL. LXVI. 2K 494 Dr. Paul Ehrlich. CROONIAN LEcTURE—“K,), the first action of radiation would be to produce a diminution of resistance. This diminution will continue to be exhibited till the maximum amount of B variety is produced. The further action of radiation now will be to reconvert B into A; but as K, Conductivit h ee ee ee we ee ae we ww ew wee we ee ew we ow oo Time of Exposure. Fie. 8.--Curve showing the Effect of Heat and Mechanical Vibration on a fatigued Iron-filings Receiver. * Application of heat. + Application of a tap. effect of mechanical disturbance a gentle tap was given to the receiver, and at once there was produced an increase of resistance due to the transformation of B into A, the receiver regaining its sensitiveness by the transformation. The action of radiation was continued, and after a few reversals the substance once more arrived at the third fatigued state, 0”. The process described above could be repeated any number of times. Effect of Heat and Mechanical Disturbance on a Negative Fatigued Substance, Experiments similar to the above were carried out with an arsenic receiver. From the curve given below (fig. 9) it is seen that the reaction of the negative substance is in every respect opposite to that of a positive substance. It will be noticed that the same cause—i.c., heating or tapping—produces, as necessary consequences of the hypo- theses previously stated, two opposite reactions in the two classes of substance. | have been able to verify this deduction by observations with nearly a dozen different substances, and have not, so far, come across any to contradict it. It thus appears that tapping restores the sensitiveness not by the separation of the electrically-welded particles (in which case tapping ought to have produced an increase of resist- 472 Prof. J. C. Bose. On Electric Touch and the Molecular ance in both the classes of fatigued substance), but by removing the strain in B and thus converting it into A. The effect of electric radiation is thus to produce rearrangement of atoms and molecules in a substance; so does light produce new atomic and molecular aggregation in a photographic plate—a subject to be dealt with in detail in a future paper. Some of my audience at the Royal Institution (January, 1897) may remember my attempt at explaining the action of the so-called coherers (which, perhaps, may be better described as “ molecular receivers”) by analogy with the photo- graphic action. I had then no proofs for the assertion. I have since been able to obtain experimental evidence that the two phenomena are Time of Exposure. Fie. 9.—Curve showing the Effect of Heat and Mechanical Vibration on a Fatigued Arsenic Receiver. 6 Effect of heat. b’ Effect of tapping. identical. The coherer may therefore be regarded as a linear photo- graphic plate; since we are more likely to understand the complex photographic action from the consideration of the much simpler action of electric radiation on elementary substances, where the effects are not complicated by secondary reactions, a photographic plate. may be regarded as merely an assemblage of “molecular receivers.” I hope also to prove that nearly all the detectors of radiation are molecular receivers in reality. The investigation of this aspect of the subject has given me some extraordinary results; they seem to connect together many phenomena which at first sight do not seem to have anything in common. Another interesting question, the consideration of which has for the present to be postponed, is, Why is it that the sensitiveness is so marked in discontinuous metallic particles? In other words, Why is the phenomenon mainly one of skin or touch? Is the phenomenon wholly unknown in continuous solids ? The experiments described in the present paper show :— Changes produced in Matter by Electric Waves. 473 (1) That ether waves produce molecular changes in matter. (2) That the molecular or allotropic changes are attended with changes of electric conductivity, and this explains the action of the so-called coherers. (3) That there are two classes of substances, positive and negative, which exhibit opposite variations of conductivity under the action of radiation. (4) That the production of a particular allotropic modification depends on the intensity and duration of incident. electric radiation. (5) That the continuous action of radiation produces oscillatory changes in the molecular structure. (6) That these periodic changes are evidenced by the corresponding electric reversals. | (7) That the “fatigue” is due to the presence of the “ radiation product,” or strained B variety. (8) That by means of mechanical disturbance or heat, the strained product can be transformed into the normal form, and the sensitiveness may thereby be restored. The method described above of detecting molecular changes is extraordinarily delicate, and is full of promise in many lines of inquiry in molecular physics. It is also seen that the phenomenon of contact sensitiveness, contrary to previous suppositions, is perfectly regular. There is no capriciousness in the response of sensitive substances to the external stimuli, which may be mechanical, thermal, or electric. The curves given above show it ; but they fail to give a fair idea of the rich- ness and variety of the molecular phenomena, seen as it were reflected in the fluttering galvanometer spot of light; of the transitory varia- tions, of the curious molecular hesitation at critical times as to the choice of the structure to be adopted, and of the molecular inertia by which the newly-formed structure is carried beyond the position of stability, and the subsequent creeping back to the more stable position. The varieties of phenomena are unlimited, for we have in each sub- stance to take account of the peculiarity of its chemical constitution, the nature of its response to ether waves, the lag and molecular viscosity. All these combined give to each substance its peculiar charac- teristic curve; it is not unlikely that these curves may give us much information as to the chemical nature and the physical condition of the different substances. I am at present trying to arrange an apparatus which will, by means of the pulsating galvanometer spot of light, automatically record the various molecular transformations caused by the action of external forces. Before concluding, I take this opportunity of expressing my grateful acknowledgments to the Royal Society for the encouragement J 474 Dr. G. L. Johnson. Contributions to the received from the Society for the last five years during which inves- tigations on Electric Radiation have been in progress at the Presidency College. I may say that the difficulties have been very numerous and disheartening, and that without this encouragement the work which it has been my good fortune to carry out would in all probability have remained unaccomplished. The Government of Bengal has also been pleased to evince a generous interest in these investigations. My assistant, Mr. Jagadindu Ray, and my pupils, Messrs. P. K. Sen, B.A., and B. C. Sen, B.A., have rendered me active assistance. | “Contributions to the Comparative Anatomy of the Mammalian Eye, chietly based on Ophthalmoscopic Examination.” By GEORGE Linpsay JouNson, M.D., F.R.C.S. .Communicated by Hans Gapow, F.R.S. Roce May 7,—Read ae Ais 1900. (Abstract.) Observations were made on the eye of the living animal, 181 dif- ferent species being examined, and frequently several individuals of the same species. The species comprise representatives of all the Mam- malian orders except the Cetacea and Sirenia. The conclusions arrived at can be summed up as follows :— The colour of the Fundus ocult in animals devoid of a Tapetum is mainly determined by reflection from the choroidal pigment; in those with a Tapetum cellulosum (Carnivores) by the colour of the retinal pigment ; in those with a Tapetum fibrosum (Ungulates) by the structural colour of the Tapetum modified by the colour of the retinal pigment. All the animals examined may be classed under three types —red, yellow, and green. The vascularisation of the retina can be summarised as follows :— _ 1. Indirect supply by means of osmosis from the vessels of neigh- bouring parts. A. Hyaloid supply. (a) The corpus vitreum is nourished by a processus falciformis, the hyaloid vessels lying well inside the corpus vitreum (Hlasmobranchs). (d) The hyaloid vessels spread over the surface of the corpus vitreum, being in consequence in the immediate vicinity of the retina (¢.g., holosteus and many tele- osteous fishes). Hereto belong also the Amphibia and most of the teptiles devoid of a pecten. B. Choroidal supply. This is probably the chief supply of the retina in those animals which possess a well- developed pecten (most Sauropsida), but are devoid of superficial hyaloid vessels. This choroidal supply by osmosis is also with cer- tainty demonstrated in the Mammalia for at least part of the thickness of the retina. 2. Direct supply. A. From the superficial hyaloid vessels. This is Comparative Anatomy of the Mammalian Eye. 475 known to be the case when the hyaloid vessels are directly continued into the retina, where they produce two vascular layers. B. From special retinal vessels cumulating in the art. centralis. This mode is restricted to the Mammalia and some of the Snakes. The vessels of the falciform process of the fishes and the central hyaline artery, wherever this occurs, are essentially the same. The faleiform process and the pecten are analogous, but not homologous, structures. In Reptiles and Birds the hyaloid artery is superseded by a new development, viz., the Pectinal system. In some of the lower Mammalia both systems actually occur side by side, but both are rendered unnecessary by the development of a third system of supply, viz., special retinal vessels, which ultimately culminate in the poses sion of an art. and vena centralis retine. Some of the normal conditions observable wm certain animals closely resemble those which we find in Man as congenital defects or vestigial relics. 1. Membrana nictitans. A fully developed nictitating membrane active enough to sweep the whole cornea, exists only in the Ungulata, and not even throughout this order. In the Carnivora and Marsupials it is much less developed, whilst throughout the Primates, Rodents, Eden- tata, and Echidna it is still more reduced, and, with rare exceptions, entirely without movement. The primary use of this third lid, viz., that of cleaning the corneal surface, is lost within the class of the Mammalia, and seems to serve chiefly to protect the eye in the animals which graze and poke their heads down into the long and sharp grass. 2. The retractor muscle of the eyeball is of frequent occurrence, chiefly in Marsupials, Edentates, Rodents, and Ungulates, z.¢., in the lower orders of Mammals. 3. Opaque nerve fibres. All stages of opacity occur congenitally in Man, and are to be found normally throughout the Mammalia. Opaque nerve fibres are most marked in some of the Rodents and Marsupials. 4. Physiological cup and congenital discoloration of the disc frequently occur in Man. An appearance similar to the physiological cup occurs in all the Felide, and in a con- siderable number of the other Carnivora; also in the Flying Squirrels and some of the other Rodents. White and grey discs occur normally in a number of animals widely separated in classification, such as the Skunk, Rhinoceros, Porcupine, Armadillo, and Echidna. 5. Structures protruding from the disc into the corpus vitreum. A. Persistent hyaline artery. This congenital defect in Man is found as a normal condition in nearly all the Ruminants and in a large number of Rodents. B. Vestiges of a pecten. In some of the Rodents, more especially in all the Agoutis, a button-shaped vascular pigmented rudimentary pecten protrudes from the disc into the vitreous. It is remarkable to find in the Mammalia a relic of this Sauropsidan organ. In a number of Marsupials vascular protuberances from the disc into the vitreous occur in different forms. 6. Colobomata. The papillary ‘ 476 Dr. G. L. Johnson. Contributions to the coloboma (Fuchs’ Coloboma) has its analogy in a white or coloured scleral ring, which is normally met with in a large number of animals. 7. Retinitis pigmentosa. In the Galagos and Lorides a spreading of pigment occurs circumferentially in the retina, which greatly resembles Retinitis pigmentosa. If these nocturnal animals are exposed for prolonged periods to daylight the pigment advances concentrically, similar to the manner in which it progresses in Man, so that the animals gradually go blind. 8. Visible choroidal vessels and stippled fundus. Visible choroidal vessels occur in most of the Simiz below _Hylobates, and in a number of the other orders. They are most marked in the Macropodide, and some of the other Marsupials, which present the appearance observable in the extreme cases of the analogous congenital defect in Man. Stippled fundi are found in the feline Douroucouli and in the Lemurs, an appearance occasionally met with in Man. 9. Ectropion of the Uvea. In a number of the Ungulates, which have large oval pupils, pigmented excrescences of the iris are met with, and these evidently serve to screen the eye against glare, since their pupils only contract moderately to light. In the Hyracoide we meet with a distinct specialised organ, which can be projected from the iris towards the cornea, like a small screen, and this I propose to call the ‘“‘ Umbraculum.,” The divergence of the optic axes follows the classification to a marked degree. The higher the order the nearer the axes approach parallel vision. Parallel vision with the power of convergence only occurs in those animals which possess a true macula, viz., Man and all the Simie. In other words, convergence appears to be the necessary out- come of a macula. This macula, which is bounded by a reflex ring, exists in all the Simize without exception, and in no other Mammals, so that it ceases with the last of the Simiz. If we eliminate the domestic animals in which the refraction varies over considerable limits in all directions, we find throughout the Mammalia, with a few notable exceptions, vision is hypermetropic. The eyes of amphibious and marine Mammals are adapted for vision in two ways. Those which live in fresh water have immensely developed ciliary muscles and proportionally increased accommodative power, enabling them to compensate for the loss of the refractive power ot the cornea when the eye is submerged. In the marine Mammals, i... Pinnipedia and Cetacea, not only is this ciliary muscle greatly developed, but there is always a large area of the cornea which is flattened in the horizontal meridian, producing an extraordinary degree of astigmatism. Binocular Vision.—It seems that if Mammals below the Simiz have binocular vision, they do not rely entirely on it. With the exception of Man and the Simia, Mammals very rarely move their eyes for the purposes of vision, but move their heads instead. Comparative Anatomy of the Manmalan Eye. ATT In all the Mammals below the Simiz which have no macula we find a larger sensitive area. Sensitive areas of restricted dimensions, omitting — those cases in which the area is limited to a macula, exist in the Carnivora, in which order the divergence is not great. In the Ungu- lates, Rodents, Edentates, and Marsupials, where we find great diver- gence of the axes, large cornee, and nearly spherical lenses, the sensitive areas are larger, and probably the degree of difference in perception over such areas, compared with the more peripheral parts, is but little. The great transparency of the retina and the extreme brilliancy of the reflecting surface of the choroid in the vast majority of Mammals, and an extraordinary prevalence of colours of every hue, lead one irresistibly to the conclusion that the rays of light do not form an image on the retina as usually taught, but that the image is formed behind the retina on the brilliant surface of the Tapetum or fusca pigment layer of the choroid, and is then reflected back on to the terminals of the bacillary layer. This arrangement for vision certainly bears a close resemblance to Lippmann’s method of obtaining coloured negatives. He obtained negatives in natural colours by placing a reflecting mercury surface in direct contact with the sensitive film, thus reflecting the light which had traversed the film on to the particles of sensitised silver. In the eye the light passes through the nearly transparent retina (which is analogous to the photographic film) to be reflected from the Tapetum, or choroidal pigment, on to the terminals of the retinal elements (which may be compared to the particles of silver haloid), In Lippmann’s device the colours are produced by interference. If we venture to carry our analogy still further, we may presume the same occurs in the eye. One difference between the two methods is that in nature the reflecting surface is always coloured, and ‘only reflects a portion of the incident light. The colour of the fundus, however, is remarkable for the absence of blues and violets and the great prominence of red, yellow, and green colours. — Yellow and orange are the prevailing colours in nocturnal animals. The peripheral area, which is characteristic of animals possessing a Tapetum, is usually dark brown, and reflects but feebly. It is probably nearly insensible to light, as it never occurs in animals having great divergence of the optic axes. The eye is no exception to the rule that domestication greatly increases variability. The colour of the Fundus oculi of domesticated races differs not only from that of the wild species from which the races are supposed to be derived, but the colour varies also individually, an occurrence almost unknown in wild species. The influence of domestication is also indicated by the frequent occurrence of myopia and astigmatism. Myopia is almost unknown in wild animals, but it may occur in wild specimens which have long been kept in captivity. 478 Mr. L. Hill, The Influence of Increased Atmospheric Although no sound classification can be based on one single organ, a striking concordance exists between an attempted arrangement of the Mammalia according to the Fundus oculi and the most modern classifi- cation. The cases of disagreement are wonderfully few. These are restricted to the following :— Chrysothrix leans towards the Arctopitheci. I find it necessary to separate the Galagos from the rest of the Lemurs—at least, as a sub- family. In the smaller Carnivores it is advisable to establish a separate family, the Cynictide, to include the otherwise viverrine genera, Cynictis and Galictis, together with Mephitis, hitherto placed with the Mus- telide. - The Sciuromorpha should be divided into Sciuride and Ptero- myidee, and Castor should decidedly be removed into the Hystricomorpha group, perhaps into the vicinity of the Octodontide. The Bats rank very low so far as the eye is concerned, possibly on account of their nocturnal habits. Among the Marsupials the Diprotodontia are decidedly lower than the Polyprotodontia chiefly on account of the high degree of development of the eyes of the Didelphide and Dasyuride. Since we meet with genera of the lowest type along with others of the highest type of retinal vascularisation, and again some without and others with the additional relic specialisation of a Tapetum, it follows that the details of the vascularisation and of the Tapetum have been developed independently in the various main branches of the Mammalia. In fine, the whole Fundus oculi affords a striking illustration of the working of progressive evolution, an example all the more valuable, since it illustrates the direct modifying effect of external factors upon a highly specialised organ—in the present case the continued influence of light upon the eye. “ The Influence of Increased Atmospheric Pressure on the Circula- tion of the Blood. (Preliminary Note.)” By Lronarp HILL, M.b. Communicated by Dr. Mort, F.R.S. Received March 22,—Read May 17, 1900. Paul Bert* recorded the arterial pressure in two dogs which he introduced, together with the kymograph, into a chamber, and sub- mitted to a + pressure of 53 cm. Hg. The atmospheric pressure was raised to this height in the course of three-quarters of an hour. The ‘mean arterial pressure rose in one dog 16 mm. Hg., in the other 46 mm. Hg.; the pulse frequency fell in the first from 216 to 200, and the respiration from 41 to 29 per minute. The respiratory oscil- lations of blood-pressure became increased. Bert ascribes the results * ‘Pression Barom#trique,’ Paris, 1878, p. 838. Pressure on the Circulation of the Blood. 479 as due to the diminution of the volume of the intestines—which results from the compression of the intestinal gas—and the consequent in- ereased play of the intrathoracic organs. He was confirmed. in this opinion by the fact that the substitution of oxygen for air made no difference in the results. | A. Smith* attributes the symptoms which arise in caisson workers partly to the effect of the increased pressure on the nervous tissue, and partly to the congestion of the blood in the neural axis. He sup- poses that the blood is driven by the compressed air from the periphery to the cranio-vertebral cavity. This mechanical explanation of caisson disease is contrary to the supposition which, theoretically, seems cor- rect, viz., that the atmospheric pressure is equally transmitted by the blood to all parts. I have put the matter to the test of experiment. Method.—An anesthetised dog or cat is placed at the bottom of an autoclave after the insertion of a cannula in the carotid artery, and, in some cases, of one.in the vena cava superior, has been carried out. The cannule are connected with Hg. monometers, and these, together with a slow-movement recording drum, are.also placed in the autoclave. The drum is then started and the monometers set to record the arterial and venous pressures. The lid of the autoclave is next screwed on, an oxygen bottle connected with the inlet tube and the pressure in the autoclave rapidly raised to + two atmospheres (30 lbs.). The outlet tap is finally opened, the pressure rapidly lowered, the lid taken off, and the record observed. ‘The times of increasing and diminishing atmospheric pressure are noted with a watch. The whole operation only takes two or three minutes. fesults—Although exposed to this rapid change of atmospheric pres- sure, the circulation of the blood scarcely varies. The arterial pressure and venous pressures either continue at the same level or very slightly fall. The respiratory oscillations are increased, and the pulse becomes slightly less frequent. The gas injected contains, roughly, 80 per cent. oxygen, and thus the oxygen tension is raised from 21 per cent. to about 190 per cent. of an atmosphere. The results are the same whether compressed air or oxygen are employed. Conclusion.—The mechanical congestion theory of caisson disease is untenable. The expenses of this research have been met by a grant from the Royal Society Government Grant. * Article “Caisson Disease,” ‘ Allbutt’s System of Medicine,’ vol. 7, p. 38, 1899. A480 Mr. L. Hill. On Cerebral Anemia and the 4 Ld “On Cerebral Anemia and the Effects which follow Ligation of the Cerebral Arteries.” By Lronarp Hitt, M.B. Com- municated by Dr. Mort, F.R.S. Received March 22,—Read May 17, 1900. (Abstract.) 1, Cerebral Anemia produced by Immobilisation in the Erect Posture, Many hutch rabbits when immobilised in the erect posture become convulsed and after a short period of time die (10’—20’) from failure of respiration.* The blood, owing to its weight, congests within the abdominal vessels, while the abdominal viscera drag on and so kink the vena cava inferior. The heart, therefore, gradually empties, and the cere- bral circulation ceases. These results are due to the flaccid and atonic nature of the abdominal wall of the hutch rabbits. Chloralisa- tion hastens the onset of death by dilating the arteries and by stopping the convulsions, for the spasms help to return the venous blood to the _ heart. Compression of the abdomen by a bandage, or immersion of the animal in a bath of water up to the neck, entirely prevents the onset of symptoms. In the case of the bath the hydrostatic pressure of the water outside balances, but not completely, the hydrostatic pressure of the blood within. At the same time the water causes the viscera to float upwards, and so removes the kinking of the vena cava inferior. Wild rabbits, owing to the better tone of their abdominal muscles, are not affected by immobilisation in the erect posture until after the lapse of some hours, and the same is the case in respect of dogs, cats, and monkeys. When the tone of the skeletal and vascular muscle becomes exhausted in these animals, owing to exposure, shock, &c., or is abolished by anesthetics, the blood congests to the lower parts. Death finally results from cerebral anemia. Immersion in a bath or compression of the abdomen has the same restorative effect on the circu- lation of these animals as on that of the hutch rabbit. Intense congestion and cedema of the lower parts, accompanied by thirst, is said to occur in men under like conditions, and death no doubt results from cerebral anzmia. Hutch rabbits, when thrown into syncope by immobilisation in the erect posture, recover almost immediately on their return to the hori- zontal position, or on immersion in a bath, or on compression of the abdomen. The animals when returned to the horizontal posture may * This fact was noted by Salathé, Lffects which follow Ligation of the Cerebral Arteries. 481 be paralysed for a few minutes, and tumble and walk on the back of the fore-paws, but these symptoms quickly disappear even though the pupils’ light reflex may have been abolished for 15’—20’. 2. Cortical Excitability after Ingation of Cerebral Arteries. The author has produced contra-lateral clonic spasms in himself by sudden compression of one carotid artery. Consciousness of the cortical discharge arises from the sensations received from the parts in movement. The cortical discharge itself is unaccompanied with consciousness. The cortex cerebri of dogs remains excitable and even hyper- excitable to electrical excitation after both carotid and vertebral arteries have been tied. The brain under these conditions is supplied with blood by the branches of the superior intercostal artery which enter the anterior spinal artery. The brain is, however, rendered profoundly anzmic by the operation, and the animals are in consequence rendered more or less demented, anesthetic, and paralysed. The paralysis results from a block established by the anemia in the sensory projection, and association fields of the cortex cerebri for the motor cells are unaffected, in so far as not only purposive movements but typical fits can be excited on stimulating the “motor area.” The “motor centres ” are clearly not autonomous, and these experiments confirm the previous deductions concerning the origin of the paralysis which was obtained by Mott and Sherrington after divi- sion of the posterior nerve-roots of a limb, and by Exner after circum- vallation of the cortex. Isolation of the cortical motor cells from sensory impulses produces paralysis, although the cells remain directly excitable. In many monkeys the two carotids, and even the two carotid and one vertebral artery, can be tied without lessening the cortical excit- ability. The ligation of all four arteries is followed within one minute by loss of cortical excitability. In some monkeys (especially those in bad condition) ligation of both carotids abolishes the excitability. 3. The Effect of Absinthe after Ligation of the Cerebral Arteries. The injection of absinthe, after ligation of the four cerebral arteries in cats and of the two carotid arteries in monkeys, produces as a rule only extensor rigidity of the fore limbs and dyspneeic respira- tion. No clonic convulsions occur. On loosening the carotids violent clonic and tonic convulsions ensue, and these can be again cut short by re-clamping the carotids. So soon as the carotids are re-clamped, the extensor rigidity reappears. It is deduced from these experiments that clonus is of cortical, and tonus of sub-cortical, and probably of cerebellar, origin. VOL. LXVI. ee 482 Mr. L. Hil On Cerebral Anemia and the 4. The Late Effects which follow Lngation of the Four Cerebral Arteries. These vary in different animals. Almost all rabbits die from failure of respiration within three minutes, after convulsions of an asphyxial type. There occur vagal inhibition of the heart, general vaso-constric- tion, and a high arterial pressure, prior to the failure of the circulatory system. ‘The symptoms are in every way similar to those produced by clamping the trachea. Cats become comatose and die within a few hours from respiratory paralysis, which is gradual in onset. Cheyne Stokes respiration some- times results, and at a later stage long-drawn spasmodic gasps of the diaphragm occur at rare intervals. Extensor rigidity often occurs before death. Cats may survive the ligation of the two carotids and one vertebral artery. Dogs recover from ligation of the four cerebral arteries, some after scarcely any symptoms, others after passing through a stage of dementia, accompanied by paralysis and anesthesia, which lasts three or four days. The dogs during this period of dementia behave exactly like the dogs in which Goltz. produced extensive destruction of the cerebrum. ‘The spatial sensations depending on the nerves of skin, joints, and muscles are no longer brought into association with the sensations which are derived from the higher senses. Reflex defence and locomotor movements alone persist. Monkeys almost all die within twenty-four hours after ligation of the two carotids and one vertebral artery. The animals become soporose and then comatose. Extensor spasms, extensor rigidity, and failure of respiration follow. Monkeys recover without symptoms after ligation of both carotids. . In one monkey, after the two carotids and one vertebral artery had . been tied, there ensued extensor rigidity and profound paralysis and dementia. This animal was kept alive by spoon feeding and continued in the same state. It was killed on the fifth day. The stage of sopor and coma or dementia may not appear in dogs or monkeys for an hour or so after the ligation of the arteries has been effected. On recovering from the anesthetic the animals may at first appear lively and intelligent. , In man the ligation of one carotid artery is not free from risk, while the ligation of both carotids is recognised as a most dangerous opera- tion. ‘The two arteries can be tied successfully at intervals of time. Attention is more particularly drawn to the following conclusions deduced from this research :— | 1. The cerebral circulation of man, in the erect posture, depends on the tone and activity of the skeletal and respiratory muscles. The blood and lymph are returned from the lower parts to the heart by the expressive action of the muscles and constant change of posture. 2. The functions of the brain may continue after a great diminution Effects which follow Ligation of the Cerebral Arteries, 483 in blood supply. ‘This substantiates previous work of the author in regard to the slight metabolism of the brain as measured by the exchange in blood gases. At the same time it does not favour the anemic theory of sleep. 3. The electrical excitability of the motor area of the cortex cerebri persists when the sensory side of the brain is to a large extent para- lysed, and the animals rendered more or‘less demented by profound cerebral anzemia. 4, The functions of the brain rapidly return so soon as efficient anastomosis is established. The period of partial paralysis and dementia lasts in dogs two or three days. Rabbits. recover after the — pupil reflex has been abolished for 15’’—20”. 5. The limits between the degree of anzmia required to produce dementia and that which paralyses the respiratory centre are extremely narrow. For example, monkeys recover without symptoms after ligation of both carotids, but as a rule die after ligation of both carotids and one vertebral artery. 6. There is considerable variation in the number of arteries which can be safely tied in various animals, ¢.g., in man, birds, goats, and horses (Mayer) one carotid ; in monkeys both carotids; in rabbits and cats both carotids and often both carotids and one vertebral ; in dogs both carotids and both vertebrals. 7. The four cerebral arteries can be safely tied in monkeys in successive operations. 8. The cortex cerebri is the place of discharge of clonic convulsions. Tonus is of sub-cortical origin. The clonic stage of an epileptic fit can be cut short by compression of both carotid arteries. Dr. Mott, to whom I am greatly indebted for help and advice in this oo ck has determined by microscopical examination of the anemic brains by Nissl’s method, that— 1. The cortical cells in the brains of the demented animals are — swollen and diffusely stained. The stichochrome granules are absent. The nuelei are swollen. The veins are congested and there may occur hemorrhages in the cortex. 2. The large pyramidal cells are least affected. 3. The changes occur very rapidly after ligation of the cerebral arteries, and disappear synchronously with the recovery of the animals from the stage of dementia. In the case of the monkey, described above, the cerebrum was softened in patches, many of the cortical cells were degenerated, and there were signs of active phagocytosis. No changes in the neurons _were displayed by the Golgi method. The expenses of this research have been met by grants from the Royal Society Government Grant. to kb P 484 Mr. H.N. Dickson. he Circulation of the “The Circulation of the Surface Waters of the North Atlantic Ocean.” By H. N. Dickson, BSe. Communicated by Sir JOHN Murray, K.C.B., F.R.S. Received March 23,—Read May 17, 1900. (Abstract.) In this paper an attempt is made to investigate the normal circula- tion of the surface waters of the Atlantic Ocean north of 40° N. lat., and its changes, by means of a series of synoptic charts showing the distribution of temperature and salinity over the area for each month of the two years 1896 and 1897. The temperature observations discussed (numbering over 16,000) were obtained from the meteorological and hydrographical departments of the countries bordering on the North Atlantic, and special arrange- ments were made with the officers of a number of ships for the con- tinuous supply of samples of surface waters for analysis. The salinity of the samples obtained was determined by volumetric estimations of the amount of chlorine present. Over 4,000 samples were dealt with in this way, and special attention was devoted to ascertaining the accuracy of the methods employed. A large number of estimations were also made of sulphates present in the waters, and the -limits of variation from a definite ratio of chloride salinity to sulphate salinity determined. The specific gravity of over 500 of the samples was determined with the pyknometer, and a formula connecting the results of these deter- minations with the salinities derived from chlorines investigated. The numerical results of the chemical and physical determinations are exhibited in a table, forming a substantial addition to the material available for the discussion of oceanographical problems of this kind. The principal conclusions arrived at with reference to the circulation may be summed up as follows :— 1. The surface waters along the whole of the eastern Biboard of North America north of (about) lat. 30° N., consisting partly of water brought from the equatorial currents by the Gulf Stream, and partly of water brought down by the Labrador current, are drifted eastward across the Atlantic towards south-western Europe, and banked up against the land outside the continental shelf. This continues all the year round, but it is strongest in summer, when the7Atlantic anti- cyclone attains its greatest size and intensity ; and the proportion of Gulf Stream water is greatest at that season. 2. The drifts in the northern part of the Atlantic area are under the control of the cyclones crossing it. The circulation set up accordingly _ reaches its maximum intensity in winter, and almost dies out in summer. In winter the drifts tend to the south-eastward from{the Surface Waters of the North Atlantre Ocean. 485 mouth of Davis Strait, eastward in mid-Atlantic, and north-eastward in the eastern region. In spring and autumn the movement is more easterly over the whole distance, and a larger quantity of water from the Labrador stream is therefore carried eastward. 3. The water banked up in the manner described in (1) escapes partly downwards, partly southwards, and partly northwards. It occupies the whole of the eastern basin of the North Atlantic, and to the north it extends westward to Davis Strait, bemg confined below 300 fathoms depth by the ridges connecting Europe, the Faeroes, Iceland, and Greenland. Above that level it escapes northward by a strong current through the Faeroe-Shetland Channel and between Faeroe and Iceland, and by the two branches of the Irminger stream, one west of Iceland the other west of Greenland. (As it seems desirable that this northerly current should have a distinctive name, it might be well to call it the European stream, and its branches the Norwegian, Irminger, and Greenland streams respec- tively.) The strength and volume of the European stream is liable to con- siderable variation, according to the form and position of the Atlantic anti-cyclone, which causes the amount of banked up water, and the proportions escaping northward and southward, to vary. It is also modified by the strength and direction of the surface drifts in its course. It is, however, always strongest in summer. 4. The Norwegian stream is by far the largest branch of the European, and it traverses the Norwegian Sea and enters the Arctic Ocean. The warm water thus sent northward melts enormous quan- tities of ice, and the fresh water derived from the ice moves southward in autumn, chiefly in a wide surface current, between Iceland and Jan Mayen, which may entirely cover other parts of the Norwegian stream. Part of the surface water also comes southward through the Denmark Strait, but the amount is much smaller, probably chiefly because the melting of the ice is slower, and the channel is longer blocked. The Greenland branch of the European current also causes melting of ice in Davis Strait, but the warm winds from the American con- tinent and the water received from the land are probably more effective in increasing the volume of the Labrador current. 5. The water from the melted ice is spread over the surface of the North Atlantic during late autumn and winter by the increasing drift circulation, and it is gradually absorbed by mixing with the underlying water. 6. The circulation described is liable to extensive irregular varia- tions, corresponding to variations in the atmospheric circulation, 486 Sir John Evans. “ Paleolithic Man in Africa.” By Sir Jonn Evans, K.C.B., F.R.S. Received May 15,—Read May 31, 1900. In April, 1896, just four years ago, I ventured to call the attention of the Society* to some paleolithic implements found in Somaliland by Mr. H. W. Seton-Karr. In doing so, I pointed out the absolute identity in form of these implements with those from the valley of the Somme and numerous other pleistocene deposits in North-western Europe and elsewhere ; and I cited others from the high land adjoining the valley of the Nile and from other places in Northern and Southern Africa. I was at the same time careful to point out that though there could be no doubt as to this identity in form, no fossil mammalian or other remains had been found with these African implements. I did not, however, hesitate in claiming them as paleolithic. Since the publication of my short note, an extensive collection of stone implements formed in Egypt by Mr. H. W. Seton-Karr has been acquired by the Mayer Museum at Liverpool. I have not had an opportunity of examining the specimens, but a detailed accountt of them, with numerous illustrations, has been published by the Director of the Liverpool Museums, Dr. H. O. Forbes. The majority of the implements are of Neolithic Age or even of more recent date, and with the account of these I need not here concern myself; but the author is at considerable pains to dispute my view that the instruments of palzeo- lithic forms belong to the Paleolithic Period. As he says, Mr. Seton- Karr’s statement that he sometimes found spear-heads “ on the ground surrounded by a mass of flakes and chips as though the people had dropped their work and fled,” is very suggestive and important. He adds, however, that “one such occurrence is almost sufficient in itself, I venture to think, to disprove the high antiquity claimed by Sir John Evans for these implements.” Were it certain that the so-called spear-heads were really of palzeo- lithic form, and had the flakes and chips been fitted on to them so as to reconstitute the original blocks of flint, as has been done in the case of undoubted paleolithic specimens by Mr. Spurrell and Mr. Worthing- ton Smith, the question would still remain to be discussed as to the condition of the localities in relation to subiéerial denudation. It is, however, hardly necessary to discuss these points, as some recent discoveries made in Algeria will, I venture to think, go a long way towards settling the question. I propose, therefore, very briefly to state their nature. About sixty miles to the south-west of the town * ‘Roy. Soc. Proc.,’ vol. 60; p. 19. + ‘Bull. Liverp. Mus.,’ II, Nos. 3 and 4 (Jan. 20, 1900); ‘ Nature,’ April 19, 1900, p. 597. Paleolithic Man win Africa. 487 of Oran, and about ten miles to the north of Tlemcen, on the plateau of Remchi, about a mile to the south of the River Isser, lies a small lake known as Lac Karar. It occupies a depression in lacustrine limestone of comparatively recent geological’ date, superimposed on beds of Lower Miocene Age. The level of the water, which is some 15° centigrade warmer than that of the ordinary springs of the dis- trict, and appears to be derived from some deep-seated source, seems to be about 600 feet higher than that of the River Isser. The lake originally filled a much larger part of the depression than it now does, and from its old bed a considerable amount of material has of late years been extracted for the Service des Ponts et Chaussées. This material consists of sand and gravel rich in iron pyrites, in the midst of which lie, pell-mell, bones of animals and stone implements fashioned by the hand of man. These have for some years been diligently collected by M. Louis Gentil, a geologist, and form the subject of a memoir that has just ap- peared in ‘]’Anthropologie’* by my friend M. Marcellin Boule, of the Galerie de Paléontologie at the Jardin des Plantes, Paris. Some 200 specimens of implements have been submitted to him, of various sizes, and all or nearly all of well-known palzolithic forms, including several with a broad chisel-like end, of which examples have been found in the laterite of Madras and the gravels of Madrid. They are for the most part formed of an eocene quartzite, though some smaller speci- mens of the type known as that of “le Moustier” are formed of flint. The facies of these latter is not so distinctly paleolithic as that of the former, of which some, through the kindness of M. Marcellin Boule, are exhibited. The most important part of the discovery is that which relates to the mammalian remains found with the implements. These are of elephant, rhinoceros, horse, hippopotamus, pig, ox, sheep, and certain cervide. I will not detain the Society with the details given in M. Boule’s memoir, but I may call attention to the fact that the elephant is not the African elephant, but one more nearly related to the quaternary or even pliocene elephants of Europe, to which the designation Atlanticus has been given. Some teeth seem closely allied to those of HL. meridionalis and even EL. armeniacus. Having regard to the whole fauna, M. Boule arrives at the conclusion that it is identical with that of the fossiliferous deposits of Algeria, which from their topographical or stratigraphical characteristics have been assigned to the Quaternary or Pleistocene Period. He also cites other instances in Algeria, such as Ternifine and a station near Aboukir, in which palzeolithic implements have been found associated with the remains of a similar pleistocene fauna. _ Altogether, these recent discoveries in Northern Africa tendimmensely * Tome XI, 1900. 488 Influence of Temperature of Iiquid Hydrogen on Bacteria. to strengthen my position with regard to the truly paleolithic character of the implements found in other parts of that vast con- tinent, and I am tempted to bring for comparison some few specimens from South Africa. One of these, found by Mr. J. C. Rickard at the junction of the Riet and Modder twenty years ago, is almost indis- tinguishable from those of the Lac Karar, as is also one from the valley of the Embabaan in Swaziland. But the most remarkable is an implement of typically paleolithic character found in 1873 under 9 feet of stratified beds at Process-fontein, Victoria West, by Mr. E. J. Dunn.* May the day be not long distant when researches for the implements of paleolithic man may again be carried on, and trenches be dug in South Africa for peaceful instead of warlike purposes. “Influence of the Temperature of Liquid Hydrogen on Bacteria.” By ALLAN Macrapyen, M.D., and Sypney Row1anp, M.A. Communicated by .Lorp LISTER, P.R.S. Received and read May 31, 1900. In a previous communication we have shown that the temperature of liquid air has no appreciable effect upon the vitality of micro- organisms, even when they were exposed to this temperature for one week (about — 190° C.).T We have now been able to execute preliminary experiments pro- jected in our last paper as to the effect of a temperature as low as that of liquid hydrogen on bacterial life. As the approximate temperature of the air may be taken as 300° absolute, and liquid air as 80° absolute, hydrogen as 21° absolute, the ratio of these temperatures roughly is respectively as 15:4:1. In other words, then, the temperature of liquid hydrogen is about one-quarter that of liquid air, just as that of liquid air is about one-quarter of that of the average mean tempera- ture. In subjecting bacteria, therefore, to the temperature of liquid hydrogen, we place them under conditions which, in severity of temperature, are as far removed from those of liquid air as are those of liquid air from that of the average summer temperature. By the kindness of Professor Dewar, the specimens of bacteria were cooled in liquid hydrogen at the Royal Institution. The following organisms were employed: Bac. acidi lactici, B. typhosus, B. diphtherie, Proteus vul- garis, B. anthracis, B. coli communis, Staphylococcus pyogenes aureus, Spirillum cholere, B. phosphorescens, B. pyocyaneus, a Sarcina, and a yeast. The above organisms in broth culture were sealed in thin glass tubes * See also a paper by M. E. T. Hamy in the ‘ Bulletin du Muséum d@’Histoire- Naturelle,’ 1899, No. 6, p. 270. t+ ‘Roy. Soc. Proc.,’ February 1, 1900; zdéd., April 5, 1900. Vapour-density of Bromine at High Temperatures. 489 and introduced directly into liquid hydrogen contained in a vacuum jacketed vessel immersed in liquid air. Under these conditions they were exposed to a temperature of about — 252° C. (21° absolute) for ' ten hours. At the end of the experiment the tubes were opened, and the contents examined microscopically and by culture. The results were entirely negative as regards any alteration in appearance or in vigour of growth of the micro-organisms. It would appear, therefore, that an exposure of ten hours to a temperature of about — 252° C. has no appreciable effect on the vitality of micro-organisms. We hope to extend these observations upon the influence of the temperature of liquid hydrogen on vital phenomena, and to make them the subject of a future communication, and to discuss their bearing upon problems of vitality. “Vapour-density- of Bromine at High Temperatures.—Supple- mentary Note.” By E. P. Perman, D.Sc., and G. A. 8. ATKINSON, B.Sc. Communicated by Professor Ramsay, F.R.S. Received April 28,—Read May 31, 1900. The authors regret that they had overlooked a monograph by C. Langer and V. Meyer, entitled ‘“ Pyrochemische Untersuchungen,” containing an account of some experiments on the vapour-density of bromine. Their method was to pass a mixture of bromine vapour and nitrogen into a porcelain tube with capillary ends placed in a furnace in a horizontal position. The bromine and nitrogen were then displaced by a current of carbon dioxide, the bromine being absorbed by potassium iodide solution, and the carbon dioxide by potash solution, while the nitrogen was collected and measured. Temperature was ascertained by displacing the tube full of air in a similar way. By this method Langer and Meyer carried out experi- ments at “‘ Zimmer-temperatur” 100°, 900°, and 1200°. The only results comparable with the authors’ are those at 900°. Their results at this temperature are 5-478, 5-414, 5-433, 5-382, 5°59, mean 5°459 (air = 1), or 78°88 (H = 1). They say that these results indicate that the vapour-density of bromine, even when diluted with eleven times its volume of air, is still normal at 900°. The mean of our results at this temperature is 78°6, and the density read from the curve (p. 17, vol. 66) is 78°8. This close agreement shows that Langer and Meyer’s results really indicate a small amount of dissociation at 900°. Diminution of pressure appears to have little effect at that temperature, as Langer and Meyer used bromine much diluted with nitrogen, while in our experiments there was no decrease of density on reducing the pressure from 767 mm. to 365 mm. It may be noted that Langer and Meyer give the boiling point of 490 Major-General J. Waterhouse. The Sensitiveness bromine as 63°, whereas we have found the boiling point of specially purified samples to be very close to 58:9°. (See Ramsay and Young, ‘ Trans. Chem. Soc.,’ vol. 49, p. 454 e¢ seq.) “The Sensitiveness of Silver and of some other Metals to Light.” By Major-General J. WATERHOUSE, LS.C. (late Assistant Sur- veyor-General of India). Communicated by Sir W. ABNEY,. K.C.B., F.RS. Received April 25,—Read May 31, 1900. During some recent investigations on the Daguerreotype process, the question presented itself as to which of the elements forming the sensi- tive surface of the plate—the silver or the halogens—the sensitiveness was due? Now, although the fact that nearly all compounds of silver, especially the haloids, are more or less sensitive to, and decomposed by, the action of light, has long been known, the sensitiveness of metallic silver itself to light, though observed in 1842, by Moser, has. never been generally fe nented either by chemmees or by pho- tographers. Moser’s Hxperiment.—Before describing my own experiments, it may be as well to give a description of Moser’s experiment taken from the. original paper in ‘ Poggendorff’s Annalen,’ vol. 56, 1842, p. 210. “A perfectly new silver plate was thoroughly cleaned and polished. A black tablet with various excised characters was fixed above it, without touching it, and the whole placed in the sun for two hours or more and directed towards it. After the plate, which naturally did not’ show the least change, was cooled, it was held over mercury, heated as usual to about 60° R. (167° F.). To my great delight a distinct image of the screen was produced in which those parts where. the sunlight (which during the course of the experiments was always weak and changeable) had acted, had attracted a quantity of mercury. This interesting experiment was repeated several times with the same result. Sometimes the plates after having been placed in the mercurial vapours were exposed to those of iodine and then placed in mee sun,, by which the images usually improved.” “If we compare this remarkable fact of the action of light upon surfaces of silver with the above-mentioned phenomena produced by contact, we can no longer doubt that light acts on all bodies, modifying them so that they behave differently in condensing the vapours of mercury. s ° | Polytremacis and the Ancestry of the Helioporide. By J. W. Gruaory, / DSe. Communicated by Professor LanxestzR, F.R.S. ‘ 5 f Gold- Aluminium Alloys. By C. T. Huycock, F.RS., and F. H. Nevinra, On the Association of Attributes in Statistics, with ani ntes from the Material of the Childhood Society, &c. By G. Upny Yous, formerly Assistant Professor of Applied Mathematics, University College, London. Com- municated by Kart Prarson, F.R.S. ‘ : , q : : u) Data for the Problem of Evolution in Man. III.—On the Magnitude of ¥ certain Coefficients of Correlation in Man, &c. By Kart Pearson, E.R.S., University College, London pre a 4 : d : aie toveriiok pe » On the Numerical Computation of the Functions Gy(a), G(x), and J,(@/2). ye By W. Steapman Apis, M.A. Communicated by Professor J. J. 3 Tuomson, F.R.S, . b : é . S A f A f : . . Price Two Shillings. | Fesrvary 9, 1900. | In future, Forms of Acknowledgment of the ‘ Proceedings ’ will only be issued with the concluding No. of each volume. NOTICE TO AUTHORS. | Authors of Papers intended for the ‘ Proceedings’ or ‘ Philo- sophical Transactions’ are urgently requested to send in all drawings, diagrams, or other illustrations in a state suitable for direct photographie reproduction. Sold by Messrs. Harrison and Sons. Royal 4to. Price 1s. LIST OF FELLOWS OF THE ROYAU SOCIETY. 30th November, 1899. No. 495): Zeosonian stag CONTENTS 4p oo 19997 ‘aia PAGE © | Meeting of December 14, 1899, and List of Papers . 44s wit 5 : Ohice Librar) | uy On the Origin of certain Unknown Lines in the Spectra of ° ofthe 6 Crucis ‘Type, and on the Spectrum of Silicon. By JosepuH Luyr, BSc., F.L.C., a Assistant Royal Observatory, Cape of Good Hope. Communicated by Hy Davin Git, C.B., F.R.S.,; Her Majesty’s Astronomer at the Cape . . 4A 4 A Note on the Electrical Resistivity of Electrolytic Nickel. By J. A. FumMina, M.A., D.Se., F.R.S., Professor of Electrical Engineering, University College, London. : ‘ ‘ : : 3 ‘ é : é - 50 Observations on the Morphology of the Blastomycetes found in Carcinomata. By Keiru W. Monsarnat, M.B., F.R.C.S.E. Communicated by Pro- fessor SHERRINGTON, F.R.S. . . , ; 3 ‘ 3 ONS Nene : Meeting of January 18, 1900, and List of Papers read. ‘ SE ye eneein - Upon the Development of the Enamel in certain Osseous Fish. By CusaruEs | | S. Tomes, M.A. F.R.S. . j 3 : ; : t b : Arn 17 1900 PAGE Py _ Meeting of February 1, 1900, and List of Pavers mad Ops i 165 #4 ye fice Livre ‘Combinatorial Analysis—The Foundations of a New Theory. By Major P. A. MacManon, R.A., D.Sc., F.R.S. RR wey aN BUR A NAF _ Wher Reihen auf der Convergenzgrenze. Von Emanven Lasker, Dr. Philos. a urther Note on the Influence of Temperature of Liquid Air on Bacteria. By Aran Macrapyen, M.D., and 8. Rowianp, M.A. Communicated by Lorp Lister, P.R.S. Saat ee : mf RL Rag iy ‘ ‘ | Report of the Kew Observatory Committee for the Year ending December 81, t 18939. ° ° Py ° r ry e ° ° 5 » » a € Price One Shilling and Stapence. No. 430. = Oommunicated by Major MacManon, F.R.S. . : i i ; ‘ phan | 323 334, 335 336 341 will only be issued with the concluding No. of ea apie oem or Uiner fiinstablens| in Las or direct photographie reproduction. “Sold by Messrs. Harrison uta Sons. want ato. Price 1s. 5 | ‘ PROCEEDINGS OF | mrae ROYAL SOCIETY. CONTENTS. Meeting of May 10, 1900, List of Candidates recommended for Election, and List of Papers read . é Result = ; : : : : ‘ Meeting of May 17, 1900 : ‘ : r : : : : _ List of Papers read May 17,1900 . Electrical Conductivity in Gases traversed by Cathode Rays. By J. C. McLennan, Demonstrator in Physics, University of Toronto. Communi- cated by Professor J. J. Tuomson, F.R.S. : ‘ * Observations on the Electromotive Phenomena of Non-medullated Nerve. By Miss 8. C. M. Sowrox. Communicated by Dr. Watuzr, F.R.S. (Plate 4) Experiments on the Value of Vascular and Visceral Factors for the Genesis of Emotion. By C.S. SHerrineton, M.A., M.D., F.R.S. On the Brightness of the Corona of April 16, 1893. Preliminary Note. By H. H. Turner, M.A., F.R.S., Savilian Professor : : i : Radio-activity of Uranium. By Sir Wittram Crooxzs, F.B.S. (Plate 5) qgoulan I nstit, f a 0, JUL 23 1900° Office (& Libra Price Two Shillings and Stxpence., _ Joby 9, 1900. VoL fe, No. 431 PAGE 374 409 ee ee eR | re eS a ERY — ——=s re Se Sie ea eee ee a ee en Se = ee ——— = In future, Forms of Acknowledgment of the ‘ Proceedings’ will only be issued with the concluding No. of each volume. NOTICE TO AUTHORS. Authors of Papers intended for the ‘ Proceedings’ or ‘ Philo- sophical Transactions’ are urgently requested to send in all drawings, diagrams, or other illustrations in a state suitable for direct photographic reproduction. sold by Messrs. Harrison and Sons. Royal 4to. Price 1s. LIST OF FELLOWS OF THE ROYAL SOCIETY. 30th November, 1899. PROCEEDINGS OF THE ROYAL SOCIETY. CONTENTS. _ ‘Meeting of May 31,1900, and List of Papersread. . . Ce OS ay Bute } Cnoontan Lroturzr.—On Immunity with Special Reference to Cell Life. By Professor Dr. Paun Euruicn, Director of the Royal Prussian Institute of Experimental Therapeutics, Frankfort-on-the-Maine (Plates 6 and 7). 424. ae | Price One Shilling. ; Hurt 24,1000. a ae In future, Forms of Acknowledgment of the ‘ Proceedings ’ will only be issued with the concluding No. of each volume. NOTICE TO AUTHORS. Authors of Papers intended for the ‘ Proceedings’ or ‘ Philo- sophical Transactions’ are urgently requested to send in all drawings, diagrams, or other illustrations in a state suitable for direct photographic reproduction. Sold by Messrs. Harrison and Sons. Royal 4to. Price Is. LIST OF FELLOWS OF THE ROYAL SOCIETY. 30th November, 1899. PROCEEDINGS OF Pam ROYAL SOCIETY. VOL. LXVI. No. 433. Zac Ta Sy ae 10, \ CONTENTS. {/ ex 19 1900” } Slice Libr at La er, a ane Annual Meeting for the Election of Fellows . PE ane Teo HTN Meeting of June 14, 1900, and List of Papers read On the Periodicity in the Electric Touch of Chemical Elements. Preliminary Notice. By Jagapis CuunpER Bosz, M.A., D.Sc., Professor of Physical Science, Presidency College, Calcutta. Communicated by Lorp Rayieien, F.R.S.. : f : ; : : p i é ‘On Electric Touch and the Molecular Changes produced in Matter by Electric Waves. By Jaagapis CuunDER Boss, M.A., D.Sc., Professor of Physical Science, Presidency College, Calcutta. Communicated by Lorp Ray- LEIGH, F.R.S.. Contributions to the Comparative Anatomy of the Mammalian Hye, chiefly based on Ophthalmoscopic Examination. By Groner Lindsay JOHNSON, M.D., F.R.C.S. Communicated by Hans Gapow, F.R.S. . The Influence of Increased Atmospheric Pressure on the Circulation of the Blood. (Preliminary Note.) ti LEONARD sea M.B. Communicated by Dr. Mott, F.B.S. . E : , ‘ On Cerebral Anemia and the Effects which follow Ligation of the Cerebral Arteries. By Lzonarp Hitt, M.B. Communicated by Dr. Mort, A a. a e The Circulation of the Surface Waters of the North Atlantic Ocean. By H. N. Dickson, B.Sc. Communicated by Sir Jonn Munray, K.C.B., E.RB.S. Palzolithic Man in Africa. By Sir Joun Evans, K.C.B., F.B.S. For continuation of Contents see 2nd page of Wrapper. Auaust 4, 1900. Price One Shilling and Sixpence. 450 452 474, 478 480 484, 486 rence of aa Hy a creel M. | ad oe Rowza’ - Lorp Lister, ER; 8." « .. . . i Vapourdenty of Bromine at High Temperatures. Gapnlemeneaeel _ By E. P. Purman, D.Sc., and @. A. 8. ATKINEON, eves! eated by Professor Ramsay, F.R.S. . “The SeHaiveHbeG of Silver and of some other Metals to Light. By Major tte Die General J. Warzrnovsz, I1.8.C. (late Assistant Surveyor-General of — y tata): Communicated by Sir'W. Asyey, K.C.B., FBS. ie | In future, Forms of Acknowledgment of ine - Pe badge Me iia only be issued with the mhiniact No. of cach volume. { Sold by Messrs. Harrison and Sons. © Royal 4to. Price Is. (E ROYAL SOCIETY. . . e . ° . © ° £ Pi “ 4 . ‘ ’ a Z “ “ i : | Prive Siaponce, i NOTICE TO AUTHORS. fora direct Potosi phils mR 8 Sold by Messrs. Harrison and Sons. Royal 4to. Price 1s. rr iahay . LIST OF FELLOWS OF THE ROYAL SOCIETY. a 30th November, 1889. | SMITHSONIAN INSTITUTION LIBRARIES TO 3 9088 01306 0249