ae ie, es a ee) a a >» BLY) » a) ue) »>)» Da. | > > DP) >a ») » Bp? yp)» DD) a. Dp PDP D >MD)> DP) REDD >» D». PPD DDD)» >») IHW » Di. ' 4 os DIED» PD ID» DY) >) WPI. i » 7p a DY DP) yy 2 yy) BaP BY i) yp ¥)) D AD GD PP dy y) yy BY J FIDL? PWIND DD? TY AD 2 D> DY DDD) D> Dy WD” »») D DP D?2 WYP?” DIP Yo 2d WIPId. ; i BD)? PPV» IPP WP PDD PID D): 2 >» D>) 3 Dep» DDD D> SIP Wp) yd)»; > D> DPDYP 2D UP D) Yd 2DY). : DP IY? ‘> py») PDP) ye PP)» wm) > »)). 7 > MW 5 ) Dip yp P “pwd BPW oy) YWPLBMW) rz Bo» DD») PB GD) PD) WD .rueD 2 Py” S32) DSly Swe 7 2» 2D DY? P DP WP D ») PDD) a 2D YD p>?! Yew yp 3 ). WY Wom pd I 2D) D DD) Py Pp PI 34 2D D> SIeDD>P > EP>yY DD wb ED)? ¢ | > > BIS» IP koa) BD WIM Rv Ee. D) Le >» 2P Woy)? BW PWS yy © DOYS Bb). >) ID yD PPB)» ) >) > EYP» ay» Bib i>) WEED »>> 9 ee > IRD)? >» Wp yd wD BD DDD? TPR) D »»»> Yom oD > wy DDD. DP Dw)» 2D DY pp 2) Pw 3 > D> WY Ww) 22P PP WD Ia D> DD > PY Py WP IP: PID IMD oP ® D> DDD)» 9:7 UMD 2D DD B- 22 D> SID Dy DP DD 2» WD - Se 2 D> bam by> DD AP >> PWD wb») TD) >> > SD Ep» DP Dw PY DVD Hy» 1D Sy I» D» vw DVD 1») VIB Wd» i= me, 4 3 » DDD DY» PR Wad 1 pa=y>») » DVD» LW Wry 12 DD) Dae yD BW. yDWO Ov VM aap SD D > AD >I DIP 1) Y> 2d [nD BW ID YT) ) DD) WDD» Y DOD ie Bb O>D ys DD YR?» » 2D <4 Lap DY > Rap)» >) MIP)» DW WOM *® | > D WD >) IED >> Pl) PI )» D VP ID) DPD D>ywp») >» WHW » DPD w mo D BY wD) PD 2» MY. LY) Wr: ; o eg p>? PDI) D > LYIID YY DW: red DM) DY» Dy )» DDP 2 DY DD PZP DBD DY TY 2) Ww BD 2D) YD DP VP 2») DWP. be >) PP» DP my P»> DIP | iw D> DPI PLD PPP WD DDD W D>) p>ypy>» 39D Mm DIB ¥ 2” WD pyryPp_ >» >» VY DY) Y- ! (ap py » Va wane ht . 2 i> Wa Wo. «OC COD Ge Cie, GG SESS, (CO i oa R4a ¢ Pi os «& = G Ca gq es: & « '§ CO #7 GG ECan OME C4 ACC OO € C0 Oa OO OM COO OCC ON COC OL OLE CC CF @. aGe FURIES FOC AO ECG HORI CO OG a Or i CE LG ( GG A GIR COCO EG ee de @oe eae” CGC GT OC 046 We” Wr we (Cm COO UC A OC be Ct Cr | cee Ge” Ee Eck Cl Git Qt Cd Ci C@ (8 dir (CUE © OMNC CLE Ct ee ic€ ft COO CUO CO CU OR Ke cas \. COE COTE CO ONE OC CORO” EET CE ure « 4 COLOR € ae © CORO EC EC ( Ceere € Teer Cae We CORE aCe (CeO COO CE UE COC COREG ECe (cere Cea HCCC (CEE Qc CL LOCOCO E CBee CoC eC Ute £ eu Ce . CQ e COCO CS & Ce Oe CS GM CO Hm, CL SS COREE | © 4a a4 x @ C¢ CC xe -€ (EQ « ae u.G i ¢ ( C.F aXe ¢ de Ge € « € ae (a. @ Ga ¢ ¢ | COLE COM Cre CO OR OE OUOE 4¢ ©, Come. CC Cee Ce OR OE MOOT CC |; OE Gee CC COG. Ye @ee¢ € Gt Gat €¢ C ¢ CO qe i CL. &} OO CK € ce ¢ ae €€ ay OCRE OC. CCE | of CU C6 Ci Cae ere eyl A ene CT! GCC OAC C& ¢ (te G& € Ei CE CK GUC CHET C@ € COO, OO UE OE CU CECE Coe € Comms, C5 ORE C CC CC eWa Cac ¢ we = y (8 Ge < ms ( 8x a(t & (. @e@ . phn —agh SEE < GE Ot CM (OGG & (CG C(O Ce @& Five ¢ Cr GGG Catan Co CCE. KC « a ¢ Ure . ¢ CM °C Gea ( & Ga &@ e, x ¢ ‘a _@ a ii ¢ Q a & ee “ob e <. KC Ci Ec€: (Ks CLC CHUM Re © ¢ @ i a ¢ (C G ¢ € @ Ce f C(@K 14 oa O(a? G C( ¢ « ( Gait L Oy ¢€ CO Ci a G c P CEES EE CEE, 6 GSE Ee Ce LU eg ©: CC. Qoc@ CC aie € PC me Ca ce COME CO ‘« € , (Gage 6 ea aa f COLE CC Oe (ELK € CO amt CC HOGEEEO CCF Qace OC GE eee Cicada «ic ay (CK CO CO, UE CUE OOO CS i ( OO CE GUE OC WO ONEO, KU ce OM OO ME MC AME © EC O_O NOC OL OME UE MC OOO FE CA OOO OAKES. OE MUO |€— EOOTE COE CE COE Gi ¢ AE COE, CE CCE G os C EQ GORE «OC GORE aC COCO EC en © OA OEE 6 COL CEO ¥ nies i i Eavt ie ; : i ( ‘ THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Petw Sevics. VOLUME XVI. r UCI 20 1999 269610 DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS & NORGATE, 14 HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1920-1922. Tue Socinry desires it to be understood that it is not answerable for any opinion, representation of facts, or train of reasoning that may appear in this Volume of its Proceedings. The Authors of the several Memoirs are alone responsible for ther contents. ERRATUM ” 2 . 63, line 2, for ‘‘ oxygen ’’ read ‘‘ water.” ’ ) t=) Dustin: Prinrep AT THE University Press RY Ponsonry AND Grpns, CONTENTS. VOL. XVI. No. I.—A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, sc.p., F.R.s., and IT’. G. Mason, m.a., so.B. (January, 1920.) 6 : 6 . . II.—The Carboniferous Coast-Section at Malahide, Co. Dublin. By Lous B. Suyvru, B.a., sc.z. (Plates I, II.) oo ~ 1920.) : : Ill.—The Application of the Food-Unit Method to the Fattening of Cattle. By James Wizson, ™.a., B.Sc. ae III, ne (February, 1920.) TV.—The Holothurioidea of the Coasts of Ireland. By Anne L. Massy. (April, 1920.) ° 2 V.—Photosynthesis and the Electronic Theory. By Henry H. * Dixon, sc.v., F.x.s., and Horace H. Pooun, sc.p. (March, 1920.) : 6 : 5 : : VI.—Note on the Decay of Magnetism in Bar Magnets. By Winu1am Brown, z.sc. (March, 1920.) , VII.—On the Inhibition of Invertase in the Sap of Galanthus nivalis. By 'l. G. Mason, m.a., sc.p. (April, 1920.) . 0 VIIL—The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Witmiam Brown, 8.sc., and Patrick O’CaLiacuan, A.R.C.SC.1., A.C. (August, 1920.) 1X.—Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methy]l- benzene). By A. G. G. Leronarp, 4.R.C.SC.1., B.SC., PH.D., and Acnrs Browne, A.8.¢.8¢.1., B.sc. (August, 1920.) X.—An Investigation into the Causes of the Self-Ignition of Ether- Air Mixtures. By the late Professor J. A. McCuriuanp, D.SCc., F.R.S.. and Rey, H. V. Gitu, s.J., D.s.0., M.C., M.A. (August, 1920.) : 6 ; ° PAGE 25 37 63 78 83 98 105 109 iv Contents. No. XI.—The Influence of Electrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By Josepx Rernxy, M.A., D.SC., ¥.R.0.80.1., and Winrrep J. Hickinporrom, B.Sc. (October, 1920.) 0 0 : XII.—Notes on some Applications of the Method of Distillation in Steam. By JosmpH .ReILLy, M.A., D.SC., F.R.C.S0.1., and Witrrep J. Hicxrnsorrom, B.sc. (October, 1920.) -. XIII.—The Determination of the Rate of Solution of Atmospheric PAGE. 120 131 Nitrogen and Oxygen by Water. By W. E. Aprnry, p.sc., _ A.R.C.S0.1., F.I.c., and H. G. Brower, A.R.0.S0.1., A.1.C. (September, 1920.) XIV.—A Determination, by means of a Differential Calorimeter, of -the Heat produced during the Inversion of Sucrose. By Henry H. Dixon, sc.p., r.r.s., and Nicrt G. Batt, B.a. (December, 1920.) 0 c : : XV.—The Measurement of very Short Time Intervals by the Con- denser-Charging Method. By Joun J. Dowuine, m.a., F.INsT.P., and Donat Donnetty, u.sc. (In conjunction with the late Prof. J. A. McCrrnuanp, r.x.s.) (February, 1921.) XVI.—A Vibrating-Flame Rectifier for High-Tension Currents. By Joun J. Dowxine, M.A., F.INST.P., and J. T. Harris, B.so. (February, 1921.) : XVII.—A Sensitive Valve Method for the Measurement of Gapnsity with some Important Applications. By Joun J. Downe, M.A., F.INST.P. (February, 1921.) i ; - XVIII.—A Direct Reading Ultra-Micrometer. By Joun J. Dowtine, M.A., M.R.I.A., F.INST.P. (March, 1921.) XIX.—Studies in the Physiography and Glacial Geology of Southern Patagonia. By HE. G. Fenton. eee V, VI, and VII.) (March, 1921.) : 0 : 0 XX.—Award of the Boyle Medal to Groren H. PreruysrineE, B.sc., PH.D., 1921. (March, 1921). XXI.—The Concentration and Purification of Alcoholic Fermentation Liquors. Part I1.—The Distillation in Steam of certain Alcohols. By JosurH REtny, M.A., D.SC., M.R.IA., F.1.C., and Witrrep J. Hicximsorrom, m.sc., atc. (August, 1921.) XXII.—The “Browning” and ‘“ Stem-Break’’ Disease of Cultivated Flax (Linum usitatissimum), caused by Polyspora lini n. gen. et sp. By H. A. Larrurry, a.r.c.sc.1. (Plates VIII-X.) (August, 1921.) : 0 148 153 165 171 175 185 189 226 233 Contents. No. XXIII.—A New Principle in Blowpipe Construction. By H. G. Brucker, A.R.C.SC.1., A1.c. (August, 1921.) . XXIV.—Uncharged Nuclei produced in Moist Air by Ultra-Violet Light and other Sources. By the late Prof. J. A. McCietnanp, ¥r.z.s., and J.J. M‘Henry, u.sc. (August, 1921.) XXV.—Biological Studies of Aphis rwnicis L. A.—Appearance of Winged Forms. B.—Appearance of Sexual Forms. By J. Davipson, p.sc. (August, 1921.) . : XXVI.—The Occurrence of Dewalquea in the Coal-Bore at Washing Bay. By T. Jounson, D.sc., r.u.s.,and Jans G. Grimorg, B.sc. (Plates XI, XII.) (August, 1921.) : XXVII.—A Simple Form of Apparatus for observing the Rate of Reaction between Gases and Liquids, and its use in determining the Rate of Solution of Oxygen by Water under different con- ditions of Mixing. By H. G. Brcxsr, a.R.c.sc.1., A.1.c. (August, 1921.) : : : 0 XXVIII. —The Occurrence of a Sequoia at Washing BBey. By T. Jounson, D.Sc., F.L.S,, and Janz G. Gitmorz,B.sc. (Plates XIII, XIV.) (August, 1921.) i XXIX.—The Sources of Infection of Potato ee with the , Blight Fungus, Phytophthora infestans. By Paut A. Murpuy, B.a., A.R.O.Sc.1. (August, 1921.) . - XXX.—Some Factors affecting the Hydrogen Ion Concentration of the Soil and its Relation to Plant Distribution. By W. R. G. ATKINS, 0.B.E., SG.D., F.1.c. (February, 1922.) XXXI.—The Hydrogen Ion Concentration of Plant Cells. By W. RB. G. ATKINS, 0.B.E., SO.D., F..c. (February, 1922.) XXXII.—Note on the Occurrence of the Finger and Toe Disease of Turnips in Relation to the Hydrogen Ion Concentration of the Soil. By W. R. G. Arnins, 0.B.E.,8¢.D., F.1.c. (February, 1922.) : : : XXXIII.—Photosynthesis and the Electronic Tee aD. By Henry H. Dixon, sc.p., r.n.s., and Nreent G. Barn, ma. Bae 1922.) 0 : XXXIV.—The Bionomies of the Conidia of Philephine W nfintere (Mont.) De Bary. By Pau A. Murpay, B.a., a.R.c.so.1. (February, 1922.) 0 6 XXXV.—On the Distribution of Activity in Radium qe under different conditions of Screening. By H. H. Poots, m.a., sc.D. (April, 1922.) PAGE 304 325 384 345 3598 369 414 427 467 vi Contents. No. PAGE XXXVI.—The Influence of Feeding on Milk Fat. By H. J. Sunny, F.R.0.8C.1., B.SC. (April, 1922.) 3 5 . 478 XXXVI. —On a New Method of Gauging the Discharge of Rivers... By J. JOLY, $¢.D., F.R.S., F.T.c.D. (April, 1922.) : . 489 XXX VITI.—On a Variety of Pinite occurring at Ballycorus, Co. Dublin. By Louis B. Suyrx, u.a., so.p. (April, 1922.) : . 492 XXXIX.—On the Life-History and Bionomics of the Flax Flea-Beetle (Longitarsus .parvulus, Payk.), with Descriptions of the hitherto unknown Larval and Pupal Stages. By J. G. RHYNEHART, A.R.C.SC.1., D.1.C., N.D.A. (Plates XV-XIX.) (April, 1922.) ; : 2 : ; . 497 THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.S.), No. 1. JANUARY, 1920. A CRYOSCOPIC METHOD FOR THE ESTIMATION OF SUCROSE. Searnsgntan Inetitcgye~ MAR © 5 1921 BY Ss *rtenen seen A HENRY H. DIXON, Sc.D., F.B.S., UNIVERSITY PROFESSOR OF BOTANY IN TRINITY COLLEGE, DUBLIN ; AND T. G. MASON, M.A., Sc.B. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN LONDON, W.C. 2. 1920. Price Sixpence. Royal Bublin Society. —— ae FOUNDED, A.D. 17381. INCORPORATED, 1749. EVENING. ‘SCIENTIFIC MERTINGS. ©. Tue Scientific Meetings of the Society are usually held at 4:15 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read ,Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and guch as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Hiditor. THE (MAR 75 4q SCIENTIFIC PROCKEDINGS THE ROYAL DUBLIN SOCIETY. ih A CRYOSCOPIC METHOD FOR THE ESTIMATION OF SUCROSE, By HENRY H. DIXON, Sc.D., E.B:S., University Professor of Botany in Trinity College, Dublin ; AND T. G. MASON, M.A., Sc.Bt Read Drecemprr 16, 1919. Published January 9, 1920. THE depression of freezing-point caused by a given quantity of sucrose in a given volume of water is approximately doubled after inversion. It is evident, therefore, that by two cryoscopic observations—before and after inversion—the sucrose-content of a solution can be determined. By using the Thermo-electric Method of Cryoscopy (2) the estimation can be effected on small amounts of fluids, and, as the presence of colloids introduces no difficulties, it should furnish a rapid and sufficiently accurate means of estimating the sucrose-content of physiological fluids. The following investigation was undertaken to work out the details of the method, and to test its applicability to plant saps. 1 Mr. Mason was enabled by a maintenance grant from the Department of Research to engage in this research. SCIENT. PROC. R.D.S., VOL. XVI, NO. T, B 2 Scientific Proecedings. Royal Dublin Society. Apparatus. he construction of the thermo-couples employed, and the general arrangements of the thermo-electric method, have already been described (2), and it is unnecessary to go into them here. Sheht modifications have, however, been introduced. Instead of single drawn pine rods to support the wires of the couple, double ones, slightly flattened on the surfaces of contact, were employed. This gives greater strength, and affords a means of pro- tecting the eureka and copper wires forming the couple, which run down the grooves on each side of the double pine supports. One of the pieces of pine is cub away close to the lower end of the support, thus forming a recess in which the junction is freely exposed to the surrounding fluid. Collodion varnish was used to waterproof the insulated wires, and to cement the pine rods together, and to keep the wires in position. In case it is desired to alter the sensitiveness of the couple, or modify the couple in any way, it is easy to dissolve off the collodion with acetone, and substitute a new eureka wire of different resistance. Calibration of the Thermo-couples, Sucrose was employed as a standard for the calibration of the thermo- couples. It was selected for this purpose, not only because it can be readily obtained of great purity, but its depressions of freezing-point, which are known with great precision from Raoult’s researches, can be more easily determined than those of other substances, whose solutions are not hydrated to the same degree. The ease with which it undergoes inversion on storage is, however, a serious disadvantage to its use; and as it is not always pos- sible to make up standard solutions for every calibration, calcium chloride solutions of suitable concentration, whose freezing-points had been deter- mined against sucrose, were used for this purpose. The importance of frequently calibrating the instrument has already been emphasized (2). Thermo-couples producing deflections of from two to three hundred milli- metres per 1°C. have been constructed with a single pair of junctions, the degree of sensitiveness varying with the resistance of the eureka wire employed. The calibration of a thermo-couple producing a deflection of 294-5 mm. per 1°C. is shown in Table I. In this approximately 30cm. of eureka wire, 0°100 mm, diam., and two leads each of copper wire, 0:200 mm. diam., were used. In the last column of the Table are given the actual freezing-points of the solutions derived from Raoult’s results quoted by Hamburger (4). Dixon anp Mason—Cryoscopic Method for Estimation of Sucrose. 3 TABLE I. | ails E | | See | Deflection Deflection Detlection | Deflection | agen | eS Grae | Obs. I. Obs. 11. | Obs. ULT. | Mean. leiRaoults i 15-75 1588 15°75 15-8 0-054° | 2 Soro rs liver 3 250 Sigal fee SO 32-0 0:108° | 3 47-13 | 48:0 | 47°38 47-5 0-163° 4 | (GH | 63-0 | 63-5 | Gee OF 2)1S2ieu| 5 80°25 80:0 | 80:75 80° 0:272° | 6 95:5 | 95:75 | 95°25 95:5 0:328° | 7 113-25 | 11975 | 1185 113-1 0-383° | | It will be observed that a thermo-couple as sensitive as that used in this calibration can only register depressions of freezing-point up to 0-849° (294-5 mm. = 1°). This limitation, of course, is imposed by the scale. Where the solution to be examined has a greater depression than this, it is necessary to substitute for distilled water a standard freezing below 0°. Solutions of calcium chloride, whese depressions have been previously deter- mined, can be employed. Though this solution may be used in place of distilled water in these cases, it is not as easy of manipulation as a pure liquid, where ice may continue to separate without alteration in freezing- point. On the whole, it has been found more satisfactory, where a greater range is desired, to construct a less sensitive thermo-couple, and to keep distilled water as the standard. Estimation of Sucrose in made-wp Solutions. To effect the inversion of sucrose, small quantities of invertase prepared by Davis’ method (1) were used. It is undesirable to dilute unnecessarily the solution, so only that quantity of invertase is added which will ensure inversion. Allowance must subsequently be made for the dilution of the solution with the water added in the invertase, and also for the depression of freezing-point produced hy the solutes in the invertase solution and the toluene added with it. The invertase solution employed had a moisture- content of 83 per cent. In each of the observations two test tubes of about 10 c.c. capacity were used. Into one approximately 5 gm. of the sucrose solution (2 gm. of sucrose B 2 4 Scientific Proceedings, Royal Dublin Society. per 100gm. of water) was weighed, and into the control tube a similar amount of distilled water. From a previously calibrated capillary tube 0°32 em. of invertase solution was introduced into each test tube. (In Obs. 2 only 0°24¢m. was used), and then a drop of toluene on a glass rod, with which the mixtures were briskly stirred. The test tubes were then tightly corked and stored at 28° in a thermostat; on withdrawal of the first six on the following day (approximately 24 hrs.) the deflections produced by the freezing-point depressions of the sucrose- invertase solutions and the controls (invertase solutions) were independently determined against distilled water. The solutions used in the subsequent five observations were left in the thermostat for another day, and then examined in a similar way. 1t is more direct, and has been found more satisfactory, to add a measured quantity of the invertase-toluene mixture to the weighed sucrose solution, keeping both at a low temperature to prevent premature inversion. Then, without allowing the temperature to rise, a determination of the freezing- point is made. The second freezing-point determination is made after incubation, and the difference of the two depressions found is a measure of the sucrose inverted. For example, in Obs. 1, Table II, the deflection produced by the solution before the addition of the invertase was 33 mm., so that 31:0 (viz. 64:0—the corrected deflection for the sugar-invertase mixture—less 35), represents the increase in deflection due to inversion; 304'] mm. with this thermo-couple represents 1°. Hence, the increase representing a difference in depression of freezing-point of only 0:101° (0°210° — 6:109°), it is evident that the inversion was not complete. Complete inversion of the sucrose present would have produced an imcrease in the depression of freezing-point of 0°108°. This stage has been reached, and is recorded in the observations made after 48 hours. The results afford an index as to the agreement that may be expected when sucrose is inverted under these conditions. As the invertase solution employed had a needlessly low depression of freezing-point, and was not of great activity, being prepared two years previously, a new preparation was made, which was used in subsequent observations. In Table III are shown the results obtained on inverting a sucrose solution containing 4 gm. of sucrose in 100 gm. of water under the conditions deseribed in the last experiment. The weight of invertase solution employed was approximately 0°38 9m. per 5 gm. suerose solution. That inversion was complete in this experiment is evident from the figures presented in the last’ column Dixon ayp Masun—Cryoscopie Method for Estimation of Sucrose. 5 TABLE II. >) 2 gms. Sucrose per 100 gms. H.O. (A = 0°109°.) DEFLECIION OF THERMO-COUPLE, 304:1 mm. per 1° C. "i Deflection of A Time in Ne: o : Novot : Sugar + Deflection of of Sugar ‘ Obser- Gm. of Sugar 5) RR: he Hours. saa Saietions Invertase Control. Solution after i ; i Solution. | Inversion. 1 | 5:07 103°3 42°35 0-2160° | Q | 5:07 102°6 39-4 0:216° | | 8 | 5-00 100-1 39°4 9°210° 24 hours 4 5:02 96°7 37-0 0°207° | | | 5 5-01 100°8 40° 0-209° 6 5°02 | 101°1 45-1 0-194° | — — i — || | 7 5:00 | 106°1 41°9 0:222° | | 8 4°00 | 106°9 42°0 0:225° | 48 hours 9 5°08 | 107°5 41°6 0:228° | 10 5:02 100:8 39°9 | O-211° Nes sistema 5:02 94:6 30°6 0°322° | Tape III. 4 om, Sucrose per 100 gm. Water diluted by addition of Invertase Solution to 3°75 gm. of Sucrose per 100 gm. of Water (A = 0:204°, Raoult). DEFLECTION OF THERMO-COUPLE, 302 mm. per 1°. l wy Deflection for apens Foe A | ane: a | Sugar and Invertase | Delleeen for | of Sugar Sol. after | amp" | Sol. after Inversion. | i Inversion. | | | oval ies _| | | 1 166:1 39°75 0:419° 2 164:7 41-0 0:410° | | 3 | 165-1 | 39:6 0°415° | fia | aes peseeess = 4 162:7 36'S 0:417° | | i) ! i [ore Scientifie Proceedings, Royal Dublin Society. Method upplicd to the Estimation of the Sucrose content of Saps. It has been demonstrated by Dixon and Atkins (4) that the applicatiou of intense cold renders tissues permeable, and so provides a means of extracting by pressure a representative sample of the sap. In their work liquid air was employed; as this, however, was not available, a freezing mixture of salt and ice has been employed. That the temperature thus obtained is capable of rendering the tissues permeable is shown elsewhere. An experiment carried out on the leaves of Galanthus spicatus will illustrate the method adopted for determining the sucrose-content, and will vive some idea of the accuracy that may be expected of it. The leaves were gathered at 9.50 a.m., April 26th, from two clumps of this plant growing side by side. From one clump light had been excluded for five days. The leaves were then packed in large test tubes, which, after corking and sealing with plasticine, were submerged in the freezing mixture [- 16°] for two hours. On removal of the corks the tissues were found to be frozen solid; they were not, however, taken from the test tubes till they were somewhat thawed. On withdrawal of the material it was wrapped in linen, and pressed in a vice between two silver discs; about 30 c.c. were collected from each sample. In order to check the immediate activity of the enzymes, the sap was collected in test tubes and jacketed in ice. Fifteen cubic centimetres of each sample were then boiled to Inll the enzymes in half-inch test tubes, fitted with reflux condensers, by immersion for one minute in a brine bath. The freezing-points of the sap before and after this treatment are shown in Table IV ; a slight lowering of the depres- sion took place in both samples. TABLE IV. Z ee | A A Description of Sample. | of sap before boiling | of sap after boiling and filtering. and filtering. Ses | Sap pressed from exposed leayes, 0 : 0: 7a0 | 0-748° | | | Sap pressed from leayes covered for five days, | 0°7038° 0-698° Two test tubes (about 10 ¢.c. capacity), each containing 5 gm. of the boiled and filtered sap of each sample, were then immersed in a freezing bath [- 2°]. Drxon ann Mason—Cryoseopic Method for Estimation of Sucrose. 7 99 As soon as the saps had taken up the temperature of the bath, 0°35 gm. of invertase solution was added from a capillary tube. The sap-invertase solutions were then incculated with a little hoar frost and their freezing- points determined. After storage overnight at 28°, the depressions were again observed. That inversion was complete was ascertained by raising to boiling-point as before, and again determining the freezing-points. In Table V are shown the depressions of the sap-invertase solutions before and after inversion. The difference in the observed depressions have been eorrected for the dilution due to the addition of the invertase solution ; in the final column are shown the equivalent concentrations of sucrose! It will be observed that no correction has to be made for the solutes introduced with the invertase solution. It is of great importance that the saps should be cooled below 0° before the addition of the invertase, or otherwise immediate inversion will take place, and the results obtaimed will he too low. TABLE V. DEFLECTION OF THERMO-COUPLE, 161-7 mm, per 1°. lea Ee meta ae | j | | Apa Al Gin. |Mois-| . | ena 5 \ : | ise q % 0 Al A2 | Corrected | Sucrose Description Os ane | of | Before | After A2-Ai. | rye || per ape: P xpt. | Inversion. | Inversion. |Dilution by) 100 Gm. | cent. Al q | Inyertase. | Sap. Sap pressed non | | ( i 0°762° 0°804° 0°048° 0°045° 0:83 leaves : covered \ | 93°53 for five days. ) | [lit | 0-763° | o-su2e | 0-039° | o0-041° | 0-76 | non o ° Bey Vic) “51° “54° «I Q Sap pressed from | | Sag ( 11 | 0°809 0:860 0-041 0:054 | 1:00 | exposed leavcs. \ | | fc 0-s12° | 0-866° -054° 0-057" | 1-06 A consideration of importance in favour of the cryoscopic method is the absence of any preliminary treatment for the purpose of clearing the sap of gums, ete.: this treatment is, of course, necessary before polarimetric or copper methods can be employed. It is, however, for work in relation to the osmotic pressure in plants that the method has been elaborated, aud is especially applicable. An approximate comparison of the delicacy of the eryoscopic method ‘The vaiues here assigned to the increase in depression are based on the assumption that the freezing-point depression of sucrose is exactly doubled on inversion ; this is not true to the thousandth of a degree, 8 Scientific Proceedings, Royal Dublin Society. with the volumetric method with Fehling solution may be made as follows :— As usually made up, 100c.c. of Fehling solution is decolourized by 0°5 gm. dextrose, or slightly less sucrose. Reading to 0:2¢.c., the probable error is 0:001 gm. With the cryoscopic method, readings of temperature may be obtained with an error of + 0:005°; 0°01° corresponds to 0:2 per cent. of sucrose. Only 2:5 e.c. is necessary for the determination, so that the probable error is 0:0016 gm. REFERENCES. 1. Davis, W. A.—Journ. Soc. Chem. Industry, Feb., 1916. Lo Dixon, H. Hi—A Thermo-electric Method of Cryoscopy. Proc. Roy. Dublin Soe., 1911, vol. xii, p. 49. 3, Dixon, H. H., and Arxins, W. R. G.—Osmotice Pressures in Plants. i. Methods of Extracting Sap. Proc. Roy. Dublin Soe., 1913, vol. xiii, p. 422. & 4. Hampurcer, H. J.—Osmotischer Druck und Ionenlehre. Wiesbaden, 1902. SCIENTIFIC PROCEEDINGS. VOLUME XVI. 4. A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, $0.D., F.R,S., and T. G, Mason, m.a., so.B. (January, 1920.) 6d. DUBLIN ; PRINTED AT THE UNIVERSITY PRESS BY PONSONRKY AND GIBBS. ‘i asa: ey Smithsonian ineg Weta MAR 5 199 oy Ro ~enan pew THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. i Vol. XVI. (N.8.), No. 2. JANUARY, 1920. THE CARBONIFEROUS COAST-SECTION AT MALAHIDE, CO. DUBLIN. BY LOUIS B. SMYTH, B.A., Sc.B., LEOTURER IN PALHZONTOLOGY, UNIVERSITY OF DUBLIN. [Authors alone are responsible for all opinions expressed in their Communications. } DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN LONDON, W.€. 2. 1920. Price One Shilling. Roval Dublin Society FOUNDED, A.D. 1731. INCORPORATED, 1749. EVENING SCIENTIFIC} MERTINGS. Tue Scientific Meetings of the Society are usually held at 4:15 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin set down for reading until examined and approved by the Science Comuuittee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Hditor. eo) Il. THE CARBONIFEROUS COAST-SECTION AT MALAHIDE, CO. DUBLIN. By LOUIS B. SMYTH, B.A., ‘S8c.B., Lecturer in Paleontology, University of Dublin. Puatzs I anv II. Read Novemuer 24, 1919. Published Fepruary 11, 1920. I.—DESCRIPTION OF SECTION.! THE section lies about nine miles N.E. of Dublin and one mile E. of Malahide, and is a little over one mile in length. It consists of limestones and shales, the former being in places dolomitized.” The rocks are exposed almost exclusively between tides. The general dip is northwards: The more argillaceous beds show cleavage dipping at 60° towards N., 85° W. Three faults interrupt the sequence. Of these, the southern occurs just under the Carrickhill martello tower, and will be referred to as the Carrick- hill fault. The middle one is 120 yards S. of the coastguard station, and . will be called the C.G. fault. The northern (N. fault) is 230 yards N. of the coastguard station, and is marked by a dyke-like mass of dolomite. The exposure is thus divided up by the three faults into four blocks, which are lettered in the accompanying inap E, F, G, H, respectively, from N. to. 8. The commonest type of rock is a very impure, shaly limestone, full of lenticular masses of purer limestone, mostly of small size. These lenticles have frequently had their original orientation altered by the pressure which produced the cleavage. In such cases the bedding is only to be made out when viewed from a distance, or when layers of shells are present. Parts of the exposure are much dolomitized, sometimes sufficiently to completely destroy the fossils. This is particularly the case in the vicinity of the northern fault. Thick beds of shale are found at the extreme northern end, and also south of the Carrickhill fault. As will be seen in the sequel, the latter beds belong to a horizon a little higher than the former. Elsewhere thin beds of shale occur occasionally, ' Reference : Geol. Suryey, Ireland, 1-inch map, sheet 102. SCIENT, PROC. R.D.S., VOL XVI, NO. U. c 10 Scientific Proceedings, Royal Dublin Society. At the northern end the beds disappear under the sand of the Malahide inlet. The next exposure in this direction occurs at a distance of one and a quarter mile, at Portraine, and consists of Bala rocks. To the south the rocks are covered by the Velvet Strand of Portmarnock, the next considerable exposure being the Lower Carboniferous of Howth, three miles away. ‘There is, however, in certain states of the sand, a reef of nodular limestone uncovered 250 yards 8. of the continuous exposure. At Carrickhill the actual fault is concealed by sand. The strike of the fault is about N. 30° W. On the N.E. side the beds have a dip to the N.W., except in close proximity to the fault, where they bend round rather sharply and dip towards the fault, making it almost certain that this is the up-throw side. S.W. of the fault the beds are greatly disturbed, the shaly beds being twisted into zig-zags, the limestones squeezed into lenticles, or fractured, and full of calcite veins. A series of parallel displacements of slight throw, having nearly the same strike as the fault, occur at intervals down to the S. end of the section. These all have the down-throw to the 8.W.,! which supports the view that this is the direction of throw in the main fault. The amount of stratigraphical throw is estimated at over 780 feet (see p. 17). The C.G. fault strikes E. and W., and is seen in the cliff to hade to the S. The beds immediately to the N. of it are dolomitized, and form a gentle anticline, thus dipping towards the fault. Immediately 8. of the fault the beds are shattered, and full of calcite veins. There is also some dolomitiza- tion for a few feet. In places the beds on this side can be seen to be sharply upturned in contact with the fault. The indications are, therefore, that it is a normal fault, with the down-throw to the southward. The amount of stratigraphical throw is estimated at 330 feet (see also p. 15). The northern fault is rather obscure. It is occupied by a thick vein of dolomite, and the rock is brecciated in its immediate neighbourhood. The beds on either side dip away from it at 35°, and are strongly dolomitized for about 30 yards, all fossils being obliterated. There is nothing to indicate the direction of hade. The amount of throw must be small, as the fauna on the two sides is almost identical. This being so, the down-throw must be to the north, as otherwise there would be a repetition of beds, which is not — the case. In the Geological Survey 1-inch map, sheet 102, and memoir to sheets 102 and 112, only one fault is recognized, namely, that at Carrickhill tower, 1A thick band of light grey limestone occurs under the martello tower. This ‘is repeated twice (as indicated on the map, PI. II) by the smaller faults, and immediately shows the amount and direction of displacement. Smytru—The Carboniferous Coast-Section at Malahide. 11 and it is taken to have the down-throw to the N. To quote the memoir referring to the beds 8. of this fault :— “These are precisely similar in lithological character to beds which are frequently found at the base of the limestone in the S. of Ireland (and also in 8S. Wales), and are called therefore the Lower Limestone Shales... . Just on the 8. side of the martello tower here there is a sudden twist in the beds, beyond which there is an equally sudden change in their litho- logical character. It is supposed, therefore, that a fault traverses the limestone here, having probably a down-throw to the N., and being perhaps connected in some way with the one just described as to be observed in the railway, where, however, the throw is in the opposite direction.” (The italics are mine.) The amount of down-throw? to the south of this railway fault is estimated on p. 56 of the memoir as greater than 750 feet—a figure curiously close to my estimate for the Carrickhill fault (see below, p. 17.). With regard to the other two faults, the C.G. fault can, as before mentioned, be seen in the cliff. It was, however, impossible to tell whether it was a dislocation of any importance, until the corresponding fossiliferous bands were discovered on either side. But it is curious that this fault is not even mentioned in the memoir.* I did not myself recognize the northern fault as such during my first visit. It seemed to be simply an anticline somewhat “stretched” at the crest. Only by noting down in detail the beds and their fossil contents in either limb of this “anticline” did it become clear that correlation was impossible. II.—Fauna! AND CORRELATION. Block E, upper part (M 1, 2, 3, 4a, b, ¢).° 290 feet = Ci. Abundant— Zaphrentis densa, Carruthers. Syringopora 0, Van. Athyris cf. glabristria, Phill. Syringothyris lanvinosa, McCoy. Conocardium fusiforme, McCoy. 1P. 56. . 2The context shows that this is ‘‘ stratigraphical ” throw. 3 By the courtesy of the officers of the Geological Survey of Ireland, I was able to consult the MS. 6-inch map, from which the l-inch was reduced. I there find three parallel lines drawn E. and W., one at this fault, the other two close to it to the S. ~ 4 Only corals and brachiopods are recorded, except in the case of a few molluses which happen to be abundant. 5 These symbols refer to groups of beds as indicated on the map. c&Z 12 Scientific Proceedings, Royal Dublin Society. Less common— Zaphrentis omaliusi, Ed. & H., Carruthers. A delanouet, Ed. & H., and parallela, Carruthers Endophyllum iregulare, sp. nov. Caninia cylindrica, Mut. Z or y. Amplexus coralloides, Sow. Michelinia aff. megastoma, Phill. Productus burlingtonensis, Hall. Rare. Pustula tenuipustulosa, 1. Thomas. Schelwienella crenistria, Phill Leptaena analoga, Phill. Rhipidomella Michelin, L’Kv. Spirifer aff. clathratus, McCoy, Vau. » Sp. (striate. ? attenuatus). Spiriferina octoplicata, Mut. 6, Vau. Syringothyris texta, Hall. ? Reticularia aff. lineata, Mart. This is the fauna referred to by Vaughan at page 16 of his “ Belgian paper”? as agreeing closely with the Petit Granit of the Kcaussines area. It is therefore of C, age. The dip is slight, and a considerable area of each bed is exposed. The surface of many of the lower layers is covered with large colonies of Syringo- pora, weathered out in dome-shaped masses, very much as they must have appeared on the ancient sea-bottom. In very many cases the clumps can be seen to be attached to enormous valves of Conocardiwm fusiforme, McCoy (as noted on the Survey Map). Shells may be found covered with the Aulopora- like bases of colonies. The following points are noteworthy :— Chonetes is entirely absent. Caninia cylindrica is small and poorly developed, but appears to be at the Mut. Z or Mut. y stage. Schelwienella crenistria has not the Mut. C type of ribbing. 1 Poorly preserved. Convex brachial valve. Strong growth halts. Alternation of ribbing rarely marked. Cardinal angles not seen. 2Q.J.G.S., vol. Ixxi, 1915. Smyra— The Carboniferous Coast-Section at Malahide. 18} Zaphrentis delanouei occurs as high up as M 3, It is small, but a section immediately below the floor of the calyx (diam. 65mm.) has the typical shape of fosstila. Block E, lower part (M4d & M5). 180 feet = Z:. Abundant— Michelinia aff. megastoma, Phill. Cyathaxonia cornu, Mich. Zaphrentis densa, Carruthers. - delanouet, Kd. & H., Carruthers. Seminula, cf. ambigua, Sow. (in one bed). Camarotoechia aff. Mitcheldeanensis, Vau. Less common— é Amplexus coralloides, Sow. Syringopora 0, Vau. Productus burlingtonensis, Hall. Pustula subpustulosa, I. Thomas. ? Schelwienella crenistria, Mut. Z, Vau. Spirifer aff. clathratus, M‘Coy, Vau. Syringothyris laminosa, M‘Coy. x sp. (cuspidate). Athyris glabristria, Phill. Melasma sp. At M 4d the fauna changes markedly. That below no longer contains vesicular Caninia, and Zaph. delanowei becomes common. I have therefore assigned these beds to Z. The presence here of a small form of Micheluma megastoma would fit in with this correlation. Camarotoechia aft. Mitcheldeanensis forms a marked element in the fauna from M 4d to the base of the Malahide exposure. Seminula, ef. ambigua, has only been found here, and appears to be con- fined to one bed. If the correlation is correct, this is the lowest record of this genus of which I am aware. No attempt has been made to correlate any of the Malahide beds with horizon y. ‘The latter is defined as the band of overlap of two faunal assemblages in the S.W.P. It ceases to have a meaning where the assemblages and conditions differ considerably from those in the 8. W. P. 14 Scientific Proceedings, Royal Dublin Society. Block F (M6, 7, 8,9). 310 feet = Z. (M7 isa repetition of M6.) Abundant— Syringopora 0, Vau. (on Luomphalus). (M 6, 7.) Productus burlingtonensis, Hall. (M 6, 7, 8, 9.) Schelwienella crenistria, Mut. Z. Vau. (M 6, 7.) Syringothyris cartert, Hall. (M 8.) Athyris glabristria, Phill. (M8, 9.) © Camarotoechia mitcheldeanensis, Vau. (M9.) aff. mitcheldeanensis, Vau. (M 7.) tbe Less common— Amplexus coralloides, Sow. Zaphrentis omaliusi, Hd. & H., Carruthers. densa, Carruthers. 3 delanouet, Kd. & H., Carruthers. , Vaughania cleistoporoides, Garwood. Productus (Avonia), cf. youngianus, Davy. sp. (? aff. bassus, Vau.). ” Syringothyris laminosa, M‘Coy. Spirivferina aff. octoplicata, Sow.~ ? Reticularia aff. lineata, Mart. Orbiculoidea sp. It seems best simply to correlate these beds with Z, without attempting to divide them. Zaphrentis densa is common at the base, which might suggest that we have only got Z,. This, however, would be an unsafe infer- ence in view of the fact that there are some unexpected phenomena at Malahide; for example, Spirifer clathratus is only abundant well up in C, and Seminula occurs below C. Syringothyris carteri, Hall, which is extremely abundant in M 8, agrees in every way with Schuchert’s description of the American shell,’ even as regards the mode of preservation, which, he says, is usual, i.e. crushed in shale. The characteristic “twilled” ornament of Syringothyris is well seen in the Malahide material. This shell is reported from ( of Burrington? and from 8 and Z of Belgium.® ; 1 Ninth Ann. Rep. State Geologist. New York, p. 30. ? Reynolds and Vaughan, Q. J. G. S., vol. xvii, 1911 (‘‘ Burrington Paper”), p. 364 (S. typa, Winchell). 3 Vaughan, Q. J. G.S., vol. lxxi, 1915 (‘‘ Belgian Paper”’), pp. 8-10. SmyvrH—The Carboniferous Ooast-Section at Malahide. 15 Block G (M 10, 11, 12). 300 feet = Z. (Repetition of M 7, 8, 9.) Abundant— Syringopora @, Vau. On ELuomphalus in M 10. On Schelwienella in M 12. Zaphrentis delanouei, Ed. & H., and parallela, Carruthers. Productus burlingtonensis, Hall. fs (Avonia) sp. (? aff. bassus, Vau.). Schelwienella crenistria, Mut. Z, Vau. Syringothyris cartert, Hall. Athyris glabristria, Phill. Camarotoechia aff. mitcheldeanensis, Van. Less common— Spirifer roémerianus, de Kon. Humetria sp. These beds are a repetition of M 7, 8, 9. Owing to the sameness of litho- logy throughout, this was not recognized until the fauna had been carefully collected. he following succession, from above downwards, was noted in each series :— (a) Syringopora growing on Huomphalus. In the centre of syncline, at M 6-7, and immediately S. of C.G. fault (M 10), in beds well exposed owing to their strike being parallel to the coast for about 200 yards. (b) Camarotoechia beds. M 7 and M 11. (c) Schelwienella bed. M7 andM 12/3. (Indicated in the map by a chain-dot line.) (d) Athyris bed. M 7 and M 12. (e) Schelwienella bed. M 7 and M 12a. M 8 contains abundant Syringothyris carteri, and M 9 Athyris glabristria. These beds should be repeated at M 12a. ‘he latter beds are heavily covered by algae and animal growths, so that it is impossible to collect thoroughly. After some trouble three fossiliferous layers were found, two of which produced only Syringothyris carteri; the other (and lowest, uncovered in the Laminaria zone during a spring tide) gave a few specimens of Athyris glabristria. : The spacing of the beds (a) to (¢) in the two series, as well as their order, is in agreement. 16 Scientific Proceedings, Royai Dublin Society. Block H (M 13). 240 feet = C;. Abundant— Caninia cornucopiae, Mich., Carruthers. (Mostly in “cornu-bovis” stage.) Zaphrentis densa, Carruthers. Densiphyllum nodosum, L. B. Smyth. Michelinia cf. tenuisepia, Phill. Spirifer aff. clathratus, McCoy, Vau. | Less common— Amplexus coralloides, Sow. Cyathaxonia cornu, Mich. Michelinia gracilis, sp. nov. Vaughania cleistoporoides, Garwood. Zaphrentis omaliusi, Hd..& H., Carruthers. Bs junetoseptata, sp. nov. Caninia cylindrica, Scouler. Lophophyllum cf. costatwm, McCoy. Productus cf. concinnus (the St. Doulagh’s form). Pustula sp. Chonetes squamata, L. B. Smyth. Leptaena analoga, Phill. Schelwienella sp. Spirifer princeps, McCoy. ? Martima pingurs, Sow. Spiriferina octoplicata, Mut. 8, Vau. Athyris ef. glabristria (shallow sinus). Ehipidomella Michelini, L’Ev. “ Rhynchonella,” sp. CORRELATION OF BEDS SOUTH OF CARRICKHILL FAULT. These beds are characterized by the abundance of Caminia cornucopiae (“cornu-bovis” stage) and Spirifer clathratus. Zaphrentis densa, Michelinia cf. tenuisepta, and Martinia pinguis are also fairly common. This agrees with y of Burrington,! except in the matter of Spirifer clathratus, which, however, is the dominant Spirifer in y of Belgium.” 1 Reynolds and Vaughan, Q. J. G.S., vol. Ixvii, 1911 (‘‘ Burrington Paper”), p. 366. ? Vaughan, Q. J. G. S., vol. Ixxi, 1915 (‘‘ Belgian Paper”), p. 11. Smyra—The Carboniferous Coast-Section at Malahide. Ne Now we must place these beds either below all others in the section, i.e. low in Z, or above all, i.e. in C. We have seen that the evidence from structure is in favour of the latter, ic. down-throw to the south (see p. 10.) The following facts seem to me to settle the matter in favour of this interpretation :— 1. Two specimens of Caninia cylindrica, Scouler, were found: one near the middle, the other in the lowest exposed beds. 2. At M 13d, the lowest beds of the group under consideration, was found a fragmentary pedicle valve, about 40mm. in length, which agrees with Spirifer princeps, MeCoy, in its shape, its broad, flat ribs splitting into two unequal parts, and its fine reticulate pattern. This species is characteristic of the Tournaisian knolls (y — C,) of Belgium and Ireland, e.g. St. Doulagh’s, County Dublin. 3. Productus cf. concinnus agrees perfectly with a form common at St. Doulagh’s. The Producti found-elsewhere in the section do not. 4, Spiriferina octoplicata has rounded cardinal angles and strong lateral ribs, as in Mut. 8, Vau. 5, Michelinia cf. tenuisepta and Chonetes squamata occur in C of Rush. 6. Lophophyllum suggests the higher, rather than the lower, level. The evidence for regarding the M13 beds as the highest in the section may be tabulated as follows :— (a) The structure immediately at the fault (see p. 10). (6) The direction of throw of neighbouring parallel dislocations (p. 10). (c) Agreement with the railway fault in direction and amount of throw (p. 11). (d@) The fauna, just stated. The only evidence against this is the similarity in lithology of M13 to the “Lower Limestone Shales ” of the S. of Ireland and 8. Wales (see above, p. 11). The thickness of the beds exposed N. of the Carrickhill fault is about 780 feet. The stratigraphical throw of the fault therefore must be at least 780 feet + the unknown (but probably small) thickness of beds cut out by the N. fault. : CORRELATION WITH RUSH AND WITH THE N. W. PROVINCE. Owing to considerable difference in the fauna, close correlation with Rush could not be made. The whole of the Malahide exposure seems to be equivalent to part of the Rush Slates, Z, - C, (see L. B. Smyth, Proc. Rk. Dub. Soe., vol. xiv (N.S.), p. 535. 18 Scientific Proceedings, Royal Dublin Society. The Malahide beds resemble the Solenopora sub-zone (y — C;) of the N.W. Proy. in the abundance of Athyris glabristria and the occurrence of Vaughania cleistoporoides. Several other species, too, are common to both (see Garwood, Q.J.G.S., vol. xviii, 1912, p. 460). II]. PALAEONTOLOGICAL NOTES. Michelinia gracilis, sp. nov. (PI. I, fig. 5, and text fig. 1) M13=C,. Description—Corallum of two to four long, flexuous, tubular corallites. Greatest length observed 5cm., but the specimen was incomplete. Diameter of largest corallites 5mm. Epitheca minutely wrinkled transversely. Group- ing of tubes often, but not always, m one plane, recalling Halysites. Tubes usually in contact, but sometimes diverging near their extremities. ‘Tabulae numerous, convex upwards, mostly complete, but often running together. No trace of septa. Mural pores few, irregularly distributed. When a corallite first appears it is D-shaped, or, if added between two others, triangular. In either case, the common wall is slightly convex towards the older AE URS aa Fig. 1.—WMichelinia gracilis, sp. noy. 1-7, serial section of one specimen; 8, longitudinal section of another specimen, showing divarication. Natural size. corallite. Comparison.—I had assigned this coral provisionally to Beawmontia until mural pores were proved. The tabulae are exactly as in B. egertoni, M.-Ed. & H., and in the divergence of the corallites it recalls B.. laa M‘Coy. SmyvtrH—The Carboniferous Coast-Section at Malahide. 19 Micheliia aff. megastoma, Phill. M4 and top of M5 = Z, - C;. Description.—Height up to at least 20mm. Calyces 20 or more, sub-_ equal, shallow, with septal striae. Average diameter of calyces, 7 or 8 mm, Abundant vesicular tissue extending from base often up to height of 17 mm. Stereoplasm moderate, not obliterating vesicular tissue. Other characters as in M. megastoma. Near the top of Md one bed contains, in addition to the larger colonies, young ones in various stages, the smallest found having but four corallites. Comparisons.—It agrees very well with de Koninck’s illustration of M. megastoma in his “ Recherches sur les Animaux Fossiles du Terr. Carb. de la Belgique,’ Pl. XIII, figs. 3-30. It is evidently the form referred by Vaughan! to his JZ. megastoma (Phill), Mut. Z,. I find it to differ, however, from his description as follows :—- Number of corallites greater, calyces shallow, vesicular tissue abundant, and not nearly concealed by stereoplasmm. From M. megastoma of the Rush Conglomerate (C,) it seems to differ only in dimensions, the average calyx diameter of the Rush form being 10-12 mm. The present form is probably an earlier mutation of that found at Rush, and possibly intermediate between it and Vaughan’s Mut. Z,. Or it may be a variety (i.e. contemporary close relative) of Mut. Z,; or even a local form of Mut. Z, due to environment. Zaphrentis. The commonest? species at Malahide is Z. densa, Carruthers, which is common at many levels throughout the exposure. A few specimens of the typical Z. omaliusi, Ed. & H., were identified. 7% ambigua, Carruthers, was not found. Z. delanouet, Kd. & H., s.s., occurs from M 3 down to M 9, and is common at M5and M11. A section below the calyx of an M 4 specimen has a diameter of 10mm. Z. parallela, Carruthers, was noted at M 4 (top) and M 11. Zaphrentis junctoseptata, sp. nov. (Pl. I, figs. 1-4). M13 = C,. Deseription.—Irvegularly cornute. Often one or more strong constric- tions. Epitheca finely wrinkled. Well marked, flattened costae. Length up to about 20mm. Diameter of calyx about 9mm. Major septa reach the centre, except just below a tabula. Cardinal fossula on convex side, bounded *Q. J. G.S., vol. Ixvii, 1911 (‘‘ Burrington Paper’), p. 371. 2 Cf. Vaughan, Q. J. G. S., vol. Ixxi, 1915 (‘‘ Belgian Paper”’), p. 34. 20 Scientific Proceedings, Royal Dublin Society. by two major septa. Cardinal septum as long as the rest. Septa more ox less concave to cardinal fossula. Minor septa long, each leaning against a major on the side remote from the cardinal fossula. The counter septum thus has two minors leaning against it. The tabulae are irregular, often very oblique. This produces the group- ing of septa seen, eg. in Pl. I, fig. 2, and to a less degree in figs. la and 4a. Symmetrical sections are uncommon. The obliquity of the tabulae has no relation to the position of the fossula (cf. Pl. I, figs. la and 2). Rejuvenescence was observed in some specimens. The youngest section obtained (38°5 mm. diam.) resembles a young 4. omaliusi, except that already there are two minor septa leaning against the counter septum, and of nearly half its length. ‘The other minors are barely indicated. Comparisons.—This coral is convergent in septal plan with Cyathaxonia cornu, with which it is associated. : It bears a strong resemblance to Zaphrentis cassedayi, M.-Ed.,' but is distinguished therefrom by the absence of spines on the epitheca. From Z. delépini, Vau.,? it is easily separated by the great development of all the minor septa. In other respects there is considerable resemblance. It differs from Z% vaughani, Douglas,’ in this same way, and also in the following points:—The fossula is on the convex side, and is bounded by a single septum on each side; the cardinal septum reaches the centre; the septa are concave to the fossula. Endophyllum irregulare, sp. nov. (PI. I, figs. 7-9.) M 2,4 = (,. Description.rLength 55mm. Growth very irregular. Strong thicken- ings at intervals. Transverse section, mature.—Diameter 17 mm. Central area occupied by tabular intersections only. The medial area contains 37 major septa, very unequal in length, sometimes reaching nearly to the centre, some- times reduced to short spikes, not thickened, often flexuous. Cardinal septum doubtfully to be distinguished. Minor septa appear as teeth on the outer wall. Outer area composed of vesicles which may all be large and — loose, with tooth-like septal projections on their outer walls; or may nearly all be interseptal, so that most of the septa reach the outer wall without 1H. Milne Edwards, Histoire Naturelle des Coralliaires, 1857-60, Pl. G1, fig 2. 2 A. Vaughan, Q. J. G. S., vol. lxxi, 1915 (‘‘ Belgian Paper”), p. 34, and PI. iv, figs. 3 and 4. 3J. A. Douglas, Q. J.'G. S., vol. Ixv, 1909, p. 577, and Pl. xxvii, fig. 11. A. Vaughan, Q. J. G. S., vol. Ixxi, 1915 (‘‘ Belgian Paper”), p. 35, and PI. iv, fig 7*. Smyra—The Carboniferous Coast-Section at Malahide. 21 discontinuity. Any intermediate state may occur, and the extremes may be shown by closely contiguous sections of one individual. The outer area is usually more developed on one side, and may be entirely absent from the other. Transverse section, young.—Only major septa present, not reaching the centre, sometimes very short. No outer area. Cardinal septum may be shorter than the rest. Longitudinal section.—Tabulae close, wide, and flat, bent down at the side, where they meet the marginal vesicles. Comparisons.—This coral closely resembles the Upper Devonian Z. pris- cum, Munster, as figured by Frech.! The latter differs from our form in having minor septa on its inner wall, and in its more regular habit. From J. burringtonense, Vau,? the present species differs in the absence of the peculiar grouping of the septa in that form, and in the smaller development of minor septa. Vaughan records his species from Malahide, but I have only found the form described above. It is interesting to compare— Endophyllum irregulare, sp. nov., C,, Malahide. Thysanophyllum pseudovermiculare, McCoy, C,.., N.W.P? Endophyllum cf. pseudovermiculare, Vaughan, lower S, Belgium.‘ Lophophyllum cf. costatwm, McCoy. (PI. I, fig. 6). M 13 = C). Description.rOnly two imperfect specimens were found, and transverse sections alone obtained. Transverse section :—Diameter of calyx, 10°5 mm.; major septa, 23 in one specimen, 20 in the other, reaching about half way to the centre. Minor septa barely indicated at rim of calyx. Cardinal septum shorter than the rest. Septa in cardinal quadrants thickened so as to be in contact. Tabular intersections, 3 or 4. Columella continuous with counter septum, very slightly thickened, with 4 to 6 rudimentary lamellae on each side. In the calyx the columella forms a strongly thickened boss, oval in section, 4 x 25mm. One row of vesicles occurs on the calyx wall. Comparison.—This coral may well be a precursor of that described by Wilmore as Lophophyllum costatum, McCoy,? from D, of the Midlands 1 Zeitsch. Deutsch. Geol. Gesellsch., vol. xxxvii, 1885, p. 76, Pl. vii, fig. 2, and Pl. x, fig. 2. * Reynolds and Vaughan, Q. J. G. S., vol. lxvii, 1911 (“‘ Burrington Paper”), p. 377, and Pl. xxx, fig. 4. 3 Garwood, Q. J. G. S., vol. Ixviii, 1912, p. 562, and PI. xlix, fig. 2. *Q.J.G.S., vol. Ixxi, 1915 (‘‘ Belgian Paper”), p. 39, and Pl. vy, fig. 3. 5Q. J. G.S., vol. lxvi, 1910, p. 573, and Pl. x1, figs. 1-4. 22 Scientific Proceedings, Royal Dublin Society. of England. It differs from that species as follows:—Smaller size; major septa fewer, shorter; minor septa scarcely developed; vesicular zone rudimentary. Camarotoechia aff. mitcheldeanensis, Vau. This only differs from Vaughan’s species in having usually 4 ribs on the fold and 3 in the sinus, and in its shghtly larger size. The internal structure is essentially the same. Athyris cf. glabristria, Phill. There are two distinct forms included under this description :— (a) A gibbous shell, whose fold and sinus appear early and become extremely marked with age. It seems to agree with the one figured by Garwood as the index of his zone A. (>) A larger, more depressed shell, whose fold and sinus appear late, and only cause a slight undulation in the valve intersection. This is evidently the form referred to by Vaughan as A. lamellosa-glabristria.? Both forms vary considerably, but I have not found a full-grown individual that could not be unhesitatingly referred to either(a) or (0). Both are usually transverse. The ornament of both is the same, and is like Davidson’s Plate XVIII, fig. 10, there being about two fringes per millimetre. The narrow lamellae, where the fringe is stripped off, show an alternation in strength, one, two, or three weak ones occurring between two stronger ones. Strongly marked growth halts, as noted by Vaughan, occur in both. Form @ is abundant at two different levels—M 1 and M 8, 9 (and their repetition M12). Form 0 is abundant and large throughout M 1, 2, 3, and 4. Seminula cf. ambigua, Sow. M5=2Z,. This is a very small shell. The largest specimen had the dimensions :— Length, 12mm.; width, 14mm.; depth, 5mm. The proportions are approxi- mately as in Davidson, Brit. Carb. Brach., Plate XV, fig. 17, but the beak is sharper and less arched. Previous Work ON THE SECTION. The late Dr. Vaughan intended at one time to publish an account of the — Malahide fauna. He has referred to it in the Burrington and Belgian papers.’ In these the following fossils and horizons are mentioned as occurring at Malahide — 1 Proc. Geol. Assoc., vol. xxvii, Pl. xii, fig. 3. 2Q. J. G.S., vol. bxxi, 1915 (‘ Belgian Paper), p. 16. 3 Reynolds and Vaughan, Q. J.G.S8., vol. Ixvii, 1911. Vaughan, Q. J. G.S., vol. Ixxi, 1915. e SmytH— The Carboniferous Coast-Section at Malahide. 23 Burrington paper (1911)— Endophyllum burringtonense. Vau. y (p. 379). Michelinia megastoma. Mut. Z,. Zz (p. 371). Belgian paper (1915)— Athyris lamellosa-glabristria. C, (p. 16). Syringothyris laminosa. C, (p. 16 and pl. vi, fig. 1; p. 48, foot-note). Michelinia, Syriugopora, and the giant Conocardiwm. C; (p. 16). Zaphrentis delanouei and Z. densa. Cy, (p. 16, foot-note). Zaphrentis delanouei and Mut. parallela. y (p. 34). Before commencing work I had at my disposal, in addition to the references just quoted, slides given me some years ago by Dr. Vaughan labelled thus :— Zaphrentis delanouet. Z, Malahide. Cyathaxonia cornu. §. of fault, Malahide = Rush Slates. Michelinia cf. tenwisepta. Rush Slates, Malahide. Zaphrentis omaliust, var. densa. Carr. Z, Malahide. Densiphyllum. Malahide, S. of fault = Rush Slates. (This is D. nodosum. See Proc., R. Dub. Soc., vol. xiv (N.S.), p. 556.) When the work for this paper was nearly complete, I was enabled, throwgh the kindnegs of Professor 8. H. Reynolds, to look through part of Vaughan’s Malahide collection which had come into his possession. It is small, and contains nothing I had not already found. Besides it is obviously not complete as, e.g., Hndophyllum burringtonense is absent. There is no key to the symbols he used, and I could only correlate them with mine in a few cases. It may interest some if I set them out. The second column contains correlations found on some of the labelled specimens. PC, Laminosa beds O N M ay, K base of y J. Yo =M5 Il=J H G WT Vn =M8 E 4Z, =M9 D C Zy B base of Z,, beds S. of fault = Rush Slates 6a, 60. =M 13 A It appears from this that the beds S. of Carrickhill fault were regarded as the lowest in the exposure. 24 Scientific Proceedings, Royal Dublin Society. SUMMARY OF THE CHIEF POINTS IN THE PAPER, 1. There are three faults of importance in the exposure, not only one, as previously believed. 2. The southern (Carrickhill) fault has the down-throw to the south, not to the north. 3. Correlation is made with C, — Z of the 8.W. Province of England. 4, Three new species are described and figured, namely :—Zaphrentis junctoseptata, Micheliniu gracilis, Endophyllum wrregulare. 5. Another form, which appears to be new, is described and figured, of which sufficient material could not be obtained for a satisfactory study, namely :—Lophophyllum cf. costatum, McCoy. 6. A Seminula (ef. ambigua) is reported in a Z, fauna. PLATE I.—Corats rrom tHe ‘‘C”’ Zone at Mananiwe. Fies. 1 to 4. Zaphrentis junctoseptata, sp. nov. M 13 (p. 19). 1, a, b, c. Photographs of three transverse sections from the same specimen. In le note the presence of two minors leaning against the counter septum, although the majors have not yet all appeared. x 2:7. 2. Camera lucida drawing of mature transverse section of a second specimen. x 6. 3. Photograph of mature transverse section of a third specimen. x 2:7. 4, a, b, Camera lucida drawings of two transverse sections of a fourth speci- men. x5. Fig. 5. Michelinia gracilis, sp. nov. M 13 (p. 18). A weathered specimen, consisting of two corallites, on a slab of limestone. Opposite the arrow a portion of the wrinkled epitleca is preserved. Just below this, some of the tabulae are exposed. x 0°6. Fic. 6. Lophophyllum ef. costatwm, M‘Coy. M 18 (p. 21). Transverse section just below floor of calyx. x 2°7. j Fies. 7 to 9. Hndophyllum irregulare, sp. nov. M 2 (p, 20). 7. Drawing of a specimen showing external characters. x 0:5. 8, a, b,c. Three sections from another specimen. x 2:7. 9. Portion of longitudinal section of a third specimen. x 2:7. Fic. 10. Caninia cylindrica, Scouler. M 138. Transverse section of immature stage. This is from the specimen referred to on page 17 as occurring at ‘‘ near the middle’”’ of M 13. It is figured because this early stage seems to be rather rarely found, and it is larger than the corresponding stage figured by Salée in ‘‘ Le Genre Caninia,” Nouv. Mém. Soc. belge Géol., Mém. No. 8 (1910). x 2:7. PLATE II. ; The left-hand figure is a plan of the outcrop at high-water level. ‘The topography is reduced from the Ordnance Survey 25-inch map. The right-hand figure is a horizontal section along the line marked on the lan. 3 HE, F,G, H. Blocks into which the section is divided by the three faults, M1-M138. Symbols for groups of beds, used in labelling the fossils. H (= M 138) is a higher horizon than M 1. G is a repetition of F. M4d. A bed taken as the top of the ‘‘ Zaphrentis”’ zone. M 12,aand 8. Two beds of Schelwienella. M 12, in G, and the same bed in F’, are marked by a chain-dot line. Plate I. Vol. XVI. NS. D Scient. Proc. R.Dublin Soc GI! ot Ap b Bemrose, Collo., Derby. Esai pt 2) iy iE 4K / a PLATE II. PLATE II. o/ SCIENT. PROC. R. DUBLIN SOC., NS, VOL. XVI, ° Vise ROBSWALLS CASTLE Mad od <4 i COASTGUARD STATION ih ELV WOES CARRICKHILL MARTELLO TOWER Grey Umestene band SCIENTIFIC "PROCEEDINGS. VOLUME XVI. 1. A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, S0.D., F.R.s., and T. G. Mason, m.a., so.p. (January, 1920.) 6d. 2. The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. SMYTH, B.A., so.B. (January, 1920.) 1s. DUBLIN.: PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIBBS. KAS te hi 7 aot) a ; iy THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 3. FEBRUARY, 1920. THE APPLICATION OF THE FOOD-UNIT METHOD TO THE FATTENING OF CATTLE... . BY y. Gmitreentan (rez > JAMES WILSON, M.A., B.Sc., \ wee 1921 ormman prewee~ PROFESSOR OF AGRICULTURE IN THE ROYAL COLLEGE OF SOIENCE, DUBLIN. (PLATES III., IV.) [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN : PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN LONDON, W.C. 2. 1920. Price One Shilling. Koval Bublin Soctety. FOUNDED, A.D. 1731. INCORPORATED, 1749. EVENING SCIENTIFIC MEETINGS. Tux Scientific Meetings of the Society are usually held at 4-15 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Hditor. a nomen (1 [wo) Or =) JOU. THE APPLICATION OF THE FOOD-UNIT METHOD TO THE FATTENING OF CATTLE. By JAMES WILSON, M.A., B.Sc., Professor of Agriculture in the Royal College of Science, Dublin. (Puares TII., LV.) Read January 27. Published Feprvary 20, 1920. Durine the last thirty years the Danes and Swedes have greatly advanced our knowledge of the use of feeding-stuffs for certain kinds of live-stock. They have worked chiefly at two sides of the problem, namely, at determin- ing (1) the relative feeding values of the usual feeding stuffs, and (2) the quantity of food required, during all stages of lactation and non-lactation, by cows of different sizes and milk-yielding capacities. Under the first head, they have found that, when they form a fair—ie., not extreme—proportion of the total ration, 2°5 lbs. of meadow hay, 4 lbs. of oat straw, 10 1bs. of mangels, and 1-2 1b. of oats have, on the average, the same feeding value as a pound of barley. ‘Then, if a pound of barley be taken as the food-unit,’ it can be said that 2°5 lbs. of meadow hay, 4 lbs. of oat straw, 10 lbs. of mangels, and 1:2 Ib. of oats each contain a food-unit. And, if a cow consume 20 lbs. of meadow hay, 80 lbs. of mangels, and 6 lbs. of oats, it can be said that she consumes 21 food-units :—20 = 25+80+10+6+12=84+8+4+5=21. Under the second head, they have found that cows’ needs vary with their weight, advance in pregnancy, and milk yield. ‘The food-units necessary to maintain barren cows or heifers of different weight, or a bullock, for that matter, in constant condition—i.e., becoming neither fatter nor leaner—are approximately as follows :— Live Weights. Food Units. 7 cwts. 55 Shee 6:25 ® 5 7 1@ 5 eS iD as 3 8°5 Shortly after becoming in calf, the cow’s needs increase till just upon calving time, when, for all weights alike, three more units are necessary. After calving, another half unit is necessary for the first ten pounds, and thereafter 1 The Swedes take a kilogramme of barley as the unit. SCIENT. PROG. R.D.S., VOL XVI, NO. III. D 26 Scientific Proceedings, Royal Dublin Society. a third of a unit for every additional pound of milk. These findings are expressed in the diagram (Plate III.). The use of this diagram may be illustrated thus:—A nine hundredweight cow giving five gallons of milk needs about 24 food-units a day. ‘This number might be extracted from 96 pounds (24 x 4) of oat straw, or 60 pounds (24 x 3:5) of meadow hay ; but, since no cow can deal with even the smaller of these quantities in a day, the total bulk of the ration must be brought down to the cow’s capacity by reducing the long fodder and substituting something which, by being readily digestible, passes quickly from the stomach, or, by having its nutritive ingredients highly concentrated, takes up little space. The necessary food-units might also be extracted from 240 pounds (24 x 10) of mangels, but, though some cows might consume this quantity, few could do so and remain in good health. Besides, the digesti- bility of the whole ration is lowered when roots or long fodder or both are fed in large quantity. Thus, to keep the whole ration within the stomach- capacity, a cow giving a large yield of milk must have concentrated foods in addition to long fodder and roots. The following may be taken as an example ration for a five-gallon cow weighing about 9 cwts., in which the limits of total bulk are not overstepped :— Meadow Hay, 173 lbs.= 7 units. Mangels, FO aN te 7, Concentrates, 10 ,, =10 ” ” As the cow’s yield decreases, or for a lower-yielding cow, straw may be substituted for hay, and, first, concentrates then mangels decreased till, when she is dry and far from calving time or not in calf at all, the ration may consist of straw alone. Since concentrated feeding stuffs came mto use, in the first half of the nineteenth century, and winter fattening of stock became possible, many fattening experiments have been carried out in England and Scotland, some of which have been described in the Journals of the English aud Scotch Agricultural Societies, the Journal of the Board of Agriculture, and reports sent out by agricultural colleges. A very carefully compiled . digest of these experiments was published in the 1909 volume of the “Transactions of the Highland and Agricultural Society of Scotland,” by Mr. Herbert Ingle, B.Sc. For each experiment Mr. Ingle gives, among others, the following data :— 1. The experimenter’s name. 2. The date of the experiment, Wison—Food-Unit Method in the Fattening of Cattle. a0 The age, breed, and sex of the experiniental animals; The duration of the experiment. Details of the average ration. The total digestible albuminoids in the ration. SID of Ee oo The starch-equivalents of the digestible fats, carbohydrates, amides, and fibre in each ration. 8. The total digestible matter and the total digestible albuminoids and starch-equivalent of the rest of the digestible matter in each ration per 1000 lbs. live weight of the experimental animals. 9. The albuminoid ratio of each ration, 10. The digestible matter consumed for every pound of live weight increase. 11. The daily live weight increase of the animals. 12. The average live weight of the animals during the experiment. Early in the latter half of the nineteenth century the theory was pro- pounded that the efficiency of a ration depended upon its albuminoid ratio, that is, upon the proportion of its digestible albuminoids to the digestible carbohydrates, fats, and fibre together. Investigators hke Wolff and Kithn suggested ratios from 1:5 to 1 :6 for fattening stock ; but, though this had the effect of raising the money value of albumen far beyond what its fattening power could have justified, the albuminoid ratio theory never seriously affected British stock fatteners, and towards the end of the century it began to lose eredit even among those who had been taught by its chief exponents. Mr. Ingle’s digest shows that it was anything but a sure and certain guide. It was then seen by some investigators that even the total amount of dry matter in a mixed ration and, still more, the total digestible matter was a better measure of the ration’s efficiency. Of course, it was realized that every animal’s albuminous waste must be made good and its needs supplied for the production of growing tissue; but, when Kellner’s work became known, it was obvious that his proposal was still better, namely, that, when sufficient albumen has been fed to meet albuminous waste and produce growing tissue, the test for the rest of the ration should be the proportion it contains of digestible starch, fat, and fibre: the two latter being reduced to their energy-producing equivalent in starch, and the three together called the “ starch equivalent.” But few farmers have the technical knowledge with which alone Kellner’s proposals are to be adapted to a particular case; still fewer when they are expressed in calories. p 2 28 Setentific Proceedings, Royal Dublin Society. In the experimental work so far carried out, one important question has not been answered. No experimenter has definitely determined the maxi- mum daily increase bullocks are capable of making and the food necessary to produce it, and farmers, being more immediately concerned, perhaps, with the mere conversion of straw and roots into manure, have not asked the question. The need for knowing the possible maximum increase may not be obvious, but it is none the less important, for, since a certain portion of his daily ration—roughly, about a third in the case of a fattening bullock, and frequently nearly the whole, sometimes more, in that of a store bullock —is required for mere maintenance of life, while the rest goes towards increase in live weight, it is clear that the bullock which fattens the quickest, and so shortens the expenditure in mere maintenance, requires, over all, the smallest amount of food for the beef produced. Missing this question, experimenters, as a rule, have usually sought to deal merely with questions of the moment which occurred to farmers in their customary practice. When turnips came into use, but before the introduction of concentrates, the first systematic experiments (Thaer’s) were devised to find how many turnips should be substituted for the hay with- drawn from the old ration of hay alone and have the bullocks continue in the old condition. In time, it was found that too many turnips could be fed, but nobody found what might be called the limit of safety. Had some one determined the optimum proportions of straw, or hay, or both, with turnips, it is possible that, when concentrates were introduced, experi- menters might have tried to determine how much extra live weight was produced by the addition of concentrates, and then raised the question: what is the possible live weight increase? British experimenters have followed no clear policy, and so no erucial question has been sufficiently investigated to become the starting-point for a successor. For the most part, the experiments have been mere trials of varying quantities of roots or straw, or of one or more food-stuffs against each other. By using the Scandinavian system of food-units, it may be possible, however, to make a great part of the recorded British experiments comple- mentary to each other, and to draw some general conclusions which may be approximately accurate and, so far, useful. This may be done by calculating the food-units contained in the rations and plotting them against the daily increases each ration produced, as in the diagram (Plate IV.). To bring out the point that the food consumed rises with their weights, the experimental animals have been divided into groups of ascending weights and the plottings for each group set out in separate sets. The weights are the averages during the experimental periods. Witson—Lood-Unit Method in the Fattening of Cattle. 29 In order to avoid extremes, only experiments which lasted over 70 and under 170 days have been made use of. Certain experiments could not be used because their food-ingredients were too loosely described. “ Roots,” for instance, may be of several kinds. In one set of experiments, “roots” were found, from the original paper, to be almost equal quantities of mangels and swedes, and the food-units are calculated on this understanding. In some experiments “common cotton cake” was fed, and, on reference to the compositions given in the original papers, this was found to be undecorticated cotton cake. One important series of experiments could not be used at all, because the amount of straw consumed was not stated. One extremely large, probably an impossible, ration, and another set, in which the gains are suspiciously large for the food consumed, have been omitted. Modern experimenters usually have the animals fasted before weighing them. It is not certain that the older experimenters had this done, but, as they are likely to have followed the same practice at both ends of the fatten- ing period, the errors resulting from weighing the animals unfasted are not vicious when different sets of experiments are being used together. There are more serious causes of variation which cannot be taken into account, as, eg. the varying capacity of cattlemen; but the results in one direction may reasonably be assumed to be cancelled by those in another. This remark applies equally, no doubt, to the experimenters themselves, and to some extent reduces the value of the deductions to be made from their work. When the data had been plotted, as in the diagram, it was obvious that a straight line could not be drawn through the plottings by the eye. Then the average food-units consumed per group and the daily gains produced were calculated and plotted on the diagrain in crosses. The spaced line which joins these crosses, however, is neither straight nor regularly curved, and, if it were to be used, the factors which make it crooked would have had to be found and allowed for. Therefore another line must be found. The ideal line would be that which should show, for all weights of cattle, the highest possible daily gains and the lowest possible food-units necessary to produce them; but such a line cannot yet be drawn. The highest daily gain shown in the diagram is 3:1 lbs., made by a lot of 12-ewt. bullocks, upon ~ 19°8 food-units a day ; but we do not know that this is the possible limit. A Scotch experimenter (Dr. J. W. Paterson) had a lot of bullocks which made 3°66 lbs. a day, over a period of eighty-eight days, upon pasture, cake, and meal; and, last summer, at Clonakilty Agricultural Station, four bullocks, 30 Scientifie Proceedings, Royal Dublin Society. over a period of six weeks, made a daily gain of 3°75 lbs. upon pasture alone. One of the four bullocks made four pounds a day. Jt may be that similar gains are not impossible upon long fodder, roots, and concentrates, and the units necessary to produce them may be less than in the proportion of 19°8 units for 3-1 lbs. increase. All that can be done in the meantime is to indicate the daily gains shown to be readily attainable—not the maximum —with a reasonable—unot the minimum—nuimber of food-units. This is done by the continuous straight line drawn across the diagram. It will be seen that this line runs well above, or just under, the numbers of food-units which have produced daily gains of two pounds or over, namely :— 14:25 units for 6-cwt. beasts. 15 4 TOW 55 15:75 os S=c wie 16°5 wi 9-cwt. ee 17°25 é 10-cwt. __,, 18 “i li-ewt. _,, 18°75 3 12-ewt. But this line must not be used without the caution that the experiments show the rations to have very variable efficiencies. It will be seen that in every group a large number of the rations do not produce two pounds a day. If lines be drawn upwards through the average of the dots in every group, the dots which fall to the left will indicate the less effective, those to the right the more effective, rations. Not only so, but the distances of the dots from the upright lines may be taken as indications of the efficiency of the rations concerned. On this understanding the following table shows the relative inefficiencies (—) and efficiencies (+) of the rations in each group :— Witsun—Food- Unit Method in the Futtening of Cattle. 31 Nore :—The rations marked vr received in addition small quantities of treacle or molasses. Ration 1N LBS. Noe nen Cea col MORES le | Biicteney. ~ | Straw. | Hav. | Roots. Concen- z as. trates. 1 (11:59 1-51 0 8-5 26-5 4-5 = 6 2 11-59 1:67 0 8:5 26-5 4-5 = § 3 | 6 cwts. 12-34 1-85 0 8 46 4:03 = 9 4 | 12-34 1:90 0 8 46 4-039 =f 5 (11-37 2-07 | 0 85) | 26:0) |= 45 +2 6 15°34 1-73 0 77 56 5-56n = § 7 14-14 1:69 © |i ilo 28 6-06r 05} 8 16-14 1-87 Oey ieelz:5 9) 242 7 = 9 9 16-07 1°81 0 | no 42 7 =) 10 16°15 1:87 0 12°5 49 6 = 9 11 7 ewts. < 13°67 1:68 0 We 65 34 - 2 12 13°67 1:76 0 ie 65 34 - 1 13 114-18 1:96 5°77 7 50 4°37 0 14 | 14:18 1:96 5°77 7 50 4-36 0 15 | 12-78 1:85 | 0 1463 | 0 6 1 al 16 (14-06 2-065 0 15 Selle 5-529 4 D 17 (18°11 1-44 8 0 110 4 eR 18 121 “96 0 0 109 0 Ne 19 16 1-42 | 13-6 0 0 10-4 BMY 20 16-35 171 0 12:5 42 6 aha 21 [aces 1:63: | 0 14 56 4-5 Lg 22 | 8 cewts.| 15-53 1-73 0 13-8 56 4-58 = 9 23 15°24 1°82 0 13:8 56 3-82 ay 24 16-58 2:14 | 10 5 50 6 Atel 25 [16-58 2-26 | 10 5 50 6 +2 26 | 163 2-21 1:57 | 7:44 | 73 5-36 a 9 27 li6-34 2-42 1°57 744 | 73 5°36 + 4 32 Scientific Proceedings, Royal Dublin Soctety. RATION IN LBS. No: lane: | consumed | Gu Salone CaN nRRR ob aes eRe sees cee ~ "| Straw. Hay. | Roots. cannes % a lbs. 28 (15°68 +88 7 0 126 0 = 10 29 20°19 1:49 7 0 130 4 - 10 30 20:08 1:43 3 7 100 5:8 = 9 31 BE 1-16 0 ql 126 0 = 9 32 [eis 1-72 4 4 105 8°84 2 6 38 20:27 1:57 7 0 130 3:67 = 8 34 lb 1:93 4 4 105 8°85 en) 35 ee L-11 8°5 0 50 7:25 = 7 36 21-72 1:81 4 4 105 8:86 = 7 37 jess 1:43 7 0 150 0 Sat) 38 17-75 1:38 3-75 | 8:25 | 955 3-49 BiG 39 20-08 1:78 3 7 100 5:3 NG 40 17-75 1:49 3°75 | 8:25 | 95-5 3:49 216 41 17-22 1:48 5 0 114 3-02 = @ 42 16-77 1:47 816 | 0 112 3 - 6 43 16:92 1-43 | 10 0 75 5:8 2G 44 20-08 1:81 3 7 100 5:8 = 6 45 | 9ewts.| <¢ 19:03 171 | 10 0 100 0 = 6 46 18-41 1-64 5 0 120 3°88 BS tg 47 18-58 1-67 5 0 108 3:88 = 6 48 13-90 1-20 9:97 | 0 50 6°67 = 6 49 15-28 1-42 9:08 | 0 5() 10 25 50 14-77 1:33 8:36 | 0 50 9°37 =. 8 51 16-77 1:61 816 | 0 112 3 a8 52 ee 1-26 7-9 0 50 8-5 par 53 16:00 1-54 | 10 0 80 4 ae) 54 15°16 1:43 8 0 84 3°88 ia 55 15°25 1:51 7 0 95 4 = 3 56 19°54 1-98 3:54 | 3°54 | 56 9°74 BS 57 17-11 1:70 | 10 0 100 4 ees 58 Poser 1:89 8 0 109 3 eel 69 | 13-75 eas |G 0 120 0 Ai 60 13-48 1:60 5 0 84 3:88 eal 61 15°83 1-94 0 16 28 65 0 62 19°24 2°38 3-07 | 3:08 | 90 6-1 4 il 63 (19:30 242 3:78 | 3:78 | 56 9°74 + 2 Witson—Food- Unit Method in the Fattening of Cattle. 33 Ration IN LBs. No. Live Food-units | Daily —-_———_ -——___.. Relative | : Weight.| consumed. | gain. i ; + l@oncene’ |) aiciency.. | re Straw. | Hay. | Roots. | Gates | ———’ 64 (24:09 1:73- | 4 4 112 8-7 = 9 65 92°33 1-69 4 112 8:7 = f 66 22-02 Teh) |) abs 0 90 6 Hy, Cie iiovertan| | ew 2°10 4 4 112 8-7 = @ 68 15:49 1°31 0 22 0 6 = 5; 69 24-97 2°38 a 0 155 3 Sie 70 | 19:05 1:97 0 0 131-11 4:6 aD | 71 | (16-41 2-48 8 0 108-75 3 Ae 6 > | 22 {22°76 1:33 4 4 112 8:6 = it 73 | 22-93 1°52 4 4 112 8-6 - 9 74 23-04 1°71 4 4 112 8:6 = 7 75 18°78 1:50 9°5 5 0 13 = 6 76 19°22 1:59 | 15 0 90 6 =. 8 ad 16°19 1°35 9-6 0 45 8 ae 78 16°51 1°50 271 7:08 29 10°23 a 79 15-61 1:37 @ 7 0 9 1 - 4 80 14-60 1:28 14-1 0 0 10 = 4 81 29°51 2-11 0 0 127-03 7-72 - 3 Bowne ees 198 | 15 0 90 8°59 - 3 83 jae 2-06 | 14 7 63 8 8 84 + 19-11 1:97 0 8°88 | 40-34 10°18 = i 85 12-18 1°35 0 17 48-4 0 = i 86 17:42 1:98 2-92 6°55 | 382-4 9-64 0 87 19:07 2-29 oO 8-88 | 40 10:18 4 i 88 oe 2-60 14 7 63 8 & 9) 89 | 16-72 2-15 4 6:97 | 39:9 9 i ® 90 eee 2335 471 6:9 39-5 9 4+ 5 91 (16:06 2°54 0 15-49 | 44-3 4:32 fy SCIENT. PROC. R.D.S., VOL. XVI, NO. Ill. E 34 Scientific Proceedings, Royal Dublin Society. | Ration IN LBs. Nos I Weignel|veonsumel: | gaits | (7) at 2 a ane uateae mmemciene fea | Straw. | Hay. | Roots. | ea 92 (23°11 1:64 9°5 5 82 | 12:5 - 8 93 19:90 1:60 | 11 5°5 AB | oa we = 6 94 23:06 1:92 7 7 70 | 10:5 25 95 23°81 2-11 5 5 We) OB = 4 96 | 2202 1:95 5 5 107 8°8 = 4 97 | 15-74 1:26 | 11 5:5 a | 947) - 4 98 | 23-11 207 | 9:5 5 m | so) = A 99 | 20°83 1-91 5b Se | a ee 3 100 18-98 1:78 | 11 5:5 56 9-4 1 So. § 101 | 18-08 1°78 8 8 49 8°8 oY 102 23-11 | 2:31 9°5 5 $2 13°5 = il WOE lam atin 19:07 1-90 0 10 95 5:35 = il 104 18-37 1:82 4 12 34-6 8-1 =a) 105 | 17:78 1:86 ® | igen 39-2 9°07 0 106 | | 1661 | 1-76 7 | 7 0 10-5 0 TOA | 19°15 | 2-04 3:5 | 10-6 56-6 871 0 108 2 23 9°39 0 | 162 40°5 9-03 A il | 109 | | | 19-02 2-20 0 14 45 819 a) 110 16,610 aa eee 3°55 766 | 32-49 9°11 4 ® na || | 19:07 | 2°38 0 10 95 5:35 + 4 112 18-19 2°30 0 1 45 8 ar IE | | 17-95 2°38 3-82 7-03 | 32-2 9-76 +5 me | (19°85 318 0 13°73 | 40-2 8-66 +10 In these experiments no lot was fed upon either long fodder alone or con- centrates alone. One lot, No. 18, was fed on roots alone (109 Ibs.), with an inefficiency which is represented by the figure — 6. Six lots were fed on long fodder and roots, always with minus efficiencies, namely :— No. Straw. Hay. Roots. Efficiency. lbs. lbs. Ibs. 28 7 0 126 - 10 Bil 0 it 126 - 9 37 7 0 150 -7 45 10 0 100 - 5 59 u 0 120 - 1 85 0 17 48-4 -1 Witson— Pood- Unit Method in the Fattening of Cattle. 35 Eight lots were fed on long fodder and concentrates, with minus efficiencies in six cases, namely : — No. Straw. Hay. Concentrates. Efficiency. lbs. lbs. lbs. 15 0 1465 6 + 1 19 13°6 0 10-4 - 6 68 0 22 6 - 9 75 9°5 5 13 - 6 19 7 7 9 - 4 80 14:1 0 10 -4 97 iu 55 9-4 - 4 106 0 loa 105 0 Two lots were fed on roots and concentrates, both with minus efficiencies, namely :— No. Roots. Concentrates. Efficiency. lbs. : lbs. 70 131-11 4:6 -2 81 127-03 T712 = 8) It is obvious that such rations are generally inefficient, more especially when they contain large supplies of roots. On running down the figures for the remaining rations, all of which contained the three ingredients, long fodder, roots, and concentrates, it will be seen that efficiency does not vary with the quantities of either long fodder or concentrates, for, among the rations falling to the right side of the upright line, the former run from 16 lbs. down to 8 lbs., and the latter from 10 lbs. down to 3 lbs. Generally the quantity of long fodder is smallest in the most inefficient rations; but the inefficiency is not due to this. The ingredient which clearly affects the efficiency of a ration and, at the same time, keeps down the consumption of long fodder, is the roots. Leaving aside the 6-cwt. and 7-cwt. groups, in which the variations are small, the most inefficient rations contained large quantities of roots, and, as the rations increased in efficiency, the quantities of roots generally declined. If a rough opinion may be hazarvded, which is all that may be done in the meantime, it is that the efficiency of a ration runs risk of being lowered when the roots in it are over 80 pounds. In Fjord’s second experiment with cows, it was found that about 50 pounds of rocts might be added to a usual ration of long fodder and concentrates, and the cows would consume the normal ration as before, plus the roots, and at the same time give two or three pounds more milk a day. Later experimenters found, however, that 36 Scientific Proceedings, Royal Dublin Society. the quantity of roots could not be increased much further without the efficiency of the whole ration being depressed. For this there are three, perhaps more, possible reasons: that the bulk of the whole ration gets beyond the manipulative capacity of the stomach, and that energy is lost, first, in heating up the water in the roots, and, second, in afterwards elimi- nating it from the system. It will also be noticed, in the rations containing less than 80 pounds of roots, that the less efficient rations frequently contain straw alone, or more straw than hay, while the quantities of these two long fodders are as frequently reversed in the more efficient rations. It might be suggested, therefore, that a ration like the following might, with 9-cwt. cattle, be a starting-point from which to discover still more efficient rations :— Oat straw, 4 lbs., 5 ; 1 food-unit. Hay, 10 lbs., : aan 2 Mangels, 70 lbs., ; mca i ¥, Concentrates, : : Sead! 5 The following are the figures, as taken from Professor Nils Hansson’s *Utfodringslara,” published in 1916, by which the various foods have been divided in order to reduce them to their equivalent in barley :— Hay, é é 25! | Linseed cake, . . ‘9 Oat straw, . : 4 | Rape cake, ; : 95 Barley straw, 4 | Gluten feed, . : 95 Wheat straw, 5 Maize or Indian meal, 95 Potatoes, . : 4 | Beans, 1 Carrots, : ‘ 8 Wheat, i Swedes, 9 Barley bran, 1:2 Mangels, 10 Oats, 1:2 Turnips, 3 : 2b) Dried grains, , 1:3 Linseed, 6 Molasses, { 11033 EKarthnut cake, : 8 Undecorticated cotton cake, 1:4 Sesame cake, 8 | Barley, 5 if 1 Decorticated cotton cake, ‘85 1 In a few cases hay is called clover hay. The figure by which pure clover hay should be divided is 2°2, but as grass and clover hay is usually called clover hay, the figure 2°5 has been used in all cases. S. ll Cwls.- I2 | i /bs) (1176-1288 |bs) ('400- Me eiibe 2 3 et. eee i i PIL, SCENT, PROC, IR, IDUIBILIIN GOW, INcSsy WAOIE. DOVE eee 6 Cwls. 7 Cwts. 8 Cwls. 9 Cwrls. 1O Cwts. I! Cwts.- WEIGHT (616-728 /bs) (728- 840 Ibs) (840-952 Ibs) (952-1064 Ibs) (1064 =1176 Ibs) (i176 ~266/bs) aie sey) /bs) NEIOHT wereasellb. 2 3 WD 3 llb. 2 3 Nb 2 lib. 2 3 IIb. 2 3 Ilb 2 BNcREAse ge EH = EEEEEH Ee ! UNITS “CC Se ee See “___) aaa Ch pole alain | | aa EE Seegeees || ie Esha aiid teal | ig aa I8 17 Stop G Leo acta 15 a = es ae PLL eee eee coe | To eee o_O aaa |. mimo lel (|_| Ih SSSa8RE8SE05 eww iil Loe a ANI, WW SCIENTIFIC PROCEEDINGS. VOLUME XVI. 1. A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, S0.D., F.R.S., and T. G. Mason, w.a., so.z, (January, 1920.) 6d. 2. The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. SuytH, B.a., so.B. (Plates I., II.) (February, 1920.) 1s. 8. The Application of the Food-Unit Method to the Fattening of Cattle. By James Wixson, m.a., B.sc. (Plates III., IV.) (February, 1920.) 1s. DUBLIN : PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIBBS. THE SCIENTIFIC PROCKEDINGS ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 4. APRIL, 1920. THE HOLOTHURIOIDEA OF THE COASTS OF IRELAND. ane Smithsonian ira is MAR 5 1927 BY ANNE L. MASSY. 86 ‘onal wusev™ [Authors alone are responsible for all opinions expressed in their Communications. } DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSLER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN LONDON, W.C. 2. 1920. Price One Shilling. Royal Bublin Society. FOUNDED, A.D. 1731. INCORPORATED, 1749, EVENING SCIENTIFIC § MEETINGS. Tue Scientific Meetings of the Society are usually held at 4:15 p.m. on the third Tuesday of every month of the Session (November to June). — Authors desiring to read Papers before the Society ae requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations im a complete form, and ready for transmission to the Editor. Cer 3 aarense Fig MAR’ 5 1924 a II THE HOLOTHURIOIDEA OF THE COASTS OF IRELAND. By ANNE L. MASSY. Read Drcremper 16, 1919. Published Aprin 8, 1920. AxBouT 25 species of this group are distributed round our shores and in the deep water off the west coast, and further research in the latter region will probably add to the number. Among the specimens taken by the naturalists of the Fisheries Branch of the Department of Agriculture and Technical Instruction for Ireland since the publication of Mr. Kemp’s paper! the following are additions to the British and Irish area :— BIO po In these tests the air-pressure, unless otherwise stated, was 2mm. The last three figures show how very much more active the zinc plate was than the chlorophyll, and indicate the presence of a certain amount of ultra-violet light, in spite of all the glass in the path of the beam. They also show the very great increase in the current caused by reducing the pressure, and also the photo-electric fatigue of the zine. The red screen used in one test was a Wratten gelatine colour filter, No. 29. This test, which was not a very satisfactory one, as the electro- meter movements were slightly irregular, shows that the ight which passes through the screen produces little, if any, effect. TESTS WITH A SOURCE RICH IN ULTRA-VIOLET LIGHT. As M‘Clelland and FitzGerald (7) obtained a considerable photo-electric effect with leaf-extracts, using the light from a spark, it seems probable that the small effect shown above is due to the ultra-violet light which passes through the glass. It appeared to be worth while to repeat some of the tests with a source rich in ultra-violet light. Accordingly the half-watt lamp was replaced by an open carbon arc, and a quartz condenser substituted for the glass one. A quartz window was fitted to the testing vessel, and a speculum metal mirror used instead of the silvered glass one. The water beaker was removed, and heating was reduced by using a much less concentrated beam than before. The rest of the apparatus was unaltered. 74 Scientific Proceedings, Royal Dublin Society. It soon became evident that the activity was enormously increased, so that the natural leak of the testing vessel was negligible. In making tests with chlorophyll the plate system was isolated, and a reading taken with the are burning, but shuttered off. The shutter was then raised for about ten seconds, as timed with a stop-watch, and then replaced. The electro- meter spot of light soon became almost stationary again, and the difference between the two readings measured the effect for the time that the light was acting. From this the displacement per minute, and so the current, was found, ‘This varied considerably in different tests, the variations being probably chiefly due to changes in the are. When a clean zinc plate was used the shortest flash of light that could be given sent the electrometer off the scale. A condenser of capacity + microfarad was put in parallel with the electrometer, thus reducing the sensitiveness to about >,55th part of its previous value. An exposure of about half a minute then produced a con- venient deflexion. The results obtained are shown below :— Substance on ‘Test Plate. Current in E.S. U. Zine plate coated with chlorophyll, . a ; 12x 10° Layer of leaf-powder about 1:5 mm. thick, . 0 UG» 55 Clean zine plate, : : ; 5 GOD Ome. It will be seen that the chlorophyll is about 2000 times, and the zine over 10,000 times, as active with this source compared with the previous one. The interposition of the red screen reduced the activity almost, if not quite, to zero. ‘lhe measurement of very small currents could not be made quite so accurately as before, as the motion of the electrometer needle was not quite so regular. The variations were probably due to draughts of slightly ionized air from the are reaching the connecting wires. Discussion or MESULTS. All the evidence points to the fact that the extremely small effect obtained with the “half-watt” lamp was due to the small amount of ultra-violet light present. The last test shows that the latter produces a pronounced effect. The light which passes through the red screen is apparently inactive. Now, this red light is very effective in causing photosynthesis, so apparently photosynthesis is caused by light whose frequency is too low to effect the. expulsion of an electron from the chlorophyll molecule. This conclusion is greatly strengthened when we consider quantitatively the results obtained. The largest photo-electric current obtained with leaf- powder illuminated by the “half-watt” lamp was about 9 x 10°=.s.u. Let us assume that the current was magnified twenty times by collision—an estimate which, in view both of the results obtained with the zine plate and of the value Dixon AnD PooLe—Photosynthesis und the Electronic Vheory. Ta f Y deduced theoretically, is almost certainly too low. The current due to the actual photo-electrons emitted must then have been 4:5 x 10-7. As the electronic charge is 4774 x 10° B.8.U., this corresponds to the emission of 940 elec- trons per second from the illuminated surface. The effective area of this was about 12°5 square centimetres, allowance being made for the part shaded by the grid, so the number of electrons per square centimetre per second works out at 75. This, which is probably an over-estimate, corresponds to the emission of 27 x 10° electrons per square metre per hour. It is most unlikely, in view of the small amount of energy available, that one electron could cause the assimilation of more than one atom of carbon. In fact, the energy quantum for red light of frequency 5 x 10“ is about 3:3 x 10 erg, while the energy required to decompose one molecule of carbon dioxide is about 66 x 10, 1.e. equal to two quanta. This figure refers to gaseous carbon dioxide, and may need modification in the case of dissolved carbon dioxide. But if we assume that each electron causes the assimilation of one atom, we are not likely to be under-estimating the quantity of carbon assimilated. Taking the mass of the hydrogen atom as 1°65 x 10“, this comes out as 5:3 x 107 gram per square metre per hour. This is an utterly negligible figure in comparison with the quantity actually assimilated in strong sunlight, which is of the order of half a gram per square metre of leaf-surface per hour. The leaf-powder used was not treated chemically in any way. The leaves were merely dried at a moderate temperature in an air-oven, and powdered. Willstatter and Stoll (9) have shown that the chlorophyll is not altered chemically by this treatment. As the energy of photo-electrons due to visible light would be so small, it seems unlikely that their range in the leaf-powder would much exceed the distance between two adjacent atoms, so we would only expect those chloro- phyll molecules which are actually on the surface to contribute to the photo-electric current. As, however, the diameter of an atom is of the order of 10° centimetre, it seems incredible that only about one in 10% of the chlorophyll molecules in a layer of powder one millimetre thick is on the surface. Even then we would have to assume that the activity of the living leaf is equivalent to that caused by all the chlorophyll in such a layer. CONCLUSION. The experiments recorded above show that those wave-lengths of light which are effective in photosynthesis are unable, to any appreciable extent, to expel electrons from the leaf-pigment complex, and hence cannot in this way produce ionization, or bring about reactions external to the pigment. 76 Scientific Proceedings, Royal Dublin Society. The action apparently occurs within the molecule of the chlorophyll itself. Possibly electrons are transferred from one atom to another, thus altering the linkage, and hence the chemical nature of the molecule, or its atomic groups. It is also conceivable that electrons might be transferred from molecules of chlorophyll to adjacent molecules, whose presence is necessary to facilitate the escape. Such an action could hardly be classed as photo-electric, using the term in its ordinary sense, and would not be detected by the method employed. If it is shown that chlorophyll acts like other sensitizers, and affects the sensitiveness of the photographic plate even when it is removed before development [Abney (1)], such a transference of electrons from the chloro- phyll to one of the other constituents of the photographic film seems a necessary assumption. The constant association of the four pigments Chlorophyll a@ (C,;H;,0;N.Mg), Chlorophyll 0 (C;;H70.NyMg), Xanthophyll (CwH502), and Carotin (CyoHss) in sensibly constant proportions, in the chloroplasts of green plants, as shown by Willstatter and Stoll(9), and the chemical relationship of these substances, elucidated by the same investigators, pointedly suggest a trans- ference of electrons between the components of the pigment complex in the chloroplast securing the ultimate reduction of the carbon dioxide and the liberation of oxygen. This transference must be effected directly or indi- rectly by the light. Any evidence as to the nature of this process and the interdependence of these substances is so desirable that it is hoped shortly to repeat our observations on pure samples of the individual pigments, instead of using the crude leaf-extract or leaf-powder. It seems, however, rather unlikely that a positive result will be obtained with the separate pig- ments, since the result with the complex was so decisively negative. For the present, then, it appears we must assume that the atomic groups of the leaf-pigment enter into the reaction of photosynthesis, and participate in the combinations and decompositions which ultimately lead to the forma- tion of carbohydrates and the evolution of oxygen. The experimental evidence seems to debar us from regarding the pigments in the chloroplasts as mechanical contrivances for effecting the ionization of molecules external to themselves. ‘Thus the chemical theories of photosynthesis, such as Hoppe Seyler’s and Willstatter’s, which assume that the chlorophyll itself enters into the reactions, are to be preferred to those suggestions, like Siegfried’s, which suppose that the reaction is accomplished externally to the chlorophyll by means of the energy absorbed and transformed by the latter. Or Dixon and Poote—Photosynthesis and the Electronic Theory. 77 REFERENCES. . ABNEY, W. DE W.—The Theoretical Aspect of Orthochromatic Photo- graphy. Journal of the Camera Club, March, 1888, vol. ii, No. 17. . ALLEN, H. $.—Photo-Electricity. London, 1915. . BecQuerEL, KpmM.—Action des rayons différemment réfrangibles sur Viodure et le bromure d’argent; influence des matieres colorantes. C.R., 1874, p. 188. . ENGELMANN, T. W.—Untersuchungen tiber die quantitativen Beziehungen g | 8 zwischen Absorption des Lichtes und Assimilation in Pflanzenzellen, Bot. Zeitung, 1884. . JoLy, J.—The Latent Image. Presidential Address to the Photographic Convention of the United Kingdom, July, 1905. Nature, vol. lxxii, p. 308. . Kyirp, H., and Minper, F.-—Ueber den Einfluss verschiedenfarbigen Lichtes auf die Kohlensaureassimilation. Zeitsch. f. Bot., 1909. . M‘CLELLAND, J. A., and F 11ZGERALD, R.—Photo- Electric Discharge from Leaves. Proc. Roy. Irish Acad., vol. xxxiii, Sect. A, No. 1, 1916. . TOWNSEND, J. S.—The Theory of the Lonization of Gases by Collision. London, 1910. . WILLSTATTER, R., and Stott, A.—Untersuchungen tber Chlorophyll. Berlin, 1913, te RS Vi. NOTE ON THE DECAY OF MAGNETISM IN BAR MAGNETS. By WILLIAM BROWN, B3Sc., Professor of Applied Physics, Royal College of Science for Ireland, Dublin. Read Fresruary 24. Published Marcu 12, 1920. SECTION 1. At a meeting of this Society on December 21st, 1909, the writer read a paper on “‘ Permanent Steel Magnets.” Section 1 of that paper dealt with the dimension-ratios of the magnets which were made of the same kind of steel, and Section 2 with the moments of magnets made of steels of various chemical composition. These magnets in Section 2 only are dealt with in the present communication; and at that time (about ten years ago) they were tested for magnetic moment per gramme at intervals for six months, and . the percentage loss in magnetic moment for a period of six months measured. The magnets were then put away from all disturbing influences, and have remained in a vertical position for about ten years, and have now,been re- tested for magnetic moment, with the results given in this note. The details of the tests and the heat treatment of the magnets are given on pp. 816-320 of the paper referred to, and the results now obtained are given and compared with the previous results, that is, the percentage loss in the magnetic moment per gramme for each magnet for a period of about ten years. The magnets were 8 cms. or 10 cms. long and 03cm. diameter, giving dimension-ratios of 27 and 33, as shown in Table I. In Table I the numbers 1 and 6 are left out, as these referred to magnets made of practically pure iron, and were used in the previous paper merely as standards of reference. 1Scient. Proc. Roy. Dublin Soc., February, 1910, vol. xii, pp. 312-820. on SCIENTIFIC ‘PROCEEDINGS. VOLUME XVI. . A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, SC.D., F-R.S., and T. G. Mason, m.a., so.B. (January, 1920.) 64d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Lous B. Suyru, B.a., scp. (Plates I., II.) (February, 1920.) 1s. . The Application of the Food-Unit Method to the Fattening of Cattle. By James Witson, wa. B.sc. (Plates III., IV.) (February, 1920.) 1s. . The Holothurioidea of the Coasts of Ireland. By Anne L. Massy. (March, 1920.) 1s. . Photosynthesis and the Electronic Theory. By Henry H. Dixon, sc.p., F.B.5., and Horace H. Poors, sc.p. (March, 1920.) 1s. —————— DUBLIN : PRINTED AT THE UNIVERGITY PRESS BY PONSONBY AND GIKES. y, sith ay A ( sc Rear PT edie aie yA aU a Tesi We EN THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 6. MARCH, 1920. NOTE ON THE DECAY OF MAGNETISM IN BAR MAGNETS. is Sauthsstlan ing ry 2 BG MAR 5 1921 < Rerveens bik pepe WILLIAM BROWN, B.Sc, i Seek PROFESSOR OF APPLIED PHYSICS, ROYAL COLLEGE OF SCIENCE FOR IRELAND, DUBLIN. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL, DUBLIN SOCIETY, LEINSLER HOUS", DUBLIN. WILLIAMS AND NORGATEH, 14, HENRIETTA STREET, COVENT GARDEN LONDON, W.C. 2. 1920. Price Sixpence. Roval Bublin Society. oe ~~ FOUNDED, A.D. 1731. INCORPORATED, 1749. SOE EVENING SCIENTIFIC MEKTINGS. Tue Scientific Meetings of the Society are usually held at 4:15 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be seat down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Editor. een Brown—Wote on the Decay of Magnetism in Bar Magnets. 79 a Bei i y ir ie = TABLE I. ME —esieaat a ae: Chemical Composition. ; ; No. 7 7 7 Be @ | iim | Si | RR | Ww I Ge A B G | wD ir 2 | 1-09 | 0-32 | 0-06 | 220-420-208 ote ys a4 3 | 0°75 | 1:00 » | 81:2 | 68 | 26-2 | 16 4 | 0:50 | 1:00 1:00 »» | 27-5 | 10-1 | 26-4 4 5 | 1:30 | 3:09 8:92 | ,, | 26-4 | 14-0 | 24-9 6 7 | 0-22 | 0-18 | 0-44 | 33 | 25-0 | 11-0 | 22-2 | 11-2 8 | 0:26 | 0-18 | 0:33 | 0-58 » (205 | 7:4 | 20°3 | 26 9 0:28 | 0-28 | 3°5 »» | 22:4 | 6:6 | 21-1 | 5:8 10 | 0-76 | 0:28 15°5 » | 88:5 | 2-7 | 89:5 | 15:5 Il | 0-40 | 2-25 | 3:25 » | 29:6 | 11-6 | 27-0 | 8:8 12 | 0-25 | 2:00] 0-75 | ,, | 30:0 | 14:0 | 26-9 | 10:3 13 | 0-48 B25 | py || B46 | BG || awe || Daw 14 1-09 | 9:50 | ,, || 27°5 | 10-4 | 93-6 | 14-2 15 “0-25 2°75 PB |) op | 15-0 | 15-2 | 143 | 4-7 | 16 | 0-31 | 2°50 1-75 foe 30°2 | 9:6 | 25-1 | 16:8 A = Magnetic moment per gramme in December, 1909, or six months after being magnetized. B = Percentage loss in magnetic moment six months after being magnetized. C = Magnetic moment per gramme in January, 1920. D = Percentage loss in magnetic moment for a period of about ten years, or from December, 1909, till January, 1920. It will be seen from the table that No. 2 magnet has lost very little of its magnetism after the first six months, and seems to have settled down to a practically permanent state. The fairly high carbon and manganese makes this a good permanent magnet, though its actual moment is not very high, Comparing Nos. 3 and 4, which have the same amount of manganese and fairly high carbon, the latter, having also 1 per cent. of nickel, has four times the retentivity of the former, though the percentage loss in the six months after being magnetized is nearly double of that in No. 3. Taking Nos. 2 and 14, which have the same amount of carbon, and the latter a high percentage of chromium, the presence of the chromium has reduced the retentivity ten times, which confirms a result given by the writer in another paper, namely, that more than about 2°5 per cent. of chromium in steel decreases its value as a magnet.? 1Scient. Proc. Roy. Dub. Soc., April, 1910, vol. xii, pp. 349-353. SCIENT. PROC. R.D.S., VOL. XVI, NO. V1. K 80 Scientifie Proceedings, Royal Dublin Society. Comparing Nos. 5 and 14, which have high carbon and a high percentage of chromium, the former’ having also over 3 per cent. of manganese, we see that the percentage loss in No. 4 is less than half that in No. 14, due no doubt to the presence of the manganese. Of the four tungsten steels, Nos. 9 to 12, inclusive, the first seems the best from the point of view of retentivity ; and it is known that a steel containing about 7:5 per cent. of tungsten makes the most permanent magnet; and No. 9, though the actual magnetic moment is low, is a fairly permanent magnet. No. 15 is a good magnet, though its moment is low compared with No. 16, which has 1 per cent. more of chromium ; the retentivity of the former is over three times that of the latter. In the paper above referred to (April, 1910), in Table LV, page 318, it was shown that the magnetic moments all increased slightly one month after they had had the final five hours’ annealing by steam, so that if we were to calculate the total loss from that time, July, 1909, till now, January, 1920, the magnets, with the exception of No. 8, arrange themselves into three natural groups, as shown in Table Il, where Column / gives the total percentage loss in ten and a half years. TABLE II. No. E 2 P 26-4 | 3 - 25:8 | 4 27-0 5 24:8 7 | 25:0 9 17-2 10 | Oey 11 21°3 12 lope 13 32:0 14 28-5 15 23-5 16 28-9 From 2-7, manganese steels, the loss is about 25 per cent.; from 9-11, tungsten steels, the loss is about 20 per cent. ; and from 12-16, chrome steels, the mean loss is over 25 per cent. 1Scient. Trans. Roy. Dub. Soc., vol. vii, Plate vi, January, 1900. Brown—WNote on the Decay of Muynetism in Bar Magnets. 81 SECTION 2. In February, 1910, the writer read a paper before this Society on “ Chrome Steel Permanent Magnets,’ where results were given for seven magnets of dimension-ratio 33, the magnets containing different percentages of chromium.' They were then tested for magnetic moment per gramme when in the condition (1) glass-hard, (2) annealed, and aiso tested for the percentage loss in the moment due to percussion. The details of the tests and the heat- treatment of the magnets are given in the paper referred to. The magnets in the annealed state were put carefully away after the tests, and have remained undisturbed from that date to this, or about ten years, when they have again been tested for the decay of magnetism, with the results here given. In Table III the results in column A are the values of the moments when the magnets were stored away ten years ago, that is, they ave the values given in the paper referred to above, minus the percentage loss due to percussion in each case. 2 TasLe III. Pe Chemical Composition. No, | a | C Si Win | Cr |) Wwe) Gar | B C D Le | 1 | 0-88 | 0:19 | 0-28 | 1-75 | 404 | 36-8 | 8-9 | 1:9 2 | 0:86 | 1:96 | 0-40 | 1-96 | 50:3 | 48:6 | 13:3 | 0-5 3 | 0-76 | 1-02 | 0-29 | 2-11 i So S) rs be rw =) oS tw wo = i) or So eo: D a ies} ee 2 ay oS iw) ~ ie) 5 | 0-85 | 0-31 | 0-50 | 5-79 183 | 36:5 | 35-0 | 41] 7-0 | | 6 | 1:36 | 0-75 | 2-60 | 9-29 427 | 40-1 | 6-3 || 1-3) | TSO MICO RL O-ON eS orsi elle 7 372 A = Magnetic moment per gramme in January, 1910. B= ” ” ” ” 1920. C = Percentage loss in the moment for a period of about ten years. D = Total percentage loss due to percussion from previous paper, p. 351. It has been stated by Mme. Curie that small quantities of silicon have little or no effect on the magnetic properties of a magnet.” 1Scient. Proc. Roy. Dub. Soc., April, 1910, vol. xii, pp. 349-353. * Bull. de la Société d’ Encouragement, pp. 36-76. 82 Serentific Proceedings, Royal Dublin Society. By comparing Nos. 1 and 2 in the table, it looks as if the high silicon in No. 2 had decreased its retentivity; the same appears to apply to No. 4. Yaking Nos. 1 and 7, with nearly the same amount of chromium, the smaller carbon and larger manganese in No. 7 seem to counteract each other ; also, the equal quantities of tungsten and copper neutralize to make the retentivity of the magnet poor. The magnets 3,5, and 6 are good, and confirm an observation made by the writer, and given in the paper referred to above, that the best proportion of chromium for a magnet was about 2°5 per cent.; the results for No. 3 magnet more nearly illustrate the fact. The actual moment per gramme of No. 5 is low, though its retentivity is high, due possibly to the presence of copper neutralizing the chromium, as well as its fairly high carbon. On the whole it is evident, as far as permanence is concerned, that the best magnets would be Nos. 3, 5, and 6. Column D in Table III gives the total percentage loss in the magnetic moment per gramme due to the magnets being allowed to fall, end on, from the height of one metre four times on to a block of glass; in every case with the true north or south pointing pole downwards. 3. Or SCIENTIFIC ‘PROCEEDINGS. VOLUME XVI. A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, SO.D., F.R.s., and I. G. Mason, m.a., scp. (January, 1920.) 6d. The Carboniferous Coast-Section at Malahide, Co. Dublin. By Lous B. Suyrn, B.a., sc.p. (Plates 1, I.) (February, 1920.) 1s. The Application of the Food-Unit Method to the Fattening of Cattle. By James Writson, m.a., Bsc. (Plates III., 1V.) (February, 1920.) 1s. The Holothurioidea of the Coasts of Ireland. By Anne L. Massy. (March, 1920.) 1s. Photosynthesis and the Electronic Theory. By Henry H. Drxon, sc.p., F.2.s., and Horace H. Pooun, sc.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Witi1am Brown, B.sc. (March, 1920.) 6d. DUBLIN : PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIBBS. ie aa art Weed EN ale ta pr) toe PE area ' Te Sed el } i ‘ ria e i ial f ‘ t THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 7. APRIL, 1920. ON THE INHIBITION OF INVERTASE IN THE SAP OF GALANTHUS NIV ALIS. BY Eunsenian ary > T. G. MASON, M.A., Sc.B. MAR 5 1921 y Sttene) nouse™ [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSLIER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN LONDON, W.C. 2. 1920. Price One Shilling. Koval Bublin Doctety. FOUNDED, A.D. 1731. INCORPORATED, 1749. EVENING SCIENTIFIC MEETINGS. Tax Scientific Meetings of the Society are usually held at 4:15 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be get down for reading until examined and approved by the Science ~ Comunittee. The copyright of Papers read becomes the property of the Society, and such ag are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Bditor. 1 : : by seid WAUE. We ON THE INHIBITION OF INVERTASE IN THE SAP OF GALANTHUS NIV ALIS. By T. G. MASON, M.A., Sc.B. Read Frenruary 24. Published Aprit 12, 1920. In view of the widespread occurrence of invertase in the leaves of plants (8), it seemed probable that an approximate estimate of the sucrose-content of the sap could be obtained by observing the increase in depression of the freezing- point after storage at a suitable temperature. Asa 1 per cent. solution of sucrose has a depression of 0:054° C., whereas after its hydrolysis to invert sugar it has approximately double this de- pression,a method is provided of recording the amount of sucrose inverted (7). The experiments were carried ovt on the Snowdrop (Galanthus nivalis), and, as starch and inulin are absent from its leaves (10), the difficulties which might be introduced by the production from these sources of maltose, dextrose, and levulose need not be considered. In the earlier experiments made on this matter it was found that other factors, which tended to limit the activity of the enzyme, were present; consequently it was considered that an investigation of their nature might prove of interest. The freezing-point determinations were made by means of Dixon’s thermo-electric method (4 and 5). The saps for the freezing-point determinations shown in Table I were pressed from tissues rendered permeable by exposure to intense cold. To do this the leaves were rolled and packed in test-tubes, which after sealing were immersed for two hours in a freezing mixture of salt and ice (-16° C.). The white bases of the leaves were rejected, and the sap, as pressed from the leaves, was collected in test-tubes fitted with an ice-jacket to prevent inversion. Sap was also pressed from leaves which had not received any treatment. SCIENT. PROC. R.D.S., VOL, XVI, NO. VII. L 84 Scientific Proceedings, Royal Dublin Society. As soon as the freezing-point determinations were made the saps were stored for approximately twenty-four hours at 29° in a thermostat. The increases in depression are shown in Table lin the last column [A, = Aj]. In the absence of other changes affecting the depression, this increase should afford a measure of the sucrose inverted. TABLE I. . . | Description of Sample. Treatment. | Ay Ay 45 — A, | 2 ~ q= ° a ro q ° Leaves collected 9.30 a.m., Witireated: eels oe Ova March 21st. Snow in}| night and on previous aay Frozen. 0-888° 0-943° 0-055° Untreated. 0°544° 0°655° O-111° Leayes collected 9.15 A.m., March 24th. Dull and _ —-— ——— ——|—_—_ —. wet previous day. Frozen. 0°834° 0:940° 0:106° That sap pressed from tissues which have been exposed to intense cold has a much greater depression of the freezing-point than that from untreated material has been already demonstrated by Dixon and Atkins (5), and does not call for further comment. On the other hand, it will be noted that the increase in depression after storage is as great in the sap ipueesiee from untreated leaves as in that from the frozen material. Assuming the increase in depression is proportional to the sucrose- content of the sap, this might be due to the greater permeability of the protoplasm to sucrose than the other solutes; or else that in the pressing of the untreated leaves the rupture of the cells has set free the contents of the larger vacuoles, where possibly the concentration of sucrose is greater than in the protoplasm. In Table II the results of a somewhat similar experiment are shown. Sap pressed from leaves, both untreated and frozen, was again employed, but in this case, with a view to determining if the acidity of the sap was responsible (by hydrolysing the sucrose) for any of the rise in depression on storage, a sample of each was immersed with its containing test-tube for one minute in boiling water ; a reflux condenser was fitted to each test-tube. Mason— Inhibition of Invertase in the Sap of Galanthus nivalis. 85 TABLE IT, Description of Sample. | Treatment. Ay | Ay Ao - Ay Untreated. 0-552° | 0°610° 0:058° Do., boiled. 0:553°, | | 05532 ae Leaves collected 9.30 a.., March 28th. Previous< | ————j/—-——_——____| —_ ji — day bright and showery. Frozen. 0-904° 0-924° 0:028° Do., boiled. 0:891° | 0:888° — | The acid of the sap has evidently not been responsible for the rise in depression observed in the samples which were not boiled. The increase in depression in the sap pressed from untreated leaves is in this case much greater than that from the frozen. It is clear from a survey of Tables I and II that, if the increase in depressions of the sap pressed from the frozen material were an index to the sucrose-content, very great fluctuations must have taken place in a comparatively short time. It is instructive to compare the increases in depression recorded in Tables I and Il with those found by Dixon and Atkins (5) (Table III) on storing sap pressed from untreated leaves of Syringa vulgaris, ‘The material for these experiments was collected towards the close of October; the saps were pressed from unfrozen leaves, and were filtered and stored at room- temperature. Tasie ITT. No. of p escripti {f Sample. A, — Experiment. Description of Sample Ay A, A A 1 j Gathered 2.30 P.m., area 975° 999° ‘ 66 and 67 {sunny morning. 1:218 1-220 0:005° 71 and 73 — 1:424° 1:540° 0°116° | Here similar great differences in the change in depression on storage are noticeable. As it was considered possible that the fluctuations observed in the increase in depression of the freezing-points on storage might not be L2 86 Scientifie Proceedings, Royal Dublin Society. unconnected with the presence of micro-organisms on the surface of the material, some leaves were immersed for six days in a little freshly boiled water. The water was then decanted off, and to 220 c.c. were added 7-7 ems. of sucrose. ‘he solution was divided into two parts, one of which was boiled. Both were saturated with toluene. After nine days’ incubation at 29° C., there was no evidence of inversion in either sample; thus the enzymes set free from micro-organisms do not appear to have been responsible for any of the rise in depression recorded in Tables I and II. In the next experiment the leaves of a clump growing in the open were divided into three parts. The leaves were gathered at 10 a.m. on April 30th. The first sample was pressed immediately without any treatment, the second was exposed to toluene vapour for five hours in a sealed chamber, while the third, enclosed in a test tube, was immersed in a salt-ice freezing mixture (-16°) for two hours. About 60 c.c. of sap were collected from each sample, and filtered and stored overnight at 0°. The depressions of freezing-point and the electrical conductivities are shown in Table IV. The conductivity observations were made at 0°, and are liable to a correction allowing for the viscosity of the saps. As the saps were saturated with toluene, 0-02° (the depression of the freezing-point of water saturated with toluene) has been subtracted from the observed depressions. Toluene was not added to the saps on which the conductivity measurements were made. TABLE LV, Treatment. A G 34 lO? Frozen, 0°818° 630 oe ec ee pes ii) ats 2 pee ee. That exposure to toluene vapour has not rendered the protoplasm completely permeable is indicated by the greater depression of the sap pressed from the frozen leaves; the difference, however, may be due toa partial destructon of sugar owing to the stimulation of respiration by the aneesthetic. On immersing the three samples of sap for one minute with their Mason— Inhibition of Invertase in the Sap of Galanthus nivalis. 87 containing test-tubes in boiling water, the variation in the amount of protein precipitated was very marked. The amount in the sap pressed from the untreated leaves was almost double that of the frozen, while that from the toluene material was very slight. On storage at room-temperature differences in the amount of precipitated colloids of the same order were observed in the untreated and frozen samples, though none could be detected in that pressed from the toluened leaves. With ferric chloride a dark green colouration was produced in all samples, but, as gelatine was not precipitated, it is improbable that a true tannin was present. The changes in the depression of the freezing-point during a storage of seventy-two hours at 29° are shown in Table V. The term “sap + sucrose” in the table refers to samples in which 25 c.c. of each sap were added to two grams of sucrose. Both the saps and the sap + sucrose solutions were stored in small flasks, previously sterilized, and fitted with rubber corks. A little toluene was added before each observation in order to ensure that the samples were completely saturated. To make an observation 5 c¢.c. were withdrawn from each flask, and stored at O° till all the observations were complete. The samples used for the determinations of the freezing-points were then returned to their respective flasks. TABLE V. | No. of Hours. Treatment. Description of Sample. | een a near 44s | An cea oe a We ay Piece San Sap, . 5 ° " ° 9 | 0:838° 0:872° 0:866° | 0°858° Frozen, es = = |e eno | Sap + Sucrose,. . . . | 1:318° 1°358° 1:352° | 1-363° ( Sap, 0:730° 0°735° 0°739° 0:741° Toluene, _———— ——— ee =— SS ————— \ Sap + Sucrose, 1:209° 1°207° 1:219° 1:216° Wie Sause Mamet) Maia) nel oteds? 0-576° 0°606° 0-582° Untreated, ; |—— =— pee eS oa bee sear = ( Sap + Sucrose, ‘ 1:015° 1-0538° 1:108° ality? 88 Scientific Proceedings, Royal Dublin Society. Figures 1, 2, and 3 show these changes in the depression of the freezing- point graphically represented. The depressions on the left and right of the figs are those of the sap + sucrose solutions and saps respectively. Fic. 1. The increase in the depression of the freezing-point of both frozen and untreated samples is, doubtless, due to the inversion of sucrose; the interpretation of the subsequent changes in the depression is not clear. The presence of factors inhibiting the hydrolysis of sucrose is, however, indicated. In the toluened samples this is very marked, but, as there is very little indication of any hydrolysis, it may be that exposure to toluene vapour has ray 0-810 0-790 0:770 fake} 40 IME IN HOURS Fie. 2. destroyed the enzyme. That this is unlikely is indicated by the fact that toluene has no toxic effect on the invertase of yeast. It may not be without significance that the amount of sucrose inverted in the three samples is roughly proportional to the colloids precipitated on storage. The clumping of the colloids of the sap after its extraction from the leaf is very striking, and has been observed in the extracted sap of a number of plants. - It was considered possible that the addition of a little saponin to the Mason—Inhibition of Invertase in the Sap of Galanthus nivalis. 89 toluened sap + sucrose solution might, by lowering the surface-tension, free the enzyme if present from any adsorbent colloid. No increase in hydrolysis could, however, be detected in the samples treated with saponin. This need not negative the adsorption view of the inhibition, as not only is there evidence that a saponin is normally present in the sap, but the adsorption might well be specific. The absence of a precipitation of colloids on storage from the toluened samples suggests that the precipitation may have taken place in the cell- vacuoles during the exposure of the leaves to toluene vapour, and that possibly the enzyme was removed with the colloids, and therefore was not extracted with the sap. yas 1105 FROM eo | 1-085 ey 40 50 TIME IN HOURS Fie. 3. As the presence of inhibiting factors seemed to be a possible explanation of the observed depressions of the freezing-point, the saps (from the frozen and untreated leaves) were placed aside in the thermostat for a further forty- eight hours in order to see whether or not by then adding invertase the hydrolysis of sucrose was complete. On withdrawing these saps (after 120 hours’ incubation at 29°), they were boiled for one minute, as described in the second experiment. To 5 gm. of the boiled and filtered saps was added 0:33 gm. of invertase solution, prepared by Davis’ method (3). Before the addition of the invertase the saps were cooled to — 2° with a view to checking inversion till the freezing- points had been determined. When the observations were completed, the samples were stored for approximately twenty-four hours at 29°, The results are shown in Table VI. 90 Scientific Proceedings, Royal Dublin Society. For comparison, the freezing-points of the saps which had been boiled on the day of extraction, and to which 0:33 gm. of invertase had been added, are also shown. The increase in depression, corrected for dilution by the invertase added, is given in the last column. TABLE VI, 45, -4 Tr BO) on art : ui 24> “0 Treatment. | Sap + 0°33 gm. Inyertase solution. Ao 404 Aon EYy ll carrented for Dilution Sap boiled on day of extraction, | 0°887° 0°958° 0:071° 0:075° Frozen, / — ee EN TS ee Ml eh os Sap boiled after 120 hours’ storage,| 0°927° 0-964° 0:037° 0-039° Sap boiled on day of extraction, 0°617° 0°661° 0-044° 0:047° Untreated, —— = ne ole Sap boiled after 120 hours’ storage,| 0°649° 0°6638° 0-014° 0:015° It will be observed that the inversion of the sucrose in the sap was not complete after 120 hours’ storage at 29°. It may be objected that a destruction of hexoses by respiratory enzymes would afford an explanation of the apparent inhibition of hydrolysis recorded in figures 1, 2, and 3. It must be admitted that such a view cannot be positively set aside; but the striking similarities in the freezing-points recorded under A,, of Table VI for the saps (to each of which °35 gm. of invertase solution was added) which were boiled on the day of extraction, and in which the activity of all the enzymes was destroyed, to those in which the enzymes were not killed until a period of 120 hours had elapsed, do not point to such a destruction of sugar. If this destruction of hexoses by respiratory enzymes had taken place during storage in the saps which were not boiled on the day of extraction, it would presumably result in their showing a smaller depression of the freezing-point than that shown by the samples in which the enzymes were destroyed by boiling on the day the sap was extracted from the leaves. That this is not so, as the figures recorded under Ay in Table VI indicate, does not exclude the possibility of a destruction of hexoses, but renders it very improbable. The interpretation of the decrease in depression shown in figs. 1 and 3 of the frozen and untreated saps is not clear; it is possibly due to a condensa- tion of the hexoses of the sap to form sucrose. Robertson, Irvine, and Dobson (11) observed an association of this nature while working with sludges prepared from the leaves of Beta vulgaris. This condensation would, of course, Mason—JInhibition of Invertase in the Sap of Galanthus nivalis. 91 be masked by the greater velocity of hydrolysis in the initial stages of storaye; but, with the introduction of factors limiting hydrolysis, a diminution in the depression of the freezing-point would be brought about. In Table VII are shown the results obtained with a sample of sap pressed from leaves which had been exposed to toluene vapour for four hours. ‘l’o 25 ¢.c. of this sap approximately 2 grams of invert sugar and sucrose respectively were dissolved. These leaves were collected at 9.30 a.m, on April 22nd. TABLE VII. No. of Hours. Description of Sample. 49 Agi avip Aik Aogr Asin Agni Hexose+Sap,| 1:471° 1°469° 1°469° 1:478° 1°478° 1°485° 1°485° Sucrose +Sap,} 1:114° erties ile? 1:128° 1-126" 1:136° 1:140° Sap, . - | 0°651° 0:651° 0.651° 0°652° 0-662° 0°662° 0°662° This experiment does not call for comment, as it confirms the results of the last experiment. There is no condensation of hexoses in the hexose-sap solution. Bonns (2), in his work on etherization and enzyme activity, found that factors inhibiting the activity of invertase were introduced by exposure to ether vapour. Possibly the slight decrease in the depression of freezing-point brought to light in these experiments may be attributed to the oxidation of hexoses. In Table VIII are given the results of an experiment designed to determine if the sap pressed from material treated with toluene exerted an inhibiting influence on the invertase of yeast. To 50 c.c. of the sap 2 grams of sucrose were added. The invertase solution was diluted with an equal volume of water, and evidently had much the same depression as the sucrose-sap solution, since its addition has produced only a very slight reduction in the depression. For comparison some of the sap was boiled before the addition of the invertase. In the last column is shown the increase in depression after storage at 29° for twenty-four hours. 92 Scientific Proceedings, Royal Dublin Society. TABLE VIII. ) — Description of Sample. | Ay Ay | Ay — Ay | | | Sap not boiled, 5 gm. No Inyertase, | 0°915° | 0-916° | 0-001° | CORN aes. t| | Sap not boiled, 5 gm. +1 drop Invertase, . | 0-911° 1:158° he 0:247° | Sap not boiled, 5 gm. + 3 drops Invertase, . | 0-910° 1-164° | 0*254° Sap boiled, 5 gm. + 1 drop Invertase, . | 0-913° 1:140° | 0°227° Sap boiled, 5 gm. + 3 drops Invertase, : 0-910° 1+155° “+ 1022452252 That the activity of the invertase added has not been limited is clear ; this, though demonstrating that the negative results obtained with the sap are not associated with the production of substances of a toxic nature, yet does not indicate whether adsorption of the leaf enzyme, either before or after extraction of the sap, or its destruction is the cause. As the factors inhibiting the activity of certain peroxidases have been removed by dialysing the sap (1), it was thought that similar treatment might possibly restore the activity of the enzyme in the sap pressed from toluened material. To 50 c.c. of the dialysed sap were added 2 gms. of sucrose; an equal weight of sucrose was also added to 25 ¢.c. of the same sap, which, however, had been stored for three days while dialysis was in progress in the first sample. The results are shown in Table IX; the results obtained with a sample of the sucrose-sap (not dialysed) diluted with an equal volume of water are also given. TaBie IX, Description of Sample. Ay oa Gy hs 4, - Ay Dialysed Sap + Sucrose, Se Oi nee 0-251° 0-259° - 0:008° Sap + Sucrose, - $ : . : 3 0°962° 0:977° 0-015° Sap + Sucrose diluted with an equal vol. of 0-474° 0:484° 0:010° water, The dialysis has evidently not been responsible for any increase. in hydrolysis ; it is possible that both enzyme and inhibitor diffused away. Mason—ZInhibition of Invertase in the Sap of Galanthus nivalis. 98 In Table X are shown the changes in the depression of the freezing-point observed on incubating some sucrose and hexose solutions to which were added samples of sap pressed from leaves gathered at 4 p.m. and midnight respectively of April 7th. The tissues were rendered permeable by the freezing treatment; the sap was not filtered, but was stored in ice till the following morning. ‘lo 25 c.c. of the sap pressed from each sample were added 75c.c. of a 6 per cent. sucrose and hexose (invert sugar) solution respectively. The solutions were kept saturated with toluene. TABLE X. ANo. of Hours. Description of Sample. ZY) | Se Se | Sip, ay PHS Ip 0°533° Day | Sap + Sucrose solution, | 0°447° | 0°455° | 0°454° | 0:462° | 0:476° | 0:496° | Day Sap + Hexose solution, | 0°654° |.0°650° | 0:650° | 0:647° | 0°662° | 0-679° | 0:690° 0-689° Night | Sap + Sucrose solution, | 0°397° | 0-417° | 0°411° | 0°410° | 0°425° | 0-441° | 0-466° | 0:479° | se Night Sap + Hexose solution, 0°622° | 0°622° | 0-608° | 0°611° | 0°638° | 0:646° | 0°655° | 0-655° | | The depressions on the left and right of figures 4 and 5 refer to the hexose and sucrose samples respectively. A Sti male ; : A 0:705 0-497 0-685 0-477 0-665 10:4.57 0645 0-437 0-625 0-417 Q605) 0-397 10) 10 20 30 60 70 80 90 40 50 TIME IN HOURS Fig. 4. 94 Scientific Proceedings, Royal Dublin Society. There is very little indication of any inhibition in the sucrose day sap, whereas there is a slight initial fall in the depression im that of the hexose TIME IN HOURS Fie. 5. sample, due possibly to a condensation of the hexoses. The subsequent increase in the depressions of both samples, we are justified in assuming, a 0-508) 0-488 a) Y 1 =a | o nw Jbl del [el | | Helge a 03487 _| S (0) 10 20 0 40 50 60 TIME IN HOURS Fie. 6. is due to the inversion of sucrose; that in the hexose sample is due to the inversion of the sucrose of the sap. In both the night samples a condensation or destruction of hexoses is Mason—JInhibition of Invertase in the Sap of Galanthus nivalis. 95 indicated ; in the sucrose sample hydrolysis of sucrose is possibly masked by this condensation. In the next experiment sap was pressed from leaves gathered at 9.30 a.m., March 25th. Saps pressed from both frozen and untreated leaves were used; 1c. and 0:5c.c. respectively of each sample were added to 6 cc. of a 6 per cent. sucrose solution. Toluene was not used. TABLE XI. Treatment. Description of Sample. Ao Aus Ay, Aso 1 c.c. of Sap + 5 c.c. Sucrose solution, : 5 Q > 0°433° 0:434° 0°437° 0-491° Frozen, —————————————————— ae ene eae eee 0-5 c.c. of Sap + 5 c.c. Sucrose SUMO, 5 : 0°388° 0-390° 0°395° 0°415° 1 c.c. of Sap + 5 c.c. Sucrose solution, ° 5 0°368° 0°395° 0°386° 0°415° Untreated, ¢_ |__| ——_——<—_ —_—__— —— 0°5 c.c. of Sap + 5 c.c. Sucrose solution, R 5 5 . 0°348° 0°357° 0°358° 0°376° The changes in depression of the sucrose sap solutions of the untreated leaves are somewhat similar to those of the night sucrose samples of fig. 4. In the frozen samples, however, the initial steep rise in depression is absent. DISCUSSION. The results obtained, though difficult to interpret, indicate that factors inhibiting the activity of the enzyme responsible for the hydrolysis of sucrose may be present in the sap. It has been pointed out that changes of a nature not yet understood (possibly, as Dr. Atkins has suggested to the writer, in the concentration of hydrogen ions) occur in the sap after its extraction from the leaf, which lead to a clumping of the colloids. It is suggested that the enzyme may be inactivated by adsorption on the coagulated colloids. Thus the inactivity shown by the toluened sap would be due to the removal of the invertase by the precipitation of the colloids before the extraction of the sap from the leaf; the absence of a colloid coagulum on storage supports this view. It is unnecessary to consider the factors which might lead to this 96 Scientific Proceedings, Royal Dublin Society. clumping of the colloids during the exposure of the leaves to the anesthetic. It is not improbable, however, that it is associated with an increase in the concentration of the hydrogen ion as a result of the stimulation of respira- tion (12). In a similar way, the greater activity shown by the untreated sap (Table V) would be due to a partial precipitation of the colloids during the freezing of the leaves; as ice separates out in the cell vacuoles, the hydrogen ion becomes more concentrated, and will, as Harvey (8) has shown, bring about a precipitation of some of the colloids. The temperature to which the leaves are exposed must clearly be an important factor, according to this view, in determining the activity of the enzyme in the sap after its extraction. The permanence of the inhibition would on this assumption depend on whether the colloid (anti-enzyme) on which the enzyme was removed was reversibly or irreversibly precipitated. It may be that the activity of the enzyme in the living cell is regulated by some such means. It has been demonstrated by numerous investigators that sucrose accumulates in the assimilating cell during photosynthesis, and, as this accumulation of sucrose is of a very pronounced character, it is obvious that it cannot be stored in the presence of an enzyme actively engaged in bringing about its hydrolysis. It would follow, therefore, that either invertase is absent from the vacuoles where storage takes place, or else a mechanism is present whereby its activity is controlled ; that the latter explanation is more probable is indicated by the experiments ioinaeal in this paper. Until such factors as temperature, traces of alkali dissolved from the glass, changes in the hydrogen ion concentration of the sap, and the effect of shaking have been considered, further discussion cannot be profitable. Opportunity for this research was provided by a maintenance grant from the Department of Scientific and Industrial Research. The writer wishes to express his indebtedness to Prof. H. H. Dixon, F.k.s., under whose direction the work was carried out. = SCIENTIFIC PROCEEDINGS. VOLUME XVI. . A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, S0.D., F.R.s., and T. G. Mason, m.a., sa.B. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. Smyrn, B.a., so.B. (Plates I., II.) (February, 1920.) Is. . The Application of the Food-Unit Method to the Fattening of Cattle. By James Witson, u.A., B.sc. (Plates III., IV.) (February, 1920.) 1s. . The Holothurioidea of the Coasts of Ireland. By Anne L. Massy. (April, 1920.) 1s. . Photosynthesis and the Blectronic Theory. By Henry H. Dixon, sc.D., F.8.8., and Horacz H. Poots, so.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Win.iam Brown, B.Sc. (March, 1920.) 6d. . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, m.a., so.p. (April, 1920.) 1s. DUBLIN : PRINTED AT THE UNIVERSITY PRESS BY PONSONKY AND GIBBS. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 8. AUGUST, 1920. THE CHANGE IN THE RIGIDITY OF-NICKEL WIRE WITH MAGNETIC FIELDS. Satnsonlan Instat MAR 5 1991 | 2 x 10° 383x105 | 4 x 10° Amps. 4. 1), Cs D. Os | D.C. AN, Os I | | 0 0 0 0 0 2 = 0:75 4 OH | = OF | | 4 + 0-5 208 | = io + 10 | 6 A oy + 25 0 | 8 + 65 + 4:7 + 1:6 + 2:5 10 + 11°6 Se al 377 dL Bobs 14 | 492-2 A148 . || & oq £ 65 18 + 32:0 dt O)-5) + 14-5 | + 10-0 20 437-0 + 265 £167 || 22 4+ 41:7 + 80-4 + 19-0 | 24 + 46-5 + 34-0 AG) — || TIGIOE : 26 + 50°8 + 38:0 + 24-3 28 + 55:2 eA ON eT 0 + 18-0 30 4 59-2 445-7 LOO | 32 #682) | E49 | 82°65 | eoneg From the values in Table I] and the curves in fig. 2 it will be seen that the behaviour of nickel wire in transverse magnetic fields is the reverse of that in longitudinal magnetic fields—that is, there is at first a slight indication of a decrease in the rigidity—and that the main effect of the application of transverse fields, for the strongest fields that could be applied (960 units), is to increase the rigidity. It is possible that for higher magnetic fields than the strongest here applied the curves may come to a maximum, then decrease, and cross the axis in the manner of those obtained with longitudinal fields. From the 104 Scientific Proceedings, Royal Dublin Society. approximate values we were able to obtain with alternating transverse fields it is evident that their action is the reverse of that for alternating longitu- dinal fields, where the changes in the rigidity were greater than for direct longitudinal fields, whereas the changes for alternating transverse fields are less than for direct transverse fields. 70 2 Microscope Readings in «ms. x 10?. o Io zo Amperes. 39 Vic. 2.--Upper Curve D. C. Field. Load 2 x 10° grm./cm?. Lower Curve A. C. Field (2 = 50). Load 4 x 10° grm./em?. Tn conclusion, it may be said :— 1. The initial increase in the rigidity of nickel is less for alternating than for direct longitudinal magnetic fields, whilst the subsequent decrease is greater for alternating longitudinal fields than for direct fields. 2. The higher the frequency of the longitudinal magnetic field used the sreater the decrease in the rigidity. 3. The changes in the rigidity of nickel wire with transverse magnetic fields are the reverse of those due to longitudinal fields, both direct and alternating. 4. The greater the load on the wire, when subjected to transverse . magnetic fields, the smaller the change in the rigidity for fields between zero and 960 units, i SCIENTIFIC PROCEEDINGS. VOLUME XVI. A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon SC.D., F.R.s., and T. G. Mason, m.a., so.z. (January, 1920.) 6d. The Carboniferous Coast-Section at Malahide, Co. Dublin. By Lous B. SmuytuH, B.a., scp. (Plates 1., 11.) (February, 1920.) 1s. . The Application of the Food-Unit Method to the Fattening of Cattle. By James Witson, u.s., B.sc. (Plates III., 1V.) (February, 1920.) 1s. The Holothurioidea of the Coasts of Ireland. By Anne L. Massy. (April, 1920.) 1s. Photosynthesis and the Electronic Theory. By Henry H. Dixon, so.p., F.R.S., and Horace H. Poon, sc.p. (March, 1920.) 1s. Note on the Decay of Magnetism in Bar Magnets. By Wiuriam Brown, B.sc. (March, 1920.) 6d. On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, u.a., sc.B. (April, 1920.) 1s. . The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Winniam Brown, B.Sc., and Parrick O’CaLLAGHAN, A.R.0.SC.1., A.L.c. 6d. <_< DUBLIN : PRINTED Att THE UNIVERSITY PRESS BY PONSONBY AND GIBES. THE SCIENTIFIC PROCKEDINGS ROYAL DUBLIN SOCIETY. Vol. XVI. (N.S.), No. 9. AUGUST, 1920. SOME DERIVATIVES OF NITROTOLUIDINE (4-nitro-2-amido-1-methyl-benzene). BY A. G. G. LEONARD, A-R.C.Sc.1., B.Sc., Pa.D. ; AND Sean aT AGNES BROWNE, A.R.C.Sc.I., B.Sc. (ws a tama) nese? [Authors alone are responsible forall opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSLER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1920. Price Sixpence. Roval Bublin Soctety, ae FOUNDED, A.D. 1731. TNCORPORATED, 1749. EVENING SCIENTIFIC MEETINGS. Tux Scientific Meetings of the Society are usually held at 4:15 p.m. on the third 'Ltuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. Lhe copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Mditor. oy 1X Heuel Muse SOME DERIVATIVES OF NITROTOLUIDINE (4-nitro-2-amido-1-methyl-benzene), By A. G. G LEONARD, A.R.C.Sc.1L, B.Sc., PH.D. ; AND AGNES BROWNE, A.R.C.Sc.1., B.Sc. Read Junv 22. Published Aueusr 10, 1920. THE nitro-ortho-toluidine C,H;(NH.) Me(NO.) [NH,:Me:NO,=1:2: 5] used for the following preparations was obtained by nitration of o-toluidine in excess of sulphuric acid (Nolting and Stoecklin, Ber. 24, 566). Three lots of 100 grams of the base gave an average yield of 83 grams of nitro- toluidine, m. pt. 107° C. Nitro-methyl-phenylazo-B-Naphthol, C,H;Me . NO, . C,H, . OH 2.) (8.) a; 1) (2.) 5 grams nitro-ortho-toluidine were dissolved in 75 c.c. warm sulphuric acid (10°/,), and the solution rapilly cooled to 0°C., whereby the-sulphate was precipitated in a fine state of division. A cold solution of 2°27 gram sodium nitrate was added slowly, the diazo solution quickly filtered and added with constant stirring to a cold solution of 4°74 grams of (-naphthol in 500 c.c. of a 3°/, solution of sodium hydroxide. During the mixing of the liquids a bright red precipitate separated. Stirring was continued for an hour, when the whole was heated to 80°C. on the water bath, cooled, filtered, and the precipitate washed with water. Yield, 9°5 grams. The product ‘thus obtained was bright red,and melted with decomposition at 204° C. It is insoluble in water, hydrochloric acid, and sodium hydroxide ; it dissolves in benzene, ether, and alcohol. and crystallizes from its solution in alcohol in long red needles. When applied to the fibre in the same way as p-nitraniline red, it dyes cotton a deep orange colour. Analysis :— (1) 2997 gram gave 33°6 c.c. moist cteak, at hee C. and 785) mm. 7, N = 18°42. (2) 3008 gram gave 34-4 ¢.c. moist nitiogen at 18°C. and 765 mm. J, N = 13°30: hee C,,H,;03 N;- requires: 13°68. SCTENT. PROC. &.D.S., VOL. XVI., NO. IX. N Tesonlan lietiiugy ley \ 106 Scientific Proceedings, Noyal Dublin Society. Nitro-methyl-benzene-diazo-amino-o- toluenc C;H;.NO,.Me.N,.NH.C,H,. Me. (9) (2) G) @) (2) 7-6 grams of nitro-toluidine were dissolved by warming with 50 c.c. cone. hydrochloric acid diluted with an equal volume of water. The solution was quickly cooled in ice, and the requisite quantity of sodium nitrite solution added. The diazo solution was rapidly filtered, diluted to 250 c.c., and excess of sodium acetate added. 5°35 grams of o-toluidine were dissolved in 30 c.c. cone. hydrochloric acid and 50 c.c. water ; the solution was diluted to 500 c.c., and excess of sodium acetate added. This solution was then cooled and added to the diazo solution previously prepared, when an orange yellow precipitate separated gradually. The substance was filtered off after standing overnight and well washed with water. The crude product melted at 108° ©., but after four crystallizations from alcohol the melting point rose to 133°C. The substance so obtained consisted of needle-shaped crystals, which decomposed on boiling with hydrochloric, sulphuric, and acetic acids, giving nitro-cresol [OH : Me: NO, =1: 2: 6], o-toluidine, and nitrogen, thus showing it to be a diazo-amino compound. Again, on heating with hydrochloric acid and (3-naphthol it yielded nitro-methyl-phenyl-azo-(-naphthol m.pt. 204° C. ; it was found that about 12°/, of the diazo-amino product reacts in this manner, the remainder yielding nitro-cresol, toluidine, and nitrogen. Analysis :— 1500 gram gave 25°8 ¢.c. moist nitrogen at 12° C. and 772 mm. 7, N = 20:86. Cy,HN,O, requires 20°74. The “diazo” nitrogen was estimated by the method of Mehner (Journ. fur Prakt. Chemie [2], 63, 304). 3000 gram gave 28 ¢.c. moist nitrogen at 20°C. and 764 mm. J, diazo nitrogen = 10°76. C.H;NO,.Me.N:N.NH. OC,H,Me requires 10°37. ° On concentration of the mother liquor from the crystallization of the above compound a small quantity of reddish crystals separated. This was picked out and crystallized from alcohol, m.pt. 144° C. Under the micro- scope the crystals were seen to be cubic. ‘The amount of the substance was too small for analysis ; it was probably the isomeric 4-amino-3 : 2’-dimethy]- 5-nitro-azo-benzene. Nitro-methyl-benzene-diazo-amino-p-toluene Cs;H;NO..Me.N..NH.O,H,. Me. ©) @ @ © (4) 1°52 gram nitro-toluidine was dissolved in 4 c.c. conc. hydrochloric acid and 50 c.c. water, cooled to 0°C., diazotized with the requisite quantity of Lronarp anp BRowNne—Some Dericutives of Netrotolurdine. 107 sodium nitrite solution, the whole filtered, and excess of sodium acetate added. The clear solution was added slowly to a solution of 1:07 gram p-toluidine in + c.c. cone. hydrochloric acid and 200 c.c. water containing excess of sodium ‘acetate. A yellow substance gradually separated, which was filtered off after standing overnight. Twice crystallized from alcohol, the crystals melted with decomposition at 131°C. It proved on examination to be a diazo amino com- pound, decomposing on treatment with acids into nitro-cresol, p-toluidine, and nitrogen. Analysis :— 1500 gram gave 25°8 c.c. moist nitrogen at 145°C. and 767 mm. 7, Nitrogen = 20-50. C,H,.N,O, requires 20°74. Diazo nitrogen :— ‘3000 grams gave 27°8 cc. moist nitrogen at 20° C. and 767 mm. 7, Diazo nitrogen 10°33. C,H;NO..Me.N:N.NH.O,H,. Me requires 10°37. Nitro-methyl-diazo-aminobenzene-p-sulphonic acid C,H, .NO,.Me.N.. NH. C,H , SO.H. @) @ @ © (4) goth molecule of nitro-o-toluidine dissolved in 75 c.c. of 10 °/, sulphuric acid and cooled to 0° OC. was diazotized by goth mol. of sodium nitrite. The solution was then added with constant stirring to a solution of j'>th mol. of sulphanilic acid in 100c¢.c. of a solution containing ;5th mol. of sodium hydroxide. After a short time a yellow amorphous substance separated, which, after washing and drying in vacuo, melted at 129°C. with decomposition. It is insoluble in water, alcohol, ether, and benzene in the cold, and enly slightly soluble on heating. When warmed with acids it decomposes with evolution of nitrogen, showing it to be a diazo-amino compound. Analysis :-— 3000 gram gave 45-2 ¢.c. moist nitrogen at 16°C. and 752 mm. “he nitrogen = ier C,3;H,.N,SO; requires 16°67. Diazo nitrogen :— 3000 gram gave 21 cc. moist nitrogen at 17°C. and 761 mm. °/, diazo nitrogen = 8°19. : C,H, . NO..Me.N:N.NH.C,H,SO,H requires 8:33. 108 Scientific Proceedings, Royal Dublin Society. Methyl-nitrodiazoamino-p-nitrobenzenc C,H;. NO,. Me.N:N.NH C,Hy. NO,. (5) (2) Q) (1) (4) sth mol. of nitrotoluidine dissolved in sulphuric acid and diazotized as, before was added with constant stirring to g>th mol. of p-nitraniline in 75 c.c. of 10°/, sulphuric acid, the whole being kept at 0°C. On standing overnight a yellow substance separated, which was filtered off, washed and dried in vacuo. It dissolves in warm alcohol, but is insoluble in ether and benzene. It melts with decomposition at 118°C. It is decomposed by boiling with acids with evolution of nitrogen, showing it to be a diazo animo compound. Analysis :— (i) 1500 gram gave 30-0 cc. of moist nitrogen at 19° C. and 762 mm. (ii) -1500 gram gave 29°6 c.c. of moist nitrogen at 17° C. and 767 mm. = °/, nitrogen = 23°23 and 23:27. Cy3Hy,N;O, requires 25° Diazo nitrogen :-— -3000 grams gave 22-4 cc. “eis nitrogen at 18°C. and 764 im. “/., diazo Bae = 88 C,H;NO,.Me.N: Ee. C,H, . NO, requires 9°30. 2-Methyl-5-nitro-2' : 4’-dihydroxyazobenzene CH; .NO,.C,H;. N..C,H;. (OH). (2) (6) (1,1) (2, 4) 2 Gi Distance of junction from end of tube, —. > 2h) Gi Diameter of tubes, . : : : F Se OuCTIN: Pressure in reservoir, —. ; 5 : . atmospheric. Length of Deflection of explosion-tube. galyanometer. 1) att 74 B55 147 @ 5 202 4, 245 D x 287 6, 310 11 295 Thus the temperature produced increased in tubes up to 6 feet long, but had decreased for a length of 11 feet. It was not convenient to use tubes between 6 and 11 feet long. 300 200 DEFLECTIONS. 100 |- Feet. Lenetu or Tuse. Curve 1.—Showing relation between deflections and length of explosion-tube. 114 Seientifie Proceedings, loyal Dublin Society. Tasue IT. Relation between the distance of junction from the closed end and the galvanometer deflection :— (a) Length of explosion-tube, : ‘ : . 3d ft. Diameter of tube, J ; : 3 5 iii, Initial pressure of air in tube, , : o & Gili, Pressure in reservoir, —. : : ‘ . atmospheric. Distance of junction from Deflection of end of tube, galvanometer. 2°5 em. 202 HAN) 5 205 OW 180 15:0)"; 154 200 ,, 111 30:0, 76 40:0, 50 5070) ,, 32 60:0 ,, Ig 100 5 13 80:0 _,, 7 85:0 6 4% 200 SI o 150 H 2 qQ 100 i oo le 1S 30 45 60 75 90 cm. DisTaNnckE OF JUNCTION FROM END or TuBE. Curve 2a.—Showing relation between deflection and distance of junction from closed end of explosion-tube. Tube 3 feet long. McC ien.anp anp Gitt—Self- [gnition of lither-Air Mixtures. 115 Tabie III. (6) Length of explosion-tube, : ee its Diameter of tube, . : ‘ : 5 BM, Initial pressure of air in tube, . 5 3 » 2 Oi, Pressure in reservoir, ‘ é : . atmospheric. Distance of junction from Deflection of end of tube. galvanometer. 1°5 em. 135 On Oars 166 OW -,, 146 SOs 102 25°0 66 30°0_,, - Al 45:0 ,, 32 59°0 23 3 200 iS} S 150 [e3} =} & & 100 | 50 15 30 45 60 cm. DisraNcre OF JUNCTION FROM END OF TunE. Curve 2b.—Showing relation between galyanometer deflection and distance of junction from closed end of explosion-tube. Tube 2 feet lone. In all cases where the junction was very near the disc closing the explo- sion-tube, the temperature as indicated by the deflection of the galvanometer fell off. This falling-off was, no doubt, due, in part at least, to the cooling effect of the neighbouring metal disc, 116 Scientific Proceedings, Royal Dublin Society. TABLE IV. relation between pressure difference in reservoir and explosion-tubes and galvanometer deflection :— Length of explosion-tube, . : : 2, Orb. Diameter of explosion-tube, . : one Distance of junction from end, . , , im, Initial pressure in tube, . ; : » 2 Git, Pressure in reservoir varied. Difference of pressure in - Deflection of reservoir and tube. gvalvanometer. 1 cm. 15 Dee 24°5 ah 39 ay 57 5) on 75 lO, 95 AO) 5, 35 BO) 5, 160 40 ,, 182 50 ,, 200 GO ., 215 HO 5 226 115 230 For the last reading the pressure of air in the reservoir was increased by means of a force-puimp. 200 100 DEPLECTION. 20 40 60 80 100 120 em. DIPreRuNCE OF PRESSURE RETWEEN Reservoir AND TUBE. Gunrve 8,—Showing relation hetween pressure difference in reseryoir and explosion-tube’ and galyanometer deflection, McCietianp anp Giut—Self Ignition of Lther-Atr Mixtures. 117 An experiment was made to determine the effect of placing a pad of cotton wool at the end of the tube. In the first experiment the junction was 8 cm. from the end of the tube. The pad was about 5 cm. thick. The pressure of air in the tube was 2cm. ‘The length of the tube was 3 feet, and the diaineter 3 inches. Without pad, . ; . Deflection, 151. With pad, 5 : 3 5 110. In the second experiment the pad was not so thick, and the junction was placed quite near it. The other conditions were the same. Without pad, . ; . Deflection, 155. With pad, Hl 143. In both cases there was apparently a diminution of temperature, which agrees with the result already described as to absence of ignition of an ether-aiz mixture under the same conditions. The part played by the connecting tube was also tested by this method: The tube usually employed was 30 inches long. This was replaced by a tube only 2 inches long. In experiments carried out under similar condi- tions, except for the lengths of the connecting tubes, there was no appre- ciable difference in the deflections due to the thermo-couple. ‘I'his result was to be expected from the conclusions already arrived at. C.— Actual temperature reached by the air in the above experiments. The numbers given above, expressed in terms of the galvanometer deflections, merely indicate the relative values of the temperature. To find the actual temperature reached we employed the following method :— A, B (fig. 2) represents a potentiometer through which a current can be sent from the cell E when the key K is closed. G isa galvanometer and J the thermal-junction nounted inside the tube into which the air is allowed torush. Mis an electro-magnet, from which the wires (a and 0) are taken to the point where the spring and lever arrangement opens the tap, allowing the air to rush from the reservoir. It is arranged that the lever opens the electro-magnet circuit, and consequently the galvanometer circuit closes at D. The contact C is varied, so that the potentiometer electromotive force balances the electromotive force of the thermal-junction. The potentiometer e.m.f. can be varied, so that the slight sudden kick of the galvanometer is in either direction when the tap is opened. Of course, the junction very soon begins to cool, and then the galvanometer moves in the direction corresponding to the emf. of the potentiometer. The exact instant of closing the galvano- meter circuit could be varied relatively to the time of opening the tap, and in SCIENT. PROC. R.D.S., VOL. XVI., NO. X. P 118 Scientific Proceedings, Royal Dublin Society. this way the maximum temperature reached by the thermal-junction could be found. A careful determination of the maximum temperature was made in this way when the junction was placed 25cm. from the end of the explosion- tube 3 feet long and 3 inches in diameter. The pressure of the air in the tube was 2cm., that of the air in the reservoir being atmospheric. About 3 feet is the shortest length of explosion-tube which was certain to give good ignition. The junction was carefully calibrated over the range indicated by the galvanometer deflections. The temperature reached by the air under these conditions varied between 185° and 193°C. —— * Fic. 2.—Potentiometer arrangement by means of which the actual temperature reached in the explosion-tube was determined. From the results of the various temperature determinations we made, it may be concluded that at a short distance from the closed end of the explosion-tube, 3 feet long, a temperature as high as 190° C. may occur when air at atmospheric pressure is allowed to rush in, the pressure of the residual air in the tube being about 2cm. of mercury. From our earlier experiments it would follow that higher temperatures would be reached with longer explosion-tubes. Assuming that when air is replaced by an ether-air mixture the tempera- ture reached is not very different, we have to consider whether this rise of temperature is sufficient to ignite the mixture. M. E. Alilaire (Comptes Rendus, tome elxviii, No. 14, Avril, 1919, p. 729) has recently studied the ignition of ether-air mixtures, and has found it possible to produce ignition ata temperature of 190°C. He states that with modifications of his apparatus he is of opinion that ignition could be obtained at even lower temperatures. The agreement is close between the lowest SCIENTIFIC PROCEEDINGS. VOLUME XVI. 1. A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, S$O.D., F.R.S., and T. G. Mason, u.a., so.B. (January, 1920.) 6d. 2. The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. Smyru, B.a., so.p, (Plates I., IJ.) (February, 1920.) 1s. 8. The Application of the Food-Unit Method to the Fattening of Cattle. By James Witson, wa. B.so. (Plates III., IV.) (February, 1920.) 1s. 4, The Holothurioidea of the Coasts of Iveland. By Anne L. Massy. (April, 1920.) 1s. 5. Photosynthesis and the Hlectronic Theory. By Henry H. Dixon, so.p., F.R.S., and Horace H. Poors, sc.p. (March, 1920.) 1s. 6. Note on the Decay of Magnetism in Bar Magnets. By Witriam Brown, B.Sc. (March, 1920.) 6d. 7. On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, u.a., sc.p. (April, 1920.) 1s. 8. The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Wituiam Brown, B.so., and Patrick O’CaLLAGHAN, A.R.0.Sc.1.,4.1.c. (August, 1920.) 6d. 9. Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. Lzonarp, A.R.0.80.1., B.Sc., PH.D., and Aenes Browne, A.R.0.SC.1., B.sc. (August, 1920.) 6d. 10. An Investigation into the Causes of the Self-Ignition of Ether-Air Mixtures. By the late Professor J. A. McCintuanp, p.sc., F.R.s., and Rey. H. V. GILL, 8.3., D.s.o., M.c., m.a., University College, Dublin. (August, 1920.) 6d. MUHLIN : PRINTED AT THE UNIVERSITY PRESS BY PONSONRY ANI GIBBS. By ghey Ria RNS: THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.S.), No. 11. OCTOBER, 1920. THE INFLUENCE OF ELECTROLYTIC DISSO- CIATION ON THE DISTILLATION IN STEAM OF THE VOLATILE FATTY ACIDS. BY JOSEPH REILLY, M.A., D.Sc. F.R.C.SeL., AND WILFRED J. HICKINBOTTOM, B.Sc. ao EN i [COMMUNICATED BY DR. F. FE, HACKETT, M.A.]| MAR 5 1921 1 Wiuse*~ [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN : PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATEH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1920. Price Sixpence. Roval Bublin Society. To FOUNDED, A.D. 17381. INCORPORATED, 1749. —_—~ EVENING SCIENTIFIC MELTINGS. Tax Scientific Meetings of the Society are usually held at 4:15 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations im a complete form, and ready for transmission to the Editor. McCLeLLAND AND GiILL—WSel/-Ignition of Ether-Air Mixtures. 119 ignition temperature found by Alilaire and the temperature we measure in the shortest tube in which we found it most satisfactory to obtain ignition. Iiscussion of Results. 1. With the type of apparatus employed we have found that when the explosion-tube is about 3 feet in length, or longer, a temperature of 190° C., or higher, may occur near the end of the tube. It has been found (Alilaire) by other methods that a temperature of 190° C. is sufficient to ignite an ether-air mixture. We may therefore conclude that the ignition observed in these experiments 1s simply due to the rise of temperature produced. 2. We may now consider how the high temperature is produced in the explosion-tube, the distribution of temperature along the tube, and the way the temperature produced depends on the dimensions of the apparatus. When the tap is opened the inrush of air, etec., from the reservoir reaches a high velocity, approximately equal to the velocity of sound. A stream of gas travels down to the further end of the explosion-tube, and the reflected stream meets the direct stream, with the result that the translational energy is quickly _e Sey ae hs é Q Bate he « 4 Mae, ie pares BOR Lre te ULE ws ADAG oy aha hie THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 12. OCTOBER, 1920. NOTES ON SOME APPLICATIONS OF THE METHOD OF DISTILLATION IN STEAM. BY JOSEPH REILLY, M.A., D.Sc., F.R.O.Sc:E.;~- AND WILFRED J. HICKINBOTTOM, B.Sc. MAR 5 1921 ational encore [COMMUNICATED BY DR. F. FE. HACKETT, M.A.] [Authors alone are responsible for all opinions expressed in their Communications.} DUBLIN : PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.-C. 2. 1920. Price Sixpence. Roval Bublin Society. ee OeEOEeeeee FOUNDED, A.D. 1781. INCORPORATED, 1749. OES ——BYENING SCIENTIFIC MEETINGS, Tae Scientific Meetings of the Society are usually held at 4:15 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring’ to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be get down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, , and such as are considered suitable for the sree will be spteltedl Hlth the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Hditor. oo \ans ni Te yi omy Hsenia netiive Ne ae gid ( MAR’ 5 1921 XI. NOTES ON SOME APPLICATIONS OF THE METHOD OF DISTILLATION IN STEAM. By JOSEPH REILLY, M.A., D.Sc., F.R.C.Sc.1., AND WILFRED J. HICKINBOTTOM, B.Sc. [COMMUNICATED BY DR. F. E. HACKETT, M.A. | [Read Aprin 26. Published Ocrozer 11, 1920.] Synopsis.— Distillation of dilute solutions; Applications. The limitations and possibilities of the distillation method for determining molecular structure are dis- cussed. A consideration of the distillation constants of a number of compounds shows that, although the molecular structure may be inferred in certain cases, this method cannot be applied generally without the support of further evidence. Determination of changes iu state of molecular aggregation. The distillation constant is deduced from Nernst’s law of distribution ; consequently any alteration in the state of molecular aggregation during distillation will alter the distillation constant. A scheme has been proposed by which it is possible to follow any change in the solute occurring in solutions the concentrations of which differ widely. A method has also been devised to differentiate butter-fat from other fats containing a comparatively high proportion of volatile acids. It is based on the presence of butyric acid in butter-fat, while this acid does not occur in appreciable amounts in other fats of the types examined. THE distillation constants of the lower fatty acids containing a normal chain rise In an approximately regular manner with increasing molecular weight. With acids in which a branched chain occurs the distillation constants are higher than those containing a normal chain. They correspond approximately with the distillation constants of the next higher member in the series of normal compounds.’ These facts suggest that the reguiarity may be general. The determinations of the distillation constants of the lower aliphatic alcohols showed that in this series there was also a parallel between the distillation constant and the molecular weight. Furthermore, the distillation constants of isomers were higher if the chain of carbon atoms was not a normal one. It appears, therefore, in the homologous series of saturated aliphatic alcohols and acids that the distillation constant of the substance may be deduced if the molecular configuration of the acid and its molecular weight are known. 1 Reilly and Hickinbottom, Proc. Royal-Dub. Soc., xv, 37, 513. SCIENT. PROC. R.D.S., VOL. XVI., NO. XI. s 132 Scientific Proceedings, Royal Dublin Sociely. Table { gives the constants calculated from values recorded in the literature! for acids other than the volatile fatty acids. Table II gives the data for the latter acids. TABLE I. kK BK Benzoic, i 0:22 p-Toluie, : : 0:33 o-Toluic, : 3 0-51 o-Hydroxy benzoic, 0:09 m-Toluic, : : 0:40 p-Hydroxy benzoic, very small The acids, glycollic, monochloracetic, oxalic, malonic, mandelic, and phenylacetic, have small distillation constants. “ Nitroglycerine” is also dis- tillable in steam, and its distillation constants are now being determined. TABLE II. 1 @ 4\* Boiling-point [ses | (elt [700mm] XO). Formic acid, F 0:00128 361 100°6 Acetic acid, 4 0:00215 3°61 118 Propionic acid, . 0:0036 1-79 140 n-Butyric acid, . 0:0055 Ie} 7/ 163 iso-Butyric acid, . 0:0073 1-44 155 * The initial volume of solution in the flask, 150 c.c. + Ramsay Shields’ association factor. The authors? have also investigated the rate of steam distillation of the -2@ following table, and the comparative values for # are also recorded :— lower alcohols. The values obtained for E log z | are given in the a TasLE IIT. real [1 “faba Methyl alcohol, . 0015 3-44 66 Ethyl alcohol, . 0-024 2-75 78°3 n-Propyl alcohol, . — 2°27 97-4 n-Butyl alcohol, . 0:040 159) 117 iso-Butyl alcohol, . 0066 - — 108 sec-Butyl alcohol, . 0-066 — 99°8 iso-Amylalcohol . 0:072 — 131 1 Of. Stein, J. Pr. Chem., 1913, 88, 83. 2 Unpublished work (J. R. and W. J. H.) which will be submitted in detail to the Royal Dublin Society later in 1920. Remy any Hicxinsorrom—Method of Distillation in Steam. 138 Hausbrand! states that the composition of vapours (from liquids which mix with water) depends, “according to certain laws, upon the composition of the boiling mixture of liquids, but, unfortunately, is not accurately known for most mixtures of liquids, although this property is utilized on the largest scale in the industries for the distillation of such liquids.” Even in the case of a liquid which does not mix with water, part of the liquid may be mechanically “taken away” with the steam, and for some substances the pro- cess of “blowing over” with saturated steam has wide industrial applications. Hausbrand gives the following approximate data for the amount of steam required to “carry over” 100 kilograms of various substances :-— Toluene requires. . 13-15 Ixilos of steam Benzene PF ; me OD Oh \iaa aire tee Fatty Acids require > 1d ene ut Tar requires ‘ Beit) 2 ® Glycerine requires . . 200 DD Nitrobenzene ,, ; » 2X00 5, 9 =» Nitrotoluene ,, 5 . 400-450,, ,, ,, (1 kilo of steam at atmospheric pressure represents 637 calories.) No differentiation is made between the cases in which mixtures of constant boiling-point, or of mechanical carrying over of the liquid in the current of steam, occur. Many other substances are purified by distillation in steam, such as various “intermediates” (e.g., aniline, orthonitrophenol, chloropicrin, etc.), petroleum products, essential oils, various esters, lactic acid, ete. It should be noted that even among compounds closely related, the vapour pressures at 100° may differ widely, and no @ priori reasoning, without actual experiment, can decide which compound will distil the more readily in steam. The results recorded above indicate in a broad manner the effect on the distillation constant of introducing various groups into the molecule. A phenyl group reduces the distillation constant, while an hydroxyl group replacing a carboxyl group causes an increase. When, however, carboxyl and hydroxyl groups are present in the molecule, the distillation constant is generally less than that of a compound containing only one of these radicles, The distillation constant is affected in a striking manner by constitutive differences in isomeric substances, and, as far as can be judged from the limited number of observations available, the influence of constitution on the distillation constant appears to be general. The constants of acids andalcohols containing a branched chain are higher 1 Evaporating, Condensing, and Cooling Apparatus, 2nd ed., pp. 19-20. 82 134 Scientific Proceedings, Royal Dublin Society. than those of isomeric substances in which the carbon chain is a normal one. In the aromatic series the position of the substituent group in the benzene ring has an influence on the distillation constant. Both in the case of the cresols and the toluic acids the ortho compound has a greater distillation constant than either the meta or the para derivatives. It is noteworthy that the distribution of o-, m-, and p-toluic acids between liquid and vapour phase in dilute solution bears the same relationship to each other and to the constant for benzoic acid as the constants of o-, m-, and p-eresol and phenol. The distillation constants of n- and dso-butyrie acids show a similar relation to those of n- and dso-butyl alcohols. The influence of constitution on the distillation constant isshown by o- and p- hydroxy-benzoic¢ acids, and by o- and p- nitro-phenols. he results in the case of substances only sparingly soluble in water may be explained by a difference in the vapour-pressure of the substances at the boiling-point of the solution. In the case of phenols and the lower fatty acids and alcohols the distilla- tion constant appears to bear no relationship to the vapour-pressure. Pure methyl alcohol at 100° exerts a.greater vapour-pressure than any of the other homologous alcohols, yet in dilute aqueous solution there is a smaller dis- tribution coefficient between liquid and vapour phases than is the case with ethyl, butyl, or amyl alcohols. It seems to indicate other factors which influence the distillation of aqueous solutions. The molecular weights of the substance distilled have an influence in comparative distillations of members of a series of compounds. In ‘lables II and III a comparison’ is made between the Ramsay-Shields’. association factor x, the distillation. constant, and the boiling-point. It is seen that as the association’ factor decreases the distillation constant increases. The association factor refers to the pure liquids; consequently no direct comparison can be made between the molecular state of the pure liquid and its’ aqueous solution. It shows, however, that for the lower members, as the homologous series is ascended, a gradual modification occurs in the attractions or forces exerted, between the molecules. The inter-molecular forces are also modified by solution owing to the force exerted by molecules of the solvent on those of the solute. For miscible liquids or soluble substances, the greater the difference between the mean attractions of the molecules of the solute for each other and the solute for the solvent, the more abnormal will be the behaviour of the solution on distillation.1 With the lower fatty acids and alcohols a gradual diminution with increasing molecular weight occurs in the attraction between molecules of ‘Cf. Berthelot, Compt, Rend., 1898, 126, 1703. Retuiy anv Hiexrysorrom— Method of Distillation in Steam. 135 the solvent and solute, using water as solvent. When such disturbing influences have been eliminated, it might be expected that the distillation constants and the vapour-pressures should be in proportion. The distillation constants of the fatty acids and the alcohols should reach a maximum, and then decrease with increasing molecular weight. Such has been found to occur in the fatty acid series. Distillations of aqueous suspensions of lauric and myristic acids were carried out. These acids are practically insoluble in water, consequently the effect of the solvent is eliminated. It was found that the distillate from’ the lauric acid suspension was richer in acid than that from the myristic acid- water mixture. The results obtained by distilling dilute solutions of the volatile fatty acids show that with the method it is possible to detect alterations in the state of molecular aggregation. The success of the method depends on — (a) Constant temperature of distillation, and freedom from variations caused by irregular heating or condensation. (6) Convenient and accurate method of estimating the solute. (c) Choice of a suitable solvent. To maintain a constant temperature, the solution must necessarily be of such a dilution that changes in concentration do not affect the boiling-point appreciably. This can be achieved by allowing only a small change in concentration, or by employing very dilute solution, such that the boiling-' point approximates to that of the solvent. It is, however, possible to correct, for change of temperature, and an apparatus has been devised so that distillation can be carried out over a considerable range of concentration and temperature of ebullition. It is intended to describe this at some future date. oe The solute need not necessarily be a volatile fatty acid, and the choice of a solute is restricted only by the solubility in the solvent and its volatility. It is, however, essential that an accurate method should be available for the estimation of the solute. By using the same solute in a series of different solvents, and employing a considerable range of concentration, evidence may be obtained concerning the actual state of substances in solution. .This applies more particularly to concentrated solutions, which cannot be con- veniently investigated by methods involving the principles of osmotic pressure. The method of distillation of dilute aqueous solutions has been. applied chiefly for the detection and estimation of mixtures of fatty acids. It is from this point of view that most of the investigations have been carried out, and the conditions determined for obtaining accurate results.! It follows . 1Upson, Plum, and Schott. J.-Amer. Chem. Soc., 1917, 39, 731; Lamb, zbid., 1917, 39, 746. Gillespie and Walters, ibid., 1917, 39, 2027 ; Richmond, Analyst, 1919, 44, 255. 136 Scientifie Proceedings, Royal Dublin Society. from the variation of distillation constant with dilution that this factor should not be neglected. Similarly, the presence of non-volatile impurities in the solution may have a considerable influence on the result. A comparison of butter-fat with other edible fats and oils shows that while most of the fats are glycerides of acids of higher molecular weight than caproic, butter contains an appreciable amount of butyric acid. It might be expected that the acids from butter-fat could he distinguished from those of the other fats by their distillation constant. Preliminary experiments confirmed this view, and a method has been worked out for determining the distillation constants of the volatile acids in fats. The principles underlying the distillation of dilute solutions have been applied in certain industrial processes. In the purification of alcohols, use is made of the fact that in dilute solutions the higher alcohols are removed more rapidly than the lower ones. This is the principle involved in the Guillaume stills for alcohol distillation. A dilute solution of the raw spirit is distilled when the fusel-oils are removed in the first portion of the distillate, while an aqueous solution of ethyl alcohol remains in the still. The addition of salts, in increasing the distillation constant of acetic acid, is the basis of one of the methods which have been employed for obtaining concentrated acetic acid from dilute aqueous solutions. It is uneconomical to do this by simple fractionation of the vinegar. ‘he method as used industrially consists in distilling the dilute solution in presence of a salt, so that the distillate is richer in acetic acid than the original solution. The Nstillation of Aqueous Solutions of Lawric and Myristie Acids. In order to determine if the increase in the distillation constant was general as the homologous series ascended, the constants of laurie and myristic acids have been determined. These acids occupy in the fatty acid series a position removed from acetic and butyric acids. They are almost completely insoluble in water; con- sequently it may be assumed that heat of solution and volume changes on mixing are eliminated. A known weight of the pure acid was added to a known volume of distilled water. The resulting mixture was then distilled in a round- bottomed flask, surrounded by a steam-jacket, in the same way as described for previous experiments. It was found that on conducting preliminary experiments, a considerable amount of the acid solidified in the condenser, and vitiated the results to some extent. The defect was remedied by using a special condenser arranged so that it could be kept at a temperature above the melting-point of the acid used, ReErLiy anv Hicxrnsorrom— Method of Distillation in Steam. 187 This could be arranged either by making the jacket of the condenser part of a warm-water circulation system, or by supplying cold water to the condenser in a very slow stream, so that the acid does not solidify. The former method is the one more easily regulated, and it is convenient for a number of determinations. Diagram shows the apparatus as finally set up, and as used for the distillation of aqueous suspension of laurie and myristic acids, [Fig. 1.] TS SSSsss : hy Fie. 1. The mixture of acid and water was collected in a series of weighed cylinders. The distillate was washed out by means of neutral alcohol, and the washings were repeated until all the acid had been dissolved. It was then titrated with standard barium hydroxide solution, using phenolphthalein as an indicator. It was considered better in the case of these acids to obtain a constant depending on the ratio of acid to water in the distillate, and to neglect the acid originally present in the flask. Supposing the system to be a mixture of immiscible and unlike liquids, the amount in the flask should not have any effect. Using a considerable excess of acid this holds approximately, but it is found that 138 Scientific Proceedings, Royal Dublin Society. when small amounts of acid are used errors are introduced, probably owing to particles of the hquid adhering to the walls of the flask. - The experimental results are given in Table IV, and constant Cin the formula has been determined : W, x My | Wages. where WV, and Wy are weights of acid and water, JZ, and M, are molecular weights of acid and water, C= ratio of vapour tensions at 100°. The constant C, when immiscible and unlike components are used, is ratio of the vapour-tensions of the two substances. The constant has therefore been used to calculate the vapour tension of lauric and myristic acids. The temperature of the liquid is taken as 100°— the boiling-point of distilled water—because the amount of the acids present and their insolubility in water render it improbable that they will affect the Gi temperature of boiling to any appreciable extent. On account of the relatively small amount of acid present, the vapour pressure is taken as atmospheric. The calculated vapour tensions are lower for myristic than for laurie acid. TABLE LV. Laurie Acid. Composition of distillate. Cc Vapour tension Water. Acid. (calculated). 3911 0:0256 7-6 x 10° 0-058 mm. Myristic Acid. 60-45 0:0327 4:3 x 10° 0:033 mm. These results are the mean of several determinations. The Distillation of Aqueous Solutions of Phenols. For the estimation of phenols in aqueous solution the method due to Messinger and Vortmann! was employed. The method. consists in adding a known amount of standard iodine solu- tion to the phenol at a suitable concentration in presence of approximately three molecular proportions of sodium hydroxide, and heated to 60°. The iodine is added until the liquid assumes a yellow colour, and a yellowish eurdy precipitate separates on shaking. The solution is cooled, diluted to a 1 Ber., 1889, 22, 2313, Of. Bergault. J. Pharm. Chem., 1908, 28, 45. Remy anp Hickinsorrom—Method of Distillation in Steam. 139 ~ known volume, and a measured portion of the diluted solution is withdrawn, filtered, and titrated with standard sodium thiosulphate, after rendering it faintly acid. From a knowledge of the amount of iodine used originally, and the volume of thiosulphate required for the final titration, the amount of phenol present in the solution can be obtained. It was found advisable during these experiments to determine the amount of iodine required for 4 known weight of each phenol before proceeding with the actual determina- tions. 10 08 2 oO ° aS log, (phenol in flask). log;, (volume in flask). Fie 2. The phenol employed was white, and had been purified before use. o-Cresol was obtained from pure o-toluidine by means of nitrous acid, and it was fractionally distilled before use. p-Cresol was bought as pure, and was not further purified It had a melting-point of 36°3° (uncorrected), and was white in colour. A solution of the phenolic body was prepared, containing a known weight of the substance. The amount of iodine required to precipitate the phenol in an aliquot part was determined, and the mean of all the determinations taken. SCIENT. PROC. R.D.S., VOL. XVI., NO. XII. T 140 Scientific Proceedings, Royal Dublin Society. 200 ¢.c. of the phenolic solution were distilled in a round-bottomed flask. heated on an electric hot plate, and surrounded by a steam-jacket, the usual precautions being taken to ensure regular ebullition, and to prevent access of cooling draughts. The distillate was collected in weighed stoppered cylinders, approximately 10 grams of distillate being collected in each fraction. The exact weight was obtained in each case by weighing the full cylinders. For the estimation of the phenolic substances in each fraction the distillate was diluted to a suitable concentration, and the estimation carried out as described above. The values log, (cresol left in flask) were plotted against log, (volume left in flask). The points obtained fell along a straight line. Then the value of p, the distillation constant, is given by the slope of the line joining the points determined experimentally. See fig. 2. The following distillation constants were determined :— TABLE V. Phenol, Wey . o-Cresol, : 4:2 m-Cresol, : Srl) ‘p-Cresol, 2:2 Comparative Distillations in Steam of Acids from Butter and other Products. As well as using the method of steam distillation for the analysis and detection of mixtures of lower fatty acids, it may be applied to detect the presence of acids of comparatively low molecular weight in the presence of higher fatty acids. Butter differs from most other fats in containing a relatively high per- centage of butyric acid. The distillation constant of the volatile fatty acids from butter will, therefore, show a lower distillation constant than those from other fats. On this fact a process has been worked out by which butter-fat may be distinguished from other fats. The Reichert-Meissl determination was carried out in the usual way, collecting 110 cc. of distillation. 100 c.c. of the filtered distillate were titrated with a standard barium hydroxide solution to determine the concentra- tion of acid in the distillate. The barium was precipitated by the addition of the calculated amount of sulphuric acid, and the solution made up to 200 ce. It was distilled in an ordinary round-bottomed flask, using a steam-jacket, and the distillate collected in 10-gram fractions, and titrated. The per- centage of acid distilling in each fraction was calculated on the acid present in 100 cc. of the original Reichert-Meissl distillate. : Remy anp Hickinsorrom— Method of Distillation in Steam. 141 The results obtained with typical fats of high Reichert-Meissl value are given in lable VI. [See also Fig. 3.] It can be seen that the values are sufficiently wide apart to differentiate between butter-fat and any of the fats employed in this investigation. 80 70 60 # 50 S = 40 ED a 230 x Ay 20 10 O Weight (in grams) of distillate. Fie. 3. TABLE VI. Percentage Acid Distilling in each Fraction. Wt. of distillate Butter Palm Kernel Babassu Cocoa Nut in grams. Fat. Oil. Fat. Oil. 10 1255 20 25 22 20 23:0 3d 42 38 30 35°5 49 57 53 40 42°5 60 70 65 50 50:0 68 80 75 The values for butter are the means of determinations on samples from three different sources. The babassu fat' was about fifteen months old, and gave the following values :— Melting-point, . : ees F eZ ORS) Todine value (Wijs), . : : é 5 55) Saponification value, . : : : . 248 Acid value, : ; ; 5 3 : 0-17 Reichert-M eissl, s P : : ; 5:8 1 We are indebted to Mr. G. van B. Gilmour for this sample of babassu fat. 142 Scientific Proceedings, Royal Dublin Society. Distillation by Blickfield’s process gave the following figures — Total volatile acids, . : 0 : eo Oiperscents Insoluble silver salts, : : : . 13°6 per cent. Soluble silver salts, .. 0 6 : ; 2-4 per cent. Melting-point of insoluble volatile acids, . 19:0° SUMMARY. The distillation constants of the fatty acids, alcohols, and phenols, soluble in water, bear a relation to the molecular structure of the substance. The distillation of solutious containing a volatile solute can be applied to detect changes in the state of molecular aggregation. The analysis of solutions of fatty acids by distillation may be extended to differentiate butter-fat from other fats containing a relatively high proportion of acids volatile in steam. 10. 11. 13. SCIENTIFIC PROCEEDINGS. VOLUME XVI. . A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, S$O.D., F.R.S., and T. G. Mason, u.a., sc.B. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. Smyru, B.a., sc.B. (Plates I., II.) (February, 1920.) 1s. . The Application of the Food-Unit Method to the Fattening of Cattle. By Jamms Witson, m.a., B.sc. (Plates III.,1V.) (February, 1920.) 1s. . The Holothurioidea of the Coasts of Ireland. By Anne L. Massy. (April, 1920.) 1s. . Photosynthesis and the Hlectronic Theory. By Hunry H. Drxon, so.d., F.2.s., and Horace H. Poots, sc.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Wituram Brown, B.sc. (March, 1920.) 6d. . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, u.a., sc.B. (April, 1920.) 1s. . The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Wini1au Brown, 8.sc., and Parrick O’CaLLaGHaN, A.R.C.SC.1.,4.1.c. (August, 1920.) 6d. . Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. LronarD, A.R.C.SC.1., B.SC., PH.D., and AGnies Browne, A.R.C.SC.I., B.sc. (August, 1920.) 6d. An Investigation into the Causes of the Self-Ignition of Ether-Air Mixtures. By the late Professor J. A. McCunuuanp, p.sc., F.z.s., and Rev. H. V. GILL, S.J., D.S.0., M.c., u.a., University College, Dublin: (August, 1920.) 6d. The Influence of Hlectrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosmpH ReiLLy, M.a., D.Sv., F.R.C.SC.1., and Witrrep J. Hicxinsorrom, B.sc. (October, 1920.) 6d. . Notes on some Applications of the Method of Distillation in Steam. By JosppH REILLY, M.A., D.SC., F.R.c.Sc.1,, and WitFRED J. HickINBOTTOM, B.SC. (October, 1920.) 6d. The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By W. H. Aprnry, D.sc., A.R.C.s0.1., F.1.c., and H. G. BECKER, A.R.C.S0.1., A.I.c. (September, 1920.) 6d. DUBLIN: PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIKBS. 7 eit i 4 ih h * ; ‘ ; : P B ; Pictvg A? Pa) 1 a oF ‘ THE SCIENTIFIC PROCKEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 13. SEPTEMBER, 1920. THE DETERMINATION OF THE RATE OF SOLUTION OF ATMOSPHERIC NITROGEN AND OXYGEN BY WATER. BY W. E. ADENEY, D.Sc. A.R.C.Sc1., FIC. ~~ ACTING PROFESSOR OF CHEMISTRY ; AND emithswntan insg ta) H. G. BECKER, A.B.C.Sc.L, A.L.C.,( MAR~5 q904_ TDEMONSTRATOR IN CHEMISTRY ; Sttenai mnsed® ROYAL COLLEGE OF SCIENCE FOR IRELAND. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1920. Price Sixpence. Roval WBublin Society. EOE FOUNDED, A.D. 1731. INCORPORATED, 1749. EVENING SCIENTIFIC MERTINGS. Tue Scientific Meetings of the Society are usually held at 4:15 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. 7 The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to- the Editor. f 14 J XIII. THE DETERMINATION OF THE RATE OF SOLUTION OF ATMOSPHERIC NITROGEN AND OXYGEN BY WATER. By W. E. ADENEY, D.Sc, A.R.C.Sc.1, F.1.C., Acting Professor of Chemistry ; AND H. G. BECKER, A.R.C.Sc.1, ALC, Demonstrator in Chemistry ; Royal College of Science for Ireland. [Read Apnit 27. Published Srpremuer 17, 1920. ] Part {11.—The Rate of Solution of Air by Quiescent Waters under Laboratory Conditions. IN previously published parts of this communication the rate of solution of gases by water, when thin films of the water are exposed to the gas or gases, and kept uniformly and rapidly mixed with the unexposed portions of it, has been dealt with, and shown to take place in accordance with a simple law. With a view to deriving, if possible, a formula for the rate of solution of air by quiescent bodies of water, the results thus obtained have been applied to the elucidation of a number of experiments which had been previously made with small volumes of still water, and the results are given in this communication. When the process of solution of air by water is considered, it is evident that it may take place in one of three ways—(1) Solution at the surface exposed to the air with thorough and rapid mixing with the unexposed portions ; (2) solution at the surface with no mixing; (3) solution at the surface with slow or imperfect. mixing. The conditions stated under section (1) are practically those obtaining in the experiments already described, and hence under these conditions the formula already deduced would apply. However, these conditions do not ordinarily occur. The conditions stated under section (2) are those which are commonly assumed to obtain, although there seems to be very little justification SOIENT, PROC. R.D.S,, VOL. XVI., NO. XIII. U 144 Scientific Proceedings, Royal Dublin Soctety. for such an assumption in practice. Obviously if there be no mixing of the water, the only process by which the air can penetrate into the mass of the water is by diffusion of the dissolved gas molecules; hence the ordinary law for the diffusion of solutes would hold under these con- ditions. But there is no experimental evidence to show that such conditions ever occur under natural conditions ; and it is doubtful if they have ever been produced artificially even in the laboratory. On the other hand, there is evidence to show that it 1s not possible to expose a mass of water to the air and still keep. the exposed and unexposed portions of the water unmixed. ‘hus, when Huefner was determining the velocity of diffusion of dissolved gases in water, he found it necessary to expose the columns of water to the gas at the lower surface through the medium of a porous plate of hydrophane, in order to avoid the mixing that he found to occur when the upper surface of the columns of water was exposed to the gas. A considerable number of experiments made by one of the authors also points to the fact that the dissolved gases do not accumulate in the upper layers of a column of water exposed to the air, as they would tend to do if there were no mixing. Hence calculations based on the law of diffusion cannot be of any practical value. The third set of conditions postulated, namely, solution at the surface with slow mixing, seems to be that which would occur most frequently in practice. Under these conditions the rate of solution will be dependent on the rate of mixing. The formula applicable in this case might be expected to be of the same form as that already derived froin the experiments with thin films, where the mixing was extremely rapid, but the constants will alter according to the rate of mixing. It will be shown that the results of the experiments quoted -here can be approximately represented by an equation of the required form, and that the constants vary with (1) the humidity of the air in contact with the water, and (2) the salinity of the water. The method of experimenting consisted in fillmg a number of tubes of about 30 cms. length and 4 cms. diameter with de-aerated tap-water, and inserting rubber-stoppers in such a way as to exclude all trace of air. The tubes were then placed in a thermostat until the water in them was at the same temperature as that of the thermostat. ‘he corks were then removed, 50 ec. water withdrawn, and the water remaining in the tubes exposed to the air for periods of different length, and the air-content determined at the end of the exposure. he initial air-content of the de-aerated water having been previously determined, the data for calculating the rate of solution was then available. Avrnny AND Bucxer—Solution of Nitrogen and Oxygen. The results of the experiments are given in 'ables I and II. Haperiments with columns of tap-water. TABLE I. Temperature, 15° C. Degree of Area Vol. of Time of Amount of air aeration in Value of ¢ 3’ exposed. water. exposure. absorbed. per cent. of const. saturation. hours 0 13:8 12°56 260 28°5 31-6 45°4 “019 12°56 260 48 55°38 69-1 “021 12°88 260 70 63°74 77:1 “019 12°88 260 98 71:6 83°6 “018 12°56 251 120 749°8 89'1 “018 12°56 260 141-5 79°7 92-1 017 Mean 12°62 258 TABLE II. Experiments with columns of sea-water, Temperature, 15° C. 12°56 12°26 12°88 12°88 12°88 12°25 12°88 12°56 12°88 Mean 12°67 21°5 24-0 24 44 72 96 96 118 168 13°7 42°5 58°3 56°71 77-0 86°5 90-0 91°8 96°1 99-3 145 | | | j | | When the amount of air dissolved is plotted against the time, the points u2 Percentage of saturation. Percentage of saturation. 100 146 Scientifie Proceedings, Royal Dublin Society. lie as shown in figs. 1 and 2, and it will be seen that they are in fair agree- ment with a logarithmic curve. Hence the results of these experiments can be dealt with in a similar way to that used in previous papers. 20 40 60 80 100 120 140 160 180 Time tn Hours. Fic. 1.—Results of experiments with tap-water. Equation of this curye:—w = (100 — w,) (1 — e> -9186¢). 20 40 60 80 100 120 140 160 180 Time 1n Hours. Fig. 2.—Results of experiments with sea-water. Equation of this curve :—w = (100 — wy) (1 — e- -0?%). Aprnry and Beckrr—Solution of Nitrogen and Oxygen. 147 The phenomenon has been shown to take place in accordance with tbe general equation dw A — = SAp - fa aan a where w = total quantity of gas in solution at any moment, S = initial rate of solution per unit area, . A = area of surface, p = pressure of the gas, J = coefficient of escape of the gas from the liquid per unit area and volume, V = volume of liquid. This can be written dw : A aE bw, where a = SAp and b =I that is, the values of the constants under the conditions of the experiment. The values of the 6 constant have been calculated for each experiment, and are given in the last column of Tables I and II. It will be seen that the experimental values obtained approximate to a constant; the deviations apparently indicate the magnitude of the experimental errors. The values of 6 for the curves shown in figs. 1 and 2 are :—- For tap-water 6 = ‘0186. For sea-water 6 = -0250. These values may be taken as the mean values of the 6 constant for the conditions of area and volume obtaining in the experiments. In order to compare the results of these experiments with those to be described it was necessary to reduce the constant to unit area and volume, and in order to do this, it was assumed that in this case, as in the case of the experiments described in previous communications, the rate of solution varied with the area exposed and inversely as the volume. : oV Since f= Si In the case of tap-water A was = 12°62 sq.cm. and V = 258cc.; hence /=°388. In the case of sea-water A = 12°67 sq.cm. and. V = 262cc.; hence f = 609. These values are an index of the rate of solution when a small body of partially de-aerated water is exposed to the air, in a quiescent condition, and kept at as uniform a temperature as possible. There was reason to think that the irregularity which was noticed in these experiments might have been due to changing atmospheric conditions, the value of / was calculated im each case. 148 Scientific Proceedings, Royal Dublin Society. particularly the humidity of the atmosphere. In order to verify this point a further experiment was made. The effect of the humidity of the air in contact with the water on the rate of solution is shown by the following experiment :— Four tubes, each 320 mm. long and 40 mm. diameter, were placed in a thermostat after being filled with de-aerated water. Three of these tubes, after 50 ce. water had been withdrawn from each, were then connected by means of rubber corks and glass tubing, so that a current of air could be drawn through the series. The fourth tube was left open to the atmo- sphere. Unfiltered air was drawn through the air-space of tube 1, and then through two U-tubes containing calcium chloride, from which the dried air passed into the air-space of tube 2, whence it passed directly into the air- space of tube 3. Thus the air in tube 2 was much drier than that in the atmosphere; and that in tube 3 was more moist, and probably nearly saturated with aqueous vapour. The results are given in the following table :— Tasse III. Experiment with air of different degrees of humidity. Temperature of Experiment, 15° C. No. of 3 Vol. of ‘Time of Degree of ; , 5 tube. area water. exposure. aeration. Velnooird. | Velma a 1 12°56 340 c.c. 43 hrs. 41:5 °/, 0125 “84 2 12°26 323 44 63:2 023 “61 3 12°26 327 44:5 31's “009 23 4 12°25 370 42 32°7 010 “29 It should be noted that tube 4 was exposed in the middle of the thermo- stat at atime when the humidity of the air was considerable. Hence the value f is rather low. pai On examining these results it will be noticed that the tube through which the dried air was drawn shows the largest proportion of dissolved air, although the time of exposure was practically the same in each case. In fact, the figures show that the water in this tube absorbed more than twice as much air as that in the tube through which the moist air was drawn. AbENEY AND Bucker—Solution of Nitrogen and Oxygen. 149 In order to reduce these experiments to a basis which will allow com- parison with the previous experiments, the values of / have been calculated in the usual way, assuming that the process of solution in this case is the same as before. These values of f are given in the last column of Table III, and on comparing them with the values already deduced for tap-water and sea-water some interesting relations become evident. Tt will be seen that the value of / for the tube through which undried air was drawn is approximately the same as that derived for tap-water from previous experiments, but the value for the tube with the dried air is very much greater, and that for the tube with the air saturated with aqueous vapour is very much less than the mean value. ‘This shows that when the air above the water is very dry the absorption of atmospheric gases takes place very much more rapidly than usual. On the other hand, when the air is saturated, or nearly saturated, with aqueous vapour the rate of solution is very greatly retarded. This would seem to indicate that the process by which the dissolved gas is carried down into the body of the liquid is affected by evaporation from the surface of the liquid, since the evaporation will be at a maximum when the air is dry, and a minimum when it is moist. In considering, therefore, this question of the mechanism of solution, there are two factors which, whatever others may affect it, are cbviously of primary importance, namely: (1) concentration of dissolved salts in the surface- layers due to evaporation ; (2) cooling of the surface-layers produced by evaporation. Both these factors are of such a nature as to result in an increase of density of the surface-layers, which would tend to set up vertical currents in the body of the water. In the case of waters containing large quantities of dissolved salts, the concentration of these salts by evaporation must play a very important part in the process of solution of air. In this connexion it is important to notice that the rate of solution of air by quiescent bodies of sea-water may be much more rapid than that by similar bodies of fresh water, as shown by the relative magnitudes of the values of f in each case. In the case of water containing only small quantities of salt in solution the effect of concentration would not be so great, but would probably be of somewhat the same magnitude as the effect produced by the cooling of the surface layers. ‘his cooling will. of course, result in an increased density of the surface layers, provided the temperature is not below 4°C., and consequently it will in most cases operate in such a way as to hasten the process of solution. ~ These two factors, tending as they do to set up a slow circulation from the 150 Scientific Proceedings, Royal Dublin Society. surface to the bottom of a mass of water, must be of much greater importance than diffusion, which is such an extremely slow process that months are required to detect its action even in the laboratory. This is further indicated by the fact that, to examine the effects of true diffusion experimentally, it would be necessary to take great care to maintain an absolutely uniform temperature, and to protect the column of water under observation from any external agency which might result in producing currents in the water. The experiments so far discussed have been made with very shallow depths of water not more than about 260mm. from the surface. It is, of course, necessary to extend them in order to investigate the question as to what depths mixing is appreciably induced by the conditions brought about by the evaporation from the exposed surfaces of quiescent waters, fresh or salt, under laboratory conditions. But such experiments would require observations to be carried on for very long periods of time, and very great care would be necessary to ensure uniformity in the conditions affecting evaporation during their continuance. The authors have not hitherto had time or opportunity of carrying out such experiments, but it is hoped to commence some experiments of this kind in this College next session. A number of preliminary experiments have been made by one of the authors with columns of water of from 18 to 24mm. cross-section! and of such length to allow of observations to be made to depths of about 1800 mm, The results of these experiments prove that mixing induced by evaporation from the exposed surfaces of columns of water does take place down to depths of at least 1800 mm., and that it occurs to a more decided extent in sea than in fresh water. Each of these experiments, however, extended over a considerable number of days, and no precautions were made to keep the conditions of evaporation from the exposed surfaces of the columns of water uniform. The results that were obtained from different experiments were consequently not sufficiently concurrent to determine whether they could be brought within the simple law found for the more shallow depths of water or not. The following experiment goes to show that the concentrated layers of salt solution, which result from evaporation at the exposed surface of a de-aerated column of sea-water, stream downwards, as they are formed, with little or no tendency to dissipation of their dissolved air-content in lateral directions. 'See ‘‘ Unrecognized Factors in the Transmission of Gases through Water.” By W, H. Adeney, Phil. Mag., March, 1905. AbgENrY AND Becker—Solution of Nitrogen and Oxygen. 151 For this experiment a double-bulbed tube of the form shown in fig. 3 was employed. The two bulbs A and & were of 175 ce. capacity each. They were connected by a narrow tube C, 6-4 mm. bore and 300 mm. long. The lower bulb 4 terminated in a small draw-off tube 6, and the upper bulb 4 was continued by the straight tube D, about 50 mm. long and 20 mm. bore. The whole tube was exhausted and filled with sea-water, the dissolved gases in which had been previously removed by boiling the water in vacuo, and extracting the gases by means of the mercury pump, described in Part 1 of this communication. The tube after being filled was securely closed from contact with the air, and placed in a thermostat, kept at the temperature of 12°7° to 13° C. When the contents of the tube had attained the tem- perature of the thermostat, the cork was removed, and a little water was drawn out so as to lower the level of the water in the tube to about 20 mm. below the mouth of VD. The tube was. then fitted with a cork, furnished with outlet and inlet tubes, so that a small air space was left above the water in D, through which a current of air could be drawn without disturbing the surface of the water. Two weighed calcium chloride tubes were attached to the outlet tube, to absorb the water-vapour brought over trom the water in the bulb with the ai-current, and to afford the means of determining the weight of water evaporated during the experiment. A current of air, previously dried by passing through calcium chloride tubes, was drawn through the air-space in the tube for six days. The dissolved air-content of the water in the upper and lower bulbs was then separately determined, with the following results :— Gases expressed in cc. at N.T.P. per 1,000 ce. sea-water. Upper Bulb. Lower Bulb. Oxygen, 340 2°80 Nitrogen, 6:83 543 The initial air-content of the water in the two bulbs may be taken as practically 0. Sea-water when saturated with air at 13°C. contains per litre :-— Oxygen, 6:06 ce. Nitrogen, 11°77 cc. SCIENT. PROC. R.D.S., VOL. XVI., NO. XIII. x 152 Scientific Proceedings, Royal Dublin Society. During the experiment 7:935 grams water were evaporated. It will be seen from the above figures that, while the water in the upper bulb was only a little more than half saturated, that in the lower bulb was also very nearly half saturated. That is, the air-content of the water was practically uniform, although the surface layer was not saturated. And the figures consequently prove that, while the dissolved gases were freely drawn down the connecting tube into the lower bulb by gravitation, they showed no tendency to spread in lateral directions in the upper bulb during their downward passage through it." The authors desire again to express their indebtedness to Dr. Hackett for the interest that he has continued to take in their investigation of the subject of this and previous communications. ‘See Phil. Mag., March, 1905. 10. 11. 12. 13. SCIENTIFIC PROCEEDINGS. VOLUME XVI. A Cryoscopie Method for the Estimation of Sucrose. By Henry H. Dixon, $0.D., F.R.s., and T. G. Mason, m.a., so.p. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. Smyru, B.a., sc.s. (Plates I., II.) (February, 1920.) 1s. . The Application of the Food-Unit Method to the Fattening of Cattle. By James Wison, m.a., B.sc. (Plates III.,1V.) (February, 1920.) 1s. . The Holothurioidea of the Coasts of Ireland. By Anns L. Massy. (April, 1920.) 1s. . Photosynthesis and the Hlectronie Theory. By Henry H. Dixon, sc.D., F.2.S., and Horace H. Poouz, sc.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Witu1am Brown, B.sc. (March, 1920.) 6d. . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, m.a., sc.B. (April, 1920.) 1s. . The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Witiiam Brown, B.so., and Parrick O’CaLLaGHaN, A.R.C.80.1.,A.1.c. (August, 1920.) 6d. . Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. Lronarp, A.R.c.SC.1., B.SC., PH.D., and AGNES Browne, A.8.0.SC.1., B.sc. (August, 1920.) 6d. An Investigation into the Causes of the Seli-Ignition of Hther-Air Mixtures. By the late Professor J. A. McCuenuanp, p.sc., F.R.s., and Rey. H. V. GILL, 8.J., D.S.0., M.c., u.A., University College, Dublin. (August, 1920.) 6d. The Influence of Electrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosrpH Reinty, m.a., D.sv., F.R.C.Sc.1., and Witrrep J. Hickinsorrom, B.sc. (September, 1920.) 6d. Notes on some Applications of the Method of Distillation in Steam. By JosepH REILLY, M.a., D.SC., F.R.c.Sc.1., and WitrrEeD J. Hickinsortom, B.Sc. (September, 1920.) 6d. The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By W. H. ApENEY, D.sc., A.R.c.sc.1., F.1.c., and H. G. BEOKER, A.B.¢.SC.1., A.1.c. (September, 1920.) 6d. DUBLIN : PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIHBS. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 14. DECEMBER, 1920. A DETERMINATION, BY MEANS OF A DIFFERENTIAL CALORIMETER, OF THE HEAT PRODUCED DURING THE INVER- SION OF SUCROSE. BY HENRY H. DIXON, Sc.D., F.B.S., Pak Ayla 1 UNIVERSITY PROFESSOR OF BOTANY IN TRINITY COLLEGE, DUBLIN ; ES enn AND Ne fv h r ) c NIGEL G. BALL, B.A. op (Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVEN GARDEN, LONDON, W.C. 2. 1920. Price Sixpence. Koval PMublin Society FOUNDED, A.D. 1781. INCORPORATED, 1749. OEE RVENING SCIENTIFIC MERTINGS. Tar Scientific Meetings of the Society are usually held at 4:15 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society,. and such ag are considered suitable for the purpose will be printed with. the least possible delay. Authors are requested to land in their MS. and! necessary Illustrations in a complete form, and ready for transmission to: the [ditor. ( 168 ) aE A DETERMINATION, BY MEANS OF A DIFFERENTIAL CALORI- METER, OF THE HEAT PRODUCED DURING THE INVERSION OF SUCROSE. By HENRY H. DIXON, Sc.D., E-B:S., University Professor of Botany in Trinity College, Dublin, AND NIGEL G. BALL, B.A. [Read Novemprr 23; published DecemBer 14, 1920.] Very little attention seems to have been paid to the heat changes accompanying enzyme reactions. Theoretically these can be deduced from the difference in the heats of combustion of the initial substances and the products of the feaction; but, as Brown and Pickering (1) have pointed out, such methods are inaccurate, as the amounts of the thermal changes to be measured are so close to the limits of experimental error for the combustion values. For example, in the case of inversion of sucrose, using the values for the heats of combustion obtained by Stohmann and Langbein (2), we get— — ie Aa Sucrose : : : . 1352°7 cal. per gm. mol. Dextrose TT xtrose wie ‘) 1349-6 : Laevulose . , O7159 Difference 3:1 cal. per gm. mol. From this it would appear that the heat of inversion of sucrose in the solid state is + 3:1 cal. per gm. mol., or + 9:1 small calories per gram inverted. This value represents only about 0:1 per cent. of the sum of the quantities measured in order to determine it, and little reliance can be placed upon its accuracy. In order to determine the heat of inversion in the dissolved condition, this value must be corrected for the differences in the heats of solution of the substances concerned, and, when this is done !Mr. Ball is indebted to the Department of Scientific and Industrial Research for a maintenance grant which was received during the progress of this research. SCIENT, PROC. R.D.S., VOL. XVI., NO. XIV. Y 154 Scientific Proceedings, Royal Dublin Society. the value + 9:1 cal. becomes, according to Brown and Pickering, - 1:0 cal. per gram of sucrose inverted. By direct experiment Brown and Pickering obtained the value + 11:2 cal. per gram inverted. In view of the difference between this result and that deduced from the heats of combustion, it seemed to us desirable to make a fresh determination of this value, using a direct method. ‘The energy changes due to enzyme action are of interest in connexion with the study of the chemical changes which take place in the plant as a result of photosynthesis. In the experiments of Brown and Pickering ordinary calorimetric methods were employed, and the temperature was measured by means of a sensitive mercury thermometer. As the temperature changes which have to be measured are very small and take place with comparative slowness, we decided that a differential method, using vacuum flasks and a sensitive thermocouple, constructed on the lines of those employed in cryoscopic measurements (Dixon (3)), would be suitable for the purpose. After our apparatus lad been set up and used for some preliminary experiments, we became acquainted with the work of A. V. Hill (4), in which a somewhat similar type of apparatus was used in certain physiological experiments. In his paper is given a clear discussion from the mathematical standpoint of the characteristics and advantages of the differential method, which it is unnecessary to repeat here. ‘The essential features of the method as employed by us to determine the thermal changes due to enzyme action will. however, be briefly explained, followed by a description of our apparatus. Two similar vacuum flasks are used, each containing one junction of a thermocouple, which is connected through a reversing switch with a sensitive galvanometer. The deflection of the galvanometer is directly proportional to the difference in temperature of the two flasks. If each flask contains the same quantity of liquid and has the same coefficient of conductivity, the changes in external temperature will affect them both to the same extent, and the differential temperature will remain the same. If heat is produced in one flask, the increase in the galvanometer deflections will be divectly proportional to the rise in temperature, and the radiation correction will be proportional to the galvanometer deflections, and will be unaffected by changes in the temperature of the surroundings. Changes in temperature due to dilution of the sugar solution by the enzyme solution in the experimental flask can be neutralized by allowing a similar dilution (using an enzyme solution which had previously been heated to 100°C.) to take place in the control flask, Drxon anp Batu— Heat produced during Inversion of Sucrose. 105 Construction of Thermocouple. ‘he thermocouple employed was constructed of copper and eureka wires, according to the method previously described (3), with the exception that the wires leading to the junctions passed through narrow glass tubes instead of being fastened to pine rods. The glass tubes used were 25 cm. long, and the wires, which were well coated with collodion varnish, projected about 1 em. from the lower ends. The junctions were protected from injury by being attached to little pieces of wood which were pushed into the tubes. The remaining spaces in the tubes were filled with paraffin-wax, which effectually sealed them and kept the wires in place. The copper wires of the thermocouple then passed to the reversing switch of the type already described (3). This switch is preferable to a mercury one, owing to complete absence of thermoelectric effects. The galvanometer used was of the Ayrton-Mather pattern, giving for one micro-volt a deflection of the spot of light on the screen one metre distant from the mirror, of 10 mm., and for one micro-ampere 206mm. By means of the reversing switch, readings were always taken on both sides of the middle of the scale, and subtracted from one another to obtain the deflection. This ensures that errors due to movement of the zero or to thermoelectric effects at the junctions of the leads with the galvanometer terminals, are completely eliminated. Calibration. The thermocouple was calibrated by using it for the determination of the freezing-point of a solution of sucrose of definite concentration, according to the method previously described (3). As the depression of freezing-point of solutions of sucrose of different concentrations are known from the researches of Raoult (5), this affords an easy and accurate method of calibration. The thermocouple used was found to give a deflection on the scale of 1 mm. per 0:00258° C. Method of Releasing Enzyme Solution. It is obviously essential that the enzyme solution, at the moment of mixing, be at exactly the same temperature as the solution on which it is to act. This can only be ensured by having it in a container immersed in the solution for some time before the commencement of the experiment. It is also necessary that the mixing be performed without opening the flasks, which might cause a disturbance in temperature of the liquids contained in them. In Brown and Pickering’s experiments (1) the enzyme solution was contained in a glass pipette, the end of which was bent upwards and then y2 156 Scientific Proceedings, hoyal Dublin Society. downwards. This was immersed in the solution in the calorimeter. By means of a combined force and suction-pump, the enzyme solution was forced out of the pipette, and then the pipette was washed out by drawing in some of the mixed liquid and again expelling it. As this method was \ \ Fie. 1. unsuitable for use with our apparatus, owing to the narrow necks of the vacuum flasks, we decided to enclose the enzyme solution in a capsule of paraffined paper, attached to one of the glass tubes of the thermocouple, a similar arrangement being used in the control flask (see fig. 1), These Drxon anp Bati—Heat produced during Inversion of Sucrose. 157 capsules, which were found to be quite satisfactory, were constructed in the following manner :— About 1 em. from the lower ends of the glass tubes containing the wires of the thermocouple little copper bands were cemented to the glass with Canada balsam, in order to support the lower ends of the capsules. The upper and lower ends of the capsules were formed of discs of cork 2 cm. in diameter. The lower ones were perforated by holes, into which the glass tubes fitted tightly. The upper ones had holes sufficiently large to allow the copper bands to pass through, and to these discs silk threads were attached, for the purpose of rupturing the capsules. Cylinders, 6 cm. long, which fitted tightly on the cork dises, were made of a double thickness of paper, with the edges cemented with seccotine. The upper ends of the paper cylinders were notched and turned over on the upper corks, leaving the lower ends open. The cylinders were then immeised in melted paraffin- wax, and the beaker which contained it was placed under the receiver of an air-pump, which was then exhausted. This removes all air-bubbles from the paper, and ensures that it is thoroughly impregnated with wax. ‘The capsules were then removed, and allowed to cool. In order to assemble the capsules on the tubes of the thermocouple, the tubes were passed through the holes in the upper corks, which were now firmly attached to the paraffined paper. ‘I'he lower corks were then pushed on to the tubes until they came into contact with the copper bands, and the edges of the discs were smeared with vaseline. The open ends of the capsules were then pushed over the lower cork discs until the ends of the paper capsules projected about 0-9 em, The whole thermocouple was then inverted, and the cup- shaped space left below the lower cork was filled with melted paraftin-wax of a low melting point. ‘Vhis effectually seals the lower ends of the capsules. For the purpose of filling the capsules, there is sufficient space between the glass tube and the edge of the hole in the upper cork to insert the end of a small pipette. The capsules used held 15 c.c. of liquid, and, after filling, the hole at the top round the glass tube was sealed with paraffin-wax. If carefully made, these capsules give no trouble due to leakage, and are easily opened by pulling the silk thread. This separates the paper cylinder from the lower cork, which is held by the copper band, and allows the enzyme solution to mingle rapidly with the surrounding liquid. Flasks. The flasks used were the ordinary so-called pint-size vacuum flasks, which are made of silvered glass, enclosed in a metal case. The conductivity constants of the two flasks were found to differ by about 9 per cent. Unless 158 Scientific Proceedings, Royal Dublin Society: the two flasks change in temperature at the same rate, owing to alteration in temperature of the surroundings, errors will be introduced should the external temperature change during the course of an experiment. In our later experiments on the inversion of sucrose the external temperature remained constant within about 01° C., so that the error introduced was negligible. But owing to the sensitiveness of the temperature measurements, it was found that slight inequalities in the radiation to which the two flasks were exposed gave rise to considerable errors. To avoid this, the apparatus was enclosed in a heavy copper cylinder, covered with a copper lid. This was placed in a wooden cask, and packed around with cork dust. The water-equivalent of the experimental flask was found to be 9°5 grams. Radiation Correction. The radiation correction was found to be affected very largely by the method adopted for stirring the flasks. In our earlier experiments the glass tubes of the thermocouple, with the capsules attached, passed loosely through holes in the corks which closed the necks of the flasks, and were attached to a cross-piece which was free to slide on a vertical rod between the two flasks. A light piece of cane was attached to the cross-piece, and passed through a hole in the copper lid of the containing vessel. By ‘this means the thermocouple and capsules could be moved up and down, thus stirring the liquid in the flasks. This method of stirring suffers from the disadvantage that the radiation error is very greatly increased owing to loss of heat from the wet surfaces of the tubes when they come out of the flasks, and was therefore abandoned. The method finally adopted was to clamp the glass tubes between split corks in the necks of the flasks, and to effect the stirring by shaking the flasks. The cask containing the apparatus was slung in a horizontal position from a bar of wood which rested on a central pivot, and could thus be rocked backwards and forwards. By this means the rate of cooling during stirring was decreased to less than one-third of its former value. The correction for radiation was then determined. In a preliminary experiment 315 c.c. of water, at about 90°C., was put into each flask, and a sensitive thermometer inserted through the cork which closed the neck. The flasks were then placed in a shaking-machine, and the rate of cooling determined. Assuming Newton’s Law of Cooling, the rate of cooling at any temperature = the excess of that temperature over the surroundings x a constant. This constant, which we may call %, was calculated from the formula :— il Lelio -k loge = 7 leg (4): Drxon anp Bati— Heat produced during Inversion of Sucrose. 159 where A is the initial temperature, 7’ the final temperature, 7) the temperature of the surrounding air, and ¢ the time in hours during which the temperature falls from A to 7. It was found that the rate of cooling of one flask was slightly greater than that of the other, but, as Hill (4) has pointed out, compensation can be made for this by adjusting the amounts of liquid in the two flasks. By putting 315 ¢.c. of water in the experimental flask, and 287 e.c. in the control flask, it was found that the rate of cooling was approximately the same in both. A more accurate determination of the value of k was made in the following manner. The apparatus was set up with the thermocouple in the flasks, the water in the experimental flask being about 1° C. warmer than that in the control flask, and the amounts of liquid in the flasks adjusted so as to make their conductivity constants approximately equal, i.e. 515 c.c. in experimental flask and 287 e.c. in the control. The rate of cooling was determined by noting the decrease in the galvanometer deflections during different periods of more than one hour while stirring was carried on, and the constant / was calculated from the formula. In this case the rate of cooling at any instant is proportional to the deflection of the galvanometer, and therefore A and 7’ are equal to the initial and final deflections respectively, and 7’, is equal to O. The value of 0°05 for & was obtained as the mean of a number of concordant observations. This means that at any temperature the rate of cooling per hour was 5 per cent. of that temperature. The smallness of the correction for radiation losses clearly demonstrates the value of this apparatus for the determination of the amount of heat produced during a chemical reaction, when the time of the reaction is necessarily prolonged. Determination of the Heat Produced during the Inversion of Sucrose. The sucrose solution which was used contained 100 gm. per litre of solution. Both experimental and control flasks were charged with 800 ¢.c. of this solution. As the temperature during the experiment could be kept constant, the same quantity of liquid was used in each flask. The solution of invertase was prepared according to Davis’ (6) method. In one capsule was put 15 ¢.c. of the enzyme solution, and-in the other 15 c.c. of the same solution which had been previously heated to 100°C. After the capsules had been sealed, the thermocouple was inserted into the flasks, which were then corked with the split corks. The silk threads passed out through grooves between the two halves of the corks. If the corks and threads are well greased, no leakage from the flasks occurs. The diagram (fig. 1) shows the arrangement of the thermocouple and flasks when assembled. Galvanometer Deflections. 160 Screntifie Proceedings, Royal Dublin Society. The apparatus was then set up and left overnight in order that the temperature inside the capsules might become the same as that of the surrounding liquid. The next morning the deflections of the galvanometer were observed at intervals, and, if constant, the capsules were ruptured and the experiment started. The fact that the deflections are constant before the commencement of the experiment is a proof that no leakage of the enzyme solution has taken place. The deflections were observed at regular intervals while the apparatus was continually shaken. When it was desired to stop the 180 | Iso} (ove 140 I pele Leelee (93% of +— oe Loa ea ENZYME| ADDED AL ‘lime in minutes. Fic. 2. inversion, the experimental flask was rapidly opened, and 10 e.c. of ammonium hydrate run in. The exact time at which this took place was noted. The amount of sucrose inverted was determined by measuring the reducing power of the solution according to Pavy’s method. A curve showing the rise in temperature was plotted, and by extrapolating this curve the exact temperature at the moment at which inversion was stopped can be deduced. From this curve another curve corrected for radiation losses was drawn, according to the method described by Hill (4). The curve obtained in Exp. No. 7 is reproduced in fig. 2. It will be 0 10 20 30 40 50 60 70 80 930 100 0 ‘120 Dixon and Batit—Heat produced during Inversion of Sucrose. 161 noticed that the temperature rises steadily from the moment mixing takes place, and after the correction is made for radiation losses the curve is very nearly a straight line throughout the course of the experiment. This is probably due to the fact that the sucrose is largely in excess of the enzyme, so that the rate of the reaction is not sensibly diminished even after about 40 per cent. of the sucrose has been hydrolysed. In calculating the results, a correction is applied for the difference in the specific heat of the sucrose solution used from that of water, and also a _ correction for the very slight reducing power of the sucrose and invertase solutions. The results of our two final experiments are shown below. ‘he tempera- ture of the solution during both these experiments was about 10° C. The values obtained in earlier experiments are not given, as, owing to the method of stirring adopted, the radiation error was very large and uncertain, and therefore the results were unreliable. | | iF ° b er age | is Ag Gm. cal. No. of Time of oor a cau ae aise Radiation a cet mene Expt: Sucrose | of Sucrose | in Gametnen per gram De 3 inverted. inverted. |'lemperature. ° inverted. | | 6 95:5 mins. | 12°98 gm. |43°3 percent.| 0°4383°C. 0:0167°C. 11°20 7 97 mins. 12°62 gm. |42-lpercent.| 0°4274°C. OMIT}, 11:27 | | | The radiation correction in Expt. 6 is slightly smaller than in Expt. 7, although in the former case the rise in temperature was greater than in the latter. This is due to the fact that, at the commencement of the experiments, the experimental flask was slightly cooler than the control in Expt. 6, and slightly warmer in Expt. 7. The values, 11:20 and 11:27, obtained in our experiments agree well with the values, 11°28 and 11:13, given in Brown and Pickering’s paper, and the mean of all these, 11°22, may be regarded as a reasonably accurate estimate of the number of calories evolved per gram of sucrose when inverted in solution by means of invertase. ‘his is equal to 3°83 calories per gram molecule. Since these experiments have been concluded, the results obtained by Barry (7 and 8) on the heat of inversion of sucrose by hydrochloric acid have been published. In this case a value of + 10-4 cal. per gram inverted at 20°C. was obtained. It seems possible that the discrepancy between this value and that obtained in our experiments is due to the fact that the energy-changes SCIENT. PROC. I..D.S., VOL. XVI., NO. XIV. Z 162 Scientific Proceedings, Royal Dublin Society. in the acid solution are different from those in a neutral solution. Barry himself has shown that the heat of solution of sucrose in acid is different from that in water. The Hydrolysis of Maltose. We also made some attempts to obtain a value for the thermal changes. which take place when maltose is hydrolysed to dextrose under the influence: of maltase. According to Stohmann and Langbein (2), the heat of combus- tion of maltose hydrate is 1339-8 cal. per gm. mol., and for two gm. mol. of dextrose it is 1347-4. This would give a value of — 7°6 cal. per gm. mol. for the hydrolysis of maltose, but such indirect determinations ave of little use. The direct determination of this value is attended by many difficulties. At air temperature maltase acts very slowly, and even at the optimum temperature, about 38°C., the action is still rather slow. A 5 per cent. solution of maltose when acted on overnight by one-twentieth of its volume of extract of dried brewers’ yeast at 38°C. was only hydrolysed to the extent of about 42 per cent. An attempt was made to obtain a value for the thermal changes accompanying hydrolysis by filling the flasks with a 5 per cent. solution of maltose at 38° C., and using an extract of dried yeast as a source of maltase. A curve showing the change in the galvanometer deflections during 1} hours was plotted, and then the capsules were ruptured. Any thermal changes produced by the action of the enzyme on the maltose should cause a. change in the direction of the curve. No change was noted, but after 15 hours the amount of hydrolysis, when determined by the polarimeter, was so small as to be within the limits of experimental error, and so no conclusions could be drawn. In order to obtain accurate results it would be necessary to obtain a much more active preparation of maltase, and to have a pair of flasks so carefully adjusted that, when filled with a sugar solution at 38°C., the temperature of the two fell at exactly the same rate. This would ensure that any change in the deflections of the galvanometer was entirely due to the action of the enzyme. Drxon ano Batt—Heat produced during Inversion of Suerose. 163 SUMMARY. Estimations of the thermal changes due to the action of enzymes when obtained from the heats of combustion are inaccurate, as the values obtained are so close to the limits of experimental error for the heats of combustion values. A differential calorimeter, in which the temperature is measured by means of a thermocouple, was used for the determination of the heat produced by the action of invertase on sucrose. A value was obtained which agrees closely with that previously given by Brown and Pickering. The mean of all the results is 3°83 cal. per gm. mol. Attempts to obtain a value for the heat of reaction during the hydrolysis of maltose have not yet been successful. REFERENCES. (1). Brown, H. T.,and Pickerine, 8. V. Thermochemistry of Carbohydrate Hydrolysis. Trans. Chem. Soc., 1897, vol. lxxi, p. 783. (2). STOHMANN and LANGBEIN. Jour. Pract. Chem. (2), xlviii, p. 305. (3). Drxon, H.H. A Thermo-electric Method of Cryoscopy. Proe. Roy. Dub. Soe., 1911, vol. xii, p. 49. (4). Hint, A.V. A New Form of Differential Micro-calorimeter for the Estimation of Heat Production in Physiological, Bacteriological, or Ferment Actions. Jour. Physiol., 1911, No. 43, p. 216. (5). Raoutt, F.-M. Compt. Rend., 1897, Tom. exxy, p. 751. (6). Davis, W. A. ‘The Use of Enzymes and Special Yeasts in Carbo- hydrate Analysis. Jour. Soc. Chem. Indus., Web. 29, 1916, vol. xxxy, p. 201. 164 Scientific Proceedings, Royal Dublin Society. (7). Barry, F. A Calorimetric Procedure for Determining the Heats of Slow Reactions. Jour. Amer. Chem. Soe., 1920, vol. xli, p. 1295. (8). ———— A Calorimetrie Procedure for Determining the Heats of Slow Reactions, II. The Calorimetry of a Slow Reaction. The Heat of Inversion of Sucrose by Acid. Jour. Amer. Chem. Soe., 1920, vol. xlii, p. 1911. 10. 11. 13. 14 SCIENTIFIC PROCEEDINGS. VOLUME XVI. . A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, So.D., F.R.s., and T. G. Mason, m.a., so.z. (January, 1920.) 6d. The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. Smytu, B.a., sce. (Plates I., 11.) (February, 1920.) 1s. . The Application of the Food-Unit Method to the Fattening of Cattle. By James Winson, m.a., B.Sc. (Plates III., 1V.) (February, 1920.) 1s. The Holothurioidea of the Coasts of Ireland. By Annm L. Massy. (April, 1920.) 1s. . Photosynthesis and the Electronic Theory. By Hunry H. Dixon, sc.p., F.R.s., and Horace H. Pootn, sc.p. (March, 1920.) 1s. Note on the Decay of Magnetism in Bar Magnets. By Wixu1am Brown, B.sc. (March, 1920.) 6d. . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, m.a., sc.p. (April, 1920.) 1s. . The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Win1iau Brown, B.sc., and Patrick O’CaLLAaGHAN, A.R.C.SC.1.,4.1.c. (August, 1920.) 6d. Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. LmonarD, 4.8.¢.8¢.1., B.sc., PH.D., and AcGnes Browne, A.R.C.SC.1., B.sc. (August, 1920.) 6d. An Investigation into the Causes of the Self-Ienition of Hther-Air Mixtures. By the late Professor J. A. McCuntuanp, p.sc., F.R.s., and Rey. H. V. GILL, S.J., D.S.0., m.c., m.A., University College, Dublin. (August, 1920.) 6d. The Influence of Electrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosnpnH Reiwiy, m.a., D.sc., ¥.R.C.Sc.1., and Witrrep J. Hrcximgortom, B.sc. (October, 1920.) 6d. . Notes on some Applications of the Method of Distillation in Steam. By JosepH ReiILLy, M.A., D.Sc., F.R.c.sc.1., and Witrrep J. Hickinsorrom, B.sc.. (October, 1920.) Gd. The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By W. HE. Apmrnny, p.sc., a.R.c.sc.1., F.1.c., and H. G. BECKER, A.B.¢.8¢.1., A.t.c. (September, 1920.) 6d. . A Determimation, by means of a Differential Calorimeter, of the Heat produced during the Inversion of Sucrose. By Huyry H. Dixon, sc.p., F.R.S., and Niczn G. Bann, p.a. (December, 1920.) 6d. DUBLIN: PRINTED AT THE UNIVERSITY PRESS BY VONSONKY AND GIKBS, ‘s ne HH THE SCIENTIFIC PROCKEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (.8.), No. 10. FEBRUARY, 1921. THE MEASUREMENT OF VERY SHORT TIME INTERVALS BY THE CONDENSER-CHARGING METHOD. BY Se JOHN J. DOWLING, M.A., Finest’ .2! | AND fe DONAL DONNELLY, M.Sc. (In conjunction with the late Pror. J. A. McCuruiayp, F'.R.S.) [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1921. Price Sixpence. “Wie ue 1AYOR ; ‘ ae 0 *) Beh ne! Rea a (Goel XV. THE MEASUREMENT OF VERY SHORT TIME INTERVALS BY THE- CONDENSER-CHARGING METHOD. By JOHN J. DOWLING, M.A., F. Lysr. P.; Payson iInstien AND If&S DONAL DONNELLY, M.Sc. Px MAR 2] (In conjunction with the late Pror. J. A. McCrsttanp, £.R,S.)! 74] ( [Read December 21, 1920. Published Frpruary 21, 1921.] SOMEWHAT similar methods have been employed by other workers, and our aim has been to examine what degree of accuracy the method was capable of when applied to measure the shortest time-interval we could reproduce with sufficient accuracy. It was, furthermore, desirable that the apparatus should be kept as simple as possible, in view of the possible application of the method in the workshop or testing room. We had intended to test the application of the method to the measurement of the velocity of explosion in explosives, but this part of the investigation was not carried out. References to previous work have been fully given by Klopsteg? in a recent paper, and need not be repeated here; while the results of a similar investigation to ours were published by Webster,’ and came to our notice in the last issue of Science Abstracts. The particular method employed is based on the charging of a condenser from a battery of steady e.m.f. through a non-inductive resistance. The charge g acquired during ¢ seconds is given by :— t a= 0(1 - oft) : : : : 3 (1) or, t = 2:3 RC[log gq, - log (gq -9)] - ; : (2) or, approximately, C= ae 3 ; i ‘ : : : (3) 0 1 The problem was submitted to us by Messrs. Nobel’s Explosives Co., and this paper is published with their permission. We desire to express our thanks to the Company and Mr. Rintoul for their courtesy. 2 Klopsteg : Physical Review, xv, 7,12. 1920. 3 Webster and Allen: Am. Phil. Soc. Proc. 58, No. 6, p. 382. 1919. SCIENT. PROC. R.D.S., VOL. XVI, NO. XV. 2a 166 Scientific Proceedings, Royal Dublin Society. To measure times of the order of 100 micro-seconds (100 x 10° sec.), if q is to be ten times g, we see, from (3), that the product RC should be about +755. Having a very well insulated mica condenser of one microfarad capacity, we made £ 1000 ohms in most of our observations; but, with a view to testing the method somewhat, we varied RF in a few cases. The actual values of R and C required are to some extent determined by the sensitivity of the galvano- meter, but the battery em.f. can also be modified so as to give g and q suitable values. To measure g, and g a ballistic galvanometer with a universal shunt was employed, so that large, and nearly equal, deflections were obtained for both g) and g. An eight-volt battery was employed throughout, and the galvanometer had a ballistic sensitivity of about 450 scale divisions per micro-coulomb. Fie. 1. The main difficulty to be overcome in practice is the means to be adopted by which the condenser is caused to commence and cease charging at the beginning and termination, respectively, of the time interval to be measured. In order that accidental and constant errors should be reduced to a minimum, it is necessary to have the commencement and termination of the charging brought about by two operations as similar as possible. We attempted to do this by arranging that these operations should each consist in the breaking of fine copper wires, in which practically the same currents would be passing. These are shown in the diagram (fig. 1) by the dotted lines. The wires were soldered tu the brass plates (A, B, D), and care was taken to have them under Dow ine anp Donnetty— Measurement of Short Time Intervals. 167 practically the same degree of tension before soldering. These were broken in turn by a small (:22) rifle bullet travelling in the direction of the arrow, and the velocity of this bullet was measured immediately afterwards by an independent method. Check observations showed that the wires produced no appreciable retardation of the bullet. In the experiments the wires were never mounted closer than 2 ems. apart, so that the slightly different lateral displacements which occur before breaking should not seriously affect the result. The bullet velocity was determined as follows:—Two circular discs of cardboard, about two feet in diameter, were mounted about two feet apart, one at either end of a stout shaft, turning between centres, and driven by a powerful electric motor at a speed of about 35 revolutions per second. To these discs somewhat larger thin paper discs were pasted, and the whole apparatus mounted so that the bullet, travelling parallel to the axle, would pierce the two paper discs in succession near their edges. A resistance in the motor circuit allowed the speed of the discs to be adjusted to a constant steady rate indicated by a stroboscope arrangement. The shot was then fired, and from a measurement of the angular displacement of the bullet holes, relative to each other, the bullet velocity could be determined. Certain obvious check experiments were carried out to determine whether, for instance, the bullet was appreciably retarded or deflected sideways during its passage through the first disc. The angle to be measured was generally about 30°, and an error of more than one or two per cent. was unlikely. The apparatus was connected up as shown by the full dimes in the diagram. (The dotted lines P, Q, 7 refer to a modification referred to later.) Observations were conducted as follows :—Having measured the separation of the wires in position, and both keys K,, K, being open, the motor speed was regulated until the standard speed of rotation of the discs was reached. The gun was imme- diately fired, breaking the wires 4 and B. The shunt being set to unity, A, was closed and the deflection observed for g. To obtain the reading for q,, kK, and K, were then operated in succession, the shunt having been set back to‘10. The observations of the angular displacement of the bullet holes in the discs were finally taken by means of a protractor attached to the axle, the punctures being sighted in turn through the rifle barrel. The following table contains the results of twelve consecutive readings obtained in this manner. In all cases rate of rotation is 34:5 revolutions per second, and distance apart of paper discs 54:8 cms. 168 Scientific Proceedings, Royal Dublin Society. TABLE I, n Ss } os a Z ea | = a Ieee x T= a5, =) =e nano fe) a eee Be |e gq | 38 3 a ree o Soy 3 c q Ae = a) 338 © Sac a SI LL aS g f] 5 2 Bu z 3° 2 BI io || os Bi Q5 ox 3 53 x & S Poe Ba x ‘dsl 2 8 z i lee 8 S sso | i; IT 3 “2 | 28 cerns aretceaille Selcamlly (ae aq |822 | a8 |Sles Og q = 1S) s§ ; 2 Ss ae | Be Pee | Re Pee | ee eka ie yal & 4 > =) = A A °4 a S S 1 2 3 4 5 6 7 8 9 10 98° 249 1:966 80:9 3630 257 | 1000 79:0 | +1:9 2-4 974° 249 1:910 | 76:8 3630 261 * 74:5 | 42:3 3+1 29° 234 1:900 | 77:8 3650 264 3 754 || 4 24 3:9 30° 226 1:762 | 77:8 3650 264 au 75-4 | + 2:4 3-9 29° 234 | 1:894 | 80-9 | 3650 515 Y, HO || eee Il Gee 27° 252 | 1:980 | 786 | 3650 144 | 2000 80-5 | -1:9 | 2-4 281° 238 | 1:949 | 81:6 | 3620 149 fa 84:0 | —2:4 | 2-9 28° 242 | 1-334 | 75:5 | 3620 132 i m2 | 41-3 | 17 284° 238 | 1:816 | 761 | 3620 132 ie 74-2). || aa-g. Ih cose 273° 249 | 1:872 | 75:0 | 3600 129 - 729 | +21 | 2:8 26° 261 | 1:882 | 72:0 | 3600 127 a m7 | 203 | 0-4 Some further experiments were carried out with a view to reducing the time interval to a still smaller value. It is not possible to do this simply by mounting the wires closer together in the arrangement just described. It would seem that a great part of the percentage error of the above results is due to unequal stretching of the wires before breaking, and, of course, this effect becomes more and more important as the separation of the wires is diminished. To avoid this, several modifications were considered, and one was tried which appeared promising. Unfortunately, the work had to be abandoned before many observations with it had been carried out. In this arrangement, instead of breaking the wires, the bullet was caused to strike them in such a manner as to /if¢ them from two metallic supports, thereby breaking the circuits. The wires were subsequently broken by the bullet, but the breaking bore no part in the actual operation of the apparatus. Dowling AND DonnELLY—Measurement of Short Time Intervals. 169 Each wire was stretched between two spring clips (aa, fig. 2), about 3 cms. apart, mounted on a rectangular piece of thin ebonite (0). Midway between the clips a large hole was pierced in the ebonite. On either side of this hole lay two stout copper wires fixed to the ebonite, and across them the fine wire (d) was stretched, and pressed lightly on them at the points of contact. One of these was then clamped over the other, so that the wires (d) were at right angles, small-distance pieces of ebonite being interposed, and the distance between the wires (d,d) was measured by a microscope with a vertical micrometer motion. The arrangement was then carefully mounted \ \« so that both wires (d) were in the line of fire, and connexions made from the support wires (¢,c) to the electrical apparatus. On either block one of the wires (c) corresponded to A or B respectively, while the other one of course took the place of Din the original arrangement. One difficulty was found, but was easily met. Owing to the resistance introduced at the points of con- tact (¢, d), there is a noticeable drop in potential across the gap (AD). This had to be balanced across the condenser, and a potential-dividing arrangement PQr was substituted for the former direct-earth connexion shown at O (fig. 1). With this refinement, some few observations were made which indicated that measurements of intervals of about 40 micro-seconds SCIENT. PROC. R.D.S., VOL. XVI, NO. XV. 2B 170 Scientific Proceedings, Royal Dublin Society. could now be made with about the same percentage accuracy as obtained for the larger intervals given in Table I. A great part of this work was completed more than a year ago, but the hope that it would be possible to push the investigation to still shorter intervals caused publication to be held back. It appears, from our own experience, and in view of the investigations of Klopsteg and of Webster already mentioned, that the “capacity” methods of measurement for short time intervals can be relied on under favourable conditions to measure to, at least, the nearest micro-second. In our first arrangement larger errors than this occurred, but the improved mounting seemed to promise readings having this accuracy. Under extremely favourable conditions, and using some special apparatus, Klopsteg reduced the errors to one-tenth of this. 10. bie 13. 14, 15. SCIENTIFIC PROCEEDINGS. VOLUME XVI. . A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, SC.D., F.R.s., and TI. G. Mason, m.a., sc.B. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Lous B. Smyru, B.a., so.B. (Plates I., II.) (February, 1920.) 1s. . The Application of the Food-Unit Method to the Fattening of Cattle. By James Winson, m.a., B.sc. (Plates III.,1V.) (February, 1920.) 1s. The Holothurioidea of the Coasts of Ireland. By Annm L. Massy. (April, 1920.) 1s. . Photosynthesis and the Electronic Theory. By Hunry H. Dixon, so.p., F.r.s., and Horacz H. Poors, sc.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Wint1am Brown, B.sc. (March, 1920.) 6d. . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, m.a., sc.B. (April, 1920.) 1s. . The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Wittiam Brown, 8.sc., and Patrick O’Catnacuan, A.R.C.S8C.1.,A.1.0. (August, 1920.) 6d. . Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G, G. Lonarp, a.B.C.SC.1., B.Sc., PH.D., and AcGnres Browne, A.R.C.SC.1., B.sc. (August, 1920.) 6d. An Investigation into the Causes of the Self-Ienition of Nther-Air Mixtures. By the late Professor J. A. McCurtuanp, p.sc., F.R.s., and Rev. H. V. GILL, S.J., D.s.o., M.c., m.a., University College, Dublin. (August, 1920.) 6d, The Influence of Hlectrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosrpH RrinLy, m.a., D.Sc., F.R.C.SC.1., and Witrrep J. Hicxinzorrom, B.sc. (October, 1920.) 6d. . Notes on some Applications of the Method of Distillation in Steam. By JosEpH REILLY, M.A., D.SC., F.R.C.Sc.J,, and WitFRED J. Hick1nsortom, B.Sc. (October, 1920.) 6d. The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By W. H: Aprnny, p.sc., a.R.¢.so.1., F.1.c., and H. G. BECKER, A.R.C.SC.1., A.1.c, (September, 1920.) 6d. A Determination, by means of a Differential Calorimeter, of the Heat produced during the Inversion of Sucrose. By Henry H. Dixon, sc.p., F.R.s., and Nicer G, Barn, B.a. (December, 1920.) 6d. The Measurement of very Short Time Intervals by the Condenser-Charging Method. By Jonn J. Downine, m.a., r.1nst.p., and Donat DonnELty, sc. (In conjunction with the late Prof. J. A, McCuenuanp, F.R.s.) (February, 1921.) 6d. DUBLIN : PRINTED AT THE UNIVERSITY PRESS BY PONSONRY ANI) GIRES. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 16. FEBRUARY, 1921. A VIBRATING-FLAME RECTIFIER FOR HIGH- TENSION CURRENTS. BY JOHN J. DOWLING, M.A., F.Inst.P. ; AND “ oval Ing ie J. T. HARRIS, B.Sc, | Ri © [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1921. Price Sixpence. miwe ae y. x 4 Sale le Ae Bie Se Ea tring [BRT EASON a erie: EN Pe Mud A sola Ul INNS es XVI. be finn 91 A VIBRATING-FLAME RECTIFIER FOR HIGH-TENSION _ CURRENTS. N\ x By JOHN J. DOWLING, M.A,, F. Inst. P.: : AND J. T. HARRIS, B.Sc. [Read Drecemper 21, 1920. Published Feprvary 21, 1921.] INCOMPLETE rectification of an alternating high-tension current can be effected by placing a small flame near one electrode of a spark-gap in the circuit. The unilateral conductivity of this arrangement is due to the large supply of negative carriers of high mobility derived from the flame. It occurred to one of the authors that, by causing the flame to vibrate in synchronism with the pulsations of electromotive force, a supply of ions would then be made avail- able when the e.m.f. was directed in one sense through the circuit, while there would be practically none available to carry the reverse component of current. Almost perfect rectification would be obtainable in this way. After a few trials, a satisfactory arrangement was devised, the simplicity and novelty of which seem to warrant this description. From the foregoing outline of the principle it will be clear that the flame- vibrations must be strictly in phase with the high-tension electromotive force, and must so occur that the flame rises and falls once, while the electromotive force makes one complete alternation. These conditions are fulfilled in the following way :—The flame-vibrations are produced by a Koénig’s manometric capsule, the membrane of which is either a thin iron plate or an india-rubber diaphragm, to the centre of which a small piece of iron is attached. Mounted close to this is an electromagnet supplied with two independent windings. Through one of these windings the full primary current of the high-tension transformer passes. The other winding carries a direct current from a storage battery, a suitable rheostat being introduced for regulating the current, as described below. The alternating current alone would cause a pulsation of the manometric flame to occur twice in each complete alternation, since the manometric dise would experience an attraction in whichever way the magnet happened to be magnetized. If the direct current in the other magnet winding be now regulated until its magnetizing effect is almost equal to the maximum value of that due to the a.c. winding, it is clear that the magnetization will no longer undergo periodic reversals, but will, on the contrary, vary between two points on the SCIENT, PROC, R.D.S., VOI. XVI, NO. XVI- 2c 172 Scientifie Proceedings, Royal Dublin Society. upper branch of the hysteresis curve. The most favourable value for this “polarizing” current is best got by trial, the manometric flame being observed in a rotary mirror, and the “polarizing” current gradually increased until the alternate maxima of the flame are seen to disappear. -A little adjustment of the relative positions of the magnet and capsule may be required to obtain the largest amplitude in the flame-vibration. The dimensions of the capsule and jet are also very important in this connexion, the best results being obtained when the capsule is rather shallow, so that considerable changes in its volume, and therefore in the gas pressure, result from the movements of the membrane in and out. A jet should be selected such that the flame will be about eight or ten centimetres high, while the supply-tube leading the gas into the capsule should either be of narrow bore or lightly plugged with cotton wool where it opens into the interior of the capsule. With a capsule of 2 c.c. volume and a jet-tube of about 1 c.c., making about 3 c.c. in alk, the rubber diaphragm being about 3 cms. in diameter, we obtain flames oscillating between 1 cm. and 10 ems., the height when not vibrating being about 9 ems. The application of the device for the purpose of rectifying was tested as follows:—A large (Newton) induction-coil was employed to transform the ordi- nary town a.c. supply (200 volt, 50 ~.) to about 6,000 volts... In the secondary circuit, as well as the rectifying spark-gap! cc, were included a small vacuum tube and a d.c, milammeter. The vacuum tube was placed in such a position that the reflected image of it in a fixed mirror was vertically beneath 1 The flat ends of two stout carbon rods were found most satisfactory for this, ~ Downine AND Harris—A Vibrating- Flame Rectifier. 173 the flame; and a rotary mirror was employed to observe the flame and the image of the tube. On exciting the coil, the vibration maxima. of the flame were seen to be strictly in phase with the discharges in the tube; and, on adjusting the polarizing current so as to cut out the intermediate maxima, the corresponding vacuum tube discharges likewise ceased. The milammeter then registered about 20 milliamperes. ‘ With a view to testing the degree of rectitication, a suitable oscillograph not being available, the following experiments were tried :—An a.c. milam- ~ meter was connected in series with the d.c. instrument already used, and the observations repeated. A rather larger current was indicated by this a.c. instrument, which seemed to point to incomplete rectification. However, an interesting explanation was found for this. Tested in series with d.c. currents, the two instruments read the same; but, having doubts as to the validity of such a test, we tried them again in series with current rendered intermittent by a tuning-fork interrupter (50 ~).. It was found that the a.c. instrument read nearly twenty per cent. higher under these conditions. As the condi- tions-are very similar in this test to those obtaining during the rectification experiment, and as the observed discrepancy there is exactly accounted for by this effect, we feel justified in assuming that the inverse current must certainly have been less than--1, milliampere, i.e. less than $ per cent. of the rectified eurrent. tiers : - Our next inquiry was directed to the possibility of using such flame- rectifiers for the purpose of charging condensers to high -steady -potentials from an-a.c. supply. For this purpose it is essential that there should be an exceedingly small inverse current, or, in other words, the resistance of- the spark-gap-must be very high during the intervals when the flame has dropped to its minimum; were this not so, the condenser would discharge with each reversal of the e.m-f., and might even take up reverse charges. To attain satisfactory conditions for a- test it-is necessary in this case-to introduce two rectifying devices in the circuit, one on each side of the con- denser, for if only one were used there would be an oscillating potential on that plate of the condenser which is connected directly to the coil. For the purpose of maintaining a steady, average high potential above earth, only one rectifier is required, the other plate of the condenser and the corresponding coil-terminal being earthed; but under these conditions the readings of an electrostatic voltmeter connected in parallel with the condenser may be mis- leading. We therefore arranged two rectifying devices, with their-a.c. coils in series, and having separate rheostats for controlling their polarizing currents. The two were, of course, arranged to pass high-tension current in the same direction, and were placed respectively in the leads from the coil-terminals to 174 Scientific Proceedings, Royal Dublin Society. those of the condenser, a leyden jar of about 90 ems. capacity. Connexions were made from the condenser-terminals by means of wet thread to the Kelvin electrostatic voltmeter. The wet thread possessed a very high resis- tance, and was employed with a view to testing whether the condenser was actually charged to a steady potential, or if, on the contrary, its potential was alternating. By allowing the flames to vibrate under the influence of the a.c. alone (polarizing current cut off), that is to say, when the condenser- potential was alternating, it was found that the voltmeter reading was zero, On polarizing the magnets, the voltmeter immediately charged up, and showed a potential about the same as that registered when connected directly to the coil. If the insulation of the condenser and voltmeter had been better than they were, we should have expected to obtain perhaps a slightly higher potential, corresponding to the peak value of the high-tension e.m.f. A somewhat more convincing demonstration of the rectifying property of the device was achieved by using a Braun cathode ray-tube as an oscillo- graph, A small coil was mounted near the cathode stream, and formed part of the high-tension circuit. With the rectifier functioning, two spots were formed on the fluorescent screen, one being due to the undeflected rays. The other spot was, of course, produced by rays deflected by the intermittent rectified current, and was several centimetres from the former. No third spot was visible, nor was the undeflected spot even noticeably widened on the side remote from the second spot. With the cutting off of the polarizing current, there was produced a band of fluorescence to each side of the centre spot, the contrast between the two conditions being very remarkable. It was hoped to apply the device in conjunction with a mercury break- interrupter, but work in this connexion has been postponed for want of leisure. In a preliminary trial of such an arrangement one interesting fact came to our notice. It is probable that a single flame will not act as an efficient rectifier with much higher voltages than those we used, since the flame is strongly acted on by intense electrostatic fields, and cannot be made to vibrate regularly in the spark-gap. Besides, the gap has to be of consider- able width, and it seems clear that several flames side by side (all vibrating in synchronism) will be necessary in such cases. There should be no great difficulty in devising an arrangement of this kind. SUMMARY. High-tension a.c. current can be rectified by interposing in the circuit one or more spark-gaps, in which the conductivity is altered in synchronism with the e.m.f. by means of a vibrating flame. Apparatus acting on this principle is described, and an account given of some tests of its operation. 10. str: 13. 14. 15. 16. SCIENTIFIC PROCEEDINGS. VOLUME XVI. . A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, sc.D., F.R.s., and T. G. Mason, m.a., sc.p. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. Smyrs, B.a., sc.B. (Plates I., II.) (February, 1920.) 1s. . The Application of the Food-Unit Method to the Fattening of Cattle. By James Wison, m.a., B.sc. (Plates III.,1V.) (February, 1920.) 1s. . The Holothurioidea of the Coasts of Ireland. By Anne L. Massy. (April, 1920.) 1s. . Photosynthesis and the Electronic Theory. By Henry H. Dixon, so.p., F.z.s., and Horace H. Poor, sc.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Wittiam Brown, B.so. (March, 1920.) 6d. - On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, m.a., sc.B. (April, 1920.) 1s. . The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Winmam Brown, 8.sc., and Parrick O’CannaGHaN, A.R.C.SC.1.,A.1.c. (August, 1920.) 6d. . Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. Lronarp, a.8.c.8c.1., B.SC., PH.D., and Acnes Brownz, A.R.0.S8C.I., B.sc. (August, 1920.) 6d. An Investigation into the Causes of the Self-Ignition of Ether-Air Mixtures. By the late Professor J. A. McCuriuanp, p.sc., F.R.s., and Rev. H. V. GILL, 8.5., D.S.0., u.c., m.A., University College, Dublin. (August, 1920.) 6d. The Influence of Hlectrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosrpH REILLY, M.A., D.Sc., F.R.C.SC.I., and Witrrep J. Hickinzorrom, B.sc. (October, 1920.) 6d. . Notes on some Applications of the Method of Distillation in Steam. By Joseph REILLY, M.A., D.SC., F.R.C.Sc.1,, and WiLFRED J. HickINBoTToM, B.Sc. (October, 1920.) 64d. The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By W. HE. Aprnny, D.sc., A.R.c.sc.1., F.1.c., and H. G. BECKER, A.B.C.SC.1., a..c. (September, 1920.) 6d. A Determination, by means of a Differential Calorimeter, of the Heat produced during the Inversion of Sucrose. By Henry H. Drxon, sc.p., F.R.S., and Nicet G. Bart, B.a. (December, 1920.) 6d. The Measurement of very Short Time Intervals by the Condenser-Charging Method. By Joun J. Dowuine, m.a., F.1nst.P., and Donat DonneLty, usc. (In conjunction with the late Prof. J. A. McCurtxanp, F.r.s.) (February, 1921.) 6d. A Vibrating-Flame Rectifier for High-Tension Currents. By Jonn J. Downe, M.A., F.INST.P., and J. T. Harris, B.sc. (February, 1921.) 6d. DUBLIN; PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIRES. Ped eit} 1 tay Nae yt ‘a 7 eh , chine) Ag cane : , fy ci BANC He ie Boia Riot ase AGA ; ARG AY brie Fay 8 et BC. ne " aE: (RATERS ree ¥ 1" i 4) ~ ij © Nees f ae Vibe. SASS Pee ee een ROSIN aE Ra ak LOxer:, 83 Las an suit ik Ps) Boe ry noua Ro Fee AY) ait Hiche p at aan RR iC sf LAE ; setfariantOe get eee THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 17. FEBRUARY, 1921. A SENSITIVE VALVE METHOD FOR THE MEASUREMENT OF CAPACITY, WITH SOME IMPORTANT APPLICATIONS. BY JOHN J. DOWLING, M.A., F.Inst.P.; [Authors alone are responsible for all opinions expressed in their Communications. } DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1921. Price Sixpence. CA OSES Ne ATEN Pat ae iy fe J XVII. A SENSITIVE VALVE METHOD FOR THE MEASUREMENT OF CAPACITY, WITH SOME IMPORTANT APPLICATIONS. By JOHN J. DOWLING, M.A., F.Insz.P. [Read Drcemser 21, 1920. Published Ferrvary 21, 1921.) I.— Outline of Principles, WHEN alternating potentials of increasing amounts are applied to the grid- circuit of a three-electrode valve, the negative terminal of the filament being connected to the a. c. supply, the plate-current is found to vary, as shown in curve It, fig. 1. This curve, of course, starts from the axis of zero-potential at the point A, where the ordinary characteristic curve for steady grid- potentials crosses this axis (curve J, fig. 1), We will call the curve m the a. c. characteristic of the valve. Tf an alternating e.m.f. # be applied to the terminals of a circuit com- prising a high resistance #& in series with a condenser of small capacity OC, there will be a drop of potential v across the terminals of 2, given approxi- mately by :— v = InnRCE, : , : 5 (1.) where x is the alternation frequency. Suppose this potential v be applied to the grid-circuit of a valve, as shown in fig. 2. A deflection of the milammeter will be obtained which depends on v, but will, in general, be somewhat less than that indicated by the a. c. characteristic for an alternating potential 7 There are two reasons for this. Since the resistance & is large—perhaps half a megohm—the valve will be operating with the grid at an average negative potential of a few volts; in the second place, the resistance of the grid-circuit, being in parallel with R, lessens the effective value of v. It will be found, however, that the milammeter deflections increase almost linearly for increasing values of the capacity C. The arrangement may consequently be utilized for the measurement of capacity if we construct a calibration curve by observing the milammeter readings due to known values of capacity. In this form, however, the apparatus has a limited range, and is neither very sensitive nor accurate. For zero capacity, or, what comes to the same thing, with no alternating e.m.f. acting, there is SCIENT. PROG. R.D.S,, VOL XVI, NO. XVII. 2nd 176 Scientific Proceedings, Royal Dublin Society. still a considerable current indicated by the milammeter. In order to obtain a sensitive arrangement it is necessary to balance this “zero-current” by an opposite steady current. We may then substitute a sensitive galvanometer for the milammeter, with the result that small capacities can be measured. 2 =) plate current —10 amp. grid potent tal -Z O +4 +8 VOLTS. Fic. 1. By an obvious extension of the same principle, if we balance the greater part of the plate-current, due to a rather large capacity in the grid-potential circuit, it becomes possible to make very accurate comparisons between two condensers of nearly equal capacity, or to study slight variations in the capacity of a given condenser. Such variations may be caused by alteration Dow.ine—Sensitive Value Method for Measurement of Capacity. 177 of the dielectric between the condenser-plates, or to movements of the latter relative to one another. It is principally in the latter connexion that the device has its widest field of application ; and I shall give in this paper a brief account of some preliminary work done in that direction. I may say, at the outset, that I have not yet tried to push the apparatus, even in its present form, to the highest possible sensitivity, but have aimed rather at determining the sensitiveness obtainable under ordinary working conditions. As is well known, in developing an apparatus of high sensitivity, in proportion to the augmentation in sensitiveness, unsteadiness and variability of the readings become more pronounced. ‘To derive the fullest measure of ck ie C Fic. 2. performance from such apparatus, a great amount of attention and care has to be devoted to the perfection of its parts, and the elimination of small disturbing influences. Even without such special precautions, the performance of the present apparatus has been remarkably good. Il.— Laperiments with Condensers. In the greater number of tests to be described, the valve, an Ediswan E.S. 1. type receiving valve, was operated with 50 volts in the plate- circuit, the filament being heated by a four-volt accumulator, the negative 2p2 178 | Scientific Proceedings, Royal Dublin Society. terminal being connected to the plate and grid-cireuits. The curves (fig. 1) represent the characteristics under these conditions. A few preliminary observations were made in the manner indicated in fig 2; but these were merely made with a view to testing the possibility of the method, and, as this mode of working is very insensitive, it is not necessary to quote any of the results obtained. ‘I'he first modification tried was to balance the current in the milammeter (fig. 2) by an opposed current from a single cell in series with an adjustable high resistance. Some experiments were made thus, a Paul Unipivot galvanometer with an rd 60 40 20 S < Q OH 400 800 1200 10” farad Fig. 3. universal shunt being employed in place of the milammeter. As it was convenient to utilize the laboratory supply for the 50-volt plate-cireuit battery, it was found impossible to connect the town supply of 200 volts a.c. directly to Ad in consequence of the insufficient insulation between the ac. and d.c. circuits. An insulated 200-volt a.c. supply was therefore obtained, using a small step-up transformer from the 12-volt a.c. galvano- meter lamp circuit. Beyond this no special care was taken as regards insulation or shielding of the apparatus. A few observations were thus made, using a graduated air-condenser of the rotating-sector type. ‘his Dow11ne—Sensitive Valve Method jor Measurement of Cupacity. 179 graduated from 100 to 1,200 micro-microfarads, and in one experiment was made 200,000 ohms. ‘lhe relation between the galvanometer deflection and capacity was almost linear, as shown in the curve (fig. 3), the condenser graduations being unreliable for the smaller values. A very sensitive moving-coil galvanometer! was then substituted for the Unipivot instrument, the universal shunt being retained, and the connexions altered to those shown in fig. 4. This had the advantage of being more compact, as only one storage battery was required. Preliminary observations were made, as before, with the graduated air-condenser, and the expected high sensitivity of the apparatus was fully obtained. For example, with a ne oe C 70,000 ohm Fie. 4. capacity of 100 micro-microfarads (107° farad) an alteration of 1 per cent. in the capacity produced a deflection of 300 scale-divisions, the galvanometer shunt being at 10.2. With this degree of sensitiveness the deflections were fairly steady, but, with a lower value of the shunt, the spot of light commenced to execute irregular movements, rendering accurate readings impossible. 1 Elliott's ‘Century ” galvanometer: 1,000-ohm coil; 2°5 x 107° amp. per scale- division. 2 In view of the fact that the resistance of the plate-circuit is very high, and that the galvanometer and shunt were in circuit with 10,000 ohms in the potentiometer- -circuit, the actual shunt values were not very different from those marked on the shunt itself. 180 Scientific Proceedings, Royal Dublin Society. ‘This unsteadiness was principally due to fluctuations in the a.c. supply, but also partly to the fact that the apparatus was insufficiently shielded and imperfectly insulated. A large sheet-metal box was procured, and the valve, resistances, condenser, and filament-battery were placed on sheets of ebonite within. ‘lhe connexions to the a.c. and d.c. supplies and to the galvanometer were, of necessity, still exposed; but, nevertheless, a considerable improvement in the steadiness of the readings was found. It appeared likely that the 46 © 48 50 psy K/0 cma Fie. 5. remaining unsteadiness was chiefly due to the a.c. fluctuations. It must be remembered that, when the greater part of the plate-current is balanced with the potentiometer, the current through the galvanometer is a difference effect. Just as a small change in the capacity C produces a large effect on the galvanometer, so will small variations in any of the other quantities which occur in equation (1). Dowiine—Sensitive Valve Method for Measurement of Capacity. 181 II1.—Application of foregoing as an ultra-micrometer and micro-pressure gauge. The first trials were made with a simple form of plate-condenser, in which one of the plates was movable normally by means of a micrometer- serew. The two plates were 2:5 cms. in diameter, and were carefully ground to a plane surface. ‘The movable plate was attached to the end of a short rod, which was ground to a smooth sliding fit in a short length of brass tube. This tube was screwed in in place of the anvil of a micrometer-screw gauge, and a couple of springs were arranged to pull the sliding-rod so that its extremity made tight contact with the end of the micrometer-screw. This arrangement was screwed to the long leg of an L-shaped piece of ebonite, so that the other plate, supported on a rod inserted in the shorter limb of the L, was opposite the first plate. By a device which I shall describe later, the two plates were made exactly parallel, and the micrometer allowed them to be moved apart any desired fraction of a millimetre. An example of the results obtained with this rough arrangement is given in fig. 5, where the abscissae give the micrometer readings in thousandths of a centimetre. The points lie nearly on a straight line over the range covered ; but the sensitiveness is not very great. The fact that the points lie a little to one side or the other of the line is obviously due to the imperfect action of the screw. It is easy to see that the sensitiveness can be greatly increased by bringing the plates closer together, and also by enlarging the areas of the plates. With plates close together and of moderate size, they form a con- denser of which the capacity is much greater than that of the grid and other connexions. In equation (1) we can thus regard Cas wholly the capacity of this condenser. C' will then be given by the formula— = A ‘ 11 C= 7 (9 x 10") farads. And we may write (1) in the form— v= Be : C 61% . (IL) if we include all the other quantities in the constant 6, Now, the slope of the a.c. characteristic being practically uniform, a given change in v will pro- duce a corresponding change of constant amount in the galvanometer reading, no matter what may be the actual value of v. Consequently, the condition for sensitiveness is that = should be as great as possible. But 3 = em Sr CET 182 Scientific Proceedings, Royal Dublin Society. hence the sensitiveness increases directly in proportion to the square of the linear dimensions of the plates and inversely as the square of their distance apart. The negative sign, of course, indicates that as the plates are moved apart the grid potential v diminishes. In the next apparatus set up, two circular metal plates 7 cms. in diameter “were utilized. One of these was cemented to the table of a Leitz microscope. The fine adjustment-screw of this had a large divided head, each division of which corresponded to a vertical movement of the microscope tube through sass cm, and + of a division was easily readable. The strong spring of the fine adjustment effectually prevented backlash. A strong glass tube was supported in the microscope tube by two stout plugs which held it very rigidly, and the lower extremity of this tube was introduced into a small metal cup attached to the centre of the upper condenser-plate. While this plate rested on its Dowiine—Sensitive Valve Method for Measurement of Capacity. 183 fellow, the cup was filled with molten fusible metal (90°C. m.p.), which rapidly cooled and cemented the metal plate to the glass support. When cool the plates were separated by turning the micrometer screw the desired amount. This condenser allowed a much higher sensitivity to be attained. Fig. 6 shows the first set of observations thus obtained. During this set the apparatus was shielded in the metal box previously mentioned, and the con- ditions were fairly steady. The points lie almost on a straight line, but the gradual increase in sensitiveness with decreasing separation of the plates is just noticeable, the points really lying on a parabolic curve. (The observations were taken with the galvanometer-shunt at 10, but the ordinates in fic. 6 have been plotted as if the galvanometer was unshunted.) The average sensitivity throughout the range is about 3200 scale divisions for s=oo7 cm.; ie. 1 division deflection corresponds to a displacement of about 3 x 107 cms, Several sets of observations were made with this apparatus with the plates at different distances apart; and it seemed probable that, under sufficiently steady conditions, it would be possible to detect a displacement of at least -d x 107% cms. Further work with the apparatus is still being carried on, and I will not, therefore, give further particulars in this paper. A sensitive pressure-gauge may easily be constructed involving the same principle. A shallow recess was turned in an ebonite disc, and a flat metal dise (6-4 cm. diameter) was cemented therein, so that its surface lay about 2mm. below the ebonite rim. A similar dise, about 6 mm. in thickness, was cemented to a sheet of india-rubber (such as is used for foot-bellows). The rubber was then cemented to the ebonite rim while slightly stretched, the metal plates being thus supported about 1°3 mm. apart. Suitable connexions were, of course, made by fine wires from each of these plates, and a tube through the ebonite gave connexion with the air-space inside. When the pressure rises within, the plates are driven apart, and the capacity of the system diminishes. The apparatus was roughly tested, and it was found that a pressure excess of 1 cm. of water resulted in a deflection of the galvano- meter (shunt = 10) of about 300 divisions—or, roughly, 3 galvanometer divisions (unshunted) corresponded to a pressure of 1 dyne percm.? It is obvious that with improved construction, larger discs both of metal and rubber, thinner rubber, and closer proximity of the plates, it should be possible to measure very minute pressure differences indeed. T have said enough to indicate the possibilities of the method, and hope to publish further particulars when I have succeeded in attaining satisfactory SCIENT. PROC. R.D.S., VOL. XVI, NO. XVII. 25 184 Scientific Proceedings, Royal Dublin Society. steadiness in its operation. So far as the present observations go, the degree of sensitiveness falls a little short of that obtained by Whiddington! for his ultra-micrometer. However, the advantages of using a deflection method would in many cases compensate for a slight lack of sensitivity. I shall, however, reserve for a future occasion a discussion of this point, and of some other considerations relative to the practical employment of the method. 1 Whiddington, Phil. Mac., vol. xl, Nov., 1920, pp. 634-639. 10. 11. 13. 14, 15. 16. 17. SCIENTIFIC PROCEEDINGS. VOLUME XVI. . A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, SC.D., F.R.s., and T. G. Mason, m.a., sc.B. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. SmuyrH, B.a., sc.B. (Plates I., II.) (February, 1920.) 1s. . The Application of the Food-Unit Method to the Fattening of Cattle. By James Winson, m.a., B.sc. (Plates III., 1V.) (February, 1920.) 1s. The Holothurioidea of the Coasts of Ireland. By Anne L. Massy. (April, 1920.) 1s. . Photosynthesis and the Electronic Theory. By Henry H. Dixon, so.p., F.n.s., and Horace H. Poot, sc.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Winuiam Brown, B.sc. (March, 1920.) 6d. . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, u.a., sc.p. (April, 1920.) 1s. . The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Wimmam Brown, 8.sc., and Parrick O’Cannacuan, A.R.¢.8¢.1.,a.1.c. (August, 1920.) 6d. . Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. LronarpD, a.B.C.8C.1., B.Sc., PH.D., and Agnes Browne, A.R.C.8C.I1., B.sc. (August, 1920.) 6d. An Investigation into the Causes of the Self-Ienition of Ether-Air Mixtures. By the late Professor J. A. McCurttanp, p.sc., F.z.s., and Rev, H. V. GILL, S.J., D.S.0., M.c., m.a., University College, Dublin. (August, 1920.) 6d. The Influence of Electrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosrpH Rernty, m.a., D.Sc., F.R.C.SC.1., and Witrrep J. Hickinzorrom, B.sc. (October, 1920.) 6d. . Notes on some Applications of the Method of Distillation in Steam. By JosEpH REILLY, M.A., D.SC., F.R.C.Sc.1,, and Winrrep J. Hicxingorrom, B.Sc. (October, 1920.) 6d. The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By W. H. Aprnny, D.sc., a.R.c.sc.1., F.1.c., and H. G. BECKER, A.R.C.SC.I., A.I.c. (September, 1920.) 6d. A Determination, by means of a Differential Calorimeter, of the Heat produced during the Inversion of Sucrose. By Henry H. Dixon, sc.p., F.R.S., and Niczn G. Ban, B.a. (December, 1920.) 6d. The Measurement of very Short Time Intervals by the Condenser-Charging Method. By Joun J. Dowzine, m.a., F.1nst.P., and Donat Donnexty, u.sc. (In conjunction with the late Prof. J. A. McCretuanp, r.z.s.) (February, 1921.) 6d. A Vibrating-Flame Rectifier for High-Tension Currents. By Joun J. Dowtine, M.A., F.InstT.P., and J, T. Harris, B.sc, (February, 1921.) 64d. A Sensitive Valve Method for the Measurement of Capacity, with some Important Applications. By Joan J. Dowzine, u.a., F.inst.p, (February, 1921.) 64d. DUBLIN : PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIRES, ried at ry a) RRS teaeae GERIER TOL THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 18. MARCH, 1921. A DIRECT READING ULTRA-MICROMETER. BY JOHN J. DOWLING, M.A., M.R.LA., F.Inst.P. [Authors alone are responsible for all opinions expressed in their Communications. } yah TQ DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY. LEINSTER HOUSE, DUBLIN. _ WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1921. Price Sixpence. AION PA AG [ 185 ] XVIII. \ A DIRECT READING ULTRA-MICROMETER. By JOHN J. DOWLING, M.A., M.R.I.A., F.Inst.P. Read January 25. Published Marcn 4, 1921. In a recent communication! to the Royal Dublin Society I described an electric valve method for the measurement of capacity, which could also be applied to the measurement of extremely small displacements; but, inasmuch as a very constant source of alternating e.m.f. was necessary for its opera- tion, further work was promised with a view to overcoming this drawback. In the course of this work, the present method suggested itself, and, on trial, was found to be even more sensitive than that previously described. It has the great advantage of simplicity, and, although capable of extraordinary sensitiveness, is singularly free from that unsteadiness to be expected in a highly sensitive apparatus. Furthermore, with a suitable adjustment of the apparatus, it is possible to maintain a high and almost constant degree of sensitiveness over a total displacement of, perhaps, one-tenth of a milli- metre. In the present communication I shall confine myself to a short descrip- tion of the apparatus in the form first set up, together with an account of some trial measurements made with it. I shall reserve for another occasion a fuller consideration of the principles involved, and of the conditions of operation of the device. It is well known that for sustained oscillations to occur jin a three- electrode valve circuit, the values of the inductances and capacity must fall within certain limits. If the self-induction and the coupling of the valve circuit shown in the figure be left unaltered, while the capacity is gradually changed, two effects are produced: the frequency of the oscillations changes and their amplitude is also, in general, affected. The former phenomenon was that availed of by Whiddington in the construction of an ultra-micrometric device.” The latter may be regarded as the effect upon which the present device depends. As the amplitude of the oscillations of potential varies, a corresponding change takes place in the plate current, and it is this property of the oscillating circuit which is utilized. “CA Sensitive Valve Method of Measurement of Capacity.” Scient: Proc. Roy. Dub. Soc., vol. xvi, No. 17, p. 175. 1921. ? Phil. Mag., Nov., 1920, vol. xl, pp. 634-639. SOIENT. PROC. R.D.S., VOL. XVI., NO. XVIII. 25 186 Scientific Proceedings, Royal Dublin Society. The connexions of the oscillating circuit shown in the figure were slightly modified in practice, a potential balancing device, similar to that described in my previous paper,’ being employed so as to permit a sensitive galvano- meter to be substituted for the milammeter A. An ordinary B.T.H. receiving valve was employed, and a negative potential of a few volts was. found necessary in the grid circuit to obtain good oscillatory conditions. In the preliminary experiments a coil of about 10° cms. inductance was used, the portion Z being usually about one-quarter of the whole. An air con- denser, having a range of 100 to 1,200 micro-microfarads was used in the course of these preliminary tests. I may mention that the currents in the plate circuit varied from one-half to several milliamperes, and that it was con- venient to employ an Ayrton-Mather universal shunt with the galvanometer to facilitate preliminary adjustments. One consideration is of importance in the working of the apparatus. With a given adjustment of the inductances, it is found that there is a loc. cit. Dowtine—A Direct Reading Ultra-Micrometer. 187 certain value of the capacity, sometimes more than one, for which the plate current reaches a maximum. It is clearly necessary to adjust the apparatus so that it is not functioning near such maximal points. The first trial of the apparatus as a micrometric device was carried out as follows:—Two circular plates of steel, 6-4 cm. in diameter, and ground reasonably flat, were mounted on the stand of a Leitz microscope. One was cemented to the platform, while the other was carried on the end of a glass rod which projected from the microscope tube, in which it was firmly supported. The two plates were adjusted to be exactly parallel. The fine adjustment screw of the microscope was furnished with a divided head, each division of which corresponded to a vertical motion of the tube through siz Im. It was thus found possible to move the upper plate through this distance with reasonable certainty, the screw adjustment being remarkably good. Observations were made as follows:—The galvanometer being shunted, the plates were screwed apart a certain number of divisions, and the sliding resistance S adjusted until the galvanometer was at zero. The shunt was then gradually reduced; slight further adjustments of S being made, if necessary, while this was being done. Except in the most sensitive settings, the shunt was reduced to unity, and observations then made of the deflection of the galvanometer resulting from the rotation of the microscope screw through one scale division. It was generally found that a deflection of at least two or three hundred divisions was thus produced; but, by selecting a suitable initial separation of the plates, much larger deflections could be obtained. It. was easy to set the apparatus so that about six hundred galvanometer divisions corresponded to one division of the screw; that is to say, a displace- ment of 5,'55 cm. produced a deflection of the galvanometer of six hundred scale divisions. In other words, a displacement of -syhao5 cm. was detect- able unmistakably. On more than one occasion a much higher sensitivity was reached, but its magnitude could not be stated with certainty. The smallest movement that could be given with any certainty to the plates was that represented by one division on the serew-head. The highest sensitivity that could be utilized was, therefore, already reached when the galvanometer deflection extended, as just mentioned, six hundred divisions from one end of the scale to the other. By shunting the galvanometer, the effect of moving the plate could be reduced, and an estimate formed of higher degrees of sensitivity. However, I am not at present certain if the sensitivity thus estimated is correctly given by multiplying the galvanometer reading by the shunt ratio. In order to produce displacements of sufficiently minute nature, another 188 Scientific Proceedings, Royal Dublin Society. plan was adopted. A brass rod, having a diameter of one-eighth inch, was suspended vertically in a wooden frame. Two circular brass plates, 10 cms. in diameter, with three-sixteenth inch holes drilled at their centres, were carefully ground to a plane surface and cemented to ebonite blocks, which were drilled to slide smoothly on the rod. The two discs were then fixed in position, so as to be separated by ‘12 mm., and to be as nearly parallel as possible. The ebonite blocks were each furnished with a set-screw, and it was assumed that if the rod were subjected to tension the plates would then be moved apart by the amount that that portion of the rod between the two set-screws increased in length. The screws were approximately 1°5 cms. apart on the rod. With this apparatus it was found that the highest sensitiveness obtain- able with this ultra-micrometer was even greater than had been suspected. For one particular adjustment of the inductances the galvanometer was quite stable in the absence of accidental vibrations, but nevertheless an extra- ordinary degree of sensitivity was displayed. Thus, it was found that a deflection of four hundred and ten divisions of the galvanometer was pro- duced when a load of five hundred grammes was suspended from the rod. A simple calculation shows that such a load applied to a brass rod one-eighth inch diameter will produce an extension of 9°6 x 10° cm. in a length of 15 cms., assuming Young’s modulus for brass to be 10” dynes per sq. cm. As this extension corresponds to four hundred and ten divisions of the galvanometer, we may, in round numbers, say that a displacement of 10 x 10° + 400, or 25 x 10° cm. was detect- able. With a more sensitive galvanometer perhaps a further limit might have been attainable, but in view of the possible unsteadiness of the readings one cannot predict what the lowest attainable limit might be. The number quoted was obtained under very unfavourable conditions, no screen- ing of any kind being placed around the apparatus, and the only precaution taken by the operator being to stand as still as possible while lifting and lowering the weight by a string. A movement of the hand towards where the apparatus lay on the table was sufficient to cause a perturbation of the galvanometer. I have no doubt, however, that by enclosing the apparatus as far as possible in a metal box, as is the usual practice with oscillatory circuits, this trouble will be eliminated. A careful series of tests with improved apparatus is now being under- taken, the results of which I hope to embody in another communication. 1 This trouble has since been completely overcome. 10. 11. 13. 14. 15. 16. 17. 18. SCIENTIFIC PROCEEDINGS. VOLUME XVI. . A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon SC.D., F.R.s., and T. G. Mason, m.a., sc.B. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. SmyrH, B.aA., sc.B. (Plates I., Il.) (February, 1920.) 1s. . The Application of the Food-Unit Method to the Fattening of Cattle. By James Winson, m.a., B.sc. (Plates III., IV.) (February, 1920.) 1s. . The Holothurioidea of the Coasts of Ireland. By Annu L. Massy. (April, 1920.) 1s. . Photosynthesis and the Hlectronic Theory. By Henry H. Dixon, sc.p., F.r.s., and Horace H. Poors, sc.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Wittiau Brown, B.sc. (March, 1920.) 6d. . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, m.a., sc.p. (April, 1920.) 1s. . Uhe Change in the Rigidity of Nickel Wire with Magnetic Fields. By Winnmiam Brown, 8.sc., and Parrick O’CaLnacHan, A.8.C.SC.1.,A.1.c. (August, 1920.) 6d. . Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. LronarD, A.R.C.SC.1., B.Sc., PH.D., and Acnes Brownk, A.R.C.SC.L., B.sc. (August, 1920.) 6d. An Investigation into the Causes of the Self-Ienition of Ether-Air Mixtures. By the late Professor J. A. McCuetuanp, p.sc., F.R.s., and Rey. H. V. Git, s.J., D.s.o., m.c., m.a., University College, Dublin. (August, 1920.) 6d. The Influence of Electrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosnrpH Rwinty, .a., D.Sc., F.R.C.SC.1., and Witrrep J. Hickinsorrom, B.sc. (October, 1920.) 64d. - Notes on some Applications of the Method of Distillation in Steam. By JosepH REILLY, M.A., D.SC., F.R.C.Sc.1., and WiLFRED J. Hickinzorrom, B.sc. (October, 1920.) 6d. The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By W. H. Aprnny, p.sc., a.R.c.sc.1., F.1.c., and H. G. BECKER, A.R.C.SC.1., A..c. (September, 1920.) 6d. A Determination, by means of a Differential Calorimeter, of the Heat produced during the Inversion of Sucrose. By Henry H. Dixon, sc.p., ¥.R.S., and Nick G. Bat, B.a. (December, 1920.) 6d. The Measurement of very Short Time Intervals by the Condenser-Charging Method. By Joun J. Downie, m.a., r.1nst.p., and Donat Donnutxy, usc. (In conjunction with the late Prof. J. A. McChenuanp, F.n.s.) (February, 1921.) 6d. A Vibrating-Flame Rectifier for High-Tension Currents. By Jonn J. Dowuine, M.A., F.InST.P., and J. T. Harris, B.sc. (February, 1921.) 6d. A Sensitive Valve Method for the Measurement of Capacity, with some Important Applications. By Joan J. Downie, .a., r.inst.e. (February, 1921.) 6d. A Direct Reading Ultra-Micrometer. By Joun J. Dowtine, .a., M.R.1A., F.inst.P, (March, 1921.) 6d. DUBLIN: PRINTED AT THE UNIVERSITY PRESS BY PONSONKY AND GIBES. Misia eae ae Soot thiwire haba 4). 34) 0 CETORABS Wy Ga PRES SKANK q THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 19. MARCH, 1921. STUDIES in toe PHYSIOGRAPHY ann GLACIAL GEOLOGY OF SOUTHERN PATAGONIA. BY a y S HG MENTON.: 6 MPS es i i ; [COMMUNICATED BY PROFESSOR G. A. J. COLE, F.B.S.] (PLATES V, VI, and VII.) \ [Authors alone are responsible for all opinions expressed in their Communications. } DUBLIN : PUBLISHED BY THE ROYAI. DUBLIN SOCIETY, LEINSIER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVEN GARDEN, LONDON, W.C. 2. 1921. Price Four Shillings and Sixpence. Roval Dublin Society. SESS OES FOUNDED, A.D. 1731. INCORPORATED, 1749. ~~ SCIENTIFIC MEETINGS. ‘Tur Scientific Meetings of the Society are usually held at 4.15 p.m. on the fourth Tuesday of every month of the Session (November to June). The attention of authors is directed to the following extracts from the regula- tions governing the presentation of communications at Scientific Meetings of the Society, and the publication of papers in the Scientific or Heonomic Proceedings :— “‘An author who desires to present a communication at a Scientific Meeting of the Society shall furnish to the Chief Executive Officer for Science a fortnight previous to that meeting the title and an abstract (not exceeding 200 words) of his communication, together with an estimate of the time required for presentation. ‘“‘An author who desires to have a communication published forthwith as a paper in the Proceedings of the Society shall furnish to the Chief Executive Officer for Science the full text and necessary illustrations at least a week before the meeting of the Committee of Science at which he desires publication to be considered. Meetings of the Committee of Science are usually held on the second Tuesday of each month from November to June inclusive. ‘‘ A paper may be published either in the Scientific Proceedings or the Iiconomic Proceedings of the Royal Dublin Society, as directed by the Committee of Science. The copyright of papers so published shall be the preperty of the Society. Pras ‘‘ Fifty copies of each paper shall be supplied free of cost to the Author or to each Joint Author. ‘‘ Additional copies may be obtained at Cost Price if applied for by the Author when returning the proofs.’ Authors are requested to apply to the Chief Executive Officer for Science for further information concerning the preparation of papers and illustrations for press. soe XIX. STUDIES IN THE PHYSIOGRAPHY AND GLACIAL GEOLOGY OF SOUTHERN PATAGONIA. By E. G. FENTON. (COMMUNICATED BY PROFESSOR (. A. J. COLE, F.R.S.) [Read January 22,1918. Published Marcu 16, 1921.] I. General Features of the Region. In the present paper I propose to deal with the quaternary and recent geological history of a certain limited portion of Patagonia in which I lived for a period of seven and a half years. The appointment which I held during - that time gave me abundant opportunities of visiting the “Camp” generally and studying the superficial features of the country. I was unfortunately never able to visit the country north of the Santa Cruz river to any great extent, or south of the Straits of Magellan; nor did I ever succeed in getting as far as the chain of the Andes or those most interesting channels or fiords which penetrate the southern cordilleras of Patagonia. However, I found in the eastern plains, even limiting myself to a small portion of their recent history, abundant materials for study. This district, lying between the Santa Cruz river and the Straits of Magellan in the one direction, and extending from the Andes to the Atlantic in the other, consists chiefly of plain or pampa, its level varying from four to five hundred feet along its eastern edge up to perhaps eight hundred or even in places a thousand feet further west. If we begin our study by taking a typical portion of the high pampa on the north side of the Gallegos river, the first thing that will strike us (Pl. VI, fig. 1) is its extreme flatness of surface. This extraordinary flatness is characteristic of all the eastern plains of Patagonia north of the Gallegos river ; and when travelling through the country the landscape becomes at times almost terrible in its monotony ; not even a single hillock can be seen to break the everlasting evenness of the horizon. SOIENT. PROC, R.D.S., VOL. XVI, NO. XIX 26 190 Scientific Proceedings, Royal Dublin Society. After travelling a certain distance, however, over the plain, the monotonous scene depicted above may change without warning. Suddenly a huge valley may be found to lie below us, hundreds of feet deep, extending for miles, and eventually rising once more in the distance to the original level. Along the bottom of this valley, which will be found to run from west to east, a small and, comparatively speaking, insignificant river will in most cases be observed, flowing sluggishly in various sinuous curves, and small, out of all proportion, to the size of the valley in which it is found. These river valleys begin, as a rule, among the hills to the west, and extend to the Atlantic ; but besides these, we occasionally come across other valleys which are found, after being followed for some miles, to begin blindly, and end in the same manner; in fact, they are hollow, cup-shaped, or, more correctly speaking, dish-shaped, depressions scooped out of the pampa, and have in their lowest parts, instead of a river, as a rule, one or more salt lagoons. These are the “bajos ” of Patagonia, and are some of the most puzzling features found in the district. here are also numbers of long, more or less narrow valleys styled “ canadones,”’ running as tributaries to the river-valleys and bajos. These vary in size, and often cut deeply back both to the north and south into the pampas. Sometimes they run for many miles, and although, as arule, they are dry, at times a small stream runs down the centre. In this eastern portion of Patagonia no trees are ever found either on the pampas, valleys, or cahadones, but the surface of the ground is everywhere clothed with a coarse grass, and in places with a scrubby bush a foot or two in height. The condition is one found generally in steppe countries. The descent from the higher valleys to the river level below is not in the form of an inclined plane, but in a series of steps or terraces. These latter vary in number in different localities ; in some places there are seven, in others only one or two, whereas in others we may find the pampa falling direct in one drop to the river below. Besides this condition of extreme flatness which I have mentioned, we shall also see, if we examine fig. 1 carefully, that there is evidence of con- siderable erosion having taken place in the surface of the country as well as the levelling action. In the foreground of fig. 1 we are looking across a large canadon which has eaten back into the country from the Atlantic. Now, if we come down on to one of the terraces we may possibly see a peculiar fat-topped hill standing out in a very bold manner some considerable distance, perhaps a mile or so from the edge of the pampa. The level of the top of one of these hills is found to be the same as the general plane of the pampa; in fact, the hillis an isolated portion of the pampa which has been able to resist the great Frnton—Physiography and Glacial Geology of S. Patagonia. 191 erosive action that cut out the river-valleysand cahadones. In one place the top of one of these hills was found to be 520 feet above the river, and this was the same as the level of the nearest high pampa. The next question in the study of the pampa is the shingle, which, as C. Darwin and others long ago pointed out, is found practically all over the plains of Southern Patagonia. In fact, I do not believe there is a single square yard of the high pampa between the Straits of Magellan and the Santa Cruz river without its shingle covering. The shingle forms a comparatively thin layer, varying from 10 to 25 feet in thickness. It lies in practically every instance on well-stratified middle tertiary rock, the latter in most cases consisting of a more or less whitish, greyish, or dull blue soft stone, locally called Tosca. ‘his peculiar formation, although apparently sedimentary, has yielded nothing but mammalian and bird fossils. The pebbles on the surface are as a rule broken and angular, as the result of repeated pampa fires ; but below the surface they are comparatively well rounded and sufficiently worn to show that they have been subjected to prolonged water action. The diversity in composition of the individual stones which make up this shingle is remarkable. Some are white, some are black, some are yellow, some are red, and some are green; but by far the greater number in most localities consists of a yellowish porphyritic variety. No particular bedding, either false or true, is to be seen in any of these layers of shingle; in fact, they give one the idea of ballast dumped higgledy-piggledy out of wagons to fill up the place. This same absence of arrangement of the stones is to be found in almost all parts of the first pampa. The stones found on the pampa are often as much as six to eight inches in length, and we might call the greater number of them three-quarters rounded, many of the stones still exhibiting the original sides, and having only the corners smoothed off. I would once more wish to impress on the reader that one of the most extraordinary characteristics of this shingle is its universal and even distri- bution ; itis spread broadcast, practically without a break, over almost the whole surface of Argentine Patagonia. It is found on the highest hills, in the lowest valleys, and even extends, as has been shown by the Beagle expedition, for hundreds of miles into the Atlantic under the sea. It never occurs in heaps, and, as seen in the cliff sections, the underlying rock shows, when completely covered by this shingle, practically no indentation or cuttings—in fact, it exhibits the same wonderful levelness that the over- hanging shingle does. The unconformity between the two is apparently very slight, and yet there must have been a very considerable hiatus in time between the depositions of the two formations. As we trace the face of a cliff down on the Atlantic coast, there is no 262 192 Scientific Proceedings, Royal Dublin Society. repetition of the shingle-layer; practically all the rock is what I have described as Tosca. It is well stratified, showing very even bedding ; in some places, however, one can see layers cropping out up to several feet in thickness, composed chiefly of marine shells. Where shingle occurs down the face of a cliff it always seems to mark the remains of old river-beds. The pebbles found are of a very different composition from those found on the pampa the layers are limited in extent laterally, and they never run along the cliff more than a few hundred yards in one place. ‘There is, as a rule, distinct and well-marked unconformity between this latter shingle and the surrounding rock; and, when we consider that these tertiary river-beds are very rare and only found in a few places, we realize that there is practically no resemblance between them and the great layer of shingle found on the pampa. From these facts we can conclude that this Jayer of pampa shingle is a unique formation, and that there is no parallel to be found in this country— at least in the earlier rocks. Now, if this shingle-layer was a beach or marine formation, as Darwin tried to prove, we should surely have found in the rocks beneath some repetition of it formed when the country began to subside under the sea or to emerge from it. Especially should this be the case when we consider that these subsidences during tertiary time extended right to the cordillera, as is proved by similar marine formation being found well among the mountains. Consequently, I think the marine theory must be abandoned. The next point of importance with regard to this shingle-layer is that in all the localities where I have examined it between the Santa Cruz river and the Straits of Magellan I have found it unfossiliferous. J understand that to the north of San Julian and Deseado in many places recent sea-shells are found on the surface of this layer; but from what I have gathered from intelligent people who have visited the locality in question, these shells are not found to any depth in the shingle. This would indicate that they represent a recent subsidence long after the shingle had been spread out over the country. I have myself noticed on many occasions mussel shells lying in heaps, mixed with the excreta of sea birds, on the pampa upwards of two miles or more from the beach, so that we should always bear in mind the possibility of carriage by birds when considering isolated instances. II. felations between the shingle and the surface on which it was laid down. If we go round the point at the mouth of the Gallegos river and turn to the north along the Atlantic coast, when we get as far as the first pampa level we see, as we have said, all along the top a darkish-brown layer, lying Funron—Physiography and Glacial Geology of S. Patagonia. 193 apparently conformably on typical tertiary rock. On examining this rock, however, it is found not to consist of pure shingle, as one would imagine from having seen the similar formation in other places, but to have a composite structure. The shingle is found above as elsewhere, but below this there is another layer of equal thickness of marine formation holding an abundance of fossilized sea-shells. ‘This marine layer has been described by Mr. J. B. Hatcher, and named by him the Cape Fairweather Bed. The unconformity between the Cape Fairweather Bed and the tertiary rock is slight but distinct, whereas there seems to be in places none between the Cape Fair- weather Bed and the shingle, the two layers passing imperceptibly into one another. In no instance in the district in question have I yet found any formation overlying this shingle except lava; the latter exists, however, in extensive sheets in many localities, and will be dealt with in a later section. We have already seen that the apparent unconformity between the shingle and the tertiary rock is on the highest level always insignificant; it will consequently be evident that, even if all the shingle were removed, the pampa would still possess that extraordinary flatness which I have mentioned as existing everywhere, and it will be further evident, moreover, that this flatness must have been a primeval condition, and must have existed before the shingle appeared. It would seem also that the shingle crept down gently and gradually over this smoothed and level pampa until it completely covered it, and that the transition fiom the condition of the climate which brought about the planing off of the pampa to those which covered the latter with sand and shingle was completely devoid of violence. ‘he tertiary rock underlying the Cape lairweather Bed has been generally looked upon as belonging to the Miocene system ; this would place the Cape Fairweather Bed either in the late Miocene, the Phocene, or early Pleistocene. The uncon- formity between the Cape Fairweather Bed and the tertiary rock is, as we have seen, slight; but it does not follow necessarily from this that there was a short hiatus in time between the periods when they were respectively deposited. It may only mean that the great planing off process extended much further back into tertiary times than we might have thought, and was operating before the Cape Fairweather Bed was formed. We have evidence from other sources that the tertiary rock suffered considerable erosion before the shingle came down, and that it once extended to a much greater altitude; it would consequently seem that the elevation and sub- sidence which occurred during tertiary times took place very gradually. This is particularly clear when we remember the complete absence of faulting and folding found in these rocks, and the perfect horizontality of the strata composing them, All these facts would point to a condition of affairs in pre- 194 Scientific Proceedings, Royal Dublin Society. shingle times as that of a dry, probably warm, climate, with high winds; in other words, an exaggeration of the climatic conditions which prevail at present, with less snow and greater warmth. At present the summers of Patagonia are very dry, and high westerly winds prevail; great quantities of dust are blown every year from the pampas into the Atlantic; still very little impression is made on the general outline of the country, as the surface is everywhere protected, not only by a coarse grass and bush, but by twenty odd feet of shingle which is spread over it. It is only near the settlements, where the grass and scrub become eaten down and trampled away, that the winds may be seen to make any impression, and it is from such places, and from the roads which are now everywhere being made in the country, that the chief part of the dust is derived. Iam informed that a few years ago, before this became a sheep-farming country, the grass was everywhere much longer than it is at present, and consequently the protection to the surface was greater. In the springs of the years, when the snow is melting, a certain amount of erosion also takes place from the floods, but this only occurs in the cafiadones and valleys, and even here very little cutting is found to have taken place, except in those cafadones where the surface has been broken by wheel- tracks or other traftic. For instance, the winter of 1915 was a very severe one, and there was much snow and rain, the floods in the spring being enormous, and yet, on carefully examining quite a number of cafadones where there was no traffic, | was surprised to find that there was practically no erosion of the surface, except a few tiny gullies, six inches wide and deep, which would be obliterated with one month’s dust and growth of vegetation. Such winters occur only once in ten years as a rule, and hence the present climatie conditions are incapable of producing the surface state of Patagonia as we now find it. If, however, the shingle were removed, and the climate slightly altered, so that the winters were more dry, and the winds in the summer perhaps a little stronger, we should have all the conditions necessary to produce that great flat country which must have existed in Patagonia in pre-shingle days. It appears that after a long period of dry, windy climate, probably in late tertiary times, the conditions began to alter, and that slowly, gradually, and without violence the shingle began to creep down over the pampas until it © reached the Atlantic. No carving or cutting of the country took place during this period, which probably lasted for a long time; in fact, at the end of the great shingle period the country was probably as flat and even as before. None of the great river valleys, in my opinion, existed then, and there were probably no bajos. As we have seen, the Cape Fairweather Bed is a marine formation, Frenron—Phystography and Glacial Geology of S. Patagonia. 195 which lies with slight unconformity on the tertiary rock, and is in turn overlain by the shingle; this arrangement, however, is not as simple as would at first sight appear, and on careful examination one will find that there is a more or less distinct transition layer between the Cape Fair- weather Bed and the shingle. In a typical locality, such as is seen m Cape Fairweather proper, we have first of all lying on the tertiary rock a layer of fifteen feet, consisting almost exclusively of oyster shells imbedded in a fine sandy matrix; over this layer is another in which there are many different varieties of sea-shells as well as oyster-shells ; the latter, in fact, are in a minority ; but in addition to shells—and this is of the greatest importance— there are to be found a fair number of typical pampa pebbles. This layer varies from two to three up to ten or fifteen feet in thickness, and the various shells, pebbles, etc., are held together by a hard rusty-brown material , in some places there is well-marked true bedding, with thin laminae of a hard stone; in others there is false bedding, as if of an old beach, while in others there seems to be a more or less homogeneous matrix holding the pebbles and shells together. The layer in which the pebbles assume a false bedded appearance, as if of an old beach, always comes between the shell-layer and the shingle, and tends to fade imperceptibly into both. Now this whole bed would give one the impression of being a shallow-water formation, and the fact that most of the shells are practically complete, with often both valves of the lamellibranchs in position, would show that it has suffered very little disturbance since it was laid down, and would indicate, at the time of its formation, calm shallow water. These facts are very significant, and tend to show that the transition from the pre-shingle to the shingle periods was gradual; in fact, when the first shingle began to be washed down from the mountains over the plains to the Atlantic coast the climatic conditions were such that these marine molluses flourished in abundance; also it is clear that calm weather prevailed, and that the floods which carried down the shingle were very mild. and had very little tearing up or grinding action, otherwise the shells mixed with the pebbles would have been all smashed up and ground to powder. Later on the shingle came down in greater quantities and with greater force, so that in the end the shell-layer became completely covered, and nothing but pure pebbles and sand remained on the surface. These facts also show that at the beginning of the shingle period the level of the main pampa was much lower than at present, and that Cape Fairweather, which is now almost five hundred feet high, was then slightly below or at the sea-level. 196 Scientific Proceedings, Royal Dublin Society. III. Indications that the Shingle was deposited during the Quaternary Ice Age. With regard to fixing the date of the first great shingle period, the facts that we have to go by are that the layer in question is found superficial to practically all the rocks of the district. The only rock overlying it is the lava, which is found over fairly large areas in many places; also it is unfossiliferous, and is unique, inasmuch as no parallel is found to it in any of the older formations. The higeledy-piggledy arrangement also gives one the idea of its having been carried down by a large volume of slowly running water. The large lava outpourings of this locality occurred during a long period, and were thrown out in great part subsequently to the beginning of the ~ ice age; so that, considering also the similarity of this shingle layer to many of the fluvio-glacial deposits found in the northern hemisphere. we are Justified in concluding that the great Patagonia pampa shingle layer belongs to the earliest portion of the Ice Age. The south of the Gallegos river is low-and undulatory, and does not seem to belong to the same horizon as the north, and there are many points of difference between the pampas in both localities. On the north, as we have seen, the pampa rises suddenly, or in a series of well-marked terraces, to about five hundred feet, and then extends away indefinitely in the form of a remarkably flat table-land, whereas on the south the plains begin only a few feet above the sea-level, and rise slowly and gradually. At first there are no apparent terraces on the south side, but a little up the Gallegos Rio Chico a terrace appears which would seem, on being traced, to correspond to the lowest terrace along the Gallegos river, and a little further still a second terrace is in many instances found. On the south bank of the Gallegos river, however, about thirty miles from its mouth, one meets once more the typical high or first pampa level, extending down in a V-shaped manner between the Gallegos river and the Rio Chico, or, more correctly speaking, in a long narrow strip a few leagues wide running down along and parallel to the former river, and separated from it by a broad terrace, about two miles wide, and the low flat of the river valley. On carefully examining this strip, one can see that it possesses all the characteristics of the pampa, such as extreme levelness of the summit, and, where eaten into, flat-topped hills. Already we have seen that at the time when the shingle began to come down, the level of the main pampas was five hundred feet lower than it is at the present day. At that time all the land to the south of this pampa line would be under water, and consequently would not be subjected to the great planing down action of the wind; this Prnron—Physiography and Glucial Geology of S. Patagonia. 197 would account for its undulatory surface, and, as we shall see later, its lack of terraces. Nevertheless, although the levels differ to the north and south of this line, the superficial formations would seem to be the same, namely, typical shingle overlying tertiary rock. Now, when we leave the Rio Chico, and travel southward towards the Straits of Magellan, after a few miles we notice a uew feature in addition to the shingle beginning to appear; this is an occasional great erratic boulder." At first such boulders are few, but after a league or two they become more numerous. Some of these erratics are found on the surface, some are half way out, whereas some are practically buried, with only their heads showing. Now, these erratics have the same compositions as the pebbles in the shingle; even amongst the lava sheets, practically none of them consist of lava,and there is a complete gradation in size from the smallest pebble up to the largest block; consequently, one is led to believe that they are part and parcel of the same formation as the shingle, and have come from the same locality. The erratics extend right to the Atlantic coast, ten miles south of the mouth of the Gallegos river, and, like the shingle, they are found equally well marked on the highest hills and in the lowest valleys. A line drawn through southern Patagonia on the Argentine side from a point on the Atlantic coast, about twelve miles north of the fifty-second parallel of lati- tude due west, to a point about twelve miles to the eastward of the seventy-first west meridian, and then curving slowly to the north, following finally up along and parallel to the cordilleras, would more or less mark the limit of the large erratic boulders. This line is not quite straight, and there are minor curves in it, which, however, are of no great extent. I have traced it as well as I can on the map. On going further to the north, along the chain of the Andes, one finds erratics extending a certain distance into the plains, but not to anything like the same distance as in the region which I have indicated ; in fact, this very abrupt extension outwards towards the Atlantic of the large erratic boulders to the south of the Gallegos river is a very remarkable fact, especially when we consider that the mountains to the west of this district are insignificant as regards height in comparison with those to the north and south. Now, although the pampas to the south of the Gallegos river, near its mouth, belong to a lower level than those to the north, yet further to the west they both merge with one another, so that at a point sixty miles from the coast all the pampas, both north and south, are on the same level; here 1 Compare B. Hatcher, American Journal-of Science, ser. 4, vol. iv, p. 348, 1897. 198 Scientifie Proceedings, Royal Dublin Society. the great erratics lie in and on the shingle belonging to the first pampa level. All the shingle extending from the south and west towards the north and east over the high pampas is one continuous layer, and belongs to the same formation, and may be divided by the line I have indicated into two divisions, a shingle area and a shingle erratic area. This evidence will confirm the view already put forward that the shingle layer belongs to the Ice Age. I have, moreover, carefully examined the erratic boulder area, and found abundance of scratched stones in it, like those found in the boulder clay of Europe. The limiting line of the erratics probably marks the former termination of a great mass of moving ice, and the further extension of the shingle over the plains was possibly due to the action of the multitudinous streams which ran away from the front of this ice-sheet. The chain of the Andes as prolonged through Tierra del Fuego and the islands to the south curves round towards the east in a hook-like manner. he limiting line of the large boulders corresponds in a striking degree with that of the Andes. These facts tend to indicate that during the first Glacial Age the snow and ice which are now limited to the top of the mountains descended some sixty miles into the pampas in the form of a uniform mass, moving outward all the time, and melting away along the line which I have tried to indicate. The ice as it descended caused considerable disruption and erosion of the mountain tops and hills, and carried with it an immense amount of broken debris, which it deposited on the lower lands; as the ice melted, the abun- dant streams formed from it carried this material further, rounding off the pebbles as they travelled, and eventually spreading them out over the flattened plains. Proceeding to the north from the lower portion of the Rio Coyle, one finds first an area of terraces extending back from the river for distances in different places up to six or even ten miles. About ten miles further across the pampa is another scarp-line, about one hundred and fifty feet high, leading up to another pampa, equally as flat as the one we have just left. The line of this pampa scarp differs from those of the terraces in that it does not run parallel to the river, and it has no corresponding scarp equal to it in height on the south side. It is also much too far back to be connected with the river. After travelling another thirty miles or so northward we come to one more scarp, which brings us on to yet another pampa. Now these two latter pampas, in addition to what I have mentioned, also differ from the terraces, in that they possess all the characteristic flatness which I have alluded to; in fact, they are pampas just as truly as that which we saw north of the Gallegos river. Hence to the north of the Coyle, unlike the same side of the Gallegos river, there are, in addition to the terraces, three distinct Frenron—Physiography and Glacial Geology of S. Patagonia. 199 pampas. Now, as we have already seen, the sudden drop from high to low level at the mouth of the Gallegos river probably marks the remains of an old coast-line; so, taking all the facts into consideration, I am inclined to believe that these latter steps belong to the pre-shingle period, and are remains of old coast-lines also. They were formed well back in tertiary times, when the high winds and dry climate prevailed, and they show a steady tendency towards elevation of the land, which, however, was inter- mittent in its action. Passing now along the north side of the Gallegos river, towards the west, we find that, sixty miles from the coast, the typical pampa still exists in all its flatness, even up to the very edges of the deep cafiadones. Another fifteen miles, however, brings about a complete change ; here the country begins to rise suddenly fifty to a hundred feet or so, and then goes on for another sixty miles before it reaches the mountains. The surface of this latter portion is completely changed from the pampas below, in that the flatness disappears and an undulatory condition takes its place. In addition, however, great stones appear here also; in fact, we find that this sudden rise in the level of the pampa marks the line I have already indicated as the termination of the erratic boulders. This line of junction between the flat pampa and the undulatory land is as a rule clear to the north of the Gallegos river, and one can trace it without much difficulty. To the south of the river, however, it is not so clear, as two actions have apparently come into play since the great shingle period which have complicated matters: one of these is the elevation of the southern coast-line, and the other is the enormous erosive action of the great river floods. The boulder shingle deposit is, undoubtedly, of glacial origin; and here it would seem that the great undulatory boulder area, the shingle of which runs up to a hundred feet in thickness, is really the remains of a great terminal moraine, and that it marks the limit of the advance of the ice during the first great ice stage. IV. The Recent Volcanic Series. About eighteen miles to the west of the port of Gallegos on the north bank of the river, we find the first evidence of former volcanic activity, and from this point for over a hundred miles in a westerly and south-westerly direction many extensive sheets of lava are to be found spread out over the surface of the plains. There is no very large volcano in this portion of the country ; but the number of small cones is immense, and they can be seen dotted all over the pampas as far as one’s eye can reach from horizon to horizon. This lava must have resulted from successive eruptions, occurring 200 Scientific Proceedings, Royal Dublin Society. during a very long period, as we have cones of all ages, from very weathered ones to others apparently quite recent.! We may begin with a short description of the most recent lava onbpouriig which I have found in this part of the country. The Cerro Diablo is a comparatively small and insignificant cone, only a few hundred feet higher than the surrounding plain; its summit consists to a large extent of scoriz and finely divided cinders; the latter exist in such quantities as to render the ascent rather difficult. Both scoriz and cinders have a very recent appearance, and give one the impression that the volcano was active at no distant date. The lava-flow from this crater is one of the most extensive in the country; it is fully fifteen miles in length, in many places upwards of fifty feet thick some miles from the cone, and nearer to the latter much more. It runs in all directions over the plain from the Cerro Diablo, and must have been poured out in a very liquid state, as it is found flowing down the small cafiadones in minor streams, even nine miles from its point of exit from the vent. The surface is covered by a kind of crust, varying from six inches up to many feet in thickness. Between this crust and the lava beneath there are empty spaces, extending here and there in all directions ; these spaces are often so large that several people could go down into them and stand erect, with a roof of lava over their heads. The lava is deeply fissured everywhere, huge crevasses gaping open, and extending many feet down into the rock. In many cases large portions of the crust have fallen in, revealing deep, gaping holes; in other places, again, the surface is practically unbroken. However, one of the most important points in our present study is that the original cuticle of the lava is absolutely intact and unabraded in the whole of this area; 1b has a recent appearance, and shows no sign of patination. ‘There is no other deposit overlying any portion of this lava-sheet, and, with the exception of a few bushes growing here and there in the corners, and some ferns, &¢., in the cracks, there is practically no vegetation. After carefully examining this lava-sheet, and walking over a considerable portion of it during several days, I have never found a pebble in any place more than ten yards from its edge. At a place called Pali Aike, situated at the side of the Rio Chico, this lava-sheet pours down the cafiadones and out into the river valley. Now, I hope presently to bring evidence to show that the lowest bed of the river valley, with the cafiadones running into it, represents the last 1 On the characters of the lavas, see Olof Backstrom, Petrographische Beschreibung einiger Basalte von Patagonien, &c., Bull. Geol. Inst. of Univ. of Upsala, vol. xiii, p. 11d. (1915.) Funron—Physiography and Glacial Geology of S. Patagonia. 201 erosive effort of the last phase of the Ice Age. If this is the case, it will be evident that the lava outpouring of the Cerro Diablo is at least post-Glacial. The other lava-sheets found on the pampas differ considerably from the above. There are no gaping fissures and no open holes to be seen anywhere in the rock, all such having been long ago filled up and obliterated; and there is no intact cuticle, nothing being found but fractured surfaces. Also where they reach the edge of a river valley or cafiadon the sheets break off abruptly, and fall in the form of cliffs. All over such lava-sheets a fair amount of recent soil occurs, with abundance of good grass and vegetation ; in fact, the hommocky masses already referred to are in many places only seen here and there jutting through the grass. Except in one or two trifling cases, which I will return to later, no shingle is found on any of the lava-sheets; the soil in which the grass and bushes grow is a rich, black loam, and does not contain any pebbles in its matrix. In the areas between the lava-sheets, where large erratics are lying about here and there all over the surface, none of these erratics consists of lava. Conse- quently the period of volcanic activity must have had its origin subsequent to the termination of the first great Ice Period. In the summer of 1914 I was fortunate enough to find two places where the typical pampa shingle could be seen under the lava, and, more strange still, in one of these places I found that between the shingle and the lava there was about one hundred and fifty feet or more of a well-stratified rock. This latter, which is of the greatest interest, I hope to describe in a subsequent section of this paper. It was formed long after the shingle was deposited on the pampas. Con- siderable erosion has occurred in many of the early lava-sheets since their formation. In one place a cafiadon, some hundreds of yards wide, and eighty to a hundred feet deep, has been cut through them. A feature of importance in the present study is that the lava-blocks are often to be found arranged in a peculiar sloping manner, rising towards the east; and on the slope facing west there are usually to be found a number of grooves running up along the stones in an oblique manner. The space between the grooves has always a smooth, ground-down appearance, and the surface is often concave. This grooving and the peculiar lie of the stones are well marked in Pl. VI, fig. 2. This peculiarity of the lava-blocks is found in widely different localities ; the grooves always run in the same oblique manner from east to west, and always on the sides of the slopes facing west. ‘The grooves may be seen on the north or south sides of the stones, but are never found on the eastern ends. The direction varies from due east to east-north- east. During several years of residence, I have visited every possible lava- sheet that I could approach in the Gallegos district within eighty miles of the 202 Scientific Proceedings, Royal Dublin Soctety. town, and have travelled many leagues over these rocks. In every place I have studied this grooving, and, as I have said, I have never failed to find it in any locality where there is lava, except in the case of the Cerro Diablo. V. Indications of Strong Westerly Winds. I have had many explanations offered to me as possible causes of the grooving of the stones mentioned in the last section of this paper. The only one, however, which is worthy of note is the action of high westerly winds carrying sand, grit, &e. One of the most characteristic climatic features at present in Patagonia, particularly during the summer months, is the strong dry winds which blow from the west. These winds blow as a rule about one to two days each week during the summer and autumn. As I have already mentioned, a large amount of sand and dust is carried along, and I have noticed surfaces of dried mud in a valley cut in grooves in the same manner as the stones after a few months of windy weather. It is noteworthy that the grooves always run up the westerly faces of the stones in an oblique manner towards the east, as would result from westerly winds blowing in dry weather and carrying gritty particles. The fact also that grooves are found down in narrow spaces between stones is strongly in favour of wind action. If, however, we examine recent lava- sheets, such as those from the Cerro Diablo, or if we examine the large blocks which have fallen down from the most recent lava cliffs bounding the present beds of the river valleys, we may search in vain for grooving. I have found abundance of grooving on all the higher levels, but on the lowest and last levels of the river. valleys, although I have searched extensively, I have never seen it in its typical and characteristic form. Another point of considerable importance is that in hundreds of localities where this grooving is found a peculiar rusty patination of the surface occurs, in marked contrast with the clean black stone revealed when the grooved surface is chipped off with a hammer. The grooved surface is now becoming weathered off the stones and not weathered on. In many localities the grooved surfaces have been overgrown with lichens, mosses, and other vegetable matter ; and they are often hidden away behind bushes and shrubs, which are growing with ease in front. The evidence, consequently, seems to point to the fact that this grooved condition of the stones was brought about during some former geological period, and was not produced since the river beds were cut down to their present levels. » Frnton—Physiography and Glacial Geology of S. Patagonia. 203 From the above we see that at some period during the time when the river valleys were being cut out, and while the rivers were still exercising an erosive action, high dry westerly winds blew over the country with great violence. Volumes of gritty particles were carried along, and all rock surfaces exposed were extensively eroded. In some places there are large, hard. basaltic fragments, which must have lost upwards of a third of their bulk in this manner, and in more than one place a stone occurs which, from its shape and general direction of its grain, must have had fully a ton weight or more of its substance removed by wind action. On going through the volcanic area, I was struck by the fact that quite a number of the cones had a long slope upwards in one direction, and a steeper descent on the other side; for instance, in the chain of the North Hills, out of eight hills, two had this slope. I have also noticed this feature in several other places, and the highest side lies always towards the east. The central craters in these sloping cones are often found to a large extent filled up, and the edges of the rims are as a rule rounded. I now conclude that the high wind was the cause of this phenomenon also. We have already seen that we have reason to believe that the great flattened condition of the pampas was produced by high winds in pre-glacial times. We see now that during inter-glacial times there were periods also when high winds prevailed. Before concluding, I wish to state one objection. At a place on the Gallegos Rio Chico, situated about thirty-five miles from its mouth, the river curves round and runs in a north-westerly direction. The terrace on the left bank is here almost a hundred feet above the bed of the river valley, and it is capped with lava, which falls in the form of a clean cliff a short distance from the top. On the summit of this cliff the usual wind-grooving runs from behind right out to the edge, and it is well marked all along. When climbing this cliff one morning, I was suddenly startled by seeing the typical grooves running down its face almost perpendicularly. These facts are a difficulty for the wind theory, as it is impossible to conceive that hard particles travelling along the top with such force as to cut the stones would not, by virtue of their momentum, be carried clear of the edge in their descent. I will return to this subject later on, when I hope to show that far outside the limits of the great moving ice-sheet there was also another form of local ice-action at work. of VATE Origin of the Terraces in Spring Floods of the Ice Age. If we take for examination any typical portion of one of the river valleys, such as that of the Gallegos river, the floor of the valley is, perhaps, a mile 204 Scientific Proceedings, Royal Dublin Society wide, in some places more, in some less, and the surface is practically level, no portion of it being more than ten to fifteen feet above the river. The sides of this valley are bounded by steep rising banks, probably eighty feet in height, which mark the beginnings of the first terraces. The floor is in most places covered with a rich black soil, on which grows an abundance of rank vegetation, except in these places which have been eaten down from over-stocking with animals. Small lakes occur here and there, and the river winds about through the valley in a very sluggish manner, now on one side and now on the other. Patches of swamp lie in different directions, yet in no place is the river very deep; in fact, almost every half mile or so it is capable of being forded. ‘The river is always shifting its position through the valley, and as it cuts a new course the old one becomes first a lagoon, afterwards a swamp, and finally becomes obliterated by vegetation, mud, and various kinds of rubbish which have fallen in. The river during the ordinary dry weather in summer runs between perpendicular banks, is perhaps about five feet below the level of the main valley, and is anything from fifty to a hundred yards in breadth. If, however, we visit it during the month of September, which is the spring of the year in Patagonia, we shall probably find that it has risen considerably, possibly flowing over the banks in places and flooding parts of the valley. These spring floods are a marked feature in this country; they vary much in different years and in different rivers, being sometimes hardly perceptible, while at other times they fill the whole valley from side to side with one great river. Seven or eight years may pass during which there are practically no floods or only very slight ones, and this period may be followed by one or two years when the whole valley may be filled. When the river rises sufficiently high to flood the whole valley, in about a week or two it usually recedes once more between its banks ; it may, however, remain in a flooded condition in its own channel for more than amonth. After the flood has gone down it is surprising how little change has taken place. We may find a few bits of twig deposited here and there along the sides, and some mud in odd places; but in most localities the valley is as it was before; in fact, if any alteration has taken place, it is that the river valley now contains somewhat more soil than it had formerly. These spring floods are always the result of the melting of the snow on the high lands towards the west, and are, consequently, greatest in those years when there is most snow. Any erosion of the valley during flood-time is more than compensated for by the deposit carried down from above, and, as erosion by wind does not take place on account of the amount of vegetation which protects the surface, we have reason to believe that under the present conditions the level of the valley is slowly rising. At any rate, as climatic Frnron—Phystography and Glacial Geology of S. Patagonia. 205 conditions exist at present, no great erosion is taking place in the river valleys; and it is quite possible, as far as we can see, that the beds of the rivers stood practically at the same level many thousands of years ago as they do to-day. The hilltops and mountains to the west are, of course, becoming worn away, but the lowlands and valleys, wherever the surface is protected by vegetation, are probably, if anything, rising. I have found abundance of evidence of this latter fact in many places, in the shape of bones of recent animals, such as guanaco, &e., some feet below the surface of the soil in the valleys. If we dig down through any of the river valleys, some distance up from the sea, we pass through a considerable depth of shingle and sand, very similar to what we find on the pampa, perhaps fifteen or twenty feet before we come on the actual tertiary rock-floor.. We also find in places a considerable depth of mud and black soil where the gradient of the valley is slight; in a tributary dry cafiadon, where the gradient is more marked, there is not as a rule much black soil or mud; but even here the floor of the cafiadon is covered by a layer of, perhaps, fifteen to twenty feet of sandy shingle. In very few cases—and this is a very important fact— is the tertiary rock-floor of the cafiadon exposed to view, the only exceptions being those cafiadones which have very steep gradients, and which drain very extensive areas. All the river valleys and cafiadones in southern Patagonia, with the exception of the Santa Cruz river, are subject to occa- sional spring floods. Those floods vary in different localities, and in different years, but they are always found at times right through the country. The Santa Cruz river is an exception, as it does not begin to rise until the summer, and does not reach its height till well into the autumn. The reason for this is that this river takes its origin in some large lakes situated well in between the cordilleras, and, as the snow on the latter melts much later than that which falls on the plains, the floods are consequently delayed. Most of the terminal branches of the Gallegos river rise in the plains; one or two of them, however, extend as far as the cordilleras, and consequently occasional summer or even autumn floods occur in the Gallegos river after periods of hot weather, when additional snow has melted on the mountains. The local rainfall has very little effect on the floods in thiscountry. Even after several days of heavy rain, such as occur occasionally, although the small streams may be slightly increased in amount, the rivers show practically no rise. Rain never fills up the lagoons, and experienced farmers will tell you that without snow they never have their land properly supplied with water. Any rain falling on the shingle-surface of the country will sink in unless the surface is made water-tight by being frozen. Now, when the snow begins to melt in the spring, the superficial layer of the soil is actually frozen, so that SOCIENT. PROC. R.D.S., VOL. XVI, NO. XIX. 2H 206 Scientific Proceedings, Royal Dublin Society. the water runs over it and escapes, some of it into the lagoons and the rest into the cafiadones, through which it eventually makes its way to the rivers. Thus the essential factor in the production of floods is not rain, but snow, and these floods are always found in the spring of the year, when the snow melts. The snow which falls on the plains and lowlands is also most productive of floods. If the climatic conditions as they exist at present became changed in such a way that the winter snows were increased to four or five times their present amount, the spring floods would be correspondingly increased, and, instead of having a depositing action, as they have at present, they would first clear all the recent soil out of the valley, and next all the gravel and sand, and finally the bed-rock would be cut into, so that the river valley would once more become deepened, and probably new terraces would come into existence. This sequence of events, namely, a change from a period of mild climate such as exists at present to one when the snow-fall was con- siderably increased and the spring floods were hugely augmented, probably occurred several times in the past, and it is likely that all the terraced river valleys were cut down intermittently in this way. It seems, then, that the present spring floods are only the shrunken remnants of what once existed on a grand scale, and that all the great river valleys and cafiadones were originally cut out by the action of spring floods, due to the melting of large masses of snow and ice. We are now in a position to adopt certain rough divisions of the past history of Southern Patagonia as far back as our story goes. We may divide it into, first, a pre-shingle period, when wind was the dominant eroding force ; second, a shingle period, which probably was identical with the first great advance of the ice—no river valleys or cafiadones existed then ; third, a long period of erosion by water, which was subject to con- siderable oscillations due to alternations in the climatic conditions; and, fourth, a genial period, which is still going on, and which has probably existed for some thousands of years. VIL. Erosion subsequent to the outpouring of the Lava. The surface of a river terrace is covered with sand and shingle, very ‘similar to that which is found on the pampa, of about the same thickness, and distributed practically as evenly. If, however, a catiadon or gully which cuts into one of these terraces be examined, it will be found that this shingle- layer is somewhat different. A cutting running into one of the terraces near the mouth of the Coyle river shows marked current bedding, the pebbles being sorted out into irregular layers, and the spaces between being filled Frnron—Physiography and Glacial Geology of S. Patagonia. 207 with sandy earth ; the latter also exhibits false bedding in thin layers. This arrangement of the terrace shingle gives one the impression of a great, fairly rapidly runing river carrying with it a large amount of silt. The terrace which we are now considering, with its fellow on the opposite side, would make a river valley perhaps five or six miles in breadth; and when we consider that the whole of this horizon is covered with a layer of sandy shingle, averaging about fifteen feet in thickness, we can form an idea of the size of the river and the maenitude of the floods which existed in former times. This terrace is about two hundred feet below the main pampa level; conse- quently the river valley of which it once formed the floor must bave been cut down very considerably through tertiary rocks before this shingle was deposited. Now on this same terrace we find an occasional flat-topped hill, rising to the level of the main pampa, and the hill is covered with pampa shingle, which shows. that the cutting-down action of the valley must have occurred subsequent to the pampa-shingle period, and the terrace shingle must be of far later date than the first Ice Age. The first phase was a cutting one, when the floods were enormous, and all the materials eroded from the valleys were swept to sea; the second occurred when the floods had con- siderably abated and the cutting action had been replaced by deposition. The map of this part of Patagonia shows that from the top of the highest ridge of the Andes to the Atlantic is not more than one hundred and eighty miles, and, as this narrow strip is intersected every thirty miles or so by a huge river valley with branching catiadones, it seems wonderful that such an area could have collected enough water to cause the floods we have been considering. if, however, this country was once covered with a great mantle of snow and ice which had been accumulating for centuries, and when this had reached its climax the climate rapidly began to get warmer, enormous floods would be prevalent every spring; this flood-epoch would last until the accumulated snow and ice would disappear in the spring andsummer. Now at first during this period the pampas and highlands would be protected by a deep layer of ice; consequently they would not suffer much erosion, but the whole force of the floods would be concentrated on the floors and sides of the river valleys, with the result that the latter would be eaten down and the debris swept away to sea. Later on, when the climate had become still more approximated to what we have at the present day, and the pampas and highlands had become denuded of their protecting mantle of ice, the floods would begin to wash down huge quantities of shingle and sand into the river valleys. The force of the floods having in the meantime considerably abated, this deposit would accumulate until the floods died down to their present-day insigni- ficance, when, the climate having become mild, vegetation would spring up 2H2 208 Scientific Proceedings, Royal Dublin Society. all over the country and protect the surface againstfurther erosion. The pampas ure sometimes frozen when the water is running freely in the river valleys; and I may mention a very striking example of this which I myself witnessed. On the north side of the Gallegos river, almost in front of the town, there is a huge cafiadon which runs back about two and ahalf miles; it is very broad, and has several tributary cafiadones, one of which, after rising rather abruptly, ends on the pampa. Inthe end of June, 1913, I was going to Coyle, and left the river-side at 9 am. A misty rain was then falling, and quite a fair- sized little river was running down the cafiadon; there was no snow below, and everything was wet and sloppy. The stream continued, although gradually diminishing in size, as I went up, until 1 got within fifty feet of the top, when it ended in snow and ice; the mist had by this time turned into sleet, and J was surprised on going about two hundred yards further and reaching the pampa to find that this latter had in its turn changed to snow. I wasalsosurprised to find the pampa dry, hard, and covered with snow and ice, with no sign of thaw. My horses here gave in, and I was compelled to return; and, when I came a half a mile down the cafiadon, I found every- thing was as before—the same running stream, the same mud, and the same misty rain. Itis also a well-known fact that after a severe winter the snow and ice will have completely disappeared from the pampas near the sea fully a month, if not more, before they begin to melt on the same levels sixty to seventy miles inland. These facts are very significant, and indicate that when the great ice mantle began to melt it did so first along the sea-coast and in some low-level bay or bight. This bay would be cut out first to a considerable extent, and would eventually mark the starting-point of a river which would cut its way back into the high ground as the ice receded. According to this method of reasoning, it would seem that the shingle covering the river terrace we have been considering came down towards the end of a flocd period, and was deposited at a time which came immediately before a phase of mild climate such as exists to-day. All along the Gallegos river valley we find numerous lava-sheets; some of them are on the pampa, but most of them are found capping the terraces, even down to the lowest. Now these lava-sheets, except where they are broken up by former ice or water action, show very even and compact distribution, and seem to have been poured out over a dry-land surface ; for, had there been floods at the time of the volcanic outpourings, the lava would have been immediately cooled as it met the water, forming huge frizzled-up heaps, and would not have become spread out in thin, hard, compact sheets. Now, except in one instance in the cafiadon Guer Aike, I have not met with any deposit of shingle on the surface of any of these lava-sheets—at Fenron—Physiography and Glacial Geology of 8. Patagonia. 209 least in the lower seventy miles of the river valley; on the other hand, the lava lies on the shingle in more than one place, and I have reason to believe that it does so in every instance. This would prove the lava to be the younger rock of the two, so that, if the lava corresponds with a resting phase of the river valley between two phases of ice and flood, the shingle must have come at the end of the flood or before the resting phase. ‘The history of a river valley in this region may be divided into four periods as follows :— first, an ice phase; second, a flood-cutting phase; third, a flood-depositing phase: and fourth, a resting phase. If now each terrace marks the base of a former river valley, we should expect this sequence of events to have been repeated as often as there are river terraces. The further discussion of this very important and fascinating subject I must leave until I am in a position to obtain additional evidence. It seems clear that a great period of ice-advance was followed by a long period of erosion by water; this flood period was intimately connected with glacial action, and the huge spring and summer floods were due to a much greater winter snow-fall in former times. There is a terrace along the Gallegos river, which is covered for a considerable distance (half a mile or more) with a mass of broken debris, consisting chiefly of huge angular fragments of basalt up to six feet or more in length. Packed between these basaltic fragments, will be found a certain amount of pampa shingle, many of the stones of which are wedged in such a way that it is clear that they came there at the same time and with the basaltic fragments. This mass of debris lies on a lava-sheet, which shows a different composition and structure from the great basaltic blocks above, and the whole mass of this deposit was evidently carried from another locality, and heaped up over the lava-sheet in question. A little further up the valley the river makes a well-marked bend round to the north, and here, on the terrace, is a peculiar knob standing by itself. This is the remnant of a lava-sheet, which once covered this terrace up to the level of its top. It is about thirty feet in height, and is removed upwards of half a mile from the nearest edge of the sheet, of which it originally formed a part. Hence extensive surface-erosion of lava occurred here at one time, and it is more than probable that it was from this surface that the lava blocks were derived which we saw heaped up on the lava-sheet below. We see from this that during the cutting down of the river valleys there are two forms of erosion of lava—namely, the ordinary erosion of the sides of the sheet during the cutting down and deepening of the river valleys, and also a form of surface- erosion, where great areas of the surface of a sheet were broken up and carried away in huge fragments. A few miles down the river, at a place 210 Scientifie Proceedings, Royal Dublin Society. called Buitreras, there are some wonderful examples of lava erosion. In a general bird’s-eye view of the river valley, looking westwards from the southern edge of the river, we see in the distance a volcanic cone forming part of the edge of the river valley. It has been eut clean in two through its centre, ike an apple divided through its middle by a knife. Pl. VII, fig. 3, gives a nearer view of this cone; it is about three hundred feet high, and of the southern half nothing whatsoever remains. Across the river valley, however, which is here a few hundred yards wide, the edge of a lava-sheet is found, which was probably at one time continuous with this cone. This means that the river valley, which has here been cut down toa depth of over eighty feet, has required for its formation the removal of not only a considerable lava-sheet, but also of half of a voleano cone three hundred feet high. Now, if we follow down the Gallegos river to its mouth some fifty miles away, we find practically not a single block of lava bigger than a few inches in length, and even these are few and far between. The river valley is throughout covered with the usual shingle, sand, and silt. Millions of tons of solid hard basalt have been removed from this area in some com- paratively recent geological time, and there is no trace of them in the river valley below. ‘The terrace which has been here cut through is the lowest in the Gallegos river valley, and consequently represents the erosive action, if not of the latest, at least of one of the latest flood periods. The basalt here is a very hard stone, and we can only wonder what has become of the millions of tons which have been torn out of this place and swept away. They may be buried in the floor of the river valley; but, if so, heaps of them should show above tle surface in some localities. This is not the case, and we may consequently conclude that these great lava blocks have either been ground into fine material or carried bodily away to sea. It is difficult to imagine that these huge fragments, many of them weighing many tons, could have been ground up into fine shingle during the last flood period. I will here quote from my note-bock: “25th March, 1915. The western end of the Buitreras table, although consisting of bed-rock (sheet of basalt), has its surface very broken. Curious hollows, ending blindly towards the west, but often opening towards the east, in the form of catiadones are found, and some small hollows are found without any opening. Some of these canadones, which are always blind towards the west, are cut deeply back into the lava towards the west, and have some small amount of shingle on their bottoms. Towards the eastern end of the western half of this table there is extensive surface- erosion of the lava; the latter seems to have been eaten into by the caiadones mentioned ; on the sides of some of these there are actual cliffs of lava up to twenty feet high.” The Buitreras table here mentioned is an elliptical- Frnton—Physiography and Glacial Geology of S. Patagonia, 211 shaped flat lava-topped piece of ground, situated on the south bank of the (rallegos river, opposite the Buitreras cone. It is about two miles long by, perhaps, half a mile broad. It is bounded on the north by the Gallegos river, and on the south by a large caiiadon over two hundred yards in breadth, which separates it from another extensive lava-sheet to the south. It falls abruptly to the west, but gradually to the east. The lava ridge is here about ninety feet high above the river valley, and the lava-sheet, which is still practically intact on the western end, is completely broken up on the eastern portion. On carefully examining it, we will find that a peculiar ridge of more or less intact lava extends down as a fringe along the northern edge of the table for a little distance. One of the above-mentioned blind cafiadones extends up behind this ridge to the south. The Buitreras table has thus suffered a considerable amount of surface-erosion during late glacial times ; and, although the erosion is almost altogether confined to the eastern two-thirds of the table, the eroded area encroaches on the intact western end in the form of a series of bights into the lava-sheet. There are also some hollows eaten out of the latter, which end blindly in all directions. What then was the agent which produced this condition of surface-erosion, which is by no means limited to this particular place? There are no signs of sea- shells, or any form of beach ; besides, all the great basaltic fragments are, as a rule, angular, and not rounded as they would have been if the sea had been playing on them for any length of time. A great flood might have been the cause ; but such is difficult to imagine. What then was the state of affairs in the Buitreras area in the particular time that we are considering ? I have imagined to myself a great mass of snow gradually settling by pressure into ice, accumulating for centuries generally over the whole country, particularly on the high pampas. This went on to such an extent that the river valleys became obliterated by ice. The country was so flat that this ice-sheet had no tendency to move, but simply lay there as a stagnant mass in the area in which it was formed. At least its movement was so slight that it caused no moraine formation, and no scratching of the stones. Asa consequence of the same stationary condition, the ice-mass was clean, and held practically no picked-up material, except perhaps a little in its lower stratum. Then occurred a change of climate, which set in rather rapidly, with the result that this great ice-sheet began to melt away. Hot summers set in, during which huge floods occurred, masses of water poured down through large crevasses, producing extensive sub-glacial erosion, and from the eastern fringe of the ice-mass immense blocks of ice became broken off, and floated away to sea on the waters. As they broke away, they carried with them large pieces of the lava, which separated along with them. When the 212 Scientific Proceedings, Royal Dublin Society. floods were at their height, these lava-laden icebergs floated away to sea, as I have said. However, when the floods diminished, most of the icebergs grounded, and melted away «i situ, shedding their lava debris. This explanation seems the only one that accounts for the condition of affairs found in this locality. Tf we now pass downwards towards the east, a few miles from the Buitreras, the high pampa level to the south is seen to be capped to a con- siderable extent with sheets of lava, which have been poured out from small cones which surmount it close by. After we pass along the highest cliff, or barranea, as it is called, for about two and a half miles below the Buitreras, this lava-topped cliff takes a sudden bend to the south, at right angles to the line of the river-valley. It runs in for about half a mile, and again runs towards the east. During this half-mile course, the lava-cliff faces east, and consequently has its back up stream. For several years I could not under- stand how even a huge river filling the whole valley, here about five miles wide, could turn this corner and undermine the lava-sheet to such an extent as to cause this cliff. There is no sign whatever of a shingle beach at the base, and all the fragments are angular, and show no signs of rounding, as they would if the sea had been acting on them for a long period. On the contrary, the whole place gives the impression of the cliff having been cut down by a large mass of water pouring over it from behind. Now, this particular cliff exists on the northern edge of the narrow tongue of pampa which I have already mentioned as extending down between the Gallegos river and the Rio Chico. This tongue is extensively cut into, and forms altogether a very small collecting area. The greatest rainfall which could be imagined would not produce sufficient running water to cut down these lava-sheets in the manner which is found to have taken place. A huge accumulation of snow, piled hundreds of feet high on top of this table, and then rapidly melting, might, however, produce sufficient running water to bring about these changes. Now, if the process outlined above was the causative fact of the surface conditions, the result would be that during a hot summer large volumes of water would be formed by melting all over the ice-sheet. This water would flow towards the east, and would fall over its limiting edge in an extensive series of cascades. Now, at first, while the ice-sheet was still very thick, these cascades, falling as they would from a considerable height, would resemble a number of miniature Niagaras, and would produce much erosion of the surface of the land on which they fall. They would also tend to hasten the backward recession of the ice-sheet by breaking away its edge. Huge fissures extending downwards through the ice as large blocks of the Frenton— Physiography and Glacial Geology of S. Patagonia. 213 latter became broken away and detached would also lead to local attack on the underlying lava, which might in that way become very much broken. A subsequent severe winter might now freeze lava-blocks and ice into one compact mass to such an extent that the ensuing summer floods might pick up large pieces of it and float them away. As, however, the ice-sheet as it recedes is also all the time diminishing in thickness, a point will sooner or later be reached when the ice-sheet will become so attenuated that its floods are negligible, and have no further eroding action. Such a condition of affairs was probably reached over the Buitreras table when the edge of the ice-sheet reached the point already indicated near its western end. It must not be supposed, however, that the series of events which ultimately took place was as simple as I have outlined above. It is much more probable that during the whole action there were extensive oscillations of temperature, not, only from summer to winter, but also greater variations when several seasons of comparatively hot weather alternated with other long periods of cold. The idea which I have outlined seems to me the most likely explanation of the phenomena which I have observed. VIII. The Problem of the Bajos. As I said in the first section, if we travel any considerable distance over the pampas we are sure to come across an occasional great hollow which is not a river-valley. A typical example is the Bajo de las Tres Lagunas, on the second pampas to the north of the Coyle river. The pampa here is upwards of a thousand feet above the sea, and has all the characteristic flatness already noted, so much so that we are almosi on the edge of the bajo before we are aware of its existence. This bajo is elliptical in shape, with the long axis running east and west; it is about five miles long, two broad, and the base of it is four hundred feet below the general level of the plain. The western extremity is deeper than the eastern end, and the three salt lagoons are situated in the western half. At the sides of these lagoons, which in the summer and autumn are practically solid salt, there are to be seen a number of springs; some of these ooze up through the mud, and are covered with a dry crust, which renders them very dangerous, as animals often sink in, and are unable to extricate themselves. The eastern end of the bajo is found to slope more gradually than the rest of the circumference, which falls rather abruptly. Asa rule there are no very great cafiadones running into it; but a number of semi-cup-shaped bights cut into the rim all along. However, in one or two places a cafiadon, 214 Scientific Proceedings, Royal Dublin Society. half a mile to a mile in length, can be seen. Some of these caiiadones are more like arms of the bajo, as the main floor-level of the bajo extends into them for almost their whole extent, or at least the floor of these arms is often not much higher than the main floor. This bajo was formed by erosion, and not by subsidence, since the strata of tertiary rock crop out on the sides, and are clean cut off at their edges, as they are on the cliffs along the Atlantic coast. Asa rule, between the bights already mentioned ridges extend downwards and outwards into the bajo from the top. In some places these ridges form three irregular steps; these latter, however, although a rough attempt at terracing, are not anything like so well marked as the terraces along the river-valleys. In some places the ridges are fairly broad, whereas in others the sides of the bajo are cut every hundred metres or so by in-running caniadones. The lower terrace is in one place so cut that it shows two or three little rounded hills, with flat tops, these being so equal and even in height that they give the idea of their marking the level of the floor as it existed at a former time. The pampa on all sides of the top of the bajo is equally level, and on the east side there is no more sign of a hill than there is on the west, so that of the great mass of material which once filed up this hollow not a trace remains. The three salt lagoons, from which the bajo takes its name, only occupy a small portion of the floor, the remainder being covered with bush, grass, and other vegetation. The cafia- dones and bights, already described, leading into this bajo are very similar to the cafiadones found elsewhere; they are clothed in vegetation, and show, like the others, practically no signs of present-day erosion. Hence it is probable that the Bajo de las Tres Lagunas has not altered much during recent centuries. Now, all the characteristics of this bajo are to be found also, to a certain extent, in all the others. Half a mile to the eastern end of this great. valley there is another one, which is very much smaller in size and depth ; this latter is about three-quarters of a mile long, and about three hundred yards broad, and about eighty feet deep; at its western end a bare, gravelly patch of a few acres in extent marks the situation of a small lagoon, which exists, as a rule, only in spring and early summer. The centre of this patch is covered with dry mud, and there is no sign of vegetation on any portion of it. There is no attempt at terracing in this bajo, and it would seem to consist of one horizon only, although, like the other, its lower portion is the western end, and it slopes slightly upwards towards the east. All the floor not occupied by the bare patch already mentioned is covered by the same vegetation as is found elsewhere, and shows no sign of recent erosion. Everywhere over the pampas semi-bare patches occur, which are often only Fenron— Physiography and Glacial Geology of S. Patagonia. 210 a few feet below the main level; these may vary trom a hundred yards or so up to four or five hundred in breadth, and are covered by short, fine grass. They have often in their centres patches of bare shingle, and no bush is found growing in any portion of them. These patches mark the sites of spring lagoons, and their semi-bareness is due to the partial sterilization of the surface produced by the water, which lies on them at times for some months. These spring lagoons are only formed after severe winters, when there is heavy snow, and sometimes they are absent for a period of six or seven years. From this it would seem that we meet all grades of bajos, from the merest shallow semi-bare patches, only a few feet below the level of the main pampa, down to great hollows several leagues in length and hundreds of feet in depth. On the floor of some of these bajos we find a class of hill which rises abruptly in the west and falls gradually towards the east. These hills, of which there are a number in some bajos, are capped with a thin layer of shingle, and their tops would seem to mark a former horizon in the formation of the valley. Bajos are found, not only on the main pampa, but on al] the lower horizons, except the floors of the river-valleys. On the side of a bajo situated on the third terrace above the floor of the Coyle valley is a very well-marked terrace about sixty feet above the lagoon, which was, at the time of my visit, of pure salt; its top is as flat and its edge is as clean cut as any of those found on the sides of the river valleys. Barometrical reading showed the surface of the lagoon in the bajo to be only twenty feet above the Coyle river. It has been suggested that these bajos probably originated through a combination of wind and water action. The water remains in a slight hollow for a few months after the melting of the snow in the spring, and causes destruction of the vegetation which normally protects the surface. The soil beneath is thereby exposed, so that, when the water has dried up, the winds, which blow very strongly in summer, attack it, eat into it, and blow it away. The next season the lagoon is a little larger, more vegetation is destroyed, more earth is blown out, and the hollow becomes deeper. This goes on year after year until a huge bajo is formed. This explanation is very plausible, and seems to agree with the facts in many instances ; in fact, I may say that I have seen in more than one place this mode of action actually working. For instance, near the town of Gallegos at least two such places occur, where the dust and sand of the hollows is found to be heaped up on the eastern sides in the form of mounds. I also know one small bajo on the high pampa between Gallegos and Coyle where such a mound can be seen on the east side of the hollow. In the vast majority of instances, however, no mounds or elevations can be seen on the sides of the bajos, and no scooping out seems to be occurring to-day. 216 Scientifie Proceedings, Royal Dublin Society. The great difficulty in the way of accepting the foregoing theory in its entirety is the presence of the shingle; and it seems unthinkable to me that wind, even assisted by water, could produce this great erosion of a surface protected by a deposit of fifteen feet of pebbles, many of them up to eight or nine inches in length. Then it may also be asked why should this erosive action only affect isolated patches when the great portion of the area is untouched? he tertiary rock below is of a clayey nature, and quite insoluble; it would consequently not suffer any erosion under fifteen feet of shingle and sand. If by any chance certain patches of the pampa remained uncovered when the shingle was washed down from the first great ice-sheet, these patches would then be exposed to the action of wind and water, and: bajos would soon result. I was talking this matter over with a friend one evening, and he suggested that perhaps there might have been a few small hills remaining in the first instance which had resisted the great planing influence which was operating in pre-shingle times, and that these hills, being higher than the surrounding plains, were left uncovered when the shingle was washed over the country generally. This explanation would account for all the facts observed, and I now beg to put it forward, with all _ due reserve, as a working hypothesis. There is, however, one serious objec- tion to this idea: there is a hill, or rather a small tableland, situated on the main pampa to the north of the Gallegos river; it falls all round, particularly towards the west, to the extent of over forty feet, and yet it is covered by the same shingle, and to the same extent, as is found on the rest of the plains. It may consequently be asked why should slight elevation above the surface render the one set of hills immune from shingle deposit, when this plateau, which is fully forty feet above the main plain, is covered com- pletely ? There seems little doubt that some of the bajos must have had their origin well back in inter-glacial times, as can be seen from their huge extent, the fact that there is more than one horizon in them, and that, as climatic conditions prevail at present, they do not seem to be suffering any marked erosion. Since they are found on all the river-valley horizons, except the lowest, factors would seem to have been present for their produc- tion in all the inter-glacial periods; these factors, moreover, were evidently absent since the last retreat of the ice. There is yet another solution of the problem that I regard as the most satisfactory of all. ‘he stagnating ice-mass postulated in Section VII would give rise to localized and often broad-fronted waterfalls during its epoch of rapid melting. The bajo may be regarded as a large pot-hole, or, rather, a vast representative of the pools in a stream with occasional waterfalls. Such pools are often deep at the head and shallow at the foot, and a bajo fifteen Frnron—Physiography and Glucial Geology of S. Patagonia. 217 miles broad and thirty miles or more in length none the less presents striking analogies with the miniature hollows eroded in the floors of ordinary streams. It may be once more pointed out that the thickness of the mass of stagnating ice resting on the pampa provideda very effective head for the water pouring in Niagara-like cascades from its surface. I have also found in one place a series of bajos extending in an irregular line from west to east for thirty or more miles. In some instances the bajos are found opening one into the other, while in others a narrow strip of intact pampa may separate them. Such series often occur, I am informed, in other parts of the country also. IX. The Buitreras Bed. As I have already said, the lava in this part of Patagonia is practically always found lying superficial to the shingle, and consequently it is believed to have been poured out after the pampa shingle was distributed. I was consequently somewhat startled to find well-marked sedimentary rock between two layers of lava running out from the cone shown in tig. 3, which is situated at the side of the Gallegos river, about fifty-five miles from its mouth, ata place called the Buitreras. On close examination, however, it was found that this rock differs considerably from any of the tertiary deposits. It is of a coarse sandy grain, generally of a yellowish colour; in some places it would seem to consist of voleanic tuff, and in most places is stratified in thin layers from one to a few inches in thickness; but the most striking fact is that it has not only an occasional angular fragment of lava embedded here and there in its matrix, but also a considerable number of typical pampa pebbles. These pebbles, in the exposure at the side of the Buitreras cone, are in all the varieties and sizes generally seen on the pampa, and are embedded in the rock in all the levels which here crop out. About a mile and a half up along the saine side the high river bank juts out at right angles to its main course for a short distance. The top of the bank is here about 170 feet above the river, and consists almost exclusively of material similar to that already described, except that there are neither pebble nor lava blocks embedded in it. It also lies above and below a sheet of lava which crops out more or less in its middle. Along this same bank a few hundred yards further the ground rises for a short distance to the level of the main pampa, and shows the usual pampa shingle on the top; in fact, it has all the characteristics of an isolated flat- topped hill, and its face, as it borders the river from summit to base, consists of the same class of rock found opposite Gallegos and along the Atlantic coast. In other words, it is typical tertiary rock. At the same place where 218 Scientific Proceedings, Royal Dublin Society. this rock was observed, although most of it, exhibits well-marked true bedding with comparatively thin layers, occasionally here and there well-marked false bedding occurs, showing that at times at least there was running water when the formation was being deposited. I was fortunate enough to find in one place in this locality the clear line of junction between this rock and the tertiary rock, and the rock we are now considering was found to lie uncon- formably on the latter; in fact, it would seem that it was deposited in a huge basin scooped out of the older tertiary rock. It is, as a rule, impossible to see what this formation rests on, as there is such a heap of rubbish generally at the bases of the cliffs. I noticed, however, all along the locality where it is found that there is a line of springs cropping out from the valley banks about ten to twenty feet above the river. This would indicate that this rock lies on shingle. In one or two places I actually found the shingle on which it lay, and it was of the usual pampa type. I have given the name of Buitreras Beds to this formation. I traced it for upwards of twelve miles along the south side of the Gallegos river valley, and it probably runs much further. At the mouth of the Gallegos Chico it is found lying directly under a huge sheet of lava which caps a table three hundred feet high. It is here over two hundred feet thick, and at the back of Bella Vista, a few miles further along, I found it up to four hundred feet above the river valley. The fact that this formation was deposited in a great hollow scooped out of the tertiary rock, and lies on a layer of shingle of the pampa type, indicates that it is of a comparatively recent origin, and was formed after the shingle was spread over the pampas. As the base of this bed is only a few feet above the floor of the present river valley, the pampa must have been cut down at least three hundred feet before the Buitreras bed began to be deposited. At least one out- pouring of the lava occurred during the time when this bed was being formed, and one river terrace came into existence. From these facts it would seem that the Buitreras beds were deposited during inter-glacial times; and, although some of the deposit may have been formed on the bottom of a lagoon, some at least was formed under running water, as typical current-bedding can be seen in places. It is evident also from the angular fragments of lava found here and there embedded in it that semi-glacial conditions must have prevailed at least during a portion of the time, as these blocks were probably carried by ice and dropped on the bottom of either a lagoon or a large slowly running river. The lava sheet, which extends through it as an even, horizontal, compact layer for some miles, proves that there was one temporary interruption in the deposi- tion, if not more, and that during that interruption the bottom of the lagoon or Frnron—Pihysiography and Glacial Geology of S. Patagonia, 219 river was dry. We find that the Buitreras beds have in places been cut down a certain distance, when there was a volcanic outbreak and a lava sheet was poured over them; a terrace which had been formed at the mouth of Gallegos Chico was in its turn cut further down bya subsequent action. ‘This extensive terrace abuts, not only on the Gallegos Chico, but on the Gallegos river also, and indicates that there were at least two periods of cutting and probably one of glaciation since the Buitreras beds were formed. Further study of this interesting deposit shows that it has suffered extensive erosivn since it was laid down; in fact, only a few isolated patches remain here and there at the sides of the huge valley that was once filled with it. It must have been upwards of three hundred feet thick, and have extended for at least fifteen miles. At the side of the Gallegos Chico (a small river which joins the Gallegos proper at a place called Bella Vista) the Buitreras beds were cut down some 15‘) feet, and a sheet of lava was poured over them. This latter was further cut through by the Gallegos Chico river another eighty feet, until the tertiary rock was reached and the floor lowered to its present level ; and then, as I will show later, a glacier descended into this last river valley, and filled its mouth completely with a terminal moraine. Finally, this moraine has now in turn been cut through by the Gallegos Chico as it runs to-day. ‘The past history of this valley was, then, something as follows :— First, the great shingle period, then a great period of erosion by floods, which may or may not have been intermittent; this lasted until the river valley was cut down to almost its present level. Then occurred a long period of deposition, which was probably intermittent, and during which there*was at least one considerable outburst of voleanic activity. When this deposition had been completed, and between two and three hundred or more feet of rock had been formed, there occurred once more a long period of erosion by water, which probably had two or three breaks, as, although there is only one terrace to be seen in the valley of the Gallegos Chico, yet a little lower down there are two more or less distinct terraces. Now, if each of these inter- missions of erosion occurred in the manner already described, it appears that there were a considerable number of advanced recessions of the ice in this valley. The cause of the deposition producing the Buitreras beds is obscure. There is no trace of a dam across the river valley lower down which might have caused a great lagoon behind it. If during a subsidence the sea came further up the river valley, the current of the flood-water would have been considerably slackened, and sedimentation would have taken place; a bar might have also been formed and have helped the sedimentation. I did not, however, find any trace of such bar or any evidence of marine formation, so I have been consequently compelled to defer the further consideration of the subject until I am in a position to obtain more evidence. 220 Scientific Proceedings, Royal Dublin Society. X. Summary of Conclusions. We are now in a position to make a general survey of the work we have gone through, and, as we saw at first, the evidence at our disposal leads us to believe that, after a long period of dry climate with high westerly winds, probably in later tertiary times, the coantry to the north of the lower reaches of the Gallegos river became so levelled off that not a hillock practically could be found for hundreds of miles. The surface was as level as the sea, and if any inclination existed it was a slight fall from the moun- tains in the west towards the Atlantic. During this time the land in question was slowly rising, but not at a uniform rate, since we find three well-marked steps from north to south, indicating three old coast-lines. The pampas between Gallegos and Coyle towards the end of this period had partly subsided again, and portion of the land at Cape Fairweather was slightly below the sea-level. When things had reached this point, a change occurred, the climate suddenly began to get colder, and the ice, which was heretofore limited to the higher levels of the mountains, began to descend into the plains. As the ice descended it caused considerable smashing up of the rocks, and it carried a huge quantity of broken debris with it. All this time the climate was getting colder and cclder, so that the spring floods were not excessive, and consequently had no great eroding power. The huge quantities of broken debris carried down by this ice-sheet became deposited partly under it and partly along its termination; this limit is shown by a line marked on the map. The spring and summer waters, melt- ing away from this great ice-mass, formed slowly running and expanded streams, and, as the country was one huge slightly inclined plain, these streams could not follow any definite valley, but ran broadcast over the country. They carried with them much of the finer sand and shingle which the ice had brought to its limits, but left all the large blocks behind. This condition of affairs lasted until the pampas were covered with a layer of shingle, which extended right to the Atlantic coast, and probably a long distance into the sea. When the pampas had become thus completely covered, a change occurred once more in the climate; the seasons began to get warmer, and each summer a greater quantity of ice and snow melted than was formed during the winter. The result of this was that the spring and summer floods increased, and, in place of depositing sand and gravel, they gave rise to huge torrential rivers, with very considerable erosive powers. By the time the ice had all melted away, and the country had settled down once more to a condition of Frnron—Physiography and Glacial Geology of S. Patagonia. 221 genial climate, the surface of the land had everywhere been cut into by large river-valleys, with many tributary cafiadones. During the genial period which followed the retreat of this great ice-mass, voleanoes began to break out, some pouring out their lava on the surface of the first pampa level, others in the great, broad river-valleys. During all this time the country was slowly rising, and, after a certain interval, which it is impossible at present to determine, the climate once more began to get cold, and the ice began to creep down from the mountains into the plains. This sequence of events was probably repeated from time to time right through the quaternary era, and, although I have only succeeded in finding four moraines attributable to different horizons, I have reason to believe that. the changes mentioned above occurred at least seven times, and in this connexion it is worthy of note, as I will presently explain, that the last moraine found, although of the small valley type, extended practically as far towards the east as the great plateau one first mentioned. APPENDIX. On the 28th March, 1915, I left Estancia Alquinta and proceeded to Douglas (Esperanza), on the north side of the Gallegos river, my intention being to study the line of junction between the great terminal moraine and the flat pampa, which I had good reason to believe existed between these two points. After two and a half hours I noticed on going up to the higher level that the pampas had still all the characteristic flatness already described as universally found to the north of the Gallegos river. Here I found myself at the foot of a long, stiff incline, which I ascended, and on the top of which I halted to rest my horse. I here quote from my notes written on the spot:—“28th March. 2 pm. Rode from Alquinta to Douglas; after entering Bella Vista camp had a long pull up to the top of a pampa. Here the barometer registered five hundred and thirty feet above Alquinta. Could see from this point the change from the flat pampas to the undula- tory. ‘There is no very marked rise of level, but the undulatory condition of the surface of the high pampa can be distinctly seen. Mouth of Gallegos Chico from here bore 8. by E. magnetic. The pampas have what might be called a drumlin appearance. The line of demarcation between the flat and SCIENT. PROC. R.D.S., VOL. XVI, NO. XIX. 21 222 Scientifie Proceedings, Royal Dublin Society. undulatory pampas can, on careful observation, be distinctly seen: the difference in level does not at first seem to be more than fifty feet or so, perhaps even less; yet the topography of the two plains is distinct. So far I have not seen an erratic, but hope to see one soon.” I restarted along the track, and my next note is as follows: “Had not gone three hundred yards beyond the point where I wrote the above when J came on erratics; not one but several, showing their tops here and there above the ground, several feet in diameter. A little further along they can be seen in all directions ; are non-voleanic, and many of them are very large.” For about two leagues at least the surface had a tendency to slightly rise, and the rather spread- out drumlin appearance was well marked. About three leagues from the edge of this great moraine is a deep cafiadon, running more or less north-west and south-east, with a floor so level that it was very difficult to say which was the direction of its outflow. Its inclination proved to be backwards, away from the valley of the Gallegos. Jn fact, this cafiadon ran into a series of bajos, which eventually, further to the west, discharged themselves into the Gallegos river through a series of cafiadones. Douglas Estancia is situated in a catiadon about thirty feet below the main pampa, about twenty miles to the west of the edge of the terminal moraine. On one or two occasions I carefully estimated its height above the river as about two hundred feet. All the pampas round about the Douglas settlement exhibit a well-marked glacial topography, showing drumlins and erratic blocks everywhere, and consequently these differences in level seemed peculiar. The river runs a very even course from the Douglas camp to the point opposite the end of the terminal moraine below Bella Vista; there are practically no rapids, and the difference in levels in the two localities cannot consequently be more than fifty feet, if so much. This will leave the end of the great moraine about two hundred and eighty feet higher than the plain at Douglas, which is undoubtedly a continuation of the same glacial feature. The particular glaciation which we are now dealing with thus left behind it a large and extensive moraine, at its eastern termination two hundred and eighty feet higher than its level twenty miles further to the west. The topography of this - area is rather complicated by the presence of a long narrow ridge which extends out into it from the cordillera in the west; it is about six hundred and fifty feet above the plain at Douglas Estancia, and the ridge is covered by moraine material and shows well-marked drumlins and erratics. It is on an average about a mile or two in breadth, and the fall on the north side is much less than that on the south. I did not estimate its height above the northern plain, but I should judge it to be about one hundred and fifty feet. Now, the tertiary rock underlying the outwash gravel on the flat plain, even close to Frenton—Physiography and Glacial Geology of S. Patagonia. 223 the fringe of the moraine, can be seen in places at no great distance, not more than thirty or forty feet, below the surface of the shingle ; consequently the surface of the tertiary rock also falls in level as we proceed towards the west, and this fall begins after it becomes overlain by the moraine proper. The accompanying diagram will explain these facts more clearly. Unfortu- nately, I have never had an opportunity of going further west than the middle of the Douglas land, so I could not study the further topography of this plain in detail; however, from the tops of several high hills I could see in the distance about twenty miles further west than Douglas Estancia another line of high ground as if a second line of terminal moraine similar to that which I have been describing. Intelligent people from whom I have made inquiries have informed me that practically the same sequence of change is found here, namely, a sudden rise of level and a gradual fall, which finally reaches the sea-level at the channels. I have always found that the ter- tiary rock, even in the middle of the volcanic area, follows a very even course and exhibits very little tendency to dip in any direction, so that I can only conclude that this sloping down to the west is here due to erosion, and that this erosion was of glacier origin. The peculiar sloping erosion of the tertiary rock coincides exactly in extent with the moraine, and we cannot help coming to the conclusion that the same agent operated as a cause in both instances. Once we get well inside the moraine area, we notice that terracing practically disappears from the river valleys. A little below the Bella Vista settlement, at the mouth of the Gallegos Chico, where it joins the Gallegos river, the lower terrace has been cut down by the former river (Gallegos Chico) to the level of the present river valley. This cutting would seem to have belonged to the last erosive phase of the CF-Cape Fairweather Bed. Rwer Valley Floor Tertiary Rock 1CO allegos Moraine. 1. Diagrammahe Section through Gallegos River Valley from Andean Ci hannels fo Atlantic. Four distinct Moraines are shown. That of the Gallegos Chico being the most recent. river-valley cutting process, and must have been subsequent to the formation of the moraine-covered terrace further along. Yet down in the mouth of 224 Scientific Proceedings, Royal Dublin Society. the Gallegos Chico, almost filling it from side to side, except where the present river has cut a recent narrow gorge, is found a most perfect little terminal moraine. In PI. VII, fig. 4, we are looking across the mouth of this valley, and it will be seen that the moraine almost closes it. The present river has cut its way through it, and is running down in a narrow channel a little beyond the horses. I consider this moraine, although very small, most instructive, as it clearly shows that the great erosive action which operated intermittently over an extensive period in the cutting out of the river valleys was accompanied to its termination by the periodic advance of ice in the form of glaciers ; and this confirms my already expressed belief that all the terraced river valleys are fluvio-glacial products even down to their lowest levels. Although I have abundantly looked for it, I have never succeeded in finding anything that would show that during the inter-glacial periods the climate was as genial as at present. There is everywhere in the country evidence of oscillations of climate, but T have never found any facts which would prove beyond doubt that there was a true inter-glacial epoch. In every terrace I have been able to examine the shingle lies directly on the underlying tertiary rock, and I have never seen any deposit of loam or mud, with animal or plant remains, intervening between them. I have found at least four separate deposits of moraine material, each of them belonging to a different horizon, and probably many more might be found if a systematic search could be made. The first I found was, as I have already described, the great terminal moraine on the high pampa, from which ran the supra-pampean shingle- layer as an outwash. The second was on a high terrace at the back of Bella Vista, and it was cut into when the river valley was lowered to the next lower level. The third was an extensive terminal moraine about. fifteen miles to the west of Douglas Estancia. Finally, I found, as men- tioned above, a very well-marked small terminal moraine filling the mouth of the Gallegos Chico river valley, where it opens into the Gallegos river, and having its base almost on the present level of the latter. Now, the Gallegos Chico valley is cut down through over two hundred feet of Buitreras bed, which, we have already seen, was probably formed on the floor of an inter-glacial lake. This shows that at the very end of the river- valley period, when their beds had been lowered to their present levels, tongues of glacier extended into the river valleys as far to the east as the first great plateau glacier which gave origin to the pampa shingle. The diagrammatic section through this portion of Patagonia from west to east shows the position of these four moraines. PLATE V. PROC. R. DUBLIN SOC. N.S, VOL. XVI. 1 Sarmiento Lago “© 20 Del. Too S Three Brothers \ @Serra Diablo ~“ -°MT AYMONT 0 ; r-} La. Blanca” SCIENT. PROC. R. DUBLIN SOC.,-N.S., VOL, XVI. PLATE VI. FIGURE 1. re pe Tw ap ae ei eS Sly ot Oar SCIENT. PROC. R. DUBLIN SOC., N.S. VOL. XVI. PLATE VII. 10. 11. 13. 14. SCIENTIFIC PROCEEDINGS. VOLUME XVI.. . A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, SO.D., F.R.s., and T. G. Mason, m.a., sc.B. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. Smyru, B.a., sc.s. (Plates I., If.) (February, 1920.) 1s. 3. The Application of the Food-Unit Method to the Fattening of Cattle. By James Winson, m.a., B.sc. (Plates III., 1V.) (February, 1920.) 1s. The Holothurioidea of the Coasts of Ireland. By Anne L. Massy. (April, 1920.) 1s. . Photosynthesis and the Electronic Theory. By Henry H. Dixon, sc.p., Fn.s., and Horace H. Poorn, sc.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Witriam Brown, B.Sc. (March, 1920.) 6d. . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, m.a., sc.B. (April, 1920.) 1s. . The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Wittiau Brown, 8.sc., and Parrick O’CannaGuan, A.R.0.8¢.1.,A.1.c. (August, 1920.) 6d. Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. Lronarp, 4.R.¢.sc.1., B.8C., PH.D., and Agnes BRowNg, A.R.C.SC.I., B.sc. (August, 1920.) 6d. An Investigation into the Causes of the Self-Ignition of Kther-Air Mixtures. * By the late Professor J. A. McCurtianp, p.sc., F.n.s., and Rey. H. V. GiLt, 8.J., D.s.o., M.c., m.a., University College, Dublin. (August, 1920.) 6d. The Influence of Hlectrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosrpH Reiuy, M.a., D.sc., F.R.C.sc.1., and Wirrrep J. Hickmzorrom, B.sc. (October, 1920.) 6d. - Notes on some Applications of the Method of Distillation in Steam. By JosepH REILLY, M.A., D.SC., F.R.C.SC.1., and WiLFRED J. Hickinzorrom, B.Sc. (October, 1920.) 6d. The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By W. EH. Aprnry, p.sc., a.R.c.sc.1., F.1.c., and H. G. B&ckER, A.R.C.SC.1., a.i.c. (September, 1920.) 6d. A Determination, by means of a Differential Calorimeter, of the Heat produced during the Inversion of Sucrose. By Hunry H. Dixon, so.p., F.R.S., and Nicet G. Batu, B.a. (December, 1920.) 6d. - The Measurement of very Short Time Intervals by the Condenser-Charging Method. By Joun J. Dowsine, w.a., F.1nst.P., and Donat Donnetty, usc. (In conjunction with the late Prof. J. A. McCurtuanp, r.R.s.) (February, 1921.) 6d. 16. 17. 18. 19. 20. SCIENTIFIC PROCEEDINGS—continued. A Vibrating-Flame Rectifier for High-Tension Currents. By Joun J. Dowtine, M.A., F.INST.P., and J. T. Harris, B,sc. (February, 1921.) 6d. A Sensitive Valve Method for the Measurement of Capacity, with some Important Applications. By Joun J. Dowxine, m.a., F.nst.p. (February, 1921.) 6d. A Direct Reading Ultra-Micrometer. By Joun J. Dowie, M.a., M.R.1A., F.inst.P. (March, 1921.) 64d. Studies in the Physiography and Glacial Geology of Southern Patagonia. By EK. G. Fenton. (Plates V, VI, and VII.) (March, 1921.) 4s. 6d. Award of the Boyle Medal to Grorcz H. Prtuysriper, B.sc., pu.p., 1921. (March, 1921). 1s. DUHLIN: PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIBES. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), No. 20. MARCH, 1921. AWARD OF THE BOYLE MEDAL TO GEORGE H. PETHYBRIDGE, B.Sc., Pu.D. 1921 VAY DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1921. Price One Shilling. Royal Bublin Society. Oe es FOUNDED, A.D. 1731. INCORPORATED, 1749. SCIENTIFIC MEETINGS. Tur Scientific Meetings of the Society are usually held at 4.15 p.m. on the fourth Tuesday of every month of the Session (November to June). The attention of authors is directed to the following extracts from the regula- tions governing the presentation of communications at Scientific Meetings of the Society, and the publication of papers in the Scientific or Economic Proceedings :— «An author who desires to present a communication at a Scientific Meeting of the Society shall furnish to the Chief Executive Officer for Science a fortnight previous to that meeting the title and an abstract (not exceeding 200 words) of his communication, together with an estimate of ‘the time required for presentation. *¢ An author who desires to have a communication published forthwith as a paper in the Proceedings of the Society shall furnish to the Chief Executive Officer for Science the full text and necessary illustrations at least a week before the meeting of the Committee of Science at which he desires publication to be considered. Meetings of the Committee of Science are usually held on the second Tuesday of each month from November to June inclusive. « To Gosomete: Llectrometer To Earth J Aluminium Terminals N to Induction Coif Air To Earth which ionization by the a-rays of uranium oxide took the place of the Rontgen rays of Becker’s experiment, and in which positive and negative ions behaved similarly, prove conclusively that nuclei and small ions are separately produced, and by their combination with each other form the large ions. The diagram gives a plan of our apparatus. Air was drawn by means of a gasometer through a cotton-wool plug, and then over a water surface or through drying tubes if desired. It then passed through a long quartz tube into an ionization tube containing two terminals. ‘The first, T,, was 50 cms. long, and generally connected to earth. The second, T., was 12 cms. long, and connected to a quadrant electrometer with a capacity in parallel. Between the two terminals was a glass tube which just fitted into the ionization tube, and which was coated on the inside with uranium oxide. 202 284 Scientific Proceedings, Royal Dublin Society. The quartz tube passed through a box lined with tinfoil and earthed. Inside the box, and opposite the quartz tube, were aluminium terminals connected to the secondary of an induction coil. The shape of the terminals enabled them to be placed, if necessary, right up against the quartz tube. A motor-driven mercury interrupter was used with the induction coil. The quartz tube was about half a millimetre thick. In most of the experiments the light also traversed 2 or 3 mms. of air. © The ultra-violet light of the spark was found, in a number of careful experiments, not to ionize the air; but as a precaution the long terminal, T,, was used {o remove to earth any ions contained in the air before its passage over the uranium oxide. The outer cylinder of the whole ionization tube AB was kept at any desired voltage by means of cells. When the air was passing through the apparatus, the ions produced by the uranium oxide took on an average about twenty or thirty seconds to reach the terminal T,. A great number were consequently lost by recombination. The current due to the uranium oxide itself we shall call C,. When C, had been measured, the spark was started. The air now passing over the uranium oxide contained the nuclei produced by the ultra-violet light. Some of these nuclei. attached themselves to small ions. In this way large ions were. produced which recombined less rapidly than the small ones. More ions consequently reached the terminal T, than when the nuclei were absent, and consequently the observed current C,, when the spark was on, was greater than the current C;. The current C, will evidently increase with the number and size of the nuclei present in the air, and thus the ratio = gives an idea of the 1 number of nuclei present. An idea as to the relative sizes of the nuclei and small ions can be got from a knowledge of their mobilities under an electric field. This mobility can be deduced from the saturation voltages in a current-voltage curve when the dimensions of the terminal and ionization tube and the quantity of air passing per second through the apparatus are known. In our experiments, the quantity of air flowing per minute varied approximately from 1000 ces. to 6000 ces. The ratio = varied from 1 about 8 for slow blasts of air to about 1:2 for quick blasts. The value of the ratio = is increased for slow velocities of the air through the apparatus, 1 not only on account of the larger time allowed for recombination, but because the time of exposure of the air to the light is also increased, so that more and larger condensation nuclei are produced. For the quickest air- blasts used the nuclei had a mobility of about ‘3 cms. per sec. per volt/em. McCuiettanp AnD M‘Huenry— Uncharged Nuclei. 285 For very slow blasts large ions were found with a mobility as low as 00024. The slower the air-blast the smaller was the mobility. Curves 1 represent the manner in which C, and C, decay with the time taken by the air to traverse the apparatus. The currents are, as throughout, in arbitrary units. The manner in which C. changes with time is also 1 Qoo 800 a BUR Vea 600 i D << ee =) S iS © Ss = = S @) 300 200. 100 100 120 OO © ie Unil nearly 1sec shown. (,, of course, is the current of small ions due to the uranium by itself when no spark is playing. ©, is the current of large ions due to the charging by the uranium ions of the nuclei produced by the ultra-violet light. The air in all these experiments passed over a water surface. For the 20 4O 60 286 Scientific Proceedings, Royal Dublin Society. 1 quickest flow of air (7,800 ccs. per min.) the value a = 106; for the 1 slowest (812 ccs. per min.), = = Sil, 1 Mobilities met with in these Kaxperiments. The mobility of the charged condensation nuclei varied with the rate of flow of the air through the apparatus. The slower the rate of flow the larger were the ions. The mobilities obtained varied from °33 to 00024 ems. per sec. under a field of 1 volt per cm. We obtained ions of mobilities -35, 18, °10, 02, 008, 0044, 0034, 0025, and 00024. The numbers were deduced from the saturation voltages in current-voltage curves, and do not claim a high degree of accuracy. The positive and negative ions under similar conditions have the same mobilities. As an example of the manner in which the mobility varied with the rate of flow of the air, the following numbers will be quoted. During the experiment the intensity of the light, &c., remained constant, only the rate of flow of the air varying :— Quantity of Air per min. Mobility. 3600 ces. “18 2050 ,, -008 1120 _,, 0034 Liffect of Varying the Intensity of the Light. C. I’. R. Wilson observed that the nuclei produced by ultra-violet light increased in size as the intensity of the light increased. Several experiments were performed by us showing that by diminishing the distance of the spark from the quartz tubes, and thus increasing the light intensity, the nuclei grew in size and became more numerous. In these experiments, keeping the current of air constant, and a voltage on the ionization tube sufficient to “saturate” the largest ions produced, the distance of the spark was varied. One of the quartz tubes employed gave an appreciable effect up to 15 ims. distance; with the second the effect stopped at 5mm. Thus in one experiment, calling the current due to the uranium alone, C, = 100, the following values were obtained for C, with McCririianp anp M‘Henry—Uneharged Nuclet. 287 the spark at different distances, the quantity of air per minute being 2,000 ces. :— P : C2 = current due to charged | Distance of Spark in mms. | ~* t due to charged nuclei. i ts a Pre | 5 213 10 162 15 117 No Spark. C, = 100 In another experiment, using a different quartz tube and a blast of air = 3,000 ces. per minute, the following numbers were obtained :— Initial uranium current, C, = 236. Final i FR C, = 238. Spark distance in mms. . | = 1d | 10 | 5 3 | 2 1 CurrentC,2 . . .,| = 238 | 250 | 250 365 625 910 90 80 C4 iS »” 70 Ne Ne < = < ae E << 2 mms. spark distance S) fe BURVES € 40 3mms spark distance 30 | Uranium Current 20 10 Volts 20 40 60 80 10 12O WO HE 0 Poo) In the latter experiment a current-voltage curve was obtained for each distance of the spark, and also one when the spark was stopped, giving the current due to the small ions alone. Curves 2 show the relative magnitude of the currents at different voltages. All the curves have the same saturation 288 Scientific Proceedings, Royal Dublin Society. point near 120 volts, indicating an ion of mobility about :009. For the larger spark distances this ion is present in small quantities only. Thus when the spark is 3mms. from the quartz tube, the largest ion is about 16 per cent. of the total; at 2 mms. 40 per cent.: at 1 mm. 61 per cent. The data from which the curves were drawn are given in the following tables :— TABLE FOR CURVES 2. Spark distance (mms.) Volts. Current. || Distance. Volts. Current. nat — ants No Spark 200 23°6 2 mms. 200 62°5 Ki ay | aa. a! 120 62:5 9 20 23 96 80 58 pe 15 mms. a 200 23°S 90 40 47°6 10 gp 200 | i 29 36°5 bitcee. 200 25 - 10 20°8 3 mms. 200 R 36°5 | ip ar GR 99 120 37 || 1mm 200 91 is 80 35-2 || a 120 89 ye 40 33.3 iie | ae 80 71-4 a 20 31°3 os 40 521 x4 10 25 es 20 35:2 - 10 18-5 Similar behaviour of (+) and (—) Ions. Our experiments showed that no difference existed between positive and negative numbers ; or, in other words, that the nuclei attached themselves indifferently to positive or negative small ions. A current-voltage curve was plotted, using positive and negative voltages alternately. A preliminary test had shown that when the first terminal T, was earthed, and the uranium removed, no ions (due to the spark), either positive or negative, reached the second terminal T,. The following table gives the numbers observed :— Voltage. Current. + 120 116 (Ci) Uranium aione, ~ 120 183 55 + 120 838 (C2) Uranium and Spark, — 120 806 ,, McCrecianp AnD M‘Henry—Uncharged Nucler. 289 Lonization Curve (Uraniwm and Spark). Volts positive. Current. Volts negative. Current. 10 288 10 246 20 400 20 382 30 467 30 500 40 575 40 543, 60 625 60 667 80 735 80 676 It will be evident from the tables above that positive and negative numbers are materially the same, indicating the equal behaviour of positive and negative ions. Results similar to the preceding were obtained by Lenard and Ramsauer (Joc. cit.). In their experiments the large ions were produced directly by Schumann—violet rays, which at once ionized the air and produced the nuclei. They found that the ions increased in size with the intensity of the light, the mobilities being measured, as in our experiments, by plotting a current-voltage curve. Increasing the time of exposure of the air to the light by diminishing the velocity of the air past the source of light also increased the size of the ions, as in our experiments. Lenard and Ramsauer give curves showing the decrease of the currents by recombination (cf. Curves 1); the rate of decrease being slower, the more vaporous impurities are contained in the air. Finally, the mobilities of the positive and negative ions were shown by them to be the same. The mobilities were deduced from the saturation points in a current-voltage curve. Lffect of Drying the Avr. The number and size of the nuclei depend upon the amount of moisture in the air. Thus the effect almost disappears in air dried by calcium chloride, but even on drying with phosphorus pentoxide a small number of nuclei seem always to be produced, as shown by a slight increase in the uranium current on switching on the spark. Curves 3 show the relative effects in moist and dry air. As will be seen, the current C,, consisting of condensation nuclei charged by the uranium ions, is in the case of dry air only slightly greater than C,, the current due to uranium alone, while in the case of moist air C, is 3 times as great as C;. The saturation point in moist air is 40 volts, indicating an ion of mobility 014. Furthermore, these ions 290 Scientifie Proceedings, Royal Dublin Society. of mobility -014 are nearly 50 per cent. of those formed in moist air. In dry air the largest ion is saturated at 8 volts, indicating an ion of mobility -07, which, as is evident from the curve, forms only about 7 per cent. of the total. 500 450 Lo 350 CURVE 3 300 Dry Air = Uranium Current 40 ko 80 100 120 140 Volts In all probability rigorous drying would remove the effect altogether, but in the experiments tried by us a small number of nuclei was always formed in dry air. The effect in ordinary air depends, of course, upon the humidity of the atmosphere, but is less than the effect in air passed over a water McCretianp ann M‘Henry Uncharged Nuclei. 291 surface. Thus in one experiment the uranium current was increased by the spark five-fold in ordinary air, and seven-fold in air passed over a water surface. The air was filtered in both cases as usual. Nature and Origin of the Nuclet. It has been pointed out that the slower the moist air passed the spark the greater were the nuclei produced. If moist air at rest be exposed to strong ultra-violet light for about thirty seconds, the nuclei grow so large that they become visible, and the path of a strong beam of light through the vessel reveals a dense blue cloud, the individual particles of which can be seen in a microscope with a dark-field arrangement. Those particles are evidently drops of H,O, and thus the nuclei which we dealt with in the preceding experiments were in all probability minute drops of water formed by the light, and growing in size with the intensity of the light and the time of exposure. C. T. R. Wilson first attributed the formation of the nuclei to the production of some substance such as H.O,, which when dissolved in water lowered the equilibrium vapour-pressure, thus causing drops to be stable and grow which would otherwise evaporate. Other experimenters have proposed oxides of nitrogen as the parent substance of the nuclei. In our experiments the difference between the formation of nuclei in dry and moist air has been shown to be very remarkable. The number and size of the nuclei were found to depend upon the quantity of water-vapour present in the air. Experimenters who used Wilson’s expansion apparatus to detect the nuclei have stated that carefully dried and moist air behave similarly as to the production of nuclei (St. Sachs. Ann. der Physik, 34, 1911, and Saltmarch, Proc. Phys. Soc., London, 27, 1915). In view of our experi- ments quoted above, this must mean either that the expansion apparatus is 30 very sensitive a method of detecting nuclei that those formed on the shght traces of moisture left in the air after careful drying cause the formation of an appreciable cloud in the cloud-chamber, but yet would not be numerous enough to affect appreciably the value of the currents in our experiments, or else that the hygroscopic parent substance (e.g., oxides of nitrogen) is formed in dry air, and forms the large nucleus when admitted into the saturated atmosphere of the cloud-chamber. On the latter supposition, if dried air is exposed to the ultra-violet light in our experiments and then passed over a water surface before entering the ionization chamber, large nuclei should be formed and detected by the increase in the current C, in the usual way. This experiment was performed, but no nuclei were detected. A further experiment was performed in which dried air was passed over a 292 Scientifie Proceedings, Royal Dublin Society. water surface at 0°C. before being exposed to the spark. A small definite quantity of water vapour was thus contained in the air, and a small number C, C, second water surface at room temp. 11°C. after the quartz tube, it was thought of nuclei were formed, being found to be 1:25. On now ineluding a x that Us would increase, as if, for example, H,O, were formed in the partly 1 moist air more nuclei would be formed in passing over the second water = was found, however, to have the value 1:24 in this case, so that ome no additional nuclei were formed. Many physicists have tried to demonstrate the presence of H,O, in the nuclei formed by ultra-violet light in moist air. Vincent (Camb. Phil. Soc. Proc., 1904, 12) used a sensitive photographic test, but failed to detect any H.O;. Miss Saltmarsh (doc. cit.) showed that prepared H,O, when introduced into the expansion apparatus caused the production of nuclei. The presence of H.O, in these nuclei could be demonstrated; as, when about 19 expansions were made, the showers falling into a small vessel containing titanic acid, the acid became yellow. On the other hand, 150 expansions on the nuclei produced by ultra-violet light caused no colouration of the acid. Miss Saltmarsh concluded that the part played by HO, was at most a very small surface. one in the formation of nuclei. On the other hand, Barkow (Ann. der Physik, 28, 1907) found that: H,O, does not assist cloud-formation unless when decomposed by strong sunlight. When the H,O, is exposed to strong sunlight, numerous nuclei are produced, which persist for long periods. Ultra-violet light is not necessary for this decomposition of H,O,, but it was not brought about by light from a strong arc lamp. Barkow tested for hydrogen peroxide in the nuclei pro- duced by ultra-violet light in moist air by drawing them slowly through a solution containing titanic dioxide in concentrated sulphuric acid. This solution turns brown with H.O., but Barkow obtained no colouration even after 150 hours, using in all 154 litres of moist air. Barkow investigated the effect of nitrogen peroxide on the condensation of water vapour. He passed nitrogen dioxide into the cloud-chamber. Spontaneous condensation immediately took place, and many expansions were required to remove the nuclei. He also ozonized oxygen, and passed it into the expansion apparatus. The characteristic blue cloud was obtained without expansion. This spontaneous condensation with ozonized oxygen, however, Barkow attributed to the presence of small quantities of nitrogen in the oxygen used, oxides of nitrogen being formed. McCiriianp and M‘Henry— Uncharged Nuclei. 293 Pringal (Ann. der Physik, 26, 1908) verified the effects of oxides of nitrogen upon the condensation of water vapour, and examined Barkow’s supposition that the condensation brought about by ozonized oxygen is really due to traces of nitrogen in the oxygen used. The question was a difficult one, owing to the task of obtaining oxygen entirely free from nitrogen; but Pringal set himself to examine the effect of successive purifications of the oxygen used. Oxygen prepared electrolytically was used, and after passing through the ozonizer was passed through a solution of NaOH to remove oxides of nitrogen. The spontaneous condensation still took place, and Pringal was led to believe that the ozone, even when free from oxides of nitrogen, was able to oxidize the traces of nitrogen in the cloud-chamber. Using a series of seven ozonizers, each fitted with a NaOH tube, he partially exhausted the cloud-chamber, and drew the ozonized oxygen, from which the last traces of nitrogen should have been removed, into the cloud-chamber. He repeated this process of exhaustion and re- placement by ozonized oxygen, the gas in the cloud-chamber becoming freer and freer from nitrogen. The process was repeated £00 times, and the spontaneous condensation gradually decreased, and finally disappeared. Still a cloud could be formed with expansion, but even the condensation with expansion dis- appeared gradually. The ozonized oxygen nowhad no effect on the condensation of the water vapour, and the previously observed condensation was attributed to the oxides of nitrogen. Barkow and Pringal suggest that nuclei formed in moist air by ultra-violet light are similarly due to the production of oxides of nitrogen, and not to hydrogen peroxide. In this connexion also Miss Saltmarsh examined the production of nuclei in moist nitrogen. When traces of oxygen are present nuclei are readily formed ; but when the residual oxygen is removed by sparking for twenty minutes and absorbing the oxides of nitrogen in NaOH, the ultra-violet light produced no nuclei in the pure moist nitrogen. The case for hydrogen peroxide, being the parent substance, rests upon a test by Bieber (Ann. der Physik, 39, 1912). He passed moist oxygen exposed to ultra-violet light through a vessel at the low temperature of — 79°C. The condensation products were then subjected to the following tests for hydrogen peroxide :— (1) Potass. iodide and starch, which, in the presence of ferrous sulphate, gives a blue colouration with hydrogen peroxide. (2) Mixture of potass. ferricyanide and ferric chloride, which gives a blue colouration with H,0,. (8) Dilute solution of titanium dioxide in strong sulphuric acid, which turns brown in the presence of H,0,, 294 Scientifie Proceedings, Royal Dublin Socicty. All three tests answered for hydrogen peroxide, and the experiment is apparently conclusive. It may be objected, however, that the hydrogen peroxide detected was not present in the nuclei, but simply formed in the air by the ultra-violet light, and not affecting the condensation. We, accordingly, tried a test in which the nuclei after being charged by uranium were driven by the field on to an electrode covered with moist paper. The paper had been dipped into a mixture of starch solution, potass. iodide, and ferrous sulphate. No blue colouration was observed even after twelve hours. The quantity of H.O, may, however, have been too small to be detected by this method. The following experiments were performed by Lenard and Ramsauer on the origin of the nuclei which were detected by their effect on the steam-jet. Air was freed from all vaporous impurities by cooling to — 76°C. No nuclei were formed in this air by ultra-violet light. Adding slight traces of water- vapour to the air caused a slight but perceptible effect on the steam-jet. Saturating the pure air with pure water-vapour by passing it over a large “ water surface produced a marked effect on the steam-jet, but not nearly so strong as that of ordinary moist laboratory air. It is evident, then, that water-vapour is only partially responsible for the production of the nuclei in ordinary air, and that cooling to —'76° C. removes, besides water-vapour, some other effective agent. Shght traces of water-vapour along with vapours from india-rubber gave a strong effect on the steam-jet, while the same amount of water-vapour alone, or of vapour from india-rubber alone, gave only a slight effect. We shall see later in Lenard’s experiments upon other vapours, besides water-vapour in air, that ammonia gave many nuclei on exposure to ultra-violet light. It is possible that ammonia is the other effective agent besides water-vapour in ordinary air. Lenard attributes the nucleation in the case of ammonia to the formation of ammonium nitrite, and ammonium nitrate, and, where water-vapour is alone present, to the formation of hydrogen peroxide. Luperiments on Air Saturated with other Vapours. Experiments were performed on air containing vapours of ethyl alcohol, methyl alcohol, and toluol. The room air, after being carefully dried, was passed over a surface of the liquid before being exposed to the ultra-violet light. C, and C, were then measured, as in the case of water-vapour, but in all three cases examined C, was equal to C;, so that nuclei similar to those obtained with water-vapour in the air were not formed. It was noticed in these experiments that if minute traces of either of the three vapours mentioned above were present they prevented the formation McCreuianp ann M‘Hunry— Uncharged Nuelci. 295 of the nuclei in moist air. In one experiment room air was filtered and ; : C, passed over a water surface and exposed to ultra-violet light. <* was found to 1 be nearly 4. On putting about }ce. of alcohol in the water, whose volume was about 1,000 ces., = dropped to the value unity, so that no nuclei were formed. The action of the alcohol-vapour in preventing the formation of the nuclei might be in either of two ways. It might prevent the formation of the hygroscopic substance (e.g., H,O,) which is the parent of the nucleus, or it might act on the minute water-drops after formation. To test this point, nuclei were produced in moist air and then passed over an alcohol surface. After passing over the alcohol surface the nuclei could no longer be detected by our methods, the action of the alcohol-vapour being just the same as if it had been contained in the air before exposure to the ultra-violet light. Air saturated with alcohol-vapour was now exposed to intense ultra- violet light in a closed vessel for some time. A narrow beam of intense are light traversed the vessel. Its path was at first invisible, but when the spark had acted for thirty seconds or so a thin cloud of large drops was seen in the beam. Using water instead of alcohol, and exposing for the same time, a cloud is also formed, but very different in appearance. It is now blue, the are light appearing as a solid blue beam through the vessel. The drops in the case of water-vapour are evidently much smaller and more numerous than when alcohol-vapour is used. The effect of adding traces of alcohol-vapour to the water-vapour in the above experiment is to prevent the formation of the characteristic blue cloud. A thin cloud of large drops is, however, formed similar to that obtained with alcohol-vapour alone. In the light of these results the reason why no nuclei were detected in our experiments by the electrical method is that, although nuclei were formed, they grew so big and were so few that their presence could not be demonstrated by our methods. Lenard and Ramsauer also investigated other vapours besides water. Alcohol, methyl, ether, benzol, chloroform, atrachinon, and alizarin were found to be ineffective. Benzine was slightly effective in forming nuclei. Ammonia, hydrogen bi-sulphide, carbon-disulphide were effective. In these experiments pure air was passed over a bulb containing water, and another containing the liquid to be examined. In addition, Lenard examined the nuclei formation in moist gases other than air. Oxygen and carbon-dioxide behaved like air; but no nuclei were formed in moist bydrogen. No nuclei were formed in water-vapour alone. 296 Seventific Proceedings, Royal Dublin Soctety. IJ.— NUCLEI DRIVEN ofr GLASS TUBES BY HB&AT. If a glass tube be heated while air is passing through it, uncharged particles are given off which can be charged by passing over uranium oxide. The apparatus previously described was used also in these experiments, the glass tube examined being included between the cotton-wool filter and the quartz tube. A small quantity of charged particles was also given off when’ a tube was heated. The tube was generally heated with a luminous gas flame extending over about an inch of the tube. The presence of the nuclei was demonstrated by an increase in the uranium current just as described in the case of the nuclei produced in moist air by ultra-violet light. The emission of the nuclei fell off rapidly, almost ceasing in about ten minutes generally, the current registered falling off to the value of the uranium current alone. The effect reappeared, however, on heating a cool place on the tube. Thus in one experiment the following numbers were obtained :— Small ion current (uranium alone) : : C, = 19°8 Current of nuclei driven by heat from glass and charged by the uranium ions : 6 C, = 90:9 After ten minutes’ heating ; : j C, = 20°8 The large increase in the current would lead us to expect that the nuclei are of great size. An attempt was then made to plot a current voltage curve. Owing to the fatigue effect it was necessary for each reading to heat a new place of the tube, beginning at that part furthest from the end where the air entered. The charged ions driven off the tube by heat were removed to earth by a second ionization tube as above described, so that the total increase in the current is due to the uncharged nuclei. Saturation occurred at 280 volts, giving a mobility of -0044. The curve also indicated the presence of a smaller ion of mobility about 016. The mobility -0048 was obtained in another experiment. Another experiment on a different tube did not give saturation at 400 volts—the highest voltage then available. This indicated an ion of mobility less than ‘0028. The experiment was repeated, using a longer terminal, and the mobility ‘0016 obtained. The nuclei obtained by heating glass tubes are much more numerous than those produced by ultra-violet light in moist air. In the latter experiments some indications were always found of unattached small ions, whereas the nuclei driven off by heat from glass tubes were so numerous that every small ion attached itself to a nucleus before reaching the terminal. The tubes which lost on heating the property of emitting uncharged McC tetianp and M‘Henry— Uncharged Nuclei. 297 nuclei regained that property after a certain interval and under certain conditions. The following experiments will be quoted as examples. A glass tube was strongly heated in a Bunsen flame and “ fatigued.” It was left open, exposed to the atmosphere for a few days. When again , : - & , examined it had recovered, the ratio = being almost as great as when C; the tube was fresh. Heating for fifteen minutes suffices to drive off nearly all the nuclei, the value of C. falling off almost to unity. It might be expected that the effect would be recovered immediately on cooling. This is not the case, nor is the effect recovered if the tube is left in the apparatus overnight closed in. Thus several glass tubes which had been fatigued were found not to recover if left for several days in the apparatus closed in, but recovered fully if left overnight fully exposed to the atmosphere. The nuclei driven off by heat from glass tubes can also be charged by exposure to ultra-violet light. The apparatus is in this case identical with that described in the previous experiments. The uranium is, of course, excluded. The tube is heated, and the nuclei so obtained are exposed during their passage through the quartz tube to the ultra-violet light of the spark, which is a few mms. distant. A current voltage curve for the positive ions thus obtained gave a mobility of -0009. When a tube is heated and fatigued, as above described, the nuclei which can be charged by passing over uranium are recovered in about twenty-four hours when the tube is exposed to the atmosphere. The property of emitting nuclei, which could be charged by ultra-violet light, is not regained so rapidly. Thus a tube which recovered in twenty-four hours the property of emitting nuclei which were charged by the uranium small ions gave off no nuclei which could be charged by ultra-violet light. The latter property, however, it regained in about a week. It may be that the nuclei are not acted upon by ultra-violet light until they reach a certain size, and that in the initial stages of recovery of a glass tube the nuclei given off are small. The ultra-violet effect is probably due to the photo-electric emissions of electrons from rather big particles. If this is the case, we should expect that many of the electrons would be lost by diffusion to the sides of the vessel before becoming attached to neutral particles. We should thus expect the positive numbers to be greater than the negative. An experiment was performed with a new glass tube to investigate this point. Positive and negative numbers were taken alternately, the spark running continuously, and the tube heated for each reading at a different SCIENT. PROC, R.D,S., VOL. XVI., NO. XXIV, 2p 298 Scientifie Proceedings, Royal Dublin Society. place. The negative numbers were very small compared with the positive the rate of flow of the air being slow (1200 ces. per min.). The saturation current was ten times as great for positive as for negative numbers. The positive ions were saturated between 320 and 400 volts, giving a mobility between ‘0015 and -0012. The negative ions had a lower saturation point, and were thus smaller than the positive. ‘The numbers obtained are shown in the following table :— Currents obtained on heating glass tube and charging nucler by ultra-violet light. Volts (+) Current (+) Volts (—) Current (—) 80 50 80 13°3 160 100 160 20-4 240 139 240 22°7 320 200 320 217 400 227 440 192 On the other hand, as might be expected, it was proved in many experi- ments that when nuclei are driven from a glass tube by heat and passed over uranium, the resulting numbers, positive and negative, are equal, any charged products from the glass tube being previously removed to earth, as in our other experiments. It was thought possible that passing dry air through the glass tube would remove the matter which gives rise to the condensation nuclei driven off by heat. A glass tube, however, through which dry air had passed for several hours was unchanged, and gave the same effect as before the drying. Whether the air drawn through the glass tube while heated is dry 01 moist makes no difference in the numbers of nuclei emitted. : In dealing with soft glass tubes a second-type of nuclei emission was encountered with stronger heating. If a soft glass tube be heated with a luminous flame, the emission falls off almost to zero in ten or fifteen minutes. Tf the fatigued part be now heated with a strong Bunsen flame, the emission is very much increased, and remains constant for a long time, no sign of fatigue being observed. In this second case the glass tube glows, and the nuclei are due probably to the decomposition of the glass itself, McCuruianp anp M‘Henry—Uncharged Nucle’. 299 III.—NvucLEI DRIVEN OFF FROM Merats By HEAT. A series of experiments were conducted on the nuclei driven off by heat from a platinum wire electrically heated. The platinum wire used was thin, having a resistance per cm. of about 15 ohms. The air after being filtered and, if necessary, dried, passed through a wide glass tube. Through corks in the ends of the tube thick copper wires were introduced supporting about 5 cms. of the platinum wire. A variable resistance and a sensitive ammeter completed the heating circuit. The wide glass tube containing the platinum wire was heated strongly in a Bunsen flame before being used, to expel all the nuclei from the glass. This precaution was, perhaps, unnecessary owing to the width of the glass tubes, whose diameter was about 6 cms., and which did not become very hot in the experiments. The remainder of the apparatus was similar to that used in the experi- ments on nuclei produced by ultra-violet light. The nuclei emitted by the platinum wire were passed over a long terminal connected to earth, then over uranium, and were detected at the short terminal connected to the electro- meter. The mobility of the nuclei when charged was deduced from a current-voltage curve as before, and gave an idea of their size. For simplicity, we shall first deal with the nuclei emitted when the wire has been kept for some time at a dull red heat. The emission then becomes very steady, and seems to be a function of the temperature alone, so that when the velocity of the air through the apparatus remains constant the number and size of the nuclei depend solely upon the current through the wire. For example, if the emission is measured at a given temperature, and the wire be kept then at a much higher temperature even for hours, on decreasing the temperature of the wire to its former value the emission of the nuclei as deducted from the ratio i It was thought that the nuclei sbfed have been formed by some catalytic action of the platinum upon the moist air, in which the platinum itself was unchanged. Whether the air was moist, however, or carefully dried over calcium chloride and phosphorus pentoxide made not the slightest difference in the numbers. The nuclei were, therefore, due in all probability to the disintegration of the platinum itself. is found to be unaltered. The size of the nuclei increased with the temperature of the wire. 2 was 1 measured for currents of 1:2, 1:3, and 1:5 amperes through the wire, the air velocity through the apparatus remaining constant. A current- 300 Scientific Proceedings, Royal Dublin Society. voltage curve was plotted for C,, and the mobilities deduced from the saturation voltages. The results are seen in the following table :— 1 “ t ‘ "947 th oieiel re. e | valet Cs) j Mobility. | i i 1-2 | 25 70 013 1-3 72 120 008 15 > 96 > 400 < *0022 When 1°5 amperes were used saturation did not occur even at 400 volts, the highest voltage obtainable at the time. The numbers, however, show clearly how the size of the nuclei depends upon the temperature of the wire. In observing the emission from platinum at a given temperature, no falling-off is observed even after hours; but if the same wire is heated from day to day, a gradual falling-off is observed, although as long as the wire is glowing nuclei are always emitted in large numbers. In addition to the nuclei given off by platinum when glowing, nuclei are also emitted at lower temperatures by fresh wires, but the effect falls off rapidly to zero, and is thus not easily detected. A fresh wire was heated by currents from °5 ampere up, and the values of C, and C, taken at each reading No nuclei were observed at ‘5 amp. nor at ‘8 amp.; but when the current was 1:0 amp., C2 was originally four times C;, but rapidly of thes current. fell off to a value near C,, showing that the emission of nuclei fell off to zero. The wire just began to glow at 11 amps., and at that low temperature, although the emission never fell off to zero, the wire, when fresh, emitted more nuclei than after it had been heated for some time. Taking another new piece of wire, the following values of C, and C, were got for the given currents through the wire. C, and C2 are given in arbitrary units :— Ci : 2 6 D 16 C2 at 0-5 Amp... : 16 » at O08 ,, i % 53 Successive readings at 08 ,, . ° 20 =f x, wat lO:8 sey5 ee: 15 Thus a large number of nuclei were emitted when 0°8 ampere passed through the wire. Three readings, each taking about one minute, were made McC.exianp anv M‘Henry— Uncharged Nuclet. 301 of C,, The numbers show that the emission fell off to zero in about three minutes. On increasing the current to 1:0 ampere more nuclei were emitted (although the wire was not glowing). The emission in this case also fell off almost to zero, but took a much longer time than that at 8 amp. The following values of C2. and C, were got at 1:0 ampere :— C; = 6 3 6 5 0 14:5 ( Ce (1st Reading) 0 ° . 52°6 ) | yi (2nd) sy, - ) : ; i 33-3 | { pp (XE Bp ) 9 5 3 28°6 ? | (athe eee) 24-4 | Loy (IR. op ; i ; 23:3 J Ci again : : ; 4 14°5 It will be seen that the rate of emission gradually decreased, but even after about five minutes was still appreciable. On leaving the current at 1-0 amp. for thirty minutes, C, was found to be 15-4 and C, 14:0, so that the emission of nuclei had practically ceased. On increasing the current through the wire to 1:2 amperes the wire glowed, and Cz was 37, while C, was 15:3. The same values were obtained after twenty minutes’ heating. The initial emission at 0°8 and 1:0 amperes was not regained by the wire even on being laid aside for over three weeks. It is probably due to the emission of occluded substances in the platinum or to surface impurities received, for example, in handling. A fresh platinum wire which was heated in a Bunsen flame before being examined as above did not emit any nuclei at the temperatures corresponding to 0:8 and 1:0 amperes. With a wire made of ni-chrome, an alloy of nickel and chromium, results were obtained similar to those described for platinum. The wire in this case began to glow at a current of about 3 amps., but nuclei were emitted from the fresh wire at 1-5 amperes. This initial emission rapidly fell off to zero. The emission at glowing temperatures remained constant for hours, but in using one wire from day to day a marked falling-off is observed. This was clearly shown by the values at 40 amps. of os which fell in about a week 1 from the value 5:4 to 2°5, 1:4, 1:25, and finally 1-2. 302 Scientific Proceedings, Royal Dublin Society. SUMMARY. 1, The nuclei produced by ultra-violet light in moist air are detected by an electrical method, and their size indicated by their mobility, when charged in an electrical field. The number and size of the nuclei depend upon the time of exposure to the light, the intensity of the light, and the amount of moisture in the air. The nuclei are probably small water-drops which owe their formation to the production in the moist air by ultra-violet light of some hygroscopic substance, such as hydrogen peroxide or oxides of nitrogen. An account of work by other experimenters is given, and the evidence, though not conclusive, points to hydrogen peroxide as the parent substance of the nucleus. Some experiments were performed in air containing, instead of water, the vapours of methyl alcohol, ethyl alcohol, and toluene respectively. No nuclei are found by our method, probably because the nuclei grow so big, and decrease so much in number, that they fail to affect the currents in our apparatus. The presence of very slight traces of either of these vapours in water-vapour prevents the formation of nuclei such as can be detected by our apparatus, and probably for the same reason. II. The nuclei which are emitted by a glass tube when heated not too strongly can be detected and measured in a manner similar to that described for the nuclei produced by ultra-violet hght. The emission of nuclei when the heating is not too strong lasts only a few minutes, but the tube “recovers” under certain conditions. The nuclei can be charged by ultra-violet light, probably by the photo- electric discharge of electrons from the rather large particles. The positive ions are found to be much more numerous and of much greater size than the negative. In all probability only particles greater than a certain size are acted on by the light. The electrons then become attached to the smaller particles or the air molecules. The negative ions thus formed are numerically more decreased by diffusion to the walls of the apparatus than the larger positive ions. oy The experiments on the recovery of glass tubes fatigued by heat seem to indicate a difference between those nuclei which can be charged by ultra- violet light and those which can be charged by the small ions produced by PLATE VIII. XVI. SCIENT. PROC. R. DUBLIN SOC., N.S., VOL SCIENT. PROC. R. DUBLIN SOC., N.S., VOL. XVI. PLATE IX. os SCIENT. PROC, R. DUBLIN SOC., N.S. VOL. XVI. PLATE X. ARKO 09 70 0 cua ¥ . Aa e 16. 17. 18. 19. 20. 21. Hee v nued. A Vibrating-F ; 1e Rectifier for on Curr 8. By Jonn J. Dowuine, M.A., F.INST.pP., vid J. T. Har (Baltuary, 1921.) 6d. A Sensitive Valve Method for tae Masurement of Capacity, with some Important Applications. By JonrJ. Dowtine, u.a., r.inst.e. (February, 1921.) 6d. A Direct Reading Ultra-Micrneter. By Joun J. Dowtine, M.a., M.R.1.A., F.InsT.P, (March, 1921,; | Studies in the Physiograph and Glacial Geology of Southern Patagonia. By H. G. Fenton. (Plat: ¥, Vi, and VII.) (March, 1921.) 4s. 6d. Award of the Boyle ladal to Groren H. Prruyprincs, B.sc., PH.D., 1921. (March, 1921). .5. The Concentratia xd Purification of Alcoholic Fermentation Liquors. Part 1.—T Distillation in Steam of certain Alcohols. By Joszpa Rerty, M.A., DO. MRA. F.a.c.. and WiLrRED J. HickINBOTTOM, M.SG., A.1.C. (Avast, 1921.) ; 22.,whe “ Browning” and ‘ Stem-Break” Disease of Cultivated Flax (Linum usitatissimum), caused by Polyspora lini n. gen. et sp. By H. A. Larrerry, 4.x.c.sc.1., Assistant in the Seeds and Plant Disease Division, Department. of Agriculture and Technical Instruction for Ireland. [Communicated by Dr. George H. Pethybridge, s.sc.| (Plates VIII-X.) (August, 1921.) 23. A New Principle in Blowpipe Construction. By H. G. Brcxmr, a.R.¢.8¢.1., a.1.c., Demonstrator in the Royal College of Science, Dublin. (August, 1921.) 24. Uncharged Nuclei produced in Moist Air by Ultra-Violet Light and other Sources. By the late Prof. J. A. McCuexuanp, F.r.s., and J. J. M‘Henry, u.sc., University College, Dublin. (August, 1921.) [ Nos. 21 to 24, price 7s.| DURLIN : PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIBBS. 10. ale 13. 14, 15. SCINTIFIC PROCEEDINGS. VOUME XVI. . A Cryoscopic Method for the Estimatiy of Sucrose. By Henry H. Dixon, sc.D., ¥.k.s., and T. G. Mason, m.a., sc.. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malaide, Co. Dublin. By Lous B. Suyvrn, B.a., sc.s. (Plates 1, II.) (Februry, 1920.) 1s. . The Application of the Food-Unit Method tothe Fattening of Cattle. By James Winson, m.a., B.sc. (Plates III, 1V.) February, 1920.) 1s. . The Holothurioidea of the Coasts of Ireland. ByAnne Li. Massy. (April, 1920.) 1s. Photosynthesis and the lectronie Theory. By Henry L. Drxon, sc.d., FR.S., and Horace H. Pootz, sc.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Winriai Brown, B.sc. (March, 1920.) 6d. . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By 1%. Mason, m.a., sc.B. (April, 1920.) 1s. . The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Wittiam Brown, B.sc., and Parrick O’CaLLacuan, A.B.0.80.1.,4.1.0. (August, 1920.) 6d. Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. LronarD, a-R.C.SC.1., B.SC., PH.D., and AGNES Browne, A.R.C.SC.1., B.sc. (August, 1920.) 6d. An Investigation into the Causes of the Self-Ignition of Hther-Air Mixtures. By the late Professor J. A. McCrrnuanp, p.so., F.R.s., and Rev. H. V. GILL, S.J., D.S.0., M.c., .a., University College, Dublin. (August, 1920.) 6d. The Influence of Electrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosmpH Reiniy, M.a., D.sv., F-R.C.SC.1., and Witrrep J. Hickinsorrom, B.sc. (October, 1920.) 6d. Notes on some Applications of the Method of Distillation in Steam. By JosEpH REILLY, M.A., D.SC., F.R.C.Sc.1., and WinFRED J. HicxinzorrTom, B.sc. (October, 1920.) 6d. The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By W. H. Aprnny, D.sc., A.R.C.sc.1., F.1.c., and H. G. BECKER, A.R.C.SC.1., A.1.c. (September, 1920.) 6d. A Determination, by means of a Differential Calorimeter, of the Heat produced during the Inversion of Sucrose. By Henry H. Dixon, sco.p., F.R.S., and Nicen G. Batt, .a. (December, 1920.) 6d. The Measurement of very Short Time Intervals by the Condenser-Charging Method. By Joun J. Dowuine, u.a., F.1nst.P., and Donat Donnexty, usc. (In conjunction with the late Prof. J. A. McCurtuanp, F.r.s.) (February, 1921.) 6d. ee THE SCLENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.S.), Nos. 25-29. AUGUST, 1921. 25.—BIOLOGICAL STUDIES OF APHJS RUMICIS L. A.—Apprar- ANCE OF WINGED Forms. B.—APPEARANCE OF SEXUAL Forms. By J. Davipson, D.Sc. Krom the Entomological Department, Institute of Plant Pathology, Rothamsted Experimental Station, Harpenden, Herts. [Communicated by Professor G. H. Carpenter.]_ 26.—THE OCCURRENCE OF DEWALQUEA IN THE COAL-BORE AT WASHING BAY. By T. Jounson, D.Sc., ¥.L.8., Professor of Botany, Royal College of Science for Ivelandyand Jane G. GILMORE, B.Sc. (Plates XI, XII.) 27.—A SIMPLE FORM OF APPARATUS FOR OBSERVING THE RATE OF REACTION BETWEEN GASES AND LIQUIDS, AND ITS USE IN DETERMINING THE RATE OF SOLUTION OF OXYGEN BY WATER UNDER DIFFERENT CONDITIONS OF MIXING. By H. G. Bacxer, A.R.C.Se.1.; A.1.C., Demonstrator _ in Chemistry, Royal College of Science, Dublin. 28.—_THE OCCURRENCE OF A SEQUOIA AT WASHING BAY. By Y. Jounson, D.Sc., F.L.S., Professor of Botany, Royal College of Science for Ireland, and Jane G. Gitmorz, B.Sc. (Plates XIII, XIV.) 29 THE SOURCES OF INFECTION OF POTATO TUBERS WITH THE BLIGHT FUNGUS, PHYTOPHTHORA INFESTANS. By Paun A. Murpuy, B.A., A-R.C.Sc.1., Assistant in the Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. [Authors alone are responsible for all opinions expressed in their Communications. | < DUBLIN : nn PUBLISHED BY THE ROYAL DUBLIN SQCIETY, LEINSLER HOUSH, DUBLIN. Vi pave WILLIAMS AND NORGATE, oR ei 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1921. Price Siz Shillings and Sixpence. Royal Bublin Soctety. FOUNDED, A.D. 1731. INCORPORATED, 1749. —~ SCIENTIFIC MEETINGS. Tan Scientific Meetings of the Society are usually held at 4.15 p.m. on the fourth Tuesday of every month of the Session (November to June). The attention of authors is directed to the following extracts from the regula- tions governing the presentation of communications at Scientific Meetings of the Society, and the publication of papers in the Scientific or Nconomic Proceedings :— «An author who desires to present a communication at a Scientific Meeting of the Society shall furnish to the Chief Mxecutive Officer for Science a fortnight previous to that meeting the title and an abstract (not exceeding 200 words) of his communication, together with an estimate of the time required for presentation. “An author who desires to have a communication published forthwith .as a paper in the Proceedings of the Society shall furnish to the Chief Executive Officer for Science the full text and necessary illustrations at least a week before the meeting of the Committee of Science at which he desires publication to be considered. Meetings of the Committee of Science are usually held on the second Tuesday of each month from November to June inclusive. “« A paper may be published either in the Scientific Proceedings or the Economic Proceedings of the Royal Dublin Society, as directed by the Committee of Science. The copyright of papers so published shall be the preperty of the Society. « Wifty copies of each paper shall be supplied free of cost to the Author or to each Joint Author. “ Additional copies may be obtained at Cost Price if applied for by the Author when returning the proofs.” Authors are requested to apply to the Clief Ixecutive Officer for Science for further information concerning the preparation of papers and illustrations for press. McCirLianp anp M‘Hrnry— Uncharged Nuclet. 303 uranium. ‘I'he difference is probably, however, one of size, as in the initial stages of recovery the particles emitted may not be so large as they become after a prolonged period of rest. Nuclei of a different type are emitted at glowing temperatures. In this case the glass does not become fatigued, and the nuclei are probably produced by the decomposition of the glass itself. IIL. When platinum wires are heated electrically, nuclei are driven off before the wire begins to glow. This emission lasts only a short time, just as in the case of glass, and is probably due to occluded impurities. When these impurities are driven off by heat, the wire emits no nuclei until it begins to glow. The emission at glowing temperatures is very constant, and does not fatigue. The size and number of the nuclei emitted increase with the temperature. Similar results were obtained with a wire of ni-chrome, an alloy of nickel and chromium. Thus in all cases examined where nuclei are driven off from a substance by heat an initial emission takes place at low temperatures which falls off rapidly. The emission of nuclei when the substance glows is practically constant. SCIENT. PROC. R.D.S., VOL. XVI., NO. XXIII. 2a f a0" | XXV. BIOLOGICAL STUDIES OF APHIS RUMICIS L. A.—APPEARANCE OF WINGED FORMS. B.—APPEARANCE OF SEXUAL FORMS. By J. DAVIDSON, D.Sc. From the Entomological Department, Institute of Plant Pathology, Rothamsted Experimental Station, Harpenden, Herts. (COMMUNICATED BY PROFESSOR G. H. CARPENTER.) [Read Aprit 26. Published Aveusr 29, 1921.] CONTENTS. PAGE < . ’ PAGE I, Inrropuction anp Exprrtentan Dara. (e) Successive viviparous generations on f i! Beans, 6 i 0 q 308 a) Fundatrix generation on Euonymus, 300 3 (a) pee § a : (£) Euonymus infected from Beans, 309 (b) The First viviparous generation on Euonymus 305 | 11. Genera Discusston. Be us 0 9 : r] a (c) Successive viviparous generations on (A) Appearance of winged forms, : 310 Euonymus 305 (B) Appearance of sexual forms, c 315 ; 0 0 : (d) Successive viviparous generations on III. Summary, . 6 8 A . 321 Rumex, » 2. « . 807 | IV. Rerzrences ce : 322 ’ J.—INTRODUCTION AND EXPERIMENTAL Data. The results embodied in this paper have been derived from extensive breeding experiments with this species in captivity during 1913-14 in Germany and 1920 at Rothamsted. Ova were found on March 25th, 1914, and were hatched out in the laboratory, the Fundatrices being transferred to small Huwonymus bushes, Rumex plants, ete., which were grown in pots, and kept covered with muslin bags. As each succeeding generation became adult, some of the offspring were transferred to new plants. In this way the history of the aphids used was known, and the exact generation with which the various plants were infected. Unfortunately, for the 1920 experiments ova could not be found, so that the colonies in these experiments were started from an early, viviparous, parthenogenetic generation found on Huonymus europaeus on the Rothamsted farm on May 10th, 1920. The plants were grown in 10” pots in a large open glass-house in summer, and a warm green-house in winter. The daily maximum and minimum temperature of the green-house throughout the winter is shown in fig. 2. Davipson—Piological Studies of Aphis rumicis L. 308 I wish to express my thanks to Dr. A. D. Imms for the helpful sugges- tions he has given me during the progress of the experiments, and to Mr. Ward Cutler for his criticism when writing the paper. (a) Fundatria Generation on Huonynvus. Young aphids {1st instar), which hatched out from ova of Aphis rumicis, from Huonymus, were reared on young twigs of Huonymus europaeus, which were kept in wet sand in the laboratory. They all developed into a.v. ? 2, being the adult Fundatrices. In six experiments it was found that Fundatrices became adult in 8 to 10 days after birth, and commenced to produce young within 2 to 4 days after the last moult. The young in all cases were apterous viviparous females. (b) Lhe First Viviparous Generation on Huonymus ewropaeus. The offspring of the Fundatrices are the first viviparous generation. In these experiments, individuals of the first viviparous generation born between April 13th and 18th, 1914, became adult in 9-11 days after birth. In all cases they gave rise toa.v. 2 @ , and the first young were produced in seven instances on the same day that the last moult took place. In the remaining cases, the first young were produced in 1 to 3 days after the last moult. They were grown on twigs of Hwonymus in wet sand. (c) Successive Viviparous Generation reared on Huonymus. Small Luwonymus bushes, about a foot high, were grown in pots and infected as required, each plant being covered with a muslin bag. Observa- tions were made as recorded below. It should be noted that when winged forms are transferred to a new plant there is often great difficulty in getting them to settle down on the plant, and several days may elapse before they produce young on it. Sometimes, out of as many as fifteen winged forms, only two or three will settle down and produce on the plant, the remainder collecting in the top of the muslin cover. By restricting the space round the plant, this difficulty can be somewhat overcome. Euonymus A.—11. 4.14. Infected with two Fundatrices reared from ova on Euonymus. 23.4.14. Adult w.v. @ 2 present; only a few a.v. 2 9 present. Euonynus A,.—23. 4.14. Infected with 4 w.v. 2 ?, Ist v. gen. from A. 12.5.14. Plant is not making young growth and is “woody.” Aphids small size, restless, and wandering over the plant; some immature w.v. @ ? pre- sent, but majority are a.v. 9? 9, somea.v. 2 ? adult. : EHuonymus Az,,—25, 4.14, Infected with 3 w.v. 2 9, Ist v. gen. from A. 2Q2 306 Scientific Proceedings, Royal Dublin Society. 8.5.14, 2nd v. gen. are all a.v. 2 9, large size, dull black in colour, with parts of legs and antennae yellowish white in contrast; the plant making young growth. 24.5.14. Plant heavily infested, aphids small size, many w.v. 2 @ present. Euonymus A;.—27. 4.14. Infected with 2a.v. 9 ?, Ist v. gen. from A. 18.5, 14. Plant is not making young growth; aphids small size; majority are w.v. @ 9, only 3 a.v. ? ? present. ( Huonynmus Ay—12.5.14. Infected with 5 a.v. 9? 9, 2nd v. gen. from Aj. 24,5,14. drd v. gen. are a.v. 9 9 and w.v. @ ?. Euonymus A;—10. 5.14. Infected with 3 a.v. 9 9, 2nd v. gen. from Az. 24, 5,14. 3rd v. gen. practically all w.v. 2 9; plant has young growth. Huonymus A,.—24. 5.14. Infected with 4 w.v. 2 ?, 3rd v. gen. from A;. 10. 6.14. 4th v. gen. small numbers but large size, all a.v. 9 9, some adult. Huonymus A;,—10. 6.14. Infected with 8 a.v. 9 @, 4th v. gen. from Ag. 20. 6.14. Some aphids of 5th v. gen. are adult; all are w.v. ? 9; plant has young growth. Luonymus Ag—20. 6,14. Infected with 6 w.v. 9 ¢, 5th v. gen. from A,. 26. 6.14. All the w.v. 2 @ have died without producing any young. Rein- fected with 8 w.v. 9 @ from A;. 6.7.14, About 20 aphids 6th v. gen. produced; killed off the winged mothers. 15.7.14. 6th v. gen. not yet adult; very small size; plant not making young growth; transferred the aphids to a new plant. 23.'7.14. Some of the aphids now adult ; all are a. v. ? @ small size; plant is not making young growth. 30.7. 14. Aphids very small size; plant woody ; several aphids of 7th v. gen. produced. Owing to the unavoidable termination of the “A” series in July, 1914, the experiments were continued during the summer of 1920. As it was not possible to find ova of Aphis rumicis, the experiments were started with an early apterous viviparous generation found on HLuonymus ewropacus on 10th May, 1920. Winged forms were present, so it might be presumed that these were probably the second viviparous generation after Fundatrices. Allowing ten days for development of each generation, this would mean that the eggs hatched about April 10th. Small Huonymus trees grown in pots and covered with muslin bags were infected as recorded below. By “cutting back” the bushes, young growth was obtained throughout the summer. Huonymus E,.—10. 5. 20. Infected with two a.v. ? 9 from Euonymus europaeus. 15.7. 20. Plant heavily infested. Huonymus E,.—15. 7. 20. Infected with 8 a.v. ? ? from Euonymus E,. 28.7. 20. About 30 aphids produced, big and healthy; plant has young growth. 10.8. 20. Many winged forms produced; afew ¢ ¢ present. Davipson— Biological Studies of Aphis rumicis L. 307 Euonymus E;.—18. 8. 20. Infected with a.v. ? 9 and winged forms from Euonymus E,. 24.8.20. A.v.? ? producing young; only one winged fort producing, remainder will not settle on the plant. 31.8.20. Only a few Aphids on the plant, small size, apterous. 14.10.20. The plant is shedding its leaves; a good number of aphids present; small size; many winged forms; ¢ ¢ and sexual @ 2 present, Fundatrix IS? V. gen. 2nd V.gen. ac V.gen. 4th V.gen. 5*"V. gen. 6°" V. gen 7A gen ethy gen. Fie. 1. Table illustrating appearance of winged and apterous forms of Aphis rumicis on Rumex. Ist v. gen., etc., = lst viviparous generation; @e =a.v. ? ?; O=w.v. 2 ?. (d) Successive Viviparous Generations reared on Rumen. Rumex plants were grown in pots and infected witha. v.? and w.v.? ? respectively, in the different generations. The first infection was to Rumex No. 1 with one Fundatrix of Aphis rumicis from Huonymus ewropaeus on 13. 4. 14. The last infections were Rumex Nos. 16 and 17 with w.v. 2 Qanda.v.? @ respectively, of the 7th v. gen. from Rumex No. 14 on 16. 7. 14. The aphids made rapid progress, and were usually of large size and dull olive green colour. In some generations only apterous forms were produced, although winged forms always appeared in later generations, when the aphids were left to re- produce on the plant. The appearance of the winged and apterous forms is shown in fig. 1. 308 Scientific Proceedings, Royal Dublin Society. It must be noted, however, that, owing to technical difficulties, where so many breeding cages are under observation, it is not possible to breed out all the offspring. The table below must therefore be taken as only expressing the general trend of the appearance of winged and apterous viviparous females. (e) Successive Generations of Aphis rumicis on Broad Beans, (Variety : Sutton’s Prolific Longpod.) Broad Beans B.—16. 6. 20. Infected with 2 a.v. 2 2; offspring of winged migrants on Broad Beans. 12.7. 20. A large number of aphids present, a. v. 9 2 and w.v. 2 @. Broad Beans B,.—10. 7.20. Infected with 4 a.v. 9 2 from B. 19.7. 20. Offspring adult. 26.7.20. One adult w.v. ? 9, remainder a.v. ? ?, large size, infestation heavy ; killed off the plant and infected a new plant with four of the a.v. 2? @. 10.8.20. Many a.v. 2 @ and w.v. 2 ? present. Broad Beans B,.—10. 7. 20. Infected with 4 w.v. 2? @ from B. 19.7. 20. Some a.v. 2 @ adult, large size. 26.7.20. Infestation heavy. Broad Beans Bs— 25.7. 20. Infected with 6 a.v ? ? from B,. 4. 8. 20. A.v. 2 @ and w.v. 2 @ present. 12.8.20. Many winged forms present. Broad Beans By—6. 8.20. Infected with 6 a.v. ? @ from B;. 27. 8. 20. Infestation heavy, a.v. 2 @ and w.v. ? ? present; killed off plant and in- fected a new plant with some of the apterous and winged forms. 14. 9. 20. Many a.v. 2? ? and w.v. 2 ? present; killed off the new plant. Broad Beans B;.—14. 9. 20. Infected with 5 a.v. 2? ? from By. 11.10. 20. A.v. 2 2; w.v. 9 ?,afew ¢ g, and two sexual ? 9 present. Broad Beans B,.—11. 10. 20. Infected with 10 a.v.? ? from B;. 29.10. 20. A fewa.v.? ?,many w.v.@ 9, and several ¢ ¢ present. Development slow owing to low temperature in glass-house. Nore.—From B, onwards the plants were kept in a warm green-house, the temperature varying as shown in chart, text (fig. 2). Broad Beans B;.—29.10. 20. Infected with 5a.v.? 9 from By, 18.11. 20. 3w.v.? ? adult, many immature winged forms, large size (sexuparae alatae), 26.11.20. All the aphids are w.v. 2 ? (sexuparae), andafew ¢ ¢. Broad Beans B;.—23. 11. 20. Infected with one of the a.v. ? mothers on B, 5.12.20. Four w.v. 2 ? and one a.v. 2 produced. Allowed w.v. ? @ to produce on the same plant. 26.12.20. The w.v. ? ? have produced oviparous ? @. Broad Beans Bs—(Two plants.) Infected one plant on 19.11.20, and the other on 25.11. 20, each with one w.v. 9 from B; Only sexual ? ? produced in each case; hence these are true female-producing sexuparae. Davinson— Biological Studies of Aphis rumicis L. 309 They produced 11 and 22 females respectively, which were adult on 5. 12. 20 and 9.12.20. A third w.v. ? produced only sexual 2 9. Broad Beans By.—5. 12. 20. Infected with a.v. 9 from B;. 23.12. 20. One a.v. ? produced, remainder are w. v. 2 @ ; total 20. Broad Beans B,,.—1. 12. 20. Infected with 10 w.v. 2 ? from B,. 26.12. 20, All the aphids produced are sexual? ?. Broad Beans By.—23. 12.20. Infected with the one a.v. 2 from By. 9.1.21. Offspring consists of 10 w.v. 2 9 adult, 2 adult a.v. 2 9, and remainder nymphs of winged forms; some 3 3. Broad Beans By3—26. 12. 20. Infected with 10 w.v. 2? @ from Bp. 1.1.21. Four w.v.2 Qhave produced young. 12.1.21. All offspring are oviparous 2 2; some are adult. Broad Beans By.—10. 1. 21. Infected with 2a.v.2 Qfrom B,,. 5.2. 21. w.v. 2 9(sexuparae), ¢ g, and 4a.v. 2 9 produced. Broad Beans B,;—10. 1. 21. Infected with 5 w.v.?2 Qfrom By. 24.1. 21, 18 oviparous $ 2 produced. Broad Beans By.—24. 1.21. Infected with 4 a.v. 2 2 from By. 3, 2. 21. 3 g dg; several w.v. 2 2 andla.v. 2 produced. Broad Beans By,,—5. 2.21. Infected with 5 w.v. 2 Ofrom By. 21. 2.21. Only one w. v. 2 reproduced, and gave 6 oviparous 9 °. Broad Beans Bys.—3. 2.21. Infected with the 5 a.v. 2 2 from By. 23. 2.21. 2a.v.2 9; several ¢ g; several w.v. 2 2 present. 19. 2. 21. The 5 original a.v. 2 mothers put on a new plant, and on 2, 3. 21 they had produced 3a.v. 2 2,some f fg, andw.v. 2°. Broad Beans Byy— 23. 2.21. Infected with 5 w.v. 2 9 from By. 10.3, 21. All oviparous 2 2 produced. Broad Beans Bx.—2. 3.21. Uhree plants infected each with one a.v. 2 from By. 26.3.21. 5a.v. 2 2, one g, and 40 w.v. 2 2 produced. (f) Huonymus Huropueus Infected from Broad Beans. Huonymus Ks—16.6.20. Infected with a.v. 2 9, offspring of winged migrants on Broad Beans. 29.6. 20. Many aphids present, mostly a.v. 2 2, but some w.v. 2 °. Euonymus E;.—10. 7.20. Infected with 25 w.v. 2 ° from Beans B. 18.7. 20. Many colonies- produced beneath the leaves. 5.8.20. Aphids small size, dark colour, many winged forms ; infestation heavy ; plant has no young growth. Euonymus E,—19. 8. 20. Infected with 15 w.v. 2 2 from Beans By. 31. 8. 20. Many aphids produced, some adult a.v. 9 2 14.9. 20. Plant heavily infested ; many aphids wandering over the plant and muslin cover. 310 Scientific Proceedings, Royal Dublin Society. 14.10. 20. Plant shedding its leaves; a.v. 9 2, w.v. 22, 5 ¢ and sexual © 2 present; ova laid. Euonymus Ey.—14. 9. 20. Infected with 15 w.v. 2 2 from Beans By. This plant has a lot of young growth. 14.10.20. Many aphids present; 3 ¢ and sexual ° 2 present. Euonymus E,—11. 10. 20. Infected with 10 w. v. 2 2 from Beans B;. 14, 10.20. Only two winged forms producing on the plant. 30.10. 20. Off- spring now adult, all sexual 2 2, 16 in number. Euonymus E,,.—(Kept in warm greenhouse.) 24.11. 20. Infected with one w.v. 2 from Beans B,. 9.12.20. Offspring all oviparous 2 2, 24 in number, some adult. 26.12.20. Ovip. 2 9, large size ; no ova produced. Euonymus E,,.—25. 2. 21. Infected with one w. v. 2 (sexupara) from Beans By. 12.3. 21. Only oviparous 2 2 produced, some nearly adult. 12. 4, 21. All the oviparous 2? $ alive, have distended abdomen; no ova laid. Ova in the bodies, but females not fertilized. II.—GENERAL DISCUSSION. A.— Appearance of Winged Forms. It was almost universally held by the earlier investigators who studied the biology of Aphids that poor sap conditions obtaining in plants resulted in the production of winged forms. Later investigations indicated that tempera- ture and humidity may also be important factors. The results obtained by these earlier workers have been so often reviewed in the literature on Aphids that it seems unnecessary to review them again. Starting therefore with Mordwilko (1907-1909), who in his extensive observations on the biology of Aphids-gives a general idea of the views obtaining at that time, we see that food conditions, probably correlated with temperature, were accepted as the important factors. Investigations during the next few years on the life cycle of Aphids from the cytological aspect lead one to the view that, although external factors may have some influence on the production of winged forms, there is in all probability some internal inherent mechanism at work which is the important factor. Later results obtained from breeding experiments, notably by Klodnitzki (1912), Baker and Turner (1916), and Matheson (1919), in which winged forms were obtained in many of the parthenogenetic generations, would seem to support this view. ‘The situation at present is, however, by no means clear, and external influences, especially of food and temperature, would appear to be very important factors in the results obtained by some recent workers, notably Ewing (1916) and Shinji (1918). In discussing this Davinson—Brological Studies of Aphis rumicis L. 311 question further, however, I hope to show with reference to the results of my experiments that those obtained by the latter authors are by no means conclusive. Ewing, working with Aphis avenae Fab., considers temperature is the determining factor. ‘The optimum temperature of 65°F. resulted in only apterous forms being produced during 20 parthenogenetic generations. Klodnitzki obtained winged forms along with apterous forms in most generations when breeding the dark green and brown varieties of Siphono- phora (Macrosiphum) vosae Koch, in normal and abnormally low temperatures. Similarly with Aphis hederae Kalt., over 35 agamic generations, winged forms appeared in low temperatures as in normal temperatures. There was a tendency for a.v. 2 2 to produce winged forms. Shinji investigated the influence of food conditions on the production of winged forms in Aphids. By growing cut stems of plants infected with aphids, in sand moistened with solutions of certain chemical substances, he obtained results which suggest that the percentage of winged forms produced was influenced. ‘his was especially the case with solutions of Mg. sulphate. The figures obtained are interesting, but a further investigation of the subject is desirable. It is necessary to know the history of the Aphids experimented with, and the number of the viviparous generation used. Comparisons between the effect on the offspring of w.v. ? 2 anda.v. ? 9 respectively should be observed. The varying individuality of the Aphids concerned, relative to an inherent tendency to wing development, should be known. Temperature and humidity factors must also be considered. It will be seen in iny experiments that winged forms appeared in most of the generations. Winged viviparous females produced as a rule only apterous viviparous females, although in one or two cases (see text, fig. 1) a few winged forms developed as well as apterous forms. Apterous viviparous females, offspring of w.v. 2 @, usually produced a varying percentage of winged forms, although sometimes only a.v. ? 9 were produced. This sequence was the same on different species of plants. Winged forms have been obtained on young succulent shoots of Huonymus ewropaeus and apterous forms on “ woody,” non-sueculent plants. Food conditions, therefore, would appear not to be the deciding factor. The tendency appears to be for a.v. 2? @ to produce a mixed progeny of w.v. | ? anda.v. ? $, or in some cases only a.v. 2? 9. W.v. 2 9,on the other hand, invariably produce a. v. 9 2, with only occasionally winged forms in later generations. It is seen that when a plant, either the winter host (Huonymus), or an intermediate host (Beans), is infected, either with a.v. 2 2 or w.v. 2 @, and the Aphids allowed to reproduce freely on the -ainqviedure} AT(ep WNWIUIM SMOYS OUT], LaMOT roinqeiadis, Ajimp wnwixvut sous aury teddg “IZGL ‘YOIVW 0} SOZEI “SIZ LOQUIOAON ‘ASNOH-NaGUH FO MIVHD TUALVUTANaY, “7 ‘Oly “HOUV]L *AUVOUAT | “RUVANVE “ATTN AIT (] “AALWAAO NT og Si Ol GS GW Se G SG v € 62 +c Gl wl 6G + O€ Sz o2 SI OI GS o€ St O@ Scientific Proceedings, Royal Dublin Society. 312 Davipson— Biological Studies of Aphis rumicis L. 313 plant, winged forms always appear in due course. Although at the time that winged forms appear in great numbers the plant may be heavily infested, the appearance of the winged forms is not necessarily connected with the condition of the plant, but due to the innate tendency of the apterous females to produce winged forms as described above. It would seem that the overlooking of this fact has clouded the observations of earlier observers, who correlated the conditions of the plant with the appearance of winged forms. The sequence of the development of winged forms on Rumex, Euonymus and Beans is well shown in figs. 1 and 3. ; The percentage of winged forms which may develop in any generation as offspring of a.v. 2 2, varies considerably. The tendency for a higher percentage of winged forms to develop in some cases on Sugar Beet, Red Beet, and Poppies, and also on some varieties of Broad Beans, has been observed in my experiments.! It is not at all clear, however, that this can he considered as due to the nature of the cell sap of the different plants acting on the Aphid metabolism. Owing to the wide variation in the numbers of winged forms produced, it seems that factors other than food and temperature must be looked for. In these experiments temperature and humidity varied, but the conditions were approximately the same for any series of experiments extending over the same period. From my observations one is led to the view that the appearance of winged forms is determined by internal factors. It seems probable, however, that environmental conditions may exert an influence by restricting or favouring metabolism, and thus affecting the production of winged forms ; but further experimental investigations in relation to temperature, humidity, and food factors and further cytological studies are necessary. It seems to the author that the methods adopted for breeding Aphids may largely account for the conflicting results obtained by various workers. When one considers the number of offspring that an Aphid mother can produce, it is obvious that only an extremely small percentage of the Aphids in any generation can be isolated to carry on further generations. From general observations on the progress of Aphis rumicis in colonies, there is a clearly indicated tendency for apterous viviparous females to eventually produce some winged forms. This inherent tendency may be more developed in some strains than in others, and a big element of chance occurs in the selection of Aphid mothers to carry on the generations. 1 Vide Davidson, J. (1921 *). 314 Scientific Proceedings, Royal Dublin Society. Before one can investigate the influence of external factors on production of winged forms, it is necessary to estimate by extensive breeding experi- ments the extent and relations of this inherent tendency to wing-production EVONYMUS BEANS MonTHS IV IV Te. ' O VI VI O Vi O Vil O Vill @ Ix (0) (3) O xX © O O O x ©) ©) O XI O O © XI er O O I ©) O I @) O © O O O aut Fig. 3. Illustrating, in successive generations of A. rwmicis, the appearance of winged and apterous forms, also of sexual forms, on Broad Beans and Euonymus. ‘The months are shown at the right- hand side. The series is being continued. == Fundatrices. @ Apterous viviparous females. O Winged viviparous females. © Winged sexuparae. @ Sexual females. 3 Males. in different agamic females. It is generally considered that the winged condition is the more primitive, and that the apterous condition has arisen Daviwson— Biological Studies of Aphis rumieis L. 315 later. In the loss of the wings, with the resulting change in body form, and in the loss of the sensoria on the antennae, is indicated a degeneration correlated with the more localized parasitic habits of the apterous form. It seems to me therefore that external factors can only be considered as likely to exert an influence in restricting or favouring what is already present as an inherent tendency, and not in producing two complex dimorphic forms. Doubtless, external factors—food and climatic seasonal conditions—extending over long periods of time, have influenced the winged and apterous condition in the agamic generations. (B).—Appearance of Sexual Forms. The early conception of the alternation of sexual and parthenogenetic generations in the life cycle of Aphids was that external conditions—food and temperature—were the chief factors accounting for the appearance of the sexual forms and for bringing the parthenogenetic generations to an end. The experiments of several early workers, such as De Geer, Kyber, and others, whereby, with certain species of Aphids, parthenogenetic reproduction was continued for a great number of generations, under favourable food and temperature conditions, seemed to support this conception. Further, the view readily fitted in with general observations on Aphids in the field, namely, the widespread distribution of agamic females in the summer months and the occurrence of sexual forms in the autumn. Comparatively little experi- mental work on the Aphididae has been carried out, however; and the views held regarding Aphids are largely influenced by the results of the numerous researches on other animals, particularly the Cladocera and Rotifera, although in these cases sex appears to be largely determined by the chromosome complex. The literature on this subject has been so frequently reviewed that it is not proposed to deal with it here. With the development of the cytological aspect of the life cycle of the Aphididae, notably by Morgan, Stevens, Tanreuther, and von Baehr, it became clear that internal factors, closely associated with the chromosomes, must be considered of first importance. Whatever influence external factors may exert, they can hardly be considered as determining the sexual forms, although in the progress of adaptation and selection these factors have, doubtless, played an important part in favouring or restricting the parthenogenetic generations. The more recent results obtained by breeding species of Aphids through several generations support the view that internal factors are the important ones concerned. 316 Scientific Proceedings, Royal Dublin Society. With individuals of some species, however, the parthenogenetic generations may be continued under favourable conditions for a long period, as has been shown in many cases. For example, Klodnitzki (1912) carried on agamic generations of Siphonophora rosae var. glauca, Buckton, for more than a year. Ewing (1916) carried on Aphis avenae for seventy-three parthenogenetic generations. It would appear that in some species the sexual forms develop after a definite number of agamic generations have been passed through ; but this is not the case in all species, as is seen in my experiments and in those of Davis (1914) with Callipterus trifolii, Monell. A further consideration is that, while certain individuals of the later summer agamic generations close the life cycle by producing only sexual forms, other individuals may also produce agamic females which continue the agamic generations under favourable conditions for a very long period. Klodnitzki carried on three cultures of Aphis hederae Kalt. over a number of generations. In two lines sexual forms appeared in the thirty-first generation, but no sexual forms appeared in the other line, even after forty-two genera- tions had been passed through. Davis (1915) observed a similar agamic series in Macrosiphum pisi Kalt. The continuation of parthenogenetic reproduction throughout winter under favourable conditions of food and temperature has been obtained in my experiments with Aphis rumicis. With the advent of unfavourable winter conditions in the field, the agamic generations would die off, but under mild conditions, and given a suitable food-plant, they may survive the winter, and slowly carry on the parthenogenetic reproduction. The finding of a colony of agamic females of Aphis rwmicis by Davidson (1914) on January 30th, 1913, on Lwonymus japonicus, which, when brought into the green-house, continued agamic reproduction throughout 1913, would seem to support the latter view. ‘This is an important consideration from the economic standpoint. It is possible, of course, that some late parthenogenetic generations, delayed in their development on the approach of winter conditions, would normally under favourable conditions lead to sexual forms. It appears that in Aphids the development of viviparous parthenogenetic generations, between the ova and the sexual forms, is an expression of adaptation and selection to environmental conditions, extending over the favourable seasons of the year. Certainly in nature the approach of winter conditions would normally be the factor limiting the extent to which the parthenogenetic phase could be extended, Davipson— Biological Studies of Aphis rumicis L. 317 In this respect it is interesting to refer to the life cycle of Aphis saliceti, a species in which the number of parthenogenetic generations is very limited, the sexual forms appearing in early summer. As shown by Klodnitzki, the ova hatch out in April and May. Some of the Fundatrices produce ¢ ¢ and sexual ? 2 ; others produce w.v. 2 %, which migrate to other willow trees, and produce ¢ ¢ and sexual 2 @. Passing now to Aphis rumicis, it is seen that the agamic generations are further extended over the favourable seasons of the year. Normally in nature, the approach of winter conditions will be the important factor in bringing these agamic generations to an end. In view of the cytological investigations in Aphids, it seems evident that some factor or factors associated with an adaptation of Aphids to seasonal conditions cause a radical change in chromozome segregation, resulting in the development of sexual forms, and the production of ova at a period which ensures the continuation of the species. It seems feasible to expect that with an extension of favourable seasonal conditions, such as would obtain in a warmer climate, there would be a corresponding extension (due to adaptation) of the agamic generations. This appears to be the case in Aphis avenae Fabr., and TVoxoptera graminum Rond. in America. It is clear from my experiments that certain of the apterous, parthenogenetic females may carry on the parthenogenetic strain throughout winter if given favourable food and temperature conditions. This cannot, however, be considered as wholly due to these two factors, because sexual forms and agamic forms appear together in each generation under the same environmental conditions. It is, I think, just a case of parthenogenetic 2 ?, which normally on the secondary host-plants would have died off on the approach of unfavourable weather conditions, being saved by keeping them in a suitable environment. One cannot say at present how far these cases occur in the Aphididae. It is highly probable that in breeding experiments, considering the very small percentage of the offspring used in each generation, it is only by chance that individuals are selected which have the continuing partheno- genetic tendency. It is seen in the series Broad Beans B, B,, B2, &c., as illustrated in fig. 3, that parthenogenetic individuals were carried on Beans from May onwards throughout the winter. In winter the plants were kept in a warm green-house, and fresh Bean plants were raised in succession for each genera- tion. Towards October, sexual forms appeared, and in some lines of the cultures the cycle closed by the production of these forms. Some of the 318 Scientifie Proceedings, Royal Dublin Society. infected plants were put into the warm green-house with the hope of carrying on the agamic generations, and the interesting result shown in fig. 3 has been obtained. After October had set in, there was a marked tendency to production of sexual forms, and the agamic forms in most cases died out. However, in the case illustrated in fig. 3 a few agamic a.v. 2 2 developed, and these were isolated, and their offspring carefully watched. In each succeeding generation large w.v. ? 2 (winged sexuparae, ¢ pro- ducers) and winged ¢ ¢ were produced; and in every case one or two agamic a.v. 2 2 also developed, which were isolated to carry on the next generation. It was not possible to follow out the exact numbers of each form produced, owing to the great number of pots and plants that would be required; but winged sexuparae were in the majority, with a comparatively small numberof ¢ g. The agamica.y. 2 2, on the other hand, were usually only one, two, or three in each of the later generations. The development period of these agamic individuals is somewhat longer than that of the summer generations, and the numbers of offspring produced considerably less. This may to some extent be due to low concentration of the cell sap of the plants, owing to conditions of winter sunshine and reduced assimilation. 1t will be seen that agamic and sexual forms in the same generations were obtained on Huonymus. The females produced by the winged sexuparae grew well on Broad Beans, and became larger than normal females; but in no case were ova laid, although on dissecting these females several large eggs were found inside them. These 2 in the experiments were not fertilized, owing to the fact that the Aphids were removed from the plants as they became adult. In one case ten oviparous ? ?, born on a Bean plant on 1st January, 1921, had not produced ova by 16th February, and by that time only two 2 2 were alive. When some of the females were dissected, they were found to contain several large eggs. Sexual forms were produced when the Aphids were kept continuously on Euonymus ewropaeus, and also when they were kept continuously on Beans. As referred to in a previous paper (1921), sexual forms have been recorded in the field on Ruwmex, Sugar Beet, and Haricot Beans. It is evident, therefore, that the food plant in itself does not determine the appearance of the sexual forms. It is very interesting that sexual forms appear on the primary or secondary host, showing that the distribution of Davipson— Biological Studies of Aphis rumicis L. 319 winged migrants over a great number of intermediate. hosts is not an essential part of the life-history, but must be regarded as an adaptation to a search for more suitable food conditions. One cannot consider, therefore, that Aphis rumicis is a true migratory species on which the influence of the intermediate host is an essential part of the life cycle. If suitable conditions are available, the life cycle can be passed through on either the primary or the secondary host. These conclusions, however, are largely based on the behaviour of Aphids in captivity. It is obvious, from the nature of the primary host and the temporary life of the secondary hosts, that the migration from the one to the other at suitable seasons is decidedly of benefit to the Aphids under natural conditions. It does not preclude, however, two very important possibilities in the life cycle of Aphis rumicis in nature, namely, (a) the occurrence of agamic females, which may continue under mild winter conditions, and carry on agamic reproduction in the following season ; (6) the occurrence of sexual forms and winter eggs on plants other than EHuonymus europaeus. These two possibilities may explain how, although one known winter host (Huonymus europaeus) of this species has a very localized distribution, the distribution of Aphis rumicis is a very wide one. Unfavourable seasonal conditions over the period when sexuparae are being produced would result in a heavy mortality of the agamic forms, a reduction in the number of fertilized eggs laid, and a smaller outbreak of Aphids the following season. It has been shown that on the approach of autumn, the influence of some important factors results in the development of sexual forms irre- spective of the number of the generation, There would appear to be at work some inherent tendency associated with adaptation to seasonal conditions. As autumn advances, an increased number of sexual forms appear. 3 6 appeared on Euonymus E, on 10. 8. 20, and ¢ ¢ and sexual 2 ¢ were present on Euonymus K, early in October. The sexual forms appeared also on Beans B; early in October. From September onwards an increasing number of sexual forms developed. In Beans B,, one agamic female continued parthenogenetic viviparous reproduction, giving rise to g ¢, winged sexuparae and twoa.v. 22. In my experiments the g g and oviparous 2 2 were produced by different mothers, the winged sexuparae producing the oviparous females, and the a. v. $ (apterous sexuparae) producing the ¢ ¢. The closing of the life cycle by the development of sexual forms SCIENT. PROC. R.D,S., VOL. XVI, NO. XXY, 2R 320 Scientific Proceedings, Royal Dublin Society. and the continuation of the agamic series may be shown in the following way — W.v. 00 [oYfe) av.oo +t a a. Vv. + avo tT Sexuparae $$ Ok ( Pp ) e019 Sosse W.V..0 Oe = W.V.00 a omage as ++ (Sexuparae) (Sexuparae) Klondnitzki found in Aphis hederae Kalt., that the sexuparae are apterous and the same mother produced both ¢ ¢ and ? ¢. Baker and Turner (1916) found in Aphis pomi De Geer, that the same mother (sexuparae) could produce both sexes; and in one case a mother produced both sexes and agamic forms. One can also consider in the case of Aphis rumicis that the same mother can give rise to both sexes, in that the winged sexuparae, which produce only sexual 2 2, are presumably predestined in the original agamic mother to produce sexual 2 9. Similarly, both sexual and agamie forms are produced by the same mother. = From a colony of agamic apterous females—Beans B,—all stages appeared in October, namely 9 3, sexual 2 9,a.v. 2 $,andw.v. 7 2. In conclusion, it is seen that in my experiments sexual forms developed under favourable conditions of food and temperature. They may arise in different generations. They were produced under the low temperatures of October in the glass-house, and later under the warm temperature in the green-house. Further, agamic and sexual forms appeared together under the same conditions of food and temperature. The autumn seasonal factor would appear to be important in accounting for the large numbers of sexual forms produced at this period. It is difficult to believe that the influence of external factors is the important one, when one considers that under the same external conditions totally different forms are produced—namely, ¢ 3, sexual 2 ?, and agamic females. Over long periods of time, external factors associated with seasonal conditions would be important, and it appears highly probable that the parthenogenetic and sexual phases are adaptations to seasonal conditions. While the generative changes appear to be determined by an internal Davipson— Biological Studies of Aphis rumicis L. 321 mechanism associated with the chromosome complex, it seems feasible to expect that changes in the environment may, by influencing the metabolism of certain individuals, indirectly affect the natural sequence of chromosome segregation, and so affect sex-production. This would allow of two conflicting views—internal mechanism and external factors—being brought into line, and would explain the result shown in figure 3. External factors, therefore, may be regarded as exerting an influence by retarding, restricting, or favouring the appearance of sexual forms, and not as determining them. I1I.—Summary. The early aphidologists considered that food and temperature conditions were the important factors influencing the apterous and winged forms in Aphids. This view appears to be upheld by the later experiments of Kwing and Shinji. From cytological investigations on Aphids and recent breeding experiments, it appears highly probable that the sequence of winged and apterous forms is largely due to some internal, inherent tendency. W.v.? 9 tend to produce a.v.? ?, and a.v. 2? 2 to produce either a.v. 9 ? ora mixed progeny, including a very variable percentage of winged forms. The apterous condition is to be regarded as an adaptation to seasonal food and temperature conditions. The great variability in the numbers of winged forms produced by apterous individuals is an important point to consider. Similarly, food and temperature conditions were considered the important factors affecting the development of sexual forms. Later cytological investigations show that the appearance of the sexual forms is associated with changes in the chromosome complex. The agamic generations appear to be interpolated between the winter egg and the sexual generation as an adaptation to seasonal conditions. The approach of winter conditions would normally be the factor affecting the bringing to an end of the partheno- genetic phase and the appearance of the sexual forms. In some cases the production of sexual forms may be superseded by continued parthenogenetic reproduction, certain agamic forms either reproducing slowly throughout a mild winter, or lying dormant and con- tinuing reproduction the following season. In isolated cases, certain agamic individuals may be affected physiologically by some factor or factors acting on their metabolism, so that they do not respond to the inherent stimulus to develop into sexual forms. These individuals normally, in winter, would die, but under favourable conditions may continue agamic reproduction. 2k 2 322 Scientific Proceedings, Royal Dublin Society. From the observations of other authors, however, it is evident that generalizations on the biology of the Aphididae, as a whole, cannot be drawn from the study of one species. Many special characteristics are exhibited by some species, owing to the complexity of the life cycle. Further investigations on the cytology of Aphids made in conjunction with breeding experiments are necessary for a clearer understanding of the biology of the Aphididae. TV.—REFERENCES. Baker, A.C., AND Turner, W. F. (1916): Journ. Agric. Res., vii, 323-443. Tbid., v, 955-993. Davipson, J. (1914): Ann. Appl. Biol. i, 118-141. Davipson, J. (1921): Bull. Ent. Res., xii, 81-89. Davinson, J. (19218): Ann. Appl. Biol. viii, 51-65. Davis, J. J. (1915): U.S. Dept. Agric. Bur. Ent. Bull. 276, pp. 1-66. Davis, J. J. (1914): U.S. Dept. Agric. Bur. Ent. Techn., Ser. 25, Pt. 11, 17-40. Ewine, H. E. (1916): Biol. Bull., xxxi, 53-112. Kiopnirzkt, J. (1912): Zool. Jahr. Abt. Syst. Geog. u. Biol. 33, 445-520. Morpwitko, A. (1907-9): Biol. Centralb., xxvii, xxvii, xxix; pp. 529, 561, 631, 747, 769, 631, 649, 82, 97, 147, 164. Moraay, T. (1909): Journ. Exp. Zool., v, 259-352. Marnesoy, R. (1919): Memoir 24. Cornell. Univ. Exp. Sta., Ithaca., N.Y., pp. 683-762. Suny, G. O. (1918): Biol. Bull. xxxy, 95--116. Wesster, F. M., and Puinuips, W. J.: U.S. Dept. Agric. Bur. Ent. Bull. 110, 1-153. f 323 | XXVI. e THE OCCURRENCE OF DEWALQUEA IN THE COAL-BORE AT WASHING BAY. : By T. JOHNSON, D.Se., F.LS., Professor of Botany, Royal College of Science for Ireland ; AND JANE G. GILMORK, B.Sc. [Pirates XI, XII.] Read May 24. Published Aveusr 29, 1921. In 1918 a boring to test the position of the concealed coalfield was begun by the Ministry of Munitions at Washing Bay, at the south-west corner of Lough Neagh. Owing to the unexpectedly enormous thickness of the beds of clay and basalt (1,500 feet instead of 250 feet estimated), the operations have been abandoned. The core has been placed in charge of the Geological Survey. Early in 1919 Mr. W. B. Wright, F.c.s., brought one of us pieces of the core containing plant remains for identification. For the past two years all available time has been devoted to the investigation of these remains, and many interesting results have been obtained. We are indebted to Professor G. A. J. Cole, F.R.s., Director of the Geological Survey, for permission to examine the core, and to Mr. Wright for the following observations on the geological features of the district :— “ The boring put down by the Ministry of Munitions in the years 1918 and 1919 in search-of the concealed coalfield supposed to exist beneath the Lough Neagh basin has yielded the first reliable evidence regarding the geological age of the Lough Neagh beds. The site selected was at Washing Bay, about four miles east of Coalisland, Co. Tyrone. The bore entered the Lough Neagh beds, beneath the glacial drift, at a depth of 48 feet, and reached the base of these beds at 1,196 feet 2 inches. The basal beds of the Lough Neagh series rested on about 70 feet of lithomarge produced by the decomposition of the Upper Basaltic lavas of Co. Antrim, which are well represented below. The series is, therefore, distintly post-basaltic in age, and is to be 1Miss Gilmore’s help was rendered possible by a grant from the Department of Industrial and Scientific Research. 324 Scientific Proceedings, Royal Dublin Society. referred to a later date than the plant-bearing beds of inter-basaltic age, which occur at Ballypallady. The Basaltic series of Co. Antrim rests as a whole on the chalk, which is of Upper Cretaceous age, and contains the zone of Belemnitella mucronata. “The facts established by the boring, combined. with certain re- lations observable at the surface and hitherto imperfectly under- stood, indicate with a fair amount of certainty that the Tertiary fold- ing and faulting of the district affect equally the Lough Neagh beds and the lavas on which they rest. The earth movements which resulted in this folding and faulting are, for reasons which cannot. be entered into here, generally considered to be of Miocene age. The Lough Neagh series is, therefore, presumably pre- Miocene. It is certainly not later than early Miocene. “The stratigraphy of the Lough Neagh beds is exhibited in the ac- companying diagram. The main horizons of identifiable plant re- mains occurred in the brown shales or clays between the depths of 881 feet and 1,008 feet 7 inches, lying mainly in the upper portions of these shales. ‘The plant remains in the white clays above and below these beds were for the most part uniden- tifiable, being highly carbonized. “From the geological relations Sand alone it is impossible to draw any more exact conclusion regarding the -SAND & GRAVEL ..LAMINATED CLAY BOULDER CLAY Grey and While Clays and Sands with Lignile ARTESIAN WATER \W°7-/0 GALS A MINUTE SSS Fo SS/L/ SBE, Rae ne MAW LIGNITE ZONE with Lignite Brown Shales While Clays& Sands ZEST) Clay Shale Lignile Scale 2e0 Fee? toan Inch Jounson & GitMorE—Dewalquea in, Coal-bore at Washing Bay. 325 age of these beds than is embraced in the general statement that they come in somewhere between late Cretaceous and early Miocene times. As, however, the lake basin in which they were laid down was probably produced by some early forerunner of the Miocene earth-movement, there is a presumption in favour of the later portions of this longsepoch. “ Tt is obvious that any palaeontolovical evidence that is obtainable becomes of very great importance in determining with more precision the horizon of the clays. Such a determination, moreover, would also assign an upper limit to the age of the great series of basaltic eruptions in north-eastern Ireland.” One of the most interesting fossils found in the coal-bore is the genus Dewalquea. Our first recognized specimen occurs in the core at a depth of 903 feet, and reveals itself as a five-lobed leaf (Pl. XI, fig. 1) in all essential features, in agree- ment with the Dewalqueas of the Kuropean Continent and America. The whole leafis 9cm. long by 9 em. broad; the petiole being only partially preserved, is not included in the measurement. The petiole divides at its distal end into three branches, the two outer ones branching again. Each of the five branches carries a lanceolate, acuminate, toothed, coriaceous leaflet. The leaf is clearly compound, quinately palmate or pedate, not a simple palmatisect or pedatisect leaf. The inedium leaflet is 8 x 1:5 cm., the inner lateral ones 7 x 1°5 em., while the outer ones are much smaller, being 0:8 cm. wide, and probably only 4cm.long. “Neither leaflet of the outer pair is completely preserved in the specimen. ‘The median petiolule is 3 mm. long, the lateral ones are 7 mm. long, and their branches very short. The base of the median leaflet is symmetrical and attenuated; the bases of the others are asymmetrical. The base of each inner lateral leaflet has the lamina prolonged downwards on its inner side beyond the outer lamina, the converse being the case with the outer pair—clearly an adaptation of shape to space requirements. The edge of each leaflet is entire in its basal part, but serrate in the upper two-thirds of its length. We have been able to restore the epidermis on both sides of the leaf. The upper and under epidermides (Pl. XII, figs. 5-8) consist of cells with more or less pronounced sinuous lateral walls. ‘The lower epidermis shows numerous sinuous cuticular striae on the outer face of the cells. Similar striae are described by Nestler (1) in Helleborus, in which they are confined to the upper epidermis. The upper epidermis of our fossil shows minute cuticular tubercles like those on the under epidermis in Hellebore. Stomata are confined to the lower epidermis. ‘they are 35-404 in diameter, sub-circular in outline, without subsidiary cells. Peltate scales, circular 326 Scientific Proceedings, Royal Dublin Society. in- outline, 60-76 wide, and consisting of a shield of 6-8 cells on a short stalk, are frequent, and are observable even with the lens on the counterpart impression through becoming detached from the leaf-surface (Pl. XII, figs. 1-2). These peltate scales and the cuticular striae have not been found previously in Dewalquea, and must be taken into account in considering its systematic position. Fic. 1.—Dewalquea hibernica 3 Restoration. Fic. 2.—D. fraxinifolia + Restoration. Vic. 3.—D. denticulata + Restoration. We propose to call the specimen just described D. hibernica. There are Jounson & GitmorE—Dewalquea in Coal-bore at Washing Bay. 327 many other examples of the genus in the core between 897-913 feet, as well. as at 787 feet. One leaf shows three leaflets, and may be named D. fraainifolia (Pl. XI, fig. 4). Another apparently simple leaf we propose to call D. denti- culata (Pl. XI, fig. 11). Most of the other specimens—some twenty in all—are fragments of leaves only. They all show the Dewalquea type of tissue, and indicate that it was common in the locality. Affinities of Dewalquea. The discovery of Dewalquea was made by Debey in the Senonian (Upper Cretaceous) of Westphalia. He appears to have sent specimens and drawings to Saporta and Marion (2), then at work on the Belgian ‘Tertiary flora, and also to Schimper (8). Debey’s ms. name of Araliophyllum, accepted by Schimper, was not, however, adopted by Saporta and Marion, who replaced it by Dewalquea (after the Belgian geologist), three specimens being described— D. haldemiana 8. and M. and D. aquisgranensis S. and M. from Westphalia, and D. gelindenensis S. and M. from the basal Eocene of Gelinden. Since 1874 other species have been described, of which Berry (4) gives a useful critical revision. They are all, with one exception, from the Upper Cretaceous, and are from America. One of the species, D. Smithi, is founded by Berry on material from the Tuscaloosa (Upper Cretaceous) beds of Alabama. D. Smitha is very similar to, but distinct from, D. insignis Hosius and von Marck from Haldem (Westphalia) (4). It differs from all previously recorded American species in being quinate, not ternate. It is similar to D. coriacea and D. pentaphylla, found by Velenovsky (6) in the Cenomanian of Bohemia. Knowlton (17), in 1917, described a second quinate specimen D. pulchella, which he considered closely allied to, if not identical with, D. insignis H. and von M. Dewalquea is an extinct type of leaf derivable from or of the same type as Polytaenia quinguesecta of the 'Turonian. It occurs in the Upper Cretaceous from Alabama to Greenland, in Westphalia and Bohemia, the Lower Eocene of Gelinden, in Belgium, and in the Oligocene of Italy. The recorded distribution in time and space adds considerably to the interest of the discovery of a Dewalquea in the coal-bore at Washing Bay. If the deposit be correctly referred to the Upper Oligocene, the Irish specimen may represent the last trace of the genus in time. The five-lobed form (Pl. XI, fig. 1) is closely allied to D. Smithi Berry. The chief obvious difference is the smaller outer leaflets of the Irish specimen. Berry shows a craspedodrome venation in his restoration, but in his description says it may be camptodrome. Our five-lobed specimen is camptodrome, and is more comparable to the small specimen of D. insignis H. and von M. (op. cit.). It is unfortunate that Dewalquea is known by its foliage only (in many cases fragmentary), and 328 Scientific Proceedings, Royal Dublin Society. that identifications have been hitherto, of necessity, based on the macroscopic characters—on the shape, mode of segmentation, and type of venation; with the exception of an observation by Saporta and Marion, nothing hitherto has been revealed of the character of the epidermis. We propose to review the affinities suggested by previous observers in the light of our observations. Helleborus Affinities. Saporta and Marion base their rejection of Debey’s reference of Dewalquea to Araliophyllum, and their own reference of it to the Helleboreae, mainly on the peculiar mode of segmentation of the leaf indicated by the term pedalo-digitate. The Dewalquea likeness-is best indicated in Helleborus foetidus, as monographed by V. Schiffner (7). Its leaf is always distinctly pedately divided, and consists of 7-10 small, lanceolate, finely sharp-toothed leaflets. We have failed to note the agreement of the venation of Dewalquea with that of Hellebore, observed by Saporta and Marion, who were not so fortunate as ourselves in the state of preservation of their specimens, judging from the remark of Hosius and von Marck that the Westphalian material showed poorly preserved secondary veins. We give actual photographic illustrations of the magnified venation of D. fraxinifolia (Pl. XII, fig. 3). The secondary veins in Hellebore arise at a very acute angle from the midrib, send off branches right and left, and end like their outer branches in the teeth of the leaflet. Each secondary vein is craspedodrome. In Dewalgquea the secondary veins come off at an angle of 60°-70°, and on nearing the margin bifurcate to join by their forks with adjoining secondaries, which then form a marginal network. They are thus camptodrome. A detailed comparison of the venation does not support the Hellebore affinities of Dewalquea. Saporta and Marion describe the epidermis of Dewalquea as formed of rounded polygonal cells, with probably sinuous walls like those in Hedleborus, and, we may add, in many other genera. Nestler gives a fully illustrated account of the structure of Helleborus. Stomata, 45 x 33n, occur scattered on the under epidermis oniy. The lateral walls of the lower epidermis cells are sinuous. Unicellular hairs occur, as in Ranunculaceae generally. The upper epidermis shows marked cuticular striation. Peltate hairs are unknown in Hellebore or any other Ranunculaceus plant. The geographical distribution is also of interest. The Lower Danube is the “Eldorado” of Helleborus, which is exclusively Old World, while Dewalquea is represented by many species in the New World, as well as by several in Europe. It would, however, be dangerous to support conclusions of affinities on present-day distribution, there being many cases in which fossil fruits prove the former wider occurrence of a genus or family. The Jounson & GitmMorrE—Dewalquea in Coal-bore at Washing Bay. 329 Ranunculaceae, consisting almost entirely of herbaceous plants, would not lend themselves to fossilization, and their past history is almost a blank in the rocks, though, as members of the Polycarpieae, they must have been one of the earliest families of the Dicotyledons to appear. Hellebore, with its coriaceous leaves in some species, would be the one most likely to be found in the fossil state. It would be distinctly satisfying if Dewalquea could be accepted as the ancestral form of the family. Its peltate scales, combined with other features, militate, however, against this view. The Araceae have been suggested as a possible line of affinity with Dewalquea. This is admittedly one of the most primitive families of the Monocotyledons, and in, e.g. Anthurium, there are several species with pedalo- digitate leaves. Their venation differs, especially in the possession of a marginal vein not found in Dewalquea. Their epidermis and stomata also differ. Peltate scales are said to occur, but they are simply saucer-like depressions, in which the secretion of the glandular epidermal cells collects. A third line of- affinity is that indicated by Debey’s name of Avaliophyllum. The Avraliaceae were rejected by Saporta and Marion on the expert advice of Decaisne that leaves showing such segmentation were not then (as they are now) known in the group. Schimper, as already mentioned, accepted the generic name, and described the fossils as Araliaceous. Peltate hairs occur in Hedera helia and in Oreopanax, but they are stellate, and do not suggest the dise-like scale-hairs of Dewalquea. P. Principi (8), in his recently published work on the fossil Dicotyledons of the Italian Oligocene, records the Belgian species, Dewalquea gelindenensis, and a new species, D. grandifolia. He follows Schimper and Zeiller (9) in referring the genus to the Araliaceae. D. grandifolia shows 5-7 leaflets, each on its own stalk, ic. a compound leaf with 5-7 pinnately arranged terminal leaflets. The Washing Bay flora is an early one from the Dicotyledonous point of view, and one would hardly expect to find highly specialized families such as the Oleaceae to be well represented in it. Thus in Mraainus excelsior the corolla has passed through the stage of gamopetaly to suppression. Mraaxinus first appeared in the Arctic regions, and gradually spread southwards in the Old and New Worlds. Heer (10) records F. praecox from the Upper Cretaceous beds of Patoot in Greenland ; and Lesquereux (11) describes #. eoceniew from the Eocene of Golden, Colorado; while Ettingshausen and Gardner (12) list F. jovis and F. prae-savinensis from Alum Bay. The seventeen European species recorded are nearly all referred to the Miocene. One of our speci- mens (Pl. XI, fig. 6), found at 909 feet, had every appearance of being the distal end of an Ash leaf, such as Fraxinus pennsylvatica var. lanceolata. 330 Scientific Proceedings, Royal Dublin Society. This identification was supported by comparison of the restored epidermis with that of Ash. The subsequent discovery of the quinately divided leaf of Dewalquea at 903 feet, with typical epidermal structure, seemed to indicate the necessity of a revision of the first view. In both Dewalquea and Fraxinus the venation, epidermis, stomata, and peltate scales (glands) all agree sufficiently to allow approximation ; and we are of opinion that Dewalquea is more naturally associated with the Oleaceae than with the Ranunculaceae. There is, however, another line of affinity which appeals more to us. Pl. XI, fig. 9, shows a well-preserved fossil leaf or leaflet, 2 x 4 cm., approximately, coriaceous, broadly oval-lanceolate, margin entire below, but sinuous-dentate in its upper two-thirds, midrib pronounced, thinning out apically ; secondary veins fairly numerous but delicate, sub-opposite ; angle of divergence, 70°-75° ; adjoining secondary veins united towards the leaf edge by loops or their own forkings, and forming a marginal network, from which veins pass into the teeth; the whole system camptodrome. There are no pronounced cross-anastomoses, but there is a fairly well-developed vascular network forming polygonal areas between the secondaries. Shortened secondary veins occur. Externally this leaf shows great resem- blance to Llea celastrina Sap. (13), from the Upper Oligocene of Saint Jean de Garquier and Armissan in the south-east of France. As the name suggests, Saporta saw affinity in his specimen to Celastrus also, and admits, in naming it, that it is fairly remote in its characters from the Ilex of to-day. If we were confined to a comparison of the external characters only, we should feel compelled to name our specimen Jlex celastrina. Restoration of the leaf tissue, however, alters our attitude. Markedly wavy walls are not found in Z/er epidermis, nor are peltate scales found in any of the Aquzfohaceae, according to Solereder (14). Of the many other possible connexions, the one combining most of the external and anatomical features is the genus Carya, e.g., the modern Carya laciniosa, in which, however, tufts of hair, not known in our fossil, occur. Our specimen shows, however, such general agreement with Dewalquea in its external features and minute structure that we feel compelled to refer it to this genus and not to Carya, and toname it Dewalquea denticulata. We are of opinion that the leaf fragment named by Heer, Pterocarya denticulata ? (15) (Juglans denticulata O. Weber) from Bovey Tracey (op. cit., Pl. LXX, fig. 5) is near to, if not identical with, D. denticulata; and that the Carya bilinica Ettings (16) is also probably nearly allied to it, though differing in form. A re-examination of the collections of fossils in our museums, after restoration by maceration, would reveal many hidden affinities. Jounson & GimMorr— Dewalquea in Coal-bore at Washing Buy. 331 Dewalquea shows certain Juglandaceous characteristics not to be overlooked in seeking to ascertain its systematic position. Principi’s D. grandifolia shows a pinnate leaf of 5-7 leaflets derivable from the types illustrated in the earlier works quoted. The transition from the pedalo- digitate to the impari-pinnate leaf is not difficult to follow. In the seedling of Pterocarya fraxinifolia Spach., the two cotyledons are deeply four or five- lobed, and may be described as pedalo-digitate. Such a form of seed-leaf makes it unnecessary to look to the Hedlebore for an explanation of Dewalquca’s leaf. Examination of the adult pinnate leaf of Pterocarya (and of other genera of the Juglandaceae) shows the presence of peltate scales or glands which bear a striking resemblance to those of Dewalquea. In Engelhardtia the likeness is such that (Pl. XII, fig. 12) one might easily be mistaken for the other. The epidermal cells show similar sinuous lateral walls, cuticular surface striae, and similar stomata. There is the same difference in the edge of the leaf as in Dewalquea. The edge may be entire or serrate in both genera— in different species in Juglans—but, according to Berry, possibly in the same species (or even plant) in Dewalquea. This difference of the leaf-margin is accompanied by a slight difference of venation, which is, generally speaking, camptodrome in Juglans. In this genus the cross-anastomoses connecting the secondary veins run straight and parallel to one another in most cases ; but in Hngelhardtia they are jointed and branched and irregular, as in Dewalquea. The secondary veins bifurcate and join together, giving a marginal network much as in Dewalquea. The Juglandaceae are a primitive group allied to the Amentaceae and Myricaceae. Though Juglans is now confined, in Europe, to the south-east corner, it was widely distributed throughout the continent in Tertiary times, having travelled southwards from the Arctic regions. Until fruits are found the question must remain open ; but one may provisionally regard Dewalquea as an ancestral Walnut. [ BIBLIOGRAPHY. 332 ~ Scientific Proceedings, Royal Dublin Society. BIBLIOGRAPHY. . NestLer, A.—Der anatomische Bau der Laubblatter der Helleboreen. Nova Acta Acad. Caes. Leop.-Carol. Germ. Nat. Curios, T. 1xi, pp. 1-44, Pl. I-III. Halle, 1894. . SAPORTA AND Marion.—Acad. Roy. des Se. des lettres et des beaux-arts de Belgique, mém. cour. et mem. des savants étrangers, T. xxxvii, fo, HZ 4, sob, jo, QZ . Scummper, W. Po.—Traité de Paléontologie végétale, T. iii, p. 38. Paris, 1874. . Berry, E. W.—The Upper Cret. and Eocene Floras of S. Carolina and Georgia. U.S.A, Geolog. Survey, Prof. Paper 84, p. 41. . Hostus aND VON DER Marcx.—Palaeontographica, vol. xxvi, p. 172, Pl. XXXII, figs. 111-113; Pl. XXXIIL fig. 109. . VELENOVSKY, V.—Die Flora der bohmischen Kvreideformation, Pt. III, 1884, pp. 11-14, Pl. I, figs. 1-9; PI. II, fig. 2; Pl. VIII, figs. 11-12. . SCHIFFNER, V.—Monographia Hellebororum. Nova Acta Acad. Caes. Leop. Carol., Deutsche Acad. der Naturforscher, T. Ix, pp. 1-198, Taf. I-VIII. . Princrer, P.—Memoire per servire Alla descrizione della carta, Geologica d'Italia, ‘I’. vi,p. 153; Tav. LIX, LX, figs. 4, d, 6. . ZeittER.—Eléments de Paléobotanique, p. 327, fig. 210. . Heer, O.—Flora fossilis arctica, vol. vii, p. 33, Taf. LXTV, fic. 2. . Lesquereux, L.—U.8.A. Geol. Survey of the ‘Territories, vol. vil, p. 229. . ErrINGSHAUSEN AND GARDNER.—Proc. Roy. Soc., London, vol. xxx, 1880, p. 238. 3. SaporTa—Etudes s. 1. vég. d. 1. France; Ann. d.8. Nat., sér. v, Pl. II, jo, WAYS Ils WOON, igG5 Tc . SOLEREDER, H.—Systematic Anatomy of the Dicotyledons. Transl. by Boodle, p. 210. 1908. . Herr, O.—Fossil Flora of Bovey Tracey. Phil. Trans. of Royal Soc., vol. clii, 1862, p. 1074, Pl. LXX, fig. 5. . ErrincsHAusen.—Die tossile Flora des ‘ertiar —Beckens von Bilin. Denkschriften der Kais. Akad. der Wissenschaften, Bd. xxix. Wien, 1869. 8.46, Taf. LI, figs. 4-6 and 13-15; laf. LIL, figs. 3-4 and 7-11. . Kyowtron, F. H.—A Fossil Flora of 8. W. Wyoming. U.S.A. Geolog. Survey, Prof. Paper, 108-F., p. 90, Jounson & GILMORE—Dewalquea in Coal-bore at Washing Bay. 333 EXPLANATION OF PLATES. Prare XI. 3. D. hibernica, sp. n. More or less complete leaf, showing five leaflets. D. fraxinifolia, sp. n., showing the three leaflets of the leaf in varying degrees of preservation. The two impressions of the same leaflet. (x 2). D. denticulata, sp. n. . Natural size. Fig. 12 (x 3). Puate XII. D. hibernica, showing the peltate scales. (x 2). D. fraxinifolia, venation in detail. (x 74.) D. denticulata, venation. (x 3.) Upper epidermis of leaf of Dewalquea, sinuous lateral walls. Under epidemis of same, showing stomata and bases of peltate scales. (7-8 (x 80), 9-10 (x 360.)) Peltate scale of Dewalquea. (x 300.) ~ Peltate scale of Hngelhardtia spicata. (x 300.) Lee XXVII. A SIMPLE FORM OF APPARATUS FOR OBSERVING THE RATE OF REACTION BETWEEN GASES AND LIQUIDS, AND ITS USE IN DETERMINING THE RATE OF SOLUTION OF OXYGEN BY WATER UNDER DIFFERENT CONDITIONS OF MIXING. By H. G. BECKER, A.R.C.Sc.1., A.I.C., Demonstrator in Chemistry, Royal College of Science, Dublin. Read May 24. Published Avausr 29, 1921. I.—INTRODUCTION. THE apparatus described in this paper was primarily devised as an alternative method of carrying out the ‘‘ Five Days’ Dissolved Oxygen Absorption Test,” and is based on a principle previously utilized by Professor Adeney for the same purpose.! In the course of the work, however, it became apparent that the principle was capable of wider application, and the apparatus was finally designed with a view to its possible use for the study of gas-liquid systems in general. The effect of different rates of stirring on the rate of solution of oxygen in water has been determined by means of the apparatus, and the results of the experiments are given in this communication, as they have a direct bearing on some previously published work.’ I].—TueE PRINCIPLE ON WHICH THE APPARATUS IS BASED. The principle utilized in the apparatus depends on the changes in pressure which oceur when a liquid is placed in contact with a closed volume of a gas which it is (\) capable of absorbing, and the action of the apparatus will be understood by reference to fig. 1, In the system shown in the diagram there is a volume of gas V, connected through a’ manometer with a volume of gas V,, there being sufficient liquid in V, to ensure saturation of the gas with vapour, and the liquid under test e Mas Il 1 Fifth Report, Royal Commission on Sewage Disposal. Appendix vi, p. 438, * Adeney and Becker; Scientific Proc., R.D,S., 1919, xv, p- 609, Brecker—Apparatus for Observing Reaction between Gases, ete. 835 is in contact with the gas in VY. When the tap is closed, any absorption of gas by the liquid from V2 results in a decrease in pressure, which is indicated on the manometer,and from the manometer reading the volume of gas absorbed can be calculated. With such an arrangement itis possible to vary the sensi- tivity of the manometer considerably. ‘he smaller the volume V, becomes in comparison with V,, the greater the movement of the manometer (within limits) for a given absorption of gas by the liquid. Further, when the tap connecting V, and JV, is opened, the manometer returns to zero, and the difference in pressure previously existing in V2 is distributed between the combined volume, thus making it a much smaller fraction of the whole. With such an arrangement it becomes possible to reset the manometer to zero several times during the course of an experiment, without allowing the pressure in V, to fall to such a value as would seriously affect the solubility of the gas. The tap also affords an easy method of replenishing the supply of one of the constituent gases of a mixture contained in V,. Thus, if V, contains atmospheric air in contact with a liquid absorbing oxygen, and V, contains pure oxygen, then, after a given absorption, the opening of tap resets the manometer, and also replaces the oxygen absorbed to an extent which will depend on the ratio a The conditions at the start of an experiment are such that the pressures in V,; and V, are equal to each other and to the atmospheric pressure P mm. Now, suppose an absorption of @ c.c. takes place in V,, and that the resulting manometer reading is py mm., then (VY, - @) ce. at P are occupying (V2 - v,) at P,, where v, = the volume of liquid which has risen above the zero point in the manometer, and V, c.c. at P are occupying (Wy + 0) Chewable. Now, since Py =P, +P, Le KU) Vath | Va oh (V.+ Vi) Vi +% +P, which gives Q=% + 3 (Vs). If the two vessels are now connected by opening the tap, P, becomes equal to P,, and p = 0. There is then V, + (V2- Q) cc. at P occupying a volume V, + Vz at P;. a P(V, + V2 - Q) V,+ V2 =P): SCIENT. PROC. R.D.S., VOL. XVI, NO. XXVII. 2s Hence IP. 336 Scientifie Proceedings, Royal Dublin Society: Hence the pressure in the air spaces after resetting the manometer will be less than that existing at the start by a quantity dependent on the ratio of the i! volume of gas absorbed to the combined A ————— ir spaces. By making WZ volume of the He® @ Ss} y king SNe the latter large this quantity can be made very small at will. II].—DEscRIPTION OF APPARATUS. The apparatus which was devised on this principle is shown diagrammatically in fig. 2. It consists of two cylindrical glass ves- sels about 30 mm. in diameter, which aie connected to two branches of a three-way tap. The lower vessel is of such a size as to contain a suitable volume (say 100 c.c.) of ©o the liquid to be tested, and still leave room = for an air space of about 50c.c.; the upper Q vessel has also a capacity of about 50 ce. The three-way tap allows of the two air spaces being connected through the mano- meter, or directly, according to its position. WITT] Hence, when an absorption has taken place, the manometer can be reset to zero by suit- ably turning the tap. aa The whole apparatus, including the manometer, is enclosed in a cylindrical tcl pa water-jacket, through which water is cir- culated from a thermostat. The handle of the three-way tap projects through a hole blown in this water-jacket, the joint being s Vv kept water-tight by means of a rubber A diaphragm. The water-jacket i s R phrag ater geckel is closed at my, AD bs each end by flanges, which may be made of LLL Y (Le Le : . . . YY brass, tin, or ebonite, according to the pur- g Y pose for which the apparatus is to be used. Ea ) y ( S S N The stirrer passes through a long sleeve in the lower flange, and is provided with two conical surfaces, one of which can be OY x} ie on tb Becker—Apparatus for Observing Reaction between Gases, etc. 837 adjusted by means of a screw, so that the whole is water-tight. In addition, the glass part of the stirrer is expanded into a cone at the lower end, which is ground to fit a corresponding cone on the bearing, so that the metallic parts of the stirrer do not come into contact with the liquid at any point. The glass part of the stirrer is so designed that, when rotated at a high speed, it causes a stream of fine bubbles of the gas to circulate through the liquid, thus ensuring equilibiium between the liquid and the gas. The manometer is provided with a scale of millimeters etched on the glass. IV.—CALIBRATION OF THE APPARATUS. In order to check the formula derived for the calculation of the volume of gas absorbed, use was made of ferrous hydrate as an absorbent for oxygen. The following experiments were made :— The apparatus was filled with 110¢.c. of distilled water which had been previously saturated with air at 20°C. by bubbling air through it for some hours, and the water circulation from the thermostat through the water- jacket started. When everything was in equilibrium, about 25 c.c. of the water was withdrawn, and in place of it was added exactly 10 c.c. ofa solution of ferrous sulphate containing 8:2 grams per litre, followed by 10 cc. of a 10 per cent. solution of sodium carbonate, the whole being carefully washed into the apparatus. The barometer was then read, and the taps of the apparatus having been suitably adjusted, the stirrer was started, and the manometer watched until the reading became constant. The maximum reading was then used to calculate the volume of oxygen absorbed by the ferrous hydrate. Experiment A.—Temperature, 20°0° C.; barometer, 768°1 mm. Maximum reading of manometer, . 114:0 divisions. Second reading after resetting to zero, . 5:5 - 114 divisions at 768°1 mm. correspond to 1°525 cc. 55 , 5 & OWS Total, 1:598 Total volume absorbed, calculated for N.T.P. = 1:504 ce. Experiment B.—Temperature, 20:0° C.; barometer, 769.9 mm. Maximum reading of manometer, 118 divisions. 118 divisions at 769°9 mm. correspond to 1°610 ¢.c., which at N.T.P. reduces to 1:520 ae. The mean of the two experiments gives 1°512 c.c. absorbed. Since the partial pressure of the oxygen in the air in the vessel was 2s2 338 Scientifie Proceedings, Royal Dublin Society. 105 ,, 20 50 50’ in the vessel before the experiments was 76871 mm., and at the end was 7580 mm. Taking the composition of atmospheric air to be approximately: nitrogen, 79 per cent., oxygen, 21 per cent., and the coefficient of absorption of oxygen as 0:0310 at 20°C., we can calculate the amounts of oxygen in solution at the beginning and at the end of the experiments. reduced from a correction for this must be applied. The pressure The pressure of aqueous vapour at 20°0° C. is 17-5 mm. The volume of oxygen in solution at the start is— V, = 0310 x 110 x se RS ‘706 cc. 50 760 The volume in solution at the end is— Ne 9 740°5 a 0310 x 110 x = x 760° Hence the difference due to change in the saturation value is ‘108 c.c., and this, when added to the volume indicated by the manometer reading, = 598 c.c. gives a total of 1°62 cc. as the actual absorption. he 10 c.c, of ferrous sulphate used should have absorbed 1°64 ¢.c. oxygen according to its titre by potassium permanganate, hence the two values agree within 1} per cent. 1-512 d ee Oxys en absorbed in CC 3 8 3 is | a [ N Ly oO or NI hb {o) 50 60 70 60 90 100 Time in Minutes Fig. 3. The graph, fig. 3, shows the course of the absorption. During the first part of the absorption the water is completely de-aerated by the excess Bucker—Apparatus for Observing Reaction between Gases, etc. 339 of ferrous hydrate, and, consequently, the oxygen is absorbed at a constant rate, as shown by the fact that the graph at this portion is practically a straight line. In the second stage, when all the ferrous hydrate has been oxidized, the absorption falls off gradually, so that the graph becomes a logarithmic curve. During this stage the water becomes completely re-aerated. In the figure, the second graph shows a more gradual approach to saturation than the first; this is due to the fact that during this experi- ment the speed of the motor driving the stirrer varied suddenly owing to variation in the voltage of the supply. This observation indicated the advisability of determining the time necessary to attain equilibrium under certain conditions. As this was intimately connected with some work which had been done previously on the rate of solution of gases by water, it was considered worth while to examine it more fully. V.—EXPERIMENTS ON THE EFFECT OF SvIRRING ON THE RATE OF SOLUTION OF OXYGEN IN WATER. The method consisted in introducing into the apparatus known amounts of a solution of ferrous sulphate and precipitating the hydrate by means of a solution of caustic potash; the stirrer was then driven at certain definite speeds by means of a small alternating-current motor (which could be relied upon not to vary in speed to any large extent), and the rate at which the oxygen was absorbed noted by observing the manometer. 340 Scientific Proceedings, Royal Dublin Society. The complete apparatus used for these experiments is shown in the photograph (fig. 4). The motor at the right of the picture drives the small circulating pump (which is immersed in the water in the thermostat), and also a stirrer in the thermostat. The water is pumped along the horizontal glass tube to the left, and into the water-jacket of the apparatus already described, from which it returns to the thermostat by the rubber tube from the top. To the extreme right will be seen one of a series of horizontal pulleys which were connected to the small alternating current motor in order to give the desired variation in speed of stirring. ‘This pulley is connected to the stirrer of the apparatus by the long belt appearing in the foreground, thus imparting motion to the stirrer at a rate which could be determined by the ratio of the various pulleys. The water circulation was continued, and the stirrer run at a constant speed during the course of each experiment. = tr CE bsorbed In ca) yeen A o Volume of Ox 03 The lower speeds of revolution were counted directly, using a stop-watch to check the time, and for the higher speeds the rate of one of the slower moving pulleys was found and that of the stirrer by calculation. Five experiments were made with different rates of stirring, and one experiment was made to compare the rate at which quiescent water absorbs oxygen with the rates at which the thoroughly stirred water absorbs it. The results of these experiments are shown in fig. 5. Bucxer—Apparatus for Observing Reaction between Gases, ete. 341 The rate of stirring in these experiments varied from zero to about 1,000 revolutions per minute. The lower rates up to about 140 revolutions per minute did not appreciably distort the surface of the water, but above this speed the surface varied in shape and area with each variation in the speed of the stirrer. The shapes assumed by the water surface are shown in fig. 6; and it will be seen that they range from the nearly flat surface of still water to the point at which the air is drawn down into the water in small bubbles. At this point the surface area becomes uncertain, so that no conclusions can be drawn from experiments at higher speeds. In the other cases the surface area was taken to be approximately equal to that of a cone with the vertical axis equal in depth to that attained by the lowest point of the water surface. Expts A.&B. Expr | Expt 2 Expts 3,485 Fig. 6. The progress of the absorption followed a linear law, as will be seen from the graph, fig. 5, and this shows that the ferrous hydrate de-oxygenated the water completely almost at once, so that the rate of solution of the oxygen remained constant until all the hydrate was oxidized. The variation in the value of the rate of solution with stirring is shown in the table and in fig: 7. On the graph the lower line shows the rate of solution under the conditions of the experiment plotted against rate of stirring, and the points lie approximately on a straight line. Hence under the conditions of these experiments the rate of solution varies directly as the rate of stirring. 342 Scientific Proceedings, Royal Dublin Society. Values of Rate of Solution for different Rates of Stirring. 005 Rate of Solution in CC. per minute. . Rate of Experiment. ee vhreas of Solution nee | rp. aa Surface, inc.c. P e | Sq.cm. per min, S = viii | A and B. 1000 19-0 030 0016 1 590 11:1 “014 00138 2 340 8°8 “008 0009 3 140 821 004 “0005 4 80 8-1 003 "0004 5 60 8:1 002 "00025 | 6 0 8-1 *0001 00001 LE 200 _ 300 400 500 Rate of Stirring in Revolutions per minute. SCALE IL 100 600 Fig. 7. = When, however, the effect of increased surface area is taken into account, and the rate of solution per unit area plotted against rate of stirring, the result is the upper line in fig. 7. This curve shows the rapid increase in the rate of solution at first, which gradually falls off as the higher values are reached, and tends to reach a maximum. gs. Manometer Readin Brcxer—Apparatus for Observing Reaction between Gases, ete. 348 VI.—CoMPARISON OF RESULTS. The results obtained in Experiment 6 are shown as a graph in fig. 8 in order to afford a comparison between the rate at which the oxygen is absorbed by quiescent water and the rate when the water is stirred. The water was at first stirred for two minutes in order to mix the ferrous hydrate with the liquid, and thus reduce the oxygen content to zero. During this period the rate of absorption was of the same order as obtained in the other experiments. The stirring was then stopped, and the rate fell off slowly as the liquid came to rest until it reached a value which remained nearly constant for three days. On the third day this suddenly changed for a higher value, which remained constant until the sixth day, by which time only about half the saturation value had been reached. The liquid was then stirred again, and the remainder of the absorption was completed in about five hours. Hence a process of absorption which may take many days to complete when the water is quiescent takes place in as many hours with moderate stirring. Do Time in Days. Fig. 8. The sudden change in the rate of solution of oxygen by the quiescent water indicated above is characteristic of the process under these conditions, for the rate of solution of gases by water is then liable to sudden and unaccountable variations within narrow limits. The greatest value for the rate of solution per unit area given in the table represents the maximum rate attainable under the conditions of these B44 Scientific Proceedings, Royal Dublin Society. experiments, without allowing the surface exposed to become indefinite : hence it is of interest to compare this with the rate of solution as determined in previous experiments. ! The rate of solution of oxygen into air-free water has been found by a totally different method to be ‘0161 ¢.c. per minute per square centimetre of area exposed. Hence for the oxygen in the atmosphere the rate should be ‘0161 x = = 0034 cc. per minute under similar conditions, since the rate of solution has been shown to be proportional to the partial pressure. This value is more than twice as great as the maximum value attained in these experiments (-0016), showing that the rate of mixing that was reached under these circumstances was not so rapid as that obtaining in the experiments referred to. This was to be expected, since in those experiments the water was exposed to the oxygen in very thin layers, and very perfectly mixed immediately afterwards. However, the fact that the two values come so near each other shows that with more perfect stirring arrangements it might be possible to approximate very closely to the maximum rate of solution. These experiments therefore afford a connecting link between the conditions obtaining in the work referred to and those obtaining in nature, They also emphasize the important part played by mixing of the water in determining the rate of solution. It will be seen on reference to the table that very gentle stirring of the under layers of the water increases the rate of solution as much as twenty-fold as compared with stationary water. This bears out the view put forward previously, that the rate of solution of air by water under ordinary conditions is largely determined by the rate of mixing. Experiments are being continued with this apparatus; and it is hoped to publish further results in the future. CHEMICAL DEPARTMENT, Roya COLLEGE OF SCIENCE, DUBLIN. 1 Adeney and Becker: Scientific Proc. R.D.S., xv, p. 625. 1919. 2 Adeney and Becker: Scientific Proc. R.D.S., xvi, p. 148. 1920. [ oe ji XXVIII. THE OCCURRENCE OF A SEQUOIA AT WASHING BAY. By T. JOHNSON, D.Sc., F.LS., Professor of Botany, Royal College of Science for Ireland ; AND JANE G. GILMORE, B.Sc. [Puates XIII, XIV.] Read Jun 28. Published Aucusr 29? 1921. In the course of examination of the Washing Bay core we have met with not only isolated seeds and pollen-grains of the genus Sequoia, the only Conifer found in it, but fortunately foliage shoots and cone-bearing twigs also, enabling us to identify the fossil. Sequoia (or Wellingtonia) has a very limited present-day peoatap ical range, being confined to the west Pacific coast of North California (and Oregon). One species, S. sempervirens (Sargent) (1), the Redwood, grows on sandstone to the height of 400 feet and a diameter of 28 feet, exposed to the warm Pacific winds and fogs, mainly in a mountain forest belt along the coast of California ; it ascends to 3,000 feet, and rarely occurs inland more than 20 miles. Sequoia gigantea or Sequoia Wellingtonia (1) (Sargent sp.), the Mammoth tree, the other living species, grows between 36°-39° N., on the western slopes of the Sierra Nevada, up to elevations of 8,000 feet, lives to 4,000 years of age, reaches a height of 320 feet or more, and a diameter of 35 feet. Though now recognized as one of the wonders of nature, the first-named of these giants of the forest was not discovered until 1847, a date it is well to keep in mind in criticizing the Conifer identifications of the earlier palaeo- botanists (e.g., of the Mull plants). Itis generally agreed that Sequoia, the only member of the family Sequoiineae, is a very ancient genus, and represents one of the earliest families of Conifers. It has been traced, at any rate, to the earliest Cretaceous beds, increasing in frequency and variety of form into the Miocene, in which it is represented by many species. It was one of the many circumpolar or Arctic genera which radiated southwards during the Cretaceous and early Tertiary epochs. Like many other genera now confined 346 Scientifie Proceedings, Royal Dublin Society. to North America or to Hast Asia, it flourished at one time in Europe, and fell before a lowered temperature and other adverse conditions. C. and E. M. Reid (2) describe and illustrate a cone from the Pliocene beds on the Dutch-German border as probably the last trace of Sequoia in Europe. E. W. Berry (3) gives a useful map (fig. 1) showing the almost world-wide distribution of Sequoia during the Tertiary epoch, in striking contrast to its present restricted area. Its occurrence in New Zealand, South-East Australia, and South America, the only localities recorded by him in the Southern Hemisphere, is worthy of note, and seems in need of critical confirmation. West and Central Europe, North-East Asia, North America, and the Arctic regions were its chief centres. Sxercu Map showing the location of Caenozoic records of Sequoia. After Berry, from ‘‘ The Plant World.”’ It will be well to state here the distinguishing characters of the foliage of the living species of Sequoia for comparison with fossil forms. Jn both species the leaves are mostly spirally arranged, and more or less clearly decurrent. In: S. sempervirens the base of the leaf is so twisted that the leaves appear distichously arranged. They are yew-like, linear, and flattened, with a single vein as midrib. ‘he apex may be acute or obtuse. In S. gigantea the spiral arrangement is undisturbed and obvious, the leaves narrow and subulate, roughly three- or four-sided (rhomboidal), with one vein. They are more or less adnate or adpressed in their lower part, spreading above. Their ridged bases are so decurrent and close together as to cover the whole surface of the stem, thus leaving no part of the stem surface proper exposed. Further, in both species there are small ovate-triangular, squamiform, keeled, closely adpressed, spirally arranged leaves. Thus both species are dimorphic. The squamiform leaves occur on the cone-bearing shoots in both species, and at intervals on the foliage shoots of S. gigantea. They occur in S. sempervirens also at the tips of the foliage shoots, making its buds “scaly ” in contrast to Jounson & GiItMoRE— Occurrence of a Sequoia at Washing Bay. 347 the “naked” buds of S. gigantea. They also occur at the bases of the foliage shoots of S. sempervirens, being in reality the persistent bud-scales through which the foliage shoot has grown and elongated. 8S. Gardner (4) notes the interesting fact that when introduced into a warmer climate, such as that of Madeira, S. sempervirens develops more of these leaves, an indication apparently of more vigorous growth in a warmer climate, with consequent formation of bigger buds. This may indicate that adpressed foliage is more adapted than the spreading foliage to a warmer climate. A reduction of transpiration and an exposure of one side only to sunlight, if bright, are obvious advantages of adpression. One would expect such foliage in ancestral Sequoias growing in the tropical or subtropical conditions of the Cretaceous or early Tertiary. It is the prevailing type in our Sequoia and in that of Bovey Tracey (5). We hope to prove that the Sequoia found in the two localities is identical, and the beds containing it contemporaneous, viz., Sequoia Couttsiae Heer in the Upper Oligocene. As already mentioned and as our illustrations (Pl. XIII, figs. 1-9) show, S. Couttsiae Heer possesses in the main an adpressed foliage. The leaves are semi-ovate or triangular, imbricate, decurrent, more or less falcate, with acute apex, and are spirally arranged. Each leaf is keeled or ribbed in correspon- dence with its solitary vascular bundle. One frequently finds at the tip of the leaf a sharp-pointed brittle cap of amber (Pl. XIII, fig. 10), which has evidently oozed out from the single resin-passage known to occur in the leaves of recent Sequoias. Under the dissecting microscope one can detach the amber tip and leave the leaf apex blunter in consequence. The leaves are readily restored. Stomata (Pl. XIII, figs. 11-12) occur on both surfaces, and on the lower side form two bands, one on either side of the midrib. With a view to testing the taxonomic value of the arrangement and character of the stomata, we made a series of preparations of the leaf of living Sequoia and other Conifers. . C. and E. M. Reid (6) were able by restoration of the epidermis of the Bovey Tracey material to show that Heer’s,species was rightly named; that it was a Sequoia,and not, as S. Gardner stated, an Athrotaris. It needs little knowledge of the Conifers to make one realize the helplessness of positive identification based on the external characters only of the foliage, owing to its pronounced heteromorphism. With differences of form supported by difference of microscopic characters there is more hope of success. A reliable way of distinguishing Sequoia gigantea from Cryptomeria japonica and S. sempervirens, and S. Langsdorfii from Taxodium distichwm microscopically would be distinctly helpful. Bertrand (7) could find no difference in the epidermis of Sequoia and Cryptomeria in the distribution or arrangement of 348 Scientific Proceedings, Royal Dublin Society. the stomata or in the structure of the epidermis. We find, however, a marked difference in the size of the stomata. Those of Sequoia are 56-64 long; those of Cryptomeria only 40-46u in our material. If this difference held good in general, it would be a useful means of discrimination. The following table gives the results of measurement of the length of the stomata :— u Sequoia gigantea, . ; 0 . 5 56-64 » Sempervirens, : : : ; 56-64 » Couttsiae, . ; : i j 52-66 » “du Noyeri,” : : : 48-60 (Cryptomerites du Noyeri Sewar an Taxodium distichum, : : 6 : 46-56 2 = mucronatum, . : : : 46-56 Cryptomeria japonica, 6 0 36-46 *Glyptostrobus heterophyllus (South Ong) : 40-46 oe s (Hong Kong), : 54-62 *Taiwania cryptomerioides, . : ; é 56-66 Athrotaxis laxifolia, : : 5 P 36-44 It will be seen that the fossils referred to Seguota agree with the living species of the genus, their smallest stomata being larger than the largest ones of Cryptomeria. The Glyptostrobus measurements seem to cast doubt on the taxonomic value of the size and arrangement of the stomata. Those in the specimen from South China, showing spreading yew-like foliage, lie parallel to the long axis of the leaf, and are of the size of those of Cryptomeria in contrast to those in the Hong Kong specimens, with adpressed foliage, which, though on the whole parallel to the long axis, are less regularly arranged, and of a size comparable to those of Sequoia. While in most of the species listed stomata occur on both upper and under surface of the leaf, in Athrotaxis Jaxifolia they are confined to the upper surface. The stomata on the under side of a leaf generally form two bands or ribbons parallel to the midrib. In the ribbon one may see in many cases distinct rows or files of stomata. When the leaf isdecurrent the bands can be traced ontothestem. These “stem” stomatal bands are particularly well marked in Giyptostrobus. In Sequoia the individual stomata lie with their long axis for the most part parallel to the midrib; in Cryptomeria they are more irregularly arranged. Florin, who has investigated the epidermal *We are indebted to Professor A. Henry for this material. Jounson & Ginmore— Occurrence of a Sequoia at Washing Buy. 349 structure of the leaf of a number of Conifers, citing Mahlert in support, notes that in &. gigantea all the stomata lie with their longitudinal axis parallel to the direction of the bundle, while in S. sempervirens numerous transitional positions also occur. In our investigation of Sequoia, before we had Florin’s results, we found that, while the stomata in S. sempervirens lie parallel to the midrib in the normal spreading foliage, they are less regularly arranged in the leaves of its fertile shoots, and in the general foliage of S. gigantea. Taxodium distichum shows the stomata in the two bands arranged length- wise in series, but also transversely in rows. Each stoma, on the whole, lies with its long axis horizontally placed, ie., at right angles to the direction of the vein. Zazxus baccata shows stomata in vertical rows parallel to one another, the individual stomata in each vertical row looking like links in a chain, with vertical longitudinal axis. The mode of arrangement of the stomata should prove a valuable diagnostic character in distinguishing between fossil Zaxodium and Sequoia, eg., T. distichwm miocenicum and S. Langsdor fii, in which we have so far failed in our attempts at restoration by maceration. While the adpressed squamiform leaf, 2-4mm. long, is the prevailing type in S. Couttsiae, narrow, acicular, or subulate, sometimes markedly divaricate leaves, 4 mm. long, also occur (Pl. XIII, figs. 6-9). We have many twigs bearing these leaves only, and were struck by their similarity to the specimens from the Interbasaltic beds of Co. Antrim, named by Baily S. du Noyeri (8). One specimen (Pl. XIV, fig. 1) from the Grainger collection in the Belfast Museum shows both types on the same shoot, and the resemblance to the fertile shoot of S. gigantea is marked. The type specimen of S. du Noyeri Baily (Pl. XIV, fig. 2) shows impressions only, in the ironstone. As our photograph of the type shows, Baily’s drawing (Pl. XIV, fig. 3) makes the leaf much too long. This figure may have led Gardner to conclude that S. dw Noyeri was really a Cryptomeria. We are of opinion that it is a Sequoia and, at the most, a form of S. Couttsiae. It is clear from Saporta’s illustrated account of S. Couttsiae var. poly- morpha, that a form of S. Couttsiae grew in the south-east of France not unlike the dw Noyert form of S. Cowttsiae in north of Ireland, the Baltic region, and Greenland. If Gardner’s drawings of cones of the Glenarm material are correct, then we must admit that Cryptomeria, now confined to China and Japan, grew in Co. Antrim in the Tertiary. It is not, however, recorded elsewhere in the fossil state. We have restored the epidermis of the leaf of a Conifer from Glenarm. It has all the characters of a Sequoia. In the same slab are winged seeds (Pl. XIV, fig. 13) indistinguishable from those of S. Cowttsiae, 350 Scientific Proceedings, Royal Dublin Society. and unlike the almost wingless seed of Cryptomeria, which was evidently not, as Gardner thought, the only Conifer at Glenarm. Pollen-grains are frequent throughout the part of the core examined. They are spherical, finely punctate, 25-3lu in diameter, and agree with the fresh pollen-grains of S. gigantea. More than once bodies like Pl. XIV, fig 4, occur, indicating a pollen mother cell in course of division. They may or may not belong to Sequoia. The branching shoot shown in Pl. XIV, fig. 5, is interesting, but, without breaking it up, it is difficult to decide as to the nature of the two swollen terminal buds. They may be scaly foliage buds, male flowers, or even young cones. No signs of pollen- erains were obtained by examination of the restored counterpart. Cones, however, occur, as well as isolated seeds, and we had the good fortune to find one cone split open—borne on a shoot carrying true S. Couttsiae leaves (Pl. XIV, figs. 6-7) —in such a way as to show its structure as clearly as if it were a median section of a fresh cone. Several seed-bearing scales are exposed. Seeds (Pl. XIV, figs. 7-10) are observable on the upper surface of each scale, overlapping one another, being inverted cr pendulous, 3°5 mm. long, 2°7 mm. broad at the base. Each seed shows a central somewhat curved “nucleus,” with two fairly broad lateral wings. The seed is pointed or apiculate at its lower or micropylar end, broader and emargi- nate at its upper basal attached end. The hilum as a disc-like scar is clearly observable. The seeds are attached at the distal end of the ovuli- ferous or bract scale. A casual inspection of Sargent’s figures of the seed scales in the two living Sequoia will suffice to show a marked difference in the arrangement of the seeds in the two. It is worthy of note that S. Couttsiae agrees with S. sempervirens, and not with S. gigantea, in the size and structure of its cone and arrangement of its seeds. The main (adpressed) type of foliage of S. Couttsiae is strictly localized in S. sempervirens, and only here and there observable in S. gigantea; otherwise it is buried in the past. It may be added that the cone of S. sempervirens ripens in one season, but, according to ‘Sargent, that of S. gigantea ripens in the second season, as in Pinus and other Abietineae. There is a certain amount of lignite in the core, and some of our prepara- tions show typical Conifer tracheides, with a single row of bordered pits (Pl. XIV, fig. 10) on their radial walls, much as in recent Sequoia wood. We hope to make a more thorough examination of the lignite later. As Sequoia and Taxodium are usually associated with lignite deposits, a second boring might reveal larger deposits of lignite, giving a fuel capable of utilization in the baking of the enormous thickness of clay which makes, we understand, a satisfactory, though coarse, kind of pottery. Jounson & GILMORE— Occurrence of a Sequoia at Washing Bay. 351 S. Couttsiac Heer seems to have suffered a good deal at the hands of investigators. Heer’s drawings, supplemented by the photographic illustra- tions of C. and E. M. Reid, leave one in no doubt of the general characters of the species. Heer gives a figure (op. cit., fig. 12) of S. Couttsiae showing linear leaves, longer and more divaricate than usual, but he shows in fig. 9 the transition to the normal type. He clearly realized the dimorphism of S. Couttsiae. We have before us material from Bovey (Pl. XIV, fig. 11) (Reid Collection, York Museum), Co. Antrim (Grainger Collection, Belfast Museum) (Pl. XIV, fig. 1), Disco (Dublin Museum) (PI. XIV, fig. 12), as well as Baily’s type specimen of S. du Noyert (I. G. Survey) (Pl. XIV, fig. 2). We are of opinion that &. Couwttsiae Heer was common to Greenland, Ireland, and Devonshire, and that S. dw Noyeri Baily and S. Whymperi Gardner should be treated as synonyms. BIBLIOGRAPHY. 1. Sarcent, C. S.—The Silva of North America, vols. x, xi. iw) . Rew, C. and EK. M.—The Pliocene Floras of the Dutch-Prussian Border. Medeo. Rijksopsporing Delfstoffen. No.6. The Hague, 1915. 3. Berry, EK. W.—The Geological History of Gymnosperms. The Plant World, vol. xix, p. 33, 1916. : 4. GARDNER, J. S.—British Eocene Flora, vol. ii, p. 35. 2 p Or . Heer.—The Fossil Flora of Bovey ‘Tracey. Phil. Trans. of Royal Soc. of London, vol. cli, 1853, p. 1051. Plates LIX, LX, LXI. 6. Rem, C. and EK. M.—The Lignite of Bovey Tracey. Phil. Trans. of royal Soc. of London, vol. cci, 1911, p. 170. 7. Bervranp, M. C. E.—Anatomie comparée des tiges et des feuilles chez les Gnétacées et les Coniféres. Annales des Sciences naturelles, 5th Série, tome xxii. Paris, 1874, p. 126. 8. Batty, W. H.—Notice of Plant-remains from Beds interstvatified with the Basalt in the County of Antrim. Quart. Journ. Geol. Soc., vol. xxv, 1869, pp. 557-861. 9. Batty, W. H.—Reports of the Tertiary (Miocene) Flora of the Basalt of the North of Ireland. 2nd Report— British Assoc. Report, 1880, p. 108, Plates Il and IIL; 3rd Report—British Assoc. Report, 1881, p. 152. SCIENT, PROC. R.D.S., VOL. XVI, NO. XXVIII. 27 352 Scientific Proceedings, Royal Dublin Society. EXPLANATION OF PLATES. Puave XIII. Figs. 1-9. Branches of Sequoia Couttsiae from Washing Bay. 1-5. Shoots with squamiform foliage. 1-3 are magnified. (x 3.) 6-9. Shoots showing acicular leaves, occasionally divaricating. (x 3.) 10. Resin tips of leaf. (x 80.) 11. Stomata from squamiform leaf of S. Couwttsiae. (x 240.) 12. Stomata from acicular leaf of S..Couttsiae. (x 240.) Puare XIV. 1. The two types of foliage on the same shoot of S. Couttsiae. (Grainger Collection, Belfast Museum.) 2. Type specimen of S, dw Noyert Baily. (Irish Geolog. Survey Collec- tions.) 3. Photograph of Baily’s restoration of specimen in fig. 2. 4, Pollen mother-cell dividing. Seguota ? 5. Branch of S. Couttstae with terminal buds. Male flowers? (x 3.) 6. Cone of S. Couttsiae. (4.) 7. Scale of cone, with overlapping seeds attached. (x 8.) 8. Shoot of Sequoia Couttsiac, showing foliage and exposed seeds. (x 3.) 9. Isolated seed of S.Couttsiae, showing micropole (m) and hilum (h). 10. Tracheides, showing bordered pits. 11. Foliage of S. Couttsiae from Bovey ‘Tracey. (York Museum.) (x 3.) - 12. Foliage of S. Couttsive from Disco Islands, Greenland. (Botanical Collections, National Museum, Dublin.) (x 3.) 13. Seed of S. Couttsiae from Glenarm. (« 3.) L Oe 7 XXIX. THE SOURCES OF INFECTION OF POTATO TUBERS WITH THE BLIGHT FUNGUS, PHYTOPHTHORA INFESTANS} By PAUL A. MURPILY, B.A., A.R.C.Sco.I., Assistant in the Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. [Read Junr 28. Published Aveust 29, 1921.] I.—PreEvious WorkK ON THE MANNER OF TUBER INFECTION. In his important paper on the control of potato blight, Jensen* distinguishes between two periods during which infection of the tubers occurs. The first is while they are still in sitw in the soil; and the decay set up by the parasite is generally obvious at digging time, or at the latest by about one week after that operation. ‘The second occurs during lifting, a copious development of Phytophthora rot setting in about one week after the potatoes have been dug. Jensen, if he was not the first to draw attention to the danger of digging the crop while any green foliage remains on diseased plants, appears to have been the first, at any rate, to emphasize the magnitude of the danger, to give a rational explanation of it, and to suggest a remedy. According to this author, “si l’on arrache les pommes de terre avant que les fanes soient complétement mortes et pendant que les feuilles portent encore des quantités de spores capable de germer, ces spores vont tomber en quantité sur les tubercules .. . et une semaine aprés ou un peu plus tard, si la température est basse, l’effet fatal s’en fera sentir.” And it is further stated: “si l’on retarde l’arrachage jusqu’’ ce que les spores soient mortes, e’est a dire environ deux semaines aprés le dessechement des fanes, tout danger est evité.” There are quoted in support of this view the results of two experiments ' Acknowledgment is made to the Department of Agriculture of the Dominion of Canada for permission to publish in advance the results of experiments carried out in that country. ; ? Jensen, J. L.: ‘‘Moyens de combattre et de détruire le Peronospora de la pomme de terre.” Mém. Soc. nat. d’Agriculture de France, t- cxxxi. 1887. SCIENT. PROC. R.D.S., VOL. XVI, NO. XXIX. 2u B04 Scientifie Proceedings, Royal Dublin Society. in which portions of plots were dug successively on September 7th, 12th, 15th, and October 11th. ‘The following percentages of the crop developed Phytophthora rot :—Ist digging, 21°5; 2nd digging, 33:5; 3rd digging, 13°0; and 4th digging, 0 per cent. As to the condition of the foliage during this period, it is stated that a certain amount of foliage was still green and covered with spores when digging was first begun; at the second and third diggings almost all the leaves were withered, but some spores were still present, while at the last date all the leaves and stalks were dried up. Rain which occurred during lifting was held responsible for the large percentage of rot oceurring in the second lot of potatoes which were dug on September 12th. While Jensen had reached the conclusion by experimental methods that soil to which spores had been added remained capable of causing infection for a longer period than the spores themselves continued to be viable while still on the leaves, and that in fact spores which had actually fallen to the ground might prove sources of infection to the tubers during the operation of digging, he over-estimated the importance of infection arising directly from diseased foliage, and under-estimated the part played by contaminated soil. His final conclusion was that the crop was likely to decay in storage if it was dug too soon, “c’est & dire pendant que les spores se trouvaient encore dans les feuilles.” To his belief in this theory is traceable the inconsistency noticeable in the practical directions which he gives for the control of tuber rot after digging. Thus, he found experimentally that only about 1 per cent. of the spores produced on the leaves survived a dry day; and that soil with which spores had been mingled induced rot in potato slices up to five days after- wards, indicating a length of spore life of five days in the soil. In spite of these results it was found necessary in practice to refrain from digging for as long as two weeks after the complete drying of the foliage. Nevertheless, he believed that where it was not possible to delay digging sufficiently long to allow an interval of even five or six days after the removal of the stalks, the danger could be considerably reduced by such precautions as refraining from digging during rain, beginning work at the windward side of the field, digging in the afternoon rather than in the morning, and especially by removing the stalks, even though digging was to follow immediately. An experimental basis for the latter conclusions, which suggest an ephemeral character for the spores of the fungus, and neglect the possibility of infection from contaminated soil, is not given. Jensen’s conclusions as to the danger of digging the crop while the foliage is blighted were tested experimentally in the United States by Jones and Mureny—The Sources of Infection of Potato Tubers. 399 Morse! The results are summarized by Jones, Giddings, and Lutman? as follows :—“ ... . where the tops are attacked by the late blight the harvest- ing of the tubers should be delayed until a week or more after the death of the tops.” In the results of two experiments quoted 55:3 per cent. and 18°4 per cent. of the potatoes rotted when they were dug early, while the percentages of rot in similar potatoes dug from three to four weeks later were 73 and 6:0 respectively. The condition of the foliage as regards blight was not stated. Jones and his co-workers did not go into the question of the source of tuber infection after the foliage was practically or completely dead, although they seemed to have assumed that such infection was possible. Jensen proved that the spores remained viable for five days in soil, but his work necessitates the assumption that tubers may contract the disease during more than double that period after the foliage is dead. If this actually takes place, it is necessary to assume an unsuspected amount of vitality in the spores of Phytophthora infestans or the existence of some hitherto undiscovered phase in the life-history of the fungus. JJ.—PxHYTOPHTHORA TUBER-ROT IN EASTERN CANADA. The present author was led to a consideration of the question of the sources of tuber infection by residence for some years in the Maritime Provinces of Canada, where the ravages of blight developing in the tubers after the crop is dug sometimes exceed the damage wrought by the disease on the growing crop. A considerable number of blighted tubers are some- times found when the crop is lifted, yet at other times such tubers may be almost absent. When, however, the apparently healthy portion of the crop is stored, in many cases an outbreak of Phytophthora rot sets in which may destroy a large proportion of the tubers. In some cases the result is that not the one-twentieth part (or thereabouts) of the normal crop required to supply “seed” for the following year survives. The sale of tubers may be almost entirely held up for a month or two in the autumn in bad years, because buyers are afraid to touch any potatoes, no matter how sound they look. After the latter part of October the danger is held to be over, as the disease will by then have developed far enough to be recognized. 1 Jones, L, R., and W. J. Morse: ‘‘ Potato Diseases and their Remedies.—Relation of date of digging to development of rot.” Vermont Agr. Experiment Station, Fifteenth Annual Report (1901-2), pp. 219-223, 1902 ; Sixteenth Annual Report (1902-3), pp.161- 162, 1903 ; Seventeenth Annual Report (1903-4), pp. 391-395, 1905. 2 Jones, L. R., N. J. Giddings, and B. F. Lutman: ‘Investigations of the Potato Fungus, Phytophthora infestans.” U.S. Dept. of Agriculture, Bureau of Plant Industry, Bull. 245. 1912. 2u4 356 Scientific Proceedings, Royal Dublin Society. The reason for the development of such a severe attack late in the season seems to be twofold. The climate in late summer and autumn is almost always such that blight is present, and yet im many cases a balance is maintained, so that both plant and parasite go on developing until up to, or nearly up to, digging time. There is frequently no frost in the coastal districts until late October or even early November; and the season is so short (June Ist to October 15th) that the stalks do not die from natural maturity. In the second place, some of the varieties grown there are standard American potatoes, which, as is well known, are on the whole very susceptible to blight,and are not suitable for cool coastal regions. Due to this combination of circumstances, the results about to be described have a significance for portions of Eastern Canada which is not, perhaps, equalled elsewhere; but, as will be shown, the same factors also operate in Ireland to some extent. The observation was made in 1915, the first season spent in Canada, that a conspicuously small amount of tuber disease followed a most severe blight attack, which completely killed all the foliage early in September. As it turned out, this was the most serious blight epidemic which occurred in the five-year period 1915-1919, yet the amount of tuber rot which followed was far lower than in any of the other years mentioned. The totalamount of Phytophthora tuber rot developed by the crop in that season, including that found in storage up to the following April, was only 1:2 per cent., while the corre- sponding average for the succeeding four years! was 26:3 per cent. In none of the latter years was the blight sufficiently severe as a rule to kill the foliage completely before the crop was dug. The attack, therefore, persisted longer, to which fact is generally attributed the greater amount of decay in the tubers, because of their being exposed to infection for a greater length of time. It is generally agreed that a long period of spore discharge (that is, a moderate and long-continued attack of blight) and severe tuber rot are associated ; and the prevalent view is that the latter is a direct consequence of the former. The assumption must then be made that more spores find their way to the soil when the blight never (or certainly not for a long time) develops sufficient intensity to kill the plants, than happens when the blight attack is much more severe but less protracted; or, alternatively, that under the former conditions the spores reach the tubers in greater numbers or under more favourable conditions for infection. These are hypotheses which 1The figures for 1919 do not record the amount of tuber rot developed after November, but they include the severe outbreak which generally occurs within the first month after harvesting, subsequently to which little or none develops, Murpeuy— he Sources of Infection of Potato Tubers. 357 should be susceptible of experimental proof, but this has not been forthcoming up to the present. In the case of protracted blight attacks late in the season it is important to distinguish between tubers in which the disease appears at, or soon after, the time of digging and those which develop disease some time after having been stored. Both outbreaks are often similarly explained (or, rather, are not distinguished), it being held that infection is in all instances contracted while the tubers are still im sitw in the soil. At other times the appearance of large amounts of disease some time after digging is explained by the assumption that the disease spreads from tubers originally infected in the field to their neighbours during storage in pits and other places. Evidence will be presented to show that the great bulk of the tubers which develop Phytophthora rot subsequent to digging were not already infected in the soil while still attached to the plant, and do not become infected in storage from contact with other tubers so attacked, but that, in fact, a distinction must be drawn between the ordinary course of infection (which may be called ‘“ subterranean infection ”’) and another method, which, for want of a better term, may be called “surface infection.” The distinction is, theoretically, not very far-reaching, but for practical purposes it is of much importance. IIl.—Tuser DISEASE IN CANADA FOLLOWING EARLY CESSATION OF SPRAYING. A further opportunity occurred in 1917 of testing the conclusion arrived at two years before, namely, that any circumstance which preserves the foliage until late in the season but then allows it to become partially blighted increases the amount of disease in the tubers. Owing to conditions over which sufficient control was not possible, a succession of sprayings was con- cluded nearly one month before the normal time on an experimental area of one acre of potatoes. The field was divided into twelve plots, of which three received no treatment, three were sprayed four times, three others three times, two plots twice, and one once. It was intended that certain of the plots should receive two further applications, but these were omitted. After spraying operations ceased the blight became serious, but the haulms in all the plots survived in part until October 3rd, when the potatoes were dug. The control plots were an exception, for they were most severely blighted and had been practically dead for some time. ‘he results are given in Table I. 358 Scientific Proceedings, Royal Dublin Society. TABLE I. Development of Phytophthora rot in sprayed and wnsprayed potatoes in Prince Edward Island, Canada. Percentage by weight of blighted tubers. ge by 3 8 No. of applications | of spray. } In storage to _ In storage to | In the field. Now One| Asal Total. | | | l | 4 | 1°6 | 13°0 Or4 | 15:0 3 1:9 jae 3¢ 2-4 nena 73 0 7:2 4:8 0-0 | 12-0 | 2 3°3 12:0 0:3 | 15°6 1 6:0 10°35 0-0 16°5 0 372 3°3 0:3 | 6:8 | 4 1:0 | 10°58 2°9 | 14°7 a Ppa} | 21-2 Oo) 23°3 0 9°6 | AGS | 0-0 14-2 3 3-7 Alea | 0-7 16-1 3 3-7 | 17-0 3:0 23-7 | 2 57 | 8:7 0-0 14-4 Average \ | 3:9 | 14:2 11 18°5 | sprayed, Average un- | (7) 4-2 0-1 11-0 | sprayed, $ | It is evident that the sprayed plots (which, incidentally, gave an increased yield) showed a considerable reduction in the amount of diseased tubers, the average for the untreated plots being 6°7 per cent.,and that for all the sprayed plots 3:2 per cent. The crop from each plot was placed separately in a storage cellar, and the potatoes were again sorted at the end of November. The result was then reversed, the potatoes from the sprayed plots showing on the average more than three times as much disease as those which had no treatment. This result was comparatively uniform throughout all the twelve plots, and, on the whole, the potatoes which were sprayed oftenest had the greatest amount of tuber rot, those sprayed less often had less, and the unsprayed plots had considerably the least. So great was the amount of disease which appeared after the potatoes were stored that it much more than neutralized the original advantage (so far as tuber disease is concerned) resulting from spraying. Murpay— The Sources of Infection of Potato Tubers. 359 A practically identical result was obtained from another acre plot in the same year (1917) which had similar treatment. The two last applications of spray had to be omitted on half the field (four plots), while the other four plots, which alternated with the first, were left untreated. There were somewhat more diseased tubers in the unsprayed plots when the crop was dug, but relatively much less after it was stored. The figures available give the amount of disease up to November 30th only, when the total for the unsprayed potatoes amounted to 12°7 per cent., while that for the sprayed potatoes reached 27-7 per cent. These results seemed again to show that more tuber disease might follow a less severe, if more protracted, attack of blight. An examination of the results does not, however, support the view that the amount of tuber rot is a function of the length of the period of spore production. The sprayed plots in which the disease lasted longest on the foliage showed the smallest amount of disease in the tubers at the time when the crop was dug. On the other hand, the potatoes from these plots developed a far greater percentage of the disease during the subsequent two months’ storage. It is possible that the disease which became evident some time after digging was contracted before that operation, but that the: rot had not advanced sufficiently to be visible. A direct answer to this objection cannot be given, because, unfortunately, the interval which elapsed between digging and the advent of the bulk of the tuber disease is not known. The supposition, however, is an improbable one. When a potato tuber becomes infected with blight, visible evidence of infection under field conditions appears in about a week. That being so, all the disease is visible on potatoes dug on any particular date except that contracted, approximately, during the previous seven days. When, as frequently happens, a serious outbreak of tuber disease follows at an interval after harvesting, it presupposes, if this hypothesis be tenable, a remarkable increase in the amount of infection in the week immediately previous to digging—an increase which is difficult to explain. There is no evidence to support the suggestion of Jones and Morse,! that early digging favours the development of the parasite in the tubers, with the result that rot follows, while in the case of potatoes dug later the fungus, though present, remains dormant. On the other hand, the great increasein the amount of tuber rot following after harvesting may, perhaps, be more reasonably connected with the changed 1 Jones, L. R., and W. J. Morse: ‘ The Relation of Date of Digging Potatoes to the Development of the Rot.” Proc. 25th Ann. Meeting of Soc. for Promotion of Agricultural Science, pp. 91-95. 1904. 360 Scientific Proceedings, Royal Dublin Society. conditions as regards exposure to infection brought about by the actual lifting of the crop. To attempt to decide between these alternatives, three series of experiments were carried out in the subsequent two years. LV.—FIELD EXPERIMENTS ON THE SOURCES OF TUBER INFECTION IN CANADA. In an experiment carried out in 1918 the theory that storage rot originates from the contact of tubers with diseased foliage, or with surface soil within five days after it becomes contaminated, was tested in two ways. The stalks, then somewhat blighted, were removed from certain plots and the potatoes were not dug until a period of five days in one case, and eleven days in another, had elapsed. Other plots in which the foliage had been allowed to remain were dug at the same time. The second test devised consisted of digging two series of plots while the leaves were considerably blighted, and immediately treating the potatoes from one series with dilute formaldehyde solution so as to prevent further infection. All the plots were laid out in duplicate. It is not necessary to go into the details of this experiment here, since the results were inconclusive for anumber of reasons. The formaldehyde treatment, from which much was hoped, was apparently too drastic; for serious injury to the tubers, involving complications in decay, resulted. Furthermore, neither interval which was allowed between the removal of the stalks and the digging of the crop (five and eleven days) prevented a large amount of blight rot in storage. If there was anything in soil infection, it seemed as though a longer period must be allowed to elapse to render the crop safe. On the whole, however, the work appeared to be on the right lines. There was least disease in the potatoes treated with formaldehyde, and most in the tubers from those plots which were dug early and while the foliage was still blighted. Profiting by this experience, two modified experiments were conducted in 1919. These are referred to asexperiments A and B. In experiment A the efficacy of leaving a period of six days between the removal of the stalks and the digging of the crop was tested. In experiment B a much longer period was allowed—namely, thirty-four days. One experiment was carried out in quadruplicate and the other in triplicate. ‘The average results are shown in Tables II and III. MurpHy—The Sources of Infection of Potato Tubers. TABLE II. 361 Experiment A (1919) on the digging of potatoes at the Experimental Station, Charlottetown, Prince Edward Island, Canada. Stalks allowed Potatoes left Fercenase weigh of | N stolen Plot to stand until— undug until— Rewane 1p aS : orataeelin (represented by (represented by is To P Nov ————)} 000000 an a00R0G00 ) In Nara | Mall Ace. 2 1 © ° g., 4 plots. field) gigs De Dateof <= 2 Geren ES = | oie, Be BR GE Ss Ss ris Say aes aa a 1 He ietergnetaes ate 0-0 35:4 | 35-4 69 Z | nacre | papers = 0-0 4:0 | 4:0 112 g | | Stored at once. | 9°0 | 40-4 | 40:4 76 : Tubers left on | 99 | 68-3 | 68-3 42 | | ground all day. a Dug after frost:| 26°7 | 17-1 | 43:8 72 | | stored at once. 8 Tubers left on | 23:1 | 21°6 | 44°7 66 s | | | ground all day. TABLE III. Experiment B (1919) on the digging of potatoes at the Experimental Station, Charlottetown, Prince Edward Island, Canada. Plot. Stalks allowed to stand until— (represented by oe Date of Digging, Bas [-.e) o no Potatoes left | Percentage weight of Net weight undug until— crop blighted. sound — (represented by Remarks. | ; potnree in opoacocoboDdose c00005)) II In Nae Total. | Avg enane | of 3 ., 3 plots. | field. | 1919, ib. Z. sari As ieee abe bila liglem ee Tubers buried in| 8:7 5:9 | 14°6 89 soil again. Saehcranatcntes — 4-7 3°3 8-0 118 ieee woeeullaeen neat aes Soil sprayed : 37 | 24 61 113 Bordeaux mixture —_— ood! || PAL CY 21-7 104 _— 0:0 | 35°6 35°6 91 Dug after frost. | 11:9 | 3:9 | 15°8 123 separately, but 1t is included in the total. 1 A very small amount of disease found in the tubers in the field was not recorded 362 Scientifie Proceedings, Royal Dublin Society. It is clear from the first table that removing the stalks six days before harvesting is of little avail in preventing tuber infection (compare plots 1 and 3). ‘This is so in spite of the fact that removing the stalks on September 24th reduced the length of the blight attack on the foliage by one-third. Removing the stalks ten days before digging reduced tuber infection to very small proportions. Here again, of course, the period during which blight was present on the foliage was still further diminished. It is probable that a considerable share of the success attained was due to this fact. The exact influence of each of the two factors cannot be apportioned, but it may be mentioned that a large amount of tuber infection resulted from allowing eleven days to elapse between stalk removal and digging in the experiment of 1918; and also a large amount of infection occurred in experiment B of 1919, where the interval was nine days. Ten days elapsed in the case under discussion, and while it is true that the weather conditions (which must probably be allowed to have an influence) differed in the different cases, the protection afforded to the crop is to be traced principally to the initial amount of blight present and not to the interval allowed before digging. Turning toexperiment B, clear proof is obtained of theorigin of Phytophthora rot in storage. The plants in plots 2, 4, and 5 retained their foliage until September 18th, when they were all moderately blighted. The last-mentioned plot was then dug at once, and the crop stored on the same day. The plants in the second and fourth plots had all their foliage removed, and one was dug nine days and the other thirty-four days later. It will be noted that the period of the blight attack was the same length in all. Notwithstanding this, 35°6 per cent. of the tubers from the early-dug plot had developed blight by November, 21°77 per cent. from the intermediate one, and 8:0 per cent. from the late-dug plot. It seems impossible to escape from the conclusion that during the thirty-four-day interval the parasite lost its power of infecting the tubers, but that, on the other hand, this power persisted for at least nine days after the last spore was shed from the leaves. As was invariably the rule in all the experiments, three in Canada and one in Ireland, the greatest amount of tuber rot developed when the crop was dug while the foliage was still suffering from blight. On the one hand this compares with a smaller amount of disease in plots, the stalks in which were removed at or about the same time as the former were dug, but the crops of which were not dug until at least two weeks later; and, on the other 1This and other points, such as the possibility of the soil of plots from which the stalks had been removed being contaminated by spores from neighbouring plots, and the possibility of tuber infection during storage, are discussed more. fully in a forthcoming Bulletin of the Dominion of Canada Department of Agriculture, entitled ‘‘ Investiga- tions on Potato Diseases.” ; MureHy—The Sources of Infection of Potato Tubers. 363 hand, with plots dug in the ordinary way late in the season. This is a very noteworthy point. That potatoes which were left in the ground exposed to blight infection from spores washed down into the soil for an added period varying from twenty-one to thirty-four days should decay less than potatoes not so exposed points to some incompleteness in the generally accepted theory of tuber infection. V.—FIELD EXPERIMENT ON THE SOURCES OF TUBER INFECTION IN IRELAND. In view of these results it was thought advisable to try a similar experiment in Ireland. While it is probably true that storage rot is seldom, if ever, as severe in this country as in portions of Eastern Canada, neverthe- less loss due to this cause is hardly ever absent, and in some cases it becomes quite serious. Portion of a field of Up-to-Date potatoes at the Albert Agricultural College, Glasnevin, Dublin, measuring twenty-five square perches, was selected for the work. This area was divided into thirty-three equal plots. Each of the eleven methods of treatment indicated in Table 1V was tried in triplicate on these plots, which were distributed as uniformly as possible over the experimental area so as to minimize inequalities in the blight and soil conditions. Different treatments were given before and after digging, as detailed in the table, the dates of digging being there indicated. Before being stored, the potatoes were most carefully sorted and all diseased and small tubers excluded (except in the case of plot 8). The remainder were stored in small pits in the open. Each pit contained about 2} ewt. of potatoes, representing the yield of each series of three plots. The pits were made in the usual way, very slightly sunk in the ground, the potatoes being covered lightly with straw, or, in certain cases, with potato stalks. Where old stalks were purposely used for covering they were replaced with straw after a few days. All the pits were finally covered with earth. They were opened and the potatoes again examined on February 15th and 16th, 1921. [TaBLE IV. 364 Scientific Proceedings, Royal Dublin Society. TABLE LV. Digging and pitting experiment at the Albert Agricultural College, 1920-21. Per cent. weight Stalks allowed Potatoes left found rotted Plot to stand until— undug until— when pits were ‘ represented b »presented b d. Se eee) | Selecmaucts: J digging (if any). Soft rot = (cause : = Date of 35 3a gS ies, undeter- Blight. Digging, 2.5 ia 6a or mined). Lo ——$—$———— None. 1:05 3°86 2 None. 0 0°68 g | | None. 0-16 | 9-60 4 None. 1:16 | 10-04 ) aS NAG eerste gcc None. 1:07 0°27 Gig | ae ee a ELA oe ee a ae None. 0 0:10 7 i ia Meh al ov oa0096999008 None. 0 0-63 Seer Covered in pit with blighted] 0°33 | 18°16 stalks. | oct crrrococoncnoct Covered in pit with blighted) 0°81 9°76 stalks. SSO Pitted with fifty marked] 1:71 0°85 | blighted tubers. DUD | ecccssccccce paSPE PEERS) Spread on plot from which | 0°66 0°33 blighted stalks had just been removed. The field in which the plots were situated had suffered severely from blight throughout the latter part of July and the month of August. By the beginning of September the plants in more than half of the field were practically dead, and those in the remainder were severely blighted, but still living vigorously. The experiment was carried out on this portion of the field. The blight made little or no headway during the comparatively dry and warm period which set in about September Ist. It had progressed no further at the time of the first digging on September 15th and 16th; and it was not until just after the second digging on September 27th and 28th that, as the result of some showery weather, it began to spread again. Thereafter the disease continued to develop, although somewhat slowly, so that there was still a certain amount of living foliage left when work was completed on October 18th. Mureuy—The Sources of Infection of Potato Tubers. 365 The potatoes reported as attacked by soft rot included those which were so far decayed that the original cause could not be determined. No effort at an analysis of this category was attempted because of its difficulty. Besides, the amount of this type of rot in all cases but one hardly exceeds one per cent., and the results are not materially altered by neglecting it altogether. In the case of plot 10, where it reaches its highest level, it is noteworthy that while fifty marked blighted tubers were placed in the pit only thirty-nine were recovered in a recognizable form. Some of these would have been included in the soft rot column had they not been marked. It is probable that the missing eleven are to be similarly accounted for. If this be true for the other plots also, all the rotted potatoes should be included in the blight column, in which case again no serious difference results. Three points are clear from this experiment. The disease which was found when the pits were opened had not spread from a few initially infected tubers. The pit (No. 10) in which blighted potatoes were placed showed hardly more blight than did No, 6 pit, which contained similarly treated potatoes without the addition of any blighted specimens. There was little evidence of any disease being communicated from tuber to tuber. In most cases the potatoes surrounding the marked blighted tubers were quite sound, although in some cases the former were covered with decaying slime. It was exceptional to find marked diseased, and unmarked potatoes which had contracted disease, in contact. As the covering of the pit was not absolutely water-tight and the interior was moist, the conditions were pre- sumably favourable for the development of blight. The most dangerous source of infection is, according to the experiment, direct contact of the tubers with blighted foliage. Two pits were covered for some days with the blighted stalks, which were afterwards removed and replaced with straw. No. 8, which was so covered, developed 13:16 per cent. of disease, as compared with 3°86 per cent. in similarly treated, but differently covered, potatoes in pit 1, an increase of 9:3 per cent. A similar difference exists between pits 6 and 9. The latter, which was covered with diseased stalks, showed 9°76 per cent. of blight. The former, not so covered, showed 0:10 per cent., the difference being 9°66 per cent. in this case. Although it is well known that the leaves are the source of the spores which infect the tubers, the use of stalks as a temporary covering for freshly dug potatoes is still too common a practice, and, as the experiment shows, it is a most dangerous one. Had there been a vigorous development of blight on the leaves, the result would probably have been even more serious. It is clear, in the third place, that direct and deliberate contact of the foliage and tubers is not necessary for the production of serious 366 Scientifie Proceedings, Royal Dublin Society. o storage rot. This is seen especially by comparing pits 3 and 6 and pits 4 and 7. The corresponding pairs of plots had been dug on the same dates, respectively. The difference consisted in that the foliage had been removed from one each of the two pairs twenty-five days and thirty-three days before digging, while it had not been removed from the other two. This operation brought about two changes in the conditions, either of which might, theoretically, be held accountable for the reduction in the disease. The duration of the blight attack on the foliage was reduced, on the one hand, and with it the supply of spores capable of being washed down to the tubers. On the other hand, the contamination of the soil ceased, and a period elapsed presumably sufficient for the spores present at or near the surface to lose their vitality. It is not possible to apportion with certainty the exact influence of each of these two factors. This is because the blight, instead of diminishing gradually with the progressive death of the foliage, rallied from the low ebb it had reached about the middle of September (when the first and second diggings were done), and was more severe at the time of the third and fourth diggings. Had it not been so, it was anticipated that there would have been a gradual diminution in the amount of tuber disease found in storage in the series of plots from 1 to 4, and a still more marked diminution in the series 5 to 7. Furthermore, the lifting of plots 4 and 7 had to be hastened owing to the exigencies of farm operations, although it was intended that they should be left until well after the death of the foliage was complete. A reduced amount of tuber disease was again anticipated from this course. While certainty was not attained, an examination of the evidence shows that the probabilities are in favour of the theory of infection at digging time. In support of this are the following points :— (1) All the plots went through a very severe blight attack, lasting from the latter part of July until the end of August, with but slight injury to the tubers (as measured by the amount of disease found in them when dug), at a time when they were immature, and theoretically more susceptible to infection. (2). The amount of blight found in the tubers in the field was slight and, comparatively speaking, uniform; and it did not vary dixectly with the date of digging or with the length of the period during which the foliage was allowed to stand. The variations were irregular, though not of great magni- tude. ‘The highest total of disease in the field was found in one of the plots dug first, and the next highest in one dug nearly last, Murruy—The Sources of Infection of Potato Tubers. 367 (3) The amount of blight appearing in the tubers in storage, generally speaking, stands in relation to the amount of blight present at and previous to the time of digging. (4) The general appearance of an increased amount of disease in the crop, which is not visible at the time of harvesting, but becomes so some time after- wards, and which must be due to infection soon before or when the crop is being lifted, can be related to no change in the conditions regularly occurring before digging, but must more probably be referred to the change brought about by that operation itself. It is not necessary to labour these points. The experiment points generally in the same direction as the series carried out in Canada, and proves that potatoes may become heavily infected with blight when they are ~ brought into contact or close relationship either with diseased stalks or contaminated surface soil. The former possibility will be generally admitted as a danger, as the result of the work of Jensen, Jones and Morse, and the present author. The view that the latter is a serious and hitherto largely unsuspected source of infection is strongly supported by evidence based on laboratory studies of the behaviour of the spores under varying conditions of environ- ment which have been made during the past winter, an account of which will be given in a separate communication. SUMMARY. Evidence is presented that more Phytophthora tuber disease may follow a less severe attack of foliage blight, occurring late in the season, than results from a severe outbreak which runs a rapid course. Conditions favourable to tuber infection may be brought about in Eastern Canada, or elsewhere if the circumstances are similar, if potatoes are sprayed in the early part of the season, but left untreated towards the later portions. Under such circumstances it is important to distinguish between the disease which appears in the tubers at or very soon after lifting and that which appears some time later. It is the later development of the disease which becomes serious following protracted or late outbreaks of blight. Evidence derived from field experiments in Canada and in Ireland is presented to show that the bulk of the infection in the case of potatoes which develop blight in storage is contracted when the tubers are being dug. 368 Scientific Proceedings, Royal Dublin Society. It is shown that direct contact of the tubers with partially blighted foliage results in serious rot in storage. The blight does not spread from tuber to tuber even in moist pits to any considerable extent, if at all. Soil contaminated by means of spores shed from the leaves continues capable of inducing blight in freshly dug tubers which are brought into contact with it over a period of at least ten days, and probably longer. The part played by contaminated soil is fully established by the results obtained in laboratory investigations, which will be published separately. XI. PLATE XVI. ScIENT. Proc. R. DuBLIN Soc., N.S., VoL. PLaTE XII. XVI. ScrEnT. Proc. R. DUBLIN Soc., N.S., VOL. Daw Seta? toy yy saa ers grey} CONOR se We A sti 9 ey Pate XIII. XVI. SciENT. Proc R. Dustin Soc, N.S., Vou a ip Lire ScrIENT. Proc. R. Dusrm Soc., N.S., Vor. XVI PLATE XIV. # hive. al Lea Th ets i 3. 10. 11. 12. 13. 14, 15. 16. SCIENTIFIC PROCEEDINGS. VOLUME XVI. . A Cryoscopic Method for the Hstimation of Sucrose. By Henry H. Dixon, sc.D., F.R.s., and 'T. G. Mason, m.a., so.B. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Louis B. Smyrn, B.A., so.B, (Plates I, II.) (February, 1920.) 1s. The Application of the Food-Unit Method to the Fattening of Cattle, By James Winson, m.a., B.sc. (Plates III, 1V.) (February, 1920.) 1s. The Holothurioidea of the Coasts of Iveland. By Anne L. Massy. (Apzil, 1920.) 1s. . Photosynthesis and the Electronic Theory. By Henry H. Dixon, so.., F.R.s., and Horacz H. Pootn, so.p. (March, 1920.) 1s. ; . Note on the Decay of Magnetism in Bar Magnets. By Wint1am Brown, B.so. (March, 1920.) 6, } . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By T. G. Mason, m.a., so.B. (April, 1920.) 1s. The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Wittiam Brown, 8.sc., and Parrick O’CaLLaGHan, A.R.¢.S0.1.,A.1.c. (August, 1920.) 6d. Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. Leonard, A.R.0.80.1., B.S0., PH.D., and Agnes Brown, 4.R.0.S0.1., B.sc. (August, 1920.) 64d. An Investigation into the Causes of the Self-Ignition of Mther-Air Mixtures. By the late Professor J. A. McCimtuanp, p.sc., F.R.s., and Rey, H. V. GILL, $.J., D.s.o., M.c., M.A., University College, Dublin. (August, 1920.) 6d. The Influence of Hlectrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosmpH Resuy, M.a., D.sc., F.n.c.sc.1., and Witrrep J. Hicxinsorrom, B.sc. (October, 1920.) 6d. | Notes on some Applications of the Method of Distillation in Steam. By JosEPH’ REILLY, M.A., D.SC., F.R.0.SC.1,, and WinrRED J. Hickinzorrom, B.Sc. (October, 1920.) 6d. The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By W. EH. Aprnny, D.sc., A.R.C.s0.1., F.1.c., and H. G. B&ckER, A.R.0.SC.1., A.r.c. (September, 1920.) 6d. A Determination, by means of a Differential Calorimeter, of the Heat produced during the Inversion of Sucrose. By Henry H. Dixon, sc.p., F.R.S,, and Niczt G. Batt, B.a. (December, 1920.) 6d. The Measurement of very Short Time Intervals by the Condenser-Charging Method. By Joun J. Dowrine, m.a., F.1vst.P., and Donat Donnexxy, M.Sc. (In conjunction with the late Prof. J. A. McCrexnanp, F.r.s.) (February, 1921.) 60d. j A Vibrating-Flame Rectifier for High-Tension Currents. By Joun J. Dowxine, M.A., F.INST.P., and J. IT. Harris, B.sc. (February, 1921.) 6d. We 18 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. SCIENTIFIC PROCEEDINGS—continued. . A Sensitive Valve Method for the Measurement of Capacity, with some Important Applications. By Joan J. Dow1ine, u.a., F.inst.p. (February, 1921.) 6d. . A’ Direct Reading Ultra-Micrometer. By Joun J. Dowtine, M.a., M.R.1A., F.insT.P. (March, 1921.) 6d. Studies in the Physiography and Glacial Geology of Southern Patagonia. By BK. G. Fenton. (Plates V, VI, and VII.) (March, 1921.) 4s. 6d. Award of the Boyle Medal to Grorcn H. Peruysriner, B.sc., pu.p., 1921. (March, 1921). 1s. The Concentration and Purification of Alcoholic Fermentation Liquors. Part I.—The Distillation in Steam of certain Alcohols. By Josnpa Rumty, M.A., D.SC,, M.RIA., F.C, and Witrrep J. HickiInBoTTOM, M.SC., A.1.C. (August, 1921.) The ‘ Browning” -and “‘ Stem-Break’” Disease of Cultivated Flax (Linum usitatissimum), caused by Polyspora lini n. gen. et sp. By H. A. Larrarry, a.R.c.sc.1., Assistant in the Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. [Communicated by Dr. George H. Pethybridge, p.sc.] (Plates VIII-X.) (August, 1921.) A New Principle in Blowpipe Construction. By H. G. Bucner, 4.R.0.8¢.1., a.1.0., Demonstrator. in the Royal College of Science, Dublin. (August, 1921.) Uncharged Nuclei produced in Moist Air by Ultra-Violet Light and other Sources. By the late Prof. J. A. McCurtuanp, F.z.s., and J. J. M‘Hewry, u.sc., University College, Dublin. (August, 1921.) (Nos. 21 to 24, price 7s.] Biological Studies of Aphis rumicis L. A.—Appearance of Winged Forms. B.—Appearance of Sexual Forms. By J. Davinson, p.sc. From the Entomological Department, Institute of Plant Pathology, Rothamsted Experimental Station, Harpenden, Herts. [Communicated by Professor G. H. Carpenter.] (August, 1921.) The Occurrence of Dewalquea in the Coal-Bore at Washing Bay. By T, JoHNSON, D.Sc., F.u.s., Professor of Botany, Royal College of Science for Treland, and Janz G. Giumors, B.sc. (Plates XI, XII.) (August, 1921.) A Simple Form of Apparatus for observing the Rate of Reaction between Gases and Liquids, and its use in determining the Rate of Solution of Oxygen by Water under different conditions of Mixing. By H.G. Bucxer, A.R.C.SC.1., A.I.c., Demonstrator in Chemistry, Royal College of Science, Dublin. (August, 1921.) The Occurrence of a Sequoia at Washing Bay. By ’. Jounson, D.sc., F.1.5., Professor of Botany, Royal College of Science for Ireland, and Janz G. Gitmorn, B.sc. (Plates XIII, XIV.) (August, 1921.) The Sources of Infection of Potato Tubers with the Blght Fungus, Phytophthora infestans. By Paun A. Murray, B.a., a.R.c.sca., Assistant. in the Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (August, 1921.) [ Nos. 25 to 29, price 6s. 6d.] DUBLIN : PRINTED AT THE UNIVERSITY PRESS RY PONSONKY AND GIBBS. THE SCIENTIFIC PROCKEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.S.), Nos. 30-34. FEBRUARY, 1922. 30.—SOME FACTORS AFFECTING THE HYDROGEN ION CON- CENTRATION OF THE SOIL AND ITS RELATION TO PLANT DISTRIBUTION. By W. R. G. Arkins, O.B.E., Sc.D., F.LC. [Communicated by Professor H. H. Dixon, Se.D., F.RB.S.] 31—THE HYDROGEN iON CONCENTRATION OF PLANT CELLS. By W. R. G. Arxkins, O.B.E., Sc.D., F.1.C. [Communicated by Professor H. H. Dixon, Sc.D., F.R.S.] 32—__ NOTE ON THE OCCURRENCE OF THE FINGER AND TOE DISEASE OF TURNIPS IN RELATION TO THE HYDROGEN ION CONCENTRATION OF THE SOIL. By W. R. G. Arxins, O.B.E., Sc.D., F.1.C. [Communicated by Professor H. H. Dixon, Se.D., F.B.S.] 33—PHOTOSYNTHESIS AND THE ELECTRONIC THEORY (11). By Henry H. Dixon, Sc.D., F.R.S., University Professor of Botany in Trinity College, Dublin; and Nicen G. Batt, M.A., Assistant to the University Professor of Botany, Trinity College, Dublin. 34.—THE BIONOMICS OF THE CONIDIA OF PHYTOPHTHORA INFESTANS (MONT.) DE BARY. By Paut A. Murpny, B.A... A.R.C.Sc.1., Assistant in the Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. | Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAI, DUBLIN SOCIKLTY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON Waa? Ins a x lon é ~y GEN 1922. Vi’ Price Seven Shallings. Roval Dublin Society. VTE FOUNDED, A.D. 1731. INCORPORATED, 1749. SCIENTIFIC MEETINGS. Tur Scientific Meetings of the Society are usually held at 4.15 p.m. on the fourth Tuesday of every month of the Session (November to June). The attention of authors is directed to the following extracts from the regula- tions governing the presentation of communications at Scientific Meetings of the Society, and the publication of papers in the Scientific or Meconomie Proceedings :— ‘¢An author who desires to present a communication at a Scientific Meeting of the Society shall furnish to the Chief Hxecutive Officer for Science a fortnight previous to that meeting the title and an abstract (not exceeding 200 words) of his communication, together with an estimate of the time required for presentation. «An author who desires to have a communication published forthwith as a paper in the Proceedings of the Society shall furnish to the Chief Executive Officer for Science the full text and necessary illustrations at least a week before the meeting of the Committee of Science at which he desires publication to be considered. Meetings of the Committee of Science are usually held on the second Tuesday of each month from November to June inclusive. *‘ A paper may be published either in the Scientific Proceedings or the Economic Proceedings of the Royal Dublin Society, as directed by the Committee of Science. ‘The copyright of papers so published shall be the preperty of the Society. “ Fifty copies of each paper shall be supplied free of cost to the Author or to each Joint Author. ‘« Additional copies may be obtained at Cost Price if applied for by the Author when returning the proofs.” Authors are requested to apply to the Chief Hxecutive Officer for Science for further information concerning the preparation of papers and illustrations for press. [ 369 = XXX, SOME FACTORS AFFECTING THE HYDROGEN ION CONCEN- TRATION OF THE SOIL AND ITS RELATION TO PLANT DISTRIBUTION. By W. 1, G. ATKINS, O.B.E., Se.D., F.1.C. (Communicated by Professor H. H. Dixon, sc.p., F.R.8.) [Read NovemBer 22, 1921. Published Fenrvary 2, 1922.] CONTENTS. PAGE Snare Introduction, 6 0 369 (@) Solubility of phosphates and occur- , “ concentration and rence of sulphur acids, 5 383 eae iy 370 (e) Solubility of iron salts, . samt oeare yey (b) Eee action : 371 (f) Effect of manuring upon soil reaction, 385 (c) The determination of Rparoeen ion (g) Acid soils, : . : : . 386 concentrations 871 (k) Aluminium salts (see under plant (d) Neutrality, 373 distribution). (ce) The bydrogen ion RRR of The hydrogen ion concentration of heated wie eee 373 and unheated soil, é 387 a The aS resented ee eal of Feros Nomenclature of soils with sean to mining hydrogen ion concentrations, . 375 Bice and acters ‘ . 888 (g) Methods of determining soil reaction, 376 e relation of the hydrogen ion concen- j Ste Pete ‘1 tration of soil to the geological forma- Agricultural application of studies on soi a Son onsihncitinn / : ies Feactlon; 3 = F The relation of the hydrogen ion concen- The relation of the ions of the soil solu- tration of the soilto plant distribution, 393 tion to its hydrogen ion content, 378 Garden cultivation’ of wild plants . 406 (a) Calcium salts, 378 The hydrogen ion concentration of natural (4) Magnesium salts, 2 380 waters in relation to plant distribution, 407 (c) Action of gypsum on alkali nad, 382 Summary, . L 419 Introduction. Ir has long been customary to speak of certain soils as acid and of others as alkaline, the terms being used as a rule only when these characteristics were - displayed to a marked degree so as to have obvious effects upon the vegetation. Concerning the nature of soil acidity, a long-standing controversy is still in being, and the views put forward have recently been summarized by Soult inst To SCJIENT. PROC, R.D.S., VOL. XVI, NO. XXX. 370 Scientific Proceedings, Royal Dublin Society. Fisher (1921). Whether the acidity is produced by humic or other organic acids, by the selective absorption of bases, by the interaction of salts with insoluble alumino-silicates, by the presence of alumino-silicic acids, or by the presence of sulphur acids derived from pyrites, there is only one method of stating precisely the effective acidity—namely, by recording the concentra- tion of hydrogen ions in the soil solution or soil extract. This may be done by giving the hydrogen ion concentration in grams per litre, symbolised as C,, or [Ht], or by using the expression pH, which is defined as the logarithm of the reciprocal of the hydrogen ion concentration expressed in grams per litre. For a discussion of the advantages of this method of stating acidity, reference may be made to Clark (1920). But to render what follows more intelligible to those biologists who are familiar neither with the methods by which hydrogen ion concentrations are determined nor with the terminology adopted, a brief account of the subject is given here. (a) Hydrogen ion concentration and acidity.—The acidity of soils has been measured in various ways, of which an account is given by Clark, but it is found that the behaviour of the soil solution or soil extract towards plants is not correctly shown by titration, which gives a measure of the total replaceable hydrogen, not of the concentration of the hydrogen ion effective at any instant. It is by the latter that the effect on the plant is governed, as well as the effect on enzyme action and the solubility of various salts. For example, equal volumes of N/100 hydrochloric and acetic acids require identical amounts of alkali for their neutralization, but the former solution has a far greater concentration of hydrogen ions than has the latter, since in dilute solution hydrochloric acid is almost completely ionized, whereas acetic is not; it is accordingly said to be a weaker acid. If it is assumed that the ionization of hydrochloric acid is complete at this dilution, its concentration in terms of hydrogen ions is Cy, or[H*] =1x 10? grams per litre. This may be written 1 1 [H*]= To? or pH = log a = — log [H'*] = 2. It may at first sight appear to be both cumbersome and unnatural to use such an expressson as pH, viz., — log [H*}, to denote hydrogen ion concentra- tion, but in practice it is extremely convenient, and gives simple numerical values which are easily remembered. Moreover, since an increase of unity in a pH value denotes a decrease to one-tenth in the hydrogen ion concentra- tion, it is obvious that for any graphical presentation of results difficulties arise when the changes are of the order of 10% or 10%, yet these are quite common in researches on this subject. For certain purposes, however, it is at Arxins— Fuetors affecting Hydrogen Ion Concentration of Soil. 371 times convenient to use C,, values. To convert the - log [H*] or pH values into Cy or [H*] values use may be made of semi-logarithmie paper, as pointed out by Roaf (1920). The first decimal points of the pH values are marked off from right to left as abscissae, the Cy values from 0°1 to 1:0 being ordinates against the logarithmic rulings. A diagonal being drawn, the C,, value of any pH value may be read off, the whole number of the latter being the negative power of ten, by which the C, value so read must be multiplied. Thus, for example, with pH 4°50 it is found that 0°30 on the abscissa corre- sponds to 0°32 on the ordinate; the C, value is then 0°32 x 10. Conversion tables are, however, given by Clark, also at shorter intervals by Schmidt and Hoagland (1919). In making a graph for converting pH into Cy values, it should be noted that the logarithmic scale, as on a slide rule, begins at 1, not 0. A further reason for using — log [H*] values will be given later on, (b) Buffer action.—lf alkali is added to acid, after a certain amount has been run in, a neutral solution is obtained. With a strong acid, such as hydrochloric, the addition of the alkali in successive portions results in a progressive diminution in the hydrogen ion concentration, as may be seen from the fact that the strong acid was largely ionized at the start. But with a weak acid the neutralization of the hydrogen ions existing at any instant results in a new equilibrium being attained by the remaining undis- sociated molecules of the acid, which become ionized to the same percentage as before. ‘he alteration in the hydrogen ion concentration is therefore much less over a considerable range. If the results are plotted with pH values as ordinates and cubic centimetres of alkali added as abscissae, the slope of the curve will be less steep for the weak acid than for the strong, except near the neutral point. Curves illustrating this are given by Clark (1920) and by Fisher (1921), and in the older literature. A weak acid thus has a considerable “buffer action’ in preventing rapid alterations in the pH values, and the same is true of weak bases. Thus peptone and albumen solutions alter in pH value comparatively slowly when acids and alkalis are added. The same is true of many soil suspensions, such as those of peat and calcareous silt; and though these are not solutions, yet they may be con- sidered as having a considerable buffer action, since they yield a continuous supply of a very dilute solution of acid or alkali till exhausted. (c) The deternunation of hydrogen ion concentrations.—From theoretical considerations Nernst developed an equation connecting the electromotive force of a concentration cell with the concentration of its ions. Platinum black deposited on platinum and immersed in a solution through which pure hydrogen is bubbled constitutes a hydrogen electrode. On the assumption that in very dilute solutions the ions obey the gas laws, it has been shown 2x 2 372 Scientific Proceedings, Royal Dublin Society. that the potential of the hydrogen electrode changes with the concentration of hydrogen ions as follows :— which on integration becomes = — log. P+ A, where A is an integration constant, & the gas constant, 7 the absolute temperature, 2 the valency of the ion, and / the faraday or quantity of electricity carried by one gram equivalent of the ion, and log,P is the natural logarithm of the partial pressure due to the hydrogen ions. Now, if two such hydrogen electrodes are connected to form a concentration cell, the electromotive force developed is equal to = log, ° where C and C” are the concentrations of hydrogen ion, since the ratio of the pressures may be taken as equal to the ratio of the concentrations. In practice such a hydrogen electrode is connected by an inverted U-tube arrangement, containing a solution of potassium chloride, with a calomel electrode, and the electromotive force of this cell is measured by means of a potentiometer, with all due precautions. The constants being evaluated, use is then made of the following equation to determine the hydrogen ion concentration :— E.M.F. (obs.) - E.M.F. (of normal hydrogen and calomel electrode cell) 0:0001983 7 Using an N/10 KCl-calomel electrode, this becomes at 25° C. E.M.F. (obs.) - 0°336 = 0:9591 log = - 0:0591 pH. The fact that the term log im is thus directly determined is another reason for expressing hydrogen ion concentrations in —- log[H*] or pH vaiues. Full proofs of these equations and directions as to the technique may be found in “The Determination of Hydrogen Ions,” by Clark. The method is used as the fundamental one for determining pH values. By its means buffer solutions of accurately reproducible pH values may be standardized for use in colorimetric determinations by means of indicators, Arkins— actors affecting Hydrogen Ion Concentration of Soil. 373 (d) Neutrality.—It has not been explained as yet what is meant by a neutral solution or neutrality. Pure water dissociates primarily into hydrogen and hydroxyl ions, and the product of the ionic concentrations is a constant at constant temperature, namely, =e [H*] x [OH] =*[H O]= &. Since the ions are produced by pure water in equal numbers, the concentra- tion is the same for both when reckoned in gram-equivalents. A solution is accordingly termed neutral when the hydrogen and hydroxy] ions are present in equivalent amounts, asin pure water. With an acid solution the hydrogen ion is in excess, and the hydroxyl is correspondingly reduced, since the product is constant. In alkaline solutions hydroxy] ions preponderate ; but as the product is constant, it is possible and convenient to state the reaction of the solution in terms of hydrogen ion concentration rather than in terms of the hydroxyl. One scale, the pH or - log H, is therefore obtained instead of two, based upon the concentrations in pure water, as a starting-point for both. (e) The hydrogen ion concentration of pure water.—It is of interest to determine the hydrogen ion concentration of pure water, but this is attended by many experimental difficulties. Those due to the solution of minute traces of glass may be avoided by the use of silica or platinum vessels, but the absorption of carbon dioxide is still a trouble. It is not possible to determine the pH value by the potentiometer on account of the great internal resistance of the hydrogen electrode half-cell when made up with pure water. The various methods adopted have recently been reviewed by Beans and Oakes (1920). They are as follow :—(1) By deduction from measurements of the electromotive force of concentration cells made up with dilute solutions. (2) By conductivity methods, giving the ionic mobilities of hydrogen and hydroxyl ions in very dilute solutions, which are considered to be valid for pure water; the conductivity of the latter is then measured, and from it and the ionic mobilities the concentration of the ions 1s calculated. This is the method used by Kohlrausch and Heydweiller (1894), and is the only one of these methods based upon an examination of pure water, but even then use is made of data derived from solutions. For pure water these workers found the hydrogen ion concentration at 26°C. to be 1:10 x 10~, or pH 6:96. The value varies with the absolute temperature, as may be seen from the equation given previously. Michaelis (1914) gave for pure water at 16°C., pH 7:10; at 22°C., pH =7:00; and at 28°C,, pH 690 (3) By methods based on the hydrolytic dissociation of salts. (4) By B74 Scientific Proceedings, Royal Dublin Society. measurements of the rates of certain reactions. (5) By the use of coloured indicators, which give definite tints at known concentrations. All these methods agree in giving values which are not very far from pH 7:1. Method (5) is used by the writer to test the purity of freshly boiled distilled water and to test glass apparatus for solubility. As shown by its behaviour to brom thymol blue and phenol red, and compared with Clark’s standard solutions, the purest, freshly boiled, thrice distilled water which has been prepared here gave pH 7:10-7:05 at about 15°C. Boiling to remove the carbon dioxide in certain types of “hard” glass tube may, however, give any value from about pH 8 upwards, so every tube has to be tested and re-tested after use with alkaline precipitates. If used in the cold, however, such “hard” glass tubes are safe, whereas soft glass tubes are not. Once the indicator has been added the liquid is no longer pure water, hence method (5) is open to objection; but if due precautions are taken in making up the indicator, and if only a very minute amount, say two drops of a 0:02 per cent. solution, is added to 10 c.c. of liquid, the error from this source is slight. Precautions are necessary on account of the fact that pure water has a negligible buffer action. An entirely new method has been introduced by Beans and Oakes (1920). In this the hydrogen electrode is set up with the purest water, and the cell Hg |HgCl. KCl | KCl | H,0 | H, is used to charge a condenser of one microfarad capacity, which it does in three to five minutes, whereas a cell of lower internal resistance in which dilute acid is substituted for pure water only requires an instant. Now, the quantity of electricity stored is equal to the product of the E.M.F. of the cell and the capacity of the condenser, viz. @ = HC. When discharged through a ballistic galvanometer a deflection d, is obtained when the pure water hydrogen electrode is used. Using a standard cell of known voltage a deflection d, is shown. Now, since d is proportional to @, and @= ZO, therefore LC d. = 1/ da B.C’ or Ey Ey = dj/dr ; and, since #, is known, #, may be found, and the pH value calculated as in the potentiometer method. The value arrived at for pure water at 25°C. is pH 7:91, or [H*] 1:23 x 10°. This differs considerably from the results previously obtained, but the method seems very simple and direct. From the biological standpoint it is not of great importance to ascertain this constant precisely, as pure water is never found in nature, and its reaction is Arkins—Muctors affecting Hydrogen Ion Concentration of Soil. 375 not necessarily the most favourable for living cells, which can, in different species, tolerate a large range of acidity and alkalinity, a reaction suitable for one being too acid for another. (f) Lhe colorimetric method of determining hydroyen ion concentrations.— In addition to the electrical method, using a potentiometer and hydrogen electrode, or the more recent method with a condenser, which requires a less expensive outfit, the colorimetric method introduced by Friedenthal (1904) and Salm (1904) is available. It is based upon the potentiometer method as a standard, the latter being used to determine the pH values of the buffer solutions, made up to be at convenient intervals on the pH scale. Measured amounts of various indicators are added to these solutions. The electrical method is also used to check sources of error, such as those due to proteins and salts, which cause the indicators to give readings higher or lower than the true values. The colorimetric method was improved and extensively tested by Sorensen (1909), who introduced new indicators, and eliminated those liable to mislead. More brilliant water-soluble indicators of the sulphone phthalein series were introduced by Clark and Lubbs (1917), as well as standard buffer solutions, having certain advantages over those of Sdrensen. The indicators and standard solutions used in this research are those described by Clark and Lubs. | 174 | p74 | ¥74 | ¥70 | t7% | 70 Kaiya nee é 6 | 70) 476) F778 | 70 | 7o | o7e | 1701 a7 | Ee | EP | BZ ISTO — | #70) 470) +70) +74 | 170 | ¥74 | t74 | 70 | B70 | oP | oF | BZ Arg, : tH | F7E | PEL HE | ELE | 7h | Be | oe | oe) 7B | BB] BD | BE SION = SS a I cH A I Aig vi? | 4 | WB | O8 | Ge | Be | GH oe | we | & 6 9 g -t108 Jo uontpuod [ULSI a8v10}8 Jo soovtd puv “SON 6 ‘8 ‘soqng ssvjs url ydey WO] UT vIpruod Jo of] Jo YJSUEeT—]]] ATAVY, 448 Screntific Proceedings, Royal Dublin Society. The longest period during which the fungus was definitely proved to have retained its vitality in soil was forty-four days. It is worthy of note that with two of the tubes (Nos. 2 and 7) infection was just as easy and certain after the lapse of this period as at the beginning. From the same two tubes infection of potato slices was again secured sparingly after sixty-six days, and a still further trace of infection was obtained from No. 2 after seventy-four days, but in none of these cases was 1. infestans recovered. Pieces of tubers inoculated after the last two intervals from other tubes (sixteen in number) and with similar non-contaminated soil (eight in number) remained sound. The conditions under which the fungus survives longest are evidently determined more by moisture than by temperature. The longest periods of activity were found in tubes which were either moist to hegin with or were kept in a saturated atmosphere, irrespective of the temperature up to and including 20°C. At 30°C. the temperature becomes important in spite of the abundance of moisture. The most fatal conditions are those which com- bine high temperature and dry soil. Attempts at direct exanination of contaminated soil—Direct examination of the soil failed to give an explanation of these findings. It was concluded, after careful search, that the fungus did not survive in the form of mycelium. On the other hand, apparently perfect conidia, among others, were still seen in soil to which they had been added more than three weeks previously. The majority of the conidia observed under such conditions appeared, how- ever, to be shrunken but ungerminated. Others were empty, and from the presence of apical openings were concluded to have produced zoospores. No germ tubes were seen. Soil samples such as these, when placed on potato slices, infected them freely, while similar slices not inoculated remained sound, so that in the apparent absence of mycelium it was concluded that the conidia themselves had the power of retaining their vitality for comparatively long periods. An attempt to prove this by placing fresh conidia on twelve dry cover glasses in a small chamber, the atmosphere of which was saturated with water-vapour, proved only partially successful. Twelve other cover glasses similarly provided with conidia and prepared at the same time were placed in a similar chamber which was kept dry. The period of life was tested by infecting potato slices (with controls) from the fifth day onwards. The conidia from the moist chamber produced disease regularly until the ninth day, but not afterwards. Neither the conidia from the dry chamber nor the controls produced disease on any day. A later experiment in which the moist chamber was closed with a water-seal led to no better result. Murpuy—Bionomies of the Conidia of Phytophthora infestans. 449 Notwithstanding these findings, based as they are on laboratory experi- ments, there are indications that the life of conidia in air may be longer under more natural conditions. When a vigorous growth of conidiophores is left undisturbed for some time, it is noticeable that the fringe of the growth contains conidia which are few in number and of poor germinative capacity. Further in towards the substratum numerous and more vigorous conidia may be found. It is probable that the air in the centre of such masses of mycelium is more moist than in the fringe, which would account, to some extent, for the result. At the same time the comparatively low vitality of conidia when removed from the conidiophores and kept in a saturated atmosphere would go to show that another factor besides moisture must be involved. It is possible that large masses of mycelium contain an abnormal proportion of carbon dioxide, arising from the hyphae themselves, the substratum, or from bacteria and other contaminations, and that this has an influence on the conidia. The bearing of a reduced oxygen supply on the germination and vitality of the fungus is discussed mere fully in later paragraphs. The view sometimes expressed (3, pp. 38 and 39; 4), however, that that the conidia may live for some weeks in dry air under certain conditions, while in moist soil their existence is shorter owing to germination supervening after the first rain, is in opposition to our results, An experiment was twice repeated in which columns of sifted loam and silver sand, each 10 cm. deep, were contained in two upright glass tubes, each haying a diameter of 28cm. The tubes were open at the top, and had an arrangement to permit of drainage water escaping at the bottom. After the materials in the tubes had settled, a small quantity of contaminated sand was placed on the surface of the loam and sand. Water in measured quantities was then allowed to drip at fixed rates on the contaminated soil in the tubes. The liquid which came through at the bottom was caught on filter papers, which were transferred at regular intervals to the cut surfaces of potato tubers. The experiments were continued for six and five days respectively. It was found in these experiments that the filter papers became sparingly pathogenic to potato tubers after the passage of water equivalent to a “ rain- fall” of 40 to 80cm. or more, and seventeen to forty-one hours after the “rain” began. In the first experiment the sand alone allowed the infecting material to pass through, this occurring after twenty-six and a half hours, and none coming through theloam. In the second test (begun twelve days after the commencement of the first, the same tubes and soils being used as before, but with the addition of fresh contaminated sand) infecting material passed through the loam filter only, and that after seventeen, twenty-four SCIENT. PROC. R.D.S., VOL. XVI, NO. XXXIV. 30 450 Scientific Proceedings, Royal Dublin Society. and forty-one hours. In the light of later experience, the possibility that this material was left in the loam from the previous experiment must not be overlooked. The most significant point was that in three cases out of four in both experiments the originally contaminated surface-soil was more pathogenic to potatoes, even after a period of three to three and a half days, during which water equivalent to 162 cm. (approximately) of “rain” had passed through it, than the drainage water was at any time. The longest period of pathogenicity of the surface-soil under the conditions of continuous “rain” slightly exceeded four and a half days. I].—GERMINATION OF CONIDIA. Previous work on the germination of the conidia.—It was conceived that the method by which the conidia germinated might have a bearing on the period of vitality, zoospores being probably more delicate than germ tubes. It was found difficult, however, to deduce from previous work which method might be expected from conidia in the soil. While Berkeley and Montagne (6) saw the early stages of zoospore formation within the sporangium, and de Payen (20) observed the emptying of the latter, the meaning of the process was unknown to them and to the earlier workers, such as Morren in 1845 (according to Hecke), Schacht (22) and Speerschneider (24), all of whom observed direct germination! It was not until 1860 that the production of zoospores was recorded by de Bary (2) for P. infestans. De Bary (3, 4) was not able to find exactly what the conditions determining the two methods of germination were. He was first of the opinion that germ tubes were formed only when the tubers were infected in the ground, but later found that this did not always hold. He regarded zoospore formation as the normal procedure, and suggested that germ tubes were produced only by old conidia of low germinative power. Strong sunlight prevented the formation of zoospores. A third method of germination described by this author (2, 3, 4) for the first time consisted of tube formation followed at once by the produc- tion of asecondary conidium. De Bary regarded this as a specific mode of germination, due to partial immersion in water. The secondary conidium had the power of producing zoospores, or a tertiary conidium might result. Marshall Ward (25), to whom we owe the fullest and most accurate description of the germination process, considered that zoospore formation was favoured by the presence of oxygen in the water and by the absence of 1That is, the production of germ tubes. When zoospores result, the germination is said to be indirect, Murpeay—Bronomics of the Conidia of Phytophthora infestans. 451 organic matter, such as jam or dung, and of too strong sunlight. Tube germination followed by the formation of secondary conidia, while depend- ing to some extent on the conidia themselves, was favoured by light, thick sowing, and by any cause which delayed germination. A temperature factor did not seem to be involved. Hallier (9) found that zoospores were formed only when food was lacking, while germ tubes resulted in nutrient solutions. Smorawski (23) apparently arrived at the same result. This conclusion Hecke (10) also largely supports ; but in so far as he found no zoospores produced in solutions containing more than 0-5 per cent. of dry matter, while germ tubes occurred not only in such solutions but were sometimes associated with zoospores in very large numbers in distilled water, he concluded that some peculiarity must lie in the conidia themselves. His conclusion was that very young conidia only (“eben erst gebildete Konidien”) can produce zoospores, but that they do so imperfectly (with partial extrusion of the contents) in nutrient solutions. Older conidia never form zoospores. ‘heir germination is poor in distilled water and generally closes with a secondary conidium, while in nutrient solutions they germinate well. Brefeld’s (8) conclusion with regard to zoospore production in the Peronosporaceae (including P. infestans) was that thorough wetting favoured this form of germination. McAlpine (14) found that the first crop of conidia produced zoospores, while the later crops were variable. A short period of drying prevented zoospore formation. Jones, Giddings and Lutman (12) found that potato juice favoured direct germination and that temperature also affected the process, “At 25°C. more than 50 per cent. of the germinations are by tubes, . . . while at 10° to 20° C. direct germination is exceptional.” According to Melhus (17) the determining factor does not consist of age or any innate peculiarity of the conidia ; nor of light, the concentration of the solution, or the amount of oxygen available. On the other hand, tempera- ture is all-important. At 13°C. the maximum of zoospore formation is found, while at about 23° C. this has practically ceased and yerm tubes are the rule. It is only in extremely strong nutrient solutions (20 per cent. dextrose) that zoospore formation is replaced by some direct germination. Fresh experiments on the germination of the conidia.—The results of previous work on the conditions regulating the manner of germination being found so conflicting that no conclusion could be drawn, the problem had to be taken up afresh. Fortunately in the course of this investigation it was 302 452 Scientific Proceedings, Royal Dublin Society. found possible to reproduce conditions in microscopical preparations under which the conidia remained viable and capable of infecting tubers for several weeks. As these conditions resemble in many respects those which occur in soil, it is reasonable to suppose that the explanations which will be given of the phenomena observed in the preparations are also applicable to natural conditions. This portion of the work also resulted in extending our knowledge of the morphology and biology of the fungus in some important particulars, one of which at least has a bearing on the fate of conidia in the soil. The findings of Hecke and Melhus were first compared. For this purpose the germination of conidia was tested in the following liquids, which were sterilized before using :—(1) distilled water, (2) tap water, (3) soil extract, prepared by the slow filtration of tap water through four inches of soil, and (4) a 1:5 per cent, solution of glucose in distilled water. To observe the effect of temperature, one set of germination tests in these media was carried out at room temperature (which during these trials never sank below 10°C. at night, or rose above 15° C. during the day, with one exception, when 22° C. was reached), while another set was carried out in a constant temperature incubator at 22°-23° C. Both sets were kept in the dark, Conidia were derived from Petri dish cultures on potato slices, which were infected from pure cultures on oat agar. Every care was taken to use uniform material from the same culture for both sets of tests. The germinations were carried out under cover glasses (except in some special cases), and the slides were kept in moist chambers to prevent drying. It is possible in this way to keep the same preparation under observation for several weeks, and to. change the liquid at will without materially disturbing the conidia. Comparison of the effects of nutrient solutions and temperature on germina- tion.—Our work on the conditions (so far as they were tested) governing the germination of the conidia agrees best with that of Marshall Ward (25). These conditions were found to be very complicated, as would be expected from the results of earlier investigations. The comparison of the effect of the solution and of the temperature was repeated seven times without achieving a constant result, as may be seen in Table LV. . of Phytophthora infestans. “a O a yf the Con ~OnoOMUCS O Mureay—B. sraquind eSivp Area @ syuasaidas (~) Toqmds ayy, “pazunoo suaqmnu DATIL[AL OY) JUasatdat sosvo OULOS UT SurMo[[oF seinsy ayy, “BIpruocd sv SutuoNoUN; 9soy} 10; Q { vISuerods sv Sutmorjoung satods tof spurs g oO 9) are a) eorry, {quo 9 evry], 5 (0) ATR 1G 8G Ajuo g pooy, *® anely, (11) 0 (9¢) 8 pooy 258) UCT 16 eI = auoN 2 ®) Tey o-9) a0wry, ae) aoe], IZ °% ‘8% - 2 eoe ~ © ef “ O arg 30) ue 8 CE aa) aout], 8) poor) = {0 100g 86-49) oer, IGG “SG “0 a0euy, ro 0), poony Ajuo 9 100g uo (6) edVLy, 1G 3 ‘FG Apuo 9 100g A[uo 9 Ue = 9uO NT Ajuo 9 UBT 1G °G “&6 38 Ie (1) 0 (#) § poor (thos poox) (1) 0 (“) § pooy, 1G ‘8% 8 poon {Tuo § poon SnS poox) (1)o () 8 pooy 16°81 8 poor) (zz) 0 (66) § poor) Tuo g pooy Tuo § pooy 1G °% “8% FES poon (11) 0 (16)'$ pooy (Ft) 0 (02) § RSAUCE (z) 0 (21) § pooy IZ 3 “9% S68 100g 8 TUT eg 100g Sats) ue IZ °% “92 {Tuo g poor) sTuo § 100g Aquo g pooy, Beas 100g 1G °G "FG (4) 0 (cz) § pooy (91) § (#2) 0 pooy = auoNy ATuo g pooy 16 6 ke “poyyeyL “MOTJBULULI9 4) *POUaTAL *WOTJBULULLAY) “poyyey “UOTPEUTONIAY) "poyiey, *UOLPUUTUIAY “quowtied xa = Jo ayeq. ‘Jos quoo aed ¢.[ esoonpy *qovayxe TIOG “19yVM ey, 0 oG1-.0T dh “TOJVM PITS, “BIPIUOD JO WOTJVUIWIES oY} UO oInyetedmo} Jo puUL SMOTINIOS JUOTIYNU Jo soUENFUT— A] ATAV], 454 Scientifie Proceedings, Royal Dublin Society. The following conclusions from this experiment seem to be justified (see Table IV) :— Temperature of 10°-15° C. 1. At a temperature of 10°-15° C. the regular method of germination is by means of zoospores. With good material the germination runs from 50 to nearly 100 per cent. 2. Generally speaking, the number of conidia producing germ tubes at this temperature is small, but it may under certain conditions exceed the number producing zoospores. 3. The influence of nutrient solutions on the production of germ tubes was not apparent in the case of the sugar solution used, but there was, perhaps, somewhat more evidence for it in the case of soil extract. A somewhat larger number of abortive (unescaped) zoospores was found in the sugar and soil solutions than elsewhere. ' 4. There is considerable evidence that the tendency to produce germ tubes at a temperature of 10°-15°C. rests in the conidia themselves (see test of 26. 2. 21 at 10°-15°C.). Temperature of 22°-23° C. 1. At 22°-23°C. the regular method of germination is by means of germ tubes. The germination seldom exceeds 50 per cent., and is generally less. 2. Ordinarily the production of zoospores is rare, but conidia with undifferentiated contents but with the papilla gone, or with contents extruded, or containing abortive zoospores, are frequent. 3. On one occasion (test of 1. 3. 21 at 22°-23°C.) the production of zoospores largely replaced that of germ tubes, indicating either a pecu- liarity on the part of the conidia then used, or that temperature does not act directly on germination, but influences it through some factor which it generally (but not always) calls into play. 4, The influence of nutrient solutions was not marked, but soil extract seemed to favour tube formation. The tests carried out at 22°-23° C. gave percentages of germination which were uniformly much lower than those found at the lower temperature. oll was evident that the conidia were near the limits of their development. It is believed that the results recorded have little, if any, bearing on the con- ditions under which blight epidemics occur. Ltfect of age on the method of germination.—Some conidia germinate by means of germ tubes under conditions which cannot be explained by reference Murpay— Bionomics of the Conidia of Phytophthora infestans.’ 455 to aération, presence of other organisms, temperature or nutrient solutions. From a consideration of the frequently different behaviour of two conidia lying close. together under apparently similar conditions, one is forced, along with most previous workers, to conclude that different potentialities must reside in the bodies themselves. This was one of de Bary’s views, and his suggestion that germ tubes are the result of age and lack of vigour has been endorsed by many of his successors without any convincing proof being adduced to support it. Evidence in support of this view can, however, now be given. Conidia were tested from an eight-months’-old pure culture on oat agar. ‘The aérial portion of the mycelium at least appeared dead, and it must have been a considerable time since any conidia had been produced. There were, how- ever, some perfect conidia present among a multitude of dead ones. The former germinated in sterile water at 10°-15°C., somewhat slowly but strongly, and entirely by means of germ tubes. A peculiar feature of their germination was that the tube or tubes (often up to five) occasionally sprang from all parts of the wall, although generally they came from the normal place beside the papilla. These conidia were also peculiar in possessing a_ yellowish oil drop and an unusually prominent papilla, just like secondary conidia. It is conceivable that this is the outward sign of their having gone through some maturation process. Similarly with conidia placed in water, whether primary or secondary, if they do not germinate within about twenty-four hours they seem practically incapable of producing zoospores. Here again in most cases (always, indeed, with secondary conidia) we get the same signs of an inward change. Further- more, when such conidia lie long in water (three weeks), the germ tube may arise from any part of the wall. An attempt, however, to distinguish between conidia from a comparatively fresh growth which germinate by means of tubes soon after being placed in water at 10°-15°C., and those with a similar origin which produce zoospores, resulted in failure. Many conidia apparently similar in all respects were marked, some of which afterwards germinated in one way and some in the other. It was observed that only the hyaline conidia germinated at all. Those which were at all dark-coloured or slightly plasmolysed failed to grow. Il1.—Errectr or LAcK OF OXYGEN ON DEVELOPMENT OF CONIDIA. Germination delayed or prevented by lack of oxygen.—During the course of these tests it was noted several times that the presence of bacteria or yeasts © 456 Scientific Proceedings, Royal Dublin Society. arrested germination. It soon became possible to distinguish between conidia - which were incapable of development and those the development of which was merely delayed. The least sign of plasmolysis is a sure index of death, but before this appears a peculiar darkening and granulation cf the protoplasm indicate, according to our experience, that the conidium is incapable of development. It was found that some slides (particularly those kept at 23°C.) became badly contaminated with bacteria and yeasts, and that under such conditions germination was confined to a zone about 4mm, wide near the edge of the cover glass. Within this ring a negligible amount of germination, or none whatever, was to be seen. If, however, the slide was very carefully and repeatedly irrigated with its appropriate solution, the process being watched under the microscope and extreme caution being exercised not to disturb the conidia, it was found that the germination of a large number of the central conidia could be brought about. ‘I'hree hours after irrigation the result was sometimes apparent, but sometimes it took twenty-four hours. While traces of indirect germination have been seen under these circum- stances, germ tubes are almost the invariable rule, irrespective of the tempera- ture or of the nature of the solution. This experience was repeated many times, at first involuntarily and afterwards of set purpose. A test in tap water at 23°C., begun on 2.3. 21, became contaminated with bacteria after twenty-four hours, and then showed only 18 conidia with germ tubes out of 198 conidia examined. Most of these were near the edge, there being very few in the centre. After irrigation and a further period of twenty-four hours, the number of conidia with germ tubes was 112 out of 271 examined. Another slide containing soil-extract and belonging to the same series also became contaminated and showed four conidia with germ tubes out of a large number examined after twenty-four hours. On irrigation the number rose to 34 in 89. That it was the presence of bacteria which prevented germination was proved repeatedly by setting up some tests under practically sterile con- ditions, while in others bacteria were introduced. The former (at 10°-15° C.) showed numberless zoospores after about three hours, and after twenty-four hours the majority of the conidia had germinated in this way all over the slide. Where bacteria had been introduced perfect zoospore formation was found only in a narrow zone near the edge of the cover glass, while there might be attempts at zoospore production elsewhere. If now fresh water were copiously introduced a great access of germination resulted in most cases, germ tubes only being produced. Control slides to which no water was added showed no change, or at most an entirely negligible amount of tube Murrayv—Bvonomices of the Conidia of Phytophthora infestans. 457 production. Corresponding results were secured with preparations kept at 22°-23° C. This experiment was repeated using motile protozoa along with conidia, which makes a beautiful preparation. The protozoa after twenty-four hours are confined to a narrow zone near the edge of the cover glass, in which they swim about freely. In this area, and extending somewhat within it, good germination is to be found, whilst still further in there may be no germina- tion whatever. Motile bacteria exhibit the same zonation to some extent, the outer ones only being in motion. Penicillium spores have also been used with somewhat similar results. In this case the rapid development of P. infestans allows some zoospores to be produced, but a great many become abortive. These results, it is believed, are to be explained by the fact that the competing organisms use up the oxygen in the water and replace it with carbon dioxide faster than the conidia can utilize it for the purpose of _ germinating. There may bea question of other excretion products having an inhibitory effect, but judging from the variety of organisms used this is hardly likely. The same effect has been produced by sealing round the edge of the cover glass with melted paraffin. Furthermore, the conidia themselves, if the slide is very thickly seeded, bring about the same result. When germination is general around the outside of a clump, there is often none at all in the centre. Conidia scattered among the hyphae of living mycelium of P. infestans frequently behave similarly. Formation of secondary conidia.—Secondary conidia, as originally described by de Bary (2, 3, 4), are the immediate products of previously existing conidia, being borne on the germ tubes of the latter. The name “ secondary conidia” is not a particularly good one, for conidia morphologically and physiologically identical are produced by hyphae which fructify in water, and identical but smaller bodies may be formed by zoospores, as will be shown. The term is used in a somewhat loose sense in this paper for the new conidia which sometimes follow after the germination both of conidia ~ and of zoospores. Such secondary conidia seem to owe the characteristic appearance which is common to them and to conidia produced by submerged hyphae to the fact that they are formed in water, on which account a word like “ hydro-conidia,” which would include all three, might be preferable. The conditions under which germination is arrested but subsequently takes place are favourable for the production of secondary conidia. These have been very common on our slides. When germ tubes are formed in the presence of bacteria, or when bacteria are introduced after their formation, secondary conidia will always be met with within a zone from about 2 to 4mm. 458 Scientific Proceedings, Royal Dublin Society. wide inside the edge of the cover glass. In preparations sealed with paraffin they are found only at the extreme edge. Their formation is constant, and is the result of specific conditions. Ata certain degree of oxygen concentration which just permits germination, or when the oxygen supply is gradually cut off after germination has taken place, secondary conidia are produced. Bacteria are the most reliable agents for bringing about this state of affairs, as may be seen by introducing them into one of two uniform preparations. At a still lower oxygen concentration ungerminated conidia are eventually (and sometimes rapidly) killed; but there is evidence derived from sealed preparations to show that germ tubes continue living for several weeks, and even increase in length, with a still scantier oxygen supply. This is possibly connected with the greater amount of exposed surface, which facilitates respiration. For some reason which has not been established, secondary conidia appear to be materially more resistant than are conidia formed in air to the adverse influences they meet with in water in competition with other organisms ; and they are, therefore, an important adaptation to certain aquatic conditions, The formation of secondary conidia has been mentioned by most workers since their discovery by de Bary (2, 3, 4), but certain new observations may be recorded. The conidia which give rise to them rapidly lose all their contents, which pass in their entirety into the smaller and less symmetrical secondary bodies. If the primary conidium has two germ tubes a secondary conidium may be produced by each. The hypha joining the primary and secondary conidia is of variable length, or may exceptionally be lacking, the second conidium being an asymmetrical prolongation of the first. The germ tubes generally grow in the direction of the greater oxygen concentra- tion. Frequently the new conidia are found in the free water surrounding the cover glass, being borne on very long tubes springing from conidia lying further in. ‘These hyphae grow straight to the margin, perhaps after a preliminary turn. The secondary conidia have thinner walls and more rounded and prominent papillae than the original ones, and they always contain one (or occasionally two) yellowish oil-drops. Even when lying in the free water around the cover glass they may remain ungerminated but viable for about three weeks. Frequently, however, they germinate sooner, and they may produce zoospores if the conditions are favourable; but germ tubes are the general rule, frequently followed by a third conidium, as de Bary and others have described. These have frequently been seen, and on several occasions even what may be called a quaternary conidium, four conidia in a chain, each somewhat smaller than the preceding one. It is conceivable that repeated germination and conidium formation may result in Mureuy— Bionomies of the Conidia of Phytophthora mfestans. 459 a further lease of life, analogous to the survival of mycelium under cir- cumstances which are fatal to fresh conidia. Zoospores and the production of “secondary” conidia.—When conidia germinate by the production of zoospores, the latter, in the presence of sufficient oxygen, settle down after a longer or shorter period of activity, develop a cell wall and immediately put out a germ tube. There is no doubt that oxygen in sufficient quantity favours this whole proceeding. When oxygen is lacking the germination is often abortive, even at favourable temperatures. The zoospores may be more or less differentiated, but none may emerge; or some or all may escape, but they do not swim far from the sporangium and soon come to rest. Under such conditions it frequently happens that no cell wall is formed, and the zoospore dies and dissolves. That it is lack of oxygen which limits the activity of the spore in these respects is shown by the presence of naked and otherwise abortive zoospores among masses of Phytophthora mycelium. The coming to rest of large numbers of zoospores in a cluster, which is sometimes seen under these conditions, is explainable on similar grounds. The chance coming of one or more into a zone of low oxygen concentration puts an end to their activity, and the accentuation of the dearth of oxygen which their presence causes entraps other zoospores which happen to swim into the locality, and so the process goes on. When the zoospores germinate under favourable conditions, the germ tube is of such enormously greater capacity than the original spore, and continues increasing in length (even though not completely filled with protoplasm), or at least remains alive and vigorous-looking so long, that a saprophytic existence is at once suggested (7). However this may be, it is certain that the zoospore germ tubes can live for one week in water, for they have been directly observed to do so for that period. The statement of de Bary (3) that they cannot live longer than twenty-four hours under such conditions is not correct. A single week, however, is not the limit of their existence. They, too, have the faculty of producing “secondary ” conidia, an observation whith does not seem to have been recorded previously. The conditions are the same as for conidium production by conidia—namely, a reduction in the amount of oxygen. The phenomenon was first observed in a sealed preparation where conidia had discharged zoospores surrounding an air bubble. Six days afterwards there was an attempt at conidium formation by a zoospore germ tube, but it did not come to fruition. Subsequently in a preparation originally set up under sterile conditions and kept for five days at 10°-15°C., in which there had been abundant zoospore formation, the 460 Scientific Proceedings, Royal Dublin Society). almost universal production of “ secondary ”’ conidia by the zoospores situated near the edge of the cover glass was observed. ‘lhe slide had become con- - taminated with bacteria. These new conidia were borne on the ends of long spirally twisted germ tubes. They contained all the protoplasm, even of the branches, where such were present. In form they exactly resembled those produced by conidia, being asymmetrical, and provided with prominent papillae and yellow oil-drops, but they were much smaller, although having a con- siderably greater volume than the original zoospore. In a companion pre- paration, which was not contaminated, the zoospores did not produce “secondary ” conidia. It is evident that the faculty by which germinating zoospores may prolong their existence under unfavourable conditions by developing directly into conidia must be of significance in the life-history of the fungus. The general and rapid production of zoospores by conidia at relatively low temperatures and in subdued light is evidently an adaptation to facilitate leaf infection during the heavy night dews of late summer and autumn. It is probable that under these conditions failure to bring about infection in the night or morning hours results in the death of the spores, at least in dry weather. In the case of conidia which fall to the ground, even though they function as sporangia, it is clear, in the light of our present knowledge, that the resulting zoospores are not nearly so delicate as to lose their vitality (unless they succeed in infecting a potato tuber) a short time after the first shower of rain which brings about their liberation. The conidia produced by zoospores have pro- bably a measure of the same power of resistance as is possessed by other secondary conidia, so that instead of a single body capable of living for days or weeks and then infecting the plant under favourable conditions, the eventual sources of infection are multiplied manifold. he extent to which this multiplication can go was not determined, but “ secondary ” conidia, the product of zoospores, were seen which were empty and open at the apex as though they, in turn, had again produced swarm spores. Relation of results obtained in microscopical preparations to soil conditions.— The germination of secondary conidia was observed twenty-four days after the original conidia were placed in water, and the germ tubes of the former were still living arter thirty-four days. Potatoes were infected many times with conidia which had been fifteen to sixteen days in water (during which time they had germinated, and im some cases produced further conidia), the entry of the germ tubes into thin slices of tubers being observed under the microscope. The hyphae grew between the cells, and fine convoluted branches were sent into the cells and between the starch grains. Where secondary conidia were not formed, the germ tubes were found to be an Mureuy— Bionomies of the Conidia of Phytophthora infestans. 461 important aid in prolonging hfe. While original ungerminated conidia produced in air were not observed to be viable after a period of five to seven days in water, germ tubes and secondary conidia are much more resistant. The low oxygen concentration in which the slow development and continued vitality of the fungus were directly observed for comparatively long periods in microscopical preparations is believed to offer an explanation of the similar resistance exhibited in the soil. It is reasonable to suppose that all the conditions reproduced artificially must be realized in the soil, as well probably as many others. According to Russell and Appleyard (21), there are two distinct atmospheres in the soil—the “ free air” and the “dissolved air.’ The former does not differ in principle from atmospheric air, but it contains a larger and fluctuating amount of carbon dioxide (about 0°25 per cent.), and is almost saturated with water vapour. The dissolved air contains practically no oxygen, and is made up of carbon dioxide and nitrogen in varying pro- portions. Itis shown that the oxygen requirements of the soil are always in excess of the supply, and thus the two bodies of air, though existing in close relationship to each other, retain their individual characteristics. 'Tempera- ture, cultivation, manuring, and rainfall in particular, however, affect the soil atmospheres. A heavy fall of rain, which is nearly saturated with oxygen, partially renews the dissolved air. This, as evidenced by the subse- quent rise in the CO, content, allows the development of aérobic organisms. Among the latter the conidia of P. infestans must find in the fluctuations of the dissolved atmosphere the conditions which enable them to survive for the long periods noted in the experiments with contaminated soil, and to germinate when conditions are made more favourable by the accession of water rich in oxygen. There is another feature in the fluctuation in the amount of carbon dioxide in the soil which may influence the course of the blight fungus. The channel along which the parasite makes its way from the tubers to the leaves in spring is not known with sufficient certainty. The stem of the potato has been proved in a few cases to be penetrable by the mycelium, but the possibility of the transference taking place through the soil has hardly been studied. This is the more curious since the downward course of the parasite in the autumn from the leaves to the tubers is admittedly through the soil. It may be significant that the time at which the parasite finds its way down- wards coincides with the minimum CO, content of the “ free ” soil atmosphere, while in the spring, when the course is upwards, the CO, content is at. its maximum. The results of an experiment to test the capacity of the blight fungus to penetrate the soil were so suggestive that further attention is being devoted to the subject. 462 Scientific Proceedings, Royal Dublin Society. SUMMARY. The conidia of Phytophthora infestans when mingled with soil and kept out of doors may remain viable and capable of infecting potato tubers for a period of between three and four weeks. The vitality of the conidia is retained longer in loam than in a quick- drying medium like silver sand. In experiments indoors conidia survive longer in comparatively dry to somewhat moist soil than in very wet soil, the respective periods being forty and twenty-six days. Temperatures up to and including 20°C. have no ill effect on the conidia as long as sufficient moisture is present. At 30°C. the temperature is not too high for the conidia to survive for twenty-six days if the atmosphere is saturated. The most fatal conditions are those which combine lack of moisture and high temperature. Under the most favourable conditions indoors, conidia in soil infected tubers just as freely after forty-four days as when fresh. Conidia when placed in soil survive without the production of obvious mycelium. ‘I'he attempt to preserve ungerminated conidia in air saturated with water-vapour met with comparative failure, nine days being the apparent limit of existence. Initially contaminated surface soil, after the continuous passage through it of large quantities of water, may remain pathogenic to potatoes for a period exceeding four and a half days, and generally infects potatoes more freely than the water after passing through 10cm. of soil does at any time during the period. As having a possible bearing on the resistance of conidia in soil, their method of germination was re-examined in the light of previous work, which is summarized. At a temperature of 10°-15°C. the conidia germinate freely, and generally produce zoospores. A small number of germ tubes may, and often do, result ; while exceptionally they may be in the majority. Nutrient solu- tions, such as soil filtrate and 1°5 per cent. glucose in water, are without material influence in favouring tube production; but such solutions hinder somewhat the perfect formation of zoospores. Abundant oxygen and the absence of competing organisms favour zoospore production. The tendency to produce germ tubes at low temperatures seems to be innate in the conidia themselves. Conidia which have remained for some time ungerminated in air and those which have had their germination in water delayed for any reason seem capable only of giving rise to germ tubes. Murreay—Bionomics of the Conidia of Phytophthora infestans. 463 At a temperature of 22°-23° C. the limit of development of the conidia is approached. The germination is comparatively poor and germ tubes are the general rule. While abortive sporangia are frequent, the production of perfect zoospores is rare. Exceptionally, zoospores may largely take the place of germ tubes. The reason for this is obscure, but is believed to lie in the conidia themselves. Secondary conidia and old but still viable conidia from air resemble each other in possessing amore prominent papilla than usual, anda large yellowish oil-drop. These appear to be signs of their having undergone some maturation process which increases their resistance. Both as a rule produce germ tubes, generally more than one, which may spring from any part of the wall. The germination of conidia in water is delayed by lack of oxygen, and may subsequently be brought about by the timely supplying of fresh water. Suspension of germination has been effected by means of bacteria, protozoa, Penicillium spores, thick seeding of conidia, and by ringing cover glasses with paraffin ; and a good measure of germination has thereafter been effected by adding fresh water. : Secondary conidia, so called, are constantly produced at a certain degree of oxygen concentration which just permits germination, or when the oxygen supply is cut off after germination has taken place. In microscopical prepara- tions they are confined to a zone near the edge of the cover glass, within which germ tubes may live and grow without having the power of forming new conidia, Ungerminated conidia from air are rapidly killed under the latter conditions. Secondary conidia appear to be conidia adapted to an aquatic existence, similar bodies being produced by hyphae which fructify in water. They are more resistant than primary conidia, and this resistance is increased by the taculty they possess of successive germination followed by fresh conidium formation, which may be repeated four times. They may also produce zoospores under favourable conditions. Zoospore germ tubes may also develop at once into “secondary ” conidia when the oxygen supply is reduced after germination has taken place. Zoospore germ tubes may live for at least seven days in water, and their life is further lengthened when new conidia result. The secondary bodies so formed may apparently again produce zoospores. The limit of existence of fresh ungerminated conidia in competition with other organisms in water is about five to seven days, and the period may be much shorter. After a germ tube is produced the resistance of the fungus is increased. Potato tubers have frequently been infected with such conidia fifteen to sixteen days after being placed in water. The germination of 464 Scientific Proceedings, Royal Dublin Society. secondary conidia has been observed twenty-four days after the original conidium was placed in water, and the resulting germ tubes lived at least ten days longer. The conditions under which the continued existence of conidia, zoospores and germ tubes in water was directly observed for periods exceeding, in some cases, thirty days are believed to represent the conditions under which existence is possible for comparatively long periods in the soil. iw) Or LITERATURE CITED. . APPEL, O.—Beitrage zur Kenntnis der Kartoffelpflanze und ihrer Krankheiten I. Arb. K. biol. Anstalt Land.- und Forstwirtsch., Band 5, Heft 7, pp. 377-435, 1907. . Bary, A. p—E.—Sur la formation de zoospores chez quelques cham- pignons. Premier mémoire. Ann. des Sciences Nat., 4 Ser. Zool., T. 13, pp. 236-251, 1860. . Bary, A. DE.—Die gegenwartig herrschende Kartoffelkrankheit, ihre Ursache und ihre Verhiitung. 75 pp. Leipzig, 1861. . Bary, A. DE.—Recherches sur le développement de quelques champignons parasites. Ann. des Sciences Nat., 4 Ser. Bot., T. 20, pp. 5-148, 1863. . Bary, A. bge.—Researches into the Nature of the Potato-fungus Phytophthora infestans. Jour. R. Agric. Soc. England, Ser. 2, vol. 12,-pp. 239-269, 1876. Also in Journal of Bot., N.S., vol. 5, pp. 105-126 and 149-154, 1876. . BERKELEY, Rev. M. J.—Observations, Botanical and Physiological, on the Potato Murrain. Jour. Hort. Soc. London, vol. 1, pp. 9-834, 1846. Includes some work of Montagne’s. . BREFELD, O.—Botanische Untersuchungen tiber Hefenpilze. Heft. 5. Leipzig, 1883 (see p. 8 ef seq.). . BREFELD, O.—Untersuchungen aus dem Gesamtgebiete der Mykologie. Band 14. Miinster 1. W., 1908 (see p. 116 et seq.). . Hattie, E.—Neue Untersuchung der durch Peronospora infestans Casp. herforgerufenen Krankheit der Kartoffeln. Zeitschr. f. Parasi- tenkunde, Band 4, Heft 3, p. 263. Ref. Just’s Bot. Jahresber., Jahrg. 3, p. 999, 1875. Mureny—Bionomics of the Conidia of Phytophthora infestans. 465 10. iil. 14. 15. 16. U7. 18. I®), Hecke, L.—Untersuchungen tber Phytophthora infestans De By. als Ursache der Kartoffelkrankheit. Jour. f. Landw., Band 46, Heft 1, pp. 71-74; Heft 2, pp. 97-142, 1898. JuNSEN, J. L.—Moyens de combattre et de détruire le Peronospora de la pomme de terre. Mém. Soc. Nat. d’Agric. de France, T. 131, 130 pp., 1887. . Jones, L. R., N. J. Grppines and B. F. Lurman. —Investigations of the Potato Fungus Phytophthora infestans. U.S. Dept. of Agric., Bureau of Plant Industry, Bull. 245,95 pp., 1912. Also issued as Bull. 168, Vermont Aer. Exp. Station. 3. KarsvEN, H.—Zweiter Bericht ther die von den landwirtschaftlichen Akademieen und Versuchstationen ... in den Jahren 1864 und 1865 ausveftihrten Untersuchungen iiber die Kartoffelkrankheit . . , Aun. d. Landw., Band 49, pp. 104-122, 1867. McAtrinet, D.—Some points of practical importance in connexion with the Life-history stages of Phytophthora infestans (Mont.) de Bary. Ann. Mycol., vol. 8, No. 2, pp. 156-166, 1910. Metnus, I. E.—The perennial mycelium of Phytophthora infestans. Centbl. f. Bakt., &e., Abt. 2, 39 Band, pp. 482-488, 1913. Metuus, I. E.—Hibernation of Phytophthora infestans of the Trish potato. Journ. Agric. Research, vol. 5, No. 2, pp. 71-102, 1915. Metnus, 1. E.—Germination and Infection with the Fungus of the Late Blight of Potato (Phytophthora infestans). Agric. Exp. Stn., University of Wisconsin, Research Bull. 37, 64 pp., 1915. Mourpuy, P. A.—Investigation of Potato Diseases. Dom. of Canada Dept. of Agric., Bull. No. 44 (Second Series), 86 pp., 1921. Muvrpny, P. A.—The Sources of Infection of Potato Tubers with the Blight Fungus, Phytophthora infestans. Se. Proc. R. Dublin Soe. vol. 16 (N.S.), pp. 353-368, 1921. . PAYEN, A. DE.—Développement et réactions du Botrytis infestans de la pomme de terre. Compt. Rend. hebd. de lAcad. d. Se, T. 25, pp. 696-699, 1846. . Russe, EK. J., and A. AppLEYARD.—The Atmosphere of the Soil: its Composition and the Causes of Variation. Journ. Agric. Science, vol. 7, Pt. 1, 48 pp., 1915. SOIENT. PROC. R.D.S,., VOL. XVI, NO. XXXIV. 3D 466 Scientific Proceedings, Royal Dublin Society. 2. Scuacut, H.—Bericht an das KG6nigliche Landes-Oeconomie-Collegium uber die Kartoffelpflanze und deren Krankheiten. 29 pp. Berlin, [1854]. 23. SMORAWSKI, J.—Zur Entwickelungsgeschichte der Phytophthora infestans (Montagne) De By. Inaugural-Dissertation Univ. Erlangen. 17 pp., 1890. Also in Landw. Jahrb., Bd. 19, 1890. bo 24. SPEERSCHNEIDER, JimDass das Faulen der Kartoffelknollen bei der sogennanten Kartoffelkrankheit durch die angestreuten und keimenden Sporen des Blattpilzes (Peronospora devastatrix) verursacht wird, durch Experimente bewiesen. Flora, N. Reihe 15 Jahrg., No. 6, pp. 81-87, 1857. Similar paper in Bot. Zeit., 15 Jahrg., No. 8, pp. 121-125, 1857. . Warp, H. M.—Illustrations of the Structure and Life-history of Phytophthora infestans, the fungus causing the Potato Disease. Quart. Journ. Microscop. Se., vol. 27, N.S., pp. 413-425, 1887. iw) OU —_ 10. li, 14, 15. 16. 17. 18. 19. 20. SCIENTIFIC PROCEEDINGS. VOLUME XVI. A Cryoscopic Method for the Estimation of Sucrose. By Henry H. Dixon, Sc.D., F.R.s., and T. G. Mason, m.a., sc.B. (January, 1920.) 6d. . The Carboniferous Coast-Section at Malahide, Co. Dublin. By Lous B. SmyrH, B.a., so.B. (Plates I, II.) (February, 1920.) 1s. . The Application. of the Food-Unit Method to the Fattening of Cattle. By James Witson, m.a., B.sc. (Plates III, 1V.) (February, 1920.) 1s. The Holothurioidea of the Coasts of Ireland. By Anne L. Massy. (April, 1920.) 1s. . Photosynthesis and the Electronic Theory. By Henry H. Dixon, so.p., F.t.s., and Horace H. Poor, sc.p. (March, 1920.) 1s. . Note on the Decay of Magnetism in Bar Magnets. By Witu1am Brown, s.sc. (March, 1920.) Gd. . On the Inhibition of Invertase in the Sap of Galanthus nivalis. By 'I'. G. Mason, m.a., sc.B. (April, 1920.) 1s. . The Change in the Rigidity of Nickel Wire with Magnetic Fields. By Wittiam Brown, B.sc., and Parriok O’CatnaGuan, A.B.0.80.1.,A.1.c. (August, 1920.) 6d. . Some Derivatives of Nitrotoluidine (4-nitro-2-amido-1-methyl-benzene). By A. G. G. LronarD, A.R.C.SC.1., B.SC., PH.D., and Agnes Browne, A.R8.¢.SC.1., B.sc. (August, 1920.) 6d. An Investigation into the Causes of the Self-Ignition of Hther-Air Mixtures. By the late Professor J. A. McCretuanp, p.sc., F.R.s., and Rey. H. V. Ginu, s.J., D.s.0., M.c., w.a., University College, Dublin. (August, 1920.) 6d. The Influence of Electrolytic Dissociation on the Distillation in Steam of the Volatile Fatty Acids. By JosnpH RetLiy, m.a., D.sc., F.R.C.SC.1., and Witrrep J. Hrcxinzorrom, B.sc. (October, 1920.) 6d. . Notes on some Applications of the Method of Distillation in Steam. By JosrpH ReILLY, M.A., D.SC., F.R.C.Sc.1., and WitrreD J. Hicxingorrom, B.so. (October, 1920.) - 6d. . The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water. By.W. HE. Aprnry, D.sc., a.R.c.sc.1., F.1.c., and H. G. BEcKER, A.R.C.SC.1., a.1.c. (September, 1920.) 6d. A Determination, by means of a Differential Calorimeter, of the Heat produced during the Inversion of Sucrose. By Henry H. Dixon, so.p., F.R.s., and Nicest G. Batt, B.a. (December, 1920.) 6d. The Measurement of very Short Time Intervals by the Condenser-Charging Method. By Joun J. Dowie, u.a., F.1nsT.P., and Donat Donnety, m.sc. (In conjunction with the late Prof. J. A. McCietianp, F.r.s.) (February, 1921.) 6d. : A Vibrating-Flame Rectifier for High-Tension Currents. By Joann J. Dowxine, M.A., F.inst.P., and J. T. Harris, B.sc. (February, 1921.) 6d. A Sensitive Valve Method for the Measurement of Capacity, with some Important Applications. By Jonn J. Dowtine, u.a., F.1nst.p. (February, 1921.) 64. A Direct Reading Ultra-Micrometer. By Joun J. Dowtine, m.a., u.R.1.a., F.nsT.P. (March, 1921.) 6d. Studies in the Physiography and Glacial Geology of Southern Patagonia. By E.G. Fenton. (Plates V, VI, and VII.) (March, 1921.) 4s. 6d. Award of the Boyle Medal to Groren H. Prrnypripen, B.sc., PH.D., 1921. (March, 1921). 1s. SCIENTIFIC PROCEEDINGS—continued. 21. The Concentration and Purification of Alcoholic Fermentation Liquors. Part I.—The Distillation in Steam of certain Alcohols. By JosrpuH Rrity, M.A., D.SC., M.R.I.A., F.I.c.. and WuitrrED J. HickINBOTTOM, M.SC., A.I.C. (August, 1921. ) 22. The “Browning” and ‘‘ Stem-Break’’ Disease of Cultivated Flax (Linum usitatissimum), caused by Polyspora lini n. gen. et sp. By H. A. Larrnrry, A.R.c.sc.1., Assistant in the Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. [Communicated by Dr. George H. Pethybridge, B.sc.] (Plates VIII-X.) (August, 1921.) 23. A New Principle in Blowpipe Construction. By H. G. Bucxur, a.z.¢.8c.1., a.1.c., Demonstrator in the Royal College of Science, Dublin. (August, 1921.) 24. Uncharged Nuclei produced in Moist Air by Ultra-Violet Light and other Sources. By the late Prof. J. A. McCreuanp, F.x.s., and J. J. M‘Henry, u.sc., University College, Dublin. (August, 1921.) _ [Nos. 21 to 24, price 7s.] 25. Biological Studies of Aphis rumicis LL. A.—Appearance of Winged Forms. B.—Appearance of Sexual Forms. By J. Davipson, p.sc. From the Entomological Department, Institute of Plant Pathology, Rothamsted Experimental Station, Harpenden, Herts. [Communicated by Professor G. H. Carpenter.] (August, 1921.) 26. The Occurrence of Dewalquea in the Coal-Bore at Washing Bay. By T. JoHNSON, D.sc., F.L.S., Professor of Botany, Royal College of Science for Treland, and Jans G. Ginmors, B.sc. (Plates XI, XII.) (August, 1921.) 27. A Simple Form of Apparatus for observing the Rate of Reaction between Gases and Liquids, and its use in determining the Rate of Solution of Oxygen by Water under different conditions of Mixing. By H.G. Brcxzr, A.R.C.SC.1., A.I.c., Demonstrator in Chemistry, Royal College of Science, Dublin. (August, 1921.) 28. The Occurrence of a Sequoia at Washing Bay. By T. Jounson, D.sc., F.L.S., Professor of Botany, Royal College of Science for Ireland, and Janz G. Gitmorz, B.sc. (Plates XIII, XIV.) (August, 1921.) 29. The Sources of Infection of Potato Tubers with the Blight Fungus, Phytophthora infestans. By Paut A. Murpny, B.a., a.R.0.8c.1., Assistant in the Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (August, 1921.) [ Nos. 25 to 29, price 6s. 6d.] 80.—Some Factors affecting the Hydrogen Ion Concentration of the Soil and its Relation to Plant Distribution. By W. R. G. ArKins, 0.B.E., SG.D., F.I.C. [Communicated by Professor H. H. Dixon, sc.p., ¥.x.s. (February, 1922.) 31.—The Hydrogen Ion Concentration of Plant Cells. ne W.R. G. Arxins, 0.B.E., sc.D., F.1.c. [Communicated by Professor H. H. Dixon, soap., F.x.s.| (February, 1922.) 32.—-Note om the Occurrence of the Finger and Toe Disease of Turnips in Relation to the Hydrogen Ion Concentration of the Soil. By W. R. G. Arxins, 0.3.5., sc.D., F.1.c. [Communicated by Professor H. H. a SC.D., F.RB.S. | (February, 1922.) 38.—Photosynthesis and the Electronic Theory (II). By Henry H. Drxon, se.v., F.R.s., University Professor of Botany in Trinity College, Dublin; and Nicet G. Bauu, m.a., Assistant to the University Professor of Botany. Trinity College, Dublin. (February, 1922.) 34.—The Bionomics of the Conidia of Phytophthora infestans (Mont.) de Bary. By Pavz A. Murpuy, B.A., a.R.c.sc.1., Assistant in the Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (February, 1922.) [Nos. 30 to 34, price 7s. | DUHLIN : PRINTED A’ THE UNIVERSITY PRESS BY PONSONKY ANID GIHBS. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XVI. (N.8.), Nos. 35-39. APRIL, 1922. 35—ON THE DISTRIBUTION OF . CTIVITY IN RADIUM THERAPY UNDER DIFFERENT CON! TIONS OF SCREENING. By H. H. Poots, M.A., Sc.D., Chief Scientific Officer, Royal Dublin Society. 36.—THE INFLUENCE OF FEEDING ON MILK FAT. By E. J. SHeEny, E.R.C.Sce.i., B.Sc. (Bio-Chemical Laboratory, D.A.T.I.) 37.—ON A NEW METHOD OF GAUGING THE DISCHARGE OF RIVERS. By J. Joy, Sc.D, Fo°S., F.T.C.D. 38—ON A VARIETY OF PINITE O: (URRING AT BALLYCORUS, CO. DUBLIN. By Louis B. yTH, M.A., Sco.B., Lecturer in Palaeontology in the University Dublin. 59.—ON THE LIFE-HISTORY AND BIONOMICS OF THE FLAX FLEA-BEETLE (LONGITARSUS PARVULUS, PAYK.), WITH DESCRIPTIONS OF THE HITHERTO UNKNOWN LARVAL AND PUPAL STAGES. By J. G. Raynenarrt, A.R.C.Sc.1L, D.L.C., N.D.A., Entomologist in the Seeds and Plant Disease Division, Depart- ment of Agriculture and Technical Instruction for Ireland. (Plates XV-XIX.) | Authors alone gre responsible for all opinions expressed in their Commdnications. Hen ay it DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY! 17) LEINSLER HOUSE, DUBLIN. WILLIAMS AND NORGATEH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1922. Price Eight Shillings. Royal MBublin Society. an FOUNDED, A.D. 17381. INCORPORATED, 1749. eae SCIENTIFIC MEETINGS. Tux Scientific Meetings of the Society are usually held at 4.15 p.m. on the fourth Tuesday of every month of the Session (November to June). The attention of authors is directed to the following extracts from the repula- tions governing the presentation of communications at Scientific Meetings of the Society, and the publication of papers in the Scientific or Hconomic Proceedings :— “An author who desires to present a communication at a Scientific Meeting of the Society shall furnish to the Chief Executive Officer for Science a fortnight previous to that meeting the title and an abstract (not exceeding 200 words) of his communication, together with an estimate of the time required for presentation. ‘¢ An author who desires to have a communication published forthwith as a paper in the Proceedings of the Society shall furnish to the Chief Executive Officer for Science the full text and necessary illustrations at least a week before the meeting of the Committee of Science at which he desires publication to be considered. Meetings of the Committee of Science are usually held on the second Tuesday of each month from November to June inclusive. “© A paper may be published either in the Scientific Proceedings or the Hconomice Proceedings of the Royal Dublin Society, as directed by the Committee of Science. The copyright of papers so published shall be the preperty of the Society. “ Fifty copies of each paper shall be supplied free of cost to the Author _ or Authors. ‘« Additional copies may be obtained at Cost Price if applied for by the Author when returning the proofs.” Authors are requested to apply to the Chief Hixecutive Officer for Science for further information concerning the preparation of papers and illustrations for [ 47 4 XXXV. ON THE DISTRIBUTION OF ACTIVITY IN RADIUM THERAPY UNDER DIFFERENT CONDITIONS OF SCREENING. By H. H. POOLE, M.A., Sc.D., Chief Scientific Officer, Royal Dublin Society. [Read January 24. Published Aprin 13, 1922.] Introduction. A GREAT deal of work has been done by various experimenters on the passage of a, 3, and y rays from radioactive substances through various kinds of matter, and on the secondary radiations set up.1 Some of the results so obtained are rather inaccessible to medical workers in this country, and, moreover, their application to specific problems is generally rendered difficult by the complexity of the radiations, and by the necessity of combining the effect of absorption with the geometvrical effect of distance. An attempt has accordingly been made to reduce the user’s problem to its simplest form by carrying out tests on the screening effects of various materials for the combined (3 and y radiation emitted by an emanation tube at a fixed distance, and combining the results so obtained with the geometrical law of distance, so as to calculate the variation of activity with distance from the source in some typical cases.” It seems probable that we may take the ionisation produced in the tissues as a measure of the physiological action. This ionisation may be due to:—(1) a rays, (2) primary #3 rays, (3) secondary (3 rays produced by 1 For a review of recent work on the absorption of y rays see ‘‘ Radium,” October, 1921. *The validity of this method has been questioned by Kroenig and Freidrich [‘‘ Strahlentherapie,” xi, p. 20, 1920]. Their direct measurements of y radiation at a depth of 10 cms. in water gave results about twice as great as they obtained by calculation. They attributed the discrepancy to scattered radiation. On the other hand, Sievert [‘ Acta Radiologica,” i, p. 89, 1921] obtained results at smaller depths according excellently with theory. It would seem unlikely that, under the conditions prevalent in radium therapy, scattered radiation could cause a large error for distances up to 10 ems. SOIENT. PROC. R.D.S., VOL XVI, NO. XXXV. 35 468 Scientific Proceedings, Royal Dublin Society. the action of the y rays on any screen used, and (4) secondary # rays produced in the tissues by the passage of the y rays. The ayvays are completely stopped by less than 0-1 mm. of flesh, or about 0:03 mm. of glass. This being much less than the usual thickness of wall of an emanation tube, they produce no effect with the common modes of treatment. If, however, emanation solution be injected, the effect of the a rays must far outweigh that of the others in all regions to which the solution finds access, as their relatively large energy and great absorbability cause, within their range, an intensity of ionisation about 100 times that due to the (3 rays. The primary /3 rays, as may be inferred from the curves shown in fig. 2, are completely absorbed by about 1:5 ems. of flesh, or 1°5 mm. of brass. Owing to this comparatively rapid absorption, they produce relatively intense ionisation, that due to unscreened (3 rays being about 100 times that due indirectly to the y rays from the same source. Secondary {3 rays, produced by the passage of y rays through ordinary matter, are similar in ionising properties to the primary (3 rays emitted by the radioactive source. They are, however, considerably less penetrating than the hardest of the primary (rays. The velocity, and hence the penetrating power of secondary ( rays, increases with the hardness of the y radiation to which they are due, so that the secondary 6 rays produced by y rays are much more penetrating than those due to X rays. If a metal screen is interposed between the source and the flesh, the surface of the latter is exposed to the secondary $8 radiation from the screen, but is deficient in secondary radiation from the flesh as compared with a point deeper down. The effect at a point in the tissue, then, varies with depth owing to the combined effect of the following causes :— (1) Absorption of primary (3 rays in the screen and the tissue. (2) Absorption of secondary (3 rays due to the screen. (3) Initial increase with depth of the secondary (3 rays due to the tissue. (4) Absorption of y rays, and consequent reduction of secondary (3 rays. (5) Geometrical effect of distance. The joint result of effects (1) to (4) has been estimated by experiments with an electroscope at a fixed distance from the source, and approximate calculations then made to include the effect of distance in some typical cases. Apparatus. The electroscope used is shown in section in fig. 1, which also shows the emanation tube mounted on a wooden block at a fixed distance below Poote—On the Distribution of Activity in Radium Therapy. 469 the electroscope. It is made of brass, with a lead base 4 mm. thick, in the centre of which is a circular window, 2°7 cms. in diameter, covered with a sheet of gelatine, 0:02 mm. thick, off an unexposed but “ fixed ” photographic plate. The electroscope is lined with leather about 2°5 mm. thick, impreg- nated with gelatine containing a little mercuric chloride to increase its electrical conductivity. This lining ensures that the secondary radiation from the walls of the electroscope resembles as nearly as possible that in animal tissue, so that the rate of motion of the gold leaf is a fair measure of the total ionisation that would occur in the tissue under the same conditions of radiation. The ea electroscope was always charged to about 250 volts by means of a Tucker hygroscopic battery before each test. The approximate constancy of the initial voltage rendered the motion of the leaf more regular. The screens to be tested were attached with a little plasticine to the under side of the lead base, Cor, I YL CLL Li . . . N SQVewl the wooden block carrying the radioactive source {SS ESN ° : ( N being so arranged that it could be removed and WN N : he Date : # N replaced in the same position. This source consisted : N LLL Leib Ll eae of one or more thin glass tubes about 15 cm. long by 0S mm. diameter, containing emanation, placed in a shallow groove in the block. The activity of the source used varied with the thickness of the Ns screen so as to maintain the rate of motion of the ~~ gold leaf within reasonable limits. The screens were placed directly in contact with the electroscope base in order that as large a percentage as possible of the secondary (9 radiation, emerging at various Fie. 1. angles from the upper surface of the screen, should enter the electroscope. The source was placed only 1:85 cms. below the base in order that the divergence of the y rays passing through the parts of the base surrounding the window might reduce the error due to these rays to a negligible quantity. An allowance was made for the gelatine film in estimating the thickness of the various screens. Experiments with a second film indicated that the absorption of unscreened radiation by such a film amounted to about 6 per cent., so this allowance was made in estimating the effect of a bare emanation tube. In all cases the thickness of the screens used was found by weight. LLL g 3 E2 470 Seventifie Proceedings, Royal Dublin Seciety. Secondary Radiation. The secondary radiation from various substances was compared by the usual method of using a compound screen of two or more materials thick enough to stop all the primary (6 rays, and observing how the ionisation current varied with the material which was: next the window. It is of course necessary that the individual screens should be thick enough to emit their full secondary radiation. In this way the combined effect due to both y rays and secondary 8 rays from the screen is directly measured. No appreciable difference could be detected between copper, zinc, brass, iron, tin, or silver. Taking the ionisation current due to y rays and secondary 3 rays from any of these screens as 100, that with an aluminium screen was 105, with leather 3 mm. thick 110, and with lead 112. These results are in accord with those of previous investigators, who have found that the “emergence” secondary radiation is a minimum for an atomic weight of about 80.1 They show that it is best to use some such metal as brass or iron, except where very light screening is required, where aluminium is the most convenient. In this case the presence of copious primary 8 radiation renders the secondary (3 radiation of small importance. Experiments were also made with a brass screen about 3 mm. thick, covered on top with different thicknesses of material of low atomic weight. For small thicknesses visiting-cards (0-040 gram per sq. cm.) were found the most suitable, six of them being almost exactly equivalent to the same mass of leather. For greater thicknesses sheets of leather were employed. It was found that adding successive cards increased the ionisation until the total thickness of card or leather amounted to about 0:3 gram per cm. [equivalent to about 3 mm. of tissue]. Further increase caused a slight fall, due to absorption of y rays. This apparently suggests that the practice of placing a material such as chamois between the metal screen and the skin is unwise; but, when allowance is made for the effect of distance, it will be shown later that this is not so. Practical experience confirms this latter view, as it is found that the presence of a sheet of chamois between the screen and the skin reduces the burning of the latter. Stopping, Power of various Substances. The results of tests on various materials are shown by the curves in fig. 2, where the ordinates represent activity, and the abscissae mass per unit area of screen. 1H.g. Hackett, Scient. Trans. R.D.S., ix, Series 1. 9, 1909, PooLtE—On the Distribution of Activity in Radium Therapy. 471 Three curves are shown for card and leather, whose absorption per unit mass is almost identical. In curve A the activity of the bare tube is taken as 100, and each unit on the horizontal axis represents 1 gram per sq. cm. Oo 1 é 3 GRAMS PER SQ 07 [i.e. 1 em. of animal tissue approximately]. The initial absorption of (3 rays is better shown on curve B, in which the horizontal scale is 10 times as open, ie. for this particular curve each unit represents 0:1 gram per sq. cm. [1 mm 472 Scientific Proceedings, Royal Dublin Society. of tissue]. In curve C the vertical scale is 10 times as open, i.e. the activity of the bare tube is now 1000. The horizontal scale is normal. The use of these varying scales is rendered necessary by the great range of activities encountered with light screens. Curves are also shown for aluminium, brass, and lead. In each case the scales are the same as for curve C, the activity of the bare tube being 1000, and the horizontal unit one gram per sq. cm. In using these curves the density of animal tissue may be taken as 1:0, that of aluminium as 2°7, brass 8:0, and lead 11:3. Thus 1 mm. of brass corresponds to 0°8 gram per sq. cm., and so on. The following points may be noted :— (1) The stopping power of a given mass of the screen for (3 rays increases with its atomic weight. (2) The primary 3 rays are completely absorbed by 14 mm. of tissue, 5 mm. of aluminium, 1°5 mm. of brass, or 1 mm. of lead. These figures are obtained by dividing the abscissae for which the various curves become approximately horizontal by the densities of the corresponding screens. (3) Owing to differences in secondary radiation, the total ionisation is not identical for various screens of just sufficient thickness to stop the primary $ rays in each case. The peculiar crossing of the brass and lead curves is to be attributed to this cause, as the latter is raised by the large secondary radiation from the lead. It subsequently falls below the former again, owing to the greater absorption of y rays by lead. Greater thicknesses of leather could not be tested with the arrangement employed; but this is not a serious disadvantage, as in the calculations which foliow the comparatively unimportant absorption of y rays at such depths can be estimated with sufficient accuracy from the results obtained for brass, or taken from the figures obtained by other workers. Tests were also made with an emanation tube enclosed in a standard serum needle 1:2 mm. external diameter, and various card and leather screens. These showed that the stopping power of the needle (apparently about 0-1 mm. thick) was very nearly equivalent to 1 mm. of tissue. This is used below. Lffect of Distance. Capillary in Needle. The effect in the tissues at various distances from a needle containing an emanation tube has been calculated, and is shown in column A of the table on p. 475, in which a@ represents the distance from the centre of the needle. In making this calculation the emanation, instead of being distri- buted throughout a tube 15 mm. long by 0:6 mm. internal diameter, is Poote—On the Distribution of Activity in Radium Therapy. 473 assumed to be concentrated at 15 axial points one millimetre apart. ‘The stopping power of the wall of the serum needle is assumed to be equivalent to 1 mm. of flesh for normal rays, and proportionally more for oblique rays. For the latter the extra thickness of glass traversed must also be allowed for. Weighing showed that 0-1 mm. is a fair figure to assume for the thickness of wall of a typical emanation tube, as issued by the Royal Dublin Society Radium Institute. This is equivalent to 0°25 mm. of flesh. The outside radius of the serum needle being 0°6 mm., the total absorption between a point O on the axis and a point P at a radial distance a from O in the tissue 1s equivalent to that of @+ 1:25 - 0:6, ie. a+ 065 mm. of flesh. This is equivalent to the bare tube plus a + 0-4 mm. of flesh. For another axial point O’ distant 6 mm. from O we have distance a OP = fa? + & and “equivalent” distance = 7 be + 0°65) — 0:25. We look up the corresponding transmission factor from the “leather” curve, taking the unscreened activity as 1000, and multiply this by ay a where m is the number of millicuries per mm. of the tube. Summing the results for the 15 axial points, we obtain the activity for any given point P. The units are such that the activity due to one millicurie at one millimetre distance, with no other screening than the normal screening of the emanation tube wall, is 1000. The figures in the table are calculated for points opposite the centre of the capillary, so that the axial points considered, with the exception of the central one, occurred in pairs, thus simplifying the arithmetic. LHifect of Distance. Thick Surface Applicator. A type of applicator which is considerably used in Dublin consists of a rectangular box made of brass, 3 mm. thick, the internal dimensions being 30 x 20 x 06 mm. To the inner surface of this a number of emanation tubes are fixed, with a little wax so as to give as uniform a distribution of activity as possible over the available area of 6 sq. ems. As the diameter of these tubes is about 0°8 mm., we may (making a small allowance for the wax) assume that the centre of the tube is 3°5 mm. from the nearer outer face of the box, and 8°5 mm. from the further outer face. In estimating the activity at various depths due to such an applicator in contact with the skin the following approximations are made to simplify the calculation :— : (1) The box is assumed to be circular, of such a diameter so as to have the same area, and the caleulations made for axial points only. 474 Scientifie Proceedings, Royal Dublin Society. (2) The emanation is assumed to be confined to a plane parallel to the face of the box in contact with the skin. (3) The activity of a parallel beam of rays which penetrate the box is assumed to be a linear function of the extra distance traversed. This will not introduce a large error for distances up to 10 cms. of tissue if a suitable absorption coefficient is chosen. (4) Scattering is neglected [as in previous calculation]. Let 7 be the radius of the box, ¢ its thickness, 0 the distance from the inner face of the box to the radioactive layer, and Z the surface-density in millicuries per sq. mm.; then obviously the activity at an axial point @ mm. below the surface is equal to (i [K - (kit + ko) sec 0] 2a Tax da J (6+¢+ a)?sec’O ? ) where x tan @ = ————_.,, b+t+a and K, ,, and &, are constants defining the absorption in brass and tissue. Hence Activity = 2rJ [ [K - (yt + ka) sec 0] tan 0 d0 0 2nT[K logesec p - (hit + kya) (sec - 1)], where fp Cs een a This expression can be evaluated rapidly and with sufficient accuracy by means of a slide-rule, provided that ¢—the angle subtended by the radius of the radioactive dise at the given point—is not too small. For small values of @ it is best to substitute tan’? tan*¢ ite payee for log. sec $, and tan? tan? ; ons — for (sec @ - 1). For very small values the expression reduces to wl tan? p [K - (Ait + kea)]. From inspection of the curves in fig. 2, and consideration of the known rate of absorption of the complex y rays from AaB and RaC, we find K = 10°56, k, = 032, and *& = 0:04 (assuming the mass coefficient of PooLe— On the Distribution of Activity in Radium Therapy. 475 absorption of brass and animal tissue for y rays to be approximately the same). In all cases considered ¢ = 3:0 mm. The figures so obtained would represent the activity if the secondary radiation in the flesh were identical with that on the skin in contact with a brass radiator, as K was determined from the “brass” curve. In the interior the increased secondary radiation due to the tissue raises the total activity by about 10 per cent., so in the results shown below this correction has been applied to all depths exceeding 3mm. Appropriate corrections are applied for smaller depths, no correction being required at the surface. Three sets cf figures have been worked out and are shown in columns B, C, and D of the following table, the corresponding conditions being explained below :— a A B C D a A B CEs aed) mm. i || mm. | | 0 — 1-24 0°62 0°0938 16 0-46 | 0°225 0°15 0°063 0-5 —_ 1:18 0-60 0-098 16 0°41 0-205 | Old | 0-051 0-6 620 — — = 17 0°37 0-19 0°18 0:049 1 227 1-12 0°58 0-098 18 0°33 0°175 | 0°12 0-047 2 58 0°99 0°525 | 0:091 19 0°30 0-16 | 0-115 | 0°045 3 23 0-87 0:475 | 0:088 20 0:27 0-16 0-105 | 0-044 4 11-6 0°76 0:425 | 0-084 25 0°17 0°105 0:076 | 0:037 ) 6-7 0°66 0°375 | 0-080 30 0-117 0-075 0-058 | 0:031 | 6 4-2 0°58 0°34 0-077 35 0-084 0-057 |3 0°044 | 0026 | 7 2°8 0°52 0-305 | 0:074 40 0-064 0-044 | 0:036 | 0:022 | | 8 1:93 0:46 0:275 | 0-071 45 0-050 0:035 | 0-030 | 0-019 | 2 Weed 0405 0°25 0-068 50 0-039 0-027 0:023 | 0-°016 TOS ent 08 0°36 0°23 0:065 60 0-026 0:019 | 0°016 | 0°012 11 0°87 0°325 0°21 0-062 70 0-018 0-014 | 0-012 | 0-009 12 0-72 0-295 0-195 | 0:060 80 0-018 0-010 | 0:009 | 0:007 13 0°61 0:27 0-18 0-057 90 0-010 0-008 0-007 | 0-006 14 0°53 0-245 0°165 | 0:055 100 | 0:008 0-006 0-005 ADS Explanation of Table. a is the distance in millimetres of the point in the tissue at which the activity is given in the other columns. For the figures in column A (emanation needle) a is the distance from the axis of the needle; for the other columns, @ is the distance from the surface of the skin. 476 Scientific Proceedings, Royal Dublin Society. The activity due to a point-charge of one millicurie at a distance of oue millimetre, with no screening except the glass of the emanation tube, is taken as 1,000. A. Activity due to 10 me. in a tube 15 mm. long inside a standard serum needle. B. Activity due to 10 me. uniformly distributed over 6 sq. cm., the emanation tubes being attached to the bottom of the brass box, 3 mm. thick, which is placed in contact with the skin. C. Activity due to 10 me. uniformly distributed over 6 sq. cm., but with the emanation tubes attached to lid of the box, leaving an air space of 5 mm. between the tubes and the inner surface of the bottom of the box. D. Activity due to 10 me. uniformly distributed over 100 sq. em., the tubes being placed as in C. General Conelusions. It is evident that the conditions specified under D could not be con- veniently realized with a single application; but a similar (though not an identical) effect can be obtained by successive applications of the brass box, covering all parts of the required area in succession for equivalent times. The total dose of millicurie hours must then be used in arriving at the resultant action. For example, if we take the figures in the table as measuring the effect produced by a ten-millicurie-hour dose, the effect of a total dose of 200-millicurie-hours uniformly distributed over 100 sq. cms. would be 20 times the corresponding figure in column D. The following points may be noted from the above figures :— (1) The very great activity of emanation in needles at moderate distances. (2) The great reduction in surface effect caused by increasing the distance of the capillaries from the skin. (3) The still greater reduction in surface effect due to increase of area. The effect of interposing material such as chamois between the applicator and the skin is nearly, but not quite, obtained by measuring “a” from the outer surface of the chamois, i.e. by treating the chamois as the outer layers of tissue in the above table. Thus, with conditions B above and 1 mm. of chamois interposed, the activity on the surface of the skin would be 1:12 instead of 1°24. This shows the effect of chamois in acting as a distance piece. This is not quite accurate, as chamois, being much less dense than living animal tissue, the increase in secondary radiation is much slower in it, so Poote—On the Distribution of Activity in Radium Therapy. 477 that the activity at the base of a sheet 1 mm. thick would be less than 1:12. It would appear that the more porous the material used between the metal and the skin the less should be the burning effect upon the latter. In the case of treatment with several parallel needles, each containing a tube 15 mm. long, the distribution of activity in a perpendicular plane through their centres can be found by summing the effects of the separate needles for any required point. Where the area of irradiation is increased by the use of long needles, which are partially withdrawn at intervals, so as tomove the emanation tube to a new position on the axis, the effect is similar to that of a long emanation tube, and could be calculated by the same method as that used above for the short tube. E ae | XXXVI. THE INFLUENCE OF FEEDING ON MILK FAT. By E. J. SHEEHY, F.R.C.Sc.1., B.Sc., M.R.LA. [Bio-Chemical Laboratory, D.A.T.1.] [Read Frsruary 28. Published Aprit 13, 1922.) EACcH individual species of mammal produces milk of composition peculiar to its own group, but among domesticated animals the proportion of some cf the constituents, especially that of fat, to the total milk yield varies con- siderably in the different varieties of each species, in the different strains inside a variety, and, further, even in the individuals of any single strain. The variation in the percentage of fat in the milk of cows and goats, for instance, is practically due to the selection exercised by man; indeed it has long been recognised that selection is the principal method available to the stock-breeder for raising milk fat percentage. The influence of feeding on milk fat has been the subject of frequent investigation, but very little precise information on the subject is available. Speir [1894, 6 and 7] compared a large number of rations, all liberal in quantity, with one another for the purpose of milk and butter fat production in the dairy cow, and came to the general conclusion that the criterion of efficiency in a ration is its total dry matter. But he was of opinion that some foods tend to increase the milk yield, while others tend to increase the milk fat principally. Fresh grass and brewers’ grains, he states, tend to increase the milk yield at the expense of the percentage of butter fat; on the other hand, vetches and decorticated cotton cake tend to increase the percentage of fat at the expense of milk yield. It is significant that both of these foods contain a high proportion of protein; in fact, Speir concludes that rations with a very high albuminoid ratio seem to have a depressing effect on the milk yield. On the other hand, Crowther [1906], reviewing the work done in Great Britain for some years previously, makes the following statement: “Such changes as may possibly be effected in the quantitative composition of milk by change of food are only very slight, provided, of course, the rations prior and subsequent to the change are suitable in nature and in quantity.” Wing and Foord [1904] record an attempt to increase the fat in milk by means of liberal feeding, and state that in a herd of poorly SHeEny—The Influence of Feeding on Milk Fat. 479 fed cows the subsequent feeding of an abundant ration resulted in an increase in the percentage of fat as well as an increase in the total milk and total fat. The experiment continued over a number of years. Yet Rihle [1919] and Pritzker [1919] find, from a large number of analyses made during the period 1914-18, no evidence to show that food deficiency causes a decrease in the fat content of milk. Data collected by the Dairy Section of the Department of Agriculture and ‘I’echnical Instruction for Ireland does not lend support to this view, but is rather in agreement with the work of Wing and Foord. Referring to the effect of water in the food on the composition of milk, Turner, Shaw, Norton, and Wright [1916] state that the water content of a ration has no effect on the composition of milk yielded therefrom; and Berry [1921] comes to the same conclusion. It is evident that there is no unanimity of opinion among previous investigators as to the effect of feeding on milk fat. Much of the work referred to was done on good condition cattle, con- tinuously getting abundant supplies of food, so that the effect of the changes in diet was difficult to detect. Besides, changes were made by completely substituting one food for another, so that usually more than one food con- stituent was involved. For instance, the substitution of brewers’ grains for cotton cake involves a change in the quantity of carbohydrates and fats, and in the quantity and quality of the proteins. Thus more than one variable was introduced, and there was not a clear comparison of the effect of one food ingredient with that of another. Realizing the necessity for conducting such an investigation under conditions approximating to laboratory control, it was decided in this investigation to substitute goats for cows. To the diet of these animals the individual food constituents could be added at comparatively small cost, while at the same time sufficient milk is yielded to afford a consider- able margin for fluctuation resulting from experimental changes in conditions, In the goat, as in the cow, the milk yield is highest at the beginning of lactation, and it declines gradually towards the end. The lactation period, in the case of the old Irish breed, usually begins in April and terminates about December. As in the cow also the percentage of fat in the milk tends to increase as lactation advances, and the weight of the animal’s body similarly increases towards the termination of the milk flow, especially with liberal feeding. The general procedure in this investigation was first to observe the quantity of food which just maintained the animal’s weight and at the same time yielded a constant flow of milk, and then test the effect of adding to this “milk and body maintenance” ration different quantities of individual food constituents. ‘Then, when the effect of an addition to the ration was determined, the effect of the withdrawal of that addition was subsequently investigated. Extra food assimilated above the “milk and 480 Scientific Proceedings, Royal Dublin Society. body maintenance” level goes to increase either the milk, or some of the constituents thereof, or the weight of the animal, or all three. Consequently — right through the period of investigation records were kept of milk, milk fat, and the animal’s weight. The milk was weighed and the fat percentage determined daily, and the animals were regularly weighed once a week. Figs. 1, 2,and 3 record the diet and changes therein of three experi- mental goats during the period of investigation, and give a graphical representation of the results from the experiment. The results obtained from this investigation are interpreted as follows :— Goar 1. 1. Reduction in daily ration (change from ration A to B) produces— (a) No change in milk yield ; as a matter of fact, the already increas- ing yleld continues to increase for a short time. (b) Reduction in daily per cent. fat of 1:25 (4°75 to 3:5). (c) Reduction in daily total fat of 0-015 Ib. (0°130 to 0°115). 2. Addition of 0°47 lbs. soya fat to daily ration (change from ration B to C) produces— (a) No change in milk yield. (0) Increase in daily per cent. fat of 0°75 (3°5 to 4:25). (c) Increase in daily total tat of 0°035 lb. (0:115 to 0-150). 2,. Withdrawal of roots from ration (change from ration C to D) pro- duces— (a) Considerable reduction in milk yield. (0) _ 5 daily per cent. fat. (¢) » ms daily total fat. 3. Addition of 0°5 lbs. soya fat to daily ration (change from ration E to F) produces— (a) No change in milk yield. (0) Increase in daily per cent. fat of 1°75 (3°75 to 5:5). (c) Increase in daily total fat of 0:05 Ib. (0:09 to 0:14). 4, Further addition of 0:5 lb. soya fat to daily ration (change from ration F to G) produces— (a) No change in milk yield. (6) Increase in daily per cent. fat of 1:0 (5°5 to 6:5). (c) Increase in daily total fat of 0-025 lb. (0:140 to 0:165). 5. Reduction in daily ration (change from ration G to H) produces— (a) No change in milk yield. (6) Reduction in daily per cent. fat of 2:0 (65 to 4:5). (c) Reduction in daily total fat of 0-075 lb. (0°165 to 0-090). Fig. 1—Goar 1. AY AUSUST Ape g 8 28 17 7 ae ‘oar DAILY MILK. DAILY TOTAL FAT. pl\ WEISHT of ANIMAL A B x C D y E F G H Dainty Ratton. A—(37 days). Hay ad lib. Mangels, 1 stone, Y—Irregular, owing to failure of supplies. Crushed oats, bran, and dried grains, 14 Ibs: E—(21 days). Hay, 1b. Turnips, 10 Ibs. Oats, B—(21 days). Hay, 1 lb. Mangels 10lbs. Skim- 0°83 lb. milk powder, 1°28 Ibs. [Reduction on pre- vious ration. ] F—(20 days). Heys 1 lb. arnipss 10 lbs. Oats, ea ee ee core 0°83 1b. Fat (soya), 0°56 lb. [0°5 lb. soya fat X—(14 days). (Transition to new ration.) added to previous ration. C—(24 days). Hay,1lb. Mangels, 10 lbs. Skim- 5 milk powder, 1°28 lbs. Fat (soya), 0°47 Ib. G—(23 days). Hay, 1 lb. Turnips, 10 lbs. Oats, [0°47 lb. soya fat added to previous ration. ] aaa Fat (soya), 1°0 lb. [0°6 lb. soya fat D—(11 days). Hay,11lb. Skim-milk powder, 1°28 gd Ge ALO bTeviousiraltione|| Ibs, Fat (soya), 0°47 lb. | Withdrawal of H—(17 days). Hay, 1]b. Turnips, 10 lbs. Oats, roots from ration. | 0°83 lb. [Reduction on previous ration.] 482 Scientific Proceedings, Royal Dublin Society. Goat 2. 1, Reduction in daily ration (change from ration A to B) produces— (a) No change in milk yield. : (6) Reduction in daily per cent. fat ef 0°25 (3°75 to 3:5). (c) Reduction in daily total fat of 0:015 lbs. (0°135 to 0-120). 2. Addition of 1:06 lbs. wheaten starch to daily ration (change from ration B to C) produces— (a) No change in milk yield. (6) Increase in daily per cent. fat of 0°75 lbs. (3°S to 4:25). (c) Increase in daily total fat of 0-03 lbs. (0°12 to 0°15). 3. Addition of 1:13 lbs. wheaten starch to daily ration (change from ration E to F) produces— (a) No change in milk yield. (>) Increase in daily per cent. fat of 0°75 (5:0 to 5:75). (c) Increase in daily total fat of 0:02 Ibs. (0°125 to 0:145). 3,, Substitution of 0°5 Ibs. fat for 1:13 lbs. starch (change from ration F to F,) produces— (a) No change in milk yield. The fact that this increase is of (>) Increase in per cent. fat. short duration, and that the yields (c) 2 fonalerats \tend to return to original level, dis- ; counts this result. 4, Substitution of 1:13 lbs. casein for 0°5 lbs. fat, and hence for 1:13 lbs. starch (change from ration F, to G), produces— (a) No change in milk yield as compared with ration F. Oe 3 daily per cent. fat ., 55 M 3 @) ~» vA daily total fat e %0 » » 5. Reduction in daily ration (change from ration G to H) produces— (a) No change in milk yield. (6) Reduction in daily per cent. fat of 0°5. (c) Reduction in daily total fat of 0:025 lbs. (0145 to 0°120). Fig. 2—Goar 2. DAILY SILK. SO i Ht hall OMY x FAT i i DAILY TOTAL MILK _ VY \A NI wey 0-05) WEISHT of ANIMAL STs 75 A Ln A 8 8 g g € £ 6 = ell g A B x Cc XG E BR F G H Dainty Rarton. A—(37 days). Hay ad lib. Mangels, 1 stone. E—(21 days). Hay, 1 lb. Turnips, 10 lbs. Oats, Crushed oats, bran, and dried grains, 13 lbs. 0°83 Ib. B—(21 days). Hay, 1 lb. Mangels, 10 Ibs. Skim- GU eta Woman te Te Hee ne aneat milk powder, 1'28 lbs. vious ration.] X—(14 days). [Reduction on pre- (Transition to new ration.) C—(24 days). Hay, 1 lb. Mangels, 10 lbs, Skim- milk powder, 1°28 lbs. Starch (wheaten), 1°06 Ibs. [1°06 lbs. starch added to previous ration,] Y—Irregular, owing to failure of supplies. SCIENT. PROC. R.D.S., VOL. XVI, NO. XXXVI. added to previous ration.] Fi—0'5 lb. fat replaces the starch of previous ration. G—(23 days). Hay, 1 1b Turnips, 10 lbs. Oats, 0°83 Ib. Casein, 1:13 lbs. [1°13 lbs. casein replaces the 0°5 lb. fat of previous feed, and the 1:13 lbs. starch of the ration pre- vious to that.] H—(17 days). Hay, 1 lb. Turnips, 10 lbs. 0°83 lb. [Reduction on previous diet.] BF Oats, 484 Scientific Proceedings, Royal Dublin Society. Goar 3 1. Reduction in daily ration {change from ration A to B) produces— (a) No change in milk yield. (b) Reduction in daily per cent. fat of 0°75 (4:0 to 3-25). (c) Reduction in daily total fat of 0:02 lb, (0:175 to 0°150). 2. Addition of 0:5 lb. butter fat to ration (change from ration B to C) produces— (a) No change in milk yield. (b) Increase in daily per cent. fat of 1:0 (5:25 to 4°25). (c) Increase in daily total fat of 0:04 lb. (0°15 to 0:19). 3. Addition of 1:13 lbs. casein to daily ration (change from ration E to F) produces— (a) No change in total milk yield. (b) Increase in daily per cent. fat of 0°75 (4:5 to 5°25). (c) Increase in daily total fat of 0:025 1b. (0-110 to 0°13). 4, Further addition of 0°5 lb. fat to daily ration (change from ration F to G) produces— (a) No change in milk. (6) Increase in daily per cent. fat of 1:0 (5:25 to 6:25). (c) Increase in daily total fat of 0-025 lb. (0°185 to 0-160). 5, Reduction in daily vation (change from ration G to H) producees— (a) No change in milk yield. (0) Reduction in daily per cent. fat of 1:5 (6°25 to 4°75). (c) Reduction in daily total fat. The foregoing results lead to the following conclusions :— 1. By considerably reducing the ration the total daily yield of fat is reduced. If, at the same time, there is little or no reduction in milk yield, the percentage of fat in the milk is lowered. ‘This occurs in the early stages of lactation, when the mammary gland is very active, and towards the end of the lactation, when the milk flow is already reduced to a minimum. Obviously, a fall in milk yield, accompanied by a fall in total fat, may leave the percentage of fat unchanged; and conceivably a considerable fall in milk occurring at the same time as a reduction in total fat may produce .an increase in the percentage of fat in the milk, Fig 3—Goar 3. APRIL AY JUNE Be" 8" 192 fe mom StS pe SE Ibs | ) TOTAL DAILY FAT. 015]! a NM Nee SOD a WEISHT oF ANIMAL 3 & 5 g A B x Cc Ys E F G H Dairy Ration. A—(37 days). Hay ad lib. Mangels, 1 stone. E—(21 days). Hay, 1 lb Turnips, 10 lbs. Oats, * Crushed oats, bran, and dried grains, 13 lbs. 0°83 Ib. B—(21 days). Hay, 1 lb. Mangels, 10 lbs. Skim- F—(20 days). Hay, 1 lb. Turnips, 10 Ibs. Oats, milk powder, 1°28 lbs. [Reduction on pre- 0°83 lb. Casein, 1°13 lbs. [1°14 lbs. casein vious ration.] added to previous ration] X—(14 days). (Transition to new ration.) G—(23 days). Hay, 1 lb. Turnips, 10 Ibs. Oats, C—(24 days). Hay, 1 1b. Mangels, 10 lbs. Whole 0:83 lb. Casein, 113 Ibs. Fat, 0°5 Ib. ( ae nGuaae) 175 Ibs. ['45 lb. butter fat (0°5 lb. fat added to previous ration.] added to previous ration.) H—(17 days). Hay, 1 lb. Turnips, 10 Ibs. Oats, Y—Irregular, owing ‘to failure of supplies. 0°83 lb. [Reduction on previous ration.) 3r2 486 Scientific Proceedings, Royal Dublin Society. 2. By adding butter fat, soya fat, wheaten starch, or casein to a “milk and body maintenance”’ ration the total fat in the daily milk is increased. In all the tests recorded in this investigation the milk flow remains unaffected by the addition of these single food substances; but the percentage of fat in the milk is increased. That is to say, the percentage of fat in the milk is raised, not only by the addition of fat to the food, but also by the addition of carbohydrate or protein. Obviously, if an increase in the milk flow occurs proportionate to the increase in the total fat, the percentage of fat in the milk is unchanged. 3. In the production of butter fat in milk, starch, casein, and fat (both butter and soya) replace one another; but the quantities which replace each other cannot yet be definitely stated. In the early part of lactation 0:45 Ib. of soya fat produced an increase of 0:035 lb. of fat in the milk of goat 1; 1:06 lb. starch (that is about two and a quarter times more starch than fat) produced an increase of 0:03 lb. milk fat in goat 2; and 0:45 lb. butter fat produced an increase of 0:04 lb. milk fat in goat 3. In the second half of lactation 0:5 lb. soya fat produced an increase in the total milk fat of 0:05 Ib. in goat 1, 1:13 1b. starch (about two and a quarter times the weight of fat), an increase of 0:02 1b. milk fat in goat 2, and 1:13 lb. casein an increase of 0:025 lb. milk fat in goat 3. In one case soya fat was just slightly better and in another it was twice as good as two and a quarter times its weight of starch. Butter fat was slightly better than soya fat for milk fat production, but the test was too limited to draw a general conclusion. It was thought that butter fat would be much superior for this purpose to soya fat, because of its content of vitamin A; but perhaps sufficient vitamin A was already supplied in the remainder of the ration to prevent a possible difference between the effects of the two fats becoming apparent. In goat 2 the substitution of fat for starch in ration F, had a peculiar result. Possibly this is temporary, and the duration of this test was too short to allow of a general conclusion. When, however, casein replaces the fat (ration G, goat 2), the total fat in the milk is similar to that produced by an equal weight of starch. Thus right through the experiment casein and starch replace one another in equal proportions. Evidently this can only hold where sufficient protein is already supplied in the “milk and body maintenance ”’ ration to contribute to the production of the protein moiety of the milk yielded, and otherwise maintain nitrogenous equilibrium. Further work remains to be done on the quantitative replacement of fats and carbohydrates in food for the purpose of milk fat production. 4, While an addition to the “milk and body maintenance” ration SHeviy—The Injluence of Feeding on Milk Fat. A87 increases the butter fat, a further similar addition does not increase the butter fat to the same extent. In goat 1, the addition of half a pound of fat to the ration (ration E to F) increased the milk fat by 0:05 1b.; a further addition of half a pound caused an increase of only 0:025 lb. milk fat. Even where the diet already contains a large proportion of protein (ration F, goat 3) the further addition of half a pound of fat to the diet gave an increase of only 0:025 lb. milk fat. The percentage of fat in milk can obviously be raised by extra feeding to a maximum; and, if fat percentage is the only consideration, it might be said that the maximum is reached according to the law of diminishing returns. But, as the ration is increased, body weight and total milk may be also increased, and a successively larger proportion of the food utilized for these purposes. Since, according to Nils Hansson [1916] and Wilson [1920], the law of diminishing returns does not hold in the production of beef and milk, neither is it likely to hold in the production of butter fat. Sunmary. In the case of a lactating goat— 1. A considerable reduction in the ration may cause a decrease in the percentage of fat in the milk. This occurs when the total fat yield is de- creased at the same time as the milk yield remains unchanged, or when the proportionate decrease in the total fat is greater than the decrease in milk. 2. The addition of a concentrate to a ration may cause an increase in the percentage of fat in milk. This occurs when the total fat yield is increased, while the milk yield remains unchanged, or when the proportionate increase of total fat is greater than the increase in mili. 3. In effecting an increase in the total andin the percentage of butter fat in milk, starch, fat (soya and butter), and protein can replace one another. After the necessary minimum of protein is supplied, starch and casein replace one another for that purpose in equal parts. 4. There is a maximum milk fat percentage, and, as this limit is approached, the extra food added to the ration produces a successively decreasing addition to the milk fat percentage, but at the same time a successively increasing addition to the body weight or milk yield, or both. 488 Scientific Proceedings, Royal Dublin Society. REFERENCES. Berry. Jr. Agric. Se, 11,78. [1921.] CrowIrHER. ——-——, 1,149. [1906 | Nits Hansson. Utfodringslara, page 487. [1916.] PrirzkEr. Schweiz. Apoth. Zte., 57, 167-9. [1919.] Ruuaie. Z. Nahr. Genussm., 38, 277-85. [1919.] Sperr. Trans. High. and Agric. Soc.; Scot., 6, fifth series, 83. [1894.] SS ———— , 8, —_— _,, 269. [1896.] 296. [1897.] TurRNER, SHAW, Norron, Wricur. Jr. Agric. Res., 6,167. [1919.] Witson. Roy. Dub. Soe. Sc. Proc., 16 (N.S.), No. 3. [1920.] Wine anp Foorp. Bull. 22, Agric. Exp. Stn., Coll. of Agric., Cornel University. [1904] (ier4son | XXXVII. ON A NEW METHOD OF GAUGING THE DISCHARGE OF RIVERS. By J: JOLY, ScD ERS, ECD: [Read Frsruary 28. Published Aprin 13, 1922. ] THE new method is based on the principle underlying “ chemical hydrometry.” This principle may be simply expressed as follows :— Suppose it were required to estimate the quantity of water contained ina large tank the volume of which it was difficult to determine by mere geometric measurements. By means of chemical hydrometry the quantity of water is easily found. A known weight of any suitable soluble salt is stirred into the tank, so that there is uniform distribution of the salt throughout the water. We then abstract a known volume V of the water, and by titration or other means estimate how much of the salt is contained in this volume of water. iL : Now, if this amounts to 7th of the salt put into the tank, the volume of water in the tank must be n x V. This method has been applied to the discharge of rivers. A tank is erected at some convenient point on the river. In this tank a large volume of water is stored, and into it is stirred a heavy charge of the salt; generally chloride of sodium. A pipe isled from the tank into the river, and the salt solution is fed slowly, and at a uniform rate, from the tank, escaping from perforations in the pipe and intermingling with the current. At a point some distance down stream samples of water are taken at intervals. These samples are titrated; and, knowing the rate at which the salt has been fed into the river, the rate oi discharge can be calculated. Very accurate results are obtained by this method. The error can be reduced below 1 per cent. It is said to be superior in accuracy to weir- measurements, and very much superior in accuracy to measurements based on current meters. A very full account of the method is given by B. F. Groat, M.AM.SOC.C.E., in the Proceedings of the American Society of Civil Engineers, vol. xli, November, 1915. A study of Mr. Groat’s paper reveals the advantages and disadvantages of the method. Chief of the latter is the cumbersome nature of the prepara- tions which have to be made, due to the large amount of salt required. 490 Scientific Proceedings, Royal Dublin Society. Moreover, the cost is considerable, owing to the scale on which the salt must be supplied. It was in connexion with the important problems arising in relation to Irish water-power utilization, as recently put forward by Sir John Purser Griffith before this Society, that I thought of the following procedure. At the time I was in ignorance of the fact that the principle involved had ever before been used or thought of. I propose to utilize, in determining river discharge, the extraordinary accuracy with which radioactive measurements can be effected. A very simple form of electroscope suffices to determine a quantity of radium to the billionth part of a gram. The apparatus costs a very few shillings. If, now, in place of introducing salt by the hundredweight into the river, we flow into the river a few litres of a solution containing a trace of radium, and taking samples down stream examine them by the electroscope, the discharge of the river may be determined. In order to reduce the working conditions to figures, I take the case of the Poulaphuca and Golden Falls on the Upper Liffey. The flow I take as 20,000 cubic feet per minute. The constant of the electroscope I take as 0°54 x 10°". That is to say, this quantity of radium, or rather the emanation in equilibrium with it, will cause the gold leaf of the electroscope to show a gain in rate of fall of 1 scale division per hour. I have made many such electroscopes. We require to know how much radium is to be put into the river, so that we can estimate the discharge to 1 per cent.; and I assume that the feed of radioactive water into the river is maintained for twenty minutes. Suppose we withdraw samples of 10 litres volume for testing by the electroscope. To evaluate the radium to 1 per cent., it must increase the rate of fall of the leaf of the electroscope by 100 scale divisions per hour. That is, the 10 tres must contain 54 x 10°? eram radium; ie., there must be 5-4 x 10°’ gram radium per litre of water passing down the river. Now, the flow of the river as assumed is about 600,000 litres per minute. The quantity of radium fed in per minute must therefore be 6 x 10° x 5-4 x 107 = 32 x 107 gram. g, p; and for twenty minutes’ flow we require 64 x 10° gram radium: that is, 0°064 milligram must be put in. Now, the most economical form in which the radium can be supplied is pitchblende, the mineral oxide of wranium. In this ore 200 milligrams of radium are associated with 1,000 kilos of the mineral, from which we find that 0°01 milligram of radium are contained in 50 grams of the ore. 300 grams are required to supply 0:06 milligram. Joty—On a New Method of Gauging the Discharge of Rivers. 491 This quantity of radium—pure—would cost about 36s.; in the ore it should be one-sixth or one-eighth part of this, i.e., five or six shillings. If a two per cent. determination was deemed sufficiently accurate, and a fifteen-minute feed into the river, a hundred grams of the crude ore would suffice, and the cost would be correspondingly reduced. As regards details, 1 propose to bring the pitchblende into solution by powdering it and adding nitric acid. It dissolves readily. The charge is then diluted with water up to about two gallons, i.e., the full of a petrol tin. In this form it is brought to the scene of operations. The stopper is then replaced by one which carries a copper tube dipping to the bottom of the tin. This tube is coupled to a similar small-bore tube to be laid out across the current. It is pierced with fine holes at intervals, say of 6 inches. Air pumped into the tin by an ordinary tyre-pump serves to expel the liquid. The current is steadied by having in circuit an indiarubber bag, which is kept inflated to a constant pressure, as read by a small air-pressure gauge. The air-pressure being considerable, the small effect of variation in hydrostatic head will be negligible. I think a very little time spent on experiments in the laboratory will suffice to bring these practical details into working order. When the samples are collected (each sample will about fill a petrol tin) and brought back to the laboratory, they are stored in ordinary boiling flasks for ten or twelve days. The emanation is then boiled off, being caught in a small exhausted bulb, from which it is introduced into the At the end of three hours the electroscope is read, and the If thought desirable, the No error is electroscope. river discharge is then immediately deducible. samples may be evaporated down to any required extent. introduced by this process. It is necessary to carry out these operations in a laboratory free from radium contamination. [ej XX XVIII. ON A VARIETY OF PINITE OCCURRING AT BALLYCORUS, CO! DUBIN: By LOUIS B. SMYTH, M.A., Sc.B., Lecturer in Palaeontology in the University of Dublin. {Read Frpruary 28. Published Apri 13, 1922.] Introduction. Some months ago Mr. J. C. Davison, of the Engineering School, Trinity College, Dublin, brought me for identification a green mineral, found at the old lead mines of Ballycorus. It appears not to have been described before. Mode of Occurrence. A visit to the locality showed that the substance occurs in the granite blocks thrown out during mining operations, and lying about in and around the large cutting, near the chimney, at the top of the hill. The excavation is at the junction of granite and schist. All stages can be found between typical granite and a thoroughly green rock. Results of Tests. The following are the results of tests made on the most translucent, lustrous fragments :— Compact. Unctuous. Jel, (Cr 2h, Lustre.—Fatty. Colour.—Oil green to pale yellowish green. Translucent. Insoluble, and does not gelatinize in hydrochloric or nitric acids, or n caustic potash. Decomposed by hot concentrated sulphuric acid, with separation of powdery silica. Alum is deposited when the filtered acid is evaporated. In the closed tube gives off water easily, and becomes white. Some fragments darken. Occasionally decrepitates slightly. BB.—Infusible, turns white, and does not colour the flame. With cobalt nitrate gives a strong blue. Details of Vests. Hardness tests required special care owing to the presence of minute grains and veins of quartz. ‘he salient angles of chips were used to scratch good cleavage surfaces oi selenite, Iceland spar, &c. In one sample, for SmytH— On a Variety of Pinite oceurring at Ballycorus, Co. Dublin. 493 instance, there were four convenient angles. Three of these scratched Iceland spar ; the fourth would not. The latter corner scratched selenite. Measurements of specific gravity were made by means of methylene iodide, diluted with xylol, on very small particles obtained by crushing. The specific gravity of the liquid was found with a Westphal balance. The highest results are considered the most reliable, for (1) the microscope showed that the chief impurity was quartz, and the specific gravity always came out higher than that of quartz; (2) dull specimens known to be impure gave low results; (3) the more finely divided material gave a higher proportion of heavy grains. Chemical Analysis. The result of a chemical analysis of five or six grams of the most carefully picked material is shown in column I of the table. One chip of the same lot was sliced for the microscope in order to gain an idea of its purity. It contains a remnant of a disintegrating felspar crystal and some minute granules of quartz, but the great bulk of it is of one kind of material of uniform texture. ils Il. TOE, SiO. _ | 49:54 | 46 47 AleO3 O 2 9 34°86 33 32 Tron Oxide. j : 0:24 4 i 3 CaO 0 é tr. 0 1 MgO ; 5 : 0-72 1 Q Na20 : ° : 0:78 1 1 K.0 6 O 9°00 9°5 8 LO . 6 B : tr. — — P205 9 = = COz 0:00 = aes F 0-00 0°5 = | Moisture : : : 0-90 | 4:80 5 6 Loss on ignition 6 ° 3°90 99°94 100 100 I.—Pinite, Ballycorus, Co. Dublin. Il.—Average of 19 analyses of muscovite quoted by Dana (Syst. Min., 6th ed., pp. 617-8, analyses 1 to 19, inclusive. 1911). I[1.—Average of 12 analyses of pinite (/oc cat., p. 622). 494 Scientific Proceedings, Royal Dublin Society. Microscopical Characters, In thin section the mineral still appears compact when viewed with a pocket lens. With a magnification of seventy diameters it exhibits a granular structure, and appears to have higher refraction than quartz. Between crossed nicols it is seen to be birefringent, the field being finely speckled. A higher power shows that in places the structure consists of rough radial aggregates, in which the components may occasionally be seen to have straight extinction. Identification. Comparing the analysis with column II of the table, it will be seen that there is good agreement with an average muscovite. The latter, however, is not affected by acids, whereas the Ballycorus mineral is completely decom- posed by hot concentrated sulphuric acid. The term Pinite is applied to a number of compact substances, which are generally alteration products, and have a composition similar to that of muscovite. Most, if not all, of them differ from muscovite, and agree with our mineral, in being decomposable by sulphuric acid. As was to be expected, since they are generally impure substances, none of the named varieties was found to agree in all points with the Ballycorus mineral, though many come near it. The best agreement seems to be with pinitoid. That mineral, according to A. Knop,’ is a basic silicate, mica-like in composition, containing water, decomposable by hot sulphuric acid, micro to eryptocrystalline, of usually earthy to dense compact habit, of a leek-, oil-, or greyish-green to whitish colour, which may pass over into various shades of red. Specific gravity, 2°788 ; H, 2:5. It is a secondary mineral formed by hydration, which frequently occurs as pseudomorphs after felspar in altered porphyry. So far this might apply to the Ballycorus mineral, except that it was not observed in shades of red. Knop’s analysis would account for this, and reveals a considerable difference in composition, for the pinitoid analysed gave 8°84 per cent. oxide of iron against 0°24 per cent. in our mineral. Pinitoid, too, is low in potash, 5°85, as compared with 9-00. Column IIT of the table is the mean of the twelve pinite analyses given by Dana.* Comparison with column I shows the close agreement. As a matter of fact, all the figures for the Ballycorus mineral are inside the extremes occurring in those analyses. 1“ Neues Jahrbuch fiir Mineralogie,” etc., p. 508. 1859. * Loc. cit., p. 622. Smyta— On a Variety of Pinite occurring at Ballycorus, Co. Dublin. 495 The mineral under discussion, then, is evidently a pinite. It seems to me unnecessary to invent any more definite term. Since killinite, another variety of pinite, occurs, also near the granite margin, at Killiney, a few miles from Ballycorus, we must distinguish it from the present mineral. It differs most distinctly (1) in being pseudomorphous after spodumene, (2) in being fusible before the blowpipe.! Origin and Relations. A number of thin slices of granite in various stages of alteration were prepared. One fresh-looking specimen contains abundant allotriomorphic microcline, with strong, typical cross-hatching, only slightly cloudy on certain laminae, and often including idiomorphic crystals of the other felspars. There is much albite, mostly speckled over with alteration products. What appears to be orthoclase is much altered, the secondary material usually occupying all the centre of the crystal, and agreeing in appearance with the pinite. The quartz is in groups of grains, and there is a fair amount of muscovite. Cataclastic structures are evident. The quartz is in places finely granular, and the mica bent and broken. In a shear zone the latter seems to be ground to a fine granular paste, which is doubtfully distinguishable in places from the pinite. A little galena is present. Slices containing a considerable amount of pinite appear to show that it originates, at least chiefly, from felspar. The latter often occurs like an archipelago surrounded by a sea of the alteration product, the islands being optically parallel. In this material the felspars have rather indefinite optical characters. Considerable fragments may appear untwinned, but a higher magnification often reveals the microcline structure on a minute scale in spots, mostly at the edges. What appears to be secondary lamellar twinning is often seen. In several cases there cross such felspars broad bands, suggesting shearing planes, distinguished by more intense alteration. Quartz is generally abundant, often showing strain shadows, which are sometimes arranged fanwise. A slice made from a piece of rock having a fairly uniform green colour, but. a dull lustre, was found to consist of pinite with abundant, but small, ragged quartz grains, a very few similar fragments of felspar, and still fewer of muscovite. A remarkable relation between quartz and pinite is observed in some 1Thos. Taylor, ‘‘ Account of a New Mineral Substance discovered at Killiney, in the vicinity of Dublin,” Trans. Roy. Ir. Acad., vol. xiii, p. 61. 1818. T. Thompson, ‘‘ Outlines of Mineralogy,” vol. i, p. 330. 1836, 496 Scientific Proceedings. Royal Dublin Society. slices, the latter being in rosette-like patches within the quartz grains in the plane of section. or biting into their edges, as if growing at the expense of the quartz. This is extremely unlikely, especially as no trace of fluorine was found in the analysed material, though looked for specially. It is probable that the silica, thrown out of the felspar as it alters into pinite, crystallises round the latter, and is moulded upon it. That quartz is so produced is well known. Rosenbusch,! referring to orthoclase “altered to muscovite (pinitoid),” states that “quartz is almost always mixed with these pseudomorphs in variable amounts.” 1 << Microscopical Physiography of Rock-making Minerals,” Rosenbusch-Iddings, 3rd ed., p. 314. 1893. XXXIX. ON THE LIFE-HISTORY AND BIONOMICS OF THE FLAX FLEA- BEETLE (LONGITARSUS PARVULUS, PAYXK.), WITH DESCRIPTIONS OF THE HITHERTO UNKNOWN LARVAL AND PUPAL STAGES. By J. G. RHYNEHART, A.R.C.Se.1., D.1.C., N.D.A., Entomologist in the Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (Pirates XV-XIX.) [Read Frpuvany 28. Published Avrin 18, 1922.] 1.—INTRODUCTION. NEARLY one hundred years ago the flax crop in parts of Ulster was almost completely destroyed while in the seedling stage asa result of the ravages of the small, black, jumping beetle, now known as the flax flea-beetle. During various intervening seasons damage due to the same cause has occurred in many parts of Ireland, and this beetle is now regarded as being by far the most dangerous insect enemy of flax, and one capable of causing losses amounting to many thousands of pounds sterling in a single season. Like the other members of the family Chrysomelidae, it is herbivorous, feeding in the adult stage on the cotyledons and young leafy shoots of flax seedlings. The destruction of the young plants gives rise to thin, uneven brairds, and the ultimate production of dwarfed, branched, and uneconomic crops of flax; indeed, it is frequently the case that the young seedlings are entirely destroyed and large areas are laid bare, thus necessitating re-sowing. Flax-growers have long been familiar with the ravages of this pest, which is popularly known as the “ fly” or “ flea”; but, despite this fact, little work has been done regarding its biology and habits, or possible remedial measures, while no details are available in entomological literature with reference to the general relations of the adult beetles and their larvae with the food plants and the soil in which the latter are growing. In July, 1920, the writer commenced, at the Department’s Field Laboratory for Flax Diseases, Coleraine, Co. Derry, an investigation on the life-history, habits, etc., of the insect, with a view to the possibility of evolving a suitable check on its ravages; and a general account of the progress made that season is incorporated in the second Report of Investigations on Flax 498 Scientific Proceedings, Royal Dublin Society. Diseases, published by the Department of Agriculture and Technical Instruc- tion for Ireland (51)! in 1921. The work was continued at Coleraine during the spring and summer of 1921, and in the present paper it is proposed to give an account of the life- history and habits of the beetle, together with detailed descriptions of the morphology of the different immature stages, and general information relating to the species in so far as it has been investigated up to the present. The writer wishes here to acknowledge his indebtedness to Professor G. H. Carpenter, of the Royal College of Science, under whose guidance the work has been carried out, for much valuable help and criticism, especially with reference to the descriptions of the different stages. 2.—HISTORY AND SYNONYMY OF THE SPECIES. The nomenclature of the flax flea-beetle has, unfortunately, been the subject of much controversy, and many descriptions under many different names have been published. late workers, however, are agreed that the correct name of the species is Longitarsus parvulus—an agreement shared by the present writer. From 1762, when Geoffroy (26) proposed for the flea-beetles the generic name Altica, until 1807, when Illiger (87) divided his genus Haltica into nine sections, and defined the characters common to each section, the position of the group was conjectural. Illiger gave to the seventh section the name “ Longitarses”; and Latreille (46), in 1829, raised this section to generic rank. Two years later Stephens (59), in a new system of classification, founded the generic name 7/yamis, and substituted it for Longitarsus Latr., while a further change was made in 1854 by Chevrolat (16), who, instead of employing either Longitarsus or Thyamis, used the new name Zeinodactyla. Longitarsus is, therefore, the legitimate generic name. The specific names parvulus and ater have been used simultaneously by various authors. Patterson (48), to whom the first record of damage to flax in Ireland by the species is due, uses the former, while several years later Janson refers specimens from Irish flax fields to the same species (61). After the publication of the catalogue of Gemminger and Harold (25), in which ater is given priority (from Altica atra F.), many writers adopted the latter specific name. Thus, it has been used by Bargagl (2) in Italy; Fowler (24), Sharp and Fowler (58), Hudson-Beare and Donisthorpe (36), and others in England; and Johnson and Halbert (39) and Carpenter (8) in Ireland. ' The numbers in brackets refer to the literature cited, Ruynenart—Life-History and Bionomics of the Flax Flea-Beetle. 499 In 1775 Fabricius (19) described Altica atra thus:—“