ia ee ee ees oe Per as Tue weet” WO AA oo eee ER ows DS bpd bd he veut vy hd evv AR AIA sicewvyh ww! ed wl ww Vv wy, ~ 4 . ~S Ess lrsruws Ste e\ SARIS eh pe beaded Sn accel Ad 1A hd ae Serre “yee io ae Sy Se , ete Vive Se \ fe ve ee ee od wee Y. Vvywwe AA vSNie AA OMA RI SSS < I I ha ner id NSS; WAS VINN TY SQN ew IIIA Shon encod ed oe SN SS hs Fg SL wees SCLC TeN es Cee e'y OSL S v AC y ws SS ew vh wwe # ~ 28s, ~s ‘S AS uwvuy ‘oly BR ~ as erNVY vw ee ee pg IN . e, a. we | a Vie Ww 'V = We oy deed ore " WOWWUU eb wey J Vow. Sone GSSESE) he aM she Stee SR a oT WOT AG AAS wo nyu’ oS Ss ay, y CER sey se oreyesttscovuyusyes™ © w" | “a : UC vew a" Ne A wo wV w Asef to ate Sie eee “Se els ~~~ SNR = ~N STAs adea oOetd wus co Wyre use wor = ee dake wowrrrr TONS wv Nea ei ca COM wer © Oe | wiv : jp hapielel eet ¥ wN Ww! ASK SS: = GET, MOS ODORS Ae IS eh wiegegis wleere esa oo ov es at toe , ee) Te ww if ; ) eihe Ss “4 YS 7 = (a) @xGhy the total surface area => 1077 (a) sq. cm., \ and the surface area per cc. = 960.2. Tn the example given the surface area = ae = 559 sq. cm. per c.c. © WATER OF K=2:4x10° @ 22 9 = FO x 10° x 2 2 2a GST x 10~& SURFACE AREA IN SQ.CMS. FER CC SPRAYING PRESSURE IN CMS. HG. — a Se ee | 15 20 2y , 30 JS 40 The results of observations made at different pressures of the sprayer are shown by the graphs of fig. 3. ‘The area of water-surface per c.c. of water sprayed increases uniformly with the pressure up to a spraying pressure of about 11 cms. Hg., and the three samples of water behave in the same way. But at higher pressures the rate of increase of surface with rising pressure is much slower and different for the three samples. Thus we have the unexpected result 8 Scientific Proceedings, Royal Dublin Society. that the sprayer working at the same pressure produces a degree of pulverization depending on the purity of the water, the purer water being broken into smaller drops. It may be remarked that the output of the sprayer (as expressed in the dotted curve of fig. 1) is the same for all samples of water. The graphs of fig. 3 have been drawn as intersecting straight lines. It is possible that more accurate observation might reveal them as rounded in the region of inflexion. The bending over of the curve may really occur at slightly different pressures for the different samples of water. The pressure at which the change occurs is certainly well removed from the pressures 16, 19, and 21cm. Hg. at which discontinuities occurred in the curves showing the charge per cc. plotted against pressure (fig. 2). The first part of the graph, which is common to the three samples of water, does not pass through the origin. It intercepts the vertical axis at a point corre- sponding to a surface of 130 sq. em. per’c.c. This is easily understood, for we have been plotting the total surface per c.c. of water, whereas what we are really interested in is the area of new water-surface. It would seem that this value of 130 represents the surface area per c.c. of the water issuing from the sprayer before it is broken up. If the issuing jet filled the whole orifice, it could be regarded as a cylinder, and its original surface per ¢.c. would be 2/r, 7 being the radius of the orifice. ‘The value of 2/7 in this case is about 70. Now, the true value of the radius of the water-jet is certainly less than that of the orifice, the water being surrounded by an annular air-blast. ‘J’o assume that the true radius of the water-jet is half that of the orifice would bring the numbers into approximate agreement. It is not unreasonable, then, to suppose that this intercept gives us the original water-surface per c.c. Charge considered as a function of the new area produced. We can now combine the results of these curves with those in fig. 2 (charge per c.c. against pressure), and, following the practice of the previous paper, plot the charge per c.c. against the area of new surface per c.c. These curves are given in fig. 4. In the previous work the points were found to lie, roughly, round a straight line passing through the origin, showing that the amount of charging was proportional to the area of new surface. In the present more accurate curves there is a rude suggestion of this relationship, especially if the proper origin of co-ordinates is taken (allowance being made for the original surface area of the water). The simple idea of direct proportionality, however, has to be abandoned. Three notable points appear from a consideration of these curves. First, the effect of purity of the water on the charge developed is most important for the smaller degrees of pulverization. For example, the purer sample, when broken into drops of a certain comparatively large size, gives a charge of 2 e.s. units per c.c. The other samples, when broken into drops of the same size, give charges of 0°8 and 0°2 es. units per c.c. respectively. In the second place, for high degrees of pulveriza- tion it would seem as if the three curves were going to fuse into one, which means that if the water is broken into small enough drops the charge per c.c. will be the same whatever the purity of the water. The only difference is that (as the curves in fig. 3 show) it is apparently more difficult to break up the impure water. Finally, the discontinuities on the curves for electric charge, which could not be associated with any value of the electric charge, or any value of the spraying pressure, appear now on all three curves for the same value of surface-area per c.c., that is, for the same average size of drop. It would appear from these curves Nouan and Enrigur—Lilectrification Produced by Breaking up Water. 9 that, as the size of the drops in the spray becomes smaller, the electrical separa- tion occurs with greater and greater facility, especially in the case of pure water, until the average size of the drop reaches a certain value (radius = 65 x 10% cms.). At this stage there is a definite check. The further stages of the curve show, in the case of the two less pure samples, a tilt up, suggesting, as we have said, that they will fuse with the curve for the purer water. In examining these curves we must consider to what extent they may have been effected by certain actions going on in the spray, which we have neglected up to the present, viz., electrical recombination and recombination of drops. The ions carrying the negative charge are not immediately separated from the spray. Some amount of recombination will take place, and the measured electric charge on the water will be correspondingly reduced. There may be two types of recombination, | Ke2-4x10° 127 R 107 is w K=30x10°° Ww 8 2 : 2 6 Ne t K=55 x10°° Wy a S i 4 a s s Wee see t + L | 0 100 200 300 100 500 600 700 AREA OF WATER-SURIFACE IN SQ.CMS PER CC. Fig. 4. an initial type resembling (but on a different scale) the initial recombination, postulated in connexion with a-particle ionisation, and a type corresponding to later stages. Jf anything of the nature of the first type exists, it is probably beyond the range of detection in any of our experiments, and consequently does not influence the shape of our curves. The second type of recombination is certainly operative, but its effect may not be very great in this case, where the ions are carried away rapidly by the air-blast, while the drops sink steadily towards the receiver. It has been found! that the ions from a water-spray, examined as tJ. J. Nolan, Royal Irish Acad. Proc., A, 33 (1916). SCIENT. PROC. R.D.S., VOL. XVII, NO. I. G 10 Scientific Proceedings, Royal Dublin Society. soon as possible after their formation, consist, to a large extent, of the slowly moving type. They are probably, as a whole, not very effective for purposes of recombination. The extent to which their activity affects our values of charge is difficult to estimate. It should not vary much for different pressures of the sprayer, but its effect is possibly less at higher pressures, owing to the more rapid removal of the air. On the whole, we conclude that the electrical recombination should not affect the shape of our curves seriously beyond a possible apparent tilting up of the electrical values at higher pressures (smaller drops). We have next to consider the possibility that coalescence takes place to some extent among the drops. This would result in our underestimating the extent of the pulverization undergone by the water. As is well known from the experi- ments of Lord Rayleigh, drops of pure water colliding will rebound if uncharged. A moderate degree of charging will promote coalescence, while a higher degree of charging will prevent it. It might be said, then, that all our measurements are affected by coalescence of the drops; that the effect is less with the purer water, owing to the higher charges developed; that the purer water being sprayed into smaller drops (fig. 3) merely means that in that case there has been less coalescence between drops. We are not in the possession of any quantitative data by which this view could be tested, but we are unwilling to believe that coalescence takes place between the drops to any serious extent, for a variety of reasons: (1) Except in the immediate vicinity of the nozzle of the sprayer, there is nothing to cause collision between drops. For most of their time, between production and capture, they are moving steadily through air that is practically at rest; and even when they are being driven forward by the air-current from the sprayer there is no appearance of whirling or violent motion which would throw them against one another. (2) When the vertical distance traversed by the drops before capture was varied from 20 to 80 ems. in steps of 10 cms., there was no appreciable change in the size of the drops examined. This experiment shows that the drops are certainly not coalescing in the later stages of their existence. (3) In fig. 3 the first part of the graph (pulverization against pressure) is common to the three samples. But this is just the stage at which the first part of the Rayleigh effect (moderate charges promoting coalescence) should show up distinctly. The charges on the three samples here differ most widely; for example, on the more impure sample the charge at 5 cm. pressure is practically zero. We should therefore be able to trace the effect of charges increasing from zero and promoting increasing degrees of coalescence, the effect developing more rapidly in the purer samples. But we have no evidence of any action of this kind. (4) If the separation of the later part of the graph (fig. 3) into three parts be taken to indicate the effect of the second part of the Rayleigh phenomenon (increasing charge preventing coalescence), it is difficult to see why the transition should occur at a certain value of the spraying pressure and size of drop, the charges at that stage being in the ratio roughly 1:3: 8. We believe therefore that recombination of drops, if it occurs at all, does not affect our measurements seriously. The tendency of the three curves in fig. 4 to join into one, which would seem to be very approximately a straight line passing through the true origin, we interpret as indicating that the charging of the water is a genuine surface effect. The charge produced is proportional to the area of new surface for any sample of water if the water is broken up to a sufficient extent. The full electrical separation is inhibited by impurities present if the degree of pulverization is small. There is apparently a critical size of drop at which the inhibiting effect of impurities becomes ineffective, The effect of the Nouan and Enricur—FLleetrification Produced by Breaking up Water. 11 impurities in modifying the electrical conditions at the surface-layer is probably connected with their effect in reducing the degree of pulverization produced at a given spraying pressure. The impure water can, to a certain extent, readjust its surface-tension under applied stresses; the pure water is, as it were, more “brittle.” Our measurements show that the quantity of electricity associated with the formation of 1 sq. em. of new water surface is about 0:02 es. units. Results considered vn connexion with Simpson’s “ breaking-drop” theory. In this connexion the outstanding feature of these experiments is the importance of purity of the water when the degree of breaking up is small. Our purest sample of water had a specific conductivity of 2:4 x 10° ohm™, which does not indicate a very high degree of purity. This water stood for days in an earthenware vessel in contact with air, and, of course, was also in contact with glass vessels. The nature of the contamination that affects the activity of the water from the electrical point of view we do not know. This is a point we hope to investigate subsequently. But it is not unreasonable to suppose that a rain-drop, formed in the well-filtered air of a thundercloud, is much purer than our purest sample. Hence, for moderate degrees of breaking up, we would expect the thunder-rain to acquire bigger charges than those observed by us. But even for the same degree of purity it can be seen that the charges observed on thunderstorm-rain can easily be reached. If a drop of radius 7 breaks up into 27 equal drops,’ the change in surface per ¢.c. is 6/7. Taking a drop of diameter 4mm., the change in surface per c.c. will be 30 sq.cm. Reference to fig. 4 shows that this would produce, in our purest sample, a charge somewhat greater than 0:2 e.s. units per c.c. We see, therefore, that for the purer water of the thunderstorm we need not demand any high degree of breaking up or any very sustained repetition of the process in order to produce the charges observed on thunder-rain. There is another process at work which tends to concentrate the charge. It is recognized that evaporation from the falling rain-drops in a thunderstorm is very rapid. Evaporation no doubt also plays a part in increasing the magnitude of the charge per c.c. on ordinary rain. We consider, therefore, that if a moderate degree of purity be assumed for rain-drops in the upper atmosphere, the theory put forward by Simpson is fully competent to account for the observed electrical phenomena of thunderstorms. It also explains the sign and magnitude of the charge on the greater part of ordinary rain. 1 See Hochschwender’s photograph, fig. 6. Lenard, loc. cit. a . | (ie | No. 2. CATAPHORESIS OF AIR-BUBBLES IN VARIOUS LIQUIDS. By THOMAS A. McLAUGHLIN, M.Sc., A.Insr.P., University College, Galway. [Read Apri 25. Printed Junz 1, 1922.] On the subject of the electrification in colloids, that is, of fine particles in suspension in a liquid medium, much research has been done and a fair under- standing obtained. Likewise, something has been done on the electrification in emulsions, but a broad field for research remains in this case of liquid suspensions in liquid. In the remaining case, of what, for the sake of continuity, may be termed gas suspensions in liquids, still less has been done, and, in view of the advanced state of knowledge on electric phenomena in gases, this seemed a promising field for research on cataphoresis phenomena. Mention is made by Quincke' of the motion of air-bubbles under an electric field. He reports that air-bubbles in water, carbon disulphide, and turpentine move under an electric field, in water towards the positive pole, in carbon disulphide and turpentine towards the negative pole. McTaggart’ verifies Quincke on air- bubbles in distilled water, and finds no motion of air-bubbles under an electric field in the pure alcohols, methyl, ethyl, propyl, and isobutyl. Otherwise, there seems to be no record of the motion of air-bubbles in pure liquids under an electric field. In the present research the plan was to make preliminary investigations into the charge on air-bubbles in pure liquids, with special reference to badly con- ducting liquids. It was thought possible, at first, that the investigation might be approached after the manner of Millikan’s oil-drop experiments, that is, that the upward tendency of an air-bubble in a liquid might be balanced, owing to the charge on the bubble, by an electric field, and from such balancing experiments the charge calculated. It was decided, however, to proceed with preliminary investiga- tions with apparatus of McTaggart’s type, and, with a fuller knowledge of the phenomena, to return to the problem by other methods of investigation. The apparatus used was a glass tube 18mm. diameter and 4 cm. long, fitted with ground brass plugs, which were air-tight. On one plug a pulley was cut, and both were supported between pivotal ends on a heavy base, fitted with levelling screws. The tube was rotated, as in McTaggart’s experiments, by a Rayleigh motor. The motor and the apparatus stand were clamped in position, so that everything was steady, and regular motion of the tube was obtained. The required air-bubble was simply introduced by removing one plug of the tube, pouring in liquid, and reintroducing the plug in time to catch a bubble which had been formed by the pouring of the liquid. The tube being in position and rotating, the bubble took up a position on the axis of the tube, and could be kept thus at rest during the period of observation by proper adjustment of the levelling screws and speed of the motor. The bubbles were illuminated by an electric lamp through a narrow 1 « Pogg Ann.” 113, 1861. 2 Phil. Mag., 27, p. 297, and 28, p. 367, 1914. SOIENT. PROC. R.D.S., VOL, XVII, NO. 9}, D 14 Scientific Proceedings, Royal Dublin Society. horizontal adjustable slit, and observed through a low-power microscope. The image of the bubble was clear and distinct, and its size could be determined on the scale- piece of the microscope, which read 26 divisions to 1 mm. The first observations taken were on air-bubbles in distilled water. The results of Quincke and McTaggart, that they moved to the positive pole, were verified. McTaggart’s results in the case of the alcohols were likewise verified. In ethyl alcohol bubbles of diameters 0:08, 0°10, 0:12, 0°16, 0°20, 0:24, and 0:28 mm. were observed under a field of 160 volts per cm., and showed no motion. In butyl alcohol bubbles of diameters 0:06, 0°12, 0°16, and 0°24 mm. showed no motion under a field of 95 volts per cm., being under observation in each case for at least sixty seconds. In methyl alcohol bubbles of 0:20 and 0:32 mm. diameter, respectively, showed no motion under 160 volts per cm. These bubbles were the only two which were success- fully observed, the time of observation being only fifteen seconds. The difficulty arose from the fact that the field had the peculiar effect of causing air-bubbles to dissolve quickly. This phenomenon is mentioned by McTaggart. The following observations may be recorded in this connexion :—With no field in action a bubble of diameter 0:08 mm. remained unchanged while observed for forty-five seconds, but dissolved completely within three seconds under a field of 160 volts per em. Again, bubbles of 0:12, 0:16, and 0:20 mm. diameters disappeared under field within “ten seconds, A bubble of 0-20 mm. diameter grew to 0°28 mm. in thirty seconds without a field, and decreased to 0'12mm. in fifteen seconds under field. Another of 0:28 mm. diameter gradually dissolved under field in forty seconds. At the same time, none of these bubbles showed any tendency to definite motion under the field. A similar effect of an electric field on the solution of the bubbles was looked for in the other alcohols and other liquids examined, but found to be absent in all cases. Observa- tions were now proceeded with on air bubbles in a number of liquids not previously examined. In wylol several bubbles, each of the following diameters, 0:08, 9°10, 0:12, 0:20, 0:24, 0:28 mm., were examined, and showed no motion under a field of 160 volts per cm. Each bubble was kept under observation with reverse fields for at least thirty seconds, except bubbles of 0:12 mm. diameter and under. These latter always dissolved quickly, while some of the larger bubbles showed a tendency to grow in the liquid, but in no case did the change in size affect the neutral attitude of the bubbles to the field. Air-bubbles in benzene were next examined. In the first sample they all moved to the negative pole under field. The benzene was probably impure, as the bottle contained some traces of red rubber cork, portion of which had probably gone into solution. In fresh pure benzene, under a field of 160 volts per cm. for periods of about thirty seconds, various bubbles of diameters 0-06, 0°12, 0:16, 0-20, 0:24, 0:28, 0°32, 0°40, and 0:48 mm. showed no motion. There was not the same difficulty with benzene as with xylol in investigating the smaller bubbles, but, as before, some bubbles grew in size while under examination, as, for example, from 0°32 to 0-48 mm. diameter, but the change did not affect the neutral attitude to the field. In foluene air-bubbles 0°32mm. and 0-48 mm. did not move under reversed fields of 95 volts per em. for thirty seconds each way. Another bubble, 0°32 mm. diameter, was under observation for two minutes in the same field, and did not move. A small bubble, 0:12 mm. diameter, gradually dissolved, but was uninfluenced by the field. It was now decided to proceed with investigations with benzene derivatives, some of which are chemically neutral, others acidic, others basic, in the hope that the presence of free ions might be the clue to the presence or absence of charge. McLaucuuwn— Cutaphoresis of Air-Bubbles in Various Liquids. 15 In bromobenzene, benzaldehyde, aniline, cinnamic aldehyde, ethyl malonate, and oleic acid, bubbles of diameters from 0:08 to 0°28 mm. did not move under a field of 95 volts per cm. Benzyl alcohol disintegrated under 95 volts per cm., but bubbles 0:08 mm. and 0-12 mm. diameters showed no motion under a field of 14 volts per cm. Lactic acid likewise disintegrated, but a bubble 0°12 mm. diameter did not move under field for thirty seconds prior to disintegration. In ethyl acetate bubbles grew, but this growth was uninfluenced by field. A bubble 0:20 mm. diameter grew to 0°28 mm. while under observation for thirty seconds each way with a field of 95 volts per cm., but showed no sign of motion, Another bubble 0:16 mm. diameter was likewise uninfluenced by field. It was not found possible to trap an air-bubble in acetone and such volatile liquids as carbon disulphide, but it was noticed that foreign matter moved to the positive pole in impure acetone. Thus, in all pure organic liquids tested so far air-bubbles showed no cata- phoresis. In one organic liquid, witrobensene, air-bubbles never failed to move under field. ‘Ihe first sample used was from ordinary laboratory supply, which was a dark yellow in colour, due to oxidation from exposure to light. It had been in store for some months. Bubbles 0:08 and 0:16 mm. diameters undoubtedly moved towards the negative pole. Likewise, other samples from same source gave to air-bubbles a motion in the same direction, the bubbles being of diameters 0:08, 0:16, 0°24, and 0°32 mm. The latter moved 20 divisions per minute and 19 divisions per minute in opposite directions under reversed fields of 95 volts per em. A fresh pure distilled sample of nitrobenzene was obtained. It was light yellow in colour, and was standing for about ten days in a dark place prior to use. A potential of 95 volts per cm. gave a motion of 4 divisions per minute in either direction, showing apparent positive charge. A later observation showed a motion of 11 divisions per minute ; the charge was still positive, but seemed to have increased. ‘The bubbles were 0:20 and 0:12 mm. diameter, respectively. A second sample from the same source gave to an air-bubble a negative charge, the motion being 5 divisions per minute under the usual field. The same sample, after standing for three days in light, again gave a negative charge to two bubbles of diameters 0:04 and 0:08 mm.; the charge had increased, the rate of motion now being 38 divisions per minute. Close examination of the liquid showed that some traces of red rubber from the cork of the bottle had got into it. A further sample of nitrobenzene was freshly dis- tilled and carefully dried over phosphorus pentoxide. In all cases bubbles showed an apparent positive charge. In the first sample bubbles of diameters 0:08, 0:12, and 0:28 mm. all moved to the negative pole with velocity of 200 divisions per minute under 95 volts per cm. Ina second sample from the same source bubbles of diameters 0°04 and 0:20 mm. moved towards the negative pole with a velocity of 140 divisions per minute under the same field. A third sample gave exactly the same motion to bubbles of 0:08 and 0:16mm. diameters. In a fourth sample two bubbles of diameters 0°08 and 0:12 mm. acted likewise. Thus, bubbles in “pure” nitrobenzene move to the negative pole, but in impure nitrobenzene the charge seems to depend, both in sign and magnitude, on the purity of the nitro- benzene—a factor which is difficult to determine. Owing to the general absence of cataphoresis in the great number of liquids tested, Quincke’s statement that air-bubbles were positively charged in turpentine is of interest. On test, however, it could not be verified. Two distinct samples of turpentine from two different sources were tested. In one an air-bubble of 16 Scientific Proceedings, Royal Dublin Society. diameter 0-4 mm. gradually dissolved, and was observed for two minutes under a field of 90 volts per em., but showed no motion. Another bubble 0-08 mm. gave the same result. In the second sample a bubble of 0:16 mm. diameter was watched gradually dissolving for two minutes under the same field, and likewise did not move. In view of these contradictory results, it is of interest to compare results on electric endosmose, reported in the same paper by Quincke, with results obtained by Perrin, and to recall Freundlich’s comment on the apparent contradiction :— “Perrin fand keine Elektroendosmose bei... Terpentinol...... dagegen fand Quincke bei Terpentin6] eine Verschiebung der Flissig- keit zum positivem Pol . . . Da schon beim Wasser kleine Zusatze grosse Anderungen der Elektroendosmose bedingen, ist es fraglos, dass Stoffe verschiedener Reinheit ein sehr verschiedenes Ergebnis zeitigen kénnen”’ (Kapillarchemie von H. Freundlich, page 241), It is of interest also to note that reference! 1s made by Coehn and Mozer to experiments showing that pure turpentine does not acquire a charge in bubbling, while they point out that Lenard found that ordinary unpurified turpentine acquired a charge. Discussion. The experiments on cataphoresis are meant to be purely of a preliminary nature, and the phenomenon is approached from the qualitative view-point rather than the quantitative. Adopting Lamb's’ equation as governing the motion of an air-bubble, then the velocity U under field 4 volts per cm. is given by where g = surface density of charge in double layer and 6 = thickness of double layer, u = viscosity of liquid. Thus the velocity of the air-bubble under a field in any given liquid is a measure of the charge in the inner double layer, i.e., the free charge which would accompany the bubble in motion in the liquid. This formula is laid down as applicable to all liquids which are not perfect insulators. Thus, absence of cataphoresis, under the conditions of the experiments, means either an entire absence of any double layer, or else that the charge is of a negligible amount as compared with that bound up in the double layer surrounding an air-bubble in distilled water. An air-bubble in nitrobenzene, or in impure liquids, is an exception. Thus, in benzene the presence of impurity brought about the cata- phoresis phenomenon. The question arises, therefore, is the phenomenon in nitrobenzene due to the same cause? The unstable nature of nitrobenzene and the doubtful nature of its purity in all cases would seem to uphold such belief. In particular the variation in the motion of an air-bubble in different samples from the same source might be explained as due to the instability of the nitrobenzene, which darkens if exposed to daylight. The phenomenon in water might be explained similarly, as distilled water cannot be regarded as perfectly pure. © It is 1Coehn u Mozer. Ann. der Physik, vol. xliii, p. 1045. 1914. 2 Lamb, Brit. Ass. Rep., 1887. McLavueGHLin— Cutaphoresis of Air- Bubbles in Various Liquids. 17 of interest also to note that the two liquids which showed the phenomenon have a much higher dielectric constant than those liquids in which it was absent, and, accordingly, are much better ionisers. This might be expected from a law laid down by Perrin as a result of experiments on electric osmose—an allied phenomenon :—‘‘ Electric osmose is only appreciable with ionising liquids; or, in other words, ionising liquids are the only ones which give strong electrification by contact.” Thus, the presence of impurities would be more hkely to bring about cataphoresis in liquids of high dielectric constant, e.g., water and nitrobenzene. The results recorded in this paper are also in keeping with Coehn and Mozer’s' results on contact electrification between gas-liquid surfaces investigated by bubbling experiments. They found that the greater the dielectric constant of the liquid, the greater was the charge recorded by bubbling a particular gas through it. It is remarkable that the charge produced by bubbling hydrogen through various liquids was much greater in the case of water and nitrobenzene than in any other tested. Benzaldehyde and aniline are recorded as giving charges in the bubbling experiments, while in these experiments they gave no charge to an air- bubble, SUMMARY. Under the conditions described in the paper air-bubbles were found to show no cataphoresis in the following liquids :—Methyl, ethyl and butyl] alcohols, xylol, benzene, toluene, bromobenzene, benzyl alcohol, benzaldehyde, aniline, cinnamic aldehyde, ethyl malonate, lactic acid, oleic acid, ethyl acetate, and turpentine. The results on turpentine are not in accordance with those obtained by Quincke. It was not found possible to trap an air-bubble in acetone and such volatile liquids as carbon disulphide. In impure acetone foreign matter moved towards the positive pole. In distilled water, air-bubbles moved towards the positive pole; im impure benzene, towards the negative pole. In “pure” nitrobenzene air-bubbles moved to the negative pole. In impure nitrobenzene the motion may be to either pole. The solution of an air-bubble in methyl alcohol under an electric field previously noted by McTaggart was coufirmed. ' Coehn u Mozer, Ann. der Physik, vol. xliii, p. 1048. 1914. CIENT. PROC. R.D.S., TOL. XVII, NO. Q. BR Ce] No. 3. ON THE AERATION OF QUIESCENT COLUMNS OF DISTILLED WATER AND OF SOLUTIONS OF SODIUM CHLORIDE. By PROFESSOR W. E. ADENEY, A.R.C.Sc.1., DSc. 1°.1.C., DR. A. G. G. LEONARD, F.ROSeL, B.Sc, ELC. AND A. RICHARDSON, A.R.C.Sce.1., AIC. [Read Arnie 25. Printed Juny 23, 1922.) Introductory. In the course of his investigations on the downward transmission of atmospheric gases through quiescent columns of water five feet in depth, one of the authors showed that, as the oxygen and nitrogen are dissolved at the exposed surface of the water, they do not remain concentrated in the surface layer, but are distri- buted through the lower layers with comparative rapidity. From the fact that the distribution of the dissolved gases, at various depths, after a comparatively short time was almost uniform, it appeared impossible to account for it on the assumption that it was entirely due to such an extremely slow process as that of the diffusion of the dissolved gases from the saturated surface layer. In addition to diffusion, it was. suggested that a process of downward “streaming” of the exposed layer of water occurs; and that it results largely, though possibly not wholly, from an increase in its density, which, in the case of distilled water, is caused by the lowering of temperature attending evaporation, and, in the case of salt solutions, by this factor and that of concentration. The downward streaming sets up a process of mixing of the constantly chanving air-saturated surface layer with the lower layers of the water. When a slow stream of dry air is continuously passed over the exposed layer of a column of water, surface density changes are constantly occurring, and comparatively rapid mixing ensues, with the result that, if the water be de-aerated at the commence- ment of the experiment, it becomes re-aerated with comparative rapidity.! The experiments described in this investigation have been carried out with the object of investigating the process of the aeration of de-aerated columns of water to a depth of ten feet. Columns of de-aerated distilled water and of solu- tions of sodium chloride were exposed to a slow stream of dry air for periods of from two to eight weeks, when samples at different depths were drawn off, and the nitrogen content of each determined. In general, it was found that re aeration proceeded more rapidly in salt soln- tions than in pure water,? and a further series of experiments was carried out ' Unrecognised Factors in the Transmission of Gases through Water.” By W. E. Adeney, Trans. R.D-S.. p. 161, 1905 ; and Phil. Mag., 1905. 2 See above references. SCIENT, PROC. R,D.S., VOL. XVII, NO. 3, EF 20 Scientific Proceedings, Royal Dublin Socvety. with solutions of sodium chloride, having a wide range of concentration, to ascer- tain the effect of concentration of the sodium chloride on the rate of aeration, Experimental. The de-aeration of the water employed in these experiments was effected by distillation in vacuo. In order to obtain water quite free from atmospheric gases, the distillation was at first carried on in a very slow current of hydrogen, or of carbon dioxide; but, although the resulting water was air-free, 1t was always found to contain undesirable traces of hydrogen or of carbon dioxide, according to which gas was employed during the distillation. It was consequently decided to rely upon distillation, under the reduced pressure obtained with the aid of a good water vacuum pump, employing a slow current of air, filtered throngh glass wool, to overcome difficulties from “bumping.” The uitrogen content of the de-aerated water obtained did not exceed 1 c.c. per litre, which, in the case of pure water, amounts to about 7 per cent. of saturation at 15°C. It was found unnecessary to determine the nitrogen content of each sample of water at the commencement of each aeration experiment, as it was found to be practically constant immediately after the de-aeration. The glass tubes used in the experiments were about 3 cms. in diameter, and varied from 9 to 12 feet in length. T-pieces were fused on, at regular intervals, along the length of the tube, for drawing off samples of the water columns at different levels, and the lower ends of the tubes were sealed. The upper end of the tube was closed by a rubber stopper, fitted with an inlet and outlet tube, by means of which connexions were made to the pump and distilling apparatus ; so that the water could be distilled directly into the experimental tube when desired. In the case of salt solutions, the required quantity of sodium chloride was placed in a five-litre flask, connected with the vacuum distilling apparatus by means of a thick-wall rubber tube, and the required volume of pure water was allowed to distil into the flask and dissolve the salt, the exact concentration of the solution being subsequently determined by analysis. A screw clip was then used to close the connexion with the pump, the flask detached with its rubber connecting tube, and attached, with the aid of the same connecting tube, to the experimental tube. When the free end of this connecting tube was completely filled with freshly distilled water, it could be attached air-free to the previously exhausted experimental tube. After attaching the flask to the tube, the latter could be filled by inverting the flask and unscrewing the clip, when the solution flowed into the tube. Experimental tubes were also employed, which were furnished with quill capillary tubes, instead of side tubes, for drawing off samples. These tubes passed upwards through a rubber stopper at the bottom of the experimental tube to different levels inside, the external ends being attached each to a piece of thick-walled rubber tubing closed by a serew chip. - Each experimental tube was attached to a stout lath of wood for facility of removal from one Jaboratory to another. he room in which the aeration was carried out was one which received no direct sunshine nor heat from artificial sources, so that the temperature prevailing in it was not subject to sudden fluctuations. In the earlier experiments, a zinc cylinder, 14 inches in diameter, and filled with water. was used as a jacket for the tubes. But it was subsequently found that a good wrapping of asbestos cloth was a sufficient guard against sudden variation in Aprnry, Leonarb, AND hicHarpson—Aeration of Water. 21 temperature. Manipulative difficulties were thereby considerably reduced, and it was possible to work with a greater number of tubes simultaneously. A recording thermometer was kept in the room, and the greatest variation in temperature in the room, during the exposure of any one tube, did not exceed 4:5 degrees centigrade. This variation, however, was exceedingly slow—too slow to give rise to convection curreuts. The air passing over the surface of the water in the tubes was first dried by calcium chloride; and the inlet tube was so arranged that the air did not play directly on to the surface. The inlet and outlet tubes were fixed at a distance of 1:5 to 2inches above the water. The moist air issuing from the tube was passed through weighed CaCl, tubes, so that the amount of water evaporated could be determined for each tube. In most of the experiments three tubes were connected in series, drying tubes being placed so that the moist air from one was dried before entering the succeeding tube. A filter pump, worked by a constant head of water, seven feet in height, was at first employed to aspirate the current of air through the tubes, but subsequently, owing to shortage of the town water supply, caused by the drought last summer, a different arrangement had to be substituted. An electrically driven small air pump was utilized to force air into a large glass vessel, which acted as an equalizer, and thence through the drying vessels and experimental tubes. This arrangement worked very satisfactorily. The experimental tubes, having been filled with de-aerated water, were fixed in a vertical position; the stoppers, which had been used during filling, were removed, and were replaced by others carrying inlet and outlet tubes for the air current. The air current was continued for two or three weeks in the case of salt solutions, and three to eight weeks for pure water. The apparatus employed for the determination of the dissolved gases was of the form devised by one of the authors.’ Method of withdrawing Samples from the Experimental Tubes. Samples of water were withdrawn from an experimental tube, without at any time allowing them to come into contact with the air, with the aid of a modification of the gas burette used for the analysis of the dissolved gases. By lowering the mercury reservoir, attached to the burette, a known volume of water was drawn from the tube into the latter, and thence transferred to a Plimpton gas holder to await examination. The depth from which the sample was drawn and the temperature of the water were at the same time noted. 50 c.c. of water were usually taken for the extraction and analysis of the dissolved gases in the case of the surface layer; and 100 cc. for samples drawn from lower levels. Determination of Satwration Values for Nitrogen of Sodiwm Chloride Solutions. In order to calculate in percentages of saturation the observed rates of solution of atmospheric nitrogen by solutions of sodium chloride of the various concentra- tions employed in this investigation, it was necessary to determine the saturation values for atmospheric nitrogen of each solution. This was done by filling large tubes, about 5 cms. in diameter and 30 cms. in length, about two-thirds full of 1 Sci. Trans. R.D.S., vol, v, Series 11, p. 548; also Supplemental vol. vi, Fifth Report of the Royal Commission on Sewage Disposal, p. 99. FQ °/, NaCl. 22 Scientific Proceedings, Royal Dublin Seciety. the salt solution. A current of air, previously passed through a glass wool filter and through distilled water to saturate it with aqueous vapour, was drawn through the solution under examination. The inlet tube reached to the bottom of the solution; and the apparatus was immersed in a thermostat. The current of air was continued for a sufficient time to ensure equilibrium being reached throughout the tube at the observed temperature. The dissolved nitrogen and the sodium chloride in the solution were then estimated. The experimental results are shown by the accompanying curve (fig. 1). ————— 30 IL ee == II 12 13 4 Saturation value of Nitrogen in c.es. per litre. Fig. 1.—Saturation values of Nitrogen in solutions of yarying NaCl concentration at 14°C. Method of expressing Experimental Results for the purpose of Comparison. The columns of water employed in these experiments were so deep—about ten feet -- and the exposure of them to the air had to be continued over such extended periods of time—two to eight weeks—that it was decided not to attempt to control the temperature obtaining during the conduct of the experiments beyond preserving the room in which they were carried out, and which was exceptionally well circumstanced for such purpose, at as closely uniform temperature as possible during a set of experiments. To have provided the means for controlling, at will, the temperature at which these experiments could be started and continued would have added very considerably to the practical difficulties of an already sufficiently laborious problem ; and it was not thought necessary to attempt to do so, since the authors had good reasons, which will be explained later on, for assuming that when the gas concentrations of a de-aerated column of distilled water, or of a solution of Apenny, Lyonarp, any Ricwarpson— Aeration of Water. 28 salt, at any time during re-aeration were expressed in percentages of saturation, they would be found to increase practically in the same proportions at different temperatures, provided that the variation in temperature did not exceed three or four degrees centigrade. This view was based upon the following considerations :— (1) Dittmar has determined the saturation values of distilled and of sea-water for atmospheric nitrogen and oxygen between 4° C. and 35° C.; and when his results are plotted against temperatures, the curves obtained approximate to parallel straight lines between the limits 8° C. and 35° C2 (2) Adeney and Becker, in their work on the rate of solution of these alinospheric gases by distilled and by sea-water, found that, when the experi- mental observations from zevo to saturation were expressed as percentages of saturation and plotted against time, the curves obtained were coincident when uniform conditions of exposure of water to the gas, of the mixing of the exposed with the unexposed portions of the water, and of temperature (within 1° C.) obtained.* It may be assumed from this that the curves showing the rate of solution of atmospheric nitrogen and oxygen by dilutions of sea-water with varying propor- tions of distilled water would also be coincident with the curves showing the same for distilled and sea-water, separately, under like conditions of exposure to air, of mixing, and of temperature varying within 1° 0, - (3) Adeney and Becker further found that the curves showing the rate of solution in water of nitrogen and of oxygen, when stated in percentages of saturation, lie very closely together ‘for differences of 5°C. within the range of temperature from 0° ©. to 30°C Consequently it is possible to convert approximately, by simple calculation, observations made at slightly different temperatures, to those that would obtain at a selected common temperature, varying even to as much as 2 or 3 degrees from them, and still obtain sufficiently accurate results for purposes of com- parison. Supplementary vol. vi to the Fifth Report, Royal Commission on Sewage Disposal, p. 59. “The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen by Water.” By W. HE. Adeney and H. G. Becker, Part 1, Sci. Proc. B.DS., 1918, and Phil. Mag., 1920, p. 385. [ TABLE. Scientific Proceedings, Royal Dublin Society. 24 | | | | | 1g = 13 GS ~ 1G 0g | Ge | i le = bce “+ “sy $% dad av yo sarqy | | | | | | | | eee FELT = IF 9SF S68 Gig || Corman ee cES FZO1 = = 19.10 UMULp IB JO S1qVT | | LFS ae | te The 69% 695 | Lig Lg axe | axe Le “suid UT UU [09 Jo WAsUeT 62-E1 1é-¢T 140-€1 16:31 66-81 96-81 60-81 160-81 61-21 86-61 1B-31 ** ‘onty aad +s9°0 ur queur ! 0 -tiadxe jo O.1NjVI9CI9} | ye Ny FO anya uotRinyzeg es = FEET | 16-81 18-€1 GLEl | 1-80 60-81 12-1 | ¥F- IT Hiei |) °° is ‘att aad | | "so°0 ul “0 oGL qv o9Aand | WO.1}* Nf JO anya uoTyBINZeVg Sera = ge¢e-8 | 86-2 G16-9 GE9-L | G1G-9 G18: 01 Os Cae |e = = 5 ur ‘daa 1938 | | | 9¢ De BE 1G 61 SI $I SI SI | Bit ial ii “* SsAup UT out, Oo YLT GAO |) FEEL i FO) cliKe 0: oF 0G ‘O 06-91 ies hei HO eee) Fo) Persie || 10) exalt "0 01 ‘sojduus Sure) uaa “dway, OD ol PI 9) ||) FOS TAN ‘O oS LT | (0 o%-LT *O 00-F1 *O of 81 | a 1D fT ‘0 o8-IT "OD SIT | (du ies way woorjoduray, | | | | | 0 0 | 910-0 | 620-0 890-0 | 9T1-0 £60 | TF-0 ¥6-0 | 16-1 608-5 ss “1OUN Jo asvquaoag | | | | c9.% G6:9) 3 Fh || 9 Ler, 96-F 99-F Hep ho Ge |) Pie GG:9 €8-8 “+ fuoruinges Jo a5uzue9 | | -aad 8B" Jo “9U09 [VIVUT | | | $19-0 90-1 819-0 | $190 819-0 819-0 819-0 | &99-0 99-0 | %99-0 6-0 rae ‘oy red | | *g9°9 UL NT JO “UdUOD [VITO | | | | *q ai | aS | 8S AS | 98 | °S us &S as) i fs Avrenry, Lronarp, AND RictArDsSoN— Aeration of Water. 25 Huperimental Results. The conditions obtaining in some experiments with columns of salt solutions and of distilled water are given in the table on opposite page. Curves showing the nitrogen content of columns of distilled water, and of salt solutions. at different depths, expressed as percentages of saturation, are given in =h g. 2. On comparing the curves D, for distilled water (exposed to the air for twenty-seven days) and §,, for a 3°3 per cent. salt solution (fourteen days), the effect of the salt on the progress of aeration is well illustrated. The curve for the distilled water column shows a fall in nitrogen content from +0 per cent. at the surface to 13 per cent. at a depth of 311 ems., whereas the corresponding values for the 3:3 per cent. salt solution ranged from 42-7 to 38-3 per cent. respectively in about half the time. °/, Saturation of Nitrogen. Jt 50 loo iso 200 250 300 Depth in centimetres. Fic. 2.—Nitrogen content expressed in percentages of saturation at different depths below the surface. Even with a solution containing only 0:0116 per cent. sodium chloride (curve S,), the effect of the salt in solution is well marked. The curves §,’, S8.’, Sy, ‘l', and T, ave plotted from results obtained from observations with columns of sea and tap water, 3 feet deep and + cms. diameter. The columns were exposed to the air at temperatures 14°, 12°, 13°5°, 13° and 13°5° C. for 7, 6°75, 11,4, and 7 days respectively. Samples of the salt solutions from the top and bottom layers of the columns, after aeration, were carefully examined by means of a Pulfrich refractometer to ascertain whether any difference in concentration of the sodinm chloride exis‘ed between the two layers, but no difference could be detected, \ 26 Scientific Proceedings, Royal Dublin Society. In order to compare more clearly the experimental results obtained for the salt solution, an approximate correction for slight variations in the length of time of their exposure to the air has been made. Eighteen days was taken as the standard time, since four out of the nine tubes were exposed for that time. Adeney and Becker's formula for calculating the rate of re-aeration was employed, viz. :— ; a [cn ee raat w= (100 = wy) l-e v jh where w = amount of gas dissolved, expressed in percentage of saturation. w, = initial concentration in percentage of saturation. f = coefficient of escape of the gas from the liquid per unit area and volune. v = volume of liquid. R a = area of surface. ¢ = time of exposure. IL ° ra SP S3 (043 9oMO ° S7(oabafoNo Cl) ° , Saturation of Nitrogen. 55 —— z S2 (1 q7}oNacir ee Sb(0n6%NoCt) °o = = eo, E 3 ° ———— 5 ; aa $5 (0 25nNoCl) Q o 3, % No() alc eee ° NaCl 425 [ a IL E 5, GSB NaCl) —|} ° so 100 150 200 250 300 Depth in censimetres. Fig. 3.—Concentration of Nitrogen at various depths after 18 days’ exposure for different salt solutions. By means of the above equation, if w be known for any time #, its value can be found for any other time ¢. ‘The mean value of the concentration of the gases in solution in the liquid was taken to be 2. The actual values at various depths were afterwards calculated. Curves were plotted for the columus of solutions, using observations made after eighteen days. ‘They are given in fig. 3. The values for the concentrations of nitrogen at a depth of 200 ams. were taken, and these values were plotted against the concentration of salt. The maximum concentration of nitrogen was found to obtain in a solution with a Aprnuy, Luonarp, anp RicHarpson—Aeration of Water. 27 concentration of about 1 per cent. of sodium chloride. The curves obtained are given in fig. 4. _ll reall °/, Saturation of Nitrogen. ° 10 20 30 40 °/, Nact. Fic. 4.—Relation between NaCl concentration and downward ‘‘streaming’’ of dissolved Nitrogen to depths of 200 cms. through solutions of that salt. > Conclusions. The following conclusions may be drawn from the experimental results obtained in this investigation :— 1. The aeration of quiescent bodies of water, fresh and salt, under natural conditions is effected by a process of mixing of the exposed layer with the unexposed portions of the water to depths of at least 10 feet. 2. The process of mixing is caused by the downward “streaming” by the constantly changing layer of water exposed to the air. This downward “streaming” sets up a process of mixing certainly to depths of 10 feet, and in all probability to much greater depths. 3. The process of mixing set up by the downward “streaming ”’ proceeds more rapidly, and more uniformly downwards, to depths of at least 10 feet, in salt water than in fresh water. SCIENT. PROC. R.D.S,, VOL. XVII, NO. 3. G 28 Scientific Proceedings, Royal Dublin Society. 4, The rate at which the “streaming” proceeds depends Jargely, though not wholly, upon the rate at which the concentration and cooling of the surface layer of the water is brought about by evaporation from it. 5. The process of “streaming,” and consequently of mixing, also proceeds more rapidly at temperatures at and above 10°C. than below it. It is distinctly less rapid and less uniform downwards to 10 feet deep, and probably to greater depths, at temperatures below 8° C., especially in fresh waters. 6. The rate at which the process of mixing from “streaming” proceeds also depends upon the concentration of salt in solution, The optimum concentration appears to be about 1 per cent. sodium chloride. This last conclusion is based upon one series of experiments only, and should be confirmed by further experiments which the authors hope shortly to carry out. CHEMICAL DIvIsIoN, ROYAL COLLEGE OF SCIENCE, DUBLIN. Ew] ON A PHYTOPHTHORA PARASITIC ON APPLES WHICH HAS BOTH AMPHIGYNOUS AND PARAGYNOUS ANTHERIDIA; AND ON ALLIED SPECIES WHICH SHOW THE SAME PHENOMENON. By H. A. LAFFERTY AND GEORGE H. PETHYBRIDGE, Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (PuatTes I anp 11.) {Read May 28. Printed Junz 28, 1922. TE work described in the present paper originated from an examination of a couple of apples of the variety “Lane’s Prince Albert,’ which were affected with an unusual form of rot, and were submitted for report in November, 1920, by Mr. E. Turner, one of the Department’s Horticultural Inspectors, from Pilltown, County Kilkenny. The diseased fruits were apparently healthy when gathered on October 14th as portion of a lot specially selected for exhibition purposes, and they began to rot about ten days later. Whether infection occurred prior to gathering or during storage is not known. I.— Nature of the Rot. The decayed apples had dark brown skins, but they were more or less firm and elastic to the touch. They showed no signs of superficial wounds, nor were any external indications of fungus growth visible on them. After they had been kept under a bell-jar in the laboratory for a few days, however, small white tufts of hyphae were present at some of the lenticels. ‘hese proved to be composed of non-septate mycelium, which was almost entirely sterile ; but prolonged search ultimately resulted in the discovery of two sporangia, suggestive of a Phytophthora. Further examination, made at a later date, revealed the presence of a few sexual organs, borne on the mycelium in the basal portions of several of the tufts. The oogonia averaged 26-4 and the oospores 244 in diameter. ‘I'he surprising thing about the sexual organs was that they were of two types. In the majority of cases the antheridia were lateral! (paragynous), but in a few instances they were observed to be of the type first discovered in Phytophthora erythroseptica, i.e., sur- rounding the base of the oogonium (ampiigynous).? The flesh of the diseased apples was brown in colour, fairly firm, and not pulpy. 1 Not necessarily lateral in the strict sense of the word, because such antheridia are frequently situated more or less below or at the base of the oogonium near the stalk, but not surrounding or penetrated by the latter. * The convenient terms, ‘‘amphigynous” and ‘‘ paragynous,”’ for expressing the position of the antheridium in relation to the cogonium, were suggested by P. A. Murphy, Ann. Bot., xxxii, 1918, p. 125. SCIENT. PROG. R.D.S., VOL. XVII, NO. 4. H 50 Scientific Proceedings, Reyal Dublin Society. In microscopical preparations it was found that rather coarse, non-septate mycelium permeated the tissues, being present both in and between the cells. White aerial mycelium was present in the cavities of the core, but it bore neither sporangia nor sexual organs. ‘Ihe latter, however, were found embedded in the soft tissues above and below these cavities; and antheridia of both the amphi- and paragynous types were present, the latter in much greater number than the former. Eighteen small pieces of affected tissue were removed aseptically, and each was placed in the centre of a Petri dish, in which a sterile nutrient agar medium had previously been allowed to set. In all cases mycelium grew from the tissue into the medium; but, in the early stages at least, with one exception, the mycelium remained barren. This exception was Quaker Oat agar. On this medium the fungus developed freely, and sexual organs with both types of antheridia were produced on the mycelium in close proximity to the original piece of tissue, whilst they were also present in the tissue itself. Photographs of each type are reproduced on P]. II, figs. 1 and 2, and the two types are figured on PI. I, figs. 1 and 2. Sub-cultures were prepared from the growth on Quaker Oat agar, and from them inoculations were made into healthy apples, the necessary controls being also provided. A rot identical with that found in the original material was produced in every case where the inoculum was introduced, whilst the controls remained sound. It appeared, therefore, that the rot was caused either by a single species of fungus having sexual organs of two types or by one or both of two associated species, each producing its own particular type of antheridium. In any case, the matter was of more than usual interest, and demanded further study. I1.—Tsolation of the Fungus and Proof of Pathogenicity. An attempt was first made to discover by direct microscopical observation whether the two types of sexual organs were borne on one and the same portion of mycelium, but this attempt gave only uncertain results. It was therefore decided to raise pure cultures. (a) Isolation from portions of single hyphae—When radiating growth had proceeded for some time on the Quaker Oat agar Petri dish cultures, mentioned above, a narrow circular band of the medium was removed just beyond the limit of growth. When the tips of the growing hyphae reached the inner margin of this annular space, they proceeded to cross it; but it was found possible, by careful manipulation with sterile needles under the low power of the microscope, to prevent the crossing of this space by all hyphae in a given region save by a single selected one. When the latter had crossed and had become well established in a completely isolated condition in the medium on the other side of the annular space, it was removed, together with a small portion of the medium in which it was embedded, and placed on a suitable medium slanted in a test tube. In this way nine pure cultures were raised, each derived from an isolated portion of a single hypha. As a means of isolating both members of a pair of fungi intimately associated with each other during growth, the above described technique is undoubtedly open to criticism. If, for example, the rate of growth of one of the pair were greater than that of the other, the former would invariably reach the annular space first, and the cultures derived, as described above, would almost certainly represent only one of the two organisms. Nevertheless, to obtain even one of them in pure culture was a step in advance, and, as the sequel will show, the method adopted served its purpose. Larrerty AND Pseruyspripgu—On a Phytophthora Parasitic on Apples. 31 Sub-cultures of the nine pure stocks raised were made in parallel series in various media, and close study of them showed that they were identical in all respects. The production of sexual organs, however, on the media employed was extremely limited. Cases were found where the antheridium was undoubtedly paragynous, and certain others occurred in the same cultures in which this organ appeared to be amphigynous, but absolute certainty on the matter could not be arrived at. In order to clear up this important point it was necessary to make cultures on a medium in which the sexual organs were produced in greater abundance. For this purpose cylinders of tissue, cut with a sterile cork-borer from a healthy raw apple under aseptic conditions, were placed in sterile plugged test tubes. One-half of the number of cylinders was inoculated from a sub-culture of one of the nine stocks alluded to above; the other cylinders were left uninoculated to serve as controls. After the lapse of about a month the cylinders were examined. The controls had remained sound, and no trace of fungus or other growth could be discovered on or in them when examined microscopically. The inoculated cylinders had rotted and become brown, while traces of aerial mycelium were visible on them. Microscopical investigation revealed the presence of no sporangia, but sexual organs were present in abundance. ‘The antheridia in the vast majority of cases were paragynous; but in two instances, at least, antheridia undoubtedly of the amphigynous type were present in the same culture along with them. A photograph of one of these is reproduced in fig. 6, Pl. II. It was thus fairly clear that only a single fungus was responsible for the apple rot, and that this fungus produced both paragynous and amphigynous antheridia. (6) Isolation from a single sporangiuvm.—tThe fungus does not produce sporangia at all abundantly, but in an old growth on oat-extract agar, derived from a piece of tissue removed aseptically from one of the original decaying apples, a few were found. The culture was in a Petri dish, and at the time was not free from bacterial contamination. In the tissue from which the growth had emanated sexual organs of both types were present. Under microscopical control six isolated sporangia were successfully removed, one at a time, by means of sterile capillary tubes, and each was washed in about twenty changes of sterile water to remove bacterial contamination as far as possible. On a healthy apple, and with aseptic precautions, six small, well-isolated cut surfaces were prepared, and on each a single sporangium was placed. A similarly prepared healthy apple, on which no sporangia were placed, was used as a control, and both fruits were kept in a moist atmosphere under the same bell-jar. Within sixteen days a rot of the first apple had started at three out of the six places where sporangia had been introduced. No rot occurred at the other three places, the sporangia in these cases being apparently not viable. The uninoculated apple remained perfectly sound. Before the rot had extended from the three centres sufficiently to become coalescent, several small portions of the affected tissues were removed from one of the centres to set media in Petri dishes, aseptic precautions being observed as before. After a few days mycelium grew into the media from the pieces of tissue, and in each case the growth was found to be free from bacterial contamination. From these growths—each the product of a single sporangium—sub-cultures in test tubes were prepared. At a later stage the pieces of tissue on the Petri dishes were examined, and H 2 32 Scientific Proceedings, Royal Dublin Society. sexual organs with both para- and amphigynous antheridia were found in them, the latter being decidedly in the minority. One of each type, isolated from this material, is illustrated in figs. 3 and 4, Plate IL. (c) Isolation from a single oospore.—Small portions of the affected tissues which had been removed from one of the original diseased apples to solidified media in Petri dishes were macerated in water, and sexual organs having antheridia of the amphigynous type, containing apparently ripe and viable oospores, were sought out under the microscope. Several were found and were removed by capillary tubes, one at a time, to films of oat-extract agar on cover-glasses, which were then inverted over excavated microscope slides in the usual way. Before removal, great care was taken to ensure that no living hyphae remained attached either mechani- cally or otherwise to the oogonia and antheridia. Hach slde thus prepared contained a single oospore ; but the films, although quite free from any other fungus, were not altogether free froin bacteria. The slides containing the oospores were kept under close microscopical observa- tion. In the majority of instances the oospores either remained in a resting condition or germinated and produced short germ tubes which soon died. In one case, however, greater success was achieved. The inner portion of the thick wall of the oospore was observed to dissolve in a manner similar to that which occurs during the germination of the oospores of Phytophthora erythroseptica. A germ- tube then arose which passed through the wall of the oogonium, grew for a short distance, and then developed a sporangium at its distal end. This sporangium produced a number of zoospores which, after actively swimming about for some time, came to rest, developed thin walls, and proceeded to form germ-tubes. At this stage the cover-slip containing the film culture was carefully inserted into a slit made in the flesh of a healthy apple, which was then kept for thirteen days in a moist atmosphere at laboratory temperature. At the end of this time a rot had developed in the apple, starting from the film side of the cover-glass, on which the germinating zoospores were originally present. After seventeen days portions of the rotted tissues were removed (aseptic precautions being still maintained) and were placed on solidified oat-extract agar in Petri dishes, and from the growths which arose in the agar medium sub-cultures were prepared. No sexual organs or sporangia had developed in the growths at the time the sub-cultures were made; but microscopical examination of the portions of tissue themselves revealed the presence in them of sexual organs with both amphi- (Plate IJ, fig. 5) and paragynous antheridia. From the foregoing it will be seen that pure cultures were raised by three different methods, and that in all cases a single fungus was derived which produced sexual organs, having both amphi- and paragynous antheridia. It might be maintained that the use of raw apple tissue in the course of the work left the question as to the absolute certainty of the results in some doubt; and, in order to obviate this, sterilised carrot tissue was subsequently employed. The results obtained were the same as before. As regards pathogenicity, the fungus in pure culture has repeatedly been inoculated into healthy apples, and always with the same result, viz., the produc- tion of a characteristic form of rot, similar to that exhibited by the original naturally infected fruits. In each case, where the attempt was made, the fungus was re-isolated in pure culture from the artificially infected apples. Non-inocu- lated controls, kept under similar conditions, remained unaffected in every case. In the diseased tissues of both naturally and artificially affected apples sexual organs occur, but not abundantly. When small portions of such tissue, however, Larrerty AND Pernysripge—On a Phytophthora Parasitic on Apples. 88 are removed to media, or even to sterilised moist filter paper in Petri dishes, sexual organs are produced within the tissues in much greater abundance. Possibly this may be the result of an increased supply of oxygen; but the matter was not further investigated. The fungus is also pathogenic to pears. When inoculated into unripe fruits a firm brown rot quickly develops. On keeping them in a moist atmosphere tufts of mycelium, sometimes bearing sexual organs, but not, so far as was observed, sporangia, arise from some of the lenticels. Sexual organs are also developed within the rotting tissue. As regards potato tubers, repeated inoculations invariably gave negative results ; hence the fungus is not pathogenic to the potato. I1.— Cultural Characteristics and Morphology of the Fungus. The fungus has been cultivated on a number of media, parallel cultures having been made with stocks raised by the various methods already described. Ona given medium, as was to be expected, the behaviour of the fungus has always been the same, regardless of the particular method by which it was isolated. On Quaker Oat agar the growth was submerged, creeping, or very slightly raised above the surface of the medium. Sporangia were few, and occurred in small localized tufts on long and rather ill-detined sporangiophores. Oospore production was very scanty, or even entirely suppressed. The oogonial wall was yellow where embedded in the substratum, but hyaline when formed on or above its surface. Yellow-walled oogonia generally imparted a faint tinge of yellow to the walls of the oospores within them. On oat-extract agar the growth was sparse, creeping, or submerged. Sporangia were few, and sexual organs absent. On this medium the mycelium, when examined microscopically, had a characteristic appearance. Towards the centre of the growth the hyphae were distinctly tuberculate, with numerous short lateral branches; while at the margins each radiating hypha branched copiously, pro- ducing a dense fan-like growth. Cooked prune-extract and cooked apple-extract agars proved to be unfavourable media for the fungus. The limited growth on each was submerged or raised slightly above the surface. Sporangia were absent, and sexual organs almost completely so. On cooked potato stalk and cooked potato plugs the fungus was completely sterile and confined to the tissues, while on cooked potato agar a very sparse and sterile growth was produced, which resembled that on oat-extract agar. Cooked or raw apple plugs and cooked carrot plugs were the media on which the best production of oospores was obtained; but on none of these media were sporangia observed. The sporangia, which, as has already been stated, are produced only sparingly by this fungus, are borne on long hyphae which branch in a sympodial manner, typical of the genus Phytophthora. Each sporangium is spherical when young, but becomes inversely pyriform as it matures. The wall is thin, except at the narrower free end, where it is thickened, but distinctly hyaline. The free end is blunt, and not papillate ; whilst at the basal end of the detached sporangium there remains a small portion of the parent hypha on which it was borne. In size the sporangia varied greatly, but were found to average 40u long by 27 broad. A single normal sporangium is illustrated on Plate I, fig. 15, and Plate II, fig. 16. When mature sporangia were placed in tap-water, a certain number of them 34 Scientific Proceedings, Royal Dublin Society. liberated their zoospores after twenty minutes. The broad, apical, hyaline, thickened portion of the wall became suddenly stretched and expanded to form a sac-like structure, while the zoospores rushed out in mass, partially filling it. Almost simultaneously the thin wall of the sac became ruptured, allowing the zoospores to escape, whilst the wall itself quickly became unrecognizable. The zoospores, which are typical of the genus Phytophthora, are lemon-shaped, and each is provided with two cilia. After some time the zoospores came to rest, rounded themselves off, and each produced a single germ tube. Individual sporangia were seen to produce from four to twenty-six zoospores; and the openings at the ends of the empty sporangia averaged 9 in breadth. In water a few sporangia produced a single germ-tube each, which generally emanated from a point a little to one side of the hyaline apical region. In many cases the germ-tube, after growing for a distance equal to the length of the parent sporangium, developed at its extremity a second sporangium, which, in turn, germinated by producing a germ-tube, and, as before, a terminal sporangium. In this way short chains of sporangia, gradually diminishing in size, were very frequently formed ; and all, with the exception of the terminal one, were devoid of contents. The sexual organs—Specimens with paragynous antheridia resemble im shape similar organs figured from time to time for species of the Cactorum group. The oogonia are pear-shaped, and arise as terminal swellings on rather short lateral hyphae. The wall of the oogonium is thin compared with that of the oospore, but considerably thicker than that of the hypha on which it is borne. The average diameter of oogonia of this type from pure cultures of the fungus on artificial media was found to be 28p. The antheridia ave small, irregularly shaped terminal swellings on short hyphae, which spring from the parent oogonial hypha or from a ueighbouring one. Antheridia were occasionally observed with one or more short, finger-like, hyphal outgrowths, which gave them a false appearance of being intercalary in origin. In the early stages of sexual reproduction an antheridium and a developing oogonium meet; the former becomes firmly attached to the wall of the latter at any point on its surface, but perhaps most frequently somewhere in the region of the oogonial stalk. During fertilization only a portion of the contents of the antheridium passes into the oosphere. The oospore, which partially or almost completely fills the oogonium, is generally hyaline, spherical, thick-walled, and filled with protoplasmic contents, in which are to be seen a large central oil-drop and a smaller highly refractive body. Oospores from sexual organs of this type were found to average 25°4u in diameter, and their walls varied from 1:5 to 24 in thickness. The sexual organs of the second type develop in asimilar mannertothat described for the first time in P. erythroseptica.. The oogonial incept grows through the antheridium, and on emergence swells rapidly. ‘The protoplasmic contents gradually contract to form an oosphere, and, after fertilization, an oospore. The average diameters of oogonia and oospores from sexual organs of this type were found to be 27°34 and 25 respectively, and the walls of the oospores varied from 1:5 to 2u in thickness. From the characters described, and as a result, of the cultural studies which will be dealt with in the following section of this paper, the fungus was ultimately recognized as being identical with Phytophthora Syringae Klebahn. Sci. Proc. Roy. Dublin Soc., xiii(N.S.), No. 35, 1918, p. 529. Larrrerry anp Peruysringe—On a Phytophthora Parasitic on Apples. 35 1V.—Previous Work on similar Types of Rot. During recent years several cases of Phytophthora attack on apples as well as on pears have been recorded both in Europe and in America. Thus, Osterwalder' described one on apples in Switzerland in 1904; Marchal,? Bubak,? Unamuno,* and Schoevers’ reported attacks on pears in Belgium. Bohemia, Asturia, and Holland respectively at various times from 1908 to 1915; Whetzel and Rosenbaum! recorded an attack on apples in America in 1916 ; Wormald’ dealt with attacks on pears and apples in Eneland in 1919; Giissow’ noted an attack on pears in Nova Scotia in 1920, while in the same year Clinton® gave particulars of attacks on both apples and pears in Connecticut, U.S.A. In all these instances the rot was attributed either to Paytophthora Cactorum (Leb. et Cohn) Schr6t. or to its synonym P. omnivora de Bary,” but in no case was any intensive study of the sexual organs of the fungus made, nor was the existence of two kinds of antheridia suspected or discovered. At the outset it appeared possible that the rot in the Irish apples was caused by a new species of Phytophthora, because neither in P. Caciorwim nor in any other species of Phytophthora hitherto described had the production of two kinds of antheridia in one and the same species been noted. On the other hand, it was not impossible that antheridia of the two types really did occur in P. Cactorum, and perhaps in other species of the genus, but that the fact had merely been overlooked or not apprehended by previous workers, owing to the comparative scarcity of cases in which the antheridia are amphigynous as compared with those in which they are paragynous. This latter surmise was strengthened as a result of a consideration of the relevant literature. Thus, Hartig,! in his description of the development of the oospore in P. Fagi, states that in exceptional cases the stalk of the oogonium swells out directly under the oogonium itself to form the antheridium, and his figure of this condition (Taf. IL], fig. 24b) suggests very strongly that this was an instance in which the antheridium may have been amphigynous. Butler and Kulkarni,” after referring to Hartig’s figure and to the statement of Himmelbaur,” that in P. Fagi the antheridium is attached to the underside of the oogonium near its base, says : “It appears probable that a penetration of the antheridium by the oogonial stalk sometimes occurs in this species.”’ 1Centralb. f. Bakt., ii, xv, 1906, p. 485. 2 Bull. Soc. Roy. de Belgique, xlv, 1908, p. 343. 5 Zeitschr. f. Planzenkrankheiten, xx, 1910, p. 257. 4Zeitschr. f. Pflanzenkrankheiten, xxi, 1911, p. 379 (abstract). © Tijdschr. over Plantenziekten, xxi, 1915, p. 153. 5 Phytopathology, vi, 1916, p. 89. 7 Ann. App. Biology, vi, 1919, p. 89. 8 Phytopathology, x, 1920, p. 50. * Conn. Agric. Expt. Station, Bull. 222, 1920, p. 406 and p. 464. 10Tt will be remembered that de Bary in 1881 assembled Schenk’s Peronospora Sempervivi and Hartig’s Peronospora Fagi under the new name Phytophthora omnivora ; and, later in the same year, also included Lebert and Oohn’s Peronospora Cactorum under the same name. (Beitr. z. Morph. u. Phys. d. Pilze, iv, 1881, and Bot. Zeit., xxxix, 1881.) Himmelbaur, however, has shown that P. Fagi and P. Cactorwmare dissimilar (Jahrb. Hamb. Wiss. Anst., xxviii, 1910), and for this reason as well as on priority grounds the combination P. omnivora should no longer be used. 11 Untersuch. forstbot. Inst. Miinchen, i, 1880, p. 49. 12 Mem. Dept. Agric., India, Bot. Series, vy, 1913, p. 257. 138 Jahrb. Hamburg Wissensch. Anstalten, xxviii, 1910, p. 450. 36 Scientific Proceedings, Royal Dublin Society. Again, some of the figures illustrating de Bary’s' own account of P. omnivora, especially figs. 25, 26, and 27 on Taf. {11 (concerning which it is stated that the point of origin of the oogonium was hidden by the antheridium), are suggestive, at least, of cases where the antheridia may have been amphigynous. Rosenbaum,” in dealing with P. Cactorwm isolated from diseased Ginseng, figures the sexual organs with paragynous antheridia only; but he states that “in some cases the stalks bearing the antheridium and oogonium are on the same side, and the antheridium then falls on the oogonial stalk. Under the microscope, such a condition may present the appearance that the oogonium has grown through the antheridium.” It seems quite likely that the “appearance” in the cases noted may, indeed, also have been the reality. In view of the uncertainty which appeared to exist in the matter, it was considered desirable to make renewed studies of P. Cactorwm and P.Fagi, and to compare the fungus isolated from the Irish rotted apples with these species, and also with others isolated from decayed apples and pears in other countries, so far as it was possible to obtain them. Later on it was found necessary to re-examine P. Syvingae. As soon as it had been established clearly that the two types of sexual organs found in the Irish apples were produced by one and the same funeus, the attention of mycologists was called to the matter in a letter to “Nature,”* and an appeal was made for cultures of species of Phytophthora from other workers. From Dr. W. F. Bewley, of the Experimental and Research Station, Cheshunt, Herts, a culture was received provisionally regarded by him as one of P. Cactorwm. The fungus had been isolated from a rotten apple purchased in Cheshunt. Professor H. H. Whetzel was good enough to send a culture of what he regarded as P. Cactorum, isolated from a decayed apple grown in his own garden in Cornell University, Ithaca, New York State, U.S.A. Dr. A. Osterwalder kindly sent from the Swiss Research Station for Fruit, Vine, and Garden Cultivation in Wadenswil a pear affected with a Phytophthora-rot, from which what appeared to be P. Cactorum was isolated without difficulty. Cultures of P. Cactorwm and P. Fagi (stated to have been isolated by Peters from Cactus and Lagus sylvatica seedlings respectively) and one of P. Syringae (originally isolated in Hamburg by Klebahn from Syringa vulgaris) were obtained from the Centraalbureau voor Schimmelcultures in Baarn, Holland, through the kindness of Dr. J. Westerdijk. Including the species isolated from the Irish apples, there were thus seven cultures of allied Phytophthoras, each from a different source; and these formed the basis of a series of sub-cultures on various media, which were stadied in considerable detail. Limitations of space preclude the description 7 eatenso of this comparative cultural work, and only the most important results can be dealt with here. In the first place, it may be stated that each of these fungi, when inoculated through wounds into healthy apples and pears, was found to be capable of setting up in them a characteristic form of brown rot, similar to that described in the present communication. Apart from minor variations, such as the number and size of the tufts of hyphae protruding through the lenticels, and so on, these forms of rot were indistinguishable from one another by the naked eye. Secondly, it was found that each of the fungi produced sexual organs of the 1 Beitr. z. Morph. u. Phys. der Pilze, iv, Frankfurt a. M., 1881, p. 22. 2 N.Y. State Coll. Agric. Cornell Univ., Bull. 363, 1915, p. 100. 3 Vol. cvii, No. 2635, April 14th, 1921, p. 204. Larrerry AND Pernysripce—On a Phytophthora Parasitic on Apples. 37 two types already described; and in all cases those with amphigynous antheridia were very much rarer than those in which the antheridia were paragynous. Illustrations of both types of these organs for the Phytophthoras under considera- tion are provided in Pl. I, figs. 1-14, and PI. II, figs. 1-12. Further, in all cases except P. Syringae and the Irish fungus, a second form of asexual reproductive body was found in varying abundance, which had not hitherto been described for P. Cactorwm or P. Fagi. With regard to the latter fungus, these bodies were not found in cultures on sterilised media, but only on cultures on living apples. These bodies, which may provisionally be called sphaero-conidia, are usually intercalary in origin, although they may sometimes be terminal. The hyphae on which they are borne are generally somewhat stouter than those which bear sporangia. They are, as a rule, uniformly spherical in shape; but, in rare instances, they develop a papilla resembling that found in a sporangium, ‘Their thin walls give the same micro-chemical reactions as those given by the sporangial walls. In diameter they vary from 33, to 40u, aud there is a tendency for them to be slightly larger when produced on living apple than on sterilised media, such as Quaker Oat agar, &c. Not infrequently short, slender hyphal outgrowths proceed from them, which, however, must not be confounded with germ tubes. Illustrations of sphaero-conidia are reproduced in fig. 15, Pl. IT.. In hanging drops of water these sphaero-conidia have been found to produce germ tubes, which may arise at any point on their surfaces. After growing a short distance a germ tube may cease to grow, or may give rise to a terminal second sphaero-conidium or to a normal sporangium. It is possible that the sphaero-conidia possessing papillae may germinate by means of zoospores; but up to the present only the germination of the much more abundant non-papillate forms has been observed. “Resting” conidia and chlamydospores have been described as occurring in Phytophthora Arecae (Colem.), P. Colocasiae But. & Kul. P. Fabert Maubl., DP. jatrophae Jens., P. Meadii M‘Rae, P. Nicotianae de H., P. parasitica Dast., P. terrestria Sherb., and P. Theobromae Colem. In some cases they are thick- walled, and are perhaps true chlamydospovres, or possibly parthenogenetic oospores. In others the walls are not so thick, and germination is sometimes by zoospores. The relation of these bodies to one another and to those described here for P. Cactorum and P. Fagi require further investigation. The sporangia of the English, American, and Swiss Phytophthoras, as well as those of P. Cuctorum and P. Fagi, resembled one another in being always papillate, the papilla being, in reality, a hyaline thickening of the sporangial wall, and forming a distinct projection or protuberance. (See Pl. I, fig. 16, and Pl. II, figs. 14 and 15.) In the Irish fungus and in P. Syringae, on the other hand, although the wall at the distal end of the sporangium is thickened and hyaline, yet there is no distinct papilla, the end being flattened or rounded. There was considerable variation in the size of the sporangia. In the case of the four first-mentioned fungi they averaged 37m x 27u on artificial media, such as Quaker oat agar, &ce.; but they were considerably larger, viz. 52u x 30m, on raw apple tissue. In the case of P. Fagi they averaged 45 x 31 on artificial media and 56u x 31u on raw apple tissue. In P. Syringae and in the Irish fungus they measured 384 x 264 and 40u x 27 respectively on artificial media, but they were not abundant enough on raw apple tissue to permit average measurements being made. Too much stress, of course, must not be laid on the sizes of sporangia as a 38 Scientific Proceedings, Royal Dublin Society. diagnostic character; but it was clear that, on the average, the sporangia of the Enelish, American, and Swiss fungi and of P. Cactorwm were considerably smaller than those of P. Fagi, whilst those of the Irish fungus and P. Syringae were intermediate in size. There was also considerable variation in the abundance of sporanginm torma- tion; but it was clear throughout the work that the Irish fungus and P. Syringae formed sporangia on the various media tried with considerably more reluctance than the five others did. Having regard to all the characters shown in cultural and inoculation trials, it was found impossible to distinguish between the English, American, and Swiss Phytophthoras and the P. Cactorwm stated to have been isolated by Peters from Cactus seedlings; and it is believed that the authors mentioned, who described the rotting of apples and pears in the instances coming under their notice as being due to P. Cactorum, were perfectly correct in so doing. That P. Fagi and P. Cactorum should not be grouped together under the name P. omnivora has already been pointed out by Himmelbaur.t The present investigations confirm this view. The sporangia of P. agi were found on the average to be distinctly longer than those of P. Cactorwm. Moreover, P. Fagi was found incapable of infecting living specimens of Sempervivum arboreum-rubrum, S. Berthelotianum, and a species of Mammilaria, whilst P. Cactorwm was patho- genetic to, and caused a rot in these plants. As the cultural work progressed it became clear that the Irish fungus was neither P. Cactorum nor P. Fagi. On the other hand, it seemed to be closely allied to, if not identical with, P. Syringae. ence, careful parallel cultures and inoculation experiments with the latter species and the Irish fungus were made, with the result that no essential differences could be found between them. The rot produced by inoculating P. Syringae into healthy apples was similar to that originally found in the Ivish apples. Further, the Ivish fungus and P. Syringae when inoculated into buds of Syringa vulgaris caused identical forms of rot; and oospores of identical character were developed in the dead tissues in both cases. Hence it is concluded that the Phytophthora which caused the rot in the Irish apples. is P. Syringae; and it is believed that this is the first record of attack by this species on apples, at any rate in Ireland.* V.—Classification of Phytophthoras. The penetration of the antheridium by the oogonial incept and the develop- ment of the oogonium proper on the summit of the antheridium in Phytophthoras were first established with certainty in the case of P. erythroseptica.’ At the same time, this method of development of the sexual organs was found to ocew in P. infestans and P. Phascoli, while it was surmised (correctly, as has since been shown by Rosenbaum) to occur also in 2. Avecae. In the paper referred to, and in 1 See foot-note, p. 35. 2 There is some reason to believe that P. Syringae occurs in Ireland on Syringa vulgaris. On two occasions within the past two years diseased Lilac leaves have been forwarded for examination on which a Phytophthora was present which, there was every reason to suspect, was P. Syringae ; but the material submitted was not sufficient or suitable for raising cultures and thus determining the species with certainty. Rosenbaum’s statement (Journ. Agric. Research, viii, 1917, p. 235), that he worked with P. Syringae isolated from Lilac in Ireland, is an error. Nor was his culture of P. Fagi derived from Fagus seedlings in Ireland. 3 Sci. Proc. Roy. Dublin Soc., xiii (N. 8.), No. 35, 1913, p. 529. Larrerry any Peraysrmee—On a Phytophthora Parasitic on Apples. 39 one! published in the following year, the various species of Phytophthora which had been described up to that time (numbering fourteen, or possibly fifteen) were enumerated. Since then the following seven species have been recorded :— P. fici Rav? P. terrestria Sherb.* P. citri Raw P. Meadii McRae?’ P. Allwi Saw.’ P. cryptogea Pethyb. and Laff. P. Melongenae Saw.’ —whilst what may possibly turn out to be three new species have been reported as attacking Paeony, Rhubarb, and Oats respectively in the United States of America,’ and another unidentified species has been found by Brittlebank, causing a disease of Papaver nudicaule, in Australia’ Thus, there would now appear to be some twenty-two species; but further study of them will probably result in a slight reduction of this number, since some of them, like P. Vheobromae and P. Faberi, P. parasitica and P. terrestria, P. Nicotianae and P. jatrophae, are probably identical. When amphieyny was first discovered, it was proposed that the genus Phytophthora should be divided into two. The generic name Phytophthora was to be retained only for those species with amphigynous antheridia, with P. infestans as type; while those in which the antheridia are lateral were to be placed in a genus tor which the name Nozemia was suggested, with NV. Cactorum as type, it being assumed, of course, that in a given species only one and the same type cf antheridium would be present. The Cactorum group then included Nozemia Ouctorwm, N. Fagi, N. Syringae, and N. Nicotianae. But it is clear from the work described in the present paper that the first three of these species cannot well remain in it. Having both amphigynous and paragynous antheridia, they constitute an intermediate group linking Phytophthora with Nozemia. On the removal of these three species the sole representative of the latter genus would now be WV. Wicotianae. It is true that de Haan’s description and illustrations of the morphology of the sexual organs in this species clearly show that the antheridia are paragynous ; nevertheless, in view of the discovery of cases of amphigyny in the other three species, it seems not at all improbable that examples of this condition might be found to occur also in WV. Nicotianae ; and further investigations on this species are highly desirable. For pure culture work, however, no medium has yet been found on which this species produces its sexual organs with certainty or in abundance; and the point cannot therefore at the moment be cleared up. Furthermore, renewed invyestivations directed to this special end might eventually show, as regards those Phytophthoras in which the antheridia ave, so far as is known at present, amphigynous, that these organs may occasionally be paragynous. In view of these and other considerations arrived at as a result of more extended research, it is now thought better that the recently erected genus 1 Journ. Econ. Biol., ix, No. 2,1914, p. 53. 7 Journ. Bombay Nat. Hist. Soc.. xxiv, p. 618. 3 Mycologia, ix, No. 4, 1917, p. 249. + Phytopathology, vii, No. 2, 1917, p. 119. ® Mem. Dept. Agric. India Bot., Series ix, No. 5, 1918, p. 219. ® Sci. Proce. Roy. Dublin Soc., xv (N.S.), No. 35, 1919, p. 487. It is of interest to note that when this species was described its sexual organs were known only from pure cultures. Since that time, however, the authors have found them in the decayed tissues of the host plant. * Science, N.S., liv,1921. p.170; Phytopathology, xi, 1921, p.55; and Science xiii, 1916, p.534. § Journ. Dept. Agric. Victoria, xvii, Pt. 2, 1919, p. 700. 9 While the present paper was in the press a reference to Ashby’s P. palmivora was noted. We have iy yet seen the paper in which it is described. (West Indian Bull., xvili, No. 1 1920, p. 61. 40 Scientific Proceedings, Royal Dublin Society. Nozemia should be abandoned. The name has as yet scarcely had time to become definitely established in the literature; and there need be therefore the less reluctance to revert to former usage, and to unite all the species in the one genus Phytophthora. Based on the mode of development of their sexual organs, the species contained in the genus may now be grouped as follows :— A.—Species in which, so far as is known at present, the antheridia when present are always amphagynous :— 1. LP. infestans (Mont.) de Bary 7. LP. terrestria Sher). 2. P. Phaseoli Thaxt. 8. P. Allw Saw. 3. P. Colocasiae Racib. : 9. P. Melongenae Saw. 4. P. Arecae (Colem.). 10. P. Mead McRae 5. P. erythroseptica Pethyb. 11. P. cryptogea Pethyb. and Laff. 6. P. parasitica Dastur B.—Species in which the antheridia are preponderatingly paragynous, but are sometimes amphigynous :— 12. P. Cactorwm (L. and C.) Schroet. 13. P. Fagi Hartig 14. P. Syringae Klebahn C.—Species in which, so far as is known at present, the antheridia are always paragynous :— 15. P. Nicotianae de Haan D.—Species in which the mode of development of the sexual organs 1s not fully known, or in which these organs have wot yet been found :— 16. P. Thalictri Wilson and Davis 20. P. jatrophae Jens. 17. P. agaves Vill. (?) 21. P. fiei Rau 18. P. Fabert Maubl. 22. P. citri Rau 19. P. Theobromae Colem. P. Thalictri is probably closely allied to P. Phaseoli, and may ultimately be found to belong to Group A. Gandara! states that a Phytophthora, specified by Villada as P. agaves, causes a disease of Agave in certain parts of Mexico ; but we have not succeeded in finding any description of this species. 2. Fabert and P. Theobromae are probably synonymous. According to Coleman, the latter species is. closely allied to P. Avecae; if this be really so, it should be placed in Group A; but antheridia are absent, or ave only rarely formed; and it is not certain whether they are amphigynous or paragynous. P. jatrophae has been issued in culture form, but apparently has not been described ; it may be identical with P. Nicotianac. P. fici and P. citri were named provisionally in 1915; but we are officially informed that nothing further has yet been published on them. 1 Mem. y. Rev. Soc, Cient. Antonio Alzale, xxv, 1908-9, p. 293. Larrerty and PeruypripGe—QOn a Phytophthora Parasitic on Apples. 41 V1I.—Practical Considerations. There is no reason to suppose that this particular form of apple-rot is likely to become a serious menace to fruit-growers in this country. It is not known how or from what source the fruits became infected. In other cases of Phytophthora- rot of apples, however, it has been observed that the affected fruit was confined to the lower branches of the trees near the ground; and it is believed that infection came from the soil as a result of rain splashes, &c. In such cases judicious propping up of the hanging branches would probably suffice to prevent infection. In some cases, too, the rot has been recorded as being particularly prevalent in fallen apples ; but whether infection took place before or after failing is not clear. Seeing that Osterwalder found P. Cactorwm (omnivora) causing a die-back of apple-shoots in Switzerland, a careful search was made amongst the trees in the orchard at Pilltown, Co. Kilkenny, from which the Irish apples came; but no trace of any such injury was to be found there. Further, an attempt made to infect apple twigs through wounds in the bark with a pure culture of the Irish fungus was not successful. Since oospores of the fungus are formed in the tissues of the rotted apples, it is clear that such sources of re-infection for the following season should not be allowed to remain in the orchard. From a general hygienic point of view, as well as in regard to this particular fungus pest, all rotted fruit should be collected and suitably destroyed. We desire to express our indebtedness to the persons named in the text of this paper who were good enough to furnish us with material for comparative study, and also to Sir Frederick W. Moore, of the Royal Botanic. Gardens, Glasnevin, who kindly placed at our disposal Cereus, Cactus, and other plants for inoculation purposes. We also desire to record our indebtedness to the Imperial Bureau of Mycology in Kew, and to Dr. H. M. Quanjer, Wageningen, for assistance in the matter of literature. VIL.— Summary. The present paper deals with a rot occurring in apples in Ireland which was found to be caused by a species of Phytophthora which proved to be P. Syringae Klebahn, and not P. Cactorum Schroet., a species that has been recorded several times as the cause of a similar rot in apples and pears in other countries. The causative fungus was isolated and studied in detail, in pure culture. It was found to produce both paragynous and amphigynous antheridia, but the latter are formed in relatively quite small numbers. Renewed studies of P. Cactorwm Schroet. and P. Lagi Hartig revealed the fact that in these species also amphigynous antheridia occasionally occur, though they are much rarer than paragynous antheridia. Apart from those species in which the sexual organs are imperfectly known or have not been discovered, P. Nicotianae de Haan is the only species of this genus in which the antheridia, so far as is known at present, are exclusively paragynous. It is suggested that renewed investigation of this species might lead to the discovery of some medium on which its sexual organs would be developed in abundance, and it is thought that amongst them examples with amphigynous antheridia would possibly be found. ‘There is also a possibility that some of the species of Phytophthora which apparently have amphigynous antheridia only may yet be found to produce paragynous antheridia occasionally. 42 Scientific Proceedings, Royal Dublin Society. It is proposed, therefore, to discard the generic name Nozemia, suggested a few years ago for those species having paragynous antheridia only, and to re-unite in the one genus Phytophthora all the species hitherto described under that name, irrespective of the types of antheridium which prevail in them. A classified list of the twenty-two species of Phytophthora recorded up to the present, with brief notes thereon, is furnished. The economic significance of the rot does not appear to be great. Supporting heavily laden, drooping branches with props and attention to orchard hygiene are suggested as preventive measures. EXPLANATION OF PLATES. Pare I. All drawings were made with the aid of a camera lucida at a time when the oospores were practically ripe, and all are magnified 645 diameters. Fig. © Sexual organs of P. Syringae in which the antheridium is amphigynous; from a naturally rotted apple. 2. The same as 1, from the same source, but with the much more common paragynous antheridium. 3. Sexual organs with amphigynous antheridium of P. Syringae, originally isolated from diseased Lilac in Germany. 4, The same as 3, from the same pure culture, but with paragynous antheridium. 5. Sexual organs of P. Cactorwm with amphigynous antheridium, from culture of Whetzel’s organism isolated from diseased apple in America. 6. The same as 5, from the same pure culture, but with paragynous antheridium. 7. Sexual organs of P. Cactorwm with amphigynous antheridium, from culture of organism isolated by Bewley from diseased apple in England. 8. The same as 7, from the same pure culture, but with paragynous antheridium. 9, Sexual organs of P. Cactorwm with amphigynous antheridium, from culture of organism isolated from diseased pear, received from Osterwalder (Switzer- land). 10. The same as 9, from the same pure culture, but with paragynous antheridium. 11. Sexual organs of P. Cuctorwm with amphigynous antheridium, from culture stated to have been isolated by Peters from Cactus seedlings. 12. The same as 11, from the same pure culture, but with paragynous antheridium. 13. Sexual organs of P. Hugi with amphigynous antheridium, from culture stated to have been isclated by Peters from seedlings of Fugus sylvatica. 14. The same as 13, from the same pure culture, but with paragynous antheridium. 15. A sporangium of P. Syringae from culture from Irish apple, showing the flattened apex. 16. A sporangium of P. Cactorwm from tuft of mycelium on apple, showing papillate apex. H LArrerty AND PeruyBripGE—On a Phytophthora Parasitic on Apples. 48 1PAatio, IOC. All figures reproduced from untouched negatives. Fig. 15 magnified 250; fig, 14, 500; the remaining figures 720 diameters. Fig. 1. Sexual organs of P. Syringac from naturally rotted apple tissue. The antheridium is paragynous and the oospore not yet ripe. 2, From the same source as 1. The antheridiuin is amphigynous and the oospore is ripe. 3, Sexual organs of P. Syringae from an apple previously inoculated with a single sporangium from a culture from the original rotted Irish apple. The antheridium is paragynous and the oospore nearly ripe. 4. Sexual organs of P. Syringae from the same source as in 3, but with amphigynous antheridium. 5. Sexual organs of P. Syringae from an apple rotted by product of germination of a single oospore derived from the original rotted apple. Antheridium aimphigynous. Contents of oospore disorganized by clearing. 6. Oogonium and amphigynous antheridium of P. Syringae from single hypha culture from original rotted apple on raw apple cylinder. , . Sexual organs of P. Syringae with amphigynous antheridium from pure culture on Quaker-oat agar originally isolated in Germany from Syringa vulgaris. 8. Sexual organs of P. Cactorwm with amphigynous antheridium on prune agar, from pure culture stated to have been isolated from Cactus seedlings. 9. The same as 8, but fungus originally isolated from rotted apple in America (Whetzel). 10. The same as 8 and 9, but fungus originally isolated from rotted apple in England (Bewley). 11. The same as 8, 9, and 10, but fungus isolated from rotted pear from Switzerland (Osterwalder). Oogonial contents plasmolysed. © 12. Sexual organs of P. Fagi with amphigynous antheridium, from culture on cooked carrot, original isolation from Cactus seedlings (Peters). 13. “Sphaero-conidia” of P. Cactorwm isolated by Bewley from rotted apple in England. The lower one has a papilla over the empty portion of the hypha on the left; it also has a hyphal outgrowth above on which the upper intercalary sphaero-conidium is borne. 14. One unripe and two ripe papillate sporangia of P. Cactorwm from rotted pear. Swellings similar to those seen in P. infestans ave present on the sporanugiophore, where the two lower sporangia are borne. 15. A single sporangium of P. Cactorwm from rotted pear, showing papilla and swelling of sporangiophore at point of attachment. 16. A single sporangium of P. Syringae from Irish rotted apple, showing flattened apex and abseuce of papilla. sie} Ee ( SCIENT. PROC. R. DUBLIN SOC., N.S., VOL. XVII. PLATE I. HAL. dul. ed, war% LAFFERTY AND PETHYBRIDGE. 1339) PEAT HenT: SCIENT. PROC. R. DUBLIN SOC., N.S., VOL. XVII. LAFFERTY AND PETHYBRIDGE. Ceo No. 2. SOME FURTHER NOTES ON THE DISTRIBUTION OF ACTIVITY IN RADIUM THERAPY. By H. H. POOLE, M.A. Sc.D), Chief Scientific Officer, Royal Dublin Society. {Read May 23. Printed Junu 21, 1922.] AN account has recently been given [Sc. Proc. R. Dub. Soc. N.s. xvi, 35, 1922] of some determinations of the total activity due to the 6B and vy rays from an emanation tube with various screens. These results were combined with the geometrical law of distance, and in this way, on certain assumptions, the activity to be expected at various depths in the flesh was calculated for a tube enclosed in a serum needle, and also for several arrangements of surface applicators. At the discussion which followed the reading of the paper it was suggested that additional figures were desirable, so as to represent the activity at various points in the neighbourhood of one or more emanation tubes. Further calculations have accordingly been made, the method already given for a single tube being extended to points off the “equatorial plane” of the tube [z.¢., the plane perpendicular to the axis of the tube at its middle point], and the effect of the use of multiple tubes worked out for a couple of typical cases. The results are shown in Tables 1 to 4, the corresponding conditions being given with the respective tables. Jn order to simplify the printing of the tables, the unit adopted has been changed. In every case the figure in the table may be taken to represent the action per millicurie hour of total dose, the action due to a dose of one millicurie hour concentrated at a point at a distance of one centimetre, without any screening except that due to the glass wall of the tube, being taken as 1000. The unit previously employed referred to a distance of one millimetre, and so was a hundred times as gveat as the new unit. On the other hand, the old figures referred to the activity of ten millicuries, or the total action produced by ten millicurie hours, so that figures in the new tables are ten times as great as the corresponding figures in the old. For example, the figure for the action on the surface of a serum needle is 6,200 in Table 1, and 620 in the table on p. 475 of the previous paper. Objection might be taken to the unit adopted, inasmuch as the screening action of the wall of an emanation tube must vary greatly. The unit, however, really refers to one particular tube which was used as a standard. Taking the “bare ”’ activity of this tube as 1,000, the activity at the same distance through 1:8 mm. of brass was 10:0. The action through this thickness of brass, which stops all the 3 rays, is sensibly independent of the thickness of the tube, and is SCIENT. PROC. R.D.S., VOL. XVII, NO. 5. I 46 Scientifie Proceedings, Royal Dublin Society. the real standard adopted in comparing the various figures. In the case of very thin screens the unavoidable variation in thickness of wall of different emanation tubes would cause appreciable errors, but for brass screens thicker than 0-2 mm. this variation would be unimportant. In each of the four tables the action is shown at various points on a plane parallel to the tube or tubes. In Tables 1 and 3 this plane passes through the axis of the tube. In Tables 2 and 4 it bisects the rectangle formed by the tubes at right angles. The relations of the various planes and distances involved may, perhaps, be described as follows :— Suppose that the emanation tube or tubes are placed horizontally in a north and south direction. Tables 1 and 3 refer to short and long single tubes respec- tively, while Tables 2 and 4 refer respectively to four parallel short tubes and six parallel long tubes, covering in each case a horizontal rectangular area. What has been called the “equatorial plane” is now a vertical east and west plane, passing through the centre of the tube or tubes. The figures in the tables refer to points in a vertical north and south plane, which, in the case of Tables 1 and 3, passes through the axis of the tube, and, in the case of Tables 2 and 4, passes midway between the two central tubes of the area: a in the tables is the vertical height of the given point above or below the tube or tubes; 6 is the horizontal distance of the given point north or south of the centre of the tube or tubes. In the case of a single tube [Tables 1 and 5] the distribution is obviously cylindrical round the axis of the tube, so that the action at any point may be found from the tables. It is only necessary to choose as the plane of the table one which passes through the axis of the tube and the required point, and measure aand bin this plane. In Table 2 the effective area covered by the four tubes is approximately square, so that the distribution in the “east and west” or “ equa- torial” plane is almost identical with that in the “north and south” plane, at distances greater than 5mm. from the tubes. In Table 4 the effective area covered by the tubes may be taken as 100 mm. x 60 mm., so that 30 mm. east or west of the centre brings us to the edge. whereas 50mm. is required to do so in the plane of the table. It would appear then that, if we wish to find the distri- bution in the equatorial plane, no large error would be introduced by taking from the table a value for the action at a point corresponding to a value of 6 20 mm. greater than the actual distance east or west of the centre. This would only apply to points at distances of 10 mm. or more from the plane of the tubes. The figures in the second column of Table 1 correspond to those in column A of the table in the previous paper. The slight discrepancies observable in some cases are due to the use of a shehtly different absorption coefficient for y rays in flesh, as a small discrepancy had occurred between the values used for column A aud those employed in columns B, C, and D of the previous table. Results are shown to two significant figures only. Some examples on the reading of the tables may be of use. Suppose a small tumour is to be treated, and it is decided to use an arrangement of needles similar to that given for ‘Table 2. If the total dose distributed over the four tubes is 50 millicurie hours, the action to be expected would be :— On the surface of a serum needle. : 50 x 1600 = 80,000 units. At a depth of 1 cm. opposite to centre of tubes 50 x Q91= 455 D »» »» » . ends Le Hx CHa S45 5 ems. Ms centre _,, 50 x 0°37 u il i m4 3 ends a HO Os = ers Poots—WNotes on the Distribution of Activity in Radium Therapy. 47 If a larger area has to be treated, an arrangement similar to that of Table 4 might be used. Suppose the total dose increased to 300 me. hrs., the actions would be :— On the surface of a serum needle 300 x 160 = 48,000 units. Ata depth of 5 cms. opposite centre of area of application 300 x 0:26 = 78 units. And so on. For a single dose with a flat surface applicator of 6 sq. em. area and 3 mm. thick (brass),in contact with the skin, the emanation tubes in the applicator being in contact with the base, and as uniformly distributed as possible, we can use the figures in column J of the table in the last paper, multiplying by 10 to allow for the change of units. Thus, for a dose of 300 me. hrs. the actions would be :— On skin at centre of applicator 300 x 12-4 = 3,720 units. At a depth of 1 em. opposite to centre of applicator 300 x 3:6 = 1,080 units. 4s 5 ems. ie 3 “ 300 x 0:27 = 81 units. In this way a comparison may be made between the actions to be expected at various depths with different arrangements and doses. As in numerous cases the per missible dose is limited solely by the damage done to the skin or flesh in contact with the applicator, it is interesting to work out the skin action through various thicknesses of brass in direct contact with the tissues. This is especially important in the case of internal tubular applicators where no intervening material is employed. The figures are shown in Table 5. Here F represents the activity at a fixed distance through ¢ mm. of brass, the bare tube being 1000. WD is the external diameter of a tubular applicator of thickness t with a bore of 1 mm. to take a single central capillary. Aq is the skin activity on the surface of such an Se or of any applicator, in which only a thickness ¢ of brass is interposed between the emanation tnbe [assumed 15 mm. long] and the flesh. Az thus also applies to a single tube in a flat applicator, or a large bore tubular applicator in which the emanation tube rests against one side. The figures in columns A, to A, represent approximately the skin activities on the surface of tubular applicators 3 3 to 7 mm. external diameter and ¢ mm. thick. The interior of the applicator, which in all these cases is at least 2 mm. in diameter, is assumed to contain as many uniformly active capillaries as it will hold, so that the distribution may be regarded as symmetrical. These values are obtained by multiplying the value of A, “for the appropriate diameter by the ratio of the values of F for the thin- and thick-walled tubes respectively. Thus, a tube 6 mm. in diameter and 2°5 mm. thick would have a bore of 1 mm., and contain a single emanation capillary. Its surface action would be 50, the value of F being 9-7. If its wall was only 0°5 mm. thick, the internal bore would be 5 mm., and would contain a large number of capillaries. The value of F for 0°5 mm. being 30, the surface action, A,, is assumed to be 30 x 50 = 9°7, ie., about 150, as shown in the table. This method is only approximate; but it seems unlikely that any errors due to it would be comparable with those due to the variations from the assumed conditions which are likely to occur in practice. 12 Scientific Proceedings, Royal Dublin Socvety. TABLE 1. Single emanation tube, 15 mm. long in serum needle. a = distance in millimetres from axis of tube. b= 7 5 i » “equatorial plane.” 25 5 75 10 15 20 25 30 40 50 mm. b 6200 6200 3100 39 6:0 27 16 1:0 052 0-29 2300 ©2300 =: 1200 47 6-2 28 1-7 11 0°57 0°33 580 550 300 47 6-4 2°8 1-7 ll 0°60 0°35 230 210 120 36 6:3 2:8 1-7 11 0°61 0°37 110 98 62 26 6-0 2°8 1-7 el 0°61 0°37 64 55 37 19 5-6 27 17 Ll 0-61 0°37 40 35 25 15 5-2 26 1-7 1 0-61 0°37 27 23 17 11 4:8 2-5 1:6 io 0-61 0°37 19 16 13) 98:9 4-4 24 1:6 11 0-60 0°36 14 12 99 74 4:0 23 15 11 0°59 0°36 10 95 80 6-2 37 2-2 1-5 1-0 059 0°35 71 6:5 5-7 4-8 3-2 2-0 1-4 0:98 0:58 0°35 5:2 4:8 4:3 3°8 2-7 1:8 1:3 0°93 0:56 0°35 3-9 3°7 34 31 2°3 16 1-2 088 0:54 0:34 3+ 3-0 2°8 2-5 2-0 15 1-1 0°84 0°51 0°33 2-5 2-4 2°83 21 1:7 1:3 1-0 0-79 0-49 0°32 16 16 15 1-4 1:2 1:0 0°81 0°65 0-43. 0-30 ll ll 1-0 1-0 0-90 0:77 065 0:54 0:38 0:27 0°61 060 059 0-57 0:53 0-48 0:43 . 0:38 029 0-22 Oo Of O8f O88 O84 OF 0°29 0:27 0:22 0-17 0°25 0:25 0:25 0-24 0:23 0:22 0-21 0:20 0-17 0-14 0-17 O17 O17 O17 O16 0-16 015 0-14 0-12 O11 0°13 OG WIR Oak Om Oa ali] 0-11 0-11 0-09 0710 0-10 0-10 0-10 0:09 0-09 0:09 0-08 0°08 0°07 0-07 0:07 0:07 0:07 0:07 ~—-0:07 0-07 0°06 0:06 0°05 Pootu—Notes on the Distribution of Activity in Radium Therapy. 49 TABLE 2. Four emanation tubes, each 15 mm. long, in coplanar, parallel serum needles equally spaced 4 mm. apart, forming a rectangle 15 mm. x 12 mm. @ = distance in millimetres from plane containing the four tubes. = » “equatorial plane.” ” ” ” The plane of the table passes midway between two tubes, so that for a = 0 and b <7°5 the nearest tubes are at 2 mm. on either side. The maximum activity on the surface of one of the needles (1:2 mm. diameter) would be 1600 on the same scale. For values of a greater than 20 the activities are the same as for the corre- sponding points in Table 1, from which they may be taken. With the exception of a layer a few millimetres thick next to the plane of the tubes, the distribution in the “equatorial plane ” would be nearly the same, d being now measured from the original plane of the table. 0 Pe) ) 7°) 10 16 20 25 30 40 50 mm. } 310 310 290 160 31 5°8 207 NEY ily 0°60 0°36 250 240 220 130 30 5°8 27 17 Her 0°60 0°36 160 150 130 80 25 a7 27 17 1-1 0-61 0°36 89 86 72 48 21 5°8 2°6 iS? Hert 0-61 0°37 56 54 47 32 17 a3 2°6 16 eA 0°61 0°37 38 36 31 23 13 5:0 25 16 TT 0°61 0°37 26 25 22 16 ll 4:6 2°5 16 ileal 0-60 0°36 19 18 16 13 3:8 4°3 2-4 16 toil 0°60 0°36 16 14 12 10 7-2 4-0 2°3 1-5 ile 0°59 0°36 ll 11 10 8-2 6:3 37 2-2 15 1:0 0-59 0°35 o1 8-8 8-0 6:9 55 34 2-1 1-4 1:0 0-58 0°35 6-4 6°3 5°8 5:3 4-4 3-0 2°0 13 0:97 0°57 0°35 4°8 4-7 4-4 4:0 35 2-6 1°8 1-2 0°91 0°55 0°34 3°8 37 35 3-2 2-9 2°2 16 oy 0°87 0°53 0-34 3°0 2-9 2°8 2°6 24 19 1-4 1-1 0°83 0°50 0°33 25 24 23 2°2 2:0 16 1-2 1:0 0-78 0°48 0°32 50 Seientifie Proceedings, Royal Dublin Society. TABLE 3. Single emanation tube in long serum needle, partially withdrawn at intervals so as to distribute the activity as uniformly as possible along a line 100 mm. long. a = distance in millimetres from axis of tube. Il b = a 3 es » “equatorial plane” (perpendicular to axis at centre point of this line). a 0 10 20 30 40 50 60 70 mm. 0 YLT Pah omens Leerect Bath Hao eS eR OR 0°6 93 930 930 930 930 470 0°80 0°30 1 350 390 350 350 350 180 0°84 0°33 2 90 90 90 90 89 45 0-88 0°35 3 38 38 38 37 37 19 0°87 0°37 4 20 20 20 19 19 959) 0°86 0°38 5 12 12 12 12 11 6°1 0-84 0°38 6 8:3 8-2 8-2 8-0 7°6 4:2 0-81 0°37 7 671 6-1 6-0 58 5-4 371 0°78 0°36 8 4a, 4-7 4°6 4-4 4°] 2:4 0°75 0°36 ¢) 3°8 3°8 37 3°6 372 2°0 0:72 0°36 10 3°2 372 31 3°0 2°6 Sy 0°69 0°35 12 2-8 2:4 2°4 2°3 2-0 1:3 0°64 0°34 14 2°0 2:0 9 1°8 1°6 ifort 0-48 0°32 16 ioe 16 1°6 15 1°3 0-91 0°53 0°31 18 14 14 1-4 1:3 Her 0:79 0-49 0:30 20 1:2 1-2 1-2 oy] 0°93 0°69 0°45 0:28 24 0-90 0°89 0°38 0-79 0°68 0°52 0°37 0°25 30 0°69 0°68 0°65 0°59 0-51 0-41 0°30 0°22 40 0°43 0:42 0:40 0°37 0°33 0:27 0:22 0:17 50 0:29 0°28 0°27 0°26 0°22 0-19 0°16 0°13 60 0:20 0°20 O19 0°18 0°16 0:14 0°12 0:10 70 0:15 0°14 0°14 0°13 0-12 0-11 0°09 0:08 80 Or1L O-11 0-10 0:10 0-09 0:08 0:07 0:06 90 0:08 0:08 0:08 0:08 0:07 0:07 0°06 0°05~ 100 0:07 0:07 0:06 0:06 0:06 0:05 0:05 0°04 Pootre—WNotes on the Distribution of Activity in Radium Therapy. 51 TABLE 4. Six emanation tubes in long, coplanar, parallel serum needles, equally spaced 10 mm. apart, and partially withdrawn at intervals so as to distribute the activity of each needle as uniformly as possible along a line 100 mm. long, thus forming a rectangle 100 mm. x 50 mm. a= 6 = distance in millimetres from plane containing the four tubes. ” ” ” “equatorial plane.” The plane of the table passes midway between two tubes, so that for a = 0 and The maximum activity on the surface of one of the needles (1'2 mm. diameter) would be 160 on the same b < 50 the nearest tubes are at 5 mm. on either side. scale. a 0 10 20 30 40 50 60 70 mm. J mm 0 4:9 4-9 48 4:7 4:5 2°5 0°59 0°31 1 4:7 4:7 4°7 4-6 4:3 2-4 0°59 0°31 2 4-3 43 42 41 3-9 2-2 0-58 0-31 3 3°8 3°8 3°7 3°6 3-4 2-0 0°57 0°31 4 3:3 3:3 3-2 31 9-9 17 0°57 0°31 5 2°8 2°8 2:8 2-7 2-4 15 0°56 0:31 6 2-5 Q4 24 2°3 2: 1:3 0-54 0°30 7 2-2 2-2 21 20 18 1-2 0°53 0°30 8 2-0 1-9 1:9 1:8 1-6 11 0°52 0°30 9 1:8 1S ney 16 1-4 0-97 0-51 0°30 10 1-7 1°6 1:6 1:5 13 0:90 0°50 0-29 12 15 14 1-4 1:3 ro 0-79 0-47 0:28 14 1:3 1:2 12 Vl 0:97 0-71 0-44 0:27 16 ll 11 11 1-0 0:86 0°64 0-41 0°27 18 1-0 1:0 0:96 0-89 0°76 0-58 0°39 0°26 20 0°91 0:90 0:87 0-80 0°68 0°53 0°36 0-25 25 0°72 0-71 0-68 0-63 0-54 0-42 0-31 0-22 30 0-57 0°56 0°53 0:49 0°43 0°35 0-27 0-20 40 0°38 0°37 0:36 0°33 0:29 0°25 0-20 0-16 50 0-26 0-26 0-25 0°23 0-21 0-18 0-15 0-12 60 0-19 0-19 0-18 0°17 0°15 0-13 0-12 0°10 70 0-14 0-14 0°13 0°12 0-11 0°10 0-09 0-08 80 0-10 0-10 0°10 0-10 0-09 0-08 0:07 0:06 90 0-08 0-08 0-08 0-07 0:07 0:06 0°06 0-05 100 0-06 0:06 0:06 0:06 0-05 0-05 0-05 0-04 52 Scientific Proceedings, Royal Dublin Society. TABLE 5. Skin action with various flat and tubular brass applicators. Thickness in mm. D = External diameter of tubular applicator of wall-thickness 7, allowing an internal diameter of 1 mm. to hold a single central emanation tube. ll F = Relative activity through ¢ mm. at any fixed distance. Aq = Skin action per me. hr. for single emanation tube in tubular applicator of thickness ¢ and diameter D, or in contact with a flat applicator of thick- ness 7. ‘This also represents the mazimum skin action for a single tube in a larger tubular applicator, in which the tube rests against one side. A, = Skin action for applicator thickness ¢, external diam. 3 mm., completely filled with emanation tubes. A, = Ditto ditto ditto 4 ditto ditto. A, = Ditto ditto ditto i) ditto ditto. A, = Ditto ditto ditto 6 ditto ditto. A, = Ditto ditto ditto i ditto ditto. t D £ Aa ds As Alg Ag Als mimi. mm. 0-1 1-2 208 5100 2200 1800 1400 1100 890 0-2 1-4 107 2000 1100 900 690 550 460 0:3 1-6 60 970 630 510 390 310 260 0-4 1:8 40 570 420 340 260 200 170 0°5 2-0 30 390 320 260 200 150 130 0-6 2-2 24 290 200 150 120 100 0:8 2°6 16:3 180 140 110 83 70 1-0 3°0 12°6 120 110 82 64 54 1-2 3:4 10-9 110 71 56 47 1-4 38 103 92 67 53 44 1°6 4-2 10°1 81 52 43 1°8 4-6 10-0 72 51 43 2-0 5:0 9-9 64 51 42 2-5 6:0 97 50 41 3-0 7:0 9°6 41 No. 6. PRELIMINARY EXPERIMENTS ON A CHEMICAL MELHOD OF SEPARATING THE ISOTOPES OF LEAD. By THOMAS DILLON, D.Sc., ROSALIND CLARKE, D.Sc.,. AND VICTOR M. HINCHY, B.Sc. (Chemical Department, University College, Galway). [Read June 27. Printed Juny 27, 1922.) THE usual methods used for the separation of the elements from one another depend upon reactions between ions, such as fractional crystallization and fractional precipitation. Reactions of this type are common to nearly all the elements, and, as the property of solubility of salts, upon which such reactions depend, often varies gradually among the elements, it is perhaps not surprising that there should be atoms which, while differing slightly in atomic weight, are so alike in all their chemical properties that they must be placed in the same position in the periodic table. Once, however, it is admitted that the atoms of any body have a different weight, it is difficult to imagine that they are absolutely identical in all chemical properties ; and it would seem to be only a matter of finding the particular chemical reaction in which they show an appreciable difference in order to be able to effect their separation. Now there is one property of the metallic elements which is of a very specialized character, and that is the property of forming organo-metallic compounds. This property is possessed by comparatively few metals, and elements standing close to one another in the periodic table show marked differences in the ease of formation and stability of their organic derivatives. Furthermore, the process of formation of the lead organic compounds is peculiar. Frankland and Lawrence! found that lead chloride reacts with zinc alkyl, half of the lead being precipitated, and the other half forming alkyl] derivatives of tetravalent lead. Pfeiffer and Truskier? were the first to substitute the Grignard reagent for the organo-zinc compounds in this reaction, and they prepared various alkyl and aryl derivatives of lead by this means. The reaction between lead chloride and the Grignard reagents is represented by the following equation :— 2 PbCl, + 4 MgRX = PbR, + 2 MgCl, + 2 MgX, + Pb. Here, some of the lead atoms change their valency from two to four, while others pass into the elemental state. It occurred to us that it would be worth while trying whether the lead atoms of different weights showed any appreciable 1 J.C. S. 35 (1879), 244. ” Ber. 37 (1904), 1125. SCIEN. PROC. R.D.S., VOL. XVII, NO. 6. K 54 Scientific Proceedings, Royal Dublin Society. choice in the direction they would take in this reaction, or, in other words, whether a separation of the isotopes of lead could be effected by its means. It should be mentioned in this connexion that Hoffmann and Wolf! in 1907 acted on lead chloride containing radium D with magnesium phenyl bromide, and found that most of the radio-activity remained in the metallic lead. Our scheme of work will be clear from the following diagram :— Na ae Poe eee) Eb PbRy | | NY VY PbvCl: (B) PbCle (B) A quantity of lead chloride is treated with the Grignard reagent, and the lead alkyl and the metallic lead are separated from one another, and converted into lead chloride. The two quantities of lead chloride are again separately treated with the Grignard reagent, and the resulting products separated as before. If there is any difference between the isotopes with regard to their tendency to form organic compounds, it will be seen that a repetition of this process, always using the metal on the extreme left and the organo-metallic compound on the extreme right of the diagram, will lead to an accumulation of one isotope at the extreme left and the other at the extreme right and the lead salts marked B and {3 in the diagram might possibly show a difference in the atomic weights of the lead which they contain. As a result of some preliminary observations, we decided to use the ethyl compound of lead as the basis of our experiments. The synthesis of the magnesium compound with this group goes smoothly and easily without the formation of troublesome bye-products, while the lead alkyl could be easily purified as far as required by us. Grutner and Krause,” who first isolated the lead tetraethyl by this method, found that it was contaminated with unsaturated lead alkyls; but for our purpose this was of no consequence. We found that a quantitative yield of lead tetraethyl could be obtained by using one mol of lead chloride to three-and-a-half mols of magnesium ethiodide. If the proportion of . lead chloride exceeded this, the yield diminished. The lead chloride used was a sample kindly given to us by Messrs. Hopkin, Williams, & Co., London. It was recovered from the manufacture of meso- thorium, and therefore consisted of a mixture of the chlorides of ordinary lead and of lead from thorite. The salt was recrystallized and dried by heating in a current of hydrochlorie acid gas. The Grignard reagent was prepared in the usual way, from 104 grams of ethyl iodide and 16 grams of magnesium. Fifty-two grams of the dry lead chloride were then added gradually. When the lead chloride had all been added, the mixture was heated for four hours under a reflux condenser, air and moisture bemg excluded by means of a mercury seal. The reaction product was poured 1 Ber. 40 (1907), 2425, 2 Ber. 49 (1916), 1415. Ditton, Ciarke, and Hincuy—Separation of the Isotopes of Lead. 5d slowly into water, and the aqueous mixture was extracted several times with ether. The ether solution was dried over calcium chloride; the ether evaporated, and the lead tetraethy] distilled in vacuo. A small residue of lead iodide remained in the distilling flask. The lead tetraethyl was next converted into lead nitrate by dropping it slowly into hot dilute nitric acid; and the lead was precipitated from the solution as sulphate by adding sulphuric acid and alcohol. The lead sulphate was filtered off, washed, and dissolved in ammonium acetate; and the solution was saturated with sulphuretted hydrogen. The precipitated lead sulphide was filtered off, washed, and dissolved in hot hydrochloric acid ; and the lead chloride which crystallized out was recrystallized from water containing hydrochloric acid. This lead chloride was dried and labelled “ Lead chloride A.” The residue from the extraction of the lead alkyl compound, consisting of basic iodide of magnesium, metallic lead, and traces of the chloride and iodide of lead, was repeatedly boiled with a solution of sodium carbonate until the carbonate solution gave no reaction for halides with silver nitrate. It was then dissolved in nitric acid, treated with sulphuric acid and alcohol; and the resulting lead sulphate was converted into lead chloride by exactly the same process as that described above. It was recrystallized from water containing hydrochloric acid, dried, and labelled “ Lead chloride a.” The A and the a chlorides were again made to react separately with the requisite quantities of magnesium ethiodide; aud the organo-lead compound was in each case extracted as before. The metallic lead from the lead chloride a and the lead tetraethy! from the lead chloride A were then treated for conversion into their respective chlorides in exactly the same manner as had been done previously, and the two samples were labelledrespectively “ Lead chloride 8” and “Lead chloride B.” Our next step was to find whether there was any appreciable difference between the atomic weights of the lead in the two soups The method adopted for this purpose was that of relative atomic weight determination as used by Soddy and Hyman! in their comparison of the atomic weights of ordinary and thorite lead. A platinum boat was heated in a current of hydrochloric acid gas ; and when the boat had cooled, the hydrochloric acid was displaced by dry aiv. The boat was weighed, and a quantity of the lead chloride under examination (a little over a gram) was placed in it. The lead chloride was fused in a current of dry hydrochloric acid ; the hydrochloric acid was displaced by air when cold, and the boat and contents were weighed. The boat was then dropped into a stoppered bottle containing one litre of water and 5 c.c. of nitric acid (which had previously been distilled over silver nitrate),.and the bottle was placed in a water bath kept at 60°C., and was shaken frequently. In this way all the lead chloride was brought into solution in about three hours. The boat was removed from the bottle, washed several times with distilled water, 100 c.c. of water being used altogether, and the washings being poured each time into the bottle. The boat was then dried in the steam oven, heated in dry hydrochloric acid, which on cooling was displaced by air, and was finally weighed again. A set of bottles, each containing one of the samples, was made up in this way, the same platinum boat being used throughout. In the first series of experiments three solutions of lead B and two of lead 3 were used. In the second series there were two of each kind. When a set of solutions had been prepared, 200 c.c. of the same silver nitrate solution were added to each from the same pipette. The bottles were well shaken, 1 J.C. S. 105 (1914), 1402. 36 Scientific Proceedings, Royal Dublin Society. and were left standing in the dark room in order to allow the precipitates to settle. On the next day the titration of the whole set was finished by dropping in the silver nitrate solution and watching by, the aid of a red lamp for the appearance of a cloud. When no cloud appeared after one minute, a reading was taken. The addition of two drops then produced no further cloud. In the first series the silver nitrate was added from a dropping funnel, which was weighed before and after the experiment, and the volume added was calculated from the specific gravity of the solution. In the second series a one c¢.c. pipette, graduated in hundredths, and having a small syringe attached to regulate the dropping, was used. In both cases the volume of one drop was about 0°03 c.c. In the first series of experiments the three B samples were titrated first, followed by the two 3 samples. In the second set the B and the (3 samples were alternated. The results of the experiments are tabulated below. The weight of lead chloride was obtained by subtracting the mean weight of the boat before and after the experiment from the weight of the boat and fused chloride. In No. 3 of the first series the weight of the boat varied by 0-4 milligram ; and in No. 1 of the second series it varied by 0°5 milligram. In all the other experiments the variation in weight did not exceed 0:2 milligram. The final column of each table gives the volume of silver nitrate required per gram of lead chloride as calculated from the experiments. First SERIES. 200 c.c. of silver nitrate solution added to each from pipette, Specific gravity of silver nitrate solution, 1-005. Wt. of Vol. of Total vol. of | Vol. of AgNO3 Wt. of | AgNOs3 AgNOs AgNOs | solution per | PbCl. solution from solution solution gram of | | burette. calculated. added. PbCle. | | | | grms. grms. | c.c. (G85 c.c. (QL) LewdB yn 1:09175 2°4074 2°3954 202°3954 185°39 2) fae pp. 26 1:09238 | 2°6126 275996 202°5996 185-47 COP RSA as 10910 | —-2:5256 2-5130 2025130 | 185-62 (4) Leads, . | 1:0893 2-44.39 2°4317 202°4317 | 1185-83 (4) os pp) 8 1:09195 2°8557 2°8415 202°8415 185°76 © \ | \ \ SECOND SERIES. | | Vol. of AgNO3 | Wt. of Vol. of AgNO3 | solution per | PbCle. | solution. eram of | | PbCle. | | ee | ae CrC Nn ECCS | (QU) eet 1°08875 | 20136 | 184-94 | | | (2) Lead g, . 1-0922 202-25 1517 | (3)) Tend BY 109275 20206 | ss-o1 | | (4) Lead B, . 1:09285 202°31 | 185°12 Ditton, Ciarkn, anv Hincny—NSeparation of the Isotopes of Lead. 57 If we omit No. 3 of the first series, the maximum difference between two samples of the same kind in either series is ‘08 c.c. per gram; whereas the minimum difference between two different samples is 0°18. If we assume that the atomic weight of the lead B is 207:1, the mean ratio of the silver nitrate required per gram of lead 8 to that required per gram of lead B would give for the lead B an atomic weight of 207-4 in the first series, and of 207°3 in the second series. Samples of the original lead chloride, together with the B and /3 chlorides, were sent to Professor John Nolan, of University College, Dublin, who very kindly examined their radio-activity, and reported that, while this property was weak in all three, the radio-activity of the (3 chloride was much greater than that of either the original or the B chloride. This confirms the observation of Hoffmann and Wolf. Unless there is an unknown source of error, the relative atomic weight determinations indicate that the different isotopes of lead are not identical in their chemical behaviour towards the Grignard reagent. If this is so, a con- tinuation of the process of chemical fractionation described above should produce such a difference in the atomic weights as would leave no doubt that a separation had taken place. Experiments in this direction are in progress. SCIENT. PROC. R.D.S., VOL. XVII., NO. 6. : L i 2] No. 7, THE LIGNITE OF WASHING BAY, CO. TYRONE. By T. JOHNSON, D.Sc., F.LS., Professor of Botany, Royal College of Science for Ireland, AND JANE G. GILMORE, B.Sc. (PratE III.) [Read Junn 27. Printed Aveusr 28, 1922.] LIGNITE or brown coal was found at various depths of the bore in isolated pieces embedded in the white clay; at a depth of 999 feet the deposit became a black mass. Much of the wood was very much crushed and twisted, to obliteration of its coniferous character, but for resin and a few pits. Some scraps occur at 712 feet, 869 feet 6 inches, and at 909 feet, which are flaky, and have the appearance of charred wood. These pieces show the structure of the pits on the walls of tracheids and of the medullary rays better than the sections obtained from the solid pieces. ‘The pits on the walls of the tracheids and the medullary rays are the same as in the solid material, but no resin parenchyma is found, with the exception of one piece at 909 feet. It seems that, in the charring, from whatever cause, the resin disappeared. Although Dicotyledonous leaves are plentiful in the bore, no trace of Dicotyledonous wood has been observed. Goeppert (6) recorded the same fact with regard to the Tertiary Flora of Silesia. The cross-section (Pl. III, figs. 1, 2) shows the wood much crushed and somewhat disorganized. The growth rings are very narrow, varying from 200 to 400u ; an occasional broad ring is seen. The spring wood consists of 10-30 rows of tracheids, which are fairly broad; the average size is 604; occasionally they reach 80u, and in one case 100m in diameter. They are thick-walled, and much flattened by pressure. The autumn wood occupies about a third of the ring. The bordered pits (Pl. III, figs. 3, 4, 5, 6) have a circular pore, but are frequently somewhat flattened top and bottom or are horizontally extended in their outer contour. They are in single or double rows; frequently three rows occur, and in one tracheid four rows of pits were seen (Pl. III, fig. 6). When in two or more rows, they are opposite, and may be either closely packed or scattered. They measure 20-24 in diameter, and when flattened measure about 4 less. Between the rows of bordered pits are very distinct Sanio’s bars, which are sometimes double. On the tangential walls of the spring and autumn tracheids small bordered pits are common (PI. ITI, fig. 7) on the whole extent of the wall. According to Penhallow (14), pitting in this position is a primitive feature, and among recent Conifers is best seen in Sequoia gigantea, making it in this respect almost unique. Resin canals are wanting. Resin parenchyma is very plentiful, and scattered throughout the wood. The average breadth of a resin cell is 50-604, and the SOIENT. PROO. R.D.S., VOL XVII, NO. 7. M 60 Scientifie Proceedings, Royal Dublin Society. length is 13 to 5 times the breadth. Short resin cells are found in wood, tending to form traumatic resin canals. The end walls of the resin cells are thin and unpitted (Pl. ITI, fig. 7). Uniseriate medullary rays, only, are present, 2-20 cells high, each cell being 20-24 in height. The horizontal and tangential walls are unpitted. The pits on the radial or lateral walls are (Pl. III, figs. 8, 9, 10, 11) usually in one, occasionally in two horizontal rows. The pits, 1-4 in the field, appear simple with oval pore, horizontally directed. They are 11-14 in length and 5-8y in breadth. At 712 feet the pits (Pl. III, fig. 11) were found to be bordered, so that the simple pits are probably due to bad preservation. ‘lhe arrangement of the pits in the ray cells is very similar to that in S. gigantea, in which, however, the pits are always bordered. In the fossiland in S. gigantea pores which are slightly oblique may occasionally be seen, and occasionally 4-5 pits in two rows in the field occur. In S. sempervirens, in common with previous observers, we find two horizontal rows of bordered pits with oblique pore in the cross-field is the normal arrangement. The identification of the wood of living and fossil Conifers has for the past seventy years been the subject of much investigation. The similarity in structure in species which are far apart, judged by their external features, and the amount of variation which may occur in any one species, render the task difficult in the case of some living, but more so in that of fossil Conifers. Recognizing the fact that leaf impressions and wood are hardly ever found together in the same deposit in organic continuity, Goeppert (7) in his Monographie der fossilen Coniferen, devised a classification of coniferous wood in which his composite genera united members of widely separated natural groups. Our wood belongs to his Cupressinoxylon type, which is characterized by bordered pits on the radial walls of the tracheids in one row, or if in more than one row then opposite; resin canals wanting (or, if present, traumatic), and abundant wood parenchyma, containing resin. Kraus, Kleeberg, Beust, Schroeter, Sanio, Schmalhausen, Conwentz, Knowlton, Penhallow, and others pursued their investigations in the hope of finding constant reliable diagnostic features for the different genera. In 1905 Gothan (8) showed that the pitting of the medullary rays is a very important feature which can be effectively utilized for identification when due regard is paid to the other characters of the wood. Gothan has founded the genus Zaaodiorylon to include fussil members of the Z'avodineae previously placed in the composite fossil genus Cupressinoxylon. Thus Tavodiorylon sequoianum includes Sequoia-like forms ; T. taxodianum, Za«odiwm-like species. Comparing our wood with the wood of living Conifers, we’find it cannot be :— Juniper or Libocedrus decurrens—since the horizontal and tangential walls of the medullary ray cells are unpitted (11); or one of the Cupressineae—since the pits on the radial walls of the tracheids are often in more than two rows (16). We conclude it is one of the Taxodineae. Schroeter (18) found in 1880 that the wood of Sequoia and Taxodium could be distinguished by the fact that in Sequoia the horizontal walls of the resin cells are of uniform thickness, and not pitted or nodulose as in 'l'axodium. Schroeter evidently appreciated the diagnostic importance of this feature, as he mentions it JoHNSoN AND GitMore—The Lignite of Washing Bay, Co. Tyrone. 61 twice in his account of Sequoia canadensis. All subsequent observers overlooked his observation until Prill (op. cit., s. 207), who has anticipated us in claiming credit for Schroeter for the discovery. Our wood is not Taaodium or Glyptostrobus, nor indeed Cryptomenia, in which the cross-walls are slightly nodulose. The species which have smooth cross-walls in the resin cells are Sequoia, Athrotaxis, and Taiwania. In Taiwania the pores of the medullary rays are oblique, sometimes almost vertical; we have not found Athrotaxis with medullary rays more than ten cells high, so that by a process of elimination we must conclude that our wood is Sequoia, and we believe it to be the wood of S. Couttsiae Heer, of which the leaves, cone, and seeds occur in the same bore, and have been already described by us (10). But for this association, it would be called Yaxodioxylon sequotanum Gothan. Heer (9) in his investigation of the Lignite of Bovey Tracey found resin plentiful, and satisfied himself that the wood was coniferous. He could make out little of its structure, but, from the occurrence of foliage and cones of S. Couttsiae, assumed that the wood belonged to that species. Beust. (2) in 1885 was a little more successful, and there is nothing in his account not in accordance with the now known characters of Sequoia wood. He states that S. Couttsiae is characterized by enormous quantities of resin, often seen in large lumps in the cells. Heer told Beust that he had found amber in large quantities in the wood. In 1869, afew years after Heer’s report on the Bovey Tracey Flora, Schenk (17) described the plant remains found in the lignite of Saxony near Leipzig, which has been assigned to the same age as Bovey ‘l'racey—i.e., the Oligocene. It is to be regretted that Schenk did not publish figures to show the cross-walls of the resin parenchyma which he said were mostly evident, and to show the roundish pits of the medullary rays. Schenk said he would have placed such a wood in the composite genus Cupressinoxylon if found alone. He examined the structure of the wood, which was in connexion with foliage shoots like those of S. giguntea, and with cones like those of S. sempervirens, and satisfied himself that the fossil was identical with S. Couwttsiae of Bovey ‘Tracey. Felix (4) holds that some of the lignite found in N.-W. Saxony and adjoining districts, and named C. protolarix, is undoubtedly Sequoia Couttsiae wood. The lignites and silicified wood of Lough Neagh have been the subjects of conjecture and study for many years. Dr. Richard Barton (1) in 1751, in what was one of the first accounts of petrifactions in the British Isles, gives particulars of the abundance of fossil wood around Lough Neagh, and of the gift to the University near Dublin of specimens, one of which weighed 150 lbs. We have not seen this fossil, but have examined another piece of silicified wood in the Geological Museum of Trinity College. This specimen has a copper cap and label on which the inscription reads: “ Brought from Lough Neagh, 1721, by Sir Wm. Fownes.” Unger (20), in 1847, named a sample of lignite from Lough Neagh, Peuce Pritchardi, and described it thus :—“ Strata concentrica minus conspicua usque ad 1 mil met lata. Vasa leptoticha versus limitem annuli paulatim angustiora. Pori disciformes minuti contigui uni-biseriales. Radii medullores simplices rarius compositi 1-25 cellulis parenchymatis amplis formati. Ductus resiniferi copiosi. Ad Lough Neagh Angliae (Andw. Pritchard).” Goeppert renamed it Pinites Pritchardt, simply repeating Unger’s description. Kraus placed it in his Cupressinoxylon group, and as such it has since been known. 62 Scientific Proceedings, Royal Dublin Society. Breadth of Tracheids. No. of Pits. Pits in Tracheids. Cupressinoxylon aequaele, Goepp (3) (Danzig). C. canadensis, Schroeter (18). 56-2u broad. 68u = radial. 40u = tangent. 12-16u 8. Couttsiae, Schenk (17). C. Holdenae, Seward (op. cit., p. 194). C. McGeei, Knowlton (12). C. polyommatum, Cramer (8). Rhizocupressinoxylon panni- conum, Felix (op. cit., p. 274). Pinites protolarix, Goepp, i (op. cit., p. 218). C, pulerum, Cramer (3, p. 171). Taxodioxylon Sequoianum (Merck.-Schmal) Gothan C. Wellingtonioides Kraiisel. (1, s. 293). (Prill) P. Pritchardi, Unger (20). d= 7dp. Felix (op. cit., s. 275). slight striation. nounced. 1 flattened and in contact. 2 1 | 20-25u. Small pits on tan- 2 gent walls. 3 1 great no. of small pits on radial 2 walls. 13°77u av. 12:24u, 3 15°3p. 4 (5) radial breadth exceeds tangent. 1 outer border ellipse with 2 greater axis parallel to med. 3 ray, @ = 18°3u. (4) 1 20°7u-24u (Saarau) Felix 2 (op. cit., s. 275). 3 hy average 22:°95u. 18°36-27-5u. (3) ihe numerous, mod. large. 3 spiral striations not very pro- 1 av, 22°95. 2 small touching. JOHNSON AND Gitmore—The Lignite of Washing Bay, Co. Tyrone. 63 Resin Canals. Resin Parenchyma. Med. rays, breadth and height. Pitting of Medullary rays. in summer wood. Tylosis in some canals. copious with brown resin. plentiful but variable. cross walls evident. present. abundant, 120-250 long, 50 | broad, slight, narrower than | tracheids. | uniseriate, 20u uniseriate, 1-8 cells high. 15°3-36°7u = height of 1 ray cell. height of 1 ray cell. 1-8 elliptical, oblique. 1-4 oblique bordered pits in 1 horizontal row. uniseriate, 2-20 cells high. uniseriate, to 30 cells high. uniseriate, 2-49 cells high. roundish, 2-4 fairly large, simple. 1-2:3 oblong, oval, simple pits. 15 x 10m. narrow. in spring wood in 1 ring. copious. | plentiful. | crowded together, cells short. copious. uniseriate, 2-26 cells high. | height varies. 2-40 cells high. 2-4 large oval pits in field, horizontal. transversely elliptical pores in two rows. | uniseriate, 4-14 cells high. 1-48 cells high. 2 large oval pits, 1 in autumn wood, sometimes bordered. Rays without pits top and bottom. 2-7 horizontal pits in spring wood, elliptical, in two rows. biseriate, not rare, to 35 cells high. uniseriate rarely compound, 1-25 cells high. 1, 23, pore somewhat inclined and not broad. 64 Scientific Proceedings, Royal Dublin Society. Macloskie (13) contributed an interesting paper in 1872, in which he gave for the first time a rough illustration of the wood. Seward (19) gives better illustra- tions of a Lough Neagh wood as Cupressinoxylon sp. ‘Two shdes in the Geological Laboratory of this College, labelled Cupressinoxylon Pritchardi Kraus, from Sandy Bay, show a wood which has the usual features of Zaaodioxylon, and one resin passage or duct (Pl. III, fig. 12) at the end of the summer wood. This feature is of diagnostic importance, as Sequoia normally possesses resin cells only; but when injured, traumatic resin passages or cysts are found in the autumn wood in S. sempervirens, according to Krausel, in the spring wood according to Penhallow (op. cit., p. 224). It would therefore appear, if these two preparations are correctly named, that C. Pritchard: Kraus may be nearly allied to, if not identical with, Taxodioxylon Sequoianum Gothan, and also with the lignite of the bore at Washing Bay. Unfortunately, the diagnosis of Unger, repeated by Goeppert and later writers, is now inadequate, and the type material needs re-examination and description in the light of present-day views. Our 1721 specimen shows occasionally three rows of bordered pits in the radial walls of the tracheids, and cannot very well be, as Gardner (5) assumes is the case for Lough Neagh henite, the wood of a Cupressus, which is distinguished by its mostly single row of bordered pits. We are at present examining microscopically lignite material from various localities in Ireland, and need hardly say that we find it is not of uniform character. We may expect to find as lignite examples of all the Conifers recorded in the Irish Tertiary. It needs little imagination to picture the presence of forests of Sequoia in N. Ireland, possibly contemporaneous with those in 8S. Devon at Bovey Tracey, the shores of the Baltic, the Rhine valley, Saxony, Silesia, and 8. France. We may yet find in Ireland large deposits of lignite or brown coal of economic value like those abroad. Among Cupressinoxylons the occurrence of 3—4 bordered pits on the radial walls of the tracheids is not common. A comparative list of species which are like the Washing Bay lignite in many respects is shown in the table, pp. 62 and 63, PuaTeE III. Figs. 1, 2. Transverse section of stem (1002 W. 1). x 50. (1002 W. 2). x 50. 3-6, Radial walls of tracheids showing bordered pits. x 50. 6. Shows four opposite pits on the wall of the tracheid (869’ 6” W. 7). 7. Tangential section showing pits on tangential walls of tracheids, smooth cross walls of resin parenchyma, and balls of resin in the cells (1005 W. 1). x 250. 8, 9. 10. Pits of medullary rays, apparently simple. x 250. 11. Pits of medullary rays with border (712’). x 250. 12, Transverse section of Cupressinoxylon Pritchardi Kraus, showing resin duct or cyst. SCIENT. PROC. k. DUBLIN SOC., N.S., VOL. XVII. PLATE III. 11. JOHNSON AND GILMORE. 16. Wo JoHNSON AND GitMorE—The Lignite of Washing Bay, Co Tyrone. 65 BIBLIOGRAPHY. . Barton, R., 1751.—Lectures in Natural Philosophy, p. 96. Dublin. . Beust, F., 1885.—Untersuchung tber fossile Holzer aus Gronland. Neue. Denksch. allgem Schweiz. Ges. gesammt Naturwiss., vol. xxix, p. 15. . Cramer, C., 1868.—Fossile Holzer der Arctischen Zone. Heer FI. foss. Arct. Bd. I, pp. 171, 174.. Ziirich. . Fexrx, J., 1882.—Beitrage zur Kenntniss fossiler Koniferenholzer. Bot. Jahrb. III, p. 269. Leipzig. . GARDNER, J. S., 1886.—A Monograph on British Hocene Flora, vol. ii, p. 82. Tondon. . Gorppert, H. R., 1852.—Beitriige zur Tertiarflora Schlesien’s. Palaeonto- graphica, Bd. II, p. 260. Cassel. . Gorpperr, H. R., 1850.—Monographie der fossilen Coniferen. Natur- werkundige Verhand. Holland. Maatschap, Wettenschappen Haarlem. Leiden. . Goruan, W., 1905.—Zur Anatomie lebender und fossiler Gymnospermen- Holzer. Abhand. K. Preuss. Geol. Landes. (N. F.) Heft., xiv. Berlin. . Heer, O., 1862.—On the Fossil Flora of Bovey Tracey. Phil. Trans. of Roy, Soe. of London, vol. elii, p. 1054. . JOHNSON AND Gi~More, 1921.—The Occurrence of a Sequoia at Washing Bay. Se. Proc. Royal Dublin Soe., vol. xvi (N.S.), p. 345. . Kigeserc, A., 1885.—Die Markstrahlen der Coniferen (Inaug, Diss.). Bot. Zeitung., Bd. xliii, p. 709. Leipzig. . Kyownton, 1889.—Fossil Woods and Lignites of the Potomac Formation, Geol. Surv., U.S.A., Bulletin 56, p. 46. Washington. 3. Mactoskiz, G., 1873—On the Silicified Wood of Lough Neagh. Proc. Belfast Nat. Hist. and Phil. Soc., p. 62. _ Pennattow, D. P., 1907.—North American Gymnosperms, p. 66. Boston, U.S.A. . PrILL AND Krause, 1919.—Die Holzer der schlesischen Braunkohle. Sond. aus dem Jahrbuch der Preuss. Geol. Landes. fiir 1917. Band xxxviil, Teil ii, Heft 1/2, p. 295. Berlin. Pritt, 1919.—Kritische Bemerkungen iiber Cupressinoxylon. Sond. aus dem Jahrbuch der Preuss. Geol. Landesanstalt. Band. xxxviii, Teil ii, Heft. 1/2, p-. 208. Berlin. Scuenk, A., 1869.—Ueber einige in der Braunkohle Sachsens vorkommende Pflanzenreste. Bot. Zeitung, p. 375. Leipzig. . ScuRoEtER, C., 1880.— Untersuchung tiber fossile Holzer aus der Arctischen Zone. Heer. FI. foss. aret., vi, p. 17. . Sewarp, A. C., 1919.—Fossil Plants, vol. iv, p. 188. Cambridge. 20. Uncer, F., 1847.—Chloris Protogaea, p. 38. No. 8. LIBOCEDRUS AND ITS CONE IN THE IRISH TERTIARY. By T. JOHNSON, D.8c., F.LS., Professor of Botany, Royal College of Science for Ireland, AND JANE G. GILMORE, B.Sc. (Pate IV.) [Read June 27. Printed Aveusr 28, 1922. THE present paucity of Conifers in the Irish Flora (Taxus and Juniperus) is most marked on comparing them with the extinct forms. We know that Pinus, Sequoia, Cryptomeria, and Cupressus once ficurished on the mountain slopes of N.-E. Ireland. We propose to add to this list the genus Libocedrus. While Sequoia is now confined to West North America, Cryptomeria to Japan and East China, and Cupressus has its nearest representative in the Mediterranean region, Libocedrus has a much wider, though discontinuous, range, and is represented by eight species, of which three occur in America, one in 8.-E. China, and the rest in New Zealand, New Caledonia, and New Guinea. It occurs to-day in much the same localities as Sequoia and Cryptomeria, but also ranges further southwards both in the Old and New World. Mr. A. Deane, the Curator of the Public Art Gallery and Museum, Belfast, kindly lent us, for comparison with the flora of the Washing Bay Bore, the collections of Jrish Tertiary fossils. Among them we discovered two slabs with cones and several slabs with foliage which came from the Interbasaltic beds of Ballypalady. We believe these to be the cones and foliage of Libocedrus salicornioides, Unger. The cone (P1. LV, figs. 1, 2) is oval oblong, 6°5 x 5mm., borne on a short lateral shoot, and apparently shows two fertile scales at right angles to two sterile ones. The oval cone of a living Libocedrus consists of four, rarely six, scales in pairs at right angles. The outer lower pair is very short in LZ. decurrens. Kach cone-scale is sub-apically mucronate. This mucro represents the free tip of the carpel of which the body is fused to the ovuliferous scale. In ZL. plumosa (L. Doniana, Pl. IV, fig. 3) the sterile pair of scales is better developed than in J. decurrens, and about one-third the length of the fertile pair, each scale having at its centre, not sub-apically, a pronounced recurved mucro. In our fossil a fertile scale is shown in surface view. The projection on the right may be either a sterile scale at right angles to it or the mucro of such a scale of the same length as the fertile one. If this latter view be accepted, ZL. plwmosa would be intermediate in this respect between JL. decurrens and the fossil. In an attempt at restoration of tissue we got scraps of wood showing bordered pits in the tracheid (Pl. 1V, fig. 4) and medullary rays with small pits (Pl. 1V, fig. 5). As far as they go, these scraps and one from a foliage shoot (PI. LV, fig. 7) agree with the structure of the living Libocedrus wood. JOHNSON AND GILMORE— Libocedrus and its Cone in the Irish Tertiary. 67 We also found a fusiform sclerotic idioblast 294 x 36u. In ZL. salicornioides from Leoben in Stygia we found an idioblast almost identical. Examination of fresh material of recent species of Zibocedrus showed similar sclerotic cells in some of them, e.g., in ZL. macrolepis. Though such cells are not by any means confined to Libocedrus, it is worthy of note that the fossil specimens agree with one another and with the living Lzbocedrus species in their possession, This Libocedrus-like cone is in continuity with a shoot which shows the characteristics of the widely recorded fossil named JL. salicornioides, specimens of which we have from the same locality. We had already examined and identified the Interbasaltic Zibocedius material when unexpected confirmation was supplied by an interesting find in the Washing Bay core (PI. IV, fig. 8). The joint-like scrap 7 x 4mm., represents an elongated grooved internode and a node flattened out, owing probably to the insertion of two opposite branches. The two lateral adpressed leaves with their decurrent bases are observable. One of the short, obtuse, slightly ridged, facial leaves is visible. To the left the basal joint of a lateral branch is recognizable. We have introduced for comparison a scrap of Libocedrus decurrens (Pl. IV, fig. 9). Fortunately the fossil yielded a little tissue which on restoration (fig. 10) shows an epidermis of oblong and polygonal cells with straight, simply pitted walls. Stomata are in single ribands about three cells apart. Stoma and ridge measure 26-36. ‘The usually solitary nucleus-like body seen on each cell represents a papilla projecting into space from the outer wall of the epidermal cell. Similar papillae occur in L. decurrens, L. macrolepis, Thuja occidentalis, and Callitiis quadrivalvis, but not in LZ. plumosa, L. Bidwilli, or Biota ortentalis. In L. chilensis and Calhiris robusta there are several in each cell. Berry (1) has found similar papillae in Frenelopsis ramosissima ; and Krausel (10) notes them in his recent description of the epidermis and stomata of Z. salicornioides. The waviness of the wall mentioned in his first account was afterwards (11) found by him to bea secondary artificial feature, due to the same cause as that resulting in the spiral striation of the tracheids of many Conifers, and in consequence of no systematic value. Our material is in general agreement with the Silesian material. The depression of the stomata naturally lowers transpiration, and this is still further lessened by the raised epidermal ring or fence which acts as a chimney or funnel. This is a xerophytic feature found in a sclerophyllous flora, of which we have many indications in the fossil Dicotyledonous leaves of Washing Bay still to be recorded. — We hoped that a comparison with other Conifers of the epidermis and stomata of our fossil, though surface views only of it were available, would throw additional light on its affinities. Hildebrand (9) found in all the many Conifers he examined that the stomata are depressed below the general level of the epidermis; that the shape of this depression, pit, or external chamber thus formed varied and was of systematic value. In the great majority of Conifers the one or two rows or rings of epidermal cells surrounding the sunk stoma have their external walls at the same level as the rest of the epidermis. In a few cases they have their external wall below the level. This holds true for Araucaria, Sequoia, most species of Abies and Picea. A third type, found more especially in the Cupressineae, shows the stoma ringed bya wall formed of the surrounding 4-6 epidermal cells raised above the level of the rest of the epidermis. The wall in some cases, as in Zhuja plicata, is due to the papillate arched thickening of the outer wall of the cell, which itself, however, is not raised above the general level. The projecting wall SOIENT, PROC. R.D.S., VOL, XVII, No. 8, N 68 Scientific Proceedings, Royal Dublin Society. is occasionally, as in Dammara and Libocedrus plumosa (L. Doniann), marked off from the general epidermis by an apparent groove or depression. In all the six species of Libocedrus we examined, except L. chilensis, it 1s of interest to note that the rampart or raised wall is observable, as it is also in our fossil. A similar wall oceurs in Thayja occidentalis, Callitris quadrivalvs, and Fokienia Hodgkinsi. he shape of the external chamber in surface view is oblong to square in the fossil as arule. The same shape is met with in L. decurrens. In L. chilensix, L. Doniana, L. tetragona, and 1’. occidentalis the prevailing shape is oblong. Callitris quadrt- valvis, C. robusta, and Libocedrus macrolepis show quadrangular or polygonal spaces, twice the size of the fossil ones. L. Bidwilli has smaller square-shaped spaces. As cross-sections of the fossil are not available, no comparison of it with the living forms as regards the shape of the outer stomatal chamber, as seen at right angles to the surface, is possible. Taking into consideration the elongated internode, the expanded whorl, the features represented by the epidermis, stomata, and papillae, we refer the Washing Bay fossil to Libocedrus salicornioides. It comes near L. decurrens and L. chilensis. Seward (16) has proposed the term Cupressinociadus to include fossil vegetative shoots of apparently Cupressincous affinity, whose generic character, especially in the absence of cones, is doubtful. As a receptacle for primitive forms of a group, not yet differentiated out into the generic types of to-day, such a term is dis- tinctly useful. In the adoption of such an all-embracing term, there is danger of placing under it, too readily, forms of recognisably distinct generic character, resulting in the introduction of an unnecessary vagueness into our knowledge of the geographical origin and distribution of forms. This seems to be the case with the fossil species of Libocedrus, as all the recorded species are referred by Seward to Cupressinocladus. Newberry (13) regards the earliest species known, Libocedrus cretacea Hr., of Greenland, as a synonym of his 7iwya cretacea. It is, however, evident from the illus- trations that the two are not identical. Heer’s fossil has the Libocedvus foliage and habit, Newberry’s the Thuya characters. In Libocedvus there is intercalated between each false whorl, formed of two pairs of opposite decussate leaves (one pair lateral. the other median or facial), a zone which has been variously described as an elongated internode, an enlarged node, or as derived from the decurrent leaf-bases. Tibovedrus has thus a more or less jointed stem, in which each joint consists of an internode covered by the decurrent bases of the false whorl of the four more or less adpressed leaves attached at its upper end. Newberry notes this difference in character of the foliage in Lbocedrus and Thuya in his text, but overlooks or ignores it in his comparison of his figure of 7° cretacea (op. cit., Pl. X, fig. 1a) with that of Heer of 1. cretacea (op. cit., Pl. XXX, fig. 3). So far as we know, the only other Cupressimeous genus which shows this inter- calated elongation of the internode in its youngest twivs is Kokienia, Henry and Thomas (8). This genus from East China shows the foliage of Libocedrus macro- lepis combined with the cone of Cupressus. In JZ. tetragona, which is nearer Thuya in habit, the intercalation takes place between each pair of facial and of lateral leaves. A joint init thus consists of only one pair of divaricate, not adpressed, leaves (either facial or lateral). Libocedrus has been described by Sargent (15) as “perhaps too closely connected with Thuya to be considered generically distinct.” Such a view must be still more applicable to early Tertiary 1 We are indebted to Professor A, Henry for recent material of Libocedrus for this com- parison. JOHNSON AND GILMORE—Libocedrus and its Cone in the Irish Tertiary. 69 forms. The seeds of Labocedrus, Thuya, and Biota are, however, easily distinguish- able, being one-sidedly, two- sidedly, or not at all winged, respective ly. The earliest. recorded Libocedrus, L. erctacea, Heer (4), is from the Upper Cretaceous beds of Atane in Greenland. Its lateral leaves are basally united for some distance. All the leaves are completely adpressed, giving the twigs, which are opposite and 2mm. wide, parallel sides. Though Heer lays stress on the oppositeness of the lateral branches, the twig he figures (op. cit., Tf. xliii, fig. 1d) shows alternate branches only. Very similar to /. cretacea and more highly suggestive of Z. decurrens of California is L. Sabiniana, Heer, from the Spitzbergen Miocene (5). The foliage and opposite branches (most Cupressineac have alternate branches) suggest Libucedrus. This view is practically confirmed by the occur- rence in the same beds of seeds which have the characteristic oblique almost one-sided wing of the Libocedrus seed, well shown in Heev’s illustrations. In his elaborate investigation of fossil woods from Greenland, Beust (2) finds, by com- parison of the fossi] aud recent woods, that Libocedrus was a common tree in the Lower Miocene of Greenland. His identification of the Greenland fossil as Libocedius is not accepted by Schenk (18). The detailed illustrated account of L. Sabiniana occurs in the second volume of Heevr’s Flora fossilis Arctica (5), but appears to have been overlooked by Seward (15), as his rejection of Heer’s identi- fication of the fossil as Zibocedrus is based on a short note in the seventh volume (6), almost confined to a statement by Heer that the fossil from the Atane Tertiary beds in Greenland agrees with that (the type) from Spitzbergen. He adds that a cone-scale from Hare Island probably belongs to Z. Sabiniana. L. gracilis, Hr. (7), which is also recorded from the Miocene of Spitzbergen, is more like L. plumosa (LZ. Doniana) of New Zealand in habit, and Z. Chilensis of South America in its associated cone-scales. The most widely distributed fossil assioned to Libocedrus is L. salicornioides Unger (17). It occurs in numerous localities in the Tertiary beds of Europe, from the Kocene to the Upper Miocene. Opinions have differed considerably as to its interpretation. ‘hus Saporta (13) finds his material from Armissan in all points identical with the type Thuytes salicornioides Unger from Croatia, but adds it is not at all certain that it is a Libocedrus, More perfectly preserved fertile specimens may, he thinks, show it is Viscwm, for which he supplies a then appropriate diagnosis. The occurrence of ZL. salicornioides in the Interbasaltic beds of North-East Ireland is not surprising, as the species is known from the amber beds of the Baltic and from Central Europe. Gardner (3) reported the genus doubtfully as Libocedrus adpressa sp.n. from the Woolwich beds at Bromley in Kent. The evidence has since been rejected as insufficient (op. cit., p. 308). It is of interest to add that Ludwig (12) records Libocedrus salicornioides from below the columnar basalt of Holzhausen. This material, like ours from Washing Bay, was found in the form of single joints. 70 ps SODA Cow Scientific Proceedings, Royal Dublin Society. Puate LV, . Cone. (x 3.) (A. 64a.) . Cone. («x 3.) (A. 648.) . Cone of LZ. plumosa (L. Doniana). (x 3.) Tracheids with bordered pits. (x 250.) (A. 64a.) . Medullary rays with small pits. (x 250.) (A. 64a.) . L. salicornioides, Unger. (x 8.) (A. 64a.) . Tracheid and medullary ray from foliage shoot. (x 250.) (A. 61.) . L. salicornioides from Washing Bay. (x 38.) (904 6”.) . L. decurrens. (x 3.) . Epidermis of JZ. salicornioides showing stomata and papillae. (x 250.) (904 6” B°.) BIBLIOGRAPHY. . Berry, E. W., 1910.—Epidermal characters of Frenelopsis ramosissima. Bot. Gaz., 50, p. 305. . Beust, F., 1885.—Untersuchung tber fossile Holzer aus Gronland. Neue Denksch. Allgem. Schweiz. Ges. gesammt. Naturwiss. Bd. xxix, p. 40. Taf. iii, figs. 10-17, Taf. v. Ziirich. . GARDNER, J. 8., 1886.—A Monograph of the British Eocene Flora, vol. i, Paleont. Soc., London, p. 25, Pl. II, figs. 17-20. . Heer, O., 1882.—Die fossile Flora der Polar Lander. Flora fossilis arctica, Bd. vi, Abt. ui, p. 49, Taf. xxix, figs. 1, 2, Taf. xliii, fig. 1d. . Heer, O., 1871.—Die Miocene Flora und Fauna Spitzbergens. Flora fossilis aretica. Bd. ii, p. 34, Taf. 11, 6-15, Taf. iv, fig. 4d. . Heer, O., 1883.—Fl]. foss. arct. Bd. vii, p. 58, Taf. Ixx, 17; Ixxxvi, 1, 2; Ixxxvii, 8. . HEER, O., 1871.—Op. cit., Taf. ii, figs. 20-24. . Henry and Tuomas, 1911.—Fokienca Hodgkinsi. The Gardener's Chronicle, vol xlix, Feby. 4, p. 67. . HILDEBRAND, 1860.—Der Bau der Coniferen Spaltdffnungen und einige Bemer- kungen tiber die Vertheilung derselben. Bot. Zeitung, p. 149. Leipzig. . Krauset, R., 1920.—Nachtrage zur Tertiirflora Schlesiens i, Sond. aus dem Jahrbuch der Preuss. Geolog. Landesanstalt fiir Band xxxix, Teil i, Heft. 3, p. 354. Berlin. . Krauseu, R., 1921—Nachtrage zur Tertiarflora Schlesiens iii, Jahrbuch der Preuss. Geolog. Landes. Band xl, Teil i, p. 376. . Lupwic, R., 1858.—Fossile-pflanzen aus dem Basalt-Tuffe von Holzhausen, bei Homberg in Kurhessen. Palaeontog. V. Taf. xxxiii, fig. 13. Cassel. . NEWBERRY AND Howick, 1895.—The Flora of the Amboy Clays, U.S.A. Geolog. Survey, vol. xxvi, p. 53, Pl. X, figs. 1, la. . Saporra, G. DE, 1865.—Etudes sur la végétation du Sud-Est de la France a Epoque tertiaire. Ann. Sc. Nat. 5°. Tome iv, p. 42, Pl. L, tig. 4. Paris. . SARGENT, C. S., 1896.—The Silva of North America, vol. x, p. 133. . Sewarp, A. C., 1919.—Fossil Plants, vol. iv, pp. 307, 308. Cambridge. . Uncer, F., 1847.—Chloris Protogaea, p. 11, Pl. II. Leipzig. . SCHENK, A. IN ZITTEL, 1891.—Traité de Paléontologie, Part i, p. 862. Paris, SCIENT. PROC. R. DUBLIN SOC., N.S., VOL. XVII. PIPARE SVE JOHNSON AND GILMORE. Lead No. 9. THE ELECTRICAL DESIGN OF A.C. HIGH TENSION TRANSMISSION LINES. By H. H. JEFFCOTY?. [Read Junr 27. Printed Avausr 4, 1922.] 1. The recent Reports on the Water-Power Resources of the British Islands have directed attention to the very important economic possibilities of hydro- electric developments in these countries. Compared with the total estimated power available, very little has yet been done in the direction of harnessing the rivers. It is hoped that, as a result of the recent investigations, active steps will be taken to bring about a great extension in the development of this valuable national asset. It appears, therefore, to be opportune to direct attention to some of the varied problems that are met with in hydro-electric engineering. The present paper gives an account of a method for calculating the per- formance of high tension transmission lines, based on direct evaluation by complex quantities. In it the phase angles and magnitudes.of the various vector quantities have not to be separately considered, and the process is almost entirely arithmetical. ‘The method is straightforward, and can largely be systematized into an arithmetical routine, which greatly simplifies the procedure. The method is illustrated in its application to three-phase overhead trans- mission lines. It may also be applied to single-phase working. 2. The problem of the electrical design of a transmission line presents itself usually in this way. At one site—the sending end—A.C. electric power is generated, transformed to high tension, and put into the line. At a second site—the receiving end—the power is to be employed, and usually step-down transformers are located. The number of phases, the frequency, the voltage, and the power to be delivered are known at the receiving end, The power factor of the load is also given, or is approximately estimated. The hours during which power is to be used and the fluctuations in the demand throughout the year are also presumed to be known approximately. The distance between the two sites is measured, and allowance is made for increased length of the wires due to sag, and also to irregularities in direction of the line both horizontally and vertically. It is then required to find the best size of conductor to use in the line, and, the size having been decided on, to determine the efficiency of transmission and the variation of voltage at the receiving end under varying load. 3. The choice of size of conductor is governed largely by the consideration that the total annual cost involved in its use shall be the least possible. It must always be provided, however, that the size of the conductor falls within the limits imposed by electrical considerations, such as ohmic heating, voltage SCIENT, PROG. R.D,S,, VOL. XVII, NO. 9. (0) 72 Scientific Proceedings, Royal Dublin Society. regulation, and corona loss, and also by mechanical considerations, such as the size of wire or cable convenient for manufacture, and its mechanical strength necessitated by its employment as an overhead line exposed to wind and weather. Tn long spans the question of mechanical strength becomes of great importance. The annual cost involved in the use of a particular size of conductor (assuming that the electrical and mechanical conditions are satisfied) depends mainly on two items. ‘The first of these is the annual value of the power lost in the line, which is greater the smaller the size of the conductor, and which also depends on the current to be transmitted, and its fluctuations in value at different times throughout the year. The second item is the interest and depreciation on the cost of the conductor ; and this is greater the larger the size of conductor employed. In the case of overhead lines the supporting poles or towers will in general increase in size to a slight extent with increase in weight of the line; but this can probably be sufficiently allowed for by a slight increase in the basic rate for cost of conductor, thus regarding the fluctuating part of the cost of poles as a small percentage added to the cost of the conductor proper. The best size of conductor then is that for which the sum of these two items of annual cost is as small as possible; and this cousideration leads to a mathe- matical determination of the best area of cross-section. 4. When the size of conductor is known, it is necessary to determine its electrical performance. The line has resistance, inductance, capacity, and leak- ance. In long distance lines it is necessary to take all these factors into account. In short lines, however, it is usually permissible to neglect the effects of capacity and leakance, with the result that the calculations are greatly simplified. The solution of the long distance high tension transmission problem commences with Heaviside’s general differential equations for the line. Their solution leads to hyperbolic functions of vector quantities. The method to be described below deals directly with these functions, and manipulates them by the laws of complex quantities. The process is illustrated by numerical examples, which indicate how the work can be systematically arranged. As a result of the calculation of the electrical performance of the line, it may be found, for example, that the voltage regulation is so bad that a change of size of the conductor must be made from that which otherwise gives theoretically the most economic performance. In other cases, the use of auxiliary apparatus, such as synchronous condensers, may be necessary. When the voltage is high and the size of conductor small, the corona loss may become relatively so great that a modification of the design is essential. 5. It is necessary now to set out symbols for the several quantities involved in the calculations. The following notation will be adopted :— Let 2 be the distance in miles measured from the sending end to any point in the line where, at a particular instant ¢ seconds from some zero of time, the instantaneous electrical pressure to neutral is v, and the current is c. Let the length of the line between stations be / miles, and the line be of over- head type. We will suppose that the system is three-phase or single-phase, with a frequency of / cycles per second. Let the circular conductor (solid or stranded) be of radius a inches, and of cross-sectional area A square inches. Let the conductors be equally spaced at 4 inches mutually apart. For each conductor let the resistance be & ohms per mile; the inductance be Z henrys per Jurrcorr—Llectrical Designof A.C. High Tension Transmission Lines. 73 mile; the leakage conductance be A mhos. per mile; and the permittance or capacity be S farads per mile. Let V volts be the effective value of the voltage between one conductor and neutral; let C’ amperes be the effective value of the current per line; and let P watts be the power transmitted per line. Let suffixes 1 and 2 applied to V, 0, P indicate their values at the sending end and receiving end of the line respectively. Let V. be the disruptive critical voltage to neutral; and let @ watts be the corona loss of power per line. Let 6 be the temperature in degrees centigrade of the atmosphere surrounding the line; and / be the barometric pressure in inches of mercury. 996% ail Let $= 350" and let y be the surface condition factor for the conductor. Further, let cos @ be the power-factor of the load, let the efficiency of trans- mission be denoted by », and the voltage regulation fraction from no to full load by p. Write i= ]-1, p= 2zf, m =,|(R + ipL)(K + ipS), hk + up K+ ips” Let the value of power be s pence per kilowatt hour, the value of conductor material be g pence per lb., and the rate of interest and depreciation on the line be 7 per annum. and n= 6. We will require to use certain identical relations connecting complex quantities, and these are written down here for reference. (@ + ix) + (b+ ty)= (a+b) +1(@+y), (a+ tw, —(b+w)=(a-6)+i(e-y), N (@ + Ie) = na + Ina, (a + ix) (b + ty) = (ab - ay) + 1 (bx + ay), a+ ab+ ay a bx — ay b+ bry “Bry (a + tx)? = (a? — x) + 2(2a2), Jit ix = 5 Gee. (alee! \? Wem es Peal one pee 4 2 cosh (a + iv) = cosha cosx + 7sinh a sin 2, sinh (@ + iv) = sinh a cosa + icosha sinz, 2 x i ae cosha=1+ 5 + ay © aq sO y 2 ve uw SMM 49 = sb a a ae dE fa oy ee)! 7! 02 74 Scientific Proceedings, Royal Dublin Society. 7. The constants for the single-phase or symmetrical three-phase line are determined by the following formule. They refer to one conductor :— “NT , 070 ae a ohms per mile for copper; or —— ohms per mile for aluminium. A 2 043 R= aA Ib = (s + 74:1 logo *) x 10° henrys per mile. K =5 x 10° mhos per mile, approximately. 3°88 x 10-8 ; ae farads per mile. logy = (ap S = 8. We commence with the determination of the best size of conductor. First we seek to determine the amount of power lost in the year and its value. The average loss of power in the line for the year depends, not on the average current, but on the square root of mean square of the current. We require to know this square root of mean square value of the current for the year. Let it be C amperes per line. If the demand for power is steady and con- tinuous night and day throughout the year, then C is the value of the current transmitted. But in general the power demand is not steady and continuous, and it is necessary to estimate the square root of mean square value of the current transmitted. Thus, for example, if the various values of the current during the year, placed in order of magnitude for their appropriate times, give a straight line, having an arithmetical mean value & and extreme variations of k’ from the mean, the effective value of the current may be found thus :— Let y be the value of the current at any time 7. ‘Then t Wik where 7'is the whole period of time or year. The root mean square value of y is ay 2 Lye AeA yrdt = fk? + 4k 0 y =(k - Wh) + 2h’ As another example: if the variations of the current transmitted throughout the year follow a sine law with a complete period of a year, or if the various values of the current during the year, placed in order of magnitude for their appropriate times, are equivalent to a sine load, we obtain |x? + 44 for the effective value of the current. Here & is the arithmetical mean value of the current for the year, and k’ is the amplitude of the sine curve. Then 2rt yak +k sin-a Jurecorr—Electrical Design of A.C. High Tension Transmission Lines. 75 and the root mean square value of y is Tiere ee eel Under the assumptions made, this is the effective value of the current so far as power loss in the line throughout the year is concerned. And similarly for other cases, the effective value of the current must be estimated from the load and time diagram. For present purposes the average power loss throughout the year is taken as O?R as a first approximation. If desired, more exact figures, resulting from the determination of the electrical performance of the line, including leakage and corona effects, may be used afterwards for greater accuracy. Let us suppose copper is the material chosen for the line. “NA2 Then & = ad ohms. 2 : 4 Cl aoa Hence power loss per line = (°F = -045 AR watts ; and the annual value of this va : l power is °376 C” Fs pence. The weight of copper is 20,300 AZ lbs. The annual charge for interest and depreciation on the copper is therefore 20,300 Algr pence. The total annual cost in pence is then *376C* Wo 20,300 Algr, and this is to be a minimum. Cosi. 4 Aas sit t : Hence wee 53,900 Agr is to be a minimum. To satisfy this requirement we find A is given by oe — 53,900 qr = 0, or A? = 00001850? — , qr . nic i ayn IS or A = -00430 le Ii q = 15 pence per lb., r = 8 per cent., s = one penny per unit, we obtain A = 00392 C. Similarly for other values of gq, 7, s. This equation gives us the most economic sectional area of copper conductor, and it will be noted that the current density in the conductor is quite small, so that there is no risk of overheating. = 0701 ; If aluminium be used for the line, we have & = Soins ohms. The weight of C? i aluminium is 6120 -A/ lbs. Hence we find that —+ 10,000 Agr is to be a minimum. Therefore 4A =:010C0 te for aluminium conductor. 76 Scientific Proceedings, Royal Dublin Society. 9. We may now presume that we know the size of conductor to be used. We seek to determine its electrical performance. The equations connecting the voltage v and current ¢ at any point distant from the sending end of the line are adv i ad = ie = (zt Je de f[ d \ -— =|KiS—)». ake \\ Saat } We are mainly concerned with the space variations of v and ¢ along the wire. We suppose a sine wave of voltage to be impressed on the line at the sending end. Differentiating the first of these equations with respect to w and using the second, we obtain ; ay d d ) —= = \(\K+S=}v. (8 1 L;,)( + Sz v dx? We may write zp for S voltage, into v%, the maximum voltage at the position considered, so that in these equations, changing v, the instantaneous = = (R + ipL) (K+ ipS) x = mp, The solution of this equation is Uo, = Age) + Brew: It Vand C be the effective values of voltage and current at the position «, we have V= Aor fi Be, where 4 and B are constants independent of z. ) : dv eae Also - = = (R + ipl) % and Fi = m(Aoe™ - Bem). ia We A mae B mx Hence C= Ror Ayn + Boe), -oL ( But fee =” 6 My = - Ayer + By ma 7 and nO = - Ae™™ + Bem, At the sending end of the line «= 0, and therefore V,=A+ Bb, nC,=- A+B; ». 24 = V, - 2Ci, 2B=V,+ nC. Hence V =1V, (er + €™) - And, (eo - eo") = V, cosh mx - nC, sinh ma. Likewise V,. C=40, (eo + om) - Vi (one —e"*) = OQ, cosh mx - —'sinh ma, 2n nN Jurrcorr— Electrical Design of A.C. High Tension Transmission Lines. 77 Put w«=/, and we find V.= V, cosh ml - nC, sinh ml, aoe CO, = OC, cosh ml — Ze sinh mil. These equations give I V, = V. cosh ml + nC, sinh ml, Wo. 0, = 0, cosh ml + —) sinh mil, Ee 1D which may also be obtained similarly to the foregoing equations by putting x =/ in the general equations which determine A and B. The power put in per phase is P, = V,C, cos (gi - $1’), Where @, and ¢, are the phase angles of VY, and (Q, respectively referred to the phase of C, as standard ; -. P, = V, cos $, x C, cos p; + %V, sin ¢, x Cy sin y’. If the receiving end is open, Vy ie OF =U; Al = exon wal, bancl (Oy 7 sinh ml, which is the charging current. If the receiving end is short-circuited, V,=0. Then V7 =nC,’sinhml and 0,” = CG, cosh ml. We thus find for the voltage regulation at the receiving end that the ratio of the voltage at the receiving end, on throwing off the load, to the full load voltage is Ve VA CVA Vencosirn the transmitting voltage remaining constant. ‘ as The efficiency is the ratio of the power taken from the line at the receiving . (? p since Vie= Vy, end to that put in at the sending end, or = a 1 10. We now turn attention to corona loss, or the loss of power from the line occasioned by the discharge from the conductor due to ionization of the surrounding atmosphere. Above a certain disruptive critical voltage this loss can take place. The magnitude of the loss depends on the amount by which the line voltage exceeds the disruptive critical voltage, and, to some extent, on the temperature and pressure of the atmosphere. For single-phase and symmetrical three-phase lines it has been shown by Peek (Trans. Aim. Inst. El. Eng., vol. xxx, 1911) to be expressible in the form @ = Bs f x AP x (V - V,)? x 10° watts, b where the disruptive critical voltage V, is given by b V,. = 123,400 yéa log,, 5 effective volts to neutral. 78 Scientific Proceedings, Royal Dublin Socicty. 6 depends on the temperature 7’ and pressure / of the atmosphere surrounding the line, and is given by 996K 273+ 6 if 0 is expressed in degrees centigrade and / in inches of mercury. y is a factor depending on the roughness of surface of the conductor, and is 1 for polished round wires, 0°98 to 0°93 for roughened or weathered wires, 0°87 té 0°83 for seven-strand cable. a, b, f,1, V have the same significations as before. 11. We can now set out in systematic form the scheme of calculation of a single-phase or symmetrical three-phase line. We divide the scheme into five sections—(1) a statement of the data of the problem, (2) the determination of the most economic section of conductor, (3) the calculation of the line constants, (4) the calculation of the electrical quantities required, and (5) the determination of the corona loss. (1) Data for one line. Frequency, f cycles per second. Voltage between wires at receiving end. Power delivered. Power factor of load, cos ¢. Effective average value of current C’ delivered throughout the year. Length of line, / miles. Conductors spaced 6 inches mutually apart. Cost of conductor material, g pence per lb., and rate of interest and depreciation thereon, 7 per annum. Value of power, s pence per unit. Average atmospheric temperature, 6 degrees centigrade, and pressure, 2 inches of mercury. (2) Determination of the most economical size of conductor. (See paragraph 8.) For copper, A’ = -0043 C’ ie sq. inches. qi For aluminium, 4’ = -010 C | a sq. inches. Choose the nearest standard size of conductor A sq. inches. ig a= th “inches for solid conductor. T (3) Calculation of line constants. _ 7048 ‘070 jis = A ohms per mile for copper = STAR ohms per mile for aluminium. a b Ib = (s + 741 x log,, 5) x 10° henrys per mile. « K = 5 x 10% mhos per mile. Jerrcorr— Electrical Design of A.C, High Tension Transmission Lines. 79 S= BSS = farads per mile. logy i p = 2nf. m = /(R + ipL)(K + ips). _ [k+l © SN K+ ips cosh ml = cosh (@ + ty) = cosh weos y + 7sinhw sin y = € + 7, sinh m/ = sinh (w + zy) = sinhacos y + 7cosh x sin y. a at ge p= at a, t+ ay tees cosh « lta a Re : ae ie oh sinha = etay ey ay te oO. . . n sinh ml. sinh ml n (4) Calculation of electrical quantities. Refer all vector quantities to the phase of the current C, as standard. The phase angles of the various vector quantities may readily be obtained if desired, since we know the components of the complex quantities. For a single-phase line :— V, = 4 x voltage between wires at receiving end = X + iu, where pu = dQ tan @. P, = 4 x total power delivered. Gh = pee) e V,cos @ In a symmetrical three-phase line :— oie 1 V, = voltage to neutral at receiving end = —— x voltage between wires at : 3 receiving end = d + tu, where u = 2 tan ¢. P, = power transmitted per phase = 4 x total power delivered. E aE C, = current per phase delivered = TCT a 5 Then V,= V,cosh ml + nCysinhml = a + 73. 2 Voltage to neutral at sending end = )/a? + 3°. C, = C,cosh ml + a sinh ml = y + w. Current per phase at sending end = ,/\?4 2. P, = ax + By. 2 n = efficiency = =, 1 80 Scientific Proceedings, Royal Dublin Society. AG vs V; a+ 13 ft é = =a = © To * ~~ cosh ml” E+ iz V, =X+ Aft. Vy ot ge Po = im = ea ne (5) Determination of corona loss. For a single-phase or symmetrical three-phase system y = 1 for polished round wires, = 98 to -93 for roughened or weathered wires, = ‘87 to ‘83 for seven-strand cables. Sis V, = 128 l ¢ = 123,400 yéa logy = volts. Loss @ = 5°53 x & ae x (V - V.)’ x 10° watts per line. b 12. When the line is comparatively short and the voltage not especially high, the capacity and leakage of the line may be neglected. The calculations then become greatly simplified. The efficiency and regulation are calculated by the following formule :— V,’ = (V, cos » + CRI)? + (V2 sin @ + CpLil)’. Nits (> War 13. Illustrations will now be given of the procedure in particular cases. 1t is convenient to follow the routine arrangement of the calculations employed below. (1) Data for one line. Line, symmetrical three-phase. Frequency, 50 cycles per second. Voltage between wires at receiving end, 66,000 volts. Greatest power delivered, 8,250 kw. Power factor of load, cos @ = °85. Effective average value of current delivered throughout a year, 31 amps. per phase, being 1:225 x mean amps. for year, since the load-curve is approximately equivalent to a sine-curve in which #’=k. (See paragraph 8.) Length of line, 7 = 25 miles. Conductors of copper, spaced 6 = 108 inches mutually apart. Cost of conductor material, g = 15 pence per lb. Interest and depreciation rate thereon, 7 = ‘08. Value of power, s = 1 penny per unit. Average atmospheric temperature, 0 = 20° C., and barometric pressure, h = 30 inches. Jurrcorr— Electrical Design of A.C. High Tension Transmission Lines. 81 (2) Determination of most economic size of conductor. O’ = 31 amps. 1 4’ = 0045 x 3 ——— = ‘122 sq. inch. £ 3x 31 «hes x08 122 sq. inch Choose conductor 19/18, having 4 =:129 sq. inch; .-. a = 0-23 inch. The current transmitted at maximum load is 85 amperes, and so the current density in the conductor is allowable. (3) Caleulation of line constants, 045 A b= 108 inches, a = :23 inch. b b : - = 470; log,, - = 2°67. a a L =(8 + 741 x 2°67) x 10% = 2:06 x 10° henrys per mile. K = 5 x 10° mhos per mile. R= = ‘34 ohms per mile. 3°8 ine ‘Sa a = 1-45 x 10°§ farads per mile. p= 247 Xx 00 = 314:16. pL = ‘648. pS = £56 x 10°. m = |(-34 + 7 x 648) (5 x 10° +7 x 4-56 x 10°), Multiplying and evaluating these complex quantities by the rules given in paragraph 6, we find m = | — 2-958 +7 x 1558 x 10° = 000437 + 4 x -00177. = 344474 :648 TTEG ZnSO La n peat es 100 /l4-24 - 4 x 747 = 388 - 7x 96:2. ml = 25 x (000437 + % x 00177) =:01092 + 7 x 0442. cosh ml = cosh ‘01092 cos :0442 + ¢ sinh :01092 sin ‘0442. cosh °01092 = 1 + 00006 = 1:00006. sinh -01092 = -01092 (1 + :00002) = ‘01092. cos :0442 = -999, sin °0442 = 0442. Hence eosh ml = 1:00006 x 999 + 7 x :01092 x 0442 = 999 + 7 x 000483. sinh ml = sinh -01092 cos 0442 + @ x cosh :01092 sin 0442 = 01092 x 999 +7 x 1:00006 x :0442 = ‘01091 +2 x 0442. n sinh ml = (388 - 4 x 96:2) (O1091 +7 x °0442) = 8:49 + ¢ x 16:1. sinh ml °01091 +7 x :0442 : a oe = = BRR o ee OD = (- 0625 +72 x 113-9) x 10°. 82 Scientific Proceedings, Royal Dublin Society. (4) Calculation of electrical quantities. Refer all vector quantities to C2 as standard. Vi = aT 66,000 = 38,000 volts at cos@ = °85, = 32,300 + 7 x 20,000. IB P, = 4x 8,250 = 2,750 kw. 2 ) 2 C. oP ONES 85 amps. * ~ 38,000 x -85 V, = V.cosh ml + nC, sinh mi. V.cosh ml = (32,300 + 2 x 20,000) (999 + 2 x 000483) = 32,260 + ¢ x 19,996. Cmsinh ml = 85 x (8°49 + 7x 161) = 722 +72 x 1,370. . V, = 32,982 + 7 x 21,366 = 39,300 volts to neutral at sending end. C, = CO, cosh ml + — > sinh ml. C, cosh mi = 85 x (999 + % x 000483) = 84:9 + ¢ x 0411. ae ah sinh ml = (32,300 + 7x 20,000) (- -0625 + 7 x 113-9) x 10° =- 2:28 +74 x 3°679. C, = 82°62 + 74 x 3:72 = 82:8 ents (Bee pee at sending end. P, = 32,982 x 82°62 + 21,366 x 3°72 watts = 2,725,000 + 79,500 = 2804 kw 2,750 0 > Sn fame + 21,3702 32,300? + 20,0002 It is thus seen that the electrical performance, calculated for maximum load, is very satisfactory. (5) Determination of corona loss. Take y = 83 for 19/13 cable. | FRO & V, = 123,400 x ‘83 x 1:02 x -23 x 2°67 = 64,300. As the working voltage to neutral (39,300) is less than this figure, there will be no corona loss. 14. Seeing that the example in the last paragraph represents a comparatively short line, and the voltage is not extremely high, it is likely the simpler formule of paragraph 12, in which & and S are omitted, would lead to close results. This may be verified. V,’ = (88,000 x °85 + 85 x 34 x 25)? + (88,000 x *d27 + 85 x 648 x 25)? = (32,300 + 722)? + (20,000 + 1,375)? = 1:546 x 10°; . V, = 39,300. JErrcor1—Llectrical Design of A.C. High Tension Transmission Lines. 88 1 r > Top = ‘978. lise 32,300 39,200 : Coat StG00 a ae Thus these results are in close agreement with those of paragraph 13. 15. As an example of long distance transmission, we will take the same data as in paragraph 15, with the exception of the leneth of line, which we will now suppose to be 200 miles. The calculation of the several quantities proceeds exactly as before till we come to m/l. Then (3) ml = 200 x (000437 + 7 x :00177) = 0874 + 7 x +354. cosh mi = cosh ‘0874 cos ‘354 + 7 sinh ‘0874 sin -354. cosh ‘0874 = 1 + 005819 + -0000006 = 1:00382. sinh -0874 = -0874 (1 + 001273) = 08751. cos ‘354 = -9580. sin 354 = 3467. cosh mf = 1:00382 x -938 + 7 x ‘08751 = ‘3467 = 9416 +7 x -03034. sinh m? = sinh:0874 cos°354 + 7 x cosh ‘0874 x sin :354 = ‘08751 x -938 + 7 x 1:00382 x -3467 = -0821 + 7 x ‘348. msinh mi = (3888 - 7 x 96-2) (0821 + 7 x 348) = 65:32 4+ 74x 127-1 sinh md -0821 +7 x °348 = (- 10:18 +2 x 894°8) x 10°. n 388 — 7 x 96-2 (4) Calculation of electrical quantities. V,, P,, C, are the same as in paragraph 13. V, = V,cosh ml + nC, sinh mi. Vy cosh ml = (32,500 + ¢ x 20,000) (9416 + 7¢ x 03034) = 29,810 + 7 x 15,810. Cyn sinh ml = 85 x (65°32 + 4 x 127-1) = 5,550 + 7 x 10,800. *. Vy = 35,360 + 7 x 30,610 = 46,750 volts to neutral at sending end. egy C, = C, cosh ml + — sinh mi. 2 n C, cosh ml = 85 (9416 + 7 x :03034) = 80-03 + ¢ x 2579. Se O00 201000) (- 10-18 + 4 x 894-3) x 10-8 = - 18:22 +7 x 28°68. . C, = 6181 +7 x 31-26 = 69:3 amps. per phase at sending end. P, = 35,360 x 61:81 + 30,610 x 31:26 = 2,183,000 + 957,000 watts = 3,140 kw. 2 e W M5 Oe a? aig le | 85,860 +2 x 30,610 ai Die OMIGHEACEO30S4 = 38,470 + 7x 31,250. _ /88,4702 + 31,250° P ~ 432,300? + 20,000? = 1:305. 84 Serentific Proceedings, Royal Dublin Society. (5) The corona loss is nil, as the disruptive critical voltage exceeds the operating voltage. Owing to the poor efficiency and bad regulation, it would be better to redesign the line for a higher voltage. 16. Using the approximate method of paragraph 12 for the example of the last paragraph, we obtain the following results :— V2 = (88,000/x 85 + 85 x -34 x 200)? + (38,000 x 527 + 85 x ‘648 x 200)? = (32,300 + 5,780)? + (20,000 + 11,000)* = 2-411 x 10°. co Ya = 49, lOO, 1 = ——— = 849. a 780 52300 49,100 Sy peeks aT P2 > 38,000 2 Comparing these results with those of paragraph 15, we find :— Par. 16 Par. 16 Voltage to neutral at sending end, 46,750 49,100 Efficiency at greatest load, ‘876 “849 Regulation fraction, 1°305 1-29 These results are somewhat discordant, and consequently the more accurate method of paragraph 14 is preferable here, except for a rough approximation. 17. Data as in paragraph 15, but let voltage between wires at receiving end be 100,000 volts, and spacing between conductors be 120 inches. (2) For the same power delivered, the effective mean current is now OF Sol x The most economic size is then A’ = ‘00392 x 20° = -0805 sq. inch. Choose conductor 19/14, having A = ‘097 sq. inch, and a@ = 0°20 inch. 66 —— = 20°5 8. 100 0:5 amps (8) Caleulation of line constants. 043 3 = ye = 45, GS, S20 Bb b = 600, log,)— = 2°78. = (8 + 74:1 x 2°78) x 10° = 2:14 x 10-*. L ea Bai Go CEE eR 108 2°78 Jurecorr— Electrical Design of A.C. High Tension Transmission Lines. 85 p = 31416. pL = ‘673. pS = 44x 10°. m = (45 +7 x 673) (005 + i x 44) x 10° = f — 2958 +7 x 1-983 x 10 = -000548 + 7 x -001806. a (eee 673 103 = WAS) > Resin 3 . ¢ “005 2 aoe. * = 1538 - 7x 102 » 10 = 4l@ 3% x eel ml = 200 (-000548 +7 x 001806) = 1096 +7 x -3612. cosh m/ = cosh ‘1096 cos °3612 + 7 sinh :1096 sin °3612. cosh 1096 = 1+ :0060 + :000006 = 1:0060. sinh ‘1096 = 1096 + 00022 = -10982. cos ‘3612 = :9354. sin 3612 = °3535. cosh mi = 1:006 x -9354 4+ 7 x 10982 x -3535 = -941 + 74 x 0388. sinh ml = ‘10982 x 9354 +7 x 1:006 x -8535 = (1027 +7 x °356. nmsin ml = (410 - 7 x 124-4) (1027 +2 x :356) = 86-4 +7 x133-9. i 102 Lx 85 . ~ aie un Ss — (- 12°27 + 7x 865) x 10>, Il (4) Calculation of electrical quantities. ) m= ae = 57,700 ab cos = -85, = 49,000 + 4 x 30,400. 2) 95 P, = S00 = gas tee Oo O00) G3 = “49,000 = 56°1 amps. V, = V.coshml + nC, sinh ml. V,cosh ml = (49,000 + 7% x 30,400) (941 +7 x 0388) = 44,920+ 7 x 30,500. Omsinn ml = 561 x (86:4 +7 x 1332) = 4,850 + 7 x 7,480. V, = 49,770 +7 x 37,980 = 62,600 volts to neutral at sending end. C, = C,cosh ml + ue sinh ml. C, cosh ml = 5671 (-941 +7 x -0388) = 52:8 + 7x 2°18. % sinh ml = (49,000 + 7x 30.400) (- 12:27 + ¢ x 865) x 10-6 = — 26-9 + i x 42-08. Cy = 25°9 +7 x 44:26 = 51:3 amps. per phase. P, = 49,770 x 25°9 + 37,980 x 44°26 = 1,290,000 + 1,680,000 = 2,970 kw, OED sores Hae 0710 ie ae , _ 49,770 + 2 x 37,980 alll V,/ = “OEE OSSSE a 54,450 + 7 x 38,150, — 94,4502 oP 38,1502 = 1-152, 7 49,000? + 40,4002 86 Scientifie Proceedings, Royal Dublin Society. (5) Determination of corona loss. 0°83 for 19/14 cable. VY = J = TOD. V, = 123,400 x :83 x 1:02 x 2 x 2°78 = 58,000. 50 x 2 Q = 553 a — es Ts x (62,600 - 58,000)? x 10-* = 46,800 watts = 468 kw. We now revise the efficiency estimate. The approximate total power put in is 2,970 + 47 = 3,017. 2, U) => syne ce 2) 18. As a further example we will assume a still higher voltage, with a view to illustrating the great increase in corona loss at high voltages. Data as in paragraph 15, but let the voltage between wires at the receiving end be 132,000 volts, and the spacing between conductors be 120 inches. (2) For the same power delivered, the effective mean current per ats 1s NOW (see paragraph 13) halved, or C’ = 15°5 amps. The most economic size is then A’ = :00392 x 15:5 = -061 sq. inch. Choose conductor 7/12 having dA =‘0606 sq. inch. .. @ = ‘186 inch. (3) Calewlation of line constants. 043 R — mo = ale Se, = 10), - 770. b log,)— = 2°89. S10 F T= (8 + 74:1 x 2:89) x 10° = 2°22 x 10% KES 8) Se AMES 33810, ee S= pag a ee . p = 2m x 50 = 814-16, pL = -699. pS = 4°23 x 10°. m = /CT71 + 7x 699) (5 x 10° + 7x £28 x 10°). = f — 296 + 24x 301 x 10% = 000794 + 2 x -00189. i —— 5x 10°+ 4x 423 x 10° = /165 - 7 x 168 x 10* = 448 - i x 188. Juercorr— Electrical Design of A.C. High Tension Transmission Lines. 87 ml = 200 (000794 +2 x 00189) = :159 + 7 x 878. cosh mi = cosh 159 cos *3878 + 7sinh:159 x sin °378. cosh 159 = 1 + -01265 + -000027 = 1-01268. sinh:159 = 159 + 00067 + 0000008 = -15967. cos 378 = 9294, sin 378 = 3691. cosh md = 101268 x -9294 +7 x 15967 x 38691 = -941 47x -059. sinh m/l = ‘15967 x -9294 + 7x 1:01268 x 3691 = :1482 + 7 x -374, nsinh ml = (448 — 7 x 188)(1482 + 7 x 374) = 136°8 +7 x 139-6. sinh ml +1482 + 7 x 374 fe) ERS IR IRS = (— 16:07 + 7 x 826) x 10°. (4) Calculation of electrical quantities. V, = Oe = 76,000, at cos » = °85, = 64,600 + i x 40,000. 2) P= = = 2,750 kw. 2,750,000 e405 amps PmecUORTES pm Grin tae V, = V2cosh ml + nC, sinh ml. V.cosh ml = (64,600 + 7 x 40,000)(-941 + 7 x (059) = 58,440 +7 x 41,450. Cm sinh ml = 42° (186°8 + 7 x 139°6) = 5,810 + ¢ x 5,930. “.V, = 64,250 + 7 x 47,380 = 79,900 volts to neutral at sending end. C, = C2 cosh mi + Pecinh ml. C,cosh ml = 42°36 (941 + 7x 059) = 40 +7 x 2°51. = sinh m/l = (64,600 + 2 x 40,000) (— 16:07 + 7 x 826) x 10° = - 34:08 +7 x 52°76. C, = 592 +%x 55°27 = 55 amps. per phase at sending end. P, = 64,200 x 5°92 + 47,380 x 55°27 = 380,000 + 2,620,000 = 3,000 kw. 2,700 ~ 3,000 — Oi , _ 64,250 + 2 x 47,380 _, . F Ve = Wil a6 xe WED = 71,300 oP Se 45,800. _ (71,800? + 45,8002 Ps 64,600" + 40,0002 = deity, (5) Determination of corona loss, y = ‘83 for 7/12 cable. 6 = 1:02. V. = 123,400 x 83 x 1:02 x 156 x 2°89 = 47,000. a x (79,900 - 47,000)? x 10-* = 2,120 kw. SCIENT. PROC. R.D.S., VOL. XVII, NO. 9. P Y = 5°53 x 88 Scientific Proceedings, Royal Dublin Society. The input at sending end, irrespective of corona loss, is 3,000 kw. Hence approximately ee TS SND AW): 7 i ‘537. Thus the efficiency is greatly reduced by corona. Here the loss from the line due to corona discharge is so great that the value of the leakage conductance K chosen above must be increased if a better deter- mination of the performance of the line be required. If so high a voltage as 132,000 be desired, it would be necessary to use a conductor of larger diameter for the sake of reducing the corona loss. Possibly aluminium might be employed instead of copper. 4 Py som No. 10. THE OCCURRENCE OF HELIUM IN THE BOILING WELL AT ST. EDMUNDSBURY, LUCAN. By A. G. G. LEONARD, F.R.C.Sc.1., Pu.D., F.LC., AND A. M. RICHARDSON, A.R.C.Sc.1., A.I.C. (PLATE V.) [Read May 23. Printed Auvcusr 29, 1922.] THE water from this well and the gases evolved from it were examined by Adeney (Proc. Roy. I. Acad., 1906}. He found the gas to contain 97°9 per cent. nitrogen, 2-1 per cent. carbon dioxide. On the suggestion of Professor Adeney, we have examined the gas for the presence of rare gases. The gases rising from the bottom of the well and giving it the appearance of ebullition were collected by the method of Ramsay and Travers (Proc. Roy. Soc., 60, p- 442, 1897). Quantities of about eight litres were collected at a time. he gases were transferred in the laboratory to glass gas-holders. Lxpervmental. Some preliminary experiments were first carried out in the preparation of argon by sparking atmospheric nitrogen with oxygen over caustic potash, and later by passing atmospheric nitrogen over red-hot calcium turnings. When using calcium for this purpose, it is necessary to heat the tube fairly strongly, and pump off any gases evolved, before commencing the combustion, as calcium turnings which are not quite fresh develop a surface film of hydroxide which loses water on heating, and this interacts with the calcium giving off hydrogen. — A qualitative examination of the gases from the well having definitely shown the presence of argon and helium, a quantitative determination was undertaken This was carried out by the removal of the bulk of the nitrogen by heated calcium and subsequent removal of argon and traces of nitrogen by charcoal immersed in liquid air. A separating funnel of capacity 838 c.cs. was fitted with a rubber stopper carry- ing two tubes with glass taps, one of which connected with the gas reservoir, and the other with the rest of the system. The stoppered stem of the funnel was connected by rubber tubing with a reservoir of water. By this arrangement it was possible without dis connection to admit several quantities of moist gas, measured at atmospheric temperature and pressure, to the apparatus. The calcium tube was SCIENT. PROC. R.D.S., VOL. xvul., No. 10. Q 90 Scientific Proceedings, Royal Dublin Society. packed with 25-30 grams of calcium turnings, and about 9 inches of its length at the end with copper oxide to eliminate hydrogen. The gas passed from the funnel through tubes containing CaCl, and P,O,;, bubbled through sulphuric acid, and then passed over heated calciuin, the issuing gas again bubbling through sul- phuric acid, which served to indicate the speed of the current of gas. The gas was then drawn into an automatic Sprengel pump, and delivered into a small gas- holder. Before passage of the gas the apparatus was exhausted with a filter pump, while the calcium tube was heated to eliminate any hydrogen evolved by interaction of the calcium with any water vapour present, and finally exhausted by the Sprengel pump. In all places, where possible, connections were made by fusing the glass tubes together. In the few places where this was not possible the rubber joints were varnished with shellac and luted with melted black rubber. The exhaustion having been completed and the apparatus proved air-tight, the measuring funnel was filled with gas and. adjusted to atmospheric pressure. The calcium was then heated and the gas passed over slowly, the pressure in the apparatus not rising above a few mms., and the residual gas being pumped off continuously until exhaustion was complete. The residual gas was now passed over heated copper oxide to remove hydrogen, and over phosphorus pentoxide, and allowed to remain fifteen minutes in contact with charcoal immersed in liquid air, which removes all gases except helium and hydrogen. The charcoal tube was previously heated to 200° C. and exhausted. From the charcoal the gas was then pumped off by a Topler pump and collected for measurement. It was found -that a further treatment with charcoal in liquid air did not appreciably affect the final volume. ‘Three estimations of the helium content were made, but the other gaseous constituents were collected together in the one charcoal tube. The volume of helium was measured by passing it into a capillary tube connected with a mercury reservoir, adjusting to atmospheric pressure, marking the level, subse- quently filling the space occupied by the gas with mercury, and then weighing the mercury. The charcoal tube was now heated to 200°C., and the gases pumped off and collected, when they were again passed over heated calcium and copper oxide. The volume of gas remaining was then measured. On examining the spectrum of this gas it showed the lines of argon strongly with some hydrogen lines. Results. The following table gives the results obtained, all volumes being reduced to N.I'.P. and expressed in c.cs. ‘The argon was estimated by uniting the residual gases from the three experiments :— | T . Vol. of gas | ee eae Vol. of Vol. of Percentage | Percentage taken. teat Helium. Argon. of Helium. of Argon. with Calcium. 7 783 9°48 0566 0-072 785 14°16 0°550 24°48 0-070 0°95 785 24-09 0°620 0-079 Some determinations by other experimenters of helium and argon in natural gas from other places are given for comparison, the values for nitrogen and carbon-dioxide in the gas from Lucan being taken from Adeney’s analysis. Soe ae ete “NOSGUVHOIY AGNV GUVNOUT ‘poseyoind (q) £ [Tea moay (My) SB Vara Deere de : = : || ae | (aviv ~ 8 G § len (Vi | ‘(M) ¥ : = | SCD) | (ANS) Vi | | a *| (a) oH ‘(M\) °H | | e | (q)°H °| ‘(A)°H | | a | ‘(d) °H | | | a | | | | | : ‘(M) °H : : | | Z *(M\) 885 epniy wm) | 110A ‘pe ‘A ALVIg TIAX “TOA “S'N “DOS NITANG “UW “DOUd “LNAIOS Leonakp ann Ricuarpson—On Helium in the Boiling Well at Lucan. 91 Oxygen. | Helium. | Argon. | Nitrogen. eS Aix-les- Bains, France, : c 0-0 =| 0:08 118 94-99 | 4:00 Badgestein, Austria, : é 14 | 0-169 1-18 97-28 trace Bath, England, : ; 0 — 0-12 — _— — Caldellas, Portugal, , : 24 Oly = wae 96-40 | 0-0 Kansas, U. 8. A., . F 5 — 0:06 | — — | — Lucan (St. Edmundsbury), Ireland, 0-0 0-074 | 0:95 96°88 | 2°10 Santinay, France, . 3 : == | 10°16 — | — — Yexas, U.S. A., . d 5 | — | 10-0 — _— = Spectroscopic examination of the gases. The photographed vacuum tube spectra are shown in Plate V. The spectrum of the original gas from the well is shown in No. 1. It gives the characteristic band spectrum of nitrogen, the helium and argon lines being completely absent. Nos. 2 and 3 are the spectra of helium from the well and of purchased helium respectively, while No. 4 shows the spectrum of the purchased helium (long) superimposed on that of the helium from the well (short). Eleven lines (dotted) appear in the latter, which are not visible in the spectrum of the purchased gas. They have been identified as lines of hydrogen, mercury, and argon. In No. 5 the purchased argon lines are photographed (long) with those of argon from the well (short) superimposed. ‘lhe bands of nitrogen appear faintly in the latter, together with some hydrogen lines. The coincidence of these lines with those of hydrogen is shown in No. 8, where the purchased argon is photographed full length, hydrogen two-thirds length, and argon from the well short. On examination of the latter spectrum with a lens, all the lines are seen to be continued into the hydrogen lines or those of the purchased argon. In No. 7 the spectrum of argon from the well (long) is superimposed on that of the purchased argon (short); the lines appearing in the latter are some of the lines of the blue spectrum of argon, which developed owing to variation in the tension of the contact-breaker of the coil. As the gases in the process of purification were passed more than once over heated copper oxide for removal of hydrogen, it is possible that the occurrence of hydrogen in the gases after separation is due to the presence of phosphorous anhydride in the commercial phosphoric anhydride employed for desiccation (Manley, Journ. Chem. Soc., 1922, 331.) Radioactivity of the water from the well. A test of the radioactivity of the water, kindly carried out by Dr. J. H. J. Poole, indicated the presence of about 0-01 x 10-’ gram of radium per c.¢., Le. rather less than the average for sea-water and rather more than that for river- water. The authors desire to express their thanks to Dr. R. Leeper, who granted them tacility of access to the well. CHEMICAL LABORATORY, RoyAL COLLEGE OF SCIENCE, DUBLIN. BSR i: 10, SCIENTIFIC PROCEEDINGS. VOLUME XVII. . Experiments on the Hlectrification produced by Breaking up Waiter, with Special Application to Simpson’s Theory of the Electricity of Thunder- storms. By Professor J. J. Nozan, m.a., D.sc., and J. ENRIGHT, B.A., M.SO., University College, Dublin. (June, 1922.) . Cataphoresis of Air-Bubbles in Various Liquids. By Tuomas A. McLaventin, M.sc., A.Inst.P., University College, Galway. (June, 1922.) . On the Aeration of Quiescent Columns of Distilled Water and of Solutions of Sodium Chloride. By Professor W. H. ApENEY, A.B.C.SC.I., D.SC., F.1.0. ; Dr. A. G. G. Leonard, F.R.C.SC.1., B.SC., F..c.; and A. RicHARDson, - A.R.C.SC1., ALC. (June, 1922.) . On a Phytophthora Parasitic on -Apples which has both Amphigynous and Paragynous Antheridia; and on Allied Species which show the same Phenomenon. By H. A. Larrerry and Grorce H. Prernyeriper, Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (June, 1922.) . Some Further Notes on the Distribution of Activity in Radium Therapy. By H. H. Poors, m.a., sc.p., Chief Scientific Officer, Royal Dublin Society. (June, 1922.) . Preliminary Experiments on a Chemical Method of Separating the Isotopes of Lead. By Tuomas Ditton, v.sc.; Rosatinn Cuarke, D.sc.; and Vicror M. Hincay, s.sc. (Chemical Department, University College, Galway). (July, 1922.) . The Lignite of Washing Bay, Co. ‘'yrone. By 'I'. JoHNson, D.sc., F.L.S., Professor of Botany, Royal College of Science for Ireland ; and Jane G. Gitmorg, B.sc. (Plate III.) (August, 1922.) . Libocedrus and its Cone in the Irish Tertiary. By T. Jounson, D.s0., F.L.S., Professor of Botany, Royal College of Science for Ireland; and Janz G. Giumorg, B.sc. (Plate IV.) (August, 1922.) . The Hlectrical Design of A.C. High Tension Transmission Lines. By H. H. Jerrcorr. (August, 1922.) The Occurrence of Helium in the Boiling Well at St. Edmundsbury, Lucan. By A. G. G. Leonarp, F.R.0.s0.1., PH.D., F.1.c., and A. M. RicHarpson, A.R.0.SC.1., A.I.d. (Plate V.) (August, 1922.) [Nos. 1 to 10, price 9s.] ———EEE———————— DUHLIN : PRINTED AT THE UNIVERSITY PRESS BY PONSONKY AND GIBKS. SCIENTIFIC PROCEEDINGS ROYAL DUBLIN SOCIETY. a ie RS re Vol. XVII, N.S., Nos. 11-13. DECEMBER, 1922. 11.—ON THE DETONATING ACTION OF a PARTICLES. By H. H. Pook, M.A., Sc.D., Chief Scientific Officer, Royal Dublin Society. 12.—_THE' VARIATIONS OF MILK YIELD WIIH THE COW’S AGE AND THE LENGTH OF THE LACTATION . PERIOD. By JamES Wixson, M.A., B.Sc. 13—A NOTE ON GROWTH AND THE TRANSPORT_OE VIC SUBSTANCES IN BITTER CASSAVA Reelin Ora SIMA). By T. G. MASON, M.A. BSc. / an ar, NOV 171923 *) y I o/ “@tignal Must’ ‘Honal WSs | Authors alone are responsible for all opinions expressed in their Communications. } DUBLIN: PUBIISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATK, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2- 1922. Price One Shilling and Sixpence. Roval Bublin Society. Oe eee 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 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 Iixecutive 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 IWconomic Proceedings of the Royal Dublin Society, as directed by the Committee of Science. ‘The copyright of papers so published shall be the pxreperty of the Society. ‘« Wifty 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 lMixecutive Officer for Science for further information concerning the preparation of papers and illustrations for press. Lote No. 11. ON THE DETONATING ACTION OF a PARTICLES. By H. H. POOLE, M.A., Sc.D., Chief Scientific Officer, Royal Dublin Society. [Read Novemper 28. Printed Decemper 7, 1922.] Ir has recently been shown! that a particles are capable of causing the detonation of iodide of nitrogen. This observation suggests two questions of some interest, the first being concerned with the very small probability of explosion—only one detonation occurring on the average per 10’ to 10° a particles, and the second with the possibility of danger arising from a similar effect with detonators and explosives in common use. This suggestion has already been put forward by the writer,2 but all the evidence recorded below appears to be of a reassuring nature. Taking first the consideration of the manner in which detonation occurs, it is obvious that mere ionisation of a molecule of the iodide is incapable of bringing it about, as in that case the first a particle striking the material would cause an explosion. Two possible explanations of the very small probability effect suggest themselves. Detonation may be caused by the combined effect of two or more a particles passing in rapid succession through the same very small region, and thus raising the local temperature to the detonation point. On this hypothesis the chance of an explosion occurring in a given time should be proportional at least to the square, and probably to a higher power, of the concentration of a particles. If, on the other hand, a single a particle is capable of causing detona- tion, probably by collision with an atomic nucleus, the chance of explosion should be directly proportional to the concentration, and the average number of particles per explosion should be independent of the concentration. Several series of observations on this point have been carried out, sets of specimen patches of iodide on filter-paper being prepared, and detonated by placing a small wire loop coated with RaC in the vicinity. The time in seconds (¢) that elapsed before detonation was noted. The number (7) of a particles striking the patch per second was estimated from the known activity of the source (obtained by y-ray test and the law of decay) and the geometrical conditions, allowance being made for small differences in the areas of the test 1G. H. Henderson, Nature, June 10, 1922. 2 Nature, July 29, 1922. SCIENT. PROC. R.D.S., VOL. XVII, NO. 11. R 94 Scientifie Proceedings, Royal Dublin Society. patches. As the activity decayed, the average interval before detonation increased, as shown below, where a typical set of results is recorded :-— nN ® t nt 2°74 x 107 0-4 11 x 107 245 ,, 24 BS) Tegagy 46 Orne CAs 1:2 PRO 4:00 x 10° 2°8 DUG bhi 2:98 3; 16°6 49 ,, BPEY) 11-6 Zola Ones 26:2 46. Mean diameter of specimen patch, 1°75 cms. Mean distance of source from specimen, 1-1 cms. in every case. An interval of over an hour elapsed between the fourth and fifth tests, so that the concentration of a particles only averages about one-tenth as great in the last four tests as in the first four. The concentration per square centimetre is nearly, but not quite, proportional to , as small variations occurred in the areas of the patches. The average value of nt for the first four tests is 4:35 x 10’, and for the last four 3°35 x 107, so the evidence here favours the view that né is approximately independent of the concentration. This result was amply confirmed by other series of observations, though the mean value of né was generally higher, probably owing to the specimens being less dry, so that some of the iodide was insensitive. A set of tests at distances -varying from 1 to 3 cms. indicated an increase in the value of nt with distance, ie. a decrease in the detonating efficiency of a particles with decrease of velocity. ‘The results are rendered somewhat indefinite by unavoidable differeuces in sensitivity among the test specimens, and by the large variations in number of particles per explosion, inherent in the effect. The detonating efficiency of an a particle at 1 em. is apparently about twice as great as that at 3 cms. We see, then, that all the evidence favours the theory that a single a particle is responsible for the detonation. It is natural to ascribe this to a nuclear collision, probably with either a nitrogen or a hydrogen atom. Such a collision would almost inevitably disrupt the molecule to which the atom belonged, and may even, as Sir Ernest Rutherford has shown, disrupt the nitrogen atom itself. The probabilities involved are quite of the order we should expect. The decrease in the detonating efficiency of an a particle with its velocity also accords well with this view. As regards the second question, it is evident that any compound containing a light element would occasionally have one of its molecules disrupted in a stream of a particles. This would appear to be certain in the case of the disruption of a nitrogen atom, and almost certain in any case of nuclear collision. In the case of iodide of nitrogen the disruption of one molecule appears frequently to entail the detonation of the whole. We would expect that the probability of detonation of any body sufficiently unstable to be exploded at all by a single molecule would be of the same order of magnitude as that for iodide of nitrogen. The following is a brief record of the various substances tested. In no case did detonation occur, though some of the bodies are notoriously sensitive. The PooLtE—On the Detonating Action of a Particles. 95 tests varied in duration from five minutes to about twenty hours, but the decay of the deposit rendered the period after the first two hours of negligible importance. The number of a particles in each case is calculated from the initial activity of the source, as measured by a y-ray electroscope, the curve of decay of the active deposit, and the geometrical conditions :— Substance. Area of Specimen. | Distance of Source. umber of Fulminate of Mercury, . 5 0 36 sq. mm. 10 mm. 7 x 109 5 2 0 6 0 36 20 2B op 6 x 1010 a0 95 36 ng I 5p 101! 96 », another specimen, 20 an 95 5 x 1010 Silver Azide, . 9 5 5 80 op Shes 3 x 1010 96 96 another specimen, : 80 ng 2 bp 1011 Dynamite, e i 5 9 0 20 06 8 99 4 x 1010 Nitro-Glycerine, F é 0 5 12 5 Zier i 1011 s ay piel sy AO ONG a 193, 5 oie 3 x 100 Potassium Picrate, . f : A 30 05 1 a5 2x 101! 55 95 : 5 3 4 30 3p Te pp 2x 101 The potassium picrate was considerably darkened in colour by the exposure to a rays. The total number of a particles involved in all these tests amounted to about 10”, which would have caused about 30,000 detonations in iodide of nitrogen, so that the chance of detonation for the other bodies tested is very small indeed compared with that for the iodide, and probably zero. It must be remembered, however, that unless the chance is absolutely non-existent, the danger remains, as the fact that 10” a particles did not produce an explosion does not in itself ensure safety. This number of a particles would be emitted in about three weeks in one ton of average rock, and so would be exceeded in a comparatively short time in a large mass of explosive if the radioactivity of the latter were at all comparable with that of ordinary materials. In conclusion, I wish to express my indebtedness to Professor Werner and Mr. J. V. Collins, of Trinity College, Dublin, for their kindness in preparing the majority of the specimens used. No. 12. THE VARIATIONS OF MILK YIELD WITH THE COW’S AGE AND THE LENGTH OF THE LACTATION PERIOD. By JAMES WILSON, M.A., B.Sc. {Read Novemper 28. Printed Decrmprr 12, 1922.] TEN or twelve years ago it was necessary to have scales by which individual milk yields could be “corrected” for the age of the cow and the length of her lactation. As it was known by that time that a cow’s capacity is indicated by her yield when it is at the maximum, a few weeks after calving. a scale was constructed from the average daily yields of the cows exhibited at the London Dairy Show during the ten or twelve years prior to 1909. Most of these Dairy Show cows are among the best of their kind and, therefore, age for age, of nearly like capacity. They are also shown when their yields are near their best; and it was expected that, if they were classified by age, their average yields, taken class by class, would indicate the variation of yield with age. Unfortunately, the data were not many, and, as there was no separate competition for young cows till a year or two before 1909, very few for the younger ages. The number of cows of different ages and their average daily yields were as follows :— Age: years, 9 0 3 4 5 6 7 8 9 10 11 and over Number of cows, . x 14 U 35 60 50 30 12 7 5 Pounds of milk a day, . 89°38 47 61:6 55:4 56:9 58 54:2 56:9 6071 A few years later another scale was constructed from better materials by Mr. W. Gavin, who was then making a statistical examination of the records which had been kept in Lord Rayleigh’s dairy herds in Essex. He was able to bring together the records of over three hundred cows through their first five, and gradually declining numbers of the same cows through three more, lactations. Having also seen that a cow's capacity is indicated by her yield at the flush, Mr. Gavin found that the daily fluctuations at this time could be smoothed out and the normal yield determined by reading “the maximum daily yield maintained or exceeded for not less than three entries in the record book.” He called this the “Revised Maximum,” and thus described how it is determined: “The three highest daily yields (whether entered weekly or daily) are first noted. Four cows, for example, might give 16, 16, 16—16, 16, 1716, 18, 18—16, 17, 18 quarts. The revised maximum is then taken as the highest yield common to the three entries. Thus, in all four cases quoted, it would be sixteen quarts.”! The following table gives the number of cows whose records, through eight 1 Journal of Agricultural Science”’ for October, 1913, p. 379. SCIENT, PROC, R.D.S., VOL. XVII., No, 12, s 98 Scientific Proceedings, Royal Dublin Society. successive lactations, were brought together by Mr. Gavin and the average revised maximum yields in quarts a day':— Lactation First Second Third Fourth Fifth Sixth Seventh Eighth Number of cows, : 320 313 326 328 323 221 148 83 Quarts of milk a day, . 9°38 12-8 14:2 14:9 15-4 15°85 15°51 15°48 Had Mr. Gavin been able to classify his cows by age rather than lactations, his scale would probably have stood for good, for cows’ yields vary with the years they have lived rather than with the lactations they have passed. As it stands, however, it needs to be only slightly revised by the figure for three-year-olds being raised in proportion as it was lowered originally through some of his “first lactation” cows being only two-year-olds. By reducing them to a common base we shall see how far the two scales agree ; and, since both are at their highest when the cows are eight years old, they can be reduced by multiplying the figures in each scale by the number which brings their highest figure to 100. Assuming Mr. Gavin’s first lactation cows to have been three-year-olds and the rest to have been a year older with each succeeding lactation, the result is :— Age, . 6 P 6 3 4 5 6 7 8 9 10 years Dairy Show cows’ yields a 67-8 81 89 955 981 100 98:5 98-1 by “58 Essex cows’ yields Whee 100 58°7 80'S 89-6 94 97:3 100 97-7 97°5 multiplied by pss’ When allowance had been made for the probability that some of the Essex first lactation cows were two-year-olds, these two scales agreed so closely that it was assumed they indicated very fairly how the cow’s yield varies with her age. But confidence in them was shaken by papers published in the “Journal of Agricultural Research” for September, 1919, by Dr. Raymond Pearland Mr. Miner, and in the “Transactions” of the Highland and Agricultural Society of Scotland for the same year by Dr. J. F. Tocher. These workers made use of the Ayrshire cow records published by the Scottish Milk Records Committee. Systematic milk-testing was begun among breeders of Ayrshire cattle in 1993 by the late John Speir, who, for the first five years, published annual reports in the “Transactions ” of the Highland Society, while subsequent reports have been published separately by the Milk Records Committee. In these reports the cows were not always classified upon the same plan. ‘Till that for 1910, the dates the cows were due to calve again were not given, nor were the yields of in-calf cows separated from those which were not in calf again. Consequently, Dr. Pearl and Mr. Minev’s results are based upon the average weekly yields of in-calf and not- in-calf cows during the time they were actually in milk. In the reports for 1910, 1911, and 1912 the yields are divided into two classes, according as the cows had “complete” or “incomplete” lactations, which are “a lactation which has concluded in (the recording year) and has been succeeded by the birth of a calf in that year,” and another “which, whether the cow was milking 1 Journal of Agricultural Science” for October, 1915, pp: 378 and 379, Witson—The Variations of Milk Yield with the Cow’s Age. 99 or not at the time of the last test in (the recording year), had not been succeeded by the birth of a calf in that year.” Dr. Tocher, who kept the figures for the two years 1911 and 1912 separate and made estimates or scales for each of these years,’ made use of both “ complete ” and “incomplete” lactation yields, and thus used data nearly parallel with those used by Dr. Pearl and Mr. Miner; but his scales are expressed in gallons per lactation instead of gallons per week. If these three scales are multiplied by the numbers which bring the figures at eight years old to 100, they can be compared, not only with each other, but with the two earlier scales. None of these three scales reached maximum at eight years old, but this does not spoil the comparison. Besides, it is doubtful, as we shall see later, if any of the scales indicates clearly the year at which the maximum is reached. Bi Sur evan ead | Aeaguige | Atala Assit 2 years — = 74:6 1/83 — 74 | 2 years Bp | Gees 58-7 75°8 80 80 bee Agee ee I tag 80°8 83-4 85°7 85:3 see Stk; 89 89:6 90:2 90°5 90 Woes ram 6, 95:5 94 95°7 94:5 94 Lge Tis 98-1 97°3 98°8 97-7 97:3 | Teas ghee 100 100 100 100 100 iy S888 gun 93:5 97.7 101°6 101°5 102 atone 10 gp 98-1 97-5 102 102-2 10333) he LOL ees Tp 99:2 102 103-9 Hit 5 12 pp 101-1 101 | nego) te Ge 13 43 1021 99-2 1G | 18 2g I 5, 97°3 96-5 one fa 160, 99:2 88-6 99°7 13 AG 5p | Seo 83-4 97 | G6 These columns, in which the five scales are reduced to a comparable basis, suggest that, since they are inconsistent with each other, the three Ayrshire scales may not indicate correctly the variation of milk-yield with age. It will be noticed that all five scales are closely agreed as to five-, six-, seven-, and eight-year-old cows, but the two earlier scales disagree with the Ayrshire ones about younger cows. In view of the excellence of Mr. Gavin’s method and of the probability that his figure for three-year-olds would have approached that for the Dairy Show cows had his cows been classified by age, it does not seem likely that scales can be accurate which make four-year-old Ayrshires proportionately higher yielders than four- year-olds of other breeds, three-year-old Ayrshires equal to other four-year-olds, and two-year-old Ayrshires much better than other three-year-olds. Still less is 1“ Transactions” of the Highland and Agricultural Society of Scotland for 1919, p. 246. 100 Scientific Proceedings, Royal Dublin Society. it likely that the gaps between two-year-old and three-year-old, three-year-old and four-year-old, and even four-year-old anc five-year-old yields should be as small as these scales indicate. Indeed, something was happening among Ayrshire cattle at the time the records which have been relied on were being taken which made them inappropriate for determining how milk yield varies with age. As already stated, the late John Speir got a small number of Ayrshire breeders to begin keeping milk records in 1903. One of the first results was that, when the good cows came to be distinguished from the bad by an accurate method, some which would have been retained before were now discarded, while others which would have been discarded were now retained; and the yields in milk- recording herds began to rise. It would be difficult to say whether the breeders were readier to discard younger or older cows which had been shown to be of poor capacity, and thus whether the average yields of older and younger cows were affected equally. Another result was that breeders began to select sires by their dams’ yields rather than by the old and generally worthless tests. This did not affect the yields at once, however, for, since sires cannot be used till they are yearlings and must be four years old when their eldest daughters are two, it could scarecly have affected the breed yields at all before 1906 or 1907. And, when it did have effect, it must have affected only two-year-olds the first year, two- and three-year-olds the second year, two-, three-, and four-year-olds the third year, and so on. Thus, though the average yields in milk-recording herds were increased by both the new way of discarding cows and the new way of selecting sires, the increase was not distributed equally over all ages, and the records after 1907 or so cannot be used to say how milk yield varies with age. All they can say is how yields varied with age in the year or years in which they were taken. The shifting of the relative positions of the average yields at different ages is indicated in the following diagram, in which the averages for two- to eight-year- old cows at four different periods are plotted out and traced together. The averages for each period have been multiplied by the figures which bring their eight-year-old averages to 100. The periods chosen are 1903 to 1907, 1908 and 1909, 1913 and 1920. The basal figures for the first period were extracted by Mr. Speir himself,1 who used them to find how yield varied with age ; those for the second are Dr. Pearl and Mr. Miner’s’ those for the third and fourth were extracted shortly after the reports for 1913 and 1920 were published. Thus, since scales based upon the Ayrshire records cannot represent the normal rise in milk yield fairly—even Mr. Speir’s was constructed too late—the scales based upon the London Dairy Show and Lord Rayleigh’s cows, or a combination of these, must stand in the meantime. Since Mr. Gavin’s scale is probably too low for three-year-olds, the combined scale would indicate that the yields of three-, four-, five-, six-, and seven-year-old cows should stand to their yields at eight years old as approximately 67, 80, 90, 95, and 98 to 100. As yet, little can be said about two-year-old yields. The early Ayrshire records give no clear help, because, while the average ages are approximately 34 years for three-year-olds, 44 for four-yeax-olds, and so on, the average age of two-year-olds is undoubtedly more than 24 years. If the combined figures got from the London Dairy Show and Lord Rayleigh’s cows were plotted to scale and produced to the left, the produced line would indicate that the yields of two-year-old cows whose ages average about 2% years should be to their yields at eight years old as about 50 to 100. 1 Report on Milk Records for Season 1908, p. 28. Witson—The Variations of Milk Yield with the Cow’s Age. 101 Nor can much more be said as to the yields of cows over eight years old. Dr. Pearl and Mr. Miner’s figures indicate that the cow comes to her maximum when she is somewhere between ten and thirteen years old, Dr. Tocher’s when she is about ten or eleven, Mr. Speir’s when she is ten or eleven; while the figures for 1913 indicate ten, those for 1919 nine to eleven, and those for 1920 eight years. (00, 90 80 Jo ls gp cows: 2 3 h. 5 6 7 & ysars Fic. 1.—Relative yields of cows of different ages at different periods. It is very doubtful, however, whether the Ayrshire records can be used to say when the cow comes to her maximum yield, for, when cows no longer young have to be discarded for age, the poor ones are likely to be discarded first, the best ones last: and the older age records become loaded with those of cows which are above the average. Mr. Gavin, though his materials were excellent and his eight-year-old cows actually gave more milk than his seven-year-olds, did not insist that the maximum is reached at eight years rather than at seven. He had fewer cows after 102 Fig. 2.—Milk yields during lactation periods of different lengths. Scientific Proceedings, Royal Dublin Society. lo 22 14 16 28 50 32 3% 36 38 4o 42 4h 46 LE witks 6 Is Ve \ their ages passed seven years, and he re- marked that their yields “ cannot be strictly comparable with the others, since as soon as one ceases to deal with the same number of cows throughout, the influence of selec- tion must comein. It is probable that only the best or the healthiest of the 336 cows would tend to remain in the herd after their fifth calf.” A scale to show how milk yield varies with length of lactation was even more necessary, for though the cow’s capacity can be told from her yield at the flush, the yield for the lactation period is what is nearly always published; and as the intervals be- tween successive calvings are not always the same, a scale had to be found which would indicate how much should be added to short time and how much subtracted from long time yields to bring them to the normal. A normal yield is one in which the calf which induced it is followed by another at about twelve months. In the absence of disturbing causes, the cow’s yield rises quickly for a week or two, remains near the maximum for two or three more, and then slowly but gradually falls. If the cow is not in calf again, the yield may continue, falling slowly all the time, for fifteen or eighteen months; if she is in calf, it begins, at a certain time, to fall more quickly, and ends sooner. This cer- tain time Mr. Gavin found to be when the cow had been in calf about sixteen weeks. Till about this time the cow’s yield is not affected by her being in calf. ‘Thus the average yield of a number of cows till they are about sixteen weeks in calf gives their average yield, over the same time, as if they were not in calf. The thick line in the following diagram, which is constructed from the average of Mr. Gavin's figures,’ shows the rise and fall in yield for cows which are not in calf. The line is con- tinued at both ends by dotted lines: at the beginning, because some of the cows were suckling calves during the two first weeks and the recorded yields are not the true yields; at the end, because Mr. Gavin’s figures stop at the thirty-sixth week. 1 “ Journal of Agricultural Science,” vol. v, p. 314. Witson— The Variations of Mitk Yield with the Cow’s Age. 1038 Had Mr. Gavin continued his observations over a longer time, he would probably have been able to indicate how yield varies with length of lactation. ‘This can be done, however, with the Ayrshire reports. The records in those for 1913, 1919, and 1920—the war period reports contained less material data—have first been divided into separate groups according to the cows’ ages; then each of these age groups has been subdivided into groups according to the lengths of the lactation periods in calendar months. Groups have been formed for all ages of cows between two years and twelve and for all lengths of lactations between nine months and eighteen. But, as the numbers in the remaining groups are few, we shall use only those con- taining the yields of three- to eight-year-old cows in eleven-, twelve-, thirteen-, fourteen-, and fifteen-month lactation periods. In some groups there are yields which are abnormally large and others which are abnormally small. They may be correct, but they are still abnormal and distort the averages. A two-year-old heifer may have given 1050 gallons of milk in a thirteen months’ lactation period; but if her yield be added to those of thirty other heifers whose average is nearer 600 gallons, the average for the group is increased by fifteen. Such yields must be neglected, and this has been done by eliminating all which were more than twenty-five per cent. above or below the average of their group as it stood when the extraction from the reports was completed. This method does not give absolute accuracy—no method does— for groups which were loaded originally with an unusual number of high or low figures may still have an unfairly high or low average when the elimination has been completed, but the unfairness is not great. The following table, which is constructed from the data in the Ayrshire Report for 1920, shows the average number of weeks cows of different ages were in milk during lactation periods of different lengths. The numbers of cows are enclosed within brackets :— Ages Lenetus or Lacration Pertops. Of ANE | Gqreeeeeeseese intense aan an TS a Dee Cows. 11 months. 12 months. 13 months. 14 months. 15 months. 3years| (174) 38-8 (235) 40-6 (206) 44:1 (145) 45 (69) 47-2 Ao (73) 38:8 (207) 40-6 (125) 42-2 (49) 43-9 (17) 46-7 5 (108) 38-1 (196) 40 (93) 42-7 (48) 44-6 (16) 46-6 Gane (72) 37-7 (182) 39:9" (77) 42 (26) 44:4 (18) 47 Flite (67) 87:9 (155) 39°5 (72) 42-1 (18) 44:7 (14) 46-9 843 (55) 37:2 (103) 39-4 (51) 41-7 (24) 44-3 [(7) 52-6]! ae 8°8 40-1 42°84 44-7 47-1 Mr. Gavin found that the yields of cows which are in calf begin to fall below those of cows which are not in calf about twenty-four weeks before the next calves are born (40 weeks in gestation, minus 16, during which the yield is unaffected, equal 24). The foregoing table indicates how long cows continue in milk in lactation periods of different lengths. If, then, as in the diagram, the points on the thick line where the yields of cows having lactation periods of different length begin to fall below those which are not in calf be joined with the points on the 1 This figure is omitted in the average. 104 Scientific Proceedings, Royal Dublin Society. base line at which the yields cease, the spaces between any two joining lines will indicate by how much the yields in lactations of different lengths should differ. According to the diagram, from which the calculation can be made, the yields for an eleven-month lactation should be about 20 gallons below, and those for thirteen-, fourteen-, and fifteen-month lactations about 35, 65, and 90 gallons above, that for a normal lactation. - The next and last table, constructed from the Ayrshire Reports of 1913, 1919, and 1920, gives the average yields of three- to eight-year-old cows in eleven-, twelve-, thirteen-, fourteen-, and fifteen-month lactations. Groups containing yields of less than thirty cows are omitted. The first column in every set of these columns gives the number of cows, the second their average yield in gallons, and the third the decrease (—) or increase (+) over the average yields in the normal lactation period of 12 months. I.eneru or Lacration Perron. co AN, 11 months. 12 months. 13 months. 14 months. 15 months. | {| 82 614 — 36 181 650 0 68 685 + 35 32 730 + 80 1913 | 3 | year 83 599 + 2 NS Hoy 203 631 + 34 141 652 4 5d 75 656 + 59 1919) | olds | 108 606 — 18 214 624 0 196 642 + 28 128 699 + 7d 57 705 +81 1920 Ais pM te D) 70 729 0 45 763 + 34 1913 | 4 year 39 708 — 8 141 717 O 95 721 +4 4 51 735 +18 1919 olds ( 70 740 + 2 198 738 0 120 751 +13 40 775 + 37 1920 (| 82 752 —15 57 767 0 1913 5 me year | 538 730 — 2 132 7387 0 64 733 -— 4 33 737 0 1919 olds 99 724 -19 186 7438 0 88 760 +17 47 778 + 26 1920 /| 38 746 —18 46 764 0 30 834 + 70 1913 6 yeur 54 709 —41 144 750 0 76 773 + 238 1919 olds 65 744 -18 176 762, 0 71. 769 + 7 1920 7 ( 88 721 — 38 Ue» ee) 55 783 + 24 1919 year olds l 64 7385 - 41 144 776 0 68 797 + 21 1920 8 ( 56 742 0 56 764 + 22 1919 year | olds | 50 749 — 43 100 782 0 47 805 +28 1920 | | aS ean es eet aaa | Average of — 22 0 +24 + 41 + 70 | ayerages These results are in fair agreement with those inferred from Mr. Gavin's data, and the truth may be taken to be somewhat near them, No. 13. A NOTE ON GROWTH AND THE TRANSPORT OF ORGANIC SUBSTANCES IN BITTER CASSAVA (MANIHOT UTILISSIMA). By T. G. MASON, M.A., B.Sc. {Read Novemner 28. Printed Decemper 28, 1922.] Introduction. A CONSIDERABLE amount of interest has in recent years been evinced in the quantitative aspects of plant growth. Possibly the most significant outcome of this work has been the correspondence revealed between the course of a. mono- molecular reaction and the rate of growth of a number of plants of a widely divergent habit (cf. Reed (5)). Asa result of this correspondence, Reed (6) has suggested that growth is some sort of a catalytic process, and that consequently the organism may be regarded as the end-product of a process in which a catalyst acts upon a substrate. By others (West, Briggs, and Kidd (9)) the similarity to a monomolecular reaction has been referred to an increasing differentiation into productive and non-productive tissues rather than to mass action. Some observations recently made by the present writer (3) on the growth of the cotton plant seem to accord with the latter rather than the former view. It was concluded as a result of this work that the correspondence with a monomolecular reaction was probably quite illusory, and that the falling-off in the rate of growth was, on the contrary, due to a correlation factor, which found expression in a deflection of the substances needed for growth from the growing-point to fruit developing on the basal fruiting-branches. After the primary axis has ceased to elongate it was found possible, for instance, to activate the dormant apical meristem by isolating it from the growth-inhibiting influence of this factor by removing it, and budding it on to a young cotton plant. The present inquiry was undertaken in order to ascertain whether there was any evidence of the presence of such a factor correlating the activity of the cells of the apical meristem of Bitter Cassava and the expansion of its tuberous roots. Experimental. Though Bitter Cassava is generally referred to as a shrubby plant, with long, thick, fleshy, starch-filled, cylindrical roots, this description is scarcely applicable to the plant as it grows in St. Vincent, West Indies, where the present observa- tions were undertaken; for here the growing-point of the primary axis seldom relinquishes its dominance over the lateral buds, which remain dormant. Con- sequently the plant normally remains unbranched. Twenty plants were grown in a row from cuttings in the ordinary way; the cuttings consisted of pieces of the stem of twelve to eighteen internodes. Only one plant was permitted to grow from each cutting. The height of the stem was determined weekly over a period of eighteen weeks. The measurements, which SOIENT, PROC. R.D.S., VOL. XVI, No. 13. T 106 Scientific Proceedings, Royal Dublin Society. commenced about five weeks after planting, were made from the apical bud to an ink-mark situated a couple of centimetres above the ground. A further period of ten weeks elapsed before the final measurements, when the plants were dug up and the tuberous portions of the roots and the stems weighed. A ring, approxi- mately half an inch wide, of the extra-xyliary tissues immediately above the ink- marks was removed from the stems of alternate plants. The ringing operation was done between the eleventh and twelfth weeks after the commencement of the measurements. The results reported in Table 1 show the rate at which the stem elongated in the two groups. The weights of the tuberous portions of the roots will be found in Table 2. In making the weight determinations of the stem, which are also recorded in Table 2, all the leaves were removed, and the stem cut into three equal segments, and then weighed. ‘lhe reproductive phase of develop- ment was not initiated during the course of the experiment. TABLE 1. Showing rate at which stem elongated in ringed and unringed plants. The figures in the body of the table show the height in centimetres at various dates. Cass. Rincep. UNRINGED. Nes ef 141/116 7 5 8 Mean.| 10 17 6 4 2 Mean. Nov. 6 | 26-9 17-5 22:0 318 268 25:0 | 20:9 9-1 211 29:0 276 21° 13 | 34:6 28:3 294 40-4 85:0 82:5 | 278 1386 30:0 371 36:0 28:9 20 | 448 30-4 387 525 45:9 42:5 | 372 192 408 47:9 47-7 38:6 27 | 56-2 383 489 643 575 53:0 | 46:9 255 506 588 591 48:2 4 | 687 48:2. 613 78:0 71:0 65-4 | 59:0 343 62:9 717 731 60-2 11 | 81-0 595 76-1 «921 85-4 «| 788 | 72:8 449 76-4 85-2 882TH 18 | 91-5 70:3 87° 103-9 98:5 90-4 | 85:4 55:6 86:2 96-7 1019 852 25 102°9 80:9 100°3 114°6 1109 101°9 97°2 64:6 96°7 107-6 113-4 95:9 1 | 117-1 941 125-3 133-6 129-3 «119-9 | 1121 76-5 1086 121-9 «133-3 110° 8 | 128-8 1043 138-7 145-2 189-8 181-4 | 1221 85:7 1185 131-7 144-4 120°5 15 | 137-9 1124 146-8 153-8 149-2 140-0 | 130° 93:8 126-4 139-0 153-2 128-6 22 | 144-9 119°2 152-1 160-7 «155°5 «1465S | «137-5 «= «99°S_— «13291455 155-9 1845 29 | 1538 126-8 159°5 169"1 164-1 154-7 | 145-4 1068 141-0 158-4 168-7 148-2 D) 164:2 135°7 167°5 179°5 173°8 164°1 154°5 114°8 150°3 162:8 178°6 152-2 12 | 170-7 1420 1740 186-5 180-1 170-7 | 161-1 119°9 157-8 169° 186-3 158-9 19 | 177-4 148°6 180-5 198-0 -188:1 177-5 | 1685 126-9 166-2 176-4 194-6 166-5 26 | 1825 155-0 185-7 199-1 193°9 183-2 | 174-4 1826 173-7 183-4 2028 173-4 5 | 187-6 160:5 189-6 205-1 199-9 188-5 | 180-9 1385 1808 191-4 211-6 180-6 16 | 232-0 2020 2356 257-6 249-2 235-3 | 253-7 193-3 2551 2690 283-7 251-0 Mason—Wote on Growth of Bitter Cassava. 107 TABLE 2, Weight of tuberous roots and stem, and height of stem in ringed and unringed plants at termination of experiment. No. of Wt. of Height Weight of stem in kilograms. No. Tuberous | Tuberous of Cass. of Roots Roots stem Plant. per in in Apical Middle Basal Total Plant. | kilograms. ems. Segment. | Segment. | Segment. | Weight. 2 14 2-34 283-7 0-24 Oy | 7 kOe 197 | ~4 10 2°40 269-0 0°24 (Ge | 1:09 1:97 | 6 13 2°59 255°1 0°24 0°57 0:98 1:79 UnninGED, 17 U 1:58 193°3 0°16 0°30 | 0°50 0°95 10 10 2°30 253°7 0°21 0:58 a) 1°75 | Mean, 10°8 2°24 250°96 0°22 0°55 0:92 1:69 i | 3 | 10 0°61 249°2 0-24 0-75 1:21 2°20 | | 5) 12 0°84 257-6 0-24 0°82 1:29 2°36 | 7 10 0°67 235°6 0-19 0°61 1:04 1:83 RINGED, 16 | 7 0:26 202°0 0:19 0°53 0°91 1°72 11 12 0°38 232°0 0:21 0°65 1:12 1:99 Mean, 10°2 0°55 235°28 0-21 0°67 Teri 2°02 The Weekly Growth Increments. The weekly growth increments, which are exhibited graphically in the figure, demonstrate that the velocity at which the axis elongated was in both groups slow initially, became more rapid, and then declined. At the end of the eleventh week the rate again increased, and then declined once more. As weekly measurements were not made between the seventeenth and the twenty-seventh weeks, owing to the temporary absence of the writer from the colony, information is not available as to how the rate of growth varied throughout this period. ‘lhe total growth made by the ringed plants in this ten-week period was, however, only 33°6 per cent. less than in the unringed. In comparing the growth made by the two groups, it will be observed that growth ran almost parallel up to the eleventh week, between which and the twelfth week the ringing operation was performed. It would seem that any divergences shown subsequently may confidently be referred to the removal of the ring of extra-xyliary tissues. For the three following weeks, that is to say, during the twelfth, thirteenth, and fourteenth, the rate of growth still continued to run almost parallel. After this the ringed plants commenced to lag somewhat behind the others. It would seem legitimate to refer this lag to the absence of growth in the xylem, occasioned by the cessation of, or a marked retardation in, the activity of the cambium below the ring. 72 108 Seientifie Proceedings, Royal Dublin Society. Inasmuch as growth was not sensibly retarded for some three weeks following the operation, and then not very markedly, it is to be presumed that both water and solutes continued to ascend the stem, and that consequently no pronounced blocking of the tracheae occurred as a result of morbid changes spreading inwards through the wood-parenchyma and rays from the injured region. The antiseptic properties of the poisonous latex are doubtless in some measure responsible for this. Inspection of the graphs in the figure will make it clear that the fluctuations in temperature and rainfall provide no adequate basis for interpreting the changes observed in the rate of growth. he) i=) weekly Growth bere mente wn Crs. 6 10 5 2 ye is K = 5 2 Rairfal/ o HC NC m x 0 g ' cn n 16 21 26 Time in Weeks Records were also kept of wind-velocity and evaporation, but are not presented, as no significant relationship can be traced between these factors and the changes in the rate of growth. The period the sun was above the horizon varied by just one hour and a quarter during the period occupied by the experiment. The subterranean environment was kept as uniform as possible; a dust mulch was maintained, and water added whenever it was suspected that growth might be limited by a reduction in the water-supplying power of the soil. Mason—WNote on Growth of Bitter Cassava. 109 The Transport of Organie Substances. Inspection of Table 2, im which are recorded the weights of the tuberous portions of the roots, and of the three segments of the stem, will show that the subterranean storage organs were more than four times as heavy in the unringed as in the ringed plants. J’rom this, and the greater weight of the basal portions of the stem in the latter group (cf. Table 2 2), it would seem that the transport of organic substances, especially carbohydrates, must have ceased, either completely or almost so, from the time of ringing; many of the tuberous roots of the ringed plants were, in fact, obviously shrunken. It cannot, of course, be inferred from this that the translocation of carbohydrates occurs-in the phloem rather than in xylem.’ When, however, one considers the results in the light of the evidence adduced by Dixon and Ball(2) in support of the view that the xylem is the channel for the longitudinal movement of carbohydrates, it would seem that though the actual translocation may take place in the xylem, yet the phloem must take an active part in this transmission; from the view-point of correlation such a hypothesis assumes a not improbable aspect. It will be patent that the movement of organic substances in the plant, now in one direction and now in another, according to the phase of development, must in some way be dependent on the mechanism which correlates the varied activities of the organism (cf. Smith (8)). It is a matter of common knowledge that the removal of the growing apex of a shoot in some way releases the buds immediately below from their condition of dormancy. The state of dominance and subordination is sometimes (cf. Reed and Halma (4)) ascribed to the transport of specific substances (hormones), and sometimes (cf. Child (1)) to the transmission of an excitation through the living protoplasm. In the course of the work reported here, it was observed that, shortly after the removal of the ring of extra-xyliary tissues, the bud immediately below the ring commenced to grow.’ It would seem that the removal of the ring of extra- xyliary tissue not only interrupted the transport of carbohydrates, but also blocked the transmission of the correlating agency, whatever its nature, which is responsible for the condition of dormancy in the lateral buds. Now, if it be admitted that the movement of organic substances in the plant is in some way dependent on the existence of this correlation factor, and if it be further granted that the presence of the phloem is necessary for the transmission of this factor, it ought to follow that, though the actual channel for the movement of carbohydrates may be the xylem, yet the removal of a ring of phloem would interrupt their translocation. The Internal Factors Controlling Growth. It is now possible to consider what are the internal factors which determine the changes in the activity of the cells of the apical meristem. In what follows it will be presumed that the changes observed in the rate at which the stem elongated were determined by similar changes in the activity of these cells. If it be assumed that their activity was controlled by the supply of organic substances available, 1 It will be evident that the downward flow of carbohydrates in the ringed plants cannot have ceased as a result of the blocking of the tracheae ; for, had this occurred, the ascent of water and the inorganic solutes necessary for growth would have been similarly checked. Nor, it may be remarked, does an examination of the xylem in the region of the ring afford any grounds for such a view. * These shoots were removed after they had attained a length of a couple of centimetres. 110 Scientific Proceedings, Royal Dublin Society. then the effect of the ringing ought to have resulted in an acceleration in the rate of growth. This, however, was not the case, for the banking up of organic substances, which was indicated by the relatively greater weight of the basal segments of the stem of the ringed plants (cf. Table 2), did not increase the rate of growth. From this it seems legitimate to conclude that the rate of growth was not influenced by the supply of organic nutrients available. It will be therefore evident that no correlation can exist between the rate at which the subterranean storage organs increased in weight and the activity of the cells of the apical meristem. It would seem that for a given complex of external conditions the rate of growth in Bitter Cassava is determined by autogenous changes within the cells of the apical meristem ; possibly, as suggested by Reed, by the catalytic activity of these cells. The experimental results are, in fact, in harmony with this view, provided it be assumed that growth proceeded in ‘two cycles, each of which followed the course of an autocatalytic reaction, and provided also it be assumed that these two cycles ran concurrently over a considerable portion of the latter period of growth. The differential equation da a7 he (a — *); which is characteristic of an autocatalysed reaction (cf. Robertson (7)), has been employed in making the computations shown on Table 3. In this equation x repre- sents the height of the stem, a the final height which would have been attained in each cycle, and / is a constant. When integrated the above equation becomes uv - loge — = Kt - #), where =ak, and ¢/ is the time at which the stem has grown to half its final height for each cycle, that is to say, when # = = By means of tables prepared by Robertson, the constanis K and the theoretical values of 2 are obtained for each cycle from the observed values. The values taken for a and?’ and the values of K derived from the observations made on the unringed group of plants will be seen in Table 3, the logarithms being reduced to the base ten and K modified accordingly. It will be observed that the calculated and the experimentally obtained values of z are on the whole in good accord. Mason—WNote on Growth of Bitter Cassava. 111 TABLE 3. Showing observed and computed values of rate at which stem elongated. Primary Cyctie. Srconpary Cycue. x observed « from x from x from both in cms. x x cycles. | tin weeks. log 160s t in weeks. log Th) a é | = +14 (¢ — 55). = +13(¢— 10). 215 0 23-2 23-2 | 28°9 1 30°4 30°4 38°6 2 39°0 | 39°0 48°2 3 49-4 | 49:4 60:2 4 61:0 ? é 60°8 73°4 5 736 73°6 85-2 6 86:4 86°4 95:9 7 98-9 98:9 11075 8 110°6 110°5 120°5 9 120°8 120°8 128°6 10 129°6 129°6 134°5 11 136°8 | 136°8 143-1 12 142°4 0 5:3 147-7 152°2 13 146°9 1 6:9 153°8 158°9 14 150°2 | 2 9-2 159°3 166-5 15 1528 | 3 12'1 1649 | 173-4 18 154-7 4 15°6 170°3 180°6 17 166:2 5 20°1 176°3 251-0 27 160 15 89:9 249-9 Summary. 1. The work reported in this paper was undertaken in order to ascertain whether there was any evidence for the presence of a factor correlating the activity of the cells of the apical meristem and the growth of the tuberous roots of Bitter Cassava. 2. Measurements of stem height were made weekly over a period of eighteen weeks, and also at the end of the twenty-seventh week. Half the plants were ringed fifteen weeks before the termination of the experiment. 3. It was found that the rate of growth of the ringed plants was not affected by the operation for a period of about three weeks; it then commenced to lag behind the unringed plants, 112 Scientific Proceedings, Royal Dublin Society. 4. The weight of the tuberous roots of the ringed plants was approximately a quarter that of the unringed ; the weight of the stem, on the other hand, was more than 1:2 times as heavy. 5. It was concluded that the activity of the cells of the apical meristem was not controlled by the supply of organic substances available, but was, on the contrary, determined by autogenous changes within the growing point. No evidence was obtained of the presence of a factor correlating the activity of the apical meristem and the growth of the tuberous roots. 6. It was pointed out that the experimental results were in accord with the view that the rate of growth of the stem was conditioned by the catalytic activity of the cells of the apical meristem. LITERATURE CITED. 1. Cuitp, C. M.—Certain Aspects of the Problem of Physiological Correlation, 1921. Amer. Jour. Bot., vol. viii, No. 6, pp. 286-295. 2. Dixon, H. H., and Batu, N. G.—Transport of Organic Substances in Plants, 1922. Nature, vol. cix, No. 2730, pp. 236-237. 3. Mason, T. G.—Growth and Correlation in Sea Island Cotton, 1922. West Indian Bul., vol. xix, No. 2, pp. 214-2538. 4, Reep, H. S., and Hata, F. F.—l'he Evidence for a Growth Inhibiting Sub- stance in the Pear Tree, 1919. Plant World, vol. xxii, No. 8, pp. 239-247. 5. Reep, H. §8.—'he Nature of the Growth Rate. Jour. of Gen. Phys., 1920, vol. ii, No. 5, pp. 545-561. 6. Rep, H. S.—Slow and Rapid Growth. Amer. Jour. of Botany, 1920, vol. viii No. 8, pp. 327-332. 7. Ropertson, T. B.—Tables for the Computation of Curves of Autocatalysis, with special reference to Curves of Growth, 1915. Univ. of California. Pub. in Physiology, vol. iv, No. 21, pp. 211-228. 8. Smrru, A. M.—On the Internal Temperature of Leaves in Tropical Insolation, with special reference to the effect of their Colour on the Temperature ; also Observations on the Periodicity of the Appearance of young Coloured Leaves growing in Peradeniya Gardens, 1909. Annals. Roy. Bot. Gard., Peradeniya, vol. iv, Part V, pp. 229-298. 9. Wust, C., Brices, G. E., and Kipp, F.—Methods and Significant Relations in the Quantitative Analysis of Plant Growth. New Phytologist, 1920, vol. xix, Nos. 7 and 8, pp. 200-207. 10. ilo 12. 13. SCIENTIFIC PROCEEDINGS. VOLUME XVII. . Experiments on the Electrification produced by Breaking up Water, with Special Application to Simpson’s Theory of the Electricity of Thunder- storms. By Professor J. J. Nouan, m.a., p.sc., and J. HinRIGHT, B.A., M.SO., University College, Dublin. (June, 1922.) . Cataphoresis of Air-Bubbles in Various Liquids. By Tuomas A. Mchaveuun, m.sc., a.inst.p., University College, Galway. (June, 1922.) On the Aeration of Quiescent Columns of Distilled Water and of Solutions of Sodium Chloride. By Professor W. H. Apmnry, A.R.C.SC.1., D.SC., F.1.C. 5 Dr. A. G. G. Leonarp, F.R.c.scJ., B.Sc., F.1.c.; and A. Rrcwarpson, A.R.C.SC.1., a.t.c. (June, 1922.) . On a Phytophthora Parasitic on Apples which has both Amphigynous and Paragynous Antheridia; and on Allied Species which show the same Phenomenon. By H. A. Larrerry and Guorcs H. Perayerien, Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (June, 1922.) . Some Further Notes on the Distribution of Activity in Radium Therapy. By H. H. Pootn, m.a., sc.p., Chief Scientific Officer, Royal Dublin Society. (June, 1922.) . Preliminary Experiments on a Chemical Method of Separating the Isotopes of Lead. By Tuomas Ditton, v.sc.; Rosaninp Crargs, p.sc. ; and Vicror M. Hincny, s.sc. (Chemical Department, University College, Galway). (July, 1922.) The Lignite of Washing Bay, Co. 'l'yrone. By 'l’. Jounson, D.Sc., F.L.S., Professor of Botany, Royal College of Science for Ireland; and Jane G. GitmorE, B.sc. (Plate III.) (August, 1922.) Libocedrus and its Cone in the Irish Tertiary. By T. Jounson, D.s¢., F.L.S., Professor of Botany, Royal College of Science for Ireland; and Jann G. GitmorE, B.sc. (Plate 1V.) (August, 1922.) . The Hlectrical Design of A.C. High Tension Transmission Lines. By Hi. H. Jerrcorr. (August, 1922.) The Occurrence of Helium: in the Boiling Well at St. Edmundsbury, Lucan. By A. G. G. Lionarp, ¥.R.c.sc.1., PH.D., F.1.c., and A. M. RicHarpson, A.R.O.SC.1., A.r.0. (Plate V.) (August, 1922.) [ Nos. 1 to 10, price 9s. | On the Detonating Action of a Particles. By H. H. Poonn, u.a., sc.v., Chief Scientific Officer, Royal Dublin Society. (December, 1922.) The Variations of Milk Yield with the Cow’s Age and the Length of the Lactation Period. By James Winson, m.a., B.sc. (December, 1922.) A Note on Growth and the Transport of Organic Substances in Bitter Cassava (Manihot utilissima). By T. G. Mason, m.a., B.sc. (December, 1922.) [Nos. 11 to 18, price 1s. 6d. } DURDLIN : PRINTED AY THE UNIVERSITY PRESS BY PONSONBY AND GIBBS. caer ae LRN ie THE SCIENTIFIC PROCEEDINGS OF THE ROVWALS DUBLIN: SOCIETY: Vol. XVII, N.S., Nos. 14-19. FEBRUARY, 1923. 14.-THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON DIPHENYLURETHANE. By Hucu Ryay, D.Sc., and ANNE DONNELLAN, M.Sc., University College, Dublin. 15.—THE ACTION OF THE OXIDES AND THE OXYACI DS NITROGEN ON ETHYL-O-TOLYLURETHANE. By Hig Rian, D.Sc., and NICHOLAS CULLINANE, PH.D., University College, Dublin, i nk IOV ia eerarg 16.—THE ACTION OF THE OXIDES AND THE OXY‘MACIDS OF NITROGEN ON ETHYL-PHENYLURETHANE. By Hbaw/Bxan,, is esi ol) D.Sc., and ANNA ConnoLLy, M.Sc., University College, Dublin. cual” — 17.—THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON PHENYL-BENZYLURETHANE. By Hucu Ryan, D.Sc., and James L. O'Donovan, M.Sc., University College, Dublin. 18.—THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON THE PHENYLUREAS. By Huau Ryan, D.Sc., and Prrer K. O’Toous, M.Sc., University College, Dublin. 19.—THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON PHENYL-METHYLUREA. By Hucu Ryan, D.Sc., and MicuarL J. Sweenry, M.Sc., University College, Dublin. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THK ROYAI DUBLIN SOCIETY LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1923. Price Four Shillings. Roval Bublin Society. FOUNDED, A.D. 1731. INCORPORATED 1749. SCIENTIFIC MEETINGS. Tun 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 recula- tions governing the presentation of communications at Scientific Meetings of the Society, and the publication of papers in the Scientific or Mconomie 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 Heonomic 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 tlie Chief Mxecutive Officer for Science for further information concerning the preparation of papers and illustrations for Fm J No. 14. THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON DIPHENYLURETHANE. By HUGH RYAN, D.Sc., AND ANNE DONNELLAN, M.Sc., University College, Dublin. [Read Drcenper 19, 1922. Printed Fasrvary 22, 1923.] INTRODUCTION. Ir has been shown by one of us, in conjunction with Miss P. Ryan (Proc. R.LA., xxxiv, B, pp. 194 and 212), that diphenylnitrosamine can be converted by nitric acid into nitro derivatives of diphenylamine and of diphenylnitrosamine more easily and more smoothly than diphenylamine. Thus diphenylnitrosamine reacted with nitric acid at the ordinary temperature and at low concentrations of the acids with formation of mononitro- and dinitro-diphenylamines with their nitro- samines, a trinitro- and a tetranitro-diphenylame. Under the same conditions diphenylamine reacted less smoothly with nitric acid with formation of similar nitro derivatives, together with relatively large amounts of brown resinous products, obviously arising from a secondary oxidation process. The difference in the behaviour of the two substances is probably due to the protective influence of the nitroso group during the nitration, and in this connexion the results which we obtained during the nitration of diphenylurethane may be of interest. By the action of concentrated nitric acid on diphenylurethane Hager(Ber. Dtsch. Chem. Ges., xviii (1885), p. 2574) obtained 10-dinitro-diphenylurethane and a syrupy substance, which proved to be 2°10-dinitro-diphenylurethane. So faras we are aware, no other nitro derivatives of the substance have been obtained, nor has its behaviour towards nitric acid at low concentrations of the interacting substances been hitherto investigated. In our experiments we found that cold concentrated nitric acid converted diphenylurethane into 4-nitro-diphenylurethane ; and we established the constitu- tion of the latter substance by converting it, by means of alcoholic potash, into 4-nitro-diphenylamine melting at 133°-134° C. With fuming nitric acid, on the other hand, we obtained the 4:10- and 2:10-dinitro derivatives, already described by Hager (Joc. cit.). Tetranitro-diphenylurethane, hitherto unknown, was formed by the action of a mixture of sulphuric and nitri¢ acids on diphenylurethane, and also on 4-nitro- diphenylurethane. Since the tetranitro compound was converted by concentrated hydrochloric acid into 2-4°8-10-tetranitro-diphenylurethane, it must be regarded ag 2-4-8-10-tetranitro-diphenylurethane. By the further nitration of the tetra- nitro-urethane symmetrical hexanitro-diphenylamine was formed. SCIENT. PROC. R.D.S., VOL, Xvi, No. 14. U 114 Scientific Proceedings, Royal Dublin Society. At low temperatures and concentrations, in acetic acid solution, apparently no reaction occurred between the acid and the urethane, even on prolonged standing. In carbon tetrachloride solution, however, the acid and the urethane reacted with difficulty. When the amount of the acid was equivalent to three molecular pro- portions and the mixture was allowed to remain at the room temperature for six weeks, 4-nitro-diphenylurethane was formed, while with four and with six molecular amounts of the acid the main products were 4:10- and 2:10-dinitro-diphenylur- ethanes. On the other hand, nitrogen peroxide, in carbon tetrachloride solution, converted it more readily into an oily mixture, from which, by decomposition with alcoholic potash, 4:10- and 2°10-dinitro-diphenylamines were obtained. EXPERIMENTAL. A.—ACTION OF NITRIC ACID‘ON DIPHENYVLURETHANE. I.—Aetion of Cold Concentrated Nitric Acid. Five grams of diphenylurethane were added slowly to seven and a half cubic centimetres of nitric acid (sp. g. 142), which was kept well cooled during the addition of the urethane. The mixture was allowed to remain for five days in a stoppered bottle at the temperature of the laboratory. During this time a dark- coloured oil separated. The contents of the flask were poured into a considerable excess of water, and the aqueous layer was decanted from the underlying heavy brown oil. ‘he latter was dissolved in warm alcohol, from which a yellowish-white solid separated on cooling. The solid was removed by filtration. It crystallised from alcohol in clusters of colourless prismatic crystals, which melted at 68° C. A mixture of this solid with the original urethane, which melted at 72° C., melted at about 35° C, It gave on analysis the following results :— 02051 g. of the substance gave 17°8 c.c. of moist nitrogen at 17°8° C. and 768 mm., corresponding to N 100. C,,H,,0,N, requires N 9°8. A small quantity of the solid was heated with alcoholic potash for a few hours on a water- bath; the solution was diluted with water, and the solid which separated was filtered off. When the solid was crystallised from a mixture of chloroform anu alcohol, yellow leafy crystals melting at 135°-134° C. were obtained. These crystals gave a violet coloration with a mixture of concentrated sulphuric acid and a trace of sodium nitrite, and their melting point was not affected by addition to them of 4-nitro-diphenylamine, with which the substance was identical. It follows, therefore, that the mononitro compound, melting at 68°C., must be 4-nitro-diphenylurethane. 4- Nitro-diphenylurethane consists of colourless prisms which are sparingly soluble in petroleum ether, and very soluble in ether, alcohol, chloroform, benzene, or acetone. l1.—Action of Cold Fuming Nitrie Acid. Five grams of diphenylurethane were added slowly, with frequent shaking of the mixture, to ten cubic centimetres of fuming nitric acid. The dark-coloured solution was allowed to remain at the laboratory temperature for two days, and was then diluted with several times its volume of cold water. The flocculent precipitate which separated was washed with water, and dissolved in hot alcohol, from which a yellow solid and a brown oil separated on cooling. Ryan & Donnettan—Action of Oxides, &c., on Diphenylurethane. 115 The solid, when recrystallised from alcohol, melted at 133°-134° C., and gave on analysis the following results :— 0:2363 g. of the substance gave 26°3 c.c. of moist nitrogen at 13°5° C.and 764 mm., corresponding to N 13-2. C,;H,,0;N; requires N 12:7. Its solution in cold alcoholic potash was at first colourless, but on standing gradually developed the violet coloration characteristic of 4:10-dinitrvo-diphenyl- amine. The crystals dissolved in concentrated sulphuric acid, forming a yellow solution, the colour of which was not affected by the addition of sodium nitrite. When the dinitro-urethane was heated with alcoholic potash and the solution was diluted with water, a yellow solid was precipitated, which, when filtered, washed, and recrystallised from alcohol, consisted of yellow crystals melting at 214°-216°C. he melting point was not affected by addition to the substance of 4:10-dinitro-diphenylamine, with which it was therefore identical. The dinitro-diphenylurethane melting at 133°-134°C. was therefore the 4-10-dinitro-diphenylurethane already obtained by Hager (Joe. cit.). 4:10-Dinitro-diphenylurethane consists of acicular faint yellow prisms, which are scarcely soluble in petroleum ether, soluble with difficulty in alcohol, soluble in ether or chloroform, and very soluble in acetone. The brown oil, which was formed together with 4:10-dinitro-diphenylurethane, and which has been mentioned above, did not crystallise. It was therefore decomposed by heating with alcoholic potash, and the resulting diphenylamine derivative, when recrystallised from xylene, was obtained mainly in the form of red crystals, which melted at 219° C., and proved to be 2°10-dinitro-diphenylamine. IIl.— Action of a Mixture of Concentrated Sulphuric and Nitrie Acids. In this experiment five grams of diphenylurethane were added to a cold mixture of 74.c.c. of concentrated nitric acid and 15 ¢.c. of concentrated sulphuric acid. ‘The mixture became hot, with separation of a dark-coloured oily substance, and was then heated on the water-bath for five or six hours. The brown solid which separated on standing overnight, was freed from the acids by diluting the mixture with cold water and filtering. The solid was washed, dried, and recrystallised from chloroform and alcohol. The yellowish-white crystals, thus obtained, melted at 184°-185° C., and gave on analysis the following results :— 0:1785 g. of the substance gave 25°8 c.c. of moist nitrogen at 19° C. and 766°5 mm., corresponding to N 16°76. C,;H10,N; requires N 16°63. When this substance was heated with alcoholic potash a black, tarry mass was obtained. It was, however, decomposed without resinification by heating it with concentrated hydrochloric acid in a sealed tube to 160°-180°C. for eight hours. The contents of the tube were neutralized with sodium carbonate, and the solid reaction product was recrystallised from benzene. It consisted of light yellow crystals, which melted at 198°-200° C., and proved to be identical with 24°8°10- tetranitro-diphenylamine. ; The tetranitvo derivative, which melted at 184°-185° ©., was therefore 2-4:8°10- tetranitro-diphenylurethane. 2-48:10-Tetranitro-diphenylurethane consists of yellowish-white platy prisms, which are sparingly soluble in petroleum ether, cold alcohol, or ether; soluble in hot alcohol, cold chloroform, or benzene; and readily soluble in acetone. 116 Scientific Proceedings, Royal Dublin Society. IV.—Aetion of Nitric Acid at low Concentrations. (a) In Acetic Acid Solutions.—(1) Two grams of diphenylurethane were dis- solved in fifty grams of glacial acetic acid, and four molecular proportions of fuming nitric acid were added. The solution, which had a light greenish colour, was allowed to remain for ten weeks at the laboratory temperature, and was then poured into water. The solid which separated was filtered and recrystallised from alcohol. ‘The colourless crystals thus obtained melted at 72° C., and proved to be diphenylurethane. As the weight of the solid recovered was almost two grams, little, if any, reaction can have occurred between the nitric acid and the diphenylurethane. (2) Three similar solutions containing respectively one, two, and three molecular amounts of nitric acid to two grams of diphenylurethane in glacial acetic acid were prepared and allowed to remain at the room temperature for ten weeks. At the end of that time the diphenylurethane was in each case recovered unchanged. It is evident, therefore, that at the ordinary temperature any nitration of diphenylurethane by means of nitric acid in an acetic acid solution which may occur must be a very slow one. (b) In Carbon Tetrachloride Solutions —(1) Two grams of diphenylurethane were dissolved in fifty grams of carbon tetrachloride, and 0°7 ¢.c. of fuming nitric acid (two molecular amounts) was added. After a few days the solution had a lemon- yellow colour, and the shght upper layer was brown. At the end of five weeks the carbon tetrachloride was evaporated, and the residue, which consisted of a colour- less solid intermixed with a small quantity of oil, was dissolved in hot alcohol. The solid which separated from the alcohol on cooling proved to be unchanged diphenylurethane, melting at 72°C. The small amount of oil which was simul- taneously formed was insufficient for further examination. (2) In a second experiment three molecular quantities of nitric acid were added to a solution, similar to the last, of two grams of diphenylurethane in carbon tetrachloride. As in the last experiment, the solution was yellow, and the upper layer was brown, in colour. After three weeks crystals began to separate, and at the end of another week the carbon tetrachloride was evaporated. The oily, crystalline solid which remained was washed with ether and alcohol. It consisted of colourless crystals which melted at 68° C., and proved to be identical with 4-nitro-diphenylurethane. (3) In another experiment four molecular quantities of nitric acid were added to the solution, which in this case rapidly acquired a deep yellow colour, and on standing overnight a brown oil separated. After the first four weeks yellowish- white crystals formed. When the mixture had remained ten weeks in all at the laboratory temperature, the solid was filtered, washed with a little carbon tetra- chloride, dried, and recrystallised from alcohol, from which it separated in the form of yellow crystals, which melted at 133°-134° C., and proved to be identical with 4:10-dinitro-diphenylurethane. When the carbon tetrachloride was evaporated from the filtrate from the crystals just mentioned, a brown oil remained. As this oil did not crystallise, it was decomposed by heating on the water-bath with alcoholic potash for about an hour. The mixture was diluted with water, and the solid which separated was filtered and dried. When heated with chloroform part of the solid dissolved ; the solution was filtered while hot, and from it 4:10-dinitro-diphenylamine, melting at 214°C., was deposited on cooling. The portion which remained undissolved was crystallised from benzene, from which it separated in the form of red crystals, melting at 219°C., and which were identical with 2°10-dinitro- diphenylamine. Ryan & Donneritan—Action of Oxides, &c., on Diphenylurethane. 117 The oil formed by the action of nitric acid on the urethane was a solution of 4:10-dinitro-diphenylurethane in 2°10-dinitro-diphenylurethane. (4) The behaviour of a solution of diphenylurethane in carbon tetrachloride, to which a quantity of nitric acid corresponding to six molecular proportions of the acid had been added, was quite similar to the last—the products formed in this case being also 4:10- and 2°10-dinitro-diphenylurethanes. : B.—ACTION OF NITRIC ACID ON 4-DINITRO-DIPHENYVLURETHANE. (1) 4-Nitro-diphenylurethane (1 gram) was added slowly with constant shaking to 4 ¢.c. of concentrated nitric acid (sp. g. 142). The solid dissolved, forming a yellow solution. On examining it after two days it was found that the nitric acid had not reacted with the nitro-urethane, which was recovered unchanged. (2) One gram of 4-nitro-diphenylurethane was added to a cold mixture of 4c.c, of concentrated nitric acid and 8 c.c. of concentrated sulphuric acid. The solution was allowed to stand at the temperature of the room for two days. On pouring it into water a white solid separated, which, when recrystallised from chloroform, melted at 184°-185° C., and proved to be 2:4°8-10-tetranitro-diphenylurethane. Hence 4-nitro-diphenylurethane on nitration by “mixed acids” in the cold goes into 2°4°8:10-tetranitro-diphenylurethane. C.—ACTION OF NITRIC ACID ON 2-4°8:10-TETRANITRO- DIPHENYLURETHANE, One gram of tetranitro-diphenylurethane was added slowly to a cold mixture of 4c.c. of concentrated nitric acid and 8 c.c. of concentrated sulphuric acid, which was then heated on the water-bath for about seven hours. On standing overnight a yellow solid separated. The mixture was diluted with water, and the solid was crystallised from chloroform, in which it was only sparingly soluble. It consisted of yellow prisms, which melted at 246° C., and were found to be identical with 2°4°6°8:10°12-hexanitro-diphenylamine. D.—ACTION OF NITROGEN PEROXIDE ON DIPHENYLURETHANE. A stream of nitrogen peroxide (generated by heating lead nitrate) was passed into a solution of two grams of diphenylurethane in fifty grams of carbon tetra- chloride. The yellow solution became wine-coloured on standing at the tempera- ture of the laboratory, and at the end of a week a brown oil separated. The solution was again saturated with nitrogen peroxide. When the solution remained at the laboratory temperature for a fortnight, the carbon tetrachloride was dis- tilled off, and a brown oil was left behind. As the oil did not crystallise, it was boiled with alcoholic potash for an hour; the mixture was diluted with water, and the solid which separated was filtered and dried. When heated with chloro- form part of the solid dissolved; the solution was filtered while hot, and from it 4-10-dinitro-diphenylamine, melting at 214°C., was deposited on cooling. The portion which remained undissolved was crystallised from benzene, from which tt separated in the form of red crystals, melting at 219°C., and which were identical with 2:10-dinitro-diphenylamine. SCIENT. PROG. R.D.S., VOL. XVII, No. 14. x 118 Scientific Proceedings, Royal Dublin Society. H.—ATTEMPTS TO SYNTHESISE THE NITRO-DIPHENYLURETHANES. (1) A small quantity of 4-nitro-diphenylurethane (0°5 gram) was dissolved in xylene, and the solution heated with 0°5 gram of chloroformic ester in a sealed tube at 160°C. for eight hours. The excess of chloroformic ester and the xylene were then distilled off, and the residue crystallised from alcohol, from which it separated in the form of yellow leaves, melting at 133°-134° C., and which proved to be the original solid—4-nitro-diphenylamine. (2) Equal quantities of 4-nitro-diphenylamine (1 gram) and chloroformic ester were heated with a little alcohol in a sealed tube at 160°C. for ten hours. In this experiment also the 4-nitro-diphenylamine was recovered unchanged. (3) In another experiment some 2°10-dinitro-diphenylamine was dissolved in xylene, and excess of chloroformic ester was added to the solution, which was then heated to boiling under a reflux condenser for ten hours. In this case also the solid recovered proved to be the original 2:10-dinitro-diphenylamine. (4) Equal quantities of symmetrical tetranitro-diphenylamine and chloro- formic ester were heated in a sealed tube at 180°C. for eight hours. Large yellow crystals were formed, which on recrystallisation from xylene melted at. 199° C., and were therefore the original solid recovered unchanged. SUMMARY. 1. Nitric acid at the ordinary temperature and at low concentrations in acetic: acid solution had apparently no action on diphenylurethane. Under similar conditions in carbon tetrachloride solution, nitric acid converted the urethane into its 4-nitro, 4:10-dinitro, and 2:10-dinitro derivatives. 2. Cold concentrated nitric acid (sp. g. 1°42) converted the urethane into its. 4-nitro derivative, while fuming nitric acid under similar conditions formed a mixture of the 4:10- and the 2:10-dinitro-diphenylurethanes. 3. The urethane, or its 4-nitro derivative, was nitrated by a mixture of con- centrated nitric and sulphuric acids into 2°4:8:10-tetranitro-diphenylurethane, and the latter reacted further with the hot mixed nitrating acids forming 2'4-6°8'10:12-hexanitro-diphenylamine. 4, Nitrogen peroxide reacted with diphenylurethane in carbon tetrachloride solution with the formation of the 4:10- and the 2-10-dinitro derivatives of the urethane. 5. Notwithstanding the similarity in structure between diphenylurethane and diphenylnitrosamine, the former substance reacts with nitric acid much less readily than the latter, the nitration at low temperatures and concentrations. stopping at the dinitro stage. In conclusion we wish to state that the above research was undertaken at the request of the Research Section of Nobel’s Explosives Company, to whom we are indebted for a grant in aid of the investigation. eel 19s8) No. 15. THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON ETHYL-O-TOLYLURETHANE. By HUGH RYAN, DSc., AND NICHOLAS CULLINANE, Px.D., University College, Dublin. [Read Decemuer 19, 1922. Printed Frsruary 22, 1923.] INTRODUCTION. Iy previous communications it was shown that a substituted diphenylamine, such as diphenylnitrosamine [H. and P. Ryan, Proc. R.I.A., xxxiv, B, pp. 194, 212], nitrates more easily and more smoothly than diphenylamine. It was, however, found later [H. Ryan and A. Donnellan, p. 113] that the urethane formed by the substitution of a carbethoxy group for the imino hydrogen atom of diphenylamine was less easily nitrated than the parent amine. It appeared of interest, therefore, to study the behaviour of other aromatic urethanes towards the oxides and the oxyacids of nitrogen at low temperatures and concentrations of the interacting substances. This study, although at. first necessarily a qualitative one, might be expected to throw some light on the influence of a urethane group on the ease of nitration of an aromatic nucleus. Ethyl-o-tolylurethane was selected for our experiments. Since the imino hydrogen atom of the parent substance, o-tolylurethane, was replaced by the ethy] radical, the possibility of the formation of an easily nitratable nitroso compound was excluded. o-Tolylurethane has been obtained by Lachmann [Ber. Dtsch. Chem. Ges., xii (1879), p. 1349; cf. Merz, zbid., vi (1873), p. 444; Cosack, did., xii (1879), p- 1450; Nevile and Winther, ) It was more expeditious, however, to obtain the urea by allowing the reaction to take place at the temperature of boiling spirit. 25 g. of diphenylurea chloride were heated to boiling with 100 cc. of methylated spirit, and a stream of ammonia was passed into the mixture for a quarter of an hour. The undissolved diphenylurea chloride went into solution, and long colourless needles of as-diphenylurea separated. On allowing the mixture to cool a further quantity of the urea separated. The solution was filtered from the separated solid. This, when well washed with water to remove ammonium chloride, melted at 189°C. Recrystallisation from alcohol did not raise the melting point. The as-diphenylurea obtained was therefore pure. By evaporating the spirit solution down to small bulk a further quantity of substance was obtained, which, when washed with water, left the as-diphenylurea. The yield was theoretical. B.—ACTION OF NITROUS FUMES ON AS-DIPHENYLUREA. as-Diphenylurea (5 g.) was dissolved in 100 ¢. of glacial acetic acid, and nitrous fumes were passed in at room temperature. The solution became red, then green, with the evolution of a gas throughout the liquid. On allowing to remain at room 150 Scientific Proceedings, Royal Dublin Society. temperature overnight a light yellow solid separated as needles. After two days the solution was filtered from the solid. The latter, after washing with alcohol, melted with decomposition at 145°-146°C. An attempt to recrystallise this substance from acetone seemed to decompose it, yielding a red substance melting with decomposition between 145°-162°C. ‘The original substance gave a violet colouration with alcoholic potash. Some of this substance was boiled for an hour with glacial acetic acid, in order to complete the decomposition effected by the acetone, during which time brown fumes were observed in the neck of the flask. This process yielded a red solution, from which, by precipitation with water, a red substance was obtained melting between 138°-155° C. By fractional crystallisation of this substance from acetic acid, a compound was obtained as prisms melting at 213°C. It was 44’- dinitro-diphenylamine, as the addition of the latter substance to it did not affect its melting point. The original compound was enecetore probably 4°4’-dinitro-diphenyl-nitrosamine (M.P. 150°C), which is very difficult to obtain pure. This would account for the low melting point of it as obtained above. Further, 4:4’-dinitro-dipheny]l-nitrosamine gives, like this compound, a violet colouration with alcoholic potash, and yields, on elimination of the nitroso radical by boilings with acetic acid, 4:4’-dinitro-diphenyl- amine. No pure substance was obtained from the original acetic acid solution. C.—ACTION OF NITROGEN PEROXIDE ON AS-DIPHENYLUREA. as-Diphenylurea (5 g.) was dissolved in 100g. of glacial acetic acid and saturated with nitrogen peroxide in the cold. A gas was evolved throughout the solution. After allowing to remain at room temperature for two days a yellow substance separated out. This was identical with the substance obtained by the action of nitrous fumes on the urea, and was therefore 4-4’-dinitro-diphenyl- nitrosamine, mixed probably with some of its isomers. No pure substance was obtained from the red glacial acetic acid fibrate. D.—ACTION OF NITRIC ACID ON AS-DIPHENYLUREA. I—Jn the Presence of Solvents. (a) Glacial acetic acid.—as-Diphenylurea (5 g.) was dissolved in 100 g. of glacial acetic acid and treated with one, two, four, and six molecular parts of nitric acid in the cold. On allowing to remain at the temperature of the room for two months, yellow colourations were developed in every case. This colouration persisted in the solids obtained by precipitation with water. The solid so obtained consisted, however, in every case of the unchanged as-diphenylurea. (a) Carbon tetrachloride.—One, two, four, and six molecular parts of nitric acid were added to 5 g. of as-diphenylurea suspended in carbon tetrachloride. The mixtures were allowed to remain at room temperature for two months. A yellow solution and a black oil were obtained in every case. In the bottles to which four and six molecular parts of nitric acid had been added a red substance appeared towards the end of the two months. On addition of ether to the bottles with one and two parts of nitric acid the oil was dissolved, with the separation of a white crystalline substance. This was the unchanged urea. By neutralizing with barium carbonate, filtering, and evaporating to dryness the ether-carbon-tetrachloride solution yielded a small quantity of black oil, from which, however, no pure solid was obtained. Ryan & O’Tooitu—Aetion of Oxides, §c., on the Phenylureas. 1651 The solutions in the bottles to which four and six parts of nitric acid had been added were filtered from the red solids. ‘The latter, after removal of the oily matter with cold glacial acetic acid, yielded in each case a red substance (M.P. 170°-190° C.), from which no erystalline substance was obtained. ‘The carbon tetrachloride solutions were neutralized with barium carbonate, filtered, and evaporated. In each case a black oil was obtained. In both cases there was much free nitric acid present. I1.—Zn the Absence of Solvents. 5g. of as-diphenylurea were added in small quantities to twelve molecular parts of nitric acid cooled in salt and ice. The urea dissolved, giving a red solution. On allowing the mixture to remain at room temperature overnight, a gas was slowly evolved, and a yellow solid separated as prisms. This, when recrystallised from glacial acetic acid, yielded yellow leaves melting at 200°C. It was 2:4:2’4’- tetranitro-diphenylamine, as the addition of the latter substance to it did not affect its melting point. A further quantity of this substance was obtained from the nitric acid solution. E,—ACTION OF NITROUS ACID ON AS-DIPHENYVLUREA. One, two, four, and six molecular parts of sodium nitrite dissolved in a small amount of water were added to 5g. of as-diphenylurea in 100g. of glacial acetic acid in the cold. The mixtures were allowed to remain at room temperature for fourteen days, when all the urea was recovered unchanged. 4,—TRIPHENYLUREA. The action of the oxides of nitrogen on triphenylurea has not been previously examined. ‘Triphenylurea in glacial acetic acid solution reacted easily with both nitrous fumes and nitrogen peroxide, givingin each case the same trinitro-triphenyl- urea. ‘he constitution of this substance was not determined. Reudler (/oc. cit.) nitrated triphenylurea in the cold, and obtained a product which gave no nitramine reactions. He obtained from it, after boiling with alcohol, 2°4:2"-4’-tetranitro-diphenylamine and 2-4-dinitro-phenylurethane. By using larger excess of nitric acid he obtained 2°4°2’-4’-tetranitro-diphenylamine and 2°4-6-trinitro-phenylurethane. Iu a similar experiment we only isolated 2°4:2’-4’- tetranitro-diphenylamine. Triphenylurea was nitrated for long periods at room temperature and at low concentrations in acetic acid solution and in carbon tetrachloride suspension. In acetic acid solution with one and three molecular amounts of nitric acid the urea was recovered unchanged; with six and nine molecular parts, however, a dinitro- and the trinitro-triphenylurea were formed in both cases. In carbon tetrachloride suspension, with one molecular amount of nitric acid, the dinitro-triphenylurea was formed ; with three and six amounts the dinitro- as well as the trinitro-triphenylurea was obtained; and with nine amounts a pentanitro-triphenylurea was formed. Nitrous acid was without action on triphenylurea in acetic acid solution even on prolonged standing. A.—ACTION OF NITROUS FUMES ON TRIPHENYLUREA. 5 g. of triphenylurea were dissolved in 100 g. of glacial acetic acid, and nitrous fumes (from arsenious acid and nitric acid) passed into saturation at room tempera- ture. The mixture was allowed to stand for a week at ordinary room temperature, 152 Scientific Proceedings, Royal Dublin Soctety. more nitrous fumes having been passed in once during the week. ‘The solution was at first green, and became brown on standing. A yellow solid also separated. The yellow solid was separated from the liquid, and when crystallised from acetic acid or acetone it gave slightly yellow-coloured leaves, melting at 205°-206°C. An analysis of this compound gave the following result :— 0:1045 g. of the substance gave 15 ¢.c. of moist nitrogen at 185° C. and 762 mm., corresponding to WN 16°67. C,,H,,O;N; required N 16°55. The compound was therefore a trinitro-triphenylurea. It is a slightly coloured substance, fairly soluble in alcohol, acetone, and acetic acid; slightly soluble in ether, chloroform, and carbon tetrachloride. A further quantity of this substance was obtained from the acetic acid solution. B.—ACTION OF NITROGEN PEROXIDE ON TRIPHENYLUREA. 5 g. of triphenylurea were dissolved in 100 g. of glacial acetic acid, and nitrogen peroxide passed into saturation at room temperature. The mixture was allowed to stand for a week at room temperature, more nitrogen peroxide having been passed in during the week. The solution, which was at first yellow, became deep brown, and a yellow solid separated. The solution was filtered from the solid, which was found to be identical with the trinitro-triphenylurea described above. From the solution, by precipitation with water, a further quantity of this substance was obtained. C.—ACTION OF NITRIC ACID ON TRIPHENYVLUREA. I.—In the Presence of Solvents. (a) Glacial acetic acid.—Triphenylurea (5 gr.) was dissolved in 100 g. of glacial acetic acid, and treated in the cold with one, three, six, and nine molecular amounts of acetic acid. The mixtures were allowed to remain at room temperature for two months. (1) The solutions in the bottles with one and three molecular amounts of nitric acid developed a purple colouration, but all the triphenylurea was recovered unchanged. (2) In the bottle to which six molecular amounts of nitric acid had been added a purple colouration was also developed in the solution. This after some time changed to deep brown. Towards the end of the two months a yellow substance began to separate. The solution was filtered from the small amount of solid. The latter was found to consist of two substances, one easily soluble and the other difficultly soluble in acetic acid. The difficultly soluble substance crystallised from glacial acetic acid in leaves melting at 205°-206° C. It was identical with the trinitro- triphenylurea obtained previously, as the addition of the latter substance to it did not affect the melting point. The more easily soluble substance was soluble in alcohol, acetone, and glacial acetic acid, from which it crystallised as yellow prisms, melting at 190°-191°C. (uncorr.). This substance, as was afterwards found, was a dinitro-triphenylurea. From the original acetic acid filtrate some of the unchanged urea was obtained, together with further quantities of the dinitro-triphenylurea. Ryan & O’'Tooty—Action of Oxides, §c., on the Phenylureas. 158 (3) In the bottle to which nine molecular amounts of nitric acid were added a purple colouration was developed in the solution. Later on this changed to a deep brown, with the separation of a yellow substance. At the end of two months a fairly large quantity of this substance had separated. As in the previous case, the solid was found to consist of a mixture of two compounds, one easily soluble and the other difficultly soluble in glacial acetic acid. The former was the yellow dinitro-triphenylurea previously obtained, as the addition of this substance to it did not affect its melting point. An analysis of it gave the following result :— 0:1308 g. of the substance gave 16°8 c.c. of nitrogen (moist) at 16° C.and 766 mm., corresponding to N 15-02. Ci5HO;N, required N 14°81. The substance was therefore a dinitro-triphenylurea. The difficultly soluble substance was identical with the above-mentioned trinitro-triphenylurea. From the acetic acid solution a further quantity of these compounds was obtained, as well as some of the unchanged urea. (b) Carbon tetrachloride.—Five g. of triphenylurea were suspended in 100 g. of carbon tetrachloride and treated in the cold with one, three, six, and nine molecular parts of nitric acid. The mixtures were allowed to remain at the temperature of the room for six weeks. (1) The bottle containing one molecular part of nitric acid gave a yellow solu- tion, with a black oil floating on the surface. The latter was separated from the solution and boiled with ether. A small quantity of a yellow substance remained undissolved. ‘I'his was soluble in acetone, from which it crystallised as yellow prisms, melting at 190°-191°C. It was dinitro-triphenylurea, as a mixture of it with this substance melted at the same temperature. ‘he ethereal extract gave on evaporation a black oil, from which no pure substance was obtained. By neutralizing with barium carbonate, filtering, and evaporating the carbon tetrachloride solution yielded a black oil, from which a further small quantity of dinitro-triphenylurea was obtained by boiling with ether. (2) The bottle to which three molecular parts of nitric acid had been added contained a yellow solid and a red solution. The solution was filtered from the solid, which, after washing with ether and crystallising from acetic acid, gave yellow leaves melting at 203°-206° C. It was identical with trinitro-triphenylurea, as a mixture of the two substances melted at the same temperature as the individual substances. From the carbon tetrachloride solution, on neutralization, filtration, and evaporation, a black oil was obtained. This, when boiled with ether, left dinitro- triphenylurea. The black oil obtained by evaporating the ethereal extract yielded no crystalline substance. (3) The bottle containing six molecular parts of nitric acid gave a yellow solid and a red solution. By methods similar to those employed in (2) the yellow solid was found to consist mainly of trinitro-triphenylurea, whilst the solution was made to yield dinitro-triphenylurea and a black oil which did not crystallise. (4) In the case of the bottle to which nine parts of the acid were added, the temperature rose somewhat during the addition of the acid. It yielded a yellow solution, with a black tarry mass in suspension. By boiling the latter with glacial acetic acid a yellow crystalline substance remained undissolved. This was well washed with glacial acetic acid. When heated it began to decompose about 180° C. 154 Scientifie Proceedings, Royal Dublin Society. with evolution of brown fumes, and melted with decomposition at 235°-236° C. (uncorr.). An analysis of this substance gave the following result :— ; 0:1237 g. of the substance gave 20 c.c. of moist nitrogen at 16° C. and 763 mm., corresponding to N 18°92. C,)HnOuN, required N 19-11. The substance was therefore a pentanitro-triphenylurea. Pentanitro-triphenylurea is a light yellow substance, and as obtaimed above crystallises in prisms. It is sparingly soluble in most organic solvents, but soluble in hot nitro-benzene. No solid was obtained either from the carbon tetrachloride or from the acetic acid solution. l—ZIn the Absence of Solvents. 5 g. of triphenylurea was added slowly to twelve molecular parts of nitric acid cooled in salt and ice. ‘The interaction was at first very violent, but towards the end became more moderated. The triphenylurea dissolved, giving a red solution. The mixture was allowed to remain overnight at the temperature of the room. On pouring into much water an oily, yellow substance was obtained. ‘This was filtered, and the solid was well extracted with hot acetene and alcohol. A small quantity of a yellow substance remained undissolved. This crystallised from glacial acetic acid as leaves, melting at 200° C. (uncorr.). This substance was therefore similar to 2°4:2’-4’-tetranitro-diphenylamine in melting point and crystalline form. ‘The addition of 2°4:2’4’ tetranitro-diphenylamine to it had no effect upon its melting point. Is was therefore 2:4:2’4’-tetranitro-diphenylamine. No other pure substance was isolated from the nitration product. D.—ACTION OF NITROUS ACID ON TRIPHENYLUREA. Triphenylurea (5 g.) was dissolved in 100 g. of glacial acetic acid and treated in the cold with one, two, four, and six molecular parts of sodium nitrite, dissolved in the smallest amount of water. The mixtures were allowed to remain at room temperature for a fortnight. Yellow colourations were developed in all cases, but the triphenylurea was recovered unchanged. SUMMARY. 1. The action of nitric acid on substituted ureas at the ordinary temperature and at low concentrations was examined. Phenylurea formed phenylurea nitrate; sym- and as-diphenylureas were unacted upon; and triphenylurea yielded a di-and a trinitro-triphenylurea, Under similar conditions, but in carbon tetrachloride suspension, phenylurea formed its nitrate, its p-nitro, and its 2°4-dinitro derivatives. Sym-diphenylurea was nitrated to its 4-nitro, its 4:4’-dinitro, and its 2-4:2"4’-tetranitro derivatives, but as-diphenylurea yielded substances from which no pure compound was obtained. Triphenylurea formed a mono-, a tri-, and a pentanitro-triphenylurea, 2. Cold fuming nitric acid acted upon all four of the substituted ureas. Phenylurea gave 2°4-dinitro-phenyl-nitro-urea ; sym-diphenylurea formed its 4-4’-dinitro and its 2:4:2’-4’-tetranitro derivatives; and from as-diphenylurea as well as from triphenylurea 2°4:2’-4’-tetranitro-diphenylamine was obtained. RYAN & O’Tootr—Action of Oxides, &c., on the Phenylureas. 155 3. In cold glacial acetic acid solution nitrous acid reacted with phenylurea te form nitroso-phenylurea, and with syim-diphenylurea to form dinitroso-dipheny]l- urea. The other two substituted ureas were apparently unacted upon under the same conditions. 4. Phenylurea, suspended in carbon tetrachloride, was converted by nitrous fumes into ortho- and para-nitrophenol. Sym-diphenylurea in acetic acid solution was converted by nitrous fumes into dinitroso-diphenylurea. By the prolonged action of nitrogen peroxide upon it 4°4’-dinitro-diphenylurea was, however, obtained. As-diphenylurea was decomposed by nitrogen peroxide at the ordinary tem- perature, forming diphenylamine derivatives, e.g., 4°-4-dinitro-dipheny]l-nitrosamine and triphenylurea under similar conditions yielded a trinitro-triphenylurea. 5. In regard to ease of nitration the phenylureas resemble closely the corresponding phenylurethanes. They are much less easily nitrated than the corresponding phenyl-nitrosamines. In conclusion we wish to state that the above research was undertaken at the request of the Research Section of Nobel’s Explosives Company, to whom we are indebted for a grant in aid of the investigation. S$OIENT. PROC, R.D.S., VOL XVII, No. 18, E [ dsy 3 No. 19. THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON PHENYL-METHYLUREA. By HUGH RYAN, DSc., AND MICHAEL J. SWEENEY, M.58c., University College, Dublin. [Read Decempenr 19, 1922. Printed Frnrvary 22, 1923. ] INTRODUCTION. Ir has already been shown that urethanes are more difficultly nitratable [H. Ryan and A. Donnellan, p. 113; H. Ryan and N. Cullinane, p. 119; H. Ryan and A, Connolly, p. 125; H. Ryan and J. O'Donovan, p. 151] than secondary amines or nitrosamines. It has been shown also that the phenylureas [H. Ryan and P. O'Toole, p. 139] resemble the urethanes in respect to difficulty of nitration at low temperatures and concentrations of the interacting substances. In the present communication we describe the results of experiments we have tried with phenyl-methylurea, which form the last of the series of comparative experiments made in this laboratory on the nitration of nitrogenous aromatic substances. Nitro derivatives of phenyl-methylurea have not hitherto been obtained, nor do they appear to be formed under the conditions which we employed in the following experiments. Nitrous acid readily converts the urea into methylaniline, and such derivatives as we have isolated in our experiments can all be obtained from this base. Nitrogen peroxide in the vapour phase converts phenyl-methylurea into tetryl, while in solution it forms on short action 4-nitro-phenyl-methyl-nitrosamine, and on more prolonged action 2°4-dinitro- and 2°4°6-trinitro-methylaniline are the chief products. Nitrous anhydride acts ov a solution of the urea in the same way as it does on the amine [cf. R. Stoermer and P. Hoffmann, Ber. d. Dtsch. Chem. Ges., xxxi (1898), pp. 2523-2541], forming on short action 4-nitro-phenyl-methyl-nitrosamine. Pure nitric acid (in the presence of urea nitrate) at low temperatures and concentrations has little, if any, nitrating action on the urea, but in the presence of nitrous acid or at more or less high temperatures, especially in carbon tetra- chloride solution, it behaves towards the urea in the same way as it does towards methylaniline, converting it into phenyl-methyl-nitrosamine, 4-nitro-phenyl- methyl-nitrosamine, 2:4-dinitro-methylaniline, 2°46-trinitro-methylaniline, and finally into tetry]. SCIENT. PROC. R.D.S., VOL. XVI, NO. 19, 2F 158 Scientific Proceedings, Royal Dublin Society. The only nitrosamines we isolated were phenyl-methy]-nitrosamine and 4-nitro- phenyl-methyl-nitrosamine. The other nitrosamines known, which include 2 4- dinitro-phenyl-methyl-nitrosamine melting at 86°C. [R. Stoermer and P. Hoffmann, Ber. xxxi (1898), pp. 2523-2541] and 2-4:6-trinitro-phenyl-methyl-nitrosamine melting at 106°5° C. [E. Bamberger and J. Miiller, Ber. xxxii (1900), pp. 100- 115], were not isolated. The same remark applies to the lower nitramine derivatives, which include the 2-nitro-phenyl-methyl-nitramine melting at 67° C. and the 4-nitro-phenyl-methyl-nitramine melting at 140° C. of KE. Bamberger [Ber. d. Dtsch. Chem. Ges., xxx (1897), pp. 1248-1263]. Attempts to prepare nitro substitution derivatives of phenyl-methylurea from nitro derivatives of methylaniline and cyanic acid have not hitherto proved successful. In conclusion, the course of the reactions between phenyl-methylurea and the oxyacids of nitrogen may be summarized as follows :— Phenyl-methylurea (M.P. 82°C.) Phenyl-methylamine Phenyl-methyl-nitrosamine ne -_ Se | 2-Nitro-phenyl-methyl-nitrosamine (M.P. 36° @.) 4-Nitro-phenyl-methyl-nitrosamine (M.P. 100° C.) bi - I 2:4-Dinitro-phenyl-methylamine (M.P. 175°C.) i 2°4°6-Trinitro-phenyl-methylamine (M.P. 111°C.) 2°4°6-Trinitro-phenyl-methyl-nitramine (tetryl). (M.P. 129°C.) EXPERIMENTAL. A.—ACTION OF NITROGEN PEROXIDE ON PHENYL-METHYLUREA. I.—In the Vapour Phase. Two shallow glass dishes containing 2 g. of phenyl-methylurea and liquid nitrogen peroxide were placed side by side on a glass plate under a bell-jar and allowed to remain for a month at the room temperature. When the dark reddish-brown liquid into which the phenyl-methylurea had been converted was allowed to remain exposed to the air overnight, it became crystalline. The solid, which melted at first about 118° C., on recrystallisation from alcohol and acetone gave faint yellow crystals, which melted at 126°-127°C., and proved to be identical with tetryl. A very small amount of a dark-coloured oil was obtained by concentrating the alcohol-acetone filtrates. Ryan & Sweenvy—Action of Oxides, &c., on Phenyl-Methylurea. 159 Il.—Jn Solution. (a) Hther.—Dry nitrogen peroxide fumes were passed at intervals through a solution of 2g. of phenyl-methylurea in 75 ¢.c. of anhydrous ether, which was allowed to remain for about two months at the temperature of the room. About 1 g. of crystals separated, which melted at 111°C., and proved to be 2°4°6-trinitro- methylaniline. By evaporating the ether filtrate and recrystallising the residue 2°4-dinitro- methylaniline was obtained in the form of yellow crystals, melting at 175°. (b) Alcohol.—In another experiment 2 g. of phenyl-methylurea were dissolved in 10 ec. of alcohol, and nitrogen peroxide vapours were passed through the solution, which was kept cold by immersing the tube containing it in water. The solution effervesced briskly and turned brown in colour. It was concentrated slightly. ‘he yellow solid which separated, after recrystallisation from alcohol, melted at 99°-100° C., and, as its melting point was not affected by mixing it with 4-nitro-phenyl-nitrosamine, it was identical with the latter substance. B.—ACTION OF NITROUS ANHYDRIDE ON PHENYL-METHYLUREA. Fumes of nitrous anhydride, prepared by the action of nitric acid on arsenious oxide, were passed for about an hour through a solution of 2 g. of phenyl-methyl- urea in 10 c.c. of absolute alcohol, kept at the temperature of the room by immersing the flask containing the solution in water. ‘he mixture effervesced, the colour changing through yellow to red. On allowing the solution to evapo- rate, 4 nitro-phenyl-methyl-nitrosamine separated in the form of yellowish crystals, melting at 99°-100°C. C.—ACTION OF NITRIC AND NITROUS ACIDS ON PHENYL- METHYLUREA. I.—In the Absence of Solvents. About 8 cc. of fuming nitric acid (sp. g. 1°51) were added slowly to 2 g. of phenyl-methylurea. ‘The dark red mixture was allowed to remain overnight and was then poured into water. ‘Ihe yellowish solid which separated was filtered, washed with water and alcohol, and then recrystallised from acetone, from which it separated in the form of nearly colourless crystals, melting at 128°- 129°C. The yield was very good. As its melting point was not affected by mixing the substance with pure, tetryl, prepared from dimethylaniline by the method described by M. C. F. van Duin [Rec. de trav. Chim. Pays Bas, xxxvii (1917), p. 111], the two bodies are therefore identical. Il.—Jn the Presence of Solvents. (a) Water: (a) At ordinary concentrations.—1. Phenyl-methylurea (13°6 g.) was dissolved in a mixture of 12 c.c. of concentrated hydrochloric acid and 24 cc. of water, and to the cold solution 7:5 g. of sodium nitrite, dissolved in a small amount of water, were added slowly and with frequent stirring. ‘he mixture rapidly became turbid, with separation of a yellowish oil, which was extracted with ether; the residue left on evaporation of the ether was reduced by means of tin and hydrochloric acid to methylaniline, which boiled at 192°-194°C. ‘The aqueous solution, from which the nitrosamine had been separated by ether, contained a small amount of the hydrochloride of methylaniline. 160 Scientific Proceedings, Royal Dublin Society. 2. An aqueous solution, containing 8-4 ¢, of nitric acid, was added slowly to a cold solution of 5 g. of phenyl-methylurea and 5 g. of sodium nitrite in 25 cc. of water. A colourless oil first separated; but when the mixture was warmed for ‘some hours on the water-bath the oily nitrosamine was gradually converted into a solid, which, when recrystallised from alcohol, melted at 99°— 100°C J., and proved to be 4-nitro- phenyl-methyl-nitrosamine. In a similar experiment, in which, however, sodium nitrite was not added to the reaction mixture, the only change observed consisted of the formation of a small amount of methylaniline. () At low concentratcons—Four solutions, each containing 1 g. of phenyl- methylurea in 50 cc. of water, to which one, two, three, and four molecular amounts of nitric acid had been added respectively, were allowed to remain in, stoppered bottles at the temperature of the room for about six weeks. The solutions, which had remained nearly colourless, were neutralized with sodium carbonate, saturated with sodium sulphate, and extracted with benzene. Almost all the phenyl-methylurea was recovered unchanged in each case. (b) Alcohol.—As phenyl-methylurea was recovered unchanged from an alcoholic solution of it to which nitric acid containing urea nitrate had been added, and which had been allowed to remain at the temperature of the room for a couple of weeks, the experiment was varied by including nitrous acid in the interacting substances. For this purpose aniline was converted into methylaniline by the method of G. T. Morgan (18.P. 102,854), and the methylaniline, without isolating it, was converted directly into its nitrosamine, which was separated from the simultaneously formed diazonium salt by means of ether; the phenyl-methyl-nitrosamine left on evaporation of the ether was further purified by distillation in a current of steam, and the product thus got was employed in the reactions described below :— A solution of 5 g. of phenyl-methyl-nitrosamine in 100 c.c. of aleohol to which 0:75 ¢.c. of nitric acid (sp. g. 1°51) had been added was heated on the water-bath for five hours. The solid which separated on cooling melted at 99°-100° C., and consisted of 4-nitro-phenyl-methyl-nitrosamine. (c) Acetic acid. (a) At ordinary concentrations.—1. ‘oa mixture of 10 g. of phenyl-methylurea and 20 cc. of glacial acetic acid 5°5 ¢.c. (two molecular amounts) of nitric acid (sp. g. 1°51) were added slowly. A solution of 5 g. of sodium nitrite in about 10 ¢.c. of water was then allowed to drop into the mixture, the temperature of which was kept below 30° C. during the addition of the nitrite. A crystalline solid separated from the solution, the latter becoming almost black in colour. After remaining overnight the solution was filtered and the solid was dissolved in hot alcohol, from which it separated in the form of almost colourless acicular prisms melting at 100°C., and consisting of pure 4-nitro-phenyl-methyl-nitro- samine. On addition of water to the acetic acid filtrate a small amount of 2'4-dinitro- methylaniline was obtained. The nitro-nitrosamine obtained in this experiment was heated with con- centrated hydrochloric acid and alcohol under a reflux condenser until the evolution of nitrous fumes had ceased. The reddish yellow solution was cooled, diluted with water, and neutralized with sodium carbonate. The yellow solid which separated was filtered, washed with water, and recrystallised from alcohol. The yellow crystals thus got melted at 151° C., and consisted of 4-nitro-methylaniline. 2, In another experiment 20 ¢.c. of fuming nitric acid (sp. g. 151) were added to a solution of 5 g. of phenyl-methylurea in 50 cc. of glacial acetic acid, Ryan & Sweensy—Action of Oxides, &c., on Phenyl-Methylurea. 161 and the mixture was heated for an hour on the water-bath. Nitrous fumes were evolved, and a dark, somewhat oily solid separated when the solution was poured into water. When this solid was crystallised from alcohol it melted, but not quite sharply, about 111° C., and consisted of 2:4-6-trinitro-methylaniline. 3. Phenyl-methyl-nitrosamine (45 g.) was dissolved. in warm acetic acid (10 c.c.), and to the solution 2°3 c.c. of fuming nitric acid (sp, g. 1:51) were added slowly. The mixture was heated to boiling for a few minutes, then diluted with water and cooled. The solid which separated was filtered, washed with water, and recrystallised from alcohol. About 3 g. of pure 2°4-dinitro-methylaniline melting at 175°C. were thus got; the parent liquid contained a small amount of an oily solid, probably consisting of the same substance. (RB) At low concentrations.—To four solutions of 5 g. of phenyl-methylurea in 100 g. of glacial acetic acid quantities of nitric acid corresponding to one, two, three, and four molecular amounts were added, and the solutions allowed to remain at the temperature of the room for about three months. In no case was there any indication of the occurrence of any appreciable amount of reaction between the constituents. No solid separated from the mixtures, and when they were poured into water the solutions remained in all cases clear. (d) Carbon tetrachloride.—At low concentrations.—As phenyl-methylurea is only sparingly soluble in cold carbon tetrachloride, it was found inconvenient in this case to carry out the reactions at the temperature of the room. Four solutions were prepared, containing in each case 5 g. of phenyl-methyl- urea in 100g. of carbon tetrachloride at 60°C. Quantities of fuming nitric acid (sp. g. 1°51) corresponding to one, two, three, and four molecular amounts were added to these respectively. In each case a vigorous reaction set in, with evolution of nitrous fumes and separation of a dark brown oily layer. The mixtures were allowed to remain at the temperature of the room for three months. They were then shaken with water, the carbon tetrachloride was in each case separated, and the aqueous washings returned to the solid or tarry matter in the flask. 1. By distilling the deep yellow carbon tetrachloride solution from the flask to which one molecular amount of nitric acid had been added about 0°5 g. of tetryl was obtained. The aqueous portion was made alkaline and extracted with ether. Methyl- aniline was extracted from the ether solution by means of hydrochloric acid, and after reconversion into the free base and extraction with ether boiled at 192°-194° C. 2. From the carbon tetrachloride solution to which two molecular amounts of nitric acid had been added we isolated, on the other hand, a small amount of 4- nitro-phenyl-methyl-nitrosamine ; while trom the mixture of the aqueous and undissolved portions in the flask we obtained a further quantity of 4-nitro-phenyl- methyl-nitrosamine and 2°4-dinitro-methylaniline. 3. In the case of the solution to which three molecular proportions of nitric acid had been added distillation of the neutralized carbon tetrachloride solution gave a small amount of a crystalline residue, which, when washed with ether, melted at 175° C., and consisted of 2°4-dinitro-methylaniline. The mixture of solid and-water in the flask was shaken with ether. The lower aqueous layer was filtered from a yellow solid (0°5 g.), which consisted of 2°4-dinitro- methylaniline (M. P. 175° C.), and the ether solution, after washing with dilute sodium carbonate, gave a further quantity (0°75 g.) of 24:6-trinitro-methylaniline. SCIENT. PROC. R.D.S., VOL. XVII, NO. 19. 26 162 Scientific Proceedings, Royal Dublin Society. 4, By distilling the carbon tetrachloride solution from the mixture to which four molecular amounts of nitric acid had been added, a small amount of nearly colourless crystals of tetryl was obtained, and by fractional crystallisation from alcohol of the crystalline solid suspended in the aqueous portion we separated some pure 2°4-6-trinitro-methylaniline from the impurity (probably tetryl) with which it was mixed. SUMMARY. 1. Under the conditions of our experiments phenyl-methylurea formed no nitro derivatives. When nitrous acid was present the phenyl-methylurea was converted into methylaniline, and hence the nitro substances we isolated were in all cases derived from the latter base. 2. Nitrogen peroxide in the vapour, phase converted phenyl-methylurea into tetryl, but in solution it formed successively 4-nitro-phenyl-methyl-nitrosamine, 2-4-dinitro- and 2°4°6-trinitro-methylaniline. 3. In the presence of urea nitrate, nitric acid had little, if any, action on phenyl- methylurea, but in the presence of nitrous acid the phenyl-methylurea was converted into phenyl-methy]l-nitrosamine, 4-nitro-phenyl-methyl-nitrosamine, 2'4- dinitro-methylaniline, 2-4°6-trinitro-methylaniline, and tetryl. Incidentally it was found that tetryl could be obtained in a good yield and in a pure condition by the nitration of phenyl-methylurea or phenyl-methyl- nitrosamine. 4. So far as ease of nitration is concerned, phenyl-methylurea bears to sym- diphenylurea a relation somewhat similar to that of ethyl-o-tolylurethane to diphenylurethane. In conclusion we wish to state that the above research was undertaken at the request of the Research Section of Nobel’s Explosives Company, to whom we are indebted for a grant in aid of the investigation. 10. SCIENTIFIC PROCEEDINGS. VOLUME XVII. . Experiments on the Electrification produced by Breaking up Water, with Special Application to Simpson’s Theory of the Electricity of Thunder- storms. By Professor J. J. Nouan, m.a., p.sc., and J. HnricHt, B.a., M.SC., University College, Dublin. (June, 1922.) . Cataphoresis of Air-Bubbles in Various Liquids. By Tuomas A. McLaueutim, M.sc., a.Inst.p., University College, Galway. (June, 1922.) . On the Aeration of Quiescent Columns of Distilled Water and of Solutions of Sedium Chloride. By Professor W. EH. Aprnny, A.R.C.SC.1., D.SC., F.1.C. ; Dr. A. G. G. Lronarp, F.R.c.sc.1., B.Ssc., F1.c.; and A. Rrcarpson, A.R.C.SC.1., A.C. (June, 1922.) . On a Phytophthora Parasitic on Apples which has both Amphigynous and Paragynous Antheridia; and on Allied Species which show the same Phenomenon. By H. A. Larrerry and Gurorcs H. Peraypripes, Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (June, 1922.) . Some Further Notes on the Distribution of Activity in Radium Therapy. By H. H. Poors, m.a., sc.p., Chief Scientific Officer, Royal Dublin Society, (June, 1922.) . Preliminary Experiments on a Chemical Method of Separating the Isotopes of Lead. By Tomas Ditton, p.sc.; Rosatinn Cuarks, D.sc. ; and Vicror M. Hincuy, s.sc. (Chemical Department, University College, Galway). (July, 1922.) . The Lignite of Washing Bay, Co. ‘l'yrone. By 'T. Jounson, D.sc., F.L.S., Professor of Botany, Royal College of Science for Ireland; and Janz G. Gitmore, B.sc. (Plate III.) (August, 1922.) . Libocedrus and its Cone in the Irish Tertiary. By T. Jounson, p.sc., F.L.s., Professor of Botany, Royal College of Science for Ireland; and Janz G. GitmorE, B.sc. (Plate 1V.) (August, 1922.) . The Electrical Design of A.C. High Tension Transmission Lines. By H. H. Jerrcort. (August, 1922.) The Occurrence of Helium in the Boiling Well at St. Edmundsbury, Lucan. By A. G. G. Lronarp, F.8.¢.sc.1., PH.D., F.1.c., and A. M. Ricaarpson, A.R.O.SC.L, A.I.c. (Plate V.) (August, 1922.) [ Nos. 1 to 10, price 9s. | . On the Detonating Action of « Particles. By H. H. Poozs, m.a., sc.p., . Chief Scientific Officer, Royal Dublin Society. (December, 1922.) . The Variations of Milk Yield with the Cow’s Age and the Length of the Lactation Period. By James Witson, u.a., B.sc. (December, 1922.) . A Note on Growth and the Transport of Organic Substances in Bitter Cassava (Manthot utilissima). By T. G. Mason, u.a., B.sc. (December, 1922.) [Nos. 11 to 18, price 1s. 6d.] 14. 15. 16. 17. 18. 19. SC IENTIFIC PROCEEDINGS—continued. The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylure thane. By Hue Dublin. Ryan, p.sc., and Anne Donnetuan, m.sc., University College, (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on Kthyl-o-Tolyl- urethane. By Hueu Ryan, v.sc., and Nicwonas CULLINANE, PH.D., University College, Dublin. (February, 1923.) The Acticn of the Oxides and the Oxyacids of Nitrogen on Mthyl-Phenyl- urethane. By Huvew Ryan, v.sc., and Anna Connouty, m.sc., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on Phenyl-Benzyl- urethane. By Hueu Ryan, p.sc., and James L. O’Donovan, m.se., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on the Phenylureas. By Hueu Ryan, v.sc, and Perer K. O’Toouz, u.sc., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on Phenyl-Methyl- urea. By Hueu Ryan, v.sc., and Micuarn J. Sweeney, u.sc., University College, Dublin. (February, 1923.) DUBLIN = [Nos. 14 to 19, price 4s. | —_— PRINTED AT THE UNIVERSITY PK ESS RY PONSONBY AND GIBBS. STi GSS SCIENTIFIC PROCEEDINGS gn << ROYAL DUBLIN SOCIEIY A * nebiig j - 4“ 4nd Vol. XVII, N.S., Nos. 20-24. 20.—ON THE CAUSE OF ROLLING IN POTATO FOLIAGE; AND ON SOME FURTHER INSECT CARRIERS OF THE LEAF- ROLL DISEASE. By Paun A. Murpeny, Se.D., A.R.C.Sc.1., Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (Plate V1.) 21—ON THE CHANNELS OF TRANSPORT FROM THE STORAGE ORGANS OF THE SEEDLINGS OF ZODOICHA, PHCNLX, AND VICL4. By Henry H. Dixon, Sc.D., F.RS., Professor of Botany in the University of Dublin; and Nicen G. Banu, M.A., Assistant to the Professor of Botany in the University of Dublin. (Plates VII-X1I.) 22 TRREGULARITIES IN THE RATE OF SOLUTION OF oXYGEN BY WATER. By H. C. Becxer, A.R.C.Sc.L, A.I.C., Demonstrator in Chemistry in the College of Science, Dublin; and E. F. Parson, A.R.C.Sc.1., Research Student. 23._THE HYDROGEN ION CONCENTRATION OF THE SOIL IN RELATION TO THE FLOWER COLOUR OF HYDRANGEA HORTENSIS W., AND THE AVAILABILITY OF IRON. By W. kR. G. Avnins, O.B-E., Sc.D., F.1.C. 24 —_THE COMPARATIVE VALUES OF PROTEIN, FAT, AND CARBO- HYDRATE FOR THE PRODUCTION OF MILK FAT. By E. J. SHEEHY, F.R.C.Sc.1., B.S¢. (Hons.), M.R.I.A., Biochemical Laboratory, D.A.T.L. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN : PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSLER HOUSE, DUBLIN. WILLIAMS AND NORGATHE, 14, HENRIETTA SIREET, COVENT GARDEN, LONDON, W.C. 2. 1925. Price Seven Shillings and Sixpence. Royal MBublin Society. See FOUNDED, A.D. 1731. INCORPORATED 1749. ESOS SCIENTIFIC MEETINGS. Tne 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 Wxecutive 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 ta 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 property 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 Hxecutive Officer for Science for farther information concerning the preparation of papers and illustrations for press. peiesey No. 20. ON THE CAUSE OF ROLLING IN POTATO FOLIAGE; AND ON SOME FURTHER INSECT CARRIERS OF THE LEAF-ROLL DISEASE. BSA PAL MWC ERYE ScD: A RUC Soule Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (Pate V1.) [Read Frpruary 27. Printed May 15, 1923.] WITHIN comparatively recent years the designation ‘‘leaf-roll’’ has, by general agreement among phytopathologists, become restricted to a single specific disease of the potato plant. It is not necessary to discuss here the taneled - history of the varying connotations of the term as used by different investigators since 1905. At the present time this name is the one applied by almost all plant pathologists exclusively to that very widely distributed disease of the potato plant which has as its most constant external symptom a distinctive thickening and upward marginal rolling of the leaflets of the lower leaves in the first instance, and which is accompanied by a more or less pronounced diminution in yield of new tubers—features which are persistent and not merely of seasonal or shorter duration.’ As thus defined, the disease corresponds to the secondary leaf-roll of Quanjer.” No general account of the disease need be given here except in so far as the results obtained by previous investigators have a direct bearing on the work now to be described. This deals mainly with some features of the physiology and histology of affected plants and with the transmission of the disease by certain classes of insects. The work was carried out principally at the Albert Agricultural College, Glasnevin, Dublin, in a laboratory and special plots pro- vided for the purpose. Acknowledgments are due to Mr. R. McKay, A.R.C.Sc.1., who acted as Outdoor Assistant during the season of 1922, for the care of the plots and for carrying out some of the experimental details. I.—Previous work on the Translocation of Food Materials in Diseased Plants. It had already been surmised by a number of workers, including Sorauer, Spieckermann, Kock and Kornouth, Doby and Quanjer, that some disturbance in the mechanism of food transference in the plant underlay the disease described by them as ‘‘leaf-roll.’’ Unfortunately it is difficult in many cases to establish with certainty the identity of the malady under discussion. The majority of 1Tt is desirable that the more indeterminate name ‘‘leaf-curl,’’? which is still unfortu- nately retained in the official publications of the English Ministry of Agriculture, should not be used for this disease. 2 Comparatively little attention was paid during the course of this work to primary leaf- roll, but it may be mentioned in anticipation that starch was found to accumulate to an abnormal extent in the upper leaves, which are the ones that first become rolled in this stage of the disease. ! SCIENT. PROC. R.D.S., VOL. XVII, No. 20. 2H 164 Scientific Proceedings, Royul Dublin Society. these investigators had in view the difficulty in upward transfer of food from the planted tubers to the developing shoot; and some evidence in support of the existence of this difficulty, based on chemical analysis, was brought forward by Spieckermann (14) and Doby (2). The same workers also found that at a later stage of growth there was an excess of food in the upper portion of the plant; whilst Quanjer (10) clearly expressed the opinion that the abnormal appearance shown by the foliage must be looked upon as being due in some way to a similar disturbance which interfered with the downward transfer of food. According to Quanjer (11), Jordi (in 1913) was the first to record the presence of abnormal quantities of starch in the stems and petioles of plants affected with leaf-roll. It was not, however, until the publication of Neger’s (6) first paper in 1918 that it was realised that the rolled leaves of diseased plants retained their starch to a very large extent instead of getting rid of it. In the following year Hsmarch (3), Hiltner (4), and Quanjer (9) published the results of investigations in which each arrived independently at the same general conclusion. According to Esmarch (le, p. 17), the leaves of healthy potato plants when darkened become free from starch within a period of from 19 to 68 hours, depending on the temperature, weather conditions, age of the leaves, and individuality of the plants. The leaves of diseased plants, on the other hand, become depleted of their starch under the same conditions either not at all or only incompletely. In Esmarch’s experiments the older leaves of diseased plants, after being darkened for from six to eight days, or in many experiments even for twelve days, still retained practically all their starch, in many cases the veins alone being more or less starch-free. In the younger leaves of diseased plants, which were still free from rolling, starch sometimes began to disappear at about the same time as in leaves of the same age from healthy plants; but cenerally it did so only after a period of from three and a half to eight days. Complete disappearance of the starch from young rolled leaves was seen only rarely. The older the leaf and the more pronounced the rolling the greater was the interference with starch translocation. The experiments carried out by Neger gave very similar results. This worker employed cut shoots which, after removal from the plants, were placed in the dark with their ends standing in water. While this method is a very convenient one for demonstrating the difference between the amount of starch in the leaves of diseased and healthy-plants (and was largely used for that purpose in the investigations now to be described), it is entirely unsuitable for experiments on the translocation or movement of carbohydrates in the plants. Consequently Neger’s conclusions on carbohydrate translocation, based, as they were, on this method, are open to serious question. Il.—Starch Accumulation in the Leaves Invariably Associated with Leaf-Roll. In taking up the further study of this subject it seemed necessary in the first place to determine whether excess of starch in the lower leaves of diseased plants was a universal concomitant of leaf-roll under Irish conditions. That this was so was established as a strong probability as a result of experiments carried out here with twenty-one varieties of potato in 1921. Two plants of each variety were selected, one diseased and the other healthy, and from a lower leaf of each, situated on the westerly side of the plant, corresponding leaflets were removed in the morning (10 a.m. to 11 a.m.), and were subjected to the well-known iodine test. In the case of all varieties but two there was a very marked difference in the amount of starch present, the affected leaflets containing an abundance of starch, while the healthy ones showed little or none, Murpuy— On the Cause of Rolling in Potato Foliage. 165 The trials were continued in 1922 with a slight variation in method. The leaflets from each pair of plants of the same variety (diseased and healthy) were cut from corresponding lower leaves in the evening (5 p.m.), and were kept in the dark with their stalks in water for a period of from sixteen to forty hours. The latter period was generally used, and was found to give a better differentia- tion. In these experiments, as in the previous year, leaflets of twenty-three varieties were submitted to the iodine test during the months of June and J uly, and in all cases without exception a pronounced difference between the diseased and healthy leaflets was observed. In some cases whole leaves were used instead of leaflets, with the same results. The healthy leaflets were generally free from starch, but sometimes they showed a slight brownish mottling, indicating its presence in small quantities; the diseased leaflets, on the other hand, presented an intense blue-black colour, which, as a rule, was practically uniform all over. The bases of the mid-ribs and adjacent lateral veins sometimes contained no starch, and more rarely practically clear patches alternated with deep blue-black ones on other portions of the leaflets. Most of the better-known varieties of potato were used in these experiments, including those of long standing and others of more recent introduction. Some of these were early, some mid-season, and some late sorts. Since the varieties from which positive results were secured in 1922 included the only two exceptional ones of 1921 (when their condition in respect of leaf- roll infection was felt to be open to some doubt on account of the exceptional drought then prevailing), it seems safe to conclude that an abnormal aceumula- tion of starch in the lower leaves always accompanies leaf-roll; and, incidentally, this circumstance provides a convenient symptom for the accurate diagnosis of the disease. Further, the iodine test has been tried and found to give satis- factory results on diseased and healthy lower leaves of the same variety purposely sent in by post in the ordinary way from the West of Ireland for examination. In such eases the difference in the colour of the two leaves, after boiling and treating with iodine, can be observed clearly by transmitted light without preliminary extraction of the chlorophyll. It was found previously that this could be done on material gathered direct and not sent through the post; and the omission of the bleaching part of the test saves both time and aleohol. Ill.—E ffects of the Accumulation of Starch on the Conformation and } Structure of the Leaves. A number of experiments were carried out to determine, if possible, the relationship in respect of cause and effect between starch accumulation and rolling of the leaves. A beginning was made by the examination during May and early June, 1922 (before any visible symptoms of leaf-roll appeared), of the starch content of the lower leaves of plants derived from diseased tubers.* All the plants examined, as was to be expected, afterwards developed normal symptoms of leaf-roll. With them were compared corresponding leaves from healthy plants, and, in most cases, similar leaves which had already begun to 1TIn the apparent absence of recognisable symptoms of the leaf-roll disease in the tubers, or at least of such as are of constant occurrence, the presence of infection in any given tuber can be assumed only when the plant of which it is the progeny was definitely known to be diseased in the previous season. Throughout this paper the expression ‘‘ diseased tubers’’ is used in this sense. Similarly ‘‘healthy tubers’’ are such as were known to be produced in the preceding season by plants which were apparently free from leaf-roll, and which were as little as possible exposed to infection, 2H2 166 Scientific Proceedings, Royul Dublin Society. roll from other diseased plants. It should be explained that the time of onset of rolling of the leaves varies considerably, even in adjacent similar plants. The comparative examinations were repeated six times from May 25 to June 8, the variety used throughout being British Queen. As a general rule, corresponding leaflets from three plants of each category were used in each examination. They were removed at 5 p.m., and kept in the dark with their cut ends in water for from sixteen to forty hours, and were then submitted to the iodine test in the usual way. The results of all the examinations may now shortly be summarised. In the first trials, on May 25, 26, and 27, there was no noticeable difference in the amount of starch present in such leaflets as came from healthy or diseased (but normal looking) plants after being kept in the dark for sixteen hours... Leaflets which had already begun to roll at this time, however, showed excess of starch ; and the more pronounced the rolling the greater was the amount of starch present. On May 30 there was clearly more starch in leaflets which had not begun to roll, but which were taken from diseased plants, than in similar healthy leaflets, both being kept sixteen hours in the dark. When, however, the leaflets were kept in the dark for forty hours, there was no difference between the two lots. This shows that although starch accumulation in the diseased leaves had begun, it had not yet reached very large proportions. Confirmatory results were secured in the case of leaves from diseased plants which were tested on June 2. The last examinations, carried out on June 7, when the majority of the plants in the plot already exhibited symptoms of leaf-roll, showed a still more striking accumulation of starch in diseased lower leaves which had not yet begun to roll as contrasted with similar healthy leaves. In this ease the difference was still easily perceptible after the leaves had been kept forty hours in the dark. This indicates that the amount of starch present in the diseased leaves had increased. Corroborative results were secured by systematic microscopical examination during the same period of bleached material taken from the plant in the morning (10 a.m.), the diseased leaves being taken from the same plant throughout. Up to June 2 there was no perceptible difference between the diseased but non-rolled plants and healthy plants. The leaves of both contained about the same amount of starch, the lower ones being generally empty except for the guard-cells and ‘‘starch-sheath,’’ and the upper ones containing a good deal of starch throughout the mesophyll. The first sign of increased starch- content in the diseased but non-rolled lower leaves was noticed on June 3. This feature had become very pronounced by June 6, there being by this time very much more starch present in the morning in the lower leaves (which had not yet rolled perceptibly) than in the upper leaves of the same plant. A general examination of diseased plants showing the earliest symptom of rolling was made at this period in comparison with healthy plants, the material being collected at 10 a.m. and examined at once. The diseased lower leaf- blades contained an abnormal amount of starch, while their petioles were not distinguishable from those of healthy plants, both containing starch only in the ‘“starch-sheath.’’ This was also the case in the stems, but here there appeared to be appreciably less starch present in the ‘‘starch-sheath”’ of the diseased plant than in that of the healthy one. No necrosis of the phloem was seen in any part of the plants, although a careful search was made. Although the methods used in the experiments described are not altogether free from objection, for the reason that the rolling of diseased leaves sets in gradually, thus rendering it impossible to say exactly when it begins, and because rolling follows rapidly after the beginning of starch accumulation, Murpuy—On the Cause of Rolling in Potato Foliage. 167 nevertheless it is sufficiently clear that the lower leaves of diseased plants, which empty themselves of their starch normally at first, gradually accumulate it in the mesophyll at a later stage; and very soon afterwards begin to roll upwards. This conclusion is confirmed by the results of two further experiments, which will now be described. Artificial retardation of starch accumulation and leaf-rolling—At 10 a.m. on June 1, 1922, before any signs of rolling in their leaves were visible, two plants of the variety British Queen growing in the field from ‘‘sets’’ derived from diseased tubers, were darkened by covering them with comparatively light- tight boxes. At that time the lower leaves were found to be practically free from starch; and they were entirely free from it after a further period of twenty-four hours, with the exception of the merest trace in the ‘‘stareh-sheath”’ and guard-cells. Carbohydrate translocation may, therefore, be assumed to have been normal at that time. The plants remained covered except for oceasional short periods, when the boxes were removed for the purpose of allowing a small amount of photosynthesis to take place. These periods generally extended from 5 p.m. to 9 a.m., but sometimes the plants were uncovered for shorter intervals in the daytime. Under those abnormal conditions the plants assumed, of course, a somewhat sickly and chlorotic appearance; but the leaves remained free from roll, and starch accumulation did not occur until a period of from thirty to thirty-six days had elapsed after all the similar, neighbouring, non-covered, diseased plants had shown both rolling of their leaves and starch accumulation. One of the covered plants was first observed to show decided rolling of its leaves on July 11, this occurring after an unduly prolonged exposure to light. Some of the leaves on the other covered plant began to roll very slightly on July 12, also following a period of exposure. The two plants were again covered on the dates mentioned, and they were not again exposed to light until seventeen days had elapsed in one case and twenty-one days in the other. In both cases periodical tests were made, and they showed that a large proportion of starch was retained by those more or less rolled leaves which remained green. The leaves which became yellow eventually showed no starch. While the badly rolled leaves of the first plant, after it had remained covered for seventeen days, turned entirely black with iodine, the bases of the veins of the slightly rolled leaves of the other covered plant began to clear at that time, and after twenty-one days the basal halves of the leaflets of this plant were quite clear, while their upper halves were full of starch (fig. 9, Pl. VI). When it was finally uncovered, the second plant developed still more pronounced symptoms of leaf-roll within twenty-four hours. Rolling of leaves and starch accumulation artificially prodwced.—In the second experiment each of four healthy tubers of the variety Up-to-Date was cut in two, and the resulting pairs of four ‘‘sets’’ were planted in two parallel rows as far removed as possible from all likely sources of leaf-roll infection. The four ‘‘sets’’ in one row were planted in mounds of soil about one foot above the general ground-level. Farmyard manure was previously placed at the bottom of each mound, and stakes were provided to which the resulting plants were afterwards tied. The corresponding four ‘‘sets’’? were planted in the other row in the usual manner, i.e. in a drill in which the manure had been placed. All the plants grew well, and appeared normal. On July 7 the soil was care- fully removed from the bases of the four plants growing on the mounds, and the stolons (which were just developing), together with most of the axillary buds and shoots, were removed from them.’ The immediate effect of this treatment was 1+This operation was repeated once during the course of the experiment. 168 Scientific Proceedings, Royal Dublin Society. the occurrence of a peculiar upward rolling of the blades of the leaflets of the upper leaves, and within three days this feature was very marked on the four treated plants. The four untreated controls remained quite normal (figs. 1 and 2, Pl. VI). The rolled leaves of the treated plants still retained their dark green colour; and tests of their starch-eontent made on the third day after treatment showed an accumulation of starch after they were kept in the dark for forty hours with their stems in water. This excess became more pronounced as time went on. On the seventh day, for example, the ‘‘artificially’’ rolled top leaves, when removed from the plants in the evening, contained obviously more starch than corresponding normal leaves taken from the untreated control plants. The starch disappeared from the normal leaflets when kept in water in the dark within sixty-four hours (perhaps in a shorter time), but within this period the similarly treated rolled leaflets had only cleared slightly at their tips (fig. 6, Pl. VI). In seven days not more than one-quarter of the area of the ‘‘artificially”’ rolled leaflets was free from starch, the clearing proceeding from the tip down- wards (fig. 7, Pl. V1). An actually diseased and rolled lower leaf from an affected plant, kept at the same time under the same conditions, began to clear, however, at its base. Similar results were obtained on the seventeenth day of the experiment, when whole leaves were used in place of leaflets. The presence of an abnormal quantity of starch in the ‘‘artificially’’ rolled leaves was confirmed by microscopical examination. The excess of starch was not confined to the lamina of the leaf, but was also very noticeable in the parenchyma, and particularly in the ‘‘starch-sheath,’’ of the petiole, in which the starch grains were unusually large. Up to the seventeenth day after excision the treated plants remained in the condition described, but the rolling had gradually become more pronounced (see photograph made on the seventeenth day, reproduced in fig. 3, Pl. VI), and had extended downwards on the plant so as to involve practically all the leaves but the lowest. At that time these plants presented a marked contrast with their controls. The symptoms of rolling obviously differed from those which accompany primary leaf-roll, the most pronounced distinction being the fact that the ‘‘artificially’’ rolled leaves retained a very dark green colour, and showed a tendency on the part of the leaf as a whole to curve downwards. On the seventeenth day, however, one of the treated plants showed suspicious symptoms of true primary leaf-roll,, and it was immediately cut off at ground- level in the hope of preventing the infection of the remaining three healthy plants, from which it was desired to secure healthy tubers or cuttings. The soil, therefore, which had previously been removed, was heaped up about the bases of the stems of these plants, and they were allowed to grow in the normal way. The hope of keeping them in a healthy condition was, however, only partially realized. Primary leaf-roll continued to appear in plant after plant, both treated and control, and each plant was removed as it became infected. On the twenty-fourth day of the experiment only three plants were left—one control, which was normal, and two of the treated plants, which were now forming tubers with extreme rapidity, and from which the ‘‘artificial’’ rolling of the leaves had almost disappeared. Two days later the leaves of the then only remaining treated plant had entirely lost their rolled appearance (see figs. 4 1Tt should be stated that during the period of the experiment primary leaf-roll had become common in neighbouring healthy plants in the same plot. All these plants were much infested with capsid bugs (Calocoris bipunctatus), which came from an adjacent weedy hedge, and infection was attributed to them. Murpny—On the Cause of Rolling in Potato Foliage. 169 and 5, Pl. VI),’ and when removed from the plant on this day and subsequently tested at intervals they were found to lose their starch fully as quickly as corresponding leaves from an untreated healthy control plant of the same variety (see fig. 8, Pl. VI), as the following details will show :— Amount of starch present alter being in dark room for— 22 hours. 25 hours. 29 hours. | 45 hours. 50 hours. | ; Top leaflets of previously Some pre- | Little pre- | Practically None. None. “artificially ” rolled Up- senti 3 sent. gone. to-Date plant. Top leaflets of healthy un- | A little more | A little more | A little more Trace. None. treated Up-to-Date plant. | than above. | than above. | than above. It appears that Véchting (15) long ago carried out a somewhat similar experiment, but not, of course, with reference to leaf-roll, and no mention seems to have been made of any resulting change in the form of the leaves, although it is apparent from some of the figures that a certain amount of rolling set in. That it was not more pronounced is probably due to the formation of aerial tubers, which provided an outlet for the large amount of starch that accumulated. Quanjer et al. (11) also refer to the presence of excess of starch in the stems and leaves of plants from which the tubers have been removed, and in those the stolons of which are attacked by Corticium vagwm; but no reference is made to any rolling of the leaves following these occurrences. From the results of the experiments described in this and in the preceding sections of this paper it is concluded that one of the earliest discernible effects of the leaf-roll disease is an abnormal accumulation of starch, principally in the lower leaves of affected plants. It has been shown that the starch disappears at the normal rate from these leaves at first. At a somewhat later period there is a gradual retardation of this process, and, following rapidly on this, the leaflets begin to roll upwards at their margins. Seeing that in the case of affected plants rolling appears to follow starch accumulation, and that the rolling of the leaves of diseased plants can be prevented for comparatively long periods by the exclusion of light, and by thus greatly reducing the amount of carbon assimilation; and further, in view of the fact that rolling and starch accumula- tion can be induced in healthy plants when they are deprived of their storage organs and most of their growing points—there is little room for doubt but that the rolling of the leaves of diseased plants is a direct consequence of the acecumu- lation in them of an excess of starch.” Distention of the spongy parenchyma.—tThis accumulation of carbohydrate in the tissues of affected leaves leads to an abnormal distention of the spongy as compared with the palisade parenchyma. Owing to the fact that the spongy parenchyma is comparatively free to expand in three directions, but is restricted 1 An unusual feature of some of the treated plants may be noted in fig. 4—namely, the prolongation of the ultimate branches of the floral shoot into shoots which bore small foliage leaves and developed new flower buds at their tips. The floral axes on these plants were un- usually stout and well developed, and in one case fruit was set, which is a very rare occur- rence in the case of Up-to-Date. These phenomena are no doubt related to the presence of an abnormal amount of food materials in the upper portion of the treated plants. 2Tt is quite probable, as appears from Neger’s work, that an abnormal amount of sugar is also present in rolled diseased leaves. This point was not investigated. 170 Setentific Proceedings, Royal Dublin Society. on its upper side by its attachment to the palisade cells, a downward and lateral extension follows, which has, as one of its consequences, the upward rolling of the margins of the leaflets. As is well known, rolled leaves give the impression of being thicker than normal ones; and on measurement this is generally found to be the ease. It is not invariably so, because it occasionally happens that the growth in length of the palisade cells of affected leaves is interfered with, and the consequent reduction may be of such a magnitude as more than to counterbalance the extension of the spongy parenchyma. It does seem to be a general rule, however, that the spongy parenchyma of diseased leaves is thicker than normal, as is proved by comparing the thickness of this tissue (including the lower epidermis, which does not vary) in similar leaves, diseased and healthy, of the same variety; or by determining what percentage of the whole thickness of the leaf is made up of spongy parenchyma in the two eases. This percentage was found to vary in mature healthy leaves between 47 and 54 per cent., while in corresponding diseased leaves it varied between 57 and 74 per cent. It is believed that this demonstrated expansion of the cells of the spongy parenchyma in a downward direction must be accompanied also by a certain amount of lateral extension, and this seems to be confirmed by the apparently more rotund shape of the cells and by a reduction in the sizes of the intercellular air-spaces between them. In view of the obviously more compact and less elastic nature of the palisade layer, to which the lower and less rigid tissue is firmly attached, the result must be an upward rolling of the leaves, just as a pair of metal strips of which the lower has the greater coefficient of expansion must, if firmly fastened together, roll upwards with a rise in temperature. The fact that a similar distention of the spongy parenchyma was observed, though not to such a marked degree, in the ‘‘artificially’’ rolled leaves of healthy plants is further evidence in support of this view. In the course of three determinations, each including five measure- ments, it was found that in the ‘‘artificially’’ rolled leaves of these healthy plants the spongy parenchyma and lower epidermis very regularly made up 57 per cent. of the total thickness of the leaf, while in corresponding normal leaves this figure varied from 46 to 53 per cent., the average being just under 50 per cent. Similar measurements made after the ‘‘artificially’’ rolled leaves had become normal showed that the spongy tissue had practically returned to its original size, the percentage it then formed of the total thickness of the leaf being 51, as against 50 per cent. in the case of similar leaves which had never rolled. Excess of starch in the leaves not confined to leaf-roll—trThere is still further evidence to connect the rolling of the leaves with the presence in them of an excessive amount of starch. Rolling of potato leaves, as is well known, is not confined to the leaf-roll disease. It may accompany other diseases, and may also occur as a result of mechanical or other injury to the plant; and in some at least of these cases the rolled leaves contain an excess of starch. This was proved to be the case in the upper rolled leaves of certain stalks of healthy plants of the variety Barley Bounty, which were partially broken across at the base, where aerial tubers were being produced in the leaf axils just above the injury. The remaining stalks and their leaves were normal in appearance. When such rolled and normal leaves were kept in the dark with their petioles standing in water for forty-two hours the normal leaves were found to be practically free from starch, except for a little, principally along the basal margins of the lower leaflets. The leaf-blades and petioles of the rolled leaves, however, contained obviously more starch at this time. It was estimated that. Murrny— On the Cause of Rolling in Potato Foliuge. 171 on the average, about one-quarter of each rolled leaflet had become clear in irregular patches, the clearing being most pronounced in the terminal leaflet as a whole, and generally in the tops of the others (fig. 10, Pl. V1). The same experiment was repeated more than once with plants of the variety Up-to-Date, which were free from leaf-roll, but which proved on examination to be affected with black-stalk rot (Bacillus atrosepticus). The rolled leaves from the top of such plants, while they retained their green colour, contained more starch than similar leaves from healthy plants; and when cut and placed standing in water in the dark became clear at about the same rate and in the same way, from tip to base, as in the case of rolled leaves from broken stalks of healthy plants. Again, excess of starch was demonstrated in the upper rolled leaves of plants of the varieties Up-to-Date and Ally, the rolling being due to obscure and apparently temporary causes, for the plants afterwards recovered. It will be observed that in all these cases an upward rolling of the leaves is an accompanying feature of starch accumulation; and, in the case of the rolling due to the breaking of the stalk at least (the only one in which the relation was determined), an increase in the depth of the spongy parenchyma was found to be another accompaniment. In the rolled leaves of this plant the spongy parenchyma and lower epidermis made up 52 per cent. of the thickness of the leaf, while the corresponding figure for healthy leaves was 47 per cent. IV.—The Cause of Starch Accumulation in Rolled Leaves of Diseased Plants. Sinee disorganization or necrosis of the phloem has been shown by Quanjer et al. (10) to be an accompanying feature of leaf-roll, and since it is believed by most plant physiologists that the carbohydrates of the plant are translocated through this tissue, the assumption was soon made that such necrosis was the cause of the accumulation of starch in the leaves; and that, in fact, the association of these two phenomena furnished a settlement of the old question as to the channel through which carbohydrates are principally distri- buted in the plant. Investigation was made as to the extent to which the phloem is disorganized in plants affected with ‘leaf-roll, and it was found to be very variable. It depends apparently on the susceptibility of the variety to the disease, the severity of the attack in the particular plant under examination, the period of the season at which the examination is made, and probably on many other factors. Thus varieties which commonly show an aggravated form of leaf-roll, such as President (which appears to be synonymous with the Dutch variety Paul Kruger) and Black Skerry, are also generally characterized by severe phloem necrosis, at least in the later stages of the disease. The same may be said for British Queen, and probably for other varieties, when the attack is unusually severe. On the other hand, when the disease occurs in its normal intensity on British Queen, Up-to-Date, and most of the varieties which are in common cultivation in Ireland, the amount of disorganization which can be seen in the phloem during the height of summer does not appear to be sufficient to account either for the external symptoms of disease or for the failure to trans- locate carbohydrates. This seems to be beyond question for the earliest recognizable stage of the disease. Hxamination made at the end of May and early in June, 1922, of plants in which starch accumulation had just begun in the lower leaves, which were showing the first signs of rolling, failed to reveal any trace of alteration in the phloem, either in the petioles of the affected leaves or in the stem below their insertion. This examination was made repeatedly 172 Scientific Proceedings, Royal Dublin Society. in comparison with similar tissues of healthy plants, and always with the same result, thus corroborating the work of Oortwijn Botjes (8), and ruling out phloem necrosis as the original or principal cause of the abnormal accumulation of starch in the leaves of diseased plants. Further, it is to be observed that there appears to be no spatial relationship - between the situation in which starch accumulation first occurs (in the mesophyll of the lower leaves) and the place at which disorganization of the phloem is supposed to be first noticeable, namely, the lower part of the stem. If the necrosis were actually the cause which precluded carbohydrate translocation, it would be reasonable to expect to find excess of starch just above the point of obstruction. This was not found to be the case. For, not only was the accu- mulation of starch at that stage confined to the blades of the lowest leaves, but, as has been stated (p. 166), less starch was found at 10 am. in the stems of diseased plants showing the first signs ofrolling than in similar healthy stems. It is also to be noticed that the disappearance of starch from the very vigorously assimilating upper leaves of affected plants is not interfered with at this time, although the destination it reaches is presumably the same as would have been reached by the starch in the lower leaves were this capable of being moved. Again, when rolled leaves are kept darkened for a considerable time, whether remaining on the plant or standing with their cut ends in water, the starch vanishes first from the neighbourhood of the lowest part of the midrib and the base of the leaflet (fig. 9, Pl. VI). This cireumstance shows that the hindrance to translocation, whether mechanical, physico-chemical, or otherwise, is less likely to be found a considerable distance away in the stem than in the leaf itself. A similar conclusion may be drawn from the fact that in the last stages of a very severe attack of leaf-roll, such as is common in the varieties President and Black Skerry, practically the only starch to be found in any part of the plant (with the exception of the tubers) was in the leaves. A systematic examination of plants of the varieties mentioned was made on July 25, 28, 31, and August 2, 1922, the material being taken at 10 a.m. to 1 p.m., and examined at once; and practically the same results were secured throughout from the two varieties. The plants selected were very badly diseased, being dwarfed and chlorotic, with upstanding leaves, which were practically all rolled, and showed considerable pinkish coloration and tip-injury. The greenest of the more or less well-developed leaves were used in the examinations. There was a great accumulation of starch in the leaves, and this was largely or entirely confined to the palisade layer. No starch was found, as a rule, in the veins; but in those portions of the spongy parenchyma furthest away from them starch was sometimes present. The midribs and petioles contained traces of starch in the form of small grains in the ‘‘starch-sheath,’’ and sometimes as isolated grains in other parenchymatous cells. The total amount of starch present in the stem was small, and was confined to the ‘‘starch-sheath,’’ with the exception of an area some little distance below the growing point, where grains oecurred in the parenchyma of the principal vascular bundles and in the pith. In proceeding from the top of the plant downwards the initially small amount of starch present in the ‘‘starch-sheath’’ became gradually less, until at about ground-level it finally disappeared. In the underground organs, with the exception of the tubers, but including the stem, stolons, and roots, starch was either entirely absent or a trace occurred in the cortex of the stolons. It should be noted that in these cases the phloem in the lower half of the stem was consi- derably disorganized. In contrast with this state of affairs, corresponding healthy leaves examined Murpay— Ou the Ouuse of Rolling in Potato Foliage. 173 at the same time showed a normal distribution of starch, there being less in the palisade layer than was observed in diseased leaves, and more present elsewhere. Similarly in healthy stems, while there was sometimes little difference in the starch-content of diseased and healthy plants near the tip, as a general rule a great deal more starch was present throughout than in corresponding tissues of diseased plants. This was particularly so in the underground portions of the stem, the stolons, and the roots. Furthermore, the lower portions of the stems of diseased plants above ground, when cut off in the afternoon and tested at once with Benedict’s solution, gave no reaction for sugar, while similar portions of healthy stems gave a strong reaction. The view is put forward, based on these results, that in the case of plants seriously affected with leaf-roll the exhaustion of starch in the regions of the leaf nearer the larger veins and the almost complete absence of starch (and sugar) along the vascular bundles in the stem in the track of movement, are probably to be explained by the fact that all the carbohydrate which can be moved has been drawn on for the most part and translocated. The remaining starch, principally in the palisade cells, is apparently not readily mobile, the reason for which is at present unknown. As a result of this, no more carbohydrate (or very little can be produced in the leaves, owing to previous starch accumulation and the disturbance it set up. The consequence is that at least in the extreme cases now being considered the plant as a whole gradually dies from starvation. This view of the case, which supposes the seat of the disturbance of starch transloca- tion in the leaf (whatever and wherever its cause may be) to lie in the leaf-blades rather than in disorganized distant tissue in the stem, is believed to be correct, because the disturbance begins.and ends in the leaves; but the problem is admittedly a complicated one. In other stages of the disease and in other varieties, when the plants are not so severely attacked and are still growing fairly vigorously, there is a very large and abnormal accumulation of starch in the petioles of the lower rolled leaves and in the stem near their points of insertion. This in itself might perhaps be accounted for by assuming it to be due to an obstruction in the phloem a little further down. But such a supposition would not account for the fact that the accumulation is known in these cases also to start in the blades of the leaves. On the whole, it seems more reasonable to consider the matter as a gradual extension of the disordered condition from the leaf-blades, where it originated, although it is possible that phloem necrosis, once it has set in, may add to the difficulty of translocation. Influence of low temperature on carbohydrate translocation.—A different theory to account for the interruption in the translocation of carbohydrate is put forward by Neger (7), and is based on experiments from which it appeared that the starch in diseased leaves—and in general also in leaves of healthy plants belonging to (assumed) susceptible varieties, or of healthy plants of (assumed) individual susceptibility—is translocated with difficulty at low temperatures (10° C.), and under :conditions of poor aeration of the leaves. The further conclusion was reached by this investigator that high temperature and good aeration promoted the translocation of starch from rolled diseased leaves. From these findings the general inference was drawn that the leaf-roll disease was the result of cold nights or of cold wet weather, and was merely the expression on the part of certain plants or varieties of their inability to trans- locate starch normally under such conditions. The experiments of Neger, which were carried out entirely with shoots cut from the plants, will be further discussed below. Before doing so, an experiment to test the effect of low temperature on starch translocation in a growing plant will be described. 174 Serentijic Proceedings, Royal Dublin Society. A healthy plant of the variety Up-to-Date (which is quite susceptible to leaf-roll) was grown in a flower-pot and placed in a position where it was as little as possible exposed to leaf-roll infection through insect agency. As a matter of fact, neither it nor its neighbours ever became infected. On August 8, when the potted plant was well developed, normal in appearance, and just showing its flower-buds, it (together with its pot) was placed during the night (from 5 p.m. to 9 a.m.) in a small ice-chamber, the temperature of which varied between 55°C. and 72°C. In the evening, when the plant was first brought in, its leaves contained a normal amount of starch, but after a night in the refrigerator they were found to contain none, and the plant was apparently in no way injured or altered by the treatment. It was then exposed to full day- light in its original position, and again placed at night in the ice-chamber, and this treatment was repeated for eleven consecutive days. The range of tem- perature at night was the same throughout. Periodical tests of the starch- content of the leaves were made during and at the end of the experiment, but an accumulation of starch was never found in the morning, and no rolling of the leaves or any other abnormal symptom appeared. In seeking for an explanation of the different results secured by Neger under apparently similar conditions, it is to be observed that this worker used cut shoots and not the whole plant in his experiments. When a normal potato shoot or leaf is cut off and placed with its end standing in water in the dark, it loses its starch in a comparatively short time; but the process is one of hydrolysis of the starch and eventual consumption of the resulting sugar in respiration, and not of translocation in the generally accepted sense of the word; for there are no normally functioning organs of storage or growth by which the carbohydrate ean be absorbed, and no soluble carbohydrate passes out of the stem into the water. Even the apparent transference of starch from the leaf-blades to the petiole is largely illusory, although some translocation may perhaps take place, because there is a gradual but less rapid diminution in the total amount of starch present in the petiole also. In the case of larger shoots there would, of course, be a correspondingly more extended field in which such limited trans- location might take place, but the conditions are so unnatural as to render doubtful the applicability of results secured under them to the normal growing plant. An even more rapid disappearance of starch is to be observed (as Neger also noted) in the case of cut leaves and shoots which are not put standing in water, and which wilt in consequence. The obvious conclusion, however, that translocation could not account for the disappearance in this case was not drawn by this worker. A similar process of hydrolysis and respiration goes on, but probably more slowly, when diseased shoots and leaves are allowed to stand in the dark with their cut ends in water; but it is not possible to measure by means of the iodine test the relative rates of carbohydrate translocation in healthy and diseased leaves or shoots kept in this way, because the starch-content of the leaves is very different to begin with, and because the diminution in starch-content is not solely or in the main due to translocation. While practically all the starch contained in a healthy leaf cut off in the evening may he dissolved during the night and retained as sugar, with the exception of that used up in respiration, it is obvious that a diseased leaf cannot under the same conditions be in a position to become free from starch in the same time, even if the starch in it were equally hydrolysable, because of the greater amount of it originally present. A further complication would also ensue, since the resulting sugar, for which there is no outlet but that of respiration, would be present in greater quantity, and would be likely to retard or prevent the final disappearance of the starch. It there- MurpHy—On the Cause of Kolling in Potato Foliage. 175 fore appears that Neger’s conclusions do not follow from his experiments. The latter refer in the main to the hydrolysis of starch in the leaves, and only in a minor degree, if at all, to the translocation of the resulting carbohydrate. The only point proved by them is that rolled leaves of diseased plants contain much more starch than similar healthy leaves.* V.—Aistological and other Symptoms of Leaf-Roll. During the course of a systematic examination of the starch-content of various portions of diseased and healthy plants certain histological differences were discovered or re-investigated. As was first pointed out by Schander and yv. Tiesenhausen (12), necrosis of the phloem, which Quanjer regards as being the underlying cause and principal symptom of leaf-roll, is not confined to plants affected with the leaf-roll disease. These two workers showed that similar injury to the phloem may be found in plants affected with a number of other diseases, including curly dwarf, black-stalk rot, and blight; and that it may even occur in maturing healthy plants. It has also been pointed out else- where by the present author (5) that a very thorough disorganization of the phloem (as well as of other tissues) occurs in streak disease. Recent work has again confirmed the fact that this type of injury is not confined to leaf-roll. It was found during the past year to occur in the lower portions of the stalks of potato plants attacked by a disease possibly due to eel-worms, and it has also been seen in the lateral veins, midribs, and petioles of potato leaves attacked by Phytophthora infestans. In neither case were the plants affected with leaf-roll. The disorganization in these diseases was not, of course, confined to the phloem, but the effect on that tissue was exactly of the same kind as in leaf-roll, only rather more pronounced. The attack began in the oldest cells, the walls of which became brown in colour, and afterwards collapsed. The walls of such cells no longer gave a cellulose reaction, but the nature of the change in com- position was not further inquired into. In the blighted plants the death of the phloem was apparently due to a toxic substance, which operated some distance in advance of the parasite itself. The first cells to be killed in the petiole were those in the epidermal and sub-epidermal layers (giving rise to the brown stripes characteristic of the disease on stems and leaf-stalks), and from them the parenchymatous ground tissue and the phloem groups nearest the surface of the stalk were attacked in turn. A difference manifests itself in the behaviour of diseased and healthy leaves in the matter of starch evacuation which does not seem to have been sufficiently emphasized. It seems to be a general rule that the disappearance of starch begins at the bases of diseased leaflets—that is, if any noticeable clearing whatever takes place—and proceeds for a less or greater distance towards the apex. This feature is most clearly seen when diseased leaflets in the earliest stage of rolling are darkened for prolonged periods (fig. 9, P]. V1). Im such cases a sharp line divides the tissue which is free from starch from that which retains it in apparently undiminished quantity. ‘So far as our experience goes, it also seems to be invariably true that the starch begins to disappear from healthy leaflets in 1Tn a subsequent paper, a note of which has just been seen (Centralbl. f. Bakt. Abt. II, 54 Band, 1921, p. 512), Neger records an experiment in which living plants diseased with leaf- roll were kept at night under different conditions. From this it appears that those kept during the night at 20°C. became entirely healthy, and showed normal (‘‘good’’) translo- cation, as measured by the application of the iodine test to the leaves. However this may be (and it should be noted that the respiration factor seems to have been left out of account), it does not prove the author’s contention that leaf-roll, or the lability of a plant to it, is an expression of inability to translocate carbohydrate at a low temperature. 176 Scientifie Proceedings, Royal Dublin Society. the first instance at or near the tip, and then gradually vanishes from the lower portions. Here there is not the same sharp line bounding the starch-free area, and the progress of hydrolysis is not so regular, but the direction in which it proceeds is unmistakable. Furthermore, in the case of healthy leaves the starch disappears first from the tip of the terminal leaflet, then from the tips of the next pair, and so on to the base of the leaf; each leaflet at any one time showing a somewhat larger area still full of starch than the leaflet immediately above. This method of evacuation is also followed by healthy leaves and leaflets which contain an excess of starch, due to causes already discussed, and which roll upwards in consequence (figs. 6 and 7, Pl. VI). The brown areas which generally arise in the course of time on rolled leaves appear to originate in the death of a single cell either of the upper or the lower epidermis, this being most frequently a guard-cell, but not invariably so. The walls of the cell become brown as well as the contents, and the discoloration spreads to the wall of any neighbouring cell. The cells involved collapse and fall in, particularly when the affected area starts on the lower surface of the leaf. All the subjacent tissues of the mesophyll are liable to be attacked in turn, until the lesion extends from one side of the leaf to the other. No parasite was seen in connexion with this injury. It may be remarked that the cells of the epidermis of diseased leaves conta apparently more vigorous nuclei, more abundant cytoplasm, and a greater quantity of starch than corresponding cells of healthy leaves. The difference is more marked in the case of the upper epidermis. V1I.—Jnsect Carriers of Leaf-Roll. A survey, which was unavoidably somewhat hurried, was made by my colleague, Mr. Rhynehart, during July, 1921, of the common insects which occurred in the experimental potato disease plots; and the writer wishes to acknowledge the help thus given. The following insects were found on the plants, and were presumed to be feeding on them :— Calocoris bipunctatus (Capsid Bug). Abundant. Typhlocyba Ulmi (Jassid). Abundant. Philaneus spumarius (Frog-hopper). Abundant. Psylliodes affinis (Potato flea-beetle). Abundant. Aphides (Genus and sp. not determined). Fairly common. Lygus pratensis var. campestris (Capsid). Fairly common. Typhlocyba sp. (Jassid). Fairly common. Anthocorus sylvestris (Anthocorid). Searee. Aetorhinus angulatus (Capsid). Searee. Idiocerus sp. (Jassid). Rare. Bythoscopus sp. (Jassid). Rare. Among these insects aphides certainly appeared to be of secondary impor- tance. They were unevenly distributed in the plots, being almost absent in some parts. This may have been due to the great prevalence of ladybirds (Coccinella sp.) and their larvae in the month of June, these being practically the only insects then to be found on the foliage. The same prevalence of these beetles and their larvae and searcity of aphides was noted in 1922. _ Estimating the aphides individual for individual against the other kinds of insects, they should probably be classed as ‘‘fairly common”’ on the average in 1921, Infection experiments with insects other than aphides.—Experiments to test Murpuy— On the Cause of Rolling in Potato Foliage. 177 the capacity of the four commonest of the insects listed above to carry leaf-roll were undertaken at once by transferring individuals taken from diseased plants to healthy plants, the latter being protected by muslin cages which had been placed in position soon after the plants appeared above ground. Each caged healthy plant was provided with a caged control plant, the two being derived from the two halves of one tuber. The caged plants were separated by a distance of four yards of unplanted ground from the nearest potatoes, which were for the most part healthy. It was found that under local conditions, and at least in the abnormal season of 1921, the caging of a healthy plant with muslin was not a guarantee of protection against leaf-roll infection unless the plant was also placed at a certain distance from diseased plants. This distance need not apparently be very great, probably because a small space of clear ground is a considerable help in preventing infestation with aphides, while the other insects (which, it will be seen, were proved to be carriers also) are com- paratively easily excluded, although much more active. The experimental infestation of the caged plants with the various kinds of insects took place on July 27, 1921, the latter being collected from diseased plants, and one lot of each, from five to twelve in number, being introduced into a cage. No insects were put in the adjacent cages containing the control plants. Owing to cireum- stances which were beyond control, it was not possible to note the behaviour of the plants during the latter part of the season, but the tubers of each of them were dug and saved separately at the usual time. The tubers were all planted in separate lots in the open field in 1922. All those from the control cages produced healthy plants, except those from cage No. 8 (two in number), both of which were diseased with leaf-roll. The plants from tubers from the corresponding cage, to which frog-hoppers had been introduced in 1921 (again two in number), were also both diseased with leaf-roll. Unfortunately there appeared in this exceptional control cage in 1921 a ‘‘volunteer’’ plant from a tuber which survived in the ground from a previous erop of the same variety as the experimental plant (Up-to-Date). This crop was very badly affected with leaf-roll in 1920; and although the plant was removed as soon as noticed, under the circumstances it is thought better to exclude this part of the experiment from consideration. The remaining results were as follow :— 1921. 1922. Healthy plant experimentally infested with 12 capsid bugs (Calocoris bipune- tutus) taken from affected plants— ) produced 5 plants, all diseased with leaf- roll. Control plant, from half of same tuber, Sate trat Dae tias Ee } produced 4 plants, all healthy. Healthy plant experimentally infested ) with 12 jassids (Lyphlocyba Uli) taken from affected plants— produced 3 plants, all diseased with leaf- roll. Soe eee ee half of same tuber, produced 4 plants, all healthy. produced 4 plants, of which one was dis- with 6 flea-beetles (Psylliodes afinis) eased with leaf-roll, and 3 were healthy. taken from affected plants— Control plant, from half of same tuber, Healthy plant experimentally infested | not infested— produced 5 plants, all healthy. 178 Scientific Proceedings, Royal Dublin Society. Where the disease occurred in 1922 in this experiment it appeared early in the season in the pronounced secondary form. It was clearly visible at the time of the first detailed examination of the plots for leaf-roll on June 20; and the contrast between the produce of the infected and the control plants was then striking. This is illustrated in the photographs taken on July 10, repro- duced in figs. 11 and 12, Pl. VI, which show one of the plants infected through the agency of capsid bugs and one of the still healthy controls beside it, the two being reduced to the same extent. The plants infected by means of jassids and their corresponding controls presented an entirely similar contrast at that time. The starch-content of the lower leaves of the diseased plants was com- pared with that of similar leaves from the healthy control plants, and the usual difference was evident. While the above results show hardly more than a suspicion that the potato flea-beetle can act as a carrier of leaf-roll, there is little room for doubt that both capsid bugs and jassids act as efficient transmitters. This is important as showing that there is no exclusive specific relationship between aphides and the actual cause of leaf-roll, which is presumably an organism. This finding might indeed have been expected in the case of capsids from the work of Oortwijn Botjes (8), if it be assumed that the German common name ‘‘ Wanze”’ refers to a species of capsid. This author seems to have been in doubt about his result, because in one experiment with these insects no infection followed, while in the other, in which it did, it was considered equally attributable to aphides, which accidentally found their way into his cages. General observations in the plots made in 1921 and 1922 showed the com- parative scarcity of aphides in both years; and yet the ease and rapidity with which infection was carried in quantity over comparatively long distances were clear. Hence it appeared that some other and more active carrier must be concerned in the matter, and that this in fact was the main problem. It may be noted that in many of the original and later experiments of Quanjer, van der Lek, and Oortwijn Botjes (10 and 8) the removal of healthy plants to a distance of from 2:5 to 4 metres from diseased individuals frequently protected them to a very large extent, if not entirely, from infection. Similar results were obtained by the writer (5) in Eastern Canada. Under the conditions under which the experiments now being described were carried out such a comparatively small degree of isolation was of little or no avail, and mass infection occurred over very much greater distances. In 1921 it is believed that jassids were the principal carriers. On account of the proximity of a number of elm trees, these insects were so abundant that if a potato plant in the vicinity of the trees were disturbed on a sunny day in July they flew into the air almost as thickly as bees in a swarm. Fortunately they appeared to be entirely absent from the plots in 1922, which were in a different situation. Capsid bugs, however, were present in that year in greater numbers than in the preceding one. This was particu- larly so near a hedge bordering on the plots. On a day following a pericd of heavy rain as many as nineteen of these insects, feeding voraciously, were counted on the exposed portions of leaves of one plant in such a situation. How many more there may have been concealed amongst the foliage was not determined. Insect infection through sprouts.—The presence of aphides feeding in late winter and in spring on the sprouts of seed potato tubers before they are planted is apparently not uncommon in some places in Ireland as well as in Great Britain, and probably elsewhere. It was known previously that they occurred thus on the farm on which the experiments described here were caried out; and since the lots of potatoes selected for planting in the plots were stored alongside of each other in small chip baskets (‘‘punnets’’), and included healthy tubers as Murpuy—On the Cause of Rolling in Potato Foliage. 179 well as others from. plants which had suffered from various diseases, it became a matter of urgency to determine whether aphides could carry infection from the sprouts of diseased tubers to those of healthy ones. This was proved to be the case so far as leaf-roll, at least, is concerned.* The details of the experiment are as foilows. The sprouts on a number of tubers which had been kept on a dish in a lobby, and which originally came from England, the past history of the tubers being not known, were found in February, 1922, to be strongly infested with aphides. One of these tubers, which was assumed to be healthy, but which gave rise later to a plant affected with leaf-roll, was selected, and eighteen aphides (Myzus Persicae, Sulzer)* from its sprouts were transferred to the sprouts of one-half of a tuber from a plant known to have been affected with mosaic in 1921. The other half of this mosaic tuber was kept separately, and no aphides were placed on its sprouts. Similarly, eighteen aphides from the same original source were placed on the sprouts of a half tuber derived from a healthy plant which was caged in 1921, while the other control half received no aphides. On two subsequent occasions further lots of twenty-five and twelve aphides respectively were transferred from the same source to the sprouts of the same two half tubers, because those originally trans- ferred did not multiply with sufficient rapidity. From the sprouts of the mosaic half tuber aphides were transferred, as their numbers increased and permitted (about twelve to twenty-four were used in each case) to the sprouts of three half tubers derived from healthy plants caged in 1921. The corresponding three half tubers were kept free from aphides. Similarly, from the sprouts of the healthy half tuber first infested from the original source aphides were placed on the sprouts of three further healthy half tubers, the corresponding halves of these receiving no aphides. Each of the half tubers was covered almost completely with sterilized soil which half filled a flower-pot, the terminal sprouts, on which the aphides fed, alone projecting. The half tubers and their sprouts were in each case enclosed in a wide glass cylinder, pressed down nearly two inches below the soil-level, and closed on top with a triple layer of muslin. The non-infested sprouted half tubers were similarly protected, and remained free from aphides. The left-hand portion of accompanying figure diagrammatically illustrates the transference of the aphides from the sprouts of the original tuber (centre) to those of the mosaic and healthy half tubers, and from both of the latter to the three healthy half tubers in each case. The controls (not infested) of all these are shown with the cut surfaces facing in the opposite direction in the top and bottom rows. The original state of health of all the tubers is indicated by shading. When the aphides had fed on'the sprouts for from 13 to 52 days (and still longer in the case of those first infested), the sprouts, including those of the control half tubers, were thoroughly fumigated with a commercial greenhouse 1Owing to unavoidable circumstances, the experiment about to be described and another similar one miscarried before any results were secured regarding the transmission of mosaic; but that this disease may also be carried in the same way is only to be expected. This con- clusion can perhaps be drawn from an experiment of Schultz and Folsom (13), in which sprouts were infected with leaf-roll and mosaic respectively by means of aphides taken from the foliage of diseased plants. These authors state that ‘‘this experiment is probably not duplicated by natural conditions’’; but storage conditions in milder climates may be such that the transmission by means of aphides of infection from the sprouts of unplanted diseased tubers to the sprouts of healthy tubers may occur and he of serious practical im- ortance. fs 2Thanks are due to Messrs. Rhynehart, Theobald, and Laing for identifying the aphides used. 5 SOIENT, PROC. R.D.S., VOL, XVII, No. 20. O11 180 Scientific Proceedings, Royal Dublin Society. fumigator, the chimneys being removed. The pots were then filled up with sterilized soil, so that the sprouts were covered, and the plants were allowed to develop. They were divided into four lots, depending on the probability of successful infection, which were separated a considerable distance from each other out of doors on the flat roof of the College of Science building in the centre of Dublin. The plants were kept under close observation, and whenev er, as happened, a few aphides appeared on the foliage of any of them, they were all fumigated again, still in separate lots. That these aphides ean have had no part im the successful infections which followed is clear from their insignificant numbers and immediate suppression; from the early development of leaf-roll ; from the fact that the disease appeared at once in the secondary form; and from the unexceptional behaviour of the control plants. aaa Idd cneed aye = LEAF-ROLL i) = MOSAIC The results of the experiment are diagrammatically represented in the right- hand portion of the figure already alluded to. The original tuber was evidently affected with leat-roll from the beginning, the resulting plant being badly diseased. The leaf-roll disease was transferred from this to both the mosaic half tuber and the healthy half tuber first infested, for the resulting plants showed leaf-roll (secondary) combined with mosaic in the one case and leaf-roll (secondary) alone in the other. The aphides taken from the former infected with leaf-roll (secondary), two out of the three plants arising from healthy half tubers on the sprouts of which they were placed. The third plant remained apparently healthy; but it should be stated that observations ceased abruptly on June 26, when the experiment had to be abandoned. Up to that date also leaf-roll had not appeared in either of the two plants derived from healthy half tubers on the sprouts of which aphides from the originally infested healthy half tuber had been placed (second horizontal row from the top on right of diagram). The third tuber in this row failed to produce a plant. In spite of Murpuy—On the Cause of Rolling in Potato Foliage. 181 the experiment being brought to a premature close, the net result was that of the seven plants produced by half tubers the sprouts of which had been infested with aphides from diseased tubers, four developed symptoms of leat- roll within about one month of their appearing above ground.’ All the controls remained entirely healthy throughout, except, of course, the plant from the half tuber known to have been diseased with mosaic. The diagnosis of leaf-roll was in all cases confirmed by making a test of the starch-content of the leaves; and it was found that all the plants which showed symptoms of leaf-roll also showed starch accumulation in the lower leaves, while the upper leaves showed none. Corresponding leaves, both upper and lower, of the plants which showed no signs of leaf-roll, when tested at the same time, reacted normally. It would appear, therefore, that the potato may be in danger of infection, certainly with leaf-roll, during its resting period. This is particularly the ease if tubers are stored in houses during the winter, or are sprouted in boxes betore planting. In all probability the more delicate and etiolated are the sprouts the better the nidus they provide for the insects; for some cases have come under observation in which improperly sprouted tubers became so infested that at planting time they were almost covered with the cast cuticles of aphides, and were slimy with their exudate. On the other hand, sprout- infestation does not appear to occur in the ease of pitted potatoes, although, of course, sprouts may develop on the tubers. In the older Enelish literature concerning ‘‘Curl’’—a term which would appear of necessity to have included leaf-roll, at any rate so far as Great Britain and Ireland are concerned—-several observers recorded the fact that potatoes stored in houses suffered more from the disease than similar potatoes stored in pits. Atanasoff (1) refers to one of these cases; and it is possible that under certain conditions the correctness of the observation might be less debatable than would appear on the surface. The presence of aphides on the sprouts of seed potatoes may be of importance in another way, because it was observed that in the course of their development such sprouts can carry the aphides above ground, and thus, no doubt, give rise to the first seasonal infestation of the foliage with these insects. Where this oceurs, the assumption made in many quarters that the infestation of potato foliage arises from the migration of aphides durimg the summer from an alternate host, such as plants of the order Rosaceae, is obviously an unnecessary one. Eyen when rose bushes infested with aphides are in the immediate vicinity of potato plants, the amount of migration that takes place is sometimes negligible, for the potatoes may remain practically free from aphides. This was the case in 1922 in the writer’s garden, where two small patches of potatoes were under observation during June, July, and part of August. A large ‘‘rambler’’ rose 1 As has been stated, the disease appeared at once in the secondary form, and therefore the dictum that secondary leaf-roll results only from the planting of tubers from infected plants requires modification. There was another exception of common occurrence in the dry, hot summer of 1921—namely, the development of the full symptoms of secondary leaf-roll on plants which had previously shown clear primary rolling, and which previous to that had all the appearance of health—all within the course of the same season. The disease took this course apparently when plants were infected through the foliage in June and July. The rolling and discoloration characteristic of the primary phase appeared first in the top of the plant, and then extended gradually downwards so as to involve all the leaves. At this stage such plants were indistinguishable from secondarily diseased plants produced from infected tubers, and carbohydrate translocation from their lower leaves appeared to be similarly affected. The course of the disease here described is evidently different from that following infection through the sprouts. 182 Scientific Proceedings, Royal Dublin Society. bush, eight feet high, which was heavily infested with aphides, grew within ten feet of the potatoes; and from this the insects dispersed to some extent to neighbouring comparatively non-infested rose trees, but only to a very slight extent to the potatoes. An effective fumigant for aphid destruction.—It has been stated that the development of aphides was feared during winter storage on the sprouts of the various seed potatoes intended for the experimental disease plots. This un- fortunately began to take place early in March, 1922; and after recourse had been had to a twice-repeated fumigation with commercial fumigants without completely satisfactory results, a successful method of treatment was discovered. The sprouting boxes containing the seed potatoes were placed in a small room having a capacity of 1,700 cubie feet. One pound of tetrachlorethane was then distributed in small lots in earthenware saucers in various parts of the room and allowed to evaporate. The room was then closed up, and was not opened for three days. The effect was immediate and lasting, for no more living aphides were seen up to the time planting was concluded, about thirty-three days later. The vapour of this chemical does not injure the tubers or sprouts, and it is perfectly safe and not unpleasant to use. VII.—Summary. It was established that the presence of an excess of starch in the rolled leaves of diseased plants is a constant symptom of leaf-roll. The rolling of the leaves of diseased plants was found to be preceded by the accumulation of starch in the mesophyll. The artificial darkening of diseased plants before their leaves rolled, and the consequent reduction of photosynthesis to a minimum, was found to prevent the rolling of the leaves for long periods. Temporary rolling of the leaves of healthy plants was brought about by depriving the latter of most of their growing points and storage organs. Accom- panying the rolling a great excess of starch was found in the rolled leaves. The rolling and excess of starch afterwards disappeared when normal growth was allowed to proceed. It is concluded that rolling of the leaves is a direct consequence of the presence in them of an abnormal amount of starch, and probably of other carbo- hydrate, and that it is caused by the distention of the spongy parenchyma, which was demonstrated. Starch accumulation in the leaves accompanies rolling due to some other causes, such as injury to the base of the stalk, attacks of black-stalk rot, and other obscure disturbances. Evidence is presented to show that the seat of the disturbance in the trans- location of carbohydrate from the leaves of diseased plants resides in the blades of the leaves, where the accumulation of starch begins and ends, and not in the disorganization of the phloem in distant tissues. re Low temperatures were found incapable of causing healthy leaves of a living plant to accumulate starch or to roll. ’ The presence of disorganized phloem was established in plants attacked by Phytophthora infestans and in others apparently suffering from an attack of eel-worms. ae / The disappearance of the starch in diseased leaflets proceeds from base to tip, but in healthy leaflets from tip to base. fest P The brown spots which develop on affected leaves originate in the death of a single cell of the epidermis. Murpuy— On the Cause of Rolling in Potato Foliage. 183 It was proved that capsid bugs (Calocoris bipunctatus) and jassids (Typhlocyba Ulm) act as carriers of leaf-roll in the field. Aphides (Myzus Persicae), when they occur on the sprouts of unplanted tubers, were also shown to be carriers of leat-roll, and to be capable of giving rise to the earliest infestation of the foliage with these insects. The vapour of tetrachlorethane was found a safe and efficient medium for ridding sprouted tubers of aphides. LITERATURE CITED. 1. Aranasorr, D.—A Study into the Literature on Stipple-Streak and related Diseases of the Potato. Med. v. d. Landbouwhoogeschool, Wageningen, Deel 26, 1922. 2. Dosy, G. (and J. Bodnar, in part).—Biochemische Untersuchungen tiber die Blattrollkrankheit der Kartoffel. Zeitschr. f. Pflanzenkrank., 21 Bd., pp. 10-17 and 321-86; 22 Bd., pp. 204-11 and 401-3; 25 Bd., pp. 4-16; 1911-15. 3. Esmarcu, F.—Zur Kenntnis des Stoftwechsels in blattrollkranken Kartoffeln. Zeitschr. f. Pflanzenkrank., 29 Bd., pp. 1-20, 1919. 4. Hmrner, L. (and G. Gentner, in part)—lUeber den Zusammenhane der Blattrollkrankheit der Kartoffel mit der Starkeanhéutung in ihren Blattern, etc. Prakt. Blat. f. Pflanzenbau, 16 Jahre., pp. 138-41; 17 Jahrg., pp. 15-19; 1918-1919. . Murpuy, P. A.—Investigation of Potato Diseases. Dominion of Canada Dept. of Agric., Bull. 44 (Second Series), 1921. 6. Necer, F. W.—Die Blattrollkrankheit der Kartoffel. Hin Beitrag zur Aetiologie der Krankheit. Deut. landw. Presse, 45 Jahrg., pp. 469-70, 1918. 7. Necer, F. W.—(Same general title.) Zeitschr. f. Pflanzenkrank., 29 Bd., pp. 27-48, 1919. OortwisJN Borges, J. G.—Die Blattrollkrankheit der Kartoffelpflanze. Inaug. Dissert. Landw. Hochschule, Wageningen, 1920. 9. QuangeR, H. M.—Sur la fonction du tissu eribblé. Med. v. d. Landbouw- _hoogeschool, Wageningen, Deel 16, pp. 95-104, 1919. 10. Quanger, H. M., H. A. A. VAN DER Lex, and J. OortwiuN BotsEs.—Nature, mode of dissemination, and control of phloem necrosis (leaf-roll) and related diseases, i. a. Sereh. Med. y. d. Rijks Hoogere Land— .. . Bouwschool, Wageningen, Deel 10, 1916. 11. Quangrr, H. M., J. C. Dorst, M. D. Dist, and A. W. v. v. Haar.—De Mozaiekziekte van de Solanaceeén hare Verwantschap met de Phloeem- necrose en hare Beteekenis voor de Aardappeleultuur. Med. v. d. Land- bouwhoogeschool, Wageningen, Deel 17, 1919. Summary with addi- tional matter in Phytopath., vol. 10, pp. 35-47, 1920. 12. Scuanver, R., and M. v. Tresennausen.—Kann man die Phloemnekrose als Ursache oder Symptom der Blattrollkrankheit der Kartoffel ansehen? Abstract in Zeitschr. f. Pflanzenkrank., 25 Bd., pp. 16-18, 1915. 13. Scuuurz, E. S., and D. Fousom.—tLeat Roll, Net-Necrosis, and Spindling- Sprout of the Irish Potato. Jour. Agr. Res., vol. 21, pp. 47-80, 1921. 14. SprecKERMANN, A.—Beitraige zur Kenntnis der Bakterienring- und Blatt- rollkrankheiten der Kartoffelpflanze. Jahresber. d. Ver. f. angew. Botanik, 8 Jahrg., pp. 1-19 and 173-7, 1911. 15. Vécutrnc, H.— Ueber die Bildung der Knollen. Bibliotheca botanica, Heft. 4, pp. 4 et seg., Cassel, 1887. SCIUNT. PROC. K.D.S., VOL. XVil, No. 20, 2k on ive) 184 Scientifie Proceedings, Royal Dublin Society. Fi 1 bo Go 10. ie 12. EXPLANATION OF FIGURES. Puate VI. g, . Leaf-rolling ‘‘artificially’’ induced in healthy Up-to-Date plant by removal of tubers and axillary shoots. Photographed on fifth day after treat- ment. . Untreated control plant corresponding to plant illustrated in fig. 1. The two were derived from the halves of a single tuber, and were photo- eraphed at same time and to same scale. . “‘Artificially’’ rolled leaf from healthy plant, as illustrated in fig. 1. Photographed on seventeenth day after original treatment of plant. One of the previously ‘‘artificially’’ rolled plants, showing normal foliage after tubers and axillary shoots had been allowed to develop. Note de- velopment of foliage leaves on floral shoot. Leaf from previously ‘‘artificially’’ rolled healthy plant after disappearance of rolling. Leaflet from ‘‘artificially’’ rolled healthy plant (on right), and from normal control plant (on left), both having been kept for sixty-four hours in dark room. The leaflet from the treated plant contains a great excess of stareh, and has cleared only very slightly at the tip. The normal leaflet is free from starch. Another pair of leaflets, one from ‘‘artificially’’ rolled plant (on right) and the other from normal control plant (on left), showing further progress of starch disappearance after a period of seven days in dark room. The resumption of normal starch translocation in previously ‘‘artificially’’ rolled foliage. Leaflet on right, from previously rolled plant, contains slightly less starch than leaflet on left from normal control plant. Both kept for twenty-nine hours in dark room. Leaflet on left taken from diseased plant which showed early stage of rolling, and was then darkened in field for twenty-one days. Note large amount of starch still present, and progression of clearing from base upwards. Leaflet on right, from healthy plant of same variety which was darkened in field for twenty-four bours, is almost free from starch. Variety, British Queen. Accumulation of starch in leaves accompanying injury to base of stalk. Leaflet on left was taken from top of injured stalk, and shows excess of starch. Leaflet on right from top of normal stalk of same plant. Both forty-two hours in dark room. Variety, Barley Bounty. Plant (of season 1922) diseased with leat-roll, being one of five diseased plants which were the progeny of a caged healthy plant on which capsid bugs (Calocoris bipunctatus) taken from diseased plants had been placed in 1921. Healthy control plant (of season 1922), corresponding to plant shown in fig. 11. This was one of four healthy plants which were the progeny of a caged healthy plant from which insects were excluded in 1921. The two original plants of 1921 were derived from the halves of one tuber. Both plants photographed to same scale. Variety, British Queen. SCIENT. PROC. R. DUBLIN SOC., N.S., VOL. XVII. PLATE VI. MuRPHY. IN@, Bil. ON THE CHANNELS OF TRANSPORT FROM THE STORAGE ORGANS OF THE SEEDLINGS OF LODOICEA, PHCENLY, AND VICLA. By HENRY H. DIXON, Sc.D., F.R.S., Professor of Botany in the University of Dublin; AND NIGEL G. BALL, M.A., Assistant to the Professor of Botany in the University of Dublin. (PLates VII--XT.) (Read January 23. Printed June 18, 1923. REASONS have been given elsewhere (8 and 4) why it is difficult to accept the generally received view that the bast or phloem is the main channel by which organic substances are transmitted from place to place in the higher plants. The suggestion was there put forward that probably the wood may be more properly regarded as this channel, and evidence which favoured this view was adduced. ' The general presence of tracheae in the connecting channels of the organs of mature plants can possibly be explained by the exigencies of water-supply; hence we cannot draw conclusive inference on this point from the development of tracheae in the bundles connecting the assimilating and storage organs of these plants. : In the case of some seedlings, however, the conditions are different. Here the storage organs are connected with the embryo, which either has an inde- pendent water-supply from the radicle, or is adequately provided for in this respect of imbibition. Hence, on the assumption that the phloem is responsible for the transport, we might expect that the xylem, beg superfluous, would either not be developed at all in these connecting organs, or would be represented by merely vestigial traces. It was then with the object of collecting evidence on this point that we have investigated the structure of the organs connecting the stores with the embryo in some seedlings. Lodoicea sechellarwm. The general course of the germination of this palm has been described by several observers. As is well known, the seed is of remarkable size, weighing, according to Sir William Hooker, 20-25 Ibs. (6). It is deeply bilobed, and is covered by a hard black shell—the fruit. In germination the embryo emerges from the depression between the lobes. It develops as a cylindrical mass, 1:5- 3-0 em. in diameter, and, turning upwards for a little, then pushes its cylindrical SCIENT. PROC. R.D.S., VOL. XVII, NO. 21. 21 186 Scientific Proceedings, Royal Dublin Society. body in a more or less horizontal direction. After a horizontal growth of 50- 300 cm. the end turns downwards and buries itself in the ground. The buried lower tip now develops as a root, while the part just above the level of the soil greatly increases in diameter. Soon a longitudinal split in this enlarged region allows the developing plumule to emerge. These relations are very well seen in fig. 1, Pl. VU, which is a photograph of a fruit of Lodoicea germinating in Kew. We are indebted to Dr. A. W. Hill, F.r.s., for this interesting picture. The whole growth of the embryo until it forms connexion with the soil is evidently made at the expense of the material stored in the seed and transmitted through the cylindrical body to the growing parts. Even for a long time after the root has entered the soil much nutriment is conveved to the embryo from the seed. Farmer (5) records a ease in which connexion was maintained between the seed and the young plant for five years. Morphologists seem agreed that the organ connecting the young plant and the seed is the cotyledon (14). The basal part of this leaf forms a complete circular attachment at the first node of the embryo, and is excavated into a conical cavity overarching the plumule. Above this cavity the walls converge and form a solid cylindrical petiole, which is continued into the seed. Within the seed its distal end expands into a bilobed, rounded mass—the haustorium. The surface of the haustorium is corrugated and thrown into many folds and ridges; secondary folds imposed upon these make the surfaces of the larger corrugations densely papillose. This papillose surface is in close contact with the store material filling the seed, and exposes a very large area to the latter, both for the excretion of dissolving enzymes and for absorption. By the kindness of Dr. A. W. Hill, we were able to make a histological examination of a germinating seed. The seed was one of four which germinated during 1922 at the Royal Gardens, Kew; and, having shown signs of retarded erowth, was most kindly placed at our disposal. The embryo was then protruding from the shell about 30 em. It had a twisted and seared appearance. When the seed was opened, the haustorium was found to be about 10 em. in diameter. It was of an irregularly rounded form, deeply grooved and papillose. It was creamy white, and contrasted with the bluish white of the endosperm. A narrow neck connected the haustorium with the petiole, apparently deeply constricted by the shell of the fruit. The distal part of the petiole with which this neck was connected was flattened by the pressure of the two lobes of the fruit. Once clear of these lobes, the petiole became approximately circular in section. The outer surface of the petiole was marked with irregular longitudinal erooves and ridges. It was brown in colour, with irregular white blotches. Its transverse section was limited by 4-6 layers of parenchymatous cells, the contents of which had disappeared, and the walls alone of which persisted. These cells in transverse section are about 0:02 X002 mm. They are about 0:07 mm. in length. Within them there follow 4-6 layers of sclerenchymatous cells, about 0:01--0:02 mm. in diameter and 0:06--0:08 mm. in length. Their thickened walls almost completely obliterate their lumen. From the sclerenchymatous layer inwards the mass of the petiole is composed of thin-walled parenchyma built of cells, which in the outer regions are comparatively small, viz., 0:01-0:03 X 0:06-0:08 mm., while in the inner regions the dimensions are 0:05 X 0:10 mm. or more. Seattered among these cells were many tannin-sacs, and throughout the tissue were inter-cellular spaces. There was no starch in any of these cells. Embedded in this fundamental tissue were the vascular bundles to the number of 800-1,200 in one cross-section. They exhibited the usual structure of the vascular bundles of xerophytic monocotyledons. One of them is repre- Dixon anv Batn— Channels of Transport in Seedlings. 187 sented in cross-section in fig. 7, Pl. VIII. The sclerenchymatous sheath of each bundle was largely developed, occupying about 75 per cent. of the area of the cross-section. Sunken in the outer surface of this sheath, rows of small cells containing siliceous spherules, as is common in palms and orchids, were observed (2 and 18). The phloem and xylem within the sheath were normally developed, the ratio of the areas of their cross-sections being on the average about 6:5. The vessels of the xylem having a diameter of 0:05-0:07 mm., were strengthened by annular and spiral supports. The diameter of the sieve-tubes averaged about 0:01 mm., their length about 0:15 mm. In the specimen examined many of the vessels contained a homogeneous slime-like substance. Here and there among the normally developed vascular bundles were found greatly reduced bundles, consisting of one or two tracheae, and a very few sieve- tubes surrounded by a thick sheath of sclerenchymatous fibres. Often all the woody elements, and even all the cellulose ones, were suppressed, and the bundle consisted of fibres alone (fig. 2, Pl. VII). These reduced bundles are probably branches of the larger bundles, and end blindly in the fundamental tissue, as has been described in the stem of Vanda teres (2). Their function appears to be mechanical only. Above the constriction already noticed, the cotyledon ‘expanded within the seed to form the haustorium. Once within the seed, the cells of the superficial ° layer presented a different appearance. They were isodiametrical, about 0:01- 0:03 mm. in diameter (figs. 12, 13, 14, Pl. X). Their walls were thin, their cytoplasm largely vacuolated, and their nuclei conspicuous. They formed a uniform layer of cells, following the lobed and papillose surface of the haustorium, generally without discontinuities. On the distal surface of the haustorium the superficial cells were somewhat smaller, and their cytoplasm was less vacuolate. They stained more deeply. In this region they presented somewhat the appearance of a columnar epithelium. Enclosed by this layer was the general mass of the haustorium, in the main composed of thin-walled fundamental tissue, built of cells with rounded contours, which were smaller, and closely packed in the outer region. In the inner parts of this tissue were enclosed large intercellular spaces, and the component cells were formed like those of the spongy parenchyma of a leaf. The spaces in the central regions were filled with air, but the smaller ones close to the surface were often infiltrated by a slimy material resembling the débris of the endosperm on the outer surface of the epithelium. This material appears, during the growth of the haustorium, to foree its way between the epidermal cells into the intercellular spaces (fig. 12). In the specimen examined the fundamental tissue was in many places traversed by the hyphae of the non-septate mycelium of a fungus, which had developed subsequently to the opening of the fruit, or possibly had established itself at some earlier stage. These hyphae, being very rich in protoplasm and multinucleate, are easily distinguished from the tubular cells which will shortly be described. The cells of the fundamental tissue contained many starch grains, and smaller starch grains were occasionally found in the superficial layer. The cells ‘of the inner fundamental tissue are comparatively large, the dimensions of their cylindrical branches being about 0:05 mm. X 0:10 mm. They are often constricted where they come into contact with one another, and the partition walls have wide pits. Here and there in this large-celled parenchyma are to be seen giant cells, having about the same diameter, but a much greater length, so that in one section they may show a length of 0:80 mm. or more (figs. 3 and 4, Pl. VII). They are tolerably straight, and seldom branch. As they pass between the other cells, they make contact with them. The areas of 188 Scientific Proceedings, Royal Dublin Soctety. contact are provided with large pits. These giant tubular cells very often form straight or slightly curved linear series connecting one bundle with another, starting in contact with or very close to the thin-walled sheath. So far as our observations go, they never contain ‘starch, but possess a readily seen proto- plasmic lining, in which is embedded a single nucleus. Often several of these tubular cells run along side by side through the loose cells of the parenchyma. The walls of these adjacent tubes cohere together, and where they cohere are marked with pits. Apparently quite distinct from these giant cells of the fundamental parenchyma, but sometimes associated with them, are narrow tubular elements, which traverse the intercellular spaces of the fundamental tissue. They have a diameter of about 0:004-0:007 mm. and a length of about 0:10-0:15 mm., and form very long linear series originating from the bundles. They pass out from the bundles into the cortical tissues, or penetrate through the inner tissues of the haustorium. Hach member of the series has a nucleus. Figs. 10, 11, Pl. IX, show a sheaf of these elements which has branched in the cortical region, and has sent one branch outwards through the cortex to the epidermis and another into the deeper tissues of the haustorium. The whole length of this tract was about 2mm. Fig. 11 shows the inner end of this tract more highly magnified. When these tubular elements reach the outer tissues of the cortex, where the inter- cellular spaces are injected with the débris of the endosperm, their ends enlarge (figs. 12, 13, Pl. X), and sometimes they push their way between the epidermal cells and come directly in touch with the food supply (fig. 14, Pl. X). Sometimes one finds their ends in intercellular spaces still within the tissue. In this case they have a rounded termination, reminding one of the tip of a root-hair or a rhizoid. The connexion of these tubular elements with the bundles may be seen in a section cutting a bundle at right angles, just at the point where these tubules emerge from it (fic. 5, Pl. VII). Here the tubules may be seen as continuations of the groups composed of two, three, or more small angular elements, which are scattered like small islands in the phloem and in the xylem-parenchyma. Fig. 9 is a tangential section of a bundle in the haustorium, and shows the transverse section of a group of these tubules emerging through the bundle- sheath. Fig. 8 is a transverse section of such a sheaf of tubules as it passes across the fundamental tissue. It also shows the tubules running im an inter- cellular space. The giant cells and the tubules seem often to accompany one another, but they also occur quite separated from one another. The tubular form of both kinds of elements and the distance of the transverse partitions in them from one another suggest that in them diffusion, or possibly some other means of translocation, will take place less hampered than in the isodiametrical parenchyma. From a causal point of view one is tempted to surmise that the exceptional abundance of organic supplies has led to a hypertrophy of two different cell- categories to form rhizoid-like internal organs. On entering the haustorium, most of the bundles bend appropriately to distribute themselves in the layers of fundamental tissue immediately next the surface. In this region they bifurcate and anastomose to form a network, the meshes of which are about 1 mm. or less across. Very often these bundles lie immediately beneath the grooves separating the papillae of the haustorium from one another, and from these bundles many of the tubular cells just described extend into the tissues of the adjacent papillae, and running along immediately beneath the epidermis, emerge at the surface between the epidermal cells (fig. 14, Pl. X). Some bundles do not turn outwards towards the surface, but after Dixon and Batt—Channels of Transport in Seedlings. 189 entering from the petiole strike across the central tissue of the haustorium, and make connexion with the network formed by the others at the distal surface of the haustorium. On leaving the petiole, the bundles soon lose their sclerenchymatous sheath, and are continued as a double strand of xylem and phloem, surrounded by an ill- defined sheath of elongated, prismatic, thin-walled cells (fig. 6, Pl. VIII). On the outside these adjoin the spongy parenchyma. Within the sheath the phloem is composed of sieve-tubes, companion-cells, and bast-parenchyma. The latter occupies the greater part of the phloem. Its cells are comparatively large, having a diameter of 0:02-0:03 mm., and being 0:10-0:20 mm. in length. The sieve-tubes and companion-cells are in small groups, and associated with them are the groups of tubular elements already noted, among the phloem-parenchyma. The whole cross-section of the phloem contains about 120 elements. In the cross-section of the xylem, on the other hand, 12-18 elements are seen. They are spiral and annular tracheae and xylem-parenchyma and one or two groups of the tubules. The ratio of the cross-section of the phloem to that of the xylem in the bundles of the haustorium is about 8:1, which is a marked contrast to the same ratio in the petiole, which is 3-0: 2:5 (ef. figs. 6 and 7, Pl. VIII). The haustorium of the specimen we examined was embedded in the softened tissue of the endosperm. This was in turn surrounded by the outer layers of the endosperm which were still white, and of stony consistency. The inner layers next the haustorium were yellowish and pasty, and could be scooped out with a spoon. Both the soft and hard layers gave a blue reaction with iodine and sulphurie acid, and the soft part reduced Benedict’s solution after standing some days with toluene (ef. 9, 10). Microscopie examination showed that in the pasty substance cell-structure was more or less obliterated. The wall-substance was jellified and the proto- plasm indistinguishable. The hard tissue was still composed of very thick- walled prismatic cells, arranged perpendicularly to the surface of the seed. Their ends were bevelled. In the comparatively small lumen there was visible a distinct lining of cytoplasm, in which was a nucleus. The vacuole contained varying quantities of mucilage. Deep and wide pits extended into the secondary layers of wall-substance to the limitmg membrane of the end and side walls (compare Gardiner’s fig. 20, in 7). The outer diameter of the endosperm cells ranged from 0:09-0:13 mm. The diameter of the lumen was 0:08-0:03 mm. The lengths of the cells varied from 0:40 mm. to 1:20 mm. Phenix dactylifera. The general course of the germination of Phanix dactylifera is well known 8, 11, 12). 2 ees Me about eight or nine bundles in the cross-section of the petiole of the cotyledon. Most of these bifureate before entering the haustorium, so that there are about sixteen at the level of the constriction below the haustorium (fig. 17). On entering the haustorium they diverge from one another, and run along just under its convex surface towards its margin. The haustorium itself is in the early stage button-shaped—oval in outline and slightly concave on its distal aspect. Figs. 16 and 17, Pl. XI, show the distal and proximal surfaces of the haustorium of an embryo just emerging from the seed. Just below the margin the bundles branch again, so as to give rise to 35-40 bundles which turn over the margin, converge, anastomose irregularly, and closely follow the concave surface towards the middle of the haustorium. Thus, close beneath the surface of the haustorium there is a network of vascular bundles, connected by about 16-20 main bundles with the vascular system of the petiole, and ultimately with 190 Scientifie Proceedings, Royal Dublin Society. that of the embryo. No bundles traverse the central regions of the haustorium, as they do in Lodoicea. The tissues of the margin and distal aspect of the haustorium retain their meristematic phase later than those towards the proximal aspect, so that in the earlier stages of germination cell and nuclear divisions are frequently found in the fundamental tissue there, and among these dividing cells the bundles are represented by procambial tracts. From these latter the tracheae are developed earlier than the sieve-tubes. When the haustorium expands in size it becomes bilobed, and the lobes turn over and overarch the depressed central concavity. In this condition the outer surface is near the surface of endosperm, while the central mass of that tissue is embraced by the haustorium, and is in contact with its inner surface. In the middle of the haustorium, between its.inner and outer surfaces, a large inter- cellular space develops. During this enlargement the network of conducting tracts, which at first had irregular meshes formed of sinuous bundles, becomes more regular and the bundles straighten. Meanwhile the surface becomes corrugated and papillose. In the younger stages of germination the epidermis of the haustorium is composed of approximately cubical cells, with dense and finely granular contents, which almost entirely fill their cavities. A large nucleus is visible. They present the appearance of the secreting cells of a columnar epithelium (fig. 19, Pl. XI, ep.). As the haustorium enlarges, their vacuoles grow and their protoplasm becomes more scanty, and loses its affinity for stains. At the same time the nucleus diminishes in size, and its granules become less conspicuous (ef. 10). The bundles in the haustorium lie extremely close to its surface, so that their outer elements are often separated from the epidermis by one or two layers only of cells. As in the ease of Lodoicea, the ratio of the area of the ecross- section of the phloem in the haustorium to that of the xylem is much greater than that in the petiole. In this case the ratio is about 5:1 in the haustorium and 1:1 in the petiole (figs. 18 and 19, Pl. XI). There are about twenty elements in the phloem both of the haustorium and of the petiole; about ten of these are sieve-tubes. In the xylem there are about four in the haustorium and twenty in the petiole. In the haustorium the sieve-tubes are comparatively wide (0:01-0:003 mm. in diam.), and their walls stain deeply with haematoxylin. Their length is 0:06- 0:09 mm. They oceupy the outer part of the phloem strand, and are separated from the xylem by several layers of cambiform cells. The tracheae of the bundles are few, and have a diameter of 0:03-0:01 mm. Phenix canariensis and P. silvestris. Later stages of germination were observed in these two palms. In the example of the first-named species which was examined there were 8--10 bundles in the transverse section of the petiole of the cotyledon. In that of P. silvestris there were only 6. Between some of these bundles there were reduced bundles (2 or 3 in all), consisting of fibres only, like those found in Lodoicea sechellarum. These died out before entering the haustorium. As in Lodoicea and in P. dacty- lifera, the strong sheath which accompanies the bundles in the petiole ceases immediately as the bundle enters the haustorium. In none of the three species of Phcenix did bundles cross the central tissue of the haustorium; as they enter it they diverge and bifureate, keeping close to its surface. Then turning over its margin and arriving on its concave surface, they converge and re-unite to form a network on its distal aspect (fig. 15, Pl. X). Dixon anp Batt— Channels of Transport in Seedlings. 191 The bundles after entering the haustorium consist of 8--10 tracheae and a few wood-parenchyma cells, forming their xylem; their phloem is composed of large phloem-parenchyma cells, which separate the two or three groups of sieve-tubes. pees this phloem and the xylem there are two or three layers of eambiform cells. Epidermal cells of haustorium, diameter ... 0-:01--0:06 mm. Sieve-tubes, diameter fap sath ... 0:01--0:02 mm. Sieve-tubes, length nee as ... 0:10--0:15 mm. Tracheae, diameter ae Ba ... 0-01--0:03 mm. Vicia faba. With a view to throwing some further light on this subject, the anatomy of the cotyledons of the broad bean, Vicia faba, was also studied. In the plant, as is well known, the cotyledons do not emerge from the seed-coat, but function solely as storage organs; and, being normally below the ground, they do not transpire. Sections of a seed which had been soaked in water for twenty-four hours were first examined. The bulk of the cotyledons is composed of large parenchymatous cells, which are closely packed with starch and protein granules. Between these cells there is a considerable development of intercellular spaces. Where the neighbouring cells come in contact with one another there are numerous large pits in the intervening walls, which possibly facilitate diffusion from cell to cell. The cotyledons are traversed by vascular bundles, which even at this stage show fully developed vessels with spiral thickenings; but there is no trace of sieve- tubes, the remainder of the bundle being composed of elongated parenchymatous cells filled with protoplasm and containing large nuclei. This early development of wood in the cotyledons may be contrasted with the conditions obtaining in the radicle and plumule, in which the presence of xylem cannot be recognized until some time after germination. At a later stage, when the main root has reached a length of about 12 em., the wood in the cotyledons shows a considerable increase in amount, and well- developed pitted vessels are present. The exact time at which the sieve-tubes are differentiated in the cotyledons is difficult to determine. At this stage they are certainly present in the petioles, and at a later stage can also be seen in the laminae of the cotyledons. The mature sieve-tubes can be easily recognized by the presence of the peculiar ellipsoidal slime-masses in the vacuole of the cell. These bodies were deseribed by Strasburger (13) and Beccarini (1) in members of the Leguminosae, and are particularly well-developed in Vicia. They appear to be quite free in the sieve-tube, and in most cases they lie close to one of the sieve-plates. The anatomy of these cotyledons, in so far as it has a bearing on the problem of translocation, may be briefly considered. The storage tissue through which the soluble organic materials must pass by diffusion is constructed so as to facilitate this function as far as possible. The vascular bundles, on the other hand, probably provide the conduit for transport to the growing points. The early formation and later development of the xylem in the cotyledons, where transpiration is negligible, is significant. It may, of course, be considered that the xylem is a purely vestigial structure, but its extremely marked development hardly seems in accordance with this view. If, however, it is functional, it would seem that it must be of use in translocation. We know that organic substances must at times be able to pass with comparative freedom from cell to 192 Scientific Proceedings, Royal Dublin Society. cell, and therefore must pass through the protoplasmic lining of the eell. Con- sequently, unless we regard the cells lying next to the vessels as being relatively more impermeable, there seems to be no difficulty in understanding how organie substances can be injected into the vessels. On the other hand, the living cells of the vascular bundles seem ill-adapted to the task of translocation. The sieve-tubes are late in development, and even when fully formed would offer a greater obstruction to diffusion than would the storage tissue, which, in addition to a large cross-sectional area, is better provided with pitted cell-walls. With a view to obtaining some evidence on the normal course of the current in the xylem of the cotyledons, the following experiments were conducted :— A seedling of Vicia faba, with a root 55 em. long, was used. The cotyledons were cut across in a direction parallel to the direction of the root, and the cut surface was placed under a solution of eosin. The root was surrounded with moist Sphagnum moss. After twenty-four hours all the bundles of the cotyledons were injected with eosin, and traces of eosin were found in the root down to about 1:5 em. below the origin of the plumule. During this period the root had elongated 0:5 em. In another experiment a seedling with a shoot 12 em. and a main root about 20 em. long was used. The root was cut across and immersed in an eosin solution for one and a half hours, and at the end of that time the whole of the wood in the cotyledons was deeply injected with eosin. Such experiments, unfortunately, give no indication of the normal course of the current in the wood. They do show, however, that if water or a solution of eosin can be drawn into the plant at any point it will be transported with an equal facility in an upward or downward direction, according to requirements. That the cotyledons have the power of absorbing water is shown by the rapidity with which the dry seeds swell wp when immersed in water. It seems probable, therefore, that water continually passes in through the cotyledons during germination; and, unless the cells lying next to the vessels are specially impermeable, a solution of organie substances would travel by means of the wood from the cotyledons in the direction of the growing points. SUMMARY. 1. In the seedlings of the palms examined there is a network of vascular bundles close below the absorbent surface of the haustorium, which is embedded in the endosperm. This network is connected with the growing embryo by bundles which traverse the basal parts of the haustorium and the petiole longitudinally. 2. Well-developed xylem, consisting largely of lignified tracheae, is found in the vascular bundles of organs connecting embryos with their stores. 3. Selerenchymatous sheaths and cords of sclerenchymatous fibres are also often found in these organs. These sclerenchymatous elements are not continued into the haustoria. 4. The total area of the cross-section of the phloem of all the bundles in the haustoria is much greater than that of the bundles in the connecting organs or petioles. There is not the same disproportionality in the cross-section of the xylem in the two organs. Sometimes the total area of the cross-section of the xylem of the bundles of the petiole approximately equals that of the bundles of the haustorium, e.g., in Phenix dactylifera. 5. In Vicia faba tracheal elements are differentiated in the petiole much earlier than are sieve-tubes. Before the latter appear, considerable transport of organic substances to the embryo must take place. Dixon anp Batt—Channels of Transport in Seedlings. 193 6. Material produced by the disintegrated endosperm finds its way into the outer intercellular spaces of the haustorium of Lodoicea sechellarum, passing between the epidermal cells. i 7. Tubular cells taking their origin at or near the sheath of the vascular bundles of the haustorium of Lodoicea sechellarum pass between the isodia- metrical cells of its spongy tissue. They form connexions with these cells, and their walls are pitted at the points of contact. They often form connexions between the sheaths of the vacsular bundles. 8. Groups of narrow tubular elements are found in the phloem and xylem of the bundles of the haustorium of Lodoicea. Here and there sheaves of these pass out of the bundles into the intercellular spaces of the surrounding tissues. Sometimes they turn inwards and push their way among the cells of the central fundamental tissue; more frequently they turn outwards and traverse the cortical region. When they come immediately beneath the epidermis, their ends expand, and may even push out between the epidermal cells to the surface. CONCLUSIONS. 1. The presence of the vascular network in haustoria and of conducting tracts in the transmitting organs suggests that the vascular bundles are the channels by which the embryo receives the organic supplies from its storage organs. 2. The development of tracheal tubes in transmitting organs of seedlings, where the transport of water is unnecessary, is in conformity with the view that these tubes convey organic store material. 3. Experiment shows that these tubes will convey fluid either in a basal or in a distal direction, according to the position of the source and sink. 4. The lateness in the differentiation of the sieve-tubes in the vascular strands of the petioles of the cotyledons of Vicia faba shows that much transport of organic substance down the petiole is effected without their assistance. During this period tracheae are available. 5. Transport of organic substance through the parenchymatous cells of the haustoria and through those of storage cotyledons has to be attributed to the permeability of the protoplasm of these cells. It seems gratuitous to assume that the cells adjacent to the tracheae are the only cells in these organs which are impermeable. If they are permeable, the tension set up in the tracheae will secure movement of organic fluids into and along the tracheae. 6. The great development of the phloem in the haustoria compared with its mass in the transmitting organ—the petiole—suggests that this tissue is chiefly concerned with the preparation of the organic substance absorbed and its trans- mission into the tracheae. 7. It may be pointed out that sieve-tubes in their mature state are by no means ideally constructed tor the transmission of organic substances. The pores in their sieve-plates are exceedingly fine, and they are mostly or entirely blocked by protoplasm or callus. Bearing in mind the semi-permeable nature of protoplasm, it would appear that the sieve-plate would present a greater obstacle to the flow of most solutions and every sol than would a simple ecell-wall. It seems quite possible that we should regard the sieve-tubes rather as minute reservoirs than as conduits. 8. The contrast of appearance of the phloem-parenchyma (cambiform cells) in this connexion is suggestive. The bundle endings in leaves and in growing points, as is well known, are without well-differentiated sieve-tubes, and their phloem is composed of cambiform cells only. These may be supposed to be effective in transmitting the products of photosynthesis into the tracheae and at the growing points to extract these supplies for the growth there. SCIENT. PROC. R.D.S., VOL. Xvi, NO. 21. Qu 194 Scientifie Proceedings, Royal Dublin Society. 9. The companion-cells of the phloem are completely filled with finely granular and easily staining protoplasm. Their nuclei are comparatively large. Their appearance is that of secreting cells. Hence we may provisionally assume that to them is assigned the function of secreting substances (probably enzymes) to prepare the organic materials and render them suitable for introduction into the tracheae and for transmission in these tubes. 10. These secretions may be stored in the sieve-tubes, whose sieve-plates prevent their passage along the bundle. The sieve-plates may be regarded as partitions which allow the protoplasm to be withdrawn through their pores, but at the same time have been rendered practically impermeable by the remnant of protoplasm and the callus left blocking the pores. REFERENCES. 1. Beccarint, P.—Intorno ad una particolarita dei vasi cribrosi nelle Papilionacee. Malpighia, vol. vi, 1892, p. 53. i) Drxon, Henry H.—On the Vegetative Organs of Vanda teres. Proce. Roy. Trish Acad., Ser. 3, vol. iti, No. 3, 1894. 3. Drxon, Henry H.—The Transport of Organic Substances in Plants. Pre- sidential Address, Section K, British Association, 1922. 4. Drxon, Henry H., and Bain, Nice, G.—Transport of Organie Substances in Plants. Nature, Feb. 23, 1922. Farmer, J. B.—Quoted by Sir W. T. Thiselton-Dyer (14). Hooker, Sir Wint1am H.—Quoted by Sir W. T. Thiselton-Dyer (14). GARDINER, W.—On the Continuity of Protoplasm through the Walls of Vegetable Cells. Phil. Trans. R.S., vol. 174, 1893. Luoyp, Fras. H.—Development and Nutrition of the Embryo, Seed, and Carpel in the Date, Phenix dactylifera L. Missouri Botanic Garden, 21st Annual Report, 1910, p. 103. 9. Ponp, R. H.—Ineapacity of the Date Endosperm for Self-digestion. Ann. of Bot., xx, 1906, p. 61. 10. Reepv, H. S—A Study of the Enzyme Secreting Cells in the Seedlings of Zea Mays and Phema dactylfera. Ann. of Bot., xviii, 1904, p. 267. lil Sacus, J—Zur Keimunesgeschichte der Dattel. Bot. Ztg., No. 31, p. 241. Ne gs co 12. Sarcant, Miss E.—-A Theory of the Origin of Monocotyledons founded on the Structure of their Seedlings. Ann. of Bot., xvii, 1903, p. 1. 13. SrraspurGeR, E.—Histologische Beitrage, Hft. III. Ueber den Leitungs- bahnen in den Pflanzen, p. 193. 14. TuIsELTON-DyeER, Sir W. T.—Germination of Double Cocoa-Nut. Ann. of - Bot., xxiv, 1910, p. 223. Dixon and Ratt—Channels of Transport in Seedlings. 195 DESCRIPTION OF FIGURES. Puate VII. Fig. 1. Germination of Lodoicea sechellarum, photographed at Kew, January 19th, 1923. The curved petiole of the cotyledon is bent down in front of the bilobed fruit. The split in the base of the petiole shows the plumule about to emerge. 2. Lodoicea sechellarum. Transverse section of sclerenchymatous strand in petiole of cotyledon. %X 509. s., sac containing siliceous nodule; t.w., thin-walled elements. 3. Lodoicea sechellarum. Transverse section of haustorium. X 50. v.b., vascular bundle; f.p., fundamental parenchyma; g.c., giant-cell. 4. Lodoicea sechellarum. Transverse section of haustorium.. X 150. Showing giant-cell and its attachments; a, with parenchyma-cells. 0. Lodoicea sechellarwm. Transverse section of vascular bundle in haustorium at the level where some tubules are emerging; is., intercellular space; p., phloem; t., tubules; x., xylem. Puate VIII. 6. Lodoicea sechellarum. Transverse section of vascular bundle close to surface of the haustorium. X 420. x., xylem; t., tubules; s.c., sieve-tubes and companion-eells; p.p., phloem-parenchyma cell; is., intercellular space filled with slime. 7. Lodoicea sechellarum. Transverse section of vascular bundle in petiole of cotyledon. X 300. «@s., sac containing siliceous nodule; t.s., trachea con- taining slime; p., phloem; x., xylem; s.s., sclerenchymatous sheath. Puate IX. 8. Lodoicea sechellarum. Transverse section of haustorium, showing tubules (t.) passing down intercellular space (is.). 250. Longitudinal tangential section of bundle sheath 9. Lodoicea sechellarum. x 250. in haustorium, showing sheaf of tubules (t.) cut transversely. 10. Lodoicea sechellarum. ‘Transverse section of haustorium. X 48. e.n., remains of endosperm; ep., epidermis; v.b., vascular bundle; t., sheaf of tubules. 11. Lodoicea sechellarum. 4 presented in fig. 10, showing its bifurcation. x 250. Inner part of the same sheaf of tubules (t.) re- is., intercellular space. 196 Scientific Proceedings, Royal Dublin Society. GAT Ne Fie. 12. 14. 15. 16. 18. iY), Lodoicea sechellarum. Transverse section of haustorium. en., remains: of endosperm; ep., epidermis; is., intercellular space filled with slime from endosperm; t., tubule in intercellular space. > 33). Lodoicea sechellarum. Transverse section of haustorium. en., remains of endosperm; t., enlarged end of tubule filling intercellular space under epidermis (ep.). Lodoicea sechellarum. Transverse section of haustorium. en., remains of endosperm; t., end of tubule filling sub-epidermal intercellular space and emerging between two epidermal cells (ep.). X 550. Phenix, silvestris. Wongitudinal section of more mature haustorium and petiole (p.). v.b., vascular bundles; is., intercellular space. % 7. Puate XT. Phenix dactylifera. Distal surface of young haustorium, showing de- veloping vascular bundles (v.b.) through the outer tissues. X 24. . Phenix dactylifera. Proximal surface of young haustorium and transverse section of petiole. v’.b’., vascular bundles in section, and v.b., through the outer tissues. X 24. Phenix dactylifera. Transverse section of vascular bundle in petiole of young cotyledon. s.s., sclerenchymatous sheath; p., phloem; x., xylem. xX 380. Phenix dactylifera. Transverse section in young haustorium. ep., epi- dermis; p., phloem; x., xylem. X 380. F SCIENT. PROC. R. DUBLIN SOC., N.S., VOL. XVII. PLATE VII. Fie. 1. 0-~-BLZO LORS Dixon AND BALL. T1Ivq GNV NOXICG w- TWA ALVId TIAX “OA “S'N “OOS NITANG ‘A ‘OOWd “LNAIOS SCIENT. PROC. R. DUBLIN SOC., N.S., VOL. XVII. PLATE IX. Fie. 10. Fig. 11. DIxon AND BALL. ‘TIVg GNV NOXIG FI STW “X LV 1d ‘TIAX “IOA “S'N “DOS NIT@NG “A ‘IO0Udd “LNYIOS SCIENT. PROC. R. DUBLIN SOC., N.S., VOL. XVII. PLATE XI. ext (2) Derg Za} Way 2 pe he aia: Dixon AND BALL. No. 22. IRREGULARITIES IN THE RATE OF SOLUTION OF OXYGEN BY WATER. By H. G. BECKER, A.R.C.Sc.1., A.L.C., Demonstrator m Chemistry in the College of Science, Dublin; AND H. F. PEARSON, A.R.C.Se.1., Research Student. (Read Marcn 27, Printed Junr 19, 1923.) In the course of some previous work one of the authors noticed that when experiments on the absorption of gases by water were made by exposing columns of air-free water to the atmosphere and then analysing the gas-content of these columns after different periods of time, various small irregularities appeared in the results, which were greater than the experimental errors involved. The effect indicated seemed to show that instead of the process of absorption of the air by the water being steady and uniform, as it is when the water is gently mixed, it varied suddenly at different times, especially towards the saturation point. This effect, though too small to be of importance on the large scale, was considered worth investigating further, with a view to elucidating the process by which a soluble gas is absorbed and distributed in a liquid. To do this it was necessary to use a method which allowed of observing the amount of gas absorbed by the liquid at any time during the whole course of the aeration. At the same time it was thought desirable to eliminate temperature variation, and also the variation in the humidity of the gas in contact with the liquid, so that the absorption of the gas would take place under conditions which were as constant as possible, and any effect due to these causes would be excluded. The apparatus shown in fig. 1 was designed for the purpose. It consists of two glass bulbs, V, and V»2, about 30 mm. diameter, and of a capacity of 60 and 200 ¢.c. respectively. The bulb V, is provided with a capillary tap (A), and is fused on to one end of the water-manometer M. The other end of the manometer is fused to a four-way piece on the top of the bulb V2. The remaining two branches of the four-way piece are fused to a capillary and tap (H) and a gas- reservoir R. The amount of gas absorbed at any time was shown by the reading of the manometer, and this could be re-set to zero at any stage of the process by allowing mercury to flow into the reservoir to replace the gas which had been absorbed. The whole apparatus was enclosed in a water-jacket 2” in diameter, through which a stream of water from a thermostat, kept constant to a tenth of a degree, was circulated by means of a pump. SCIENT. PROG. R.D.S., VOL. XVII, NO. 22. 2N 198 Scientific Proceedings, Royal Dublin Society. Previous to an experiment the three taps (H), (A), and (KX) were connected by means of a three-way piece, and the whole apparatus thus exhausted by a good filter-pump. The manometer was kept at zero i “A during this process by manipulating the taps (H) | and (A). When the apparatus was completely i exhausted, oxygen from a gas-holder was allowed to iJ, enter through the same connexions until the pressure | was again atmospheric. These operations were re- Ale peated three times in succession in order to ensure that no appreciable amount of air would remain in the apparatus. The oxygen used was taken from a cylinder of the gas, and was filtered through a tube containing glass-wool to retain traces of dust, and then passed through a small flask immersed in»the thermostat, and containing a little water, in order to saturate the gas with water-vapour at the working temperature (25:8° C.). When the apparatus was full of oxygen, the air- free water (which had been prepared by boiling dis- tilled water in an apparatus, previously described, and stored in a bulb of about 250 cc. capacity over mereury) was allowed to enter. The bulb containing the water was attached to the tap (B) and the apparatus exhausted; on opening the tap, the water was displaced over and allowed to rise to the 60 e.c. mark. The oxygen was then allowed to enter until the pressure was atmospheric, when the taps (H) an@ _. (A) were opened simultaneously and closed after about . two seconds, thus rendering the pressure in both bulbs exactly the same, and the observations of the mano- meter were then started. The barometer was read at the time of starting an experiment, and the volume of gas absorbed calculated by means of the formula given in a previous paper. It was originally intended to make a very complete study of the phenomenon by means of a large number of experiments over a long period of time, but the occupation of the College of Science for purposes other than scientific interfered with this scheme, and finally interrupted the work completely. As it will not be possible to resume the work for some time, the results so far obtained are presented in this paper, since they are of interest in that they clear up some points which were previously in doubt. The results of six experiments are given in the table, and the variation of gas-content with time is shown in fig. 2. The points indicated by dots were obtained with an apparatus which was supported rigidly on a slate wall-bracket (such as is used for balances), and braced in every direction with wire stays to eliminate vibration as completely as possible. The points indicated by crosses were obtained with an apparatus supported in the ordinary way by a retort- stand. Becker AND Psarson—Irregularities in Solution of Oxygen by Water. 199 TABLE OF RESULTS. Time. °/, Sat. ‘Time. Gg | Time. °/, Sat. EXPERIMENT 2a. ExpermMEnt 10a. | EXPERIMENT 18a. 55 10-0 1-5 OH) | 1:0 2-5 22:5 34-0 3-0 59 | - 18-75 30-0 24-5 36-0 19°75 Bye) | 20-5 32:7 26:0 39-2 23:0 34:8 | 21:5 34:3 EXPERIMENT 3a. 25-0 37:6 | 23:25 36:0 1-6 2-2 28-0 42:0 | 43-0 64-0 2-0 3-0 43°75 57-0 | 44-0 64-2 3:0 5°5 45°75 58°8 | Experiment 110. 4:25 7-3 47-75 600 | 16-0 40-8 5-0: 8-4 51-25 620 | 18-0 41-7 29-0 85:5 67-75 S00 mio: 44-7 Experiment 12h, 68:75 82:0 | 29-0 48:0 17-5 38:6 74-0 S70 23-6 50°7 22-5 47-0 91-75 Oise yee. 40:0 724 24-4 49-0 96-5 96-0 44-75 73:3 425 68°8 99°5 96°5 | 47-5 75:0 45°5 724 115-75 98:0) | 66°75 80:8 66-5 79-8 118-0 98-6 112°5 86-0 The curves drawn on the graph are reference curves, calculated by means of the formula w+=100(1 - ¢*), which has been shown to apply when the water is mixed, however gently. The value of the coefficient ‘‘b’’ for the upper curve is 0:028, and that for the lower curve is 0:018, and these values represent the upper and Jower limits of the rate of solution under these conditions. It will be seen that up to a value of about 60 to 70 per cent. of saturation the experimental values agree well with the logarithmic curves, but beyond that the divergences are wide. In the case of the upper curve this is shown by a marked falling-off in the rate of solution, while in the case of the lower curve a marked increase followed by a falling-off is recorded. The fact that the process follows the logarithmic curve during the early stages of the absorption would seem to indicate that the water is kept slowly but steadily mixed during this stage, while the uncertain behaviour after this 200 Scientific Proceedings, Royal Dublin Society. shows that the force causing the mixing has become so extremely small as to be capricious in its action. ; ae 80 ee fe) Lo} ration 5) £ [o} Percentage of Satu n oO 20 40 60 80 100 120 Time in Hours Fic. 2. In considering the possible agencies which could affect the process under the experimental conditions, it must be remembered that these were kept as constant as possible. The mass of gas used was completely enclosed, and both gas and water were maintained continually at the same temperature. The gas was given the correct humidity for the working temperature before admission to the apparatus, and therefore evaporation from the surface layers of the water was prevented. The possibility of slight convection currents, due to cooling of the surface layers, which might occur when the water is exposed to the atmosphere, is eliminated. All the more obvious causes which might produce mixing have thus been excluded in these experiments, and the fact that it is present shows that the cause must be more fundamental in its nature, and further careful experiments will be necessary to discover its precise nature. Whatever the cause, these experiments clearly show that pure distilled water exposed to oxygen at a uniform steady temperature absorbs the gas in such a way as to indicate that very slow mixing of the water occurs even under these conditions during the early stages of the absorption, but that towards the saturation point this mixing tends to become uncertain in its action. CHEMICAL LABORATORY, COLLEGE OF SCIENCE, DUBLIN. f 20n | No. 238. THE HYDROGEN ION CONCENTRATION OF THE SOIL IN RELATION TO THE FLOWER COLOUR OF IYDRANGEA HORTENSIS W., AND THE AVAILABILITY OF IRON. By We BG: ATKINS, OBE. ScD, HEC (Read Marcu 27. Printed June 18, 1923. Flower Colour and Seit Reaction. When studying the relation between the distribution of plants and the hydrogen ion concentration of the soil (Atkins, 1922, 1), it was noticed that pink hydrangeas were found on alkaline soil and blue on acid. A number of situations were examined to test the validity of this observation; the results are given in the following table :— Locality. Colour of hydrangea. | pH of soil. | Notes. | | Cornwood, Devon. . | Blue. : 6 3 6:0 | Several plants. ue Re . | Pink. : : 3 76 | One plant, blue when in | | previous site. Plymouth Hoe. . | Pink. : : -| 8 | Several plants, various sites. Falmouth, near sea. | Pink. . 6 | 7:5 | One plant. Dublin. : 2 || Libmlk : : : | 8:0 | One plant. Mt. Edgcumbe, Corn- Blue. 4 5 -| 5°75 | Numerous plants, blue. wall. “ 5 . | Pink and blue. - | 5:9 | One plant. Antony, Cornwall. . | Pink; someblue,and | 7:3 | Two plants. intermediate. | Templemichael, Co. | Blue, pink,andinter- | 6-2 Plant, when bought and Waterford. mediate. | planted, had bright pink flowers only. There is in these examples undoubted proof of the relation between soil reaction and flower colour. They include the case of a plant originally bearing blue flowers changing to pink when in an alkaline habitat, as at Cornwood, and the converse change, pink to blue and some pinkish, at Templemichael. Some plants bear blue and pink fiowers; these are found in situations with a reaction of pH 75 and under. It must be added that the soil tested was from SCIENT. PROG. R.D.S., VOL. XVII, NO. 23. 20 202 Scientific Proceedings, Royal Dublin Society. the first four inches, which is insufficient to get near the absorbent portions of the roots. Sap-soluble Pigments as Possible Indicators. Since the sap-soluble colouring matters:of many plants act as indicators, as shown by Haas (1916), it seemed possible that the diverse colours in hydrangeas were due to varyine degrees of acidity, the more so as it is known that certain colour changes from red to blue, according as the flower fades, are explained correctly by this hypothesis. As against this must be set the remarkable con-. staney in pH value given by the leaves, stems, and roots respectively of members of the same species. To test the matter directly, pink and blue flowers were treated with dilute acetic acid, but the pink did not change to blue. In fact, no very marked change was noticed in either case. Furthermore, flowers were obtained from a single plant bearing both pink and light mauve or blue. Petals were crushed with an agate pestle, and two drops of water, and one of indicator, added to each. Using brom phenol blue as indicator, both appeared to be close to pH 40, and were indistinguishable. This indicator is, however, dichroic, so an exact com- parison in a turbid drop is not easy to make. Methyl orange covers a somewhat similar range, and with it both the pink and blue petals were ascertained to be at pH 42, and were, as before, indistinguishable. It is accordingly clear that the colours of hydrangea flowers are not due to the natural pigment acting as an indicator. Flower Colour and Availability of Iron Salts. Among gardeners the practice of potting with iron nails is well known as a means of producing blue hydrangeas. It therefore seemed probable that the solubility of iron salts might be the direct cause of the production of the blue. Attempts were made to induce cut flowers to change from pink to blue by placing the stalks in dilute solutions of ferrous and ferric salts. A deep dark green appeared in the stems, and spread slowly into the petioles and veins of the petals. The flowers then withered. Possibly with iron salts in much smaller amount a blue might have resulted. Culture experiments by Dugegar (1920) have shown that certain salts ordinarily considered as insoluble are quite effective as plant nutrients, since the solids yield a continual supply of a minute amount in solution. There is, however, a limit to the availability of ‘‘insoluble’’ compounds, as shown by McCall and Haag (1921) to be the case with ferric salts. These workers found that culture solutions containing ferric salts were adequate for nutrition of wheat when the reaction was pH 40 or more acid, but solutions from pH 40 to pH 70 gave rise to chlorosis in the plants grown in them. Patten and Mains (1920) have shown that ferric hydroxide is precipitated in quantity between pH 35 and pH 60, at which the process is complete. In nature, however, few, if any, soils are as acid as pH 35; yet plants grow im natural solutions, according to species, up to pH 8 or 9. As previously suggested by the writer (1922, 1), this is probably due to the presence of iron in the ferrous condition. Further work (1922, 2) has shown that ferrous hydroxide does not begin to be precipitated until pH 51; and though it comes down in quantity at pH 55 to pH 65, yet even beyond pH 71 a small amount exists in the solution, and is slowly precipitated as it becomes oxidised to the ferric state. It thus appears that under the reducing conditions met with in soils, especially perhaps in badly aerated acid soils, in which the colloids are not aggregated as in the presence of calcium bicarbonate, iron may be readily available at pH 6, and even, though in Arxins—Hydrogen Ion Concentration of the Soil. 208 greatly diminished amount, at pH 7 to 8. It accordingly seemed probable that the blue colour of the flower stalks and petals, shown in habitats of the more acid type, was due to the presence of traces of ferrous salts not required in the general metabolism of the leaves. The ferrous salt in excess mav then react with the natural anthocyan pigment, which is pink, giving the blue colour which is observed. This conclusion, as it has since been found by the writer, was previously arrived at by Molisch (1897), whose work is discussed in a following section. On this view the hydrangea flower may be taken as an indicator of the availability of iron, and indirectly of the soil reaction. Such variations of flower colour with habitat do not appear to be common. An interesting example has been examined by Boresch (1920), who found that the Cyanophycean Phormidium Retzti var. nigro-violacea has normally an olive-green colour; this, however, becomes violet or yellowish-brown in cultures deficient in iron salts, addition of which restores the usual colour. Qualitative and Quantitative Tests for Iron in the Hydrangea Flower. The facts and considerations of the preceding pages rendered it of interest to see whether it was possible by direct tests to demonstrate the presence of a greater amount of iron in blue hydrangea flowers than in pink. For this purpose the delicate hematoxylin test of Macallum (1897) was tried. Dried flowers (no fresh were available) were boiled in twice distilled water containing a dilute solution of well-washed hematoxylin crystals. Such a solution is a pale yellow, and iron, where present, acts as a mordant for the stain. Pink flowers and others, blue to mauve, all from the same plant, were used. Control tests without. hematoxylin showed that the blue flowers always appeared slightly darker than the pink long after their natural colour had been extracted. With hematoxylin there was, however, a very definite darkening, amounting in parts to the development of the typical blue purple of iron hematoxylin, but in the blue or bluish flowers only. No trace of the blue purple was ever seen in the pink flowers, though some possibly showed an almost imperceptible darkening. Since this test is only given by iron in inorganie, or in at least ionisable condition, the absence of such iron in the pink hydrangea and its presence in the blue may be taken as demonstrated. It seemed that quantitative results were desirable, so flowers, including the blue or pink portions of the stalks, were dried and incinerated. The almost white ash was then dissolved in hydrochlorie acid, and after oxidation with nitric acid was treated as usual for the precipitation of ferric hydroxide. None could be seen, however, so the white precipitates were redissolved, made up to the same concentration with respect to ash, and the traces of iron estimated colorimetrically with ammonium thiocyanate. Standard N/10,000 ferric iron was diluted, and it was seen then that the solution from the blue hydrangea ash lay near a 0:5 dilution, viz., 2°8 milligrams of iron per litre, whereas that from the pink corresponded to 0:3 dilution, or 1:7 milligrams. Since the volume of the blue hydrangea solution was approximately one-eighth of a litre, the iron in the ash was only about 0:35 mgrm., an amount too small for gravimetric analysis. The ash of the blue hydrangea flowers was found to be 5-9 per cent. of the material taken after drying at 95° and over sulphuric acid. It contained 0:24 per cent. of iron. The pink hydrangea gave 43 per cent. of ash, of which iron amounted only to three-fifths as much as in the blue, namely, 0-14 per cent. It seems probable that both types of flower have a certain amount of iron in the form of complex organic substances, but in the blue only is there an excess available*to combine with the natural sap-soluble pigment, 204 Seventific Proceedings, Royal Dublin Society. Calculated on the dry weights of the flowers used, the foregoing analyses give approximately 140 parts of iron per million for the blue flowers and 60 p.p.m. for the pink, when the difference in ash-content is taken into consideration. The analyses of Maquenne and Cerighelli (1921) show that the iron in plant tissues varies from about 20 to 150, rising exceptionally to much higher values, such as 362 p.p.m.; thus the quantities found for hydrangea flowers are quite normal. These workers draw attention to the fact that iron accumulates in tissues such as bark, leaves, ete., as they grow old. Such a difference, however, ean hardly be held to explain the varying iron-content of two flowers in full bloom. Moreover, were this the cause, flowers which open pink should later on become blue, but such a change dees not oecur—the pink remain pink and the blue remain blue, though the intensity of colour may change somewhat. Further- more, the evidence of the hematoxylin test is against it, as young blue flowers often give the reaction intensely. ; Correlation of Results with those of Previous Workers. It had long been recognised that certain soils possessed the property of producing a blue colour in the hydrangea; and Charles Darwin recorded that alum influenced the flower colour. Molisch grew no less than four hundred hydrangeas in various soil mixtures with or without added salts and metallic iron and iron oxides. Since the flowers were normally all red, it is evident that his standard earth must have been alkaline. The addition of iron and iron oxides failed to induce any alteration in colour. This is in keeping with the results for alkaline soils. In peaty soils the plants grew excellently, and produced blue flowers. Various metallic salts were tried, and either had no effect or were poisonous to the plant. Ferrous sulphate, however, led to the production of blue flowers, as did also alum and aluminium sulphate. The action of these salts is to increase the acidity of the soil, for quite dilute ferrous sulphate solution may be as acid as pH 48. Pure aluminium sulphate in dilute solution gives an acidity of about pH 4, the value for alum being almost identical, as previously mentioned. Since iron in some form is always present in the soil, an increase in soil- acidity renders more iron available for the plant, so the addition of alum or aluminium sulphate increases both the soluble iron and aluminium salts. To which of these, then, is due the production of the blue flowers? As already mentioned, ferric salts are precipitated as hydroxide between pH 3:5 and pH 6.0, ferrous from about pH 5-1 onwards to about pH 8; aluminium, as hydroxide, begins to be precipitated at pH 39 to 42, a precipitate is still obtained at pH 54, but on filtering the solution the filtrate at pH 64 fails to give any further precipitate or trace of turbidity when rendered less acid. The quantity of aluminium salts in solution at pH 6 to 64 must therefore be very minute, though hydrangeas are normally blue at pH 6, tending towards mixed colours at pH 64. These considerations, taken in conjunction with the qualitative and quantitative tests for iron in the flower, render it probable that iron rather than aluminium is the metal that reacts with the pink anthocyanin to give the blue in the growing plant. An experiment performed by Molisch certainly does not leave this conclusion entirely free from doubt. Having failed, as did also the writer, to obtain an extract of the pigment, which quickly becomes decolorized, Molisch found that longitudinal sections of the flower stalks gave a blue colour when treated with alum, aluminium, sulphate, or ferrous sulphate. Since indicator changes in the pigment have been ruled out, the following possibilities present themselves as interpretations. Firstly, that under the Arxins—Hydrogen Ion Concentration of the Soil. 205 conditions of the experiment iron and aluminium form blue salts with the red anthocyanin. Secondly, that the action of the aluminium is in part responsible for the blue colour in nature, though iron also plays a large, or possibly the main, part. Thirdly, that the action of the acidity of the alum and aluminium sulphate was to liberate iron already present either as an insoluble deposit (which in some plants has been shown by Gile and Carrero (1916, 1) to exist) or from organie combination. Variation in colour due to the addition of salts to a number of anthocyanin pigments has been studied by Shibata, Shibata and Kasiwagi (1919). The extracts examined by them do not include that of the hydrangea, yet they state that “‘the colour change of hydrangea and other flowers caused by iron salts and alum... is nothing but the complex formation, as we see with the extracts containing anthocyanins.’ Culture e experiments carried out by Kraemer (1906, 1909) with Hydrangea otaksa, which normally has red flowers, showed that plants grown with alum eave blue flowers, as did also those with aluminium sulphate and calcium hydroxide. In the latter case the excess of the aluminium was probably the active portion of the mixture. Kraemer further discovered that the paldion of potassium carbonate to plants grown in sand resulted in the production of blue flowers. It appears, therefore, that the availability of iron may somehow be increased by a markedly alkaline reaction after passing through a minimum value. It is desirable that such experiments should be repeated, and the pH value of the soil examined at intervals during the growth of the plants. The possibility of an inerease of the absorption of iron in the region of marked alkalinity is in keeping with the results obtained by Arrhenius (1922) for the intake of salts in general. Working with wheat and radish in well-aerated water cultures, he found that, at maximal growth, the intake of the salts is at a minimum. Chlorosis and Availability of Iron Salts. The relation between chlorosis and a deficiency of iron was established as long ago as 1843 by the work of Eusétbe Gris. The disease became of con- siderable economic importance in France through its manifestation in American vines, and numerous researches showed that treatment with ferrous sulphate sprayed on the leaves was beneficial. The subject has been reviewed at length by Roux (1900). Roux’s own experimental cultures are of great interest, and a series of photographs illustrates the growth of nine species in soils containing from 0 to 25 per cent. of calcium carbonate. More recently Tansley (1917) has demonstrated how in ealeareous soils Galium saxatile becomes chlorotic, and either dies or cannot withstand the competition of G. sylvestre, when the two are grown in absence of other plants. Chlorosis has received attention also frem Gile and Carrero (1916), who conclude that iron is not easily translocated from old leaves where it has accumulated. They (1916, 1920), moreover, made an elaborate study of the chlorosis of the rice plant, and showed that the ash of chlorotic rice plants was low in iron; the condition was met with in calcareous soils with a normal amount of water, though not when the same soils were submerged. They suggested that special roots better fitted to absorb iron were developed. An explanation in keeping with present knowledge of the pH values at which ferrous and ferric salts are precipitated is as follows. The water standing over the rice tends to lessen soil aeration, to intensify the reducing action of the soil, and possibly slightly to imerease its hydrogen ion concentration in 206 Scientific Proceedings, Royal Dublin Society. the immediate neighbourhood of the roots, through retention of carbon dioxide in solution; as a consequence, ferrous salts become more available, and chlorosis disappears. It appears, therefore, that chlerosis in certain plants and the development of hydrangeas with pink flowers are closely related phenomena, since both depend on the same factor, the low availability of ferrous salts in the soil. Jamestone soils may have pi values from 82 downwards to pH 7-6, or even less. In these the solubility of ferrous salts are much reduced compared with scils at pH 6 to 7, and so chlorosis may develop. The work of van Alstine (1920) and of Jones and Shive (1921) is of interest in this connexion, ‘and experiments showed that ferric phosphate gave rise to chlorosis at pH 41, though satisfactory growth could be made with ferrous sulphate up to much higher pH values. Further work in this line has been carried out by Arndt (1922). Heretofore the pH values for precipitation of hydroxides have been econ- sidered; but in culture solutions, and even in the soil, phosphates may have to be considered in this connexion. It was mentioned previously that M‘Call and Haag (1921) found their nutrient solutions free from all traces of ferric iron below pH 3:14, yet Patten and Mains (1920) give pH 35 to 60 as the limits for precipitation of ferric iron. The explanation lies in the presence of phosphate in quantity sufficient to precipitate the iron completely. Aluminium, too, has a highly insoluble phosphate, for, according to V. D. Elst (1922), it is precipitated at pH 2:6 to 3-7, the range“for the hydroxide being given as pH 3 to 5. In the writer’s own experience this hydroxide is not completely precipitated till beyond pH 5-4. Conner (1921), too, states that at pH 39 aluminium is more completely precipitated as phosphate than it is at pH 60 as hydroxide. Beyond pH 6-4 the writer could not obtain any further precipitate of aluminium hydroxide. The work of Lipman (1921) on the relation of the soil to chlorosis in citrus trees appears to indicate that the lack of available iron may not always be the cause of chlorosis. Effects of Excess of Soluble Iron Salts im the Soil. There are a number of records of an excess of iron salts being the cause of injury to growing plants, and the subject has been considered experimentally by workers mentioned previously. An effect on the soil itself remains to be considered. Swedish chemists have studied the processes taking place under acid humus soils. The rain-water percolates, and, according to Arrhenius (1922), becomes acidified and dissolves iron compounds, so that in time a bleached earth is found below the humus. The water, as it penetrates, reaches a less acid region, where its iron and aluminium salts are deposited, giving rise to an ‘‘ortsten’’ layer. Since peat gives to aqueous extracts acidity equivalent to pH 46, and the subsoil is far less acid and acts as a buffer, the reason for precipitation is clear. Further, under the reducing conditions of the humus, most of the iron is probably in the ferrous condition. Frosterus (1914), in his study of the soils of Finland, gives illustrations in colour of these changes, as well as analyses of the different layers. The contrasts are very striking. In a schematic presentation of his results Frosterus distinguishes between the red iron ‘‘podsol’’ and the brownish humus ‘‘podsol.’’ The red earth is low in humus, rich in iron and bases, but poor in aluminium. The brown is richer in humus, containing 3 to 11 per cent., considerably more acid, richer also in aluminium. Now, since ferric salts are precipitated as hydroxides from about pH 3:5 to 60, and aluminium salts from pH 3:9 to somewhere above Arkins— Hydrogen Ion Concentration of the Soil. 207 pH 54 and below pH 6-4, one would expect that they would be washed out in much the same manner. Arrhenius, indeed, in describing the process, classes iron and aluminium together. It is interesting, therefore, to find that, in the two types of podsol analysed for Frosterus, the iron and aluminium have to a certain extent been separated. The explanation appears to be that the iron present as a ferrous salt is not precipitated at all till over pH 5, and appreciable amounts are still in solution at pH 7. It is, therefore, carried further into the soil before precipitation is complete. The explanation of iron podsol formation advanced by Arrhenius may be employed to explain the formation of iron pan in this country. The conditions under which it is formed are described by Hall (1910). At about the level to which the soil is ordinarily aerated, a layer of hydrated ferric oxide accumulates in acid clay and sandy soils, which are apt to be water-logged. The conditions prevailing in certain land in the Ballyhoura Mountains, Co. Cork, may be cited as an example. For these details the writer is indebted to Mr. A. C. Forbes: “The soil consists chiefly of glacial drift on Old Red Sandstone. some of the more irregular are aggregates of quadrilaterals. Only two or three were seen with a pentagonal shape, and no other outlines were observed definite enough to be given a name. In a number of cases there are notches in the corners of quadrilaterals, with linear sides approximately parallel to the sides of the quadrilaterals. On the same bedding plane are a specimen of Oldhamia antiqua, the cast of an arched worm burrow about 3 mm. in diameter, another horizontal one about 0:3 mm. in diameter, and a shallow funnel-shaped depression surrounded by a tumid elevation raised about 0:5 mm. above the general level, and 1 mm. above the bottom of the depression. Horizontal and transverse sections were made through the spots. The latter (Pl. XII, figs. 4, 5) showed that the spots are tabular bodies, with a thickness varying little from 0:3 mm. The rock is very fine grained, with well-marked bedding, certain levels being marked by dark lines. The bodies in question le upon. one of the bedding planes so marked, and have their upper surfaces on the plane along which the slab was split. The light green shale is apparently composed chiefly of flakes of sericite and chlorite and minute grains of quartz. The dark bands are due to an increase in chlorite. The bodies we are studying, when examined with the microscope in thin section, are seen to differ from the rest of the rock in that the flakes of sericite and chlorite are, on the whole, larger, the chlorite more abundant, and interstitial quartz more evident. A horizontal section of a reticulate specimen shows a marginal accumulation of chlorite and an increase of colourless constituents forming the spots. In some of the transverse sections there is a light-coloured interior, with, apparently, much quartz, a chloritic layer reuree it above and below, as well as at the sides (Pl. XII, fig. 4 left, and fig. 5 right). With regard to the origin of Thee curious markings, it is not easy to suggest a satisfactory explanation. It seems very unlikely that they are of organic origin. The possibility of their being crushed specimens of primitive ‘eystids was considered. The tabular form, the angular outlines, and the apparent presence of a thick middle layer having a less firm structure than the rest, are points of resemblance. But the general quadrilateral, instead of pentagonal or hexagonal shape, the angular notches, and the complete absence of any characteristic structures, such as pores or marginal folds, seem to rule out this suggestion. The angular nature of the markings appears incompatible with the activities of worms. Turning to inorganic agencies, concretionary action could not produce such sharp angles and straight edges. An inspection of some of the more reeular outlines, with their occasional notches and ‘‘outworks’’ (Pl. XII, fig. 6), stronely suggests crystals. Fig. 7, Pl. XII, shows camera lucida drawings of the shapes of pseudomorphs of halite crystals in Triassic marl. Some of the shapes are remarkably like those we are considering. It is suggested, therefore, that the spots originated as crystals embedded in the sediment. These subsequently became dissolved away, and the cavities filled by mud. If the erystals had been those of an iron-bearing mineral, the pseudomorphs might have contained a residue of iron, especially along their boundaries, which would help to account for the greater abundance of chlorite in the spots, and its segregation at the edges in some of them. It seems unnecessary to assume that the original crystals were tabular, since pseudomorphs in mud of equidimensional crystals would presumably be squeezed flat by the compression of the sediment into shale. This would also SCIENT, PROC. R. DUBLIN SOC., VOL. XVII. PLATE XII. SMYTH. SmytH—Problematie Structure in the Oldhamia Rocks of Bray Head. 231 account for distortion of angles. Pyrites suggests itself as containing iron, having a cubic form, and being common in argillaceous sediments. But, unless some orienting influence can be assumed, one would expect more varied shapes to result from the compression of cubes. EXPLANATION OF PLATE XII. Figs. 1 & 2.—Portions of the bedding plane of the shale, photographed wet. x2. Fig. 3.—A portion, photographed dry. Oldhamia antiqua on the right; worm burrow below. Note the slight differentiation of the marks when not wetted. X17. Figs. 4 & 5.—Transverse sections through the markings. Each figure contains two of the marks; the one on the right in fig. 4 is the same as that on the left m fig. 5. X 12. Fig. 6.—Camera lucida drawings of some of the spots from various parts of the specimen. X 0d. Fig. 7.—Camera lucida drawings of pseudomorphs of halite in Triassic marl. x18. *That such distortion may be remarkably small was shown by experiments with plasticine. Cubes of red plasticine were cut and embedded in green plasticine, the latter being carefully built up round the cubes in order to secure as perfect contact as possible. The whole was then strongly compressed between two parallel boards, and sections made in various directions. [ eng No. 28. THE HYDROGEN ION CONCENTRATION OF THE SOIL AND OF NATURAL WATERS IN RELATION TO THE DISTRIBUTION OF SNAILS. By W. R. G. ATKINS, O.B.E., Sce.D., F.LC., AND M. V. LEBOUR; D.Sc. [Read Junr 26. Printed Juny 27, 1923. ] Iv has been shown that the hydrogen ion concentration of the soil exercises a profound effect upon the distribution of plants and upon the growth of crops. Accordingly it seemed possible that similar effects might be observed in the case of animals, such as worms and snails. With the object of exploring this possibility, snails were collected in diverse situations. In order to get an idea of the relative abundance of the members of each species, every snail found by careful searching within a small area, about four square feet, was collected without any sort of selection. In the table of results the actual numbers are given. It was considered that this was better than to give the number of situations at which each species was found. The numbers given must not, however, be considered as precise statistical values, but as an attempt to give a rough quantitative record. Many of the more acid sites examined were very poor in snails. The hydrogen ion concentration of the soil was determined, for each situation in which snails were gathered, by means of the colorimetric method, as previously employed by one of us (Atkins, 1922) when studying plant distribution. The results are recorded, as is customary, in pH values. The symbol pH denotes the logarithm of the reciprocal of the number of grams of hydrogen ion per litre, viz., pH =log = Only a few observations relate to aquatic species. In order to economise space, the pH values of typical soil situations and waters are recorded in Table 1. A detailed account of the determinations and the supposed causes of the reactions found is given in the paper already mentioned. It may be mentioned that the R. Yealm, where examined, supports Asellus aquaticus, but not Gammarus pulex. The latter crustacean abounds, however, in the alkaline streams of the district. In water from these it can live and breed in small glass vessels, but when placed: the R. Yealm water it has invariably died. Determinations of soil pH values are made to pH 0-1, but in tabulating the values obtained, the results are grouped to nearest pH 0:5. In doing this, pH 68-72 is called pH 7:0, and pH 7:3-7'7 is given as pH 7-5. SCIENT. FROC. R.D.S., VOL, Xvi, NO. 28. 2x 234 Scientific Proceedings, Royal Dublin Society. Taste 1 Solution. pH. Notes. Peat, S. of Ireland bog. : 4-1 Saturated solution. », Dartmoor : ; 4-6 5p 5p ,, Dublin Mts. : 2 4:6 99 » Disintegrating granite, Dartmoor 5-2 No vegetation. Light soil, over slate 0 54 Bracken thrives. Surface soil, copse, in alkaline district. 6-4 Surface is more acid than subsoil. Clay soils, heavy 5°4-6'8 These become less heavy if limed. Garden soil : 6°5-7:5 Common values. Limestone soil : 6 7:6-8:2 Where limestone is not leached out. Bog pool, Dartmoor 5:0 Typical amber water, tadpoles abundant. Bog stream, Dartmoor 5:5 Light amber water, no recent rain. Rain water : 5:9 Acid from carbonic acid. R. Yealm, Dartmoor : 6:4-6:8 Examined where streams from sedi- mentary rocks flow in. Stream, Devon 8:2 Calcium bicarbonate saturation value. Sea water 8:2 Slight seasonal changes occur. Small reservoir 7:6-9:2 Depending on season, most alkaline when photosynthesis is active. TABLE 2. j | Number of specimens found at each pH value. Noset pH values of soil. 5-0 5:5 6-0 6:5 7-0 15 8-0 DUS ZONITIDAE. Vitrina pellucida 0 0 0 1 3 1 3 B) Hyalinia erystallina 0 0 1 0) 4 0 1 5 HH. nitidula 0 tv) l 4 3 4 8 13 H. fulva 5 0 0 0 0 0 0 1 H. pura 4 0 b) 0 0 0 0 2 H. vadiatula 4 0 1 0 0 0 0 2 H. alliaria 3 0 2 0 0 0 0 2 HELICIDAE. | Pyramidula rupestris . | 0 0 0 0 1 0 3 2 P. rotundata 0 1 by) A 9 8 7 18 Felicella virgata 0 0 0 9 1 9 74 15 Hl. caperata 0 0 0 0 0 23 36 ll H. itala 0 0 0 0 many v. many 0 2 Cochlicella barbara 0 0 0 0 0 0 8 1 Hygromia hispida. | 0 0 0 29 10 4 5 17 H, striolata a 0 0 0 1 0 12 16 7 Vallonia pulehella 0 0 0 1 1 0 0 2 Helix aspersa | 0 0 1 8 4 5 3 11 HH. nemoralis | 0 0 2 0 2 1 2 B) HH. hortensis | 0 0 2 2 1 0 0 3 STENOGYRIDAE. Cochlicopa lubrica 0 0 Bi) 17 8 2 0 8 PuripAr. Pupa umbilicata 0 0 7 20 16 0 iil 9 P. anglica 0 0 0 0 3 0 0 1 CLAUSILIIDAE. Balea perversa 0 0 il 0 0 2 2 Clausilia bidentata 0 0 0 0 5) 7 0 D) SuccinEipar. | Succinea putris : 0 0 0 0 4 0 0 1 AUKICULIDAE. Carychium minimum . 0 0 0 3 8 0 0 3 No. of species found 4 4 13 15 20 16 14 (total 27) Arkins anp Lesour—Hydrogen Ton Concentration of the Soil. 236 The results of Table 2 are shown graphically in figs. 1 and 2. The record of H. virgata, 9 at pH 6:5, must be explained. One was found on acid soil near the sea, the situation, being exposed to blown spray, so it must have been close to an alkaline region. The other eight were on bracken and nettles in a coastal situation where alkaline soil, as tested, was within fifty yards. probably less. ntl Fie. 1. It must be pointed out that the bottom line of Table 2 records the number of species for each pH value, each species being assumed to occur for every value within its extreme range. Thus H. nemoralrs, found at pH 6 and pH 7, is presumed to occur at pH 6°5 also. 25 Ml §o0 $5 6-0 65 7° 7S Fo Fig. 2. Viewed as a record of species of snails for the districts concerned, mainly the Plymouth district and the country round Bray Head, Co. Wicklow, the collection ranks as very imperfect. Sufficient work has, however, been done to show that the pH value of the soil is a factor which is important im regulating the distribution of snails in any district. Certain minute Hyalinias are 236 Scientific Proceedings, Royal Dublin Soctety. found on soil as acid as pH 48 over quartzite rock. Other Hyalinias are found at pH 8. The genus therefore covers a very wide range. Apparently species with hyaline shells can inhabit districts free from calcium salts, or where at least very minute amounts are found. On the other hand those with markedly caleareous shells, such as Hellicella caperata and H. virgata, appear to be strictly confined to regions which are predominantly alkaline. Again, there are a number of species which are plentiful in a neutral or slightly acid habitat, though they are found in alkaline regions also; these include Hygromia hispida, Pwpa wmbilicata, and Helix aspersa. When one recalls the fact that snails possess powers of locomotion, and that within a narrow area considerable differences in reaction may be found, it would be unreasonable to expect very hard and fast rules as to distribution. For example, owing to more thorough leaching out of calcium carbonate, a eully is usually more acid than the land it drains, and, even on alkaline land, plant remains and leaching may together render the surface soil slightly acid. It is precisely on such soil that H. hispida abounds around grass roots. The results for various species are shown graphically in fig. 1, as are also those for the numbers of species at different pH values in fig. 2. The authors are indebted to Mr. F. W. R. Brambell, of Trinity, College, Dublin, for collecting specimens on and near Bray Head, Co. Wicklow; also to Miss Worsnop for certain Plymouth specimens. Mr. J. W. Taylor, of Leeds, kindly identified the species collected by Mr. Brambell. Not included in the list of Table 2 are the following :—Limnaea truncatula at pH 7, L. stagnalis in abundance, in a pond at Kew (it is absent from Devon and Cornwall), at pH 7. It must be added that the pond contained hard water, rich in calcium salts. This could become much more alkaline, about pH 9, through abstraction of carbonic acid by water plants. Other ponds near were more alkaline, being richer in water plants. Bathyomphalus contortus also was abundant in a marshy pool at Plymouth, at pH 8. One specimen of Valvata piscinalis was obtained from the same situation, in which it may be remarked the large ciliate Spirostomum ambiguum was very plentiful. We are indebted to Mr. W. C. De Morgan for this identification. It has not been possible to collect many of the aquatic species as yet, but it may be said that they are either very poorly represented, or altogether absent from the upper reaches of the R. Yealm and the more acid waters listed in Table 1. As is well known, snails are more abundant in limestone regions than in most others. The hydrogen ion determinations put this observation on a quantitative basis; furthermore, many situations upon caleareous sandstones and shales are quite as alkaline as similar situations upon limestone. It may be mentioned that boulder clay, covering limestone rock, is often acid, as it is genetically unrelated to the underlying strata. Through leaching by rain most limestone regions are acid in parts, so a greater variety of soil reaction may be met with in them than in those where the rocks are granitic or quartzite. Thus Roebuck (1921) in his census records 97 species of land and fresh water mollusea from Co. Dublin, where limestone lowland and plutonic uplands are found. But the neighbouring Co. Wicklow, which is also maritime, has only 74 recorded species, Co. Wexford, on the other side of Wicklow, having 87. The absentees from Wicklow include Limnaea stagnalis, Valvata piscinalis, and Bathyomphalus contortus, though these are present in Wexford and Dublin. The records for these three species, as shown in this paper, are from alkaline water. Wicklow contains no extensive limestone areas, if indeed any limestone is to be found. The uplands consist of plutonic rocks, and most of the rest is made up of altered and unaltered Silurian strata, with the Bray Head Cambrian; Avxrns anp Lesour—lHydrogen Ion Concentration of the Soil. 287 Wexford is in part similar, but contains, like Dublin, a tract of limestone. It seems probable that the absence of these three species from Wicklow is not due to want of complete investigation, as Stelfox (1912) records for Dublin, Wicklow, and Wexford, 96, 73, and 80 species for 1910, which have only been increased to 97, 74, and 87 in Roebuck’s list nine years later. Moreover, Bathyomphalus (Planorbis) contortus is found in every other county in Ireland according to Stelfox. There is, however, a discrepancy as regards records for V. piscinalis, the species being listed by Stelfox as present in Wicklow, though not in Roebuck’s list. It must at least be a rarity in Co. Wicklow. One of the writers collected Mollusca at Dalbeattie, Kirkeudbrightshire, where the rock is chiefly grey granite. Very few occurred on the granite, but about two and a half miles from Dalbeattie, on the Palnackie road, near Munches, where the rock was basalt, the snails were strikingly numerous. The following is a list of those taken from near the road where the basalt begins :— Helix nemoralis juv. Arianta arbustorum juy. Pyramidula rotundata, many. Acanthinula aculeata, common among fallen leaves. A. lamellata, common among fallen leaves. Punctum pygmacum, common among fallen leaves. Hyalinia excavata, many. HT. cellaria, fairly common. H. crystallina, common. H. mitidula, fairly common. HT, alliavia, common. H. pura, common. Huconulus fulvus, common. Vitrina pellucida, a few. Pupa wmbilicata, common. Sphyradium edentula, very common. Sphyradium edentula v. substriata, common. Cochlicopa lubrica, common. Balea perversa, on walls, common. Acme lineata, one, among leaves. Carychiwm minimum, very common. Also the following slugs were present :—Avrion hortensis, Agriolimax agrestis, common, A. laevis, common. It may be pointed out that soil over granite is normally markedly acid, since this rock gives off inappreciable amounts of alkali to neutralize plant remains. With basalt it is, however, otherwise, for it has been found that while a weathered basalt boiled with distilled water gives usually a slightly acid solution, yet a fresh basalt surface may quickly yield sufficient alkali to give a reaction of pH 8:0. Out of eight basalts tested, on further boiling one remained acid, pH 5-4, six were at pH 7:0--73, and one at pH 83. It is highly probable therefore that the situation where snails were plentiful over basalt was markedly less acid than the granite region. It was most probably slightly acid or even almost neutral. The species found notably in the alkaline region were absent from the basalt, viz., H. virgata, H. caperata, H. itala, and Hygromia striolata. H. hispida was also absent, but species of Hyalinia, H. pura, H. (Euconulus) fulva and H. alliaria, known to occur even on highly acid soil, were present. _ While the authors feel convinced that the acidity or alkalinity of a situation is an important, and often a dominant, factor in limiting the distribution of 238 Scientific Proceedings, Royal Dublin Society. species of snails, they are fully aware that other factors may also be important or dominant. The amount of salt in the soil is tentatively put forward as one of these for land species near salt marshes, as the concentration of salt obviously is for water mollusea. Furthermore, the dry or wet nature of a habitat is also of much importance, though hard to express quantitatively. This factor and the relation between various plant associations and snails have been studied recently by Kendall (1921, 1922). The plant associations are themselves regulated by the soil reaction. Salisbury and Tansley also (1921) have pointed out that the mollusean fauna of certain woods on the Wenlock limestone are intermediate in character between the plentiful fauna of a caleareous beech wood, and the restricted list of species obtainable from an acid oak wood (Salisbury, 1918). In order to examine possible factors limiting the distribution of snails, other than the hydrogen ion concentration, a detailed study was made of three loealities, similar im soil reaction and general situation. From these large numbers of snails were collected, and since the sites were calcareous it was possible to get the specimens in a small area. For each site the pH value was determined, also the salt content as shown by the electrical conductivity of a mixture of one part of air-dry soil with five of water. The soil was sieved, and the portion passing the hundred mesh to the inch was used in the determin- ations. The mixtures were shaken at intervals for three hours on a rotating wheel, by which time it was judged that all the readily soluble salts had gone into solution. On standing, however, the conductivity increased and attained an equilibrium value, probably due to the solution of calcium carbonate as bicarbonate, occasioned by bacterial production of carbonic acid. The conduc- tivity was measured at 0°C., and the apparatus was standardized against one hundredth normal potassium chloride. The snail population is given in percentages, the total number collected being also shown. The sites’ were as follows :— A. Plymouth Hoe, grassy bank, near laboratory, gentle slope, south aspect. B. Plymouth Hoe, grassy bank, near sea, steep slope, south aspect. C. Fort Stamford, near Plymouth Sound, east side, grassy and stony bank, with moss and red valerian (Centranthus ruber), at east end, south aspect. D. Oreston Quarry, near Plymouth, on estuary of R. Plym, limestone rubble and grassy banks, aspects south and east. A. iB: C. D. Helicella caperata. . . S8percent. 2 percent. 72 percent. 54 percent JE, OOPGOHE 9 o 5 a 6 WO a 51 ie 18 - 0 4 Pupa umbilicata 6 5 23 i 1 BS 10 + Helix aspersa. 2 a if 3 4 ra 2 43 Helix nemoralis 0 ‘ 0 a 0 ae 1 - A, pulchella . 0 ry 0 33 0 " 1 ie Hygromia hispida . 0 5 iL is 2 Ks 3 = HT. striolata Sites 0 . 0 ms 2 55 12 i Pyramidula rotundata 1 as 0 5 0 “ 5 5 Hyalinia mtdula 4 . 0 ' 0 5, 7 . HT. crystallina 0 i 0 a 0 * 2 * Cochlicopa lubrica . 0 i 0 1 ie 3 ri Vitrina pellucida 0 5 0 ~ 0 i. 1 Fy Cochlicella barbara 0 f 22 ‘ 0 7“ 0 es Total count 3 dil? 348 150 159 Species 6 6 ri 12 DEL A RT BATE te 78 7-5-8:0 15-77 77-8:0 C X 10°, 3 hours 12 12 13 11 C X 10°, 11 days 37 31 31 27 Arxrins AnD Lesour—AHydrogen Ion Concentration of the Soil. 239 Site A appeared to be the best soil, growing good grass with leguminous plants, and conductivity measurements indicate that it was less washed out than were the other soils. The three-hour values show that none of the soils have more than small amounts of salts, the figures found being less than one hundredth part of that for sea-water. So the soils from B are not markedly salt, as it was thought they might be. Chlorides are, however, present in B in larger traces than in the others, and give an opalesence with silver nitrate though inland soils often give nearly as much. : It is evident that though the pH values are all between pH 75 and 80 there is no definite sequence in them. All the sites may be considered as more or less uniformly of slight alkalinity. With regard to exposure to wind, B is most exposed, followed by A, C, and D. The distribution of the snails is strikingly different, for example, in B and C, where the physico-chemical values are almost identical. Again, though D has twelve species out of the total fourteen, H. virgata and C. barbara are absent, in spite of the fact that they constitute 73 per cent. of the population of B and HH. virgata 79 per cent. of A. There appears to be no specially marked difference in temperature or moisture in the sites, all being dry and very warm, A probably retaining moisture a little better than the others. Of all the characteristics of the sites perhaps the force of the wind is the most noticeable difference, though it is not easy to see why on this score Oreston should have no H. virgata. According to the ‘‘ Age and Area’’ theory of Willis the extent of the distribution of a species depends upon the interval of time since its introduction into the region under consideration. This conception is obviously modified by the existence of ecological factors inimical to any particular species. On the other hand ecological factors alone must be incapable of explaining the distri- bution of newly arrived species. On turning to the map given in Roebuck’s census it is seen that Cochlicella barbara is widely distributed in Ireland, with inland as well as coastal records. In England, Wales, and Scotland, however, it is only found as a maritime species on the west coasts, and along parts of the south of England. These facts point to its having a south-western origin—from Europe or by Atlantie drift possibly. The area colonised by Theba cantiana is to be accounted for similarly in all probability, for it occupies the east of England, extending some way down the south coast and up as far as Northumberland, where it has been found by one of us. This species is absent from western England, Wales, Scotland, and Ireland, and may be considered to have an eastern origin. SUMMARY. 1. The hydrogen ion concentration of the soil is a factor limiting the distribution of snails. 2. Snails are more numerous at pH 7-8 than they are elsewhere. 3. The number of species of snails found in the districts studied increases from pH 5 four species, to pH 7 twenty species, falling at pH 8 to fourteen out of the total twenty-seven species found. 4. Snails with hyaline shells may be found in any portion of the range, but those with calcareous shells are limited to the more alkaline end. Granite and quartzite regions have few species, basaltic districts have a more numerous fauna, and in limestone areas both species and numbers of individuals give high values. 240 Scientific Proceedings, Royal Dublin Soctety. 5. There remain over a number of puzzling cases in which, within an area of two square miles, certain species are altogether absent from one locality, though abundant in others, in spite of similarity i pH value, salt content (as shown by electrical conductivity measurements of soil extracts), and aspect. ‘A difference only in exposure to wind could be noted. 6. The distribution of some species within the British Isles is probably explained by the ‘‘ Age and Area ”’ theory of Willis rather than by a limitation through unfavourable ecological factors. Cochlicella barbura appears to have a western, and Theba cantiana an eastern, origin, REFERENCES. Arkins, W. R. G—Some factors affecting the hydrogen ion concentration of the soil, and its relation to plant distribution. Sci. Proc. R. Dublin Soe., 1922, 16 (N.S.), No. 30, and Notes from Bot. School, T.C.D., 1922, 8, No. 133-177. Kenpaty, C. E. Y—The mollusea of Oundle. J. of Conchology, 1921, 16, 240-244 and 1922, 16, 248-250. Rorsuck, W. D.—Census of the distribution of British land and fresh water Mollusea. J. of Conchology, 1921,16, 165-212. Sauispury, E. J—The oak-hornbeam woods of Hertfordshire, Pts. 3 and 4. J. of Ecology, 1918, 6, 14-52. Sauispury, E. J. and Tanstry, A. G.—The durmast oak-woods (Querceta sesstliflorae) of the Silurian and Malvernian strata- near Malvern. J. of Ecology, 1921, 9, 19-38. Sretrox, A. W.—A list of the land and fresh-water mollusks of Ireland. Proce. Royal Irish Acad., 1912, 298, 65-164. oars No. 29. IMPROVED METHODS OF EVAPORATION IN THE LABORATORY. By H. G. BECKER, A.R.C.Sc.1., A.1.C., Demonstrator in Chemistry, College of Science, Dublin. (Read June 26. Printed Aucust 15, 1923.) Owi1ne to the frequency with which the operation occurs in chemical work, it is desirable that the methods of evaporation used in the laboratory should be as efficient as possible. On the industrial scale evaporation has been brought to a high degree of efficiency, largely owing to the fact that in this case it is not necessary to distinguish between true evaporation and ebullition. On the laboratory scale, however, the problem is complicated by the requirement that the liquid must vaporise without ebullition in order to avoid loss by spirting, and this has led to the usual method of heating the liquid to a low temperature on a water bath, and allowing the evaporation tc proceed slowly over long periods of time. Fie. 1. The object of the work here described was to determine experimentally to what extent this quiet evaporation could be hastened by allowing it to take place under the most favourable conditions, and whether the saving in time effected would justify the use of the more complicated apparatus which might be required. For practical purposes the determining factors in the vaporisation of liquids are (1) the rate of heat transmission to the liquid (which is determined by the temperature of the source of heat and the conductivity of the containing vessel), and (2) the rate at which the vapour is removed (which may be fixed by diffusion, a current of air, or by the removal of atmospheric pressure). The ratio between the two rates ‘determines the temperature of the liquid, and also whether vaporisation takes place with or without ebullition. From the point of view of quiet evaporation the removal of the vapour is the more important of the two, since the rate at which the vapour can escape from the free surface of an adequately heated liquid is the determining factor in preventing the temperature rising to the boiling point. The most favourable conditions for rapid, quiet evaporation are therefore that the rate of removal of the vapour should be as SCIENT. FROG. R.D.S., VOL. XVII, NO. 29. 2y 242 Scientific Proceedings, Royal Dublin Society. great as possible, and the heat supply so adjusted that it keeps the temperature of the liquid just at its boiling point without any danger of superheating. The first part of this work was therefore directed towards finding how far these conditions could be realised for evaporation in open dishes. Tn order to observe the rate of evaporation up to the boiling point, it was necessary to use an oil-bath for heating the dishes. The bath was made of copper, in the form of a rectangular box, 10” square and 5” deep, and was built into the lower surface of a copper wind-tunnel, of which the front could be closed by glass plates. As shown in fig. 1, the whole was designed so that the dish containing the water was almost completely immersed in the oil, and the water surface formed a continuation of the bottom of the tunnel. The cross- section of the tunnel was 10” by 8”, and it projected 14” on each side of the bath. The bath could accommodate dishes up to 8” diameter, and was provided with a stirrer, a thermo-regulator, and a thermometer. For the lower tempera- tures one bunsen burner was used for heating, but for the higher temperatures two were necessary. Dish Temperature 60 80 100 120 140 160 180 200 220 Bath Temperature Fic. 2. The dishes used were glass crystallising dishes, with straight sides, and the rate of evaporation was measured by the fall in level of the dish. The level was observed by providing a small glass float drawn out to a fine tip and kept in a vertical position by a small stand made of brass rod, so as to rise and fall with the surface of the liquid, the position of the tip of the float being read by means of a cathetometer. This method allowed of a great number of observations being taken in a short time, so that the mean rate of evaporation could be obtained from a graph of the observations. At the left-hand end of the wind-tunnel an aluminium fan, 8” diameter, was fixed, which could be driven by an 4 H.P. motor. By this means air-currents up to 1,000 feet per minute could be produced, the speeds being measured by a small anemometer. With this apparatus the rate of evaporation of distilled water was measured over a range of 30° to 98° C, in still air, and from 30° to 70° C. in air currents. BrcxEr—Improved Methods of Evaporation in the Laboratory. 248 As the oil was maintained at a steady temperature by the thermo-regulator, it was possible to note the difference of temperature between the oil-bath and the water in the dish during the evaporation. This is an important point, as it is an indication of the rate at which heat is being supplied to the liquid. As the loss of heat becomes ereater at higher temperatures, this difference naturally increases rapidly with increasing temperature. The temperature of the water for different bath temperatures is shown in fig. 2, where it will be seen that a bath temperature of 100° C. gives a water temperature of about 70°C. in still air (curve a), about 60° C. in a moderate draught (curve b), and 54° C. in a strong draught (curve ¢c). These temperatures, therefore, represent the maximum obtainable under the given conditions with a water bath as source of heat. However, by using a higher bath temperature it is possible to maintain the water at, or near, its boiling point, and thus obtain the great advantage of the higher vapour pressure. The graph shows that to maintain water at 100° C. in still air the bath must be at 170° C., in a moderate air current it must be at 197° C., and in a strong draught at 215°C. This graph thus shows the approximate temperature at which any given bath should be maintained in order that water may evaporate at a certain temperature under the given conditions. oO Qo Rate of Evaporation co) S 9) 20 40 60 80 100 Temperature Fie. 3. Owing to the difficulty of maintaining the higher temperatures with the oil- bath, the evaporation from 70° to 100° C. im air currents was measured in a slightly different way. An aluminium dish was supported on a powerful ring burner in close proximity to the fan, so that the air could be blown across its 244 Scientific Proceedings, Royal Dublin Society. surface. The burner was lighted, the fan started, and the burner regulated until the water came to the steady temperature required, when the observations of the float were made as before. By this means the temperature of the water could be maintained at the boiling point even in a draught of 1,000 feet per minute. The results of these experiments are shown plotted as a graph in fig. 3. The effect of the vapour pressure is shown by the upward sweep of the curve in the region 70° to 100° C. Thus the increase in rate of evaporation due to the rise from 90° to 100° is equal to that due to the rise from 30° to 90°, so that, roughly speaking, each degree rise in the higher region is six times as effective as a degree in the lower region. This is further emphasised in the graph fig. 4, where the rate of evaporation is plotted against vapour pressure. It will be seen that the rate is proportional to the vapour pressure up to 90° C., but above this it increases more rapidly. 12 (oe) (o} oo poration Rate of Eva Vapour Pressure Fig. 4. The region 90° to 100° is therefore particularly effective in producing rapid evaporation, and wherever possible the water should be kept within this tem- perature range. It is also interesting to note how the results for evaporation in still air compare with figures calculated by means of an approximate formula put forward by Hinchley,’ namely, rate of evaporation in kilograms per sq. metre per 22 P 1:2 hour from water surface = om where P, = vapour pressure of the liquid in mm. Of mercury Pg = pressure of the water vapour in the air. 1¢«The General Problem of Evaporation,’’? J. W. Hinchley. Jour. Soc. Chem. Ind., 41, 242'T., 1922. BrckEr—IJmproved Methods of Evaporation in the Laboratory. 246 The two series of values agree fairly closely up to 90°, but above this the experimental values are higher than the calculated. From both these facts it appears that, as the boiling point is approached, the rate of evaporation increases more rapidly than it would if it were proportional to the vapour pressure. This may be due to the fact that at the higher temperatures the convection currents in the vapour over the surface of the liquid become much more rapid, thus leading to a more effective removal of the vapour. The effect of a current of air in hastening evaporation is also shown by the two upper curves (b) and (ce) in fig. 3. The effect is most marked at low dish temperatures, and falls off as the temperature rises. Thus at 50° C. the rate is 2:8 times as ereat in a current of 500 feet per minute, and 3:8 times in 1,000 feet per minute; at 80° C. the figures are 2:0 and 2:5 respectively, while at 100° C. they are 1:7 and 2:2. ‘These figures also show that, while at low temperatures the rate is approximately proportional to the velocity of the current, at high temperatures the moderate current is almost as effective as the higher speed. s The relatively high efficiency of the air current at low temperatures is probably due to its action in preventing the vapour from accumulating over the surface of the liquid, as it is liable to do, owing to the weakness of the convection currents. On the other hand, the falling-off in efficiency of the current at high temperatures is probably due to the fact that the draught removes the heat from the surface layers faster than it can be supplied by conduction from the body of the liquid; this results in a surface cooling, which slightly reduces the rate of evaporation. When the air is heated before being led over the surface of the liquid this surface cooling is avoided, and slightly higher rates of evaporation are obtained. Thus Aldrich’ has described an apparatus for evaporating solutions at 20° to 30°, in which an average rate of evaporation of ‘01 ¢.c. per sq. em. per minute is obtained in a draught of warm air, while the rate obtained with the draught at room temperature is about ‘007. In most of the formulae put forward for rate of evaporation in a draught the factor connecting it with evaporation in still air is assumed to be a constant, but this probably arises from the fact that the experimental results were mostly obtained at low temperatures. Thus Leonard Hill found that the evaporation from a muslin surface was doubled in a wind velocity of ‘55 metre per second (107 feet per minute); and Carrier arrived at a value’ of 1:17 metre per second (228 feet per minute) for a similar doubling. The present experiments show that the factor cannot be regarded as constant over the range 40° to 100° C. From a consideration of these results the best conditions for evaporation from open dishes can be laid down. The rate of evaporation on the water bath (dish temperature 70° C.), as at present in use, is approximately -01 ¢.c. per minute per sq. em. Raising the temperature of the water to 95° C. raises the rate to 04, without leading to any increased risk of loss, while, if the water be kept at the same temperature in a draught, the rates increase to 07 for a draught of 500 feet per minute, and to 087 for 1,000 feet per minute. In order, therefore, to cut down the time required for evaporation to one- seventh of its present value it is merely necessary to make arrangements to keep the water at 95° C. in a moderate current of air.’ This may be done by means + J. Biol. Chem., 23, 255, 1915. * See Hinchley, loc. cit. “It is important to note that it is not sufficient to heat the water to 95° C. and then start the draught, as the cooling effect of the air then practically neutralises its efficacy in removing the vapour; the heat supply must be sufficient to maintain the liquid at 95° C. in the draught. 246 Scientific Proceedings, Royal Dublin Society. of a thermostatically controlled air, oil, or sand bath provided with a suitable fan. With such an arrangement an evaporation which would take a day with existing apparatus could be completed in an hour, so that the saving in time would amply repay any extra trouble in setting up the bath in the first instance. The main difficulty in evaporation, as will be seen from the foregoing, is to supply the heat to the liquid sufficiently rapidly to produce rapid evaporation and yet prevent ebullition. The latter is to be avoided on two grounds: (1) its very existence shows that the liquid is being superheated locally; and (2) it may lead to loss of the liquid by spirting. When liquid is heated in a vessel with only the convection currents to keep it mixed, the bottom of the vessel may become heated much above the boiling point of the liquid, and, since the convection currents are unable to mix the liquid sufficiently rapidly, ebullition of more or less violence occurs if the heat is supplied at more than a certain rate. eee Tp ] A G = 3 KW Le A Fic. 5. Tt is well known that this tendency to superheating of the liquid is diminished by stirring or shaking, but this leaves the liquid still in contact with the highly heated bottom of the vessel. Any method which would keep the liquid mixed and at the same time prevent the heating ot the containing vessel to a temperature greatly in excess of the boiling point of the liquid would provide almost ideal conditions for rapid evaporation. Such a method is incorporated in the design of the evaporator described below. In its simplest form this consists of a glass bulb containing the liquid, and kept in constant slow rotation while it is heated by a gas burner. Owing to the adhesion of the water to the glass, it.is drawn up on the side of the bulb which is rising, and a continuous thin film is spread over the upper part of the bulb. This ensures very thorough mixing of the liquid, and also materially increases the surface of liquid from which free evaporation can take place. At Brecxer—Improved Methods of Evaporation in the Laboratory. 247 the same time, the rotation of the bulb over the burner prevents any portion of it getting superheated; and, in fact, the temperature of the liquid can be regulated very exactly by the height of the flame used. With a plain bulb of this type water may be rapidly boiled away with very slight formation of bubbles, and consequently no loss by spirting. If a current of air be blown through the bulb, a similar effect to that noted in the experiments recorded in the first part of this paper is produced—namely, that the liquid may be much more strongly heated without the formation of bubbles. The air current may be produced in two ways: either by making use of the rotation of the bulb to eject the mixture of air and vapour centrifugally, or by blowing in air directly by, means of a filter-pump or bellows. Fig. 5, shows the modification designed to use the first method, consisting of the addition of a wide bore T-piece fused to one end of the bulb. When the bulb is revolved at a moderate speed (about 150 r.p.m.), the mixture of air and vapour is thrown out through the arms of the T by centrifugal force, and the air is consequently drawn in through the cther end. In this way the escaping vapour forms its own blower, and materially hastens the evaporation. Fig. 5, B shows the modifi- cation suitable for use with an outside current of air. The air-jet is placed in the bell-mouth of the tube, so that it draws in with the primary air a considerable volume of secondary air, thus producing a current which completely removes the vapour as it is formed. The latter method can be further amplified by supplying to the jet any particular gas in which the liquid being evaporated may be most stable. Thus easily oxidisable substances can be evaporated in an atmosphere of coal-gas or hydrogen, while readily hydrolysable chlorides may be evaporated in hydrogen chloride. When the gas used is scarce or expensive, arrangements can be made for circulating it continuously through the bulb, so that a limited volume would suffice for a large volume of liquid. When the liquid has to be evaporated to dryness and the solid recovered, as in the determination of silica in analytical work or in the reecrystallisation of salts, a form of apparatus with a very wide opening at one end is used. This is shown in fig. 5, c, and in its simplest form may consist of a round-bottom flask of which the bottom has been softened and blown out, the neck cut off short, and the whole mounted on a piece of glass tubing by means of a rubber stopper. This allows the bulb to be revolved and a current of air blown through at the same time, while the solid left behind can be scraped or washed out at the conclusion of the evaporation. Incidentally it may be noted that, owing to the action of the rotating bulb, solutions never become supersaturated, but the salt cerystallises out in very fine small crystals, even at the high temperature, as the liquid becomes concentrated. In this way no difficulty is experienced in erystallising such difficult substances as potassium hydroxide and _ ferric chloride. Owing to the complete absence of superheating, the apparatus is particularly suited for use in evaporating at reduced pressure, as it obviates the violent bumping which usually gives such trouble at low pressures. A form which has been found effective for this purpose is shown in fig. 5, p. The revolving bulb is connected to the fixed end tubes by means of carefully selected pieces of black rubber tubing, of such a size that when slightly moistened they allow the bulb to revolve freely. When the whole is properly adjusted, the power required to revolve the bulb is not large, even when the atmospheric pressure is acting on the rubber joints, a small electric motor of 1/20 H.P. being quite sufficient, 248 Scientific Proceedings, Royal Dublin Society. The following rates of evaporation have been obtained with revolving bulb evaporators of different types :— Bulb with one end open to air (type C) ... °142 ¢@¢. per sq. em. per minute. Bulb with one end open to air, with air-jet ... 237 ,, BY ea fe Bulb with centrifugal cross-arm (type A) ooo let) 3) ” ”? ? The rates of evaporation are calculated on the area of the water surface when at rest. ; CHEMICAL LABORATORY, CoLLEGE OF SCIENCE, DUBLIN. p 249 |] No. 30. A RAPID GASOMETRIC METHOD OF ESTIMATING DISSOLVED OXYGEN AND NITROGEN IN WATER. By H. G. BECKER, A.R.C.ScL., A.LC., AND W. E. ABBOTT, A.R.C.Sc.1., A.L.C., B.Sc. (Read JuNE 26. Printed August 15, 1923.) Introduction. Tue transient milkiness of water in which potash is dissolving, due to minute bubbles of dissolved air, suggested the possibility of utilismg this phenomenon for the estimation of dissolved gas. Consultation of the literature’ encouraged this view, since it appeared that the solubility of a gas in ‘dilute solutions is generally considerably less than in pure water. While little was known regarding the solubility of gas in concentrated solutions, it appeared probable that it should be nearly zero in very concentrated solutions of certain electrolytes. It was proposed, therefore, to work out a method whereby the water should be brought in contact with the solid electrolyte under such conditions that the displaced gas could be readily collected and measured. Preliminary attempts made to displace the dissolved oxygen and nitrogen from tap-water showed that the gas evolution was fairly complete, though rather slow at ordinary pressure. With a view to hastening the process, it was considered advisable to arrange that the gas should be liberated under a greatly diminished pressure. The choice of a solute was the first point to be considered. Caustic potash had been shown by the preliminary experiments to be very good in most respects. Owing to its high cost, however, attempts were made to find a cheaper substance which would be equally effective. Such materials as were not available in a compact form suitable for testing were prepared for use in our apparatus by fusion, and subsequent solidification. Caustic soda was as good as potash, except for a tendency to crystallize and block the capillaries during manipulation. Magnesium chloride, and calcium nitrate in large quantity expelled practically all the dissolved gas, but at a very slow rate. Fused potassium chloride expelled about 95 per cent. of the dissolved gas, but twenty to thirty minutes were required for this evolution, as against one to two minutes when using potash. Sodium sulphate, citric acid, calcium chloride, sodium carbonate, sodium thio- sulphate, zine chloride, and alum proved unsatisfactory. The rate of evolution of gas when using potash or soda is very remarkably greater than that which follows the use of any other of these materials. Glycerol, and strong sulphuric acid in large amounts give a rapid and good, but possibly not quite complete, evolution of gas. It was finally decided to adopt. potash as the solute in the absence of a satisfactory substitute; but we recognise that we have not investi- gated the possibilities of substitutes as fully as we would desire. Thus it has le.o., G. Geffcken Z. physik. Chem. 49, 271, 1907. SCIENT. PROC. R.D.S., VOL. XVII, No. 30. 260 Scientific Proceedings, Royal Dublin Society. been shown’ that a litre of a normal solution of ammonium chloride only contains 0:07 ¢.c. of dissolved oxygen, and, if this substance can readily be prepared in a suitable form, it would probably prove satisfactory. On the other hand, ammonium chloride dissolves with absorption of heat, while potash evolves heat on solution, and this might militate against ammonium chloride as compared with potash, since heat evolution undoubtedly greatly accelerates the displace- ment of gas. Description of Apparatus. The final form of apparatus used is shown in the figure. O is a capillary 3-way mercury-sealed tap, joined to a tube AB of 2:5 to 3 mm. bore. This igs Y () oo joined to a bulb of 20 to 30 ce. capacity, connected by a short tube of about 1 em. bore to a vessel of the shape shown, about 9 em. long by 4 em. in diameter. 1 MacArthur, J. Phys. Chem., 20, 495, 1916. Becker & Assott—WMethod of Estimating Dissolved Oxygen, Fe. 251 The side tube Z, for introducing solid potash into the apparatus, is of 2 em. bore, and is carefully shaped so that a rubber cork will fit it perfectly. Above the 3-way tap one of the capillaries is connected to an absorption pipette containing alkaline pyrogallol, while the other is bent sideways and downwards, and projects in front of the tap. W, Y, and Z are rests for the mercury reservoir on the wooden stand to which the apparatus is fastened. Y was at such a height that when the reservoir was placed on it the mereury level fell to H, leaving a vacuum in the upper part of the apparatus. Marks were etched on the glass at A, B, and C, and the volumes OA, OB, and OC were determined by weighing their contents of mercury. Experimental Procedure. The mercury level is brought to rest at H, and secured in this position by a screw clip at G. Then the potash sticks, broken up into pieces 1 to 14 inches long, are introduced. The potash used was Merck’s (85 to 86 per cent. KOH). The same amount can be introduced into the apparatus for each estimation by using the same length of stick, but in nearly all our experiments a quantity sufficient to ensure saturation was used. As a general rule it may be taken that 1 gram of potash should be used for each cubic centimetre of water sample, although possibly less than this may suffice. The cork is replaced and secured by a metal clip M passing around the bulb. The mercury level is then raised until the capillary S is full of mercury, whereupon the tap is closed. On placing the mercury reservoir in the bottom rest Y the mercury falls to the vicinity of E, leaving the potash sticks exposed and in a partial vacuum. When the reservoir is raised, a bubble collects in OB, and is expelled from the apparatus via S. This operation is repeated until no more air can be detected in the apparatus. Three exhaustions usually suffice. The apparatus and capillary S being full of mercury, the sample to be tested is drawn in through S until the mereury level reaches the mark C. By this means a definite volume of water OC is measured. Having closed the tap, the reservoir is placed in Y, whereupon the water sample falls on the potash sticks, and the dissolved gas is expelled with great vigour in one to two minutes. When boiling ceases, the mereury level is made the same inside and outside the apparatus, and the bubble is measured by making its top correspond with O, and measuring the distance of its bottom from A. The height of the potash column is also measured. The bubble and a little of the potash are then forced into P, and the rest of the potash is removed through 8. After gently rocking the stand for one to two minutes, the bubble is drawn back to OB and re-measured. This operation is repeated to ensure complete absorption of oxygen. Slightly over five minutes are required for the total gas determination, while the oxygen determination inereases the time by about ten minutes. From the measurements it is easy to calculate the volume of gas per 100 c.c. of sample. The alkaline pyrogallol reagent is best prepared by dissolving 5 grams of pyrogallol in 30 cc. of water, and saturating with potash in the apparatus, in the same way as a water sample. The reagent can then be driven into the pipette without contact with the air. Before starting a new determination it is essential to wash all traces of pyrogallol from OC. It is also advisable to wash the tap with water and acid after a series of experiments. 252 Scientific Proceedings, Royal Dublin Society. Discussion of Results. The new method was standardised at first by comparing the results obtained by it with those obtained by Winkler’s iodometric method, using distilled water saturated in a thermostat under definite conditions. At a later date it was tested by comparing the results of analyses obtained by its use and by boiling out. Many series of experiments, with modifications of apparatus and procedure, were performed, and the results are given in the accompanying table. Hach figure represents the mean of three determinations. Results in cc. gas at N.T.P. per 100 cc. water. Each figure average of 3 determinations. | TOTAL GAS. S OXYGEN. New Calculation Boiling | Winkler Calculation Boiling Method. from Tables. Out. Naw MIS HoG 'Determination.| from Tables. Out. i i} 1:92 1:76 — 0-590 | 0:613 0-611 -= 1:90 1:73 _ 0-604 0-621 0-604 —_ 1:83 1:72 = 0°562 | - 0:580 0:595 — 1:81 1:69 ut 0579 | 0-597 0:590 Bs 1:86 1:73 == 0:591 0-605 0:594 — 184 1-75 Me 0589) | 0:598 0-600 bah 1:87 1:76 — 0-606 0-620 0:603 —_ 1:86 1:75 — 0:595 0°595 0:604 — Mean 1:86 Mean 1:74 = Mean 0°590 Mean 0:603 | Mean 0:600 — Te tet rv a ] | 2:34 2-27 2304 || 0-750 a6) O00 0-808 2-25 215 215 OLER | = 0°750 0-747 | | It is ‘seen that the total oxygen and nitrogen by the new method is con- sistently greater than the standard values. On the other hand, the figure for oxygen recorded by the new method is always less than that given by the standards. The following appear to be the possible sources of error :— (a) Gas contained in potash sticks. —The magnitude of this error was determined by performing blank experiments with air free water and different. amounts of potash in the apparatus. Minute bubbles were evolved in all cases, and measured in the fine capillary tubing of the tap. The size of the bubble was proportional to the quantity of potash dissolved. When 264 cc. of water were saturated with potash the bubble measured 0-01 to 0:015 ce. This would only account for about 0:06 ¢.c. of the difference of 0:12 ¢.c. between the experimental and calculated values for total gas in the first part of the table. It would, however, just suffice to account for the difference in results in the second part. (b) Defective oxygen absorption—At first the low value for the oxygen figure was put down to imperfect evolution of oxygen. When, however, this Brecker & Asporr—Aethod of Estimating Dissolved Oxygen, §¢. 258 figure continued low in further series of experiments, and, moreover, the total gas figures remained higher than the tabular value, defective oxygen absorption appeared to be the cause. The literature revealed a widespread dissatisfaction with alkaline pyrogallol as an oxygen absorbent for very accurate work, and very varied recipes are given by different authors. In our work the difficulty inherent in alkaline pyrogallol is no doubt greatly aggravated by the minute size of the bubble to be tested. The formula we finally adopted approximates to that recommended by Anderson.’ It is, nevertheless, not wholly reliable, as is amply proved by the unsatisfactory results recorded in the second last series in the table; obtained when using a freshly prepared sample of this reagent. It was further noticed that the reagent does not work very well on the first absorption of a series if it has remained idle for some time. This difficulty, which is mentioned by other workers, can be overcome by absorbing an unknown volume of oxygen from air before beginning a series of determinations. Chromous chloride was tried as an alternative absorbent, but had to be abandoned owing to the precipitation which occurred when the acid solution met with a trace of potash. (c) Incomplete evolution of gas.—The facts discussed under («) indicate that it is probable that the error introduced by this cause is negligible. Conclusions. The work so far completed shows that the use of a very soluble substance, such as potash, to expel dissolved gases has many advantages. ‘The value of direct gasometric measurements is very generally recognised, and volumetric methods are only resorted to in order to avoid the tedious process of boiling out the gases and measuring them. ‘This method gives all the advantages ot the gasometric method, and the further important one of rapidity. The modifications which must be introduced into the simple Winkler process, for instance, when dealing with heavily polluted liquids, or in the presence ot chlorides, not only complicate the manipulation, but add very greatly to the time required. On the other hand, there is no reason to suppose that the action of the potash would be attected by the ordinary impurities in a water, and an estimation could be done as easily as with distilled water. Where the quantity of the sample is limited, the method only requires 20 to 00 cc. of the water, as compared with the 250 «ec. used in the volumetric processes. The disadvantage (from a commercial point of view) of the cost of the potash may quite possibly be removed by a more minute search for a suitable substance than the present authors have had opportunity to make. The method certainly works much better than its simplicity would lead one to expect, and it is hoped that 1t may be possible to develop it still further in the future. CHEMICAL LABORATORY, COLLEGE OF SCIENCE, DUBLIN. 1 Journ. Ind. Hny. Chem., 1915, p. 587. SOIENT, PROG. R.D.S., VOL. XVI, No. 30, BA ay} sg idle MRC eNia ARs | pe ase ae Fe uF ED Cod [ 255 J No. 31. LIGNEOUS ZONATION AND DIE-BACK IN THE LIME (CITRUS MEDICA, VAR. ACIDA) IN THE WEST INDIES. + By T. G. MASON, M.A., Sc.D., Botanist, West Indian Agricultural College. (Piates XIII--XV1) (Read May 29. Printed August 28, 1923.) Introduction. TxHouGH the Lime thrives on the whole in the more humid of the Lesser Antilles, yet its cultivation in those islands which are subjected to periods of pronounced desiccation is fraught with considerable hazard. A progressive dying back of the terminal shoots is very prone to occur in habitats in which the aridity of the plant’s environment is liable to fluctuate markedly. Nowell’s (8) view is that the young trees become established and do well for about ten years, attaining to a good size and bearing normal crops. Then, in a uniform field of this nature, the trees most exposed to the desiccating blast of the Trade Winds begin to show signs of die-back, which sooner or later extends widely, and in two or three years may involve the whole field, or may leave for a time groups of less affected trees here and there. The loss of dominance of the apical buds and the replacement of the mother shoots by laterals become especially accentuated. The terminal part of the mother shoot beyond the daughter shoot ceases to grow, sheds its leaves, and dies. For a time the losses incurred in this way may be balanced by the production of new shoots. Partial recovery may even occur, provided the water balance of the plant (6) becomes stabilized, but normally the decline, when once initiated, continues. The tree dies branch by branch; a process of self-pruning, which is accelerated, as a rule, by the presence of scale insects and the infestation of the dying shoots by Diplodia, ete. ; Etiology. Nowell (8) states that the underlying cause of the decline appears to be one of insufficiency and irreeularity of the water supply, and that the duration and completeness of the dry season seem to have more effect that the intensity of the wet season. He also points out that a heavy crop of fruit, especially in dry weather, frequently leads to a loss of branches. Spasmodice applications of manure or of cultivation are generally attended by similar results. Plan of Work. As the writer recently (April, 1921) had occasion to visit a number of Lime estates in the islands of Montserrat and Dominica, a collection of shoots was made from trees showing the die-back condition and from those in a more SCIENT, FROC. R.D,S., VOL, XVI, NO, 31, 3B 256 Scientific Proceedings, Royal Dublin Society. healthy state. It is probably unnecessary to point out that nearly every stage between trees in perfect health and those showing the most pronounced dying back of the branches is to be encountered. The collection of shoots was made with a view to an examination of the differentiation of the xylem cylinder. Tt was areued that if pronounced fluctuations in the water balance of the plant (6) were the cause of the affection, these fluctuations might be registered by different types of zonation in the woody cylinder. Periodicity of Growth. Before deseribing the results of this examination it seems desirable to say something concerning the periodicity of growth and its relation to moisture conditions. Temperature fluctuates so.little in these islands in the course of the year that it does not cbtrude itself as a factor. Reference to fig. 1 will br 60 (8) O I Ox nip Coefficient ability. | 40 (0) 50 ro Coefficient of Variability of Monthly Rainfall aN 60 Q S 9 fo Frust Production in ( aHOnS of f ime:duice. / 20 Mean Monthly Rainfall in Inches. Ga/lons of Lime-Juice per Acre ) v : : Secs SR SS. UR Sy ag Cone oy aie Came So 18s gah SS Sa ee oer ee my Fic. 1. show that there are two periods in the course of the year at which the rate of fruit production becomes markedly accelerated. In other words, there are two crops, both of which are gathered in the wet season. Inasmuch, however, as the fruit requires some five months to mature, the earlier portion of the first crop may develop under rather arid conditions. The large values for the coefficients of variability in rainfall from February till June, when considered in relation to the meagre rainfall at this time, will serve to indicate how frequently spells of drought oceur. Vegetative growth (leaf production and shoot elongation) commences some time after the maturation of the fruit, and is generally most active between January and August—that is to say, in the dry season. It will thus be seen that the period of maximum vegetative activity is also the time at which spells of drought are especially liable to occur, : “NOSVIN “OL %G ‘o ‘OLy ‘00Ud “LNHIOS TIAX “IOA “S'N “OOS NITEHNG “A ULV 1d “IIIX Mason—Ligneous Zonation and Die-Back in the Lime, West Indies. 267 In a word, vegetative activity takes place in the dry months of the year, and the maturation of fruit in the wet. The aridity fluctuates rapidly and markedly in the course of the dry season, while the vegetative meristems are most active. There is generally a bare sufficiency of soil moisture and frequently a pronounced shortage at this season, while the desiccating power of the aerial environment becomes particularly accentuated as a result of the vigour with which the Trade Winds blow about this time. The Zonation of the Xylem Cylinder. The different types of zonation presented by the wood of the shoots from the various habitats will be now considered. As the habitats and the condition of the trees supplying the shoots, which are dealt with in this paper, have been recently deseribed by Hardy (2) in the course of his ‘‘Studies in West Indian Soils,’ it is possible to consider the zonation in relation to his description, which is reproduced below in italics. The sections, which were prepared by polishing, are enlarged approximately nine diameters. THe MuLtcHEep PLor.—RosrAau Lime EXPERIMENT STATION oe 2, Puate XIII). Lime trees completely free from die-back. The mulched plot, which supports a vigorous and. eatin crop of lime trees, is well drained, and is composed of soil in good tilth and of considerable depth. It is adequately sheltered from winds by the proximity of lime trees of neighbouring plots, and by a dense artificial wind-belt that lines the nearest hill- crest. Mean Annual Rainfall—80 inches. The environment is on the whole conducive to the maintenance of the water balance of the plant throughout the year, though rather dry periods may occur between February and May. The presence of the mulch does much, however, to stabilize the soil moisture conditions. Xylem Cylinder. Growth zones can be distinguished, though they are not very marked. They are indicated by circumferential areas, in which the large vessels are more densely grouped. Somewhat ill-defined tangentially disposed bands of parenchyma tend to oceur within these areas. This section does not suggest that the autonomy of the plant was markedly influenced by the adversities of the environment. The tendency for the vessels to be grouped is presumably associated with periods of more rapid leaf- production. The parenchyma bands, it is conceived, may indicate intermittent checks in the activity of the cambium as a result of the production of con- siderable tension (1) in the water-tracts of the plant; in other words, they record intermittent periods at which the aridity of the environment (4) became pronounced. OLVESTON VALLEY, Montserrat (rig. 3, Puate XIII). Lime trees free from die-back. The soil in this valley near the sea is typical of the valley soils of the northern part of Montserrat. It is rocky but deep, especially in the pockets between the boulders. The situation is sheltered, and the soil at the time of sampling (April, 1921) was moist to the touch. The weeds growing between the lime trees were tall and vigorous. Mean Annual Rainfall—60 inches. 258 Scientific Proceedings, Royal Dublin Society. The environment in this valley is less conducive to the maintenance of the water balance of the plant than in the Lime Experiment Station at Roseau, Dominica; the absence of a surface mulch is especially important. Xylem Cylinder. The zonation is here very marked. The large vessels are rather densely grouped circumferentially conferring the appearance of ‘‘spring wood,’’ within which there occur bands of tangential parenchyma; these bands, it will be seen, tend to be grouped radially. It will be evident that there was a rather definite periodicity in the activity of the cambium. The presence of ‘‘spring wood’’ may indicate that leaf-production was vigorous over a rather limited period; that, in horticultural parlance, there was a flush of growth. It will be remem- bered that growth is generally most active during the dry season. The presence of the radially grouped parenchyma bands may be due to intermittent periods of desiccation about this time. O’GaARRA’S HiILLSLoPE, MonTSERRAT (FIG. 4, PLaTE XIV). > ? Lime trees free from die-back. The steep talus slope above O’Garra’s is dissected longitudinally by deep ghauts, bordered by steep round-backed ridges. In the ghauts a certain amount of shelter from winds is experienced, but the slopes of the divides are exposed and dry. The lime plot examined occurs in one of these sheltered ghauts, at an elevation of some 250 feet. The soil is a brown deep loam, and is characterized by the presence over its surface of a very well-marked layer of volcanic cinders and small stones. They form a loose stone-mulch on the soil, and, when removed, expose a plexus of lime rootlets, which ramify through a morst soil. Mean Annual Rainfall—60 inches. Xylem Cylinder. A rather definite periodicity in the activity of the cambium is indicated. The parenchyma bands, it will be observed, again are radially grouped, but they tend to be distributed in the ‘‘autumn’’ rather than in the ‘‘spring’’ wood. This is intelligible, if it be assumed that leaf-production was deferred until the close of the dry season, and that the tangential parenchyma resulted from intermittent desiccation of the cambium during these months. GRovE Boranic Station, MONTSERRAT (FIG. 5, PLATE XIV), LInme trees free from die-back. The soil is very wuform in texture, deep, and uncompacted. A heavy mulch of cane grass, cut from a neighbouring plot, covered the soil surface. In con- sequence of this, and partly also because of the sheltered position, the soil at the time of sampling (April, 1921) was moist to the touch. Well-grown specimens of the leguminous shrub, Gliricidia maculata, occurred between the lime trees. Mean Annual Rainfall—60 inches. Xylem Cylinder. The differentiation into zones simulating ‘‘spring’’ and ‘‘autumn’’ wood is here scarcely perceptible. The zonation is mainly due to the parenchyma bands, which are on the whole distributed at regular intervals radially; this is more “NOSV]V @ ‘ON “AIX FLV Id TIAX “1OA “OOS NITANG “A “OO’Ud “INHIOS SCIENT. PROC. R. DUBLIN SOC., VOL. XVII. MASON. PLATE XV. Mason—Ligneous Zonation and Die-Back in the Lime, West Indies. 259 marked in the right-hand side of the section. These bands are again grouped, and are not especially marked. It would seem that they originate as a result of recularly recurring periods of desiccation. A certain similarity between this section and that from the Lime Experiment Station at Roseau, Dominica, can be traced in the radial distribution of the vessels; the vessels are, however, distinetly smaller in diameter, and, moreover, the parenchyma bands better defined. Both characters may be due to a more pronounced aridity in the environment (cf. 6). BELLE Firup, Grove Esratrr, Montserrat (ric. 6, Pharm XY). Lime trees exhibiting die-back. The soil is of the talus slope type, free from large stones, and in good tilth. The situation is exposed to winds, and the existing artificial wind-belts appear to be inadequate. It was here that certain cultural and spraying experiments, designed with a view to suppressing attacks of scale imsects, were inaugurated am 1915. Mean Annual Rainfall—58 inches. Xylem Cylinder. This section resembles that just described from the Grove Botanic Station, Montserrat, in the diameter and radial distribution of the vessels. It differs, however, in one respect. The parenchyma bands are more sharply delimited, and do not oceur in groups along the radu. If it be again assumed that these bands of tangential parenchyma originate during periods of pronounced desic- cation, it may possibly be inferred from their more irregular distribution that the activity of the cambium was checked at irregular intervals in the course of the year, and not alone for a short period during the dry season. It is also necessary to assume that these checks in growth were of greater duration than in the case of the Grove Botanic Station. Tt will be observed that Hardy states that the situation is exposed to winds. It is quite conceivable, therefore—for the conductivity of the wood is evidently small—that growth may have been checked at any period in the course of the year, even though the water-supplying power (5) (7) of the soil was adequate. OLVESTON, MONTSERRAT (FIG. 7, PLATE XV). Lime trees exhibiting die-back. This locality is situated on the gently sloping side of an exposed low hill near the sea-coast. The soil is very shallow and rocky. The lime trees have been planted chiefly in pockets of soil. At the time of sampling (April, 1921) the sow was very dry and compact. Mean Annual Rainfall—60 inches. Xylem Cylinder. Tn this section there is little or no regularity exhibited in the zonation. It is possible to infer only one thing, namelv, that the activity of the cambium was repeatedly interrupted, presumably as a result of irregularly recurring periods of desiccation, due, no doubt, to the exposed position of the field. SauToun, 9-AcRE PLoT, Dominica (Fic. 8, Puate XVI). Lime trees exhibiting die-back. The soil at Saltoun is shallow, and overlies sheet rock. The subsoil is a 260 Scientific Proceedings, Royal Dublin Society. yellow clay. The drainage 1s mainly superficial, the land sloping wm the direction of a natural water-course. Asa result of a very high ramfall, the sow has been subjected to leaching and washing. : It is markedly acidic in reaction (pH £7), and is almost continuously water- logged. Mean Annual Rainfall—200 inches. Vylem Cylinder. The zonation here is essentially similar to that in the former section. The erowth otf the eambium has evidently received repeated checks. The conductivity ot the wood is small. The soil in which the trees were rooted was indubitably physiologically dry for the greater part of the year as a result of a low oxygen- supplying power, and, possibly, the presence of toxic substances, and not, of course, from an excess of soluble salts. It is probable that the numerous irregularly distributed parenchyma bands are due to the drying blast of the Trade Winds. O’GarRa’s CoasTan Region, Montserrat (F1g. 9, PLATE XVI.) LInme trees exhibrting die-back. This locality is situated on low ground near the extremity of a steep talus slope. The soil rs somewhat shallow and stony, with pockets between the rocks. [t was dry and compacted at the time of sampling (April, 1921), and tts weed flora was withered. Under drainage appears not to be good, for the presence of nut grass (Cyperus rotundus) and Commelina nudiflora, growing amongst the xerophytic short-lived annuals that make the natural plant association, indicates that the land is subjected to periodic water-logging. Mean Annual Rainfall—43 inches. Xylem Cylinder. There is here a very marked differentiation into zones simulating ‘‘spring”’ and ‘‘autumn’’ wood. It will be observed how very sharply these zones are delimited from one another. It will be evident that the rate of growth must have been checked with great rapidity. This is possibly the result of the periodic water-logging to which Hardy refers. Conclusions. The interpretation of the zonation exhibited by the woody cylinder of the shoot of an arborescent evergreen, particularly in a tropical climate, is evidently a task beset with no little hazard. As yet, it would seem, the problem has never been approached experimentally. It is but natural that this should be so, for the prosecution of work of this nature would call for resources and equipment at present possessed by no tropical, and few northern, botanical departments. It has been generally assumed by paleontologists that rings of growth in the stems of fossil plants indicate the existence of a Temperate Zone climate and their absence a more equable one. In the woody eylinder of the shoot from the Lime Experiment Station of Dominica, where the moisture conditions are more generally favourable for growth than in any of the other habitats visited in the course of this survey, zonation is scarcely appreciable. It is indicated merely by circumferential areas, where the large vessels are somewhat more densely distributed than elsewhere, and within which ill-defined tangential bands of parenchyma occur. This grouping of the vessels, which culminates in some PLATE XVI. SCIENT, PROC. R. DUBLIN 'SOC., VOL. XVII. Jens Bo Fig. 8. Mason. Ds Us eis) Relhean as Mason—Ligneous Zonation and Die-Back in the Lime, West Indies. 261 of the sections in areas presenting all the characters of ‘‘spring wood,’’ is evidently not especially associated with the absence of the die-back condition, for it is particularly marked in the wood from the coastal region of O’Garra’s, where the trees were badly affected. Inasmuch as nothing definite is known coneerning the factors which determine the anatomical differences between “‘spring’”’ and ‘‘autumn’’ wood, even in temperate regions, it would be obviously unprofitable to attempt to discuss this matter further. It will be also clear that the diameter of the vessels is subject to considerable variability, and that, though it is generally greater in healthy trees than in those showing the die-back condition, yet no significant relationship is to be traced. It was suggested in the preceding section that the parenchyma bands originated during periods of desiccation. Jeffrey (3) has pointed out that the impulse towards production of terminal parenchyma was probably supplied in the past by climatic cooling. It would seem that the impulse needed for the production of tangential parenchyma may also be provided by desiccation of the cambium. It is not improbable, of course, that the production of parenchyma as a result of climatic cooling is in the last analysis effected by desiccation. Tt will be evident that, if the interpretation of the zonation adopted is correct, namely, that the zones of vessels correspond to periods during which the plant’s water-supply was adequate, and that the bands of parenchyma were conditioned by desiccation of the cambium, their frequent association must indicate very rapid changes in the aridity of the environment. Rather sudden fluctuations do, in fact, characterize the dry season. ‘There is generally at this time a bare sufficiency of moisture for growth. It may be suggested that the production of tangential parenchyma is due to autogenic changes in the activity of the cambium, or that it is possibly related to the cireumvasal parenchyma, but there would seem to be no foundation for such a view. It will be recalled that the parenchyma bands were generally more pro- nounced and more irregular in their radial distribution in those habitats in which the aridity of the environment fluctuated most markedly. It is in these habitats that the die-back condition is most conspicuous. It would seem that the rapid and repeated desiccation of the meristems, which is recorded in the ease of the cambium by tangential parenchyma, results in the premature loss of dominance of the apical over the lateral buds, as a result of which daughter shoots grow out and replace the mother shoot. A repeated dying back of shoots of this nature is characteristic of trees affected with die-back. The growth of the Lime in the West Indian islands of: St. Vincent and Carriacou is of some interest in this connexion. The islands lie only some fifty miles apart, but differ very markedly in two respects. The annual rainfall in St. Vincent is about 100 inches, and the soil is shallow and porous, drying out with great rapidity whenever a couple of weeks pass without rain. The annual precipitation in Carriacou, on the other hand, is only about 40 inches, but the soil is very retentive, and dries out slowly. The consequence is that in the dry season St. Vincent undergoes rapid though normally short-lived spells of drought, whereas in Carriacou drought conditions are initiated relatively slowly as a result of the retentive nature of its soil, but are frequently prolonged. Now the Lime cannot be grown in St. Vincent, though repeated attempts have been made to do so, while it thrives in Carriacou, where the aridity is many times greater. It would seem that this plant can tolerate considerable aridity and still thrive, but that it is unable to withstand a rapid desiccation of its tissues. The writer is unaware of any other hypothesis that so well accords 262 Scientific Proceedings, Royal Dublin Society. with the actual distribution of Lime cultivation in the West Indies. That the tree can thrive under a wide range of; relatively static edaphic conditions (hydrogen ion concentration, ete.) has been convincingly shown by Hardy (2). There can be little doubt that the rate, rather than the amount or duration of desiccation is im many cases of extreme importance in determining the distri- bution of plants. It is eee that this is one reason why eertain plants will erow vigorously in clay soils and fail completely where the colloid content is small. Tn conclusion, the writer wishes to express his indehtedness to Mr. Robson, in Montserrat, and Mr. Keys, in Dominica, for makine observations on the periodicity in the growth of the Lime tree. Summary. 1. In this paper the results of an examination of the ligneous zonation of a number of Lime shoots from trees erowing in different parts of the West Indian islands of Dominica and Montserrat are considered. Half of the shoots were collected from trees affected with die-back. 2. The zonation in the woody eylinders from healthy trees indicates a rather definite periodicity in the activity of the eambium. Tangential bands of parenchyma are generally distributed within the more porous zone of vessels. 3. Inasmuch as the Lime tree in these islands makes its vegetative growth during the dry season, it was inferred that the zones of vessels registered the production of leaves at this period, while the water-supply was adequate, and that the tangential parenchyma, which is distributed within these zones originated at periods of relatively great eridity. Rapid fluctuations in the aridity of the environment, which the association of the porous vessels and the circumferential parenchyma indicates, are characteristic of the climate in the dry season. 4. The wood of shoots from trees affected with die-back exhibited considerable irreeularity in the distribution of the parenchyma bands. All the sections suggested that the eambium had been exposed to sudden checks in its activity. 5. It was tentatively concluded that an important factor in causing the dying back of the shoots was rapid and repeated desiccation of the meristems. In the terminal meristem this resulted in a premature loss of the dominance of the apical bud, and its replacement by daughter shoots, which in turn suffered the same fate; and in the eambium by the production of tangential parenchyma. LITERATURE CITED. 1. Drxon, H. H., 1914.-- Transpiration and the Ascent of Sap in Plants. London. : 2. Harpy, F., 1922.—Studies in West Indian Soils. West Indian Bul. xiv, 2, 189-213. 3. Jerrrey, EH. C., 1917.—The Anatomy of Woody Plants. Chicago. 4. Livineston, B. E., and Hawkins, I. A., 1915.—The Water Relation between Plant and Soil. Carnegie Inst. of Washineton, Pub. 204. 5. Livineston, B. E., and Koxrrsu, R., 1920.—The Water-Supplying Power of the Soil as related to the Wilting of Plants. Soil Science, 9, 469-485. 6. Mason, T. G., 1921—The Water Balance of the Plant and its Significance in Crop Production. West Indian Bul. xviii, 4, 157-184. 7. Mason, T. G., 1922.—The Physiological Humidity of the Soil and its Direct Determination. West Indian Bul. xix, 2, 137-155, 8. Nowrtu, W.—|[In the Press. | 3 10. 11. 12. 13. 14, 15. SCIENTIFIC PROCEEDINGS. VOLUME XVII. . Experiments on the Electrification produced by Breaking up Water, with Special Application to Simpson’s Theory of the Electricity of Thunder- storms. By Professor J. J. Nouan, .a., D.sc., and J. Hwricut, B.A., M.Sc., University College, Dublin. (June, 1922.) . Cataphoresis of Air-Bubbles in Various Liquids. By Tuomas A. McLavennim, M.SO., A.InsT.P., University College, Galway. (June, 1922.) On the Aeration of Quiescent Columns of Distilled Water and of Solutions of Sodium Chloride. By Professor W. H. ADENEY, A.R.C.SC.1., D.SO., F.1.C. ; Dr. A. G. G. Lonarp, F.8.¢.SC.., B.SC., Fa.c.; and A. RioHaRDson, A.R.O.SC.1., A.C. (June, 1922.) . Ona Phytophthora Parasitic on Apples which has both Amphigynous and Paragynous Antheridia; and on Allied Species which show the same Phenomenon. By H. A. Larrerty and Grorce H. Prruypriner, Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (Plates land Il.) (June, 1922.) . Some Further Notes on the Distribution of Activity in Radium Therapy. By H. H. Poors, m.a., sc.p., Chief Scientific Officer, Royal Dublin Society. (June, 1922.) Preliminary Experiments on a Chemical Method of Separating the Isotopes of Lead. By Tsomas Ditnon, p.sc.; Rosatinp Chars, p.sc.; and Vioror M. Hincay, s.sc. (Chemical Department, University College, Galway). (July, 1922.) . The Lignite of Washing Bay, Co. 'yrone. By ‘I’. Jounson, D.so., F.L.s., Professor of Botany, Royal College of Science for Ireland ; and Janz G. Gitmore, B.sc. (Plate III.) (August, 1922.) . Libocedrus and its Cone in the Irish Tertiary. By T. Jounson, p.sc., r.u.s., Professor of Botany, Royal College of Science for Ireland; and Jann G. GinmorzE, B.sc. (Plate lV.) (August, 1922.) . The Klectrical Design of A.C. High Tension Transmission Lines. By H. H. Jzerroorr. (August, 1922.) The Occurrence of Helium in the Boiling Well at St. Edmundsbury, Lucan. By A. G. G. Lonarp, F.R.¢.sc.1., PH.D., F.t.c., and A. M. Ricwarpson, A.R.O.SC.1., A.C. (Plate V.) (August, 1922.) [Nos. 1 to 10, price 9s.] On the Detonating Action of a Particles. By H. H. Pootn, m.a., so.n., Chief Scientific Officer, Royal Dublin Society. (December, 1922.) The Variations of Milk Yield with the Cow’s Age and the Length of the Lactation Period. By Jamms Wison, M.a., B.sc. (December, 1922.) A Note on Growth and the Transport of Organic Substances in Bitter Cassava (Manihot utilissima). By 'T. G. Mason, m.a., B.sc. (December, 1922.) [Nos. 11 to 18, price 1s. 6d.] The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylurethane. By Hue Ryan, p.sc., and Annz Donnetnan, .sc., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on Rthyl-o-Tolyl- urethane. By Hue Ryan, p.sc., and Nuicuonas CoLLINANE, PH.D., University College, Dublin. (February, 1923.) 16. 17. SCIENTIFIC PRO CEEDINGS—continued. The Action of the Oxides and the Oxyacids of Nitrogen on HWthyl-Phenyl- urethane. By Hueu Ryan, p.sc., and Anna Connozty, u.sc., University College, Dublin. (February, 1925.) The Action of the Oxides and the Oxyacids of Nitrogen on Phenyl- Decne ~» -urethane. By Huen Ryan, v.sc., and James L. O'Donovan, u.se., 18. 19. 20. al. 22. 23. 30. bl. University College, Dublin. (Hebruary, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on the Phenylureas. By Hueu Ryan, p.sc, and Parmr K. O’Toorz, u.sc., University College, Dublin. (February, 1928.) The Action of the Oxides and. the Oxyacids of Nitrogen on Phenyl- Methyl. urea. > By Huen Ryan, -p.sc., and Micuant J. Swneney, .sc., University College, Dublin. (February, 1928.) [Nos. 14 to 19, price 4s.] On the Cause of Rollmg in Potato Foliage; and on some further Insect Carriers of tlie Leaf-roll Disease. By Paun A. Murpry, sc.b., a.R.C.S6.1., Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (Plate VI.) (May, 1923.) On the Channels of Transport from the Storage Organs of the Seedlings of Lodoicea, Phenix, and Vicia. By Hrnry H. Drxov, sc.., r.n.s., Professor of Botany in the University of Dublin; and Niern G. Batt, u.a., Assistant to the Professor of Botany in the University of Dublin. (Plates VII-X1.) (June, 1923.) Irregularities in the Rate of Solution of Oxygen by Water. By H. G. Brcxnr, A.R.G.8C.1., A.1.c., Demonstrator in Chemistry in the College of Science, Dublin; and EH. F. Pearson, a.r.c.sc.1., Research Student. (June, 1923.) The Hydrogen Ion Concentration of the Soil in relation to the Flower Colour of Hydrangea Hortensis W., and the Availability-of Iron. By W. Rh. G. ATKINS, 0.B.E., SC.D., F.1.c. (June, 1923.) . The Comparative Values of Protein, Fat, and Carbohydrate for the Production of Milk Fat. By EH. J. Suuuny, r.x.c.sc.1., B.Sc. (Hons.), M.R.1.A., Biochemical Laboratory, D.A.T.1. (June, 1923.) [Nos. 20 to 24, price 7s. 6d. | . The Utilisation of Monomethylaniline in the Production of Tetryl. By Tuomas JosrpH Nonan, D.sc¢., F.1.c., and Henry W. Cuapsam, Nobel Research Laboratories, Ardeer. (Communicated by Prof. H. Ryan.) | (July, 1923.) . Hyidence of Displacement of Carboniferous Strata, Co. Sligo. . By: Arruur HW. Cuarx, B.a., Trinity College, Dublin. (Communicated by Mr. L. B. Smyru.) (July, 1923.) : On a Problematic Structure im the Oldhamia Rocks of Bray Head, County Wicklow. By Louis B. Suyzu, m.a., sc.s. (Plate XII.) (July, 1923.) . The Hydrogen Ion Concentration of the Soil and of Natural Waters m relation to the Distribution of Snails. By W. R. G. ArxKmns, 0.B.z., Sc.D., r.ic., and M. V. Lezour, p.sc. (July, 1923.) ! . Improved Methods of Hvyaporation in the Laboratory. By H. G. Broxsr, A.R.0.80.1., A.1.c., Demonstrator in Chemistry, College of Science, Dublin. (August, 1923.) A Rapid Gasometric Method of Estimating Dissolved Oxygen and Nitrogen in Water. By H. G. Broxer, a.r.c.scu., atc., and W. EH. Asgorr, A.R.C.SC.1., A.I.C., B.Sc. (August, 1923.) Ligneous Zonation and Die-Back in the Lime (Citrus medica, var. Acida) im the West Indies. By T. G. Mason, .a., so.p., Botanist, West Indian Agricultural College. (Plates XIII-XVI.) (August, 1923.) [Nos. 25 to 31, price 6s. 6d. | DUNLIN : PRINTED AT THE UNIVERSITY PKESS BY PONSONBY AND GIBBS, . Sy QO 6 i Lyf SCIENTIFIC PROCEEDINGS ROYAL DUBLIN SOCIETY. Vol. XVII, N.S., Nos. 32-41. MARCH, 1924. 32.—ON THE EXTRACTION OF SAP FROM LIVING LEAVES BY MEANS OF COMPRESSED AIR. By Henry H. Drxon, Sc.D., F.R.S., Professor of Botany in the University of Dublin; and Nicet G. Batt, M.A., Assistant to the Professor of Botany in the University of Dublin. 33.—_SOME EXPERIMENTS ON THE CONVECTION OF HEAT IN VERTICAL WATER COLUMNS. By H. H. Poon, Sc.D. 34.—ON THE SUPPOSED HOMOLOGY OF THE GOLGI ELEMENTS OF THE MAM- MALIAN NERVE CELL, AND THE NEBENKERN BATONETTES OF THE GENITAL CELLS OF INVERTEBRATES. By F. W. Rogers Brampenn, B.A., Sc.B. (DuBL.); and J. BrontT# GavTENBy, M.A. (OMEN) D.PuHIn. (Oxon.),; D.Sc. (Lonp.). (Plate XVII.) 35.—PHOTOTROPIC MOVEMENTS OF LHEAVES.—THE FUNCTIONS OF THE LAMINA AND THE PETIOLE WITH REGARD TO THE PERCEPTION OF THE STIMULUS. By Nicrn G. Batt, M.A., Assistant to the Professor of Botany in the University of Dublin. 36.—THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON PHENYLBENZYLETHER. By HueH Ryan, D.Sc.; and JOHN KEANE, PH.D., University College, Dublin. 37.—THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON ETHYL-g-NAPHTHYLETHER. By HucH Ryan, D.Sc.; and JoHN KEANE, PH.D., University College, Dublin. 38.—_ THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON DIPHENYLETHYLENEETHER. By HucH Ryan, D.Sc.; and TERENCE Kunny, M.Sc., University College, Dublin. ; 39.—THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON DIPHENYLETHER. By Huee Ryan, D.Sc.; and Peter J. Drum, M.Sc, University College, Dublin. 40.—THE ACTION OF THE OXIDES AND THE OXYACIDS OF NITROGEN ON DIPHENYLENE OXIDE. By Hucu Ryan, D.Sc.; and NicHonas CULLINANE, Pu.D., University College, Dublin. 41.— THE HABITATS OF LIMN AEA TRUNCATULA AND ZL. PEREGER IN RELATION TO HYDROGEN ION CONCENTRATION. By W. RB. G. Arxins, O.B.E., F.LC.; and Mariz V. Lesour, D.Sc., Marine Biological Laboratory, Plymouth. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAI, DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, manque \> Y USS ap aa , HENRIETTA STREET, COVENT GAKDEN, LOxvoN, WO) Gg; ‘ / a ees (SUNT P1904 31 : Ye é Price Five Shillings and Sixpence. tig nal M sewn Roval Wublin Society, ~ FOUNDED, A.D. 17381. INCORPORATED 1749. ee 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 Mconomie 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 eslomate 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 iJgonomie Proceedings of the Royal Dublin Society, as directed by the Committee of Science. The copyright of papers so puislisined 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 HWxecutive Officer for Science for further information concerning the preparation of papers and illustrations for press. fen SCIENTIFIC PROCEEDINGS. VOLUME XVII. 1. Experiments on the Electrification produced by Breaking up Water, with Special Application to Simpson’s Theory of the Hlectricity of 'hunder- storms. By Professor J. J. Nouan, m.a., p.sc.,and J. ENRIGHT, B.A., M.SO., University College, Dublin. (June, 1922.) 2. Cataphoresis of Air-Bubbles in Various Liquids. By Tuomas A. Mchauesum, M.SO., A.INSsT.P., University College, Galway. (June, 1922.) 3. On the Aeration of Quiescent Columns of Distilled Water and of Solutions of Sodium Chloride. By Professor W. EH. ApEnny, A.R.0.SC.1., D.SO., F.1.C. ; Dr. A. G. G. Lronarp, F.R.0.8c.1., B.Sc., F.1.0.; and A. RicHarpson, A.R.C.S0.J., Arc. (June, 1922.) [Continued on p. 3 of cover. (2630 ayn No. 32. ON THE EXTRACTION OF SAP FROM LIVING LEAVES BY MEANS OF COMPRESSED AIR. By HENRY H. DIXON, Sc.D., F.B.S., Professor of Botany in the University of Dublin, AND NIGEL G. BALL, M.A., Assistant to the Professor of Botany in the University of Dublin. [Read NovEMBER 27. Printed DECEMBER 7, 1923.] Ir has already been pointed out (1 and 2) that there are many reasons which favour the suggestion that the xylem provides the conducting channel for the transport of organic substances in plants. According to this theory a backward current of water containing dissolved carbohydrates, etc., must pass, either continually or intermittently, from the leaves to places of growth and storage. By the application of a method which was used by one of us (8) a long time ago for the direct measurement of osmotic pressure in plants, it was found possible to cause a backward current to flow in the xylem from the leaves towards the lower parts of the plant. It seemed desirable, therefore, to test whether this current, although produced artificially, would be utilised by the plant for the purpose of transporting carbohydrates. The apparatus used is represented in the diagram. It consisted of a thick- walled cylinder of specially annealed glass about 40 em. long by 12 em. external diameter. (Owing to the bursting of this cylinder it was subsequently replaced by a similar one made of mild steel.) The ends of this cylinder were closed by strong castings secured by three long bolts, the joints between the cylinder and the castings being made air-tight by means of leather washers impregnated with wax. The cut end of the branch was passed through a tubulure in the lower casting, the joimt being secured by thick rubber tubing wired on and covered with glue. Compressed air, supplied through a lead tube of narrow bore, was admitted at the top of the cylinder. In order to obtain the necessary pressure a pump was used in the earlier experiments, but later this was replaced by a cylinder of compressed air. The pressures were measured by means of a Bourdon pressure gauge. The experiments were carried out as follows:—a small branch was fitted with its cut end protruding through the lower casting. The cylinder was then assembled and bolted up tightly, the waxed-leather washers being previously rendered soft and limber by heating. The compressed air was then admitted. At a pressure of from 5 to 8 atmospheres drops began to appear at the cut end. By gradually increasing the pressure to a maximum of 20 atmospheres something more than 3-5 ¢.c. of liquid could be obtained from a small branch. SCIENT. PROC, R.D.S., VOL. XVII, NO. 32. 30 264 Scientific Proceedings, Royal Dublin Society. July, 1922. Branches of Sambucus mgra and of Tilia americana were subjected to pressures up to 15 atmospheres. From 1 to 3 ec. of liquid was obtained from each branch; but when boiled with Benedict’s solution this liquid, whether inverted with hydrochloric acid or not, showed complete absence of sugars. Some of the branches were kept in the dark for 24 hours before the experiment, but this did not affect the result. | WELLL (WILLE September, 1922. Branches of Tilia americana were again subjected to pressure. The liquid obtained gave a very slight reduction with Benedict’s solution. The amount of reduction was not increased by previous inversion of the sample with hydrochloric acid. Dixon anp Batt— Extraction of Sap from Living Leaves. 265 In one experiment a branch weighing 42 g. was exposed to 20 atmospheres for 20 minutes, i.e. until no more liquid was obtained. The amount of liquid yielded was 2:25 ¢.c. The pressure was then reduced and a few c.c. of toluene was introduced into the cylinder. After 14 hours a pressure of 20 atmospheres was again applied and continued until no more liquid was obtained. The amount yielded this time was 65 ¢c. This liquid was brown in colour and, after Inversion with invertase, was titrated with Benedict’s solution. About 5 per cent. of sugar was found to be present. The branch was then removed and dried. Its dry weight was found to be 17 g. The amount of liquid extracted represented therefore about 35 per cent. of the total amount of water in the branch. June, 1923. An experiment with Z%lia americana gave a similar result to that obtained in September, 1922. It seemed possible that the liquid, which was obtained before the cells were rendered permeable with toluene, represented only that which was present in the wood at the time of the experiment, being forced out by the sap pressed from the leaf cells. To obtain this latter with as little dilution as possible the following arrangement was made:—Three small branches supporting together 18 large and 4 small leaves were fixed in the lower casting and enclosed together in the apparatus. From these about 3 ¢.c. were extracted under a pressure of 20 atmospheres. The volume of the wood, excluding the negligible amount in the veins of the leaves, was estimated by making cross- sections of the twigs at various levels, and found to be about 2 cc. About 50 per cent. of this volume would be represented by the walls of the tracheae; so that about 2 ce. of the liquid obtained must have come from the cells of the leaves. The first few drops of the liquid which were collected showed the presence of a trace of reducing sugars, due probably to contamination from the cut surface of the twigs; but the remainder gave no reduction with Benedict’s solution either before or after inversion with hydrochloric acid. It seems clearly established by these experiments that, even when the cells of the leaves contain considerable quantities of sugar, no part of this sugar can be driven by external pressure from the tracheae of the supporting branch. This may be due to the continued impermeability of the leaf cells under the conditions of the experiment, or possibly the cells in contact with the conducting channels at lower levels extract what sugars pass into these channels from the leaves, and prevent the carbohydrates appearing at the cut surface of the stem. In order that a backward transport of organic substances from the leaves to the stem may take place through the wood, two conditions must be fulfilled :— (1) The cells, owing to internal or external causes, must be rendered permeable ; (2) a backward flow must take place through the tracheae. It is probable that the second condition. would follow as a result of the first, since then any tension existing in the tracheae would be no longer resisted by the osmotic pressure of the leaf cells. The experiments described above seem to show, however, that the first condition is not of necessity a result of the second. At the present state of our knowledge it would be useless to speculate on the underlying causes which, under normal conditions, effect changes in the permeability of plant cells. No matter what theory regarding the mechanism of transport be held, it seems to be necessary to postulate the occurrence, either gradually or suddenly, of such changes. It is probable that further progress towards the elucidation of the problem of the transport of organic substances 266 Scientific Proceedings, Royal Dublin Society. in plants can only be made when the conditions governing the permeability of plant cells are better understood. SUMMARY. Branches of Tilia americana and Sambucus nigra were enclosed in a strong eylinder in such a way that their cut ends protruded. Compressed air at pressures up to 20 atmospheres was admitted into the cylinder, and the liquid which exuded from the cut end of the branch was collected. This liquid was found to be completely, or almost completely, free from sugars. Experiments carried out in early and late summer gave similar results. After the leaf cells had been made permeable by means of toluene vapour the sugar in the expressed sap amounted to about 5 per cent. REFERENCES. 1. Drxon, Henry H.—The transport of Organic Substances in Plants. Presidential Address, Section K, British Association, 1922. 2. Drxon, Henry H. and Batu, Nice G.—Transport of Organic Substances in Plants. Nature, Feb. 23, 1922. 3. Dixon, Henry H.—On the Osmotic Pressure in the Cells of Leaves. Proce. Royal Ivish Academy, 1896, Vol. IV, ser. 3, p. 61. No. 33. SOME EXPERIMENTS ON THE CONVECTION OF HEAT IN VERTICAL WATER COLUMNS. By H. H. POOLE, Sc.D. [Read Novremser 27. Printed Drcrmprr 7, 1923.] Introduction. In view of the importance of radioactive heating in modern geological theories, considerable interest attaches to the question as to how far the estimated upward flow of heat through the earth’s crust accords with that to be expected from the known radioactivities of the surface materials. The figure obtained by multiplying the thermal conductivity of dry rock by the average temperature gradient just below the surface is surprisingly small. The question has recently been discussed by the writer elsewhere.t It would appear that there are several factors tending to reduce the gradient near the surface, one of these being the effect of underground water, which may bring heat to the surface either by means of thermal springs rising from considerable depths or by convection currents in ‘‘stagnant’’ water in pores, pockets, or fissures in the rocks, thus adding to the effective thermal conductivity of water-logged strata. An account is here given of some experiments on the convection of heat in vertical water columns, made with a view to estimating, if possible, the probable magnitude of the heat flow in stagnant water. As might be expected, the effect appears to be extremely complex, depending greatly on the size, and especially on the diameter, of the column. Experimental Details. Fig. 1 represents in section the general outlines of the apparatus used, the scale bemg 1/5. The column of water under test was contained in a glass tube A, open at both ends, connecting a copper vessel B, about 6 em. deep by 4 em. in diameter, with an open upper vessel C, the cork stoppers through which A passed being rendered water-tight with paraffin wax. The outer cylindrical surface of B was uniformly covered with a_ heating winding of silk-covered 28 S.W.G. Nichrome wire, wound directly on the vessel, and covered with molten paraffin wax. This winding extended from near the bottom of the vessel to with 1 cm. of the lower surface of the cork. Inside the vessel were placed 24 copper rods, each 5 em. by 0-4 em., 16 of them being approximately vertical, reaching from the top to the bottom of the water space, near the wall of the vessel, and the remaining S crossing it diagonally at an angle of about 45° to the vertical. This arrangement of rods, which is not shown in the figure, left a conical region just below A unobstructed, while ensuring that the water im it was at an approximately uniform temperature. CO was kept cool by a stream of tap-water 1Pphil. Mag., September, 1923. SCIENT. PROC. R.D.S., VOL. XVI, NO. 33. 30D 268 Scientifie Proceedings, Royal Dublin Society. Wy LU N Soop eooSsssesaocosccs Se ee NS SSNS LWA SERS SSN Fig. 1. Poor.e—On the Convection of Heat in Vertical Water Columns. 269 flowing through a flat metal worm immersed in it. In most of the experiments, the water, before entering the worm, passed through a coil of metal tubing immersed in the lower part of the large water-bath ZL. A and B were enclosed in an inverted bell-jar D, the intervening space being filled with granulated cork, and B further screened by a Dewar vessel, as shown in the figure. Metal ballast in the bottom and a few small weights on top of the cork kept D floating vertically in the water-bath, which was enclosed in a barrel F’, the intervening space being filled with granulated cork. The worm tube in C was independently supported. The difference of temperature between the ends of A, ranging from a small fraction of a degree to about 10 degrees, was read by a thermo-couple, whose junctions were enclosed in thin glass tubes passing through the annular corks into B and C respectively, so that the junctions were level with and close to the ends of A. The temperatures of C and H were read by a pair of accurate mereury thermometers, one being independently supported with its bulb close to the upper thermo-junction, and the other floating in H with its bulb rather below the level of B. No stirring was used, as it was desirable to work with the water in A as quiescent as possible, but even so the water in E was generally at a very nearly uniform and constant temperature throughout a test. This arrangement was almost entirely made up of apparatus which chanced to be available. A thermostat would obviously have been an improvement, but a rather complicated system of temperature control would have been necessary, as it was desirable to work below 20° C., and considerable quantities of heat had to be got rid of. Even if C and # were maintained at constant temperatures, the mean temperature of the water in A would vary with the heat supplied to its base, so, to keep the latter temperature constant, C should he cooled as the power supply was increased. In view of the essentially rough character of the tests such a complication did not appear to be justified. The thermo-couple consisted of 5 pairs of iron and constantan wires, the total resistance being 85 ohms, and was connected through a Morse telegraph key to a galvanometer of resistance about S00 ohms, the key being arranged so that the galvanometer was normally short-circuited. The couple was inserted in the cireuit by depressing the key, which was enclosed in a wooden box and operated by a long vuleanite handle to reduce thermal effects. The zero was re-set, when necessary, by moving the galvanometer lamp, the scale being fixed. In this way thermal effects at the galvanometer were almost completely eliminated. It was found that leakage from the battery supplying the heating current might cause a shift of, a few divisions in the zero, but did not appreciably jaffect the deflection produced by depressing the key. This was tested by reversing the heating current, and also by momentarily disconnecting the battery. In the later experiments the zero shift was entirely eliminated by earthing one end of the heating winding, which was electrically connected to the vessel B. Identical results were obtained whether this earth connexion was made or not. The couple and galvanometer were calibrated throughout the scale of 500 divisions by comparison with the two mereury thermometers, these, in turn, being compared with a standard. The deflection per degree varied from 32:9 seale divisions for small deflections to 32:0 for large ones, and was sensibly independent of the mean temperature of the couple over the range (10°C. to 25° C.) covered. The current and the P.D. across the heating coil were both measured, and it was found that over a large part of the range of power used, the resistance 270 Scientific Proceedings, Royal Dublin Society. of the coil remained sensibly constant. The power supplied ranged from 0-04 watt to 25 watts. Two sets of tests were carried out with the heating and cooling vessels filled with water, but the connecting tube, or tubes, corked and empty. First the initial rate of rise of temperature was noted when a known power was turned on, so that the effective thermal capacity of the heating vessel and its contents could be found. This allowed a small correction to be applied, when necessary, for any slight changes of temperature during a convection test. Secondly, by noting the final temperature attained by the heating vessel for a given power, it was found that the loss of heat through the cork and the vacuum vessel was proportional to the difference of temperature between the heating vessel and the surrounding water-bath. A correction for this loss was thus made. This was also small, except in cases of low power supply with a narrow tube, when it became very serious. The stopper was then removed from the top of the tube, which was filled with water, and the convection tests started. The usual procedure was to turn on a certain current, start the cooling water flowimg in the worms, and leave the apparatus for several hours, until a steady temperature had been attained. A series of readings of the voltmeter, ammeter, galvanometer, and thermometers over a period of 8 to 10 minutes was then made. From these the rate of flow of heat up the tube for a given temperature difference was found. In allowing for the effect of small variations in temperature the effective thermal capacity of the heating vessel and contents was taken as that already found with the convection tube empty plus half the volume of the tube. The most satisfactory results were those obtained towards the middle of the power range. For small powers the correction for heat loss becomes very important ‘and somewhat uncertain, as a very appreciable part of the heat flowing up the lower part of the convection tube must escape laterally through the cork jacket. On the other hand, near the top of the range the convection currents sometimes became unsteady, causing the temperature difference indicated by the couple to vary rapidly over a range of perhaps 20 per cent. Five sets of tests were carried out with single columns of various diameters and lengths, and three sets with double columns, consisting of two parallel glass tubes of as nearly as possible the same dimensions, with their ends at the same level. In these cases circulation almost certainly occurred, warm water rising in one tube and cool water descending in the other. The tubes were separated by about 5 mm. of cork, which would form a fairly efficient thermal insulation. The last pair of tubes tested beimg 28 em. in diameter, necessitated the use of larger heating and cooling vessels, a larger Dewar vessel, and a larger floating vessel. In this case the heating coil was wound in several sections in parallel, the joint resistance being 9 ‘ohms, and the power range from 0:4 watt to 66 watts. As it was not possible to carry out all the tests with the same mean temperature of the convecting column, some experiments were made with heated water flowing through the worms in the water-bath and the cooling vessel. In this way the average temperature of the convecting water was raised without inereasing the temperature gradient in it. The tests were in general less satisfactory than those with cold circulating water, as the temperatures were more variable. With a single column the rather surprising result was obtained that the heat flow for a given gradient did not vary very much for a range of mean temperature from 135° to 235°C. There was some Poo1.n—On the Convection of Heat in Vertical Water Columns. 271 evidence of a maximum at about 17°C. followed by a fall for higher temperatures. With double columns the results were much more consistent, the heat flow for a given gradient being approximately proportional to the excess of the mean temperature above 4° C. We should expect, at first sight, that the flow of heat should increase with the temperature at about this rate, owing to the increase in the coefficient of expansion and the decrease in the viscosity. The contrary results obtained with the single column are probably to be ascribed to the increased mixing of the ascending and descending currents with rise of temperature. In the case of the double tube this mixing cannot occur. SYVELE TUBE 22:-8em XK 2-82cm 2 Dey, 25 Ocm X 2OTc» 23 -Len X ¢-O7 077 10202 X 2:77 CI 10: 2$¢tm X/-O$F cp? LOG GC Dre. 2 Results. The results for single columns are plotted in fig. 2, and those for double columns in fig. 38, the lengths and diameters being shown on the figures. Since it was desired to estimate the probable magnitude of the effect of convection 272 Scientific Proceedings, Royal Dublin Society. on the thermal conductivity of the earth’s crust, it appeared to be most convenient to state the results in terms of a quantity analogous to the conduc- tivity of a solid. This quantity, which may be ealled the ‘‘convectivity,’’ is denoted by the letter C in the figures, and is defined as the heat flow per sq. em. of total horizontal area per sec. divided by the gradient G. As, however, the extreme range of C in the plotted results is from 5 X 10° to over 10, and that of G@ from 5 X 10-° to over 1, all the results could not conveniently be plotted directly on one scale, so their logarithms are plotted instead. / S S S 2 N (0) © (Sans) y) © fe) (C) (O} O DoveBLé TUBE 22-8tm % 2-80 ci. fe) >? 2? 25:Ocm X 1:08 crn. fo) apy, 2? 10-Stro X 1:10CIM- -/ ZOG.C 5) -/ (8) Fic. 3. In the case of the results shown in fig. 2 no correction has been made for the mean temperature of the water column, as the effect of variations of a few degrees at about air temperature does not seem to be very important for single columns. The general average of the mean temperatures for these results may be taken as about 15° C., the temperature generally increasing with increase of gradient. The results for the double columns shown in fig. 3 have all been reduced to a mean temperature of 17° C. by assuming that the convectivity is proportional to the excess of the mean temperature above 4°C. As unavoidable small differences occurred in the dimensions of the glass tubes, the figures given are, in each ease, the means for the pair. It will be observed that the convectivity generally rises rapidly with rise in gradient, the effect becoming more and more marked as the diameter is Poote—On the Convection of Heat in Verticas Water Columns. 273 reduced. The length of the column has also a great effect on the convectivity for a given gradient; in fact, the flow of heat up a tube of given diameter, for a given temperature difference between the ends, was generally about the same for short and long tubes, in spite of the fact that with the former the gradient was much greater. Since the gradient in the earth’s crust is about 3 X 10* we have to extrapolate the results down to a value for logG@ of — 352. It is evident that no reliance could be placed on the actual figures thus found, but they should furnish some idea of the magnitude of the convection effect in cavities of various sizes in the earth’s crust. If we take the curves for the three long tubes, each about 23 em. in length, which gave much more consistent results than the shorter ones, and assume that their slopes remain unchanged for small gradients, we obtain for the convectivity at earth gradient, 16 X 107+, 15 X 10-*, and 4 X 10°48, for single columns, 2:8, 2:0, and 11 cm. in diameter, respectively. For a double column each 2:°8 em. in diameter the corresponding figure is 2:4, and for a pair 11 em. in diameter, 14 X 10°. The shorter columns would, in each ease, yield lower results. The enormous importance of area of cross-section, especially in the case of single columns, is at once apparent. It would seem that for single columns less than 2 em. in diameter, and for double ones less than 1 em., the effect of convection at earth gradient should be negligible. Thus we should expect that, in the upward flow of heat through a water-logged porous rock, convection would play a very small part. The presence of water in the pores would doubtless greatly reduce their thermal resistance and prevent the rock from behaving as a very bad conductor of heat like dry pumice, but there is no reason to suppose that it would raise the effective thermal conductivity even as high as that of an otherwise similar rock devoid of pores. On the other hand, the presence of water-filled fissures a couple of centi- metres across, or more, should add very considerably to the upward flow of heat, especially if their shapes were such as to favour continuous circulation. The results obtained with the largest pair of tubes give a value of the convectivity some 400 times as great as the conductivity of common rocks. We might expect that a similar circulation would occur in a single fissure if its extension in both directions in its plane were considerable compared with its width. t The evidence of these experiments is, then, in favour of the view that convection in stagnant water in the earth’s crust is of importance only in regions where fissures of appreciable size exist in the strata. In conclusion I wish to express my gratitude to Prof. W. E. Thrift, r.1.c.p., for the laboratory facilities which he so kindly granted to me in Trinity College, Dublin. SUMMARY. Experiments are described on the convection of heat in single and also in double vertical water columns. The results show that the flow of heat in most cases increases much more rapidly than the temperature gradient. The smaller the column the more rapid is the rise of heat flow with rise of gradient. It is concluded that, for the small gradient existing in the earth the effect of convection in water-logged porous rocks would be negligible. Where, however, water-filled fissures occur, we should expect an appreciable increase in the vertical flow of heat, [ ore 4 No. 34. ON THE SUPPOSED HOMOLOGY OF THE GOLGI ELEMENTS OF THE MAMMALIAN NERVE CELL, AND THE NEBENKERN BATONETTES OF THE GENITAL CELLS OF INVERTEBRATES. - By F. W. ROGERS BRAMBELL, B.A., Sc.B. (Dust); AND J. BRONTE GATENBY, M.A. (Dust.), D.Pum. (Oxon.), D.Sc. (Lonp.). (Piuate XVII.) [Read NovVEMBER 27. Printed DECEMBER 10, 1923.] I.—INTRODUCTION AND SUMMARY OF PREVIOUS WORK. In the year 1898 Camillo Golgi (5) demonstrated, for the first time, the ‘‘apparato reticulare interno’’ in the nerve ganglion cells of the spinal cord. About the same time, but quite independently, Veratti achieved a similar result. Since then much work has been done on the Golgi apparatus of the somatic cells of vertebrates, and it has, in consequence, been demonstrated in all the cells that have been carefully examined by competent observers. Weigl, Hirschler, Gatenby, and others have applied the silver and osmium methods to the germ cells of invertebrates, and have demonstrated in them an apparatus which can be followed through the various stages of maturation, fertilization, and develop- ment. This apparatus is, in all probability, an integral part of every animal cell at some period of its life. The apparatus of the reproductive cells of the invertebrates, the ‘‘nebenkern’’ of some older writers, is considered by many to be strictly homologous to the Golgi apparatus of the vertebrate ganglion cells. In this paper we have endeavoured to state the evidence, already known, bearing on this question, and to put forward some more which we have gleaned by a careful study of the ganglion and other cells of Helix. We venture to believe that the whole body of evidence is sufficient to prove beyond doubt the homology of the ‘‘Golgi’’ bodies found in the various animal cells already mentioned. IJ.—MetnHops AND TECHNIQUE. The following work was carried out on Helix aspersa. As it was carried out in winter, the specimens employed had to be roused from their hibernation by immersing them in warm water. When a specimen was required for work, its head was cut off with a razor, without the use of any anesthetic, and the cephalic ganglion was quickly excised and transferred immediately to a capsule of the fixative. Da Fano’s Golgi apparatus technique was employed with success. The best results were obtained with this method by fixing for two hours, washing quickly in aq. dist., and transferring to 1:5 per cent. silver nitrate solution. After being kept in the dark for two days the specimens were quickly washed in aq. dist. SCIENT. PROG. R.D.S., VOL. Xvil, No. 34, 3 2716 Scientific Proceedings, Royal Dublin Society. and transferred to Cajal’s reducer for two days. They were then again quickly washed and embedded in paraffin. The method of Cajal was also employed with success. Champy-Kull fixation followed by Champy-Kull staining or iron-hematoxylin was very valuable, as was also Elemming-without-acetic-acid and iron- hematoxylin technique. Good results were obtained by the Mann-Kopsch method. The material was fixed in the osmo-sublimate solution for one hour and impregnated with 4 per cent. OsO, for 14 days. Subsequently some of the sections were extracted with turpentine for some days, and stained with Altmann’s acid-fuchsin. Material was also fixed in 10 per cent. neutral formalin for two days, or in 90 per cent. alcohol for one day, both methods being followed by iron-hematoxylin staining. Petrunkewitsch was used, followed by toluidin blue, warmed on the slide. The live ganglion cells were carefully examined in their own lymph, and also stained with Soudan III. in 70 per cent. aleohol, Dahlia 0:75 per cent. in 0:75 per cent. salt sol., Bismark brown in 1 per cent. acetic acid, Janus green 1 in 10,000 (approximately) and 1 per cent. osmic acid. III.—GENERAL DESCRIPTION OF THE GANGLION CELLS. Neurones at every stage of development can be found, at the same time, in the cephalic ganglion of the adult Helix. The smallest of these measure little over 7 in diameter, while the largest attain a size of over 110 wu. In preparations of the living ganglion cell, mounted in its own lymph, the axon and numerous branched dendrons are well seen. The cytoplasm of the body of the living cell appears coarsely granular. On the average the diameter of the nucleus is about two-thirds of that of the whole cell; this proportion remains approximately constant no matter what the actual size of the cell may be. In all the successful silver preparations it was observed that the majority of the nuclei did not take the silver, while other nuclei were deeply impregnated, the chromatin granules showing up well. The nucleoli failed to take the silver even in the most deeply impregnated nuclei. IV.—THE GouGi APPARATUS. The silver methods of Da Fano and Cajal and the Mann-Kopsch osmium technique impregnate well the Golgi apparatus in the nerve cells of Helix. In the smallest neurones observed, the Golgi apparatus was in the extra- centric juxta-nuclear position (Pl. XVII, fig. 1). At this stage it consists of a large number of curved rods, lying on an archoplasmie sphere. This apparatus, at a slightly later stage, divides, the resulting portions tending to pass around the nucleus (fig. 2). The breaking up and scattering continues, the apparatus passing through a stage (fig. 3) consisting of a number of archoplasmie discs, each surrounded by three or four rods. At the end of this process all the rods, each with a portion of archoplasm filling its concave side, are separated and seattered completely around the nucleus. After this no further change takes place, but with the growth of the cell they greatly increase in number (figs. 4 ~ and 5). Ina large neurone the number of Golgi rods, some of which are straight, others curved or ring-shaped, is enormous. They form a zone around the nucleus, and pass out some distance into the dendrons, but are not found in the periphery of the cell. They are not found in the axon or axon-hillock, and are much fewer and markedly smaller in the region between the hillock and the nucleus (fig. 5). BRAMBELL & GatTENBY— Golgi Elements of Mammalian Nerve Cell. 277 In material fixed in Champy-Kull, Flemming-without-acetic-acid, Petrunkewitsch, 10 per cent. formalin, or 90 per cent. alcohol, the Golgi rods appear as unstained ghosts. On account of the highly granular nature of the eytoplasm, it is impossible, in the living cell, to identify the Golgi rods with certainty, but highly refractive bodies, which are probably they, and resemble the nebenkern batonettes of the living Spermatocyte, can sometimes be distinguished. Attempts were made, by centrifuging the ganglia in their own lymph, to segregate the various cytoplasmie inclusions into distinct regions of the cell. It was hoped that by this method the Golgi rods could be made to occupy a definite region, and could, in consequence, be more easily identified ‘‘intra vitam.’’ The apparatus, however, remained scattered around the nucleus in Da Fano preparations, which had been fixed immediately after centrifuging for half an hour at 4,000 revolu- tions per minute, occupying exactly the same position as in the normal cell.2 V.—TueE MitrocHonnvris. The mitochondria appear as numerous minute golden specks scattered through the cytoplasm after the silver methods of Da Fano and Cajal. They appear brown in Mann-Kopsch preparations, and dark grey in material fixed in Champy- Kull or Flemming-without-acetic-acid, and stained in iron-hematoxylin. The mitochondria are therefore not unique, but closely resemble those described from the ganglion cells of many vertebrates (Cowdry). VI.—TuHe NISSL AND OTHER GRANULES. In order to see if the tigroid substance could be identified in the neurones of Helix, some material fixed in Petrunkewitsch, and stained by Seott’s hematoxylin-eosin method (see Bolles Lee, 1922), or in toluidine blue, was studied. No marked granulation was present, but in all the cells fine floeculent granules, taking the blue stain deeply in each ease, were seattered all through the cytoplasm of the cell body, and were in places more definite and better developed. It is probable that these basophil granules do represent the tigroid body of the neurones of vertebrates. In the live cell a number of granules of uniform size are easily seen (fig. 7). They are scattered through the cytoplasm, but may be more or less aggregated in places. These granules are present in every neurone, and appear to be a definite constituent of these cells. They have a strong affinity for Soudan III, and also take Janus green and Neutral red. They do not stain perceptibly in Bismark brown, Dahlia, or 1 per cent. osmie acid. They do not appear after fixation in Da Fano’s, Cajal’s, Mann-Kopsch’s, or Petrunkewitsch’s fluids, or in 90 per cent. alcohol, or 10 per cent. formalin. They are preserved in Champy- Kull’s fixation or Flemming’s fluid without acetic acid, and stain well if these methods are followed by the iron-hematoxylin technique (fig. 6). They take the Fuchsin strongly after Champy-Kull staining. On centrifuging they collect at the top of the cell. It is difficult to state the nature of these granules from the above results; they are possibly lecithin or an allied substance. VIL—ARGENTOPHIL ZONES. A dark perinuclear zone is present in many of the neurones of Helix prepared by the silver methods of Da Fano and Cajal. This zone is approximately coincident with the distribution of the Golgi rods, and is narrower and less dense 1We would like to take this opportunity of thanking Professor H. H. Dixon, of Trinity College, for his kindness in allowing us to use the centrifuge in the School of Botany. 278 Scientific Proceedings, Royal Dublin Society. at the end of the nucleus near the axon-hillock (fig. 5), where, as already remarked, the rods are few and unusually small. This perinuclear zone is not seen ‘‘intra vitam,’’ nor is it shown by any of the other techniques employed. The zoning is similar to that described by one of the authors (J.B.G.) in the oocytes of Saccocirrus prepared by Da Fano’s method. In many of the neurones, in which the Golgi apparatus is in process of fragmentation and spreading around the nucleus, dark zones, similar to those around the nuclei of the larger cells, can be observed around each rod or group of rods (figs. 2-4). This appears to indicate that the Golgi rods, not the nucleus, as suggested in the ease of Saccocirrus, are connected with their formation. In material fixed in Petrunkewitsch’s fluid and stained by Scott’s hematoxylin-eosin method (see Bolles Lee, 1922), the cytoplasm of the neurone is seen to be mainly basophil, with the exception of the axon, which is oxyphil. In medium-sized neurones, however, a juxta-nuclear oxyphil cloud appears. It seems possible that this cloud is connected with the argentophil clouds deseribed in silver preparation, but this question is more fully dealt with by one of the authors in a separate paper (F'. W. R. B., 2). VIII.—HouMGREN’s CANALS. Canals, similar to those originally described by Holmeren, are present in the neurones of Helix, and were observed in many of our preparations. They are of considerable length and sometimes branched (fig. 8), thus in no way resembling the Golgi rods, as seen in our most thoroughly impregnated material. We are not prepared, in this case, to affirm or deny the extra-cellular connection of these canals, but in some neurones they were so related to processes of the sub- capsular cells that they might be interpreted as continuous with them. TX.—ON THE VISIBILITY OF THE GoLel APPARATUS OF GENITAL CELLS “INTRA VITAM.”’ If the ovotestis of Helix be teased out ‘‘intra vitam”’ and examined in its own lymph without staining, the nebenkern batonettes of the sperm cells of all stages can be seen clearly. The apparatus is also visible while in the eccentric position in the young oocytes, but, as it scatters ih growth, gradually it becomes indistinguishable. These observations support those from fixed material, in which the nebenkern elements alter in chemical constitution during dispersal through the cytoplasm. This chemical change is accompanied by a change in refractivity, which tends to make them invisible ‘‘intra vitam,’’ although they are still freely demonstrable by the osmie and silver nitrate techniques. The fact that in most cells the Golgi elements are invisible ‘‘intra vitam’’ has been urged as evidence that in fixed preparations the so-called Golgi apparatus is merely an artifact. That this argument is not convincing may easily be shown, for, in the first place, the micro-chemical methods used for demonstrating the Golgi apparatus reveal the latter both in cells in which the Golgi elements are visible ‘‘intra vitam’’ (e.g. snail spermatocytes), and in cells in which the tolei elements are not visible ‘‘intra vitam’’ (e.g. snail neurones). Moreover, we have instanced the case of snail oocytes where, in the young cells, the apparatus is visible ‘‘intra vitam,’’ but, as the yolk begins to form, the apparatus undergoes certain chemical changes, and is then no longer visible, but is still easily demonstrable after fixation. X.—DISCUSSION. From the time of its ‘discovery the Golgi apparatus of the nerve cell of mammals was recognised as a branching network, each element tending to BraMBett & GatenBy— Golgi Hlements of Mammahan Nerve Cell. 279 anastomose with its neighbour. Many observers who have worked almost exclusively on mammalian material, when shown the peculiar banana-shaped batonettes of molluse, annelid, insect, and other invertebrate cells, have expressed doubts as to the homology of the Golgi apparatus of their material with these structures in the cells of invertebrates. The evidence for the view that the two sets of structures are truly homologous © falls into the following groups :— 1. In the neurones of invertebrates the Golei apparatus techniques reveal an apparatus formed mainly of isolated rods and batonettes. This apparatus occupies the same position in both young and old neurones as does the true Golgi apparatus in young and old mammalian nerve cells. 2. The micro-chemical evidence shows that the silver nitrate techniques of Golgi, Cajal, and Da Fano consistently reveal both the dendriform apparatus of the mammals and the batonette or isolated elements of the lower animals. 3. The osmium techniques of Kopsch, Mann-Kopsch, and Sjovall likewise consistently impregnate these two sets of bodies in the vertebrate and the invertebrate animals. 4. The embryological evidence shows that the so-called Golgi bodies of eggs, during segmentation and histogenesis, are divided or sorted out among the blastomeres and their tissue derivatives, and that finally these ege Golgi bodies form the Golgi apparatus in the cells of the nervous system. Thus, eytologically the Golgi bodies of the egg have been traced from the small eccentric batonettes of the primitive germ cells, and embryologically their derivatives in the egg have been traced into the nerve cells of the new organism. The chain is therefore complete. 5. In many vertebrate animals (eg. the frog) the Golgi apparatus of the nerve cell is formed of elements intermediate in shape and appearance between the banana-shaped genital element of invertebrates and the tree-like or net-like elements of the mammalian neurone. It must then be admitted, especially in’ view of the embryological evidence (Hirschler and Gatenby, 3 and 6), that there is a complete continuity between the genital Golgi element and the nerve Golgi element. Some observers have expressed doubts as to the reliability of the various silver nitrate techniques; our evidence, however, does not depend solely on similarity of micro-chemical reaction. On the other hand, we recognise fully that the techniques of osmice acid and silver nitrate often reveal wide differences in micro-chemical reactions between the batonettes of invertebrate genital cells and the Golgi apparatus of mammalian cells: this is not a serious matter, however, because equally wide micro-chemical differences exist not only between the Golgi elements of different animals, but -also between the Golgi elements of different tissues of the same animal. We believe that the evidence adduced in this paper is sufficient to prove that the batonettes or nebenkern rods of such an animal as Helix, the snail, are homologous with the remarkable Golgi inner network of the mammalian neurone. SUMMARY. 1. The Golgi apparatus in the smallest neurones of Helix was in the perinuclear extra-centrie position, surrounding an archoplasmic sphere. In larger neurones it becomes dispersed around the nucleus and the individual elements become much more numerous. 2. Basophil granules, probably representing the tigroid body, and also lecithin (?) granules, are described in the neurones. SCIENT. PROC. R.D.S., VOL. XVI, NO. 34. OF 280 Scientific Proceedings, Royal Dublin Society. 3. In silver preparations dark zones are found around the Golgi elements. These probably represent a product of its activity. 4. Long and sometimes branched Holmgren canals were found in the neurones. They were separate and distinct from the Golgi elements. There was some evidence for considering them processes of the subcapsular cells, but not sufficient to come to any definite conclusion. 5. From the position occupied by the apparatus in nerve and germ cell, from its similarity of micro-chemical reaction in both, and from embryological evidence, it is believed that the nebenkern batonettes of the invertebrate germ-cells are homologous to the Golgi network of the mammalian neurone. BIBLIOGRAPHY. 1. Boutes Lee.—The Microtomist’s.Vade Mecum. Gatenby, 1922. 2. BramBetL, F. W. Rocgmrs.—The Activity of the Golgi Apparatus in the Neurones of Helix aspersa. Jour. Physiol., Aug., 1923. 3. Gatensy, J. Brontii.—The Cytoplasmic Inclusions of the Germ-cells. Part V. The Gametogenesis and Early Development of Limnaea stagnalis (LL), with special Reference to the Golgi Apparatus and the Mitochondria. Quart. Jour. Micr. Se., Vol. 63, Part IV. 4, The Gametogenesis of Saccocirrus. Quart. Jour. Mier. Se., Vol. 66, 1922. Goner, CamiLo.—tInterno alla structura delle cellule nervose. Boll. Soe. Med. Chr. di Pavia, 1898. 6. Hirscuier, J.—Ueber den Golgischen Apparat embryonaler Zellen. Arch. fur Mikr. Anat. Bd. 91, Abt. 1. oO DESCRIPTION OF PLaTE XVII. Key to lettering :—Golgi apparatus in extra-centrie position (G.), and in subsequent stages of breaking up (G, and G,), Golei Rods (G.R.), Ghosts of Golgi Rods (G.G.R,), Argentophil zone surrounding Golgi elements (A.Z.), Mitochondria (M.), Holmegren’s canals (H.C.), Granules (Lecithin?) (A.), Nucleus (N.), Plasmasome (P.), Region of Axon hillock (H.). These drawings were made with the help of a camera lucida. Fig. 1. Small neurone with Golei apparatus in perimuclear extra-centrie position. Da Fano preparation. > 4000. 2. Small neurone with Golgi apparatus beginning to spread. Da Fano preparation. X 4000. 3. Further stage in spreading out of Golgi apparatus. Da Fano preparation. x 4000. 4, Medium-sized neurone with Golgi apparatus spread around nucleus. Da Fano preparation. 2500. 5. Large neurone. Golgi rods small and scaree im region of Axon hillock (H.). In this and the preceding figure the argentophil zones around the rods are shown. Da Fano preparation. > 600. 6. Large neurone showing leeithin granules (A) and ghosts (G.G.R.) where the Golgi rods have been dissolved away. Champy-iron-haematoxylin preparation. X 1300. . Medium-sized neurone showing lecithin granules. From living cell. 8. Portion of large neurone showing Holmgren’s canals in cytoplasm. % 1300. ~ PLATE XVII. SCIENT. PROC. R. DUBLIN SOC., N.S., VOL. XVII. F.W.RB. det. BRAMBELL AND GATENBY. fie se No; Bo: PHOTOTROPIC MOVEMENTS OF LEAVES.—THE FUNCTIONS OF THE LAMINA AND THE PETIOLE WITH REGARD TO THE PER- CEPTION OF THE STIMULUS. By NIGEL G. BAL, M.A., Assistant to the Professor of Botany in the University of Dublin. (Read DeceMBER 18. Printed DECEMBER 28, 1923.) In spite of a considerable amount of work which has been done from time to time, the various factors involved in the orientation of leaves in response to light are by no means clearly understood. This applies especially to the question as to how far the lamina exerts a directing influence on the petiole, and so far unanimity on this point has not been attained. Darwin (1) first showed that in experiments on leaves of Tropaeolum majus and Ranunculus ficaria, in which the blades were covered with black paper, the petioles became curved towards the light as completely as those of unprotected leaves. Vo6chting (2), on the other hand, concluded that in the case of Malva verticillata the movements of the petiole are controlled by the lamina. Darwin’s experiments on Tropaeolum were later confirmed by Rothert (3), who considered that illumination of the lamina was without influence on the bending of the petiole, and Krabbe (4) obtained similar results with Fuchsia and Phaseolus. Haberlandt (5) repeated the experiments of Darwin and Rothert on Tropaeolum, but found that the correct diaphototropie position was not reached as completely or as accurately by the leaves covered with black paper as by those which were uncovered. He therefore concluded that while the main movement of the leaves of Tropaeolum was effected owing to the phototropic response of the petiole, the finer adjustments were controlled by the lamina. Im the ease. of Begonia discolor and Monstera deliciosa he found that even when the petioles were covered with tinfoil the leaves were able to reach the correct phototropic position, and concluded that in these plants the lamina is the sole percipient organ, and that it transmits the stimulus to the petiole. Working with Phaseolus, however, he confirmed Krabbe’s (4) results, and found that leaves in which the lamina was covered with black paper reached the correct phototropie position as rapidly and as completely as uncovered leaves. More recently Wager (6) concluded that the ‘‘perception of light is located not in the leaf-blade but in the leaf-stalk.’’ As a result of the experiments which have been carried out by Boysen- Jensen (7), Padl (8), and others, on the transmission of the phototropic stimulus in seedlings, phototropic reactions have lately acquired a new interest, and further work on the movements of leaves seemed desirable. In the first experiments leaves of Oxalis macra Small were used. A number of freshly-plucked leaves were fixed with the ends of the petioles in water in small glass jars. The petioles passed loosely through holes in corks fitting the necks of the jars, and were firmly fixed in these holes by means of SCIENT. PROC. R.D.S., VOL. XVI, NO. 35. ; 3G 282 Scientific Proceedings, Royal Dublin Society. plasticine. These jars were placed in front of a window and suitably shaded, so that they were illuminated from one side only. In some eases the leaves were left intact, and in others the leaflets were cut off where they jomed the petiole. The amount of bending was deduced by taking measurements of the horizontal displacement of the top of the petiole. As a result of a number of experiments, it was found that the petioles from which the leaflets had been removed responded to phototropie stimuli as readily as those of mtact leaves. The power of response of the decapitated petioles, however, was lost after a period of one to two days. An experiment in which both the growth in length and the amount of bending in the intact and decapitated petioles were measured, showed that the loss of the power of response in the decapitated petioles could be correlated with cessation of erowth. In all cases the region of greatest growth and bending was about 4-6 em. from the top of the petiole. — In order to obtain evidence regarding the vexed question as to whether phototropic stimuli can be perceived by the lamina and transmitted to the petiole, experiments were carried out on Sparmannia africana. This plant was found to be very suitable for the purpose as the leaves readily set them- selves in the optimum position relative to incident light. The larger leaves are, on an average, about 15 em. in length by 12 em. broad, being supported on petioles about 12 em. in length. Fic. 1. Plant of Sparmanmia africana, viewed from above, showing the methods adopted in shading the leaves. Two plants were chosen which were growing in pots and had been exposed to light from above in a greenhouse, with the result that the leaf-blades were all horizontal. In each plant a pair of similar leaves lying close together in the same horizontal plane was selected. A piece of cardboard was supported by means of a clamp attached to a retort stand just above, but not touching, the adjacent edges of the two leaves in the manner shown in fig. 1, a. The pots containing the plants were standing on the floor of a greenhouse, and were surrounded by tall screens so that they were illuminated from above only. Under these circumstances the shades affected small portions only of the leaf- blades, and were without influence on the light falling on the petioles. Batt—Pholotropic Movements of Leaves. 283 The distance between the tops of the two petioles and also the distance between the two leaf-tips were measured from day to day and the means of these two distances are plotted in fig. 2, which shows the results obtained from the two plants (Hxpts. I and ID), both experiments being carried on simultaneously. At the end of five days the shades were removed and replaced by similar shades which covered the outer edges of the pairs of leaves, as shown in fig. 1, b. The experiments were continued for a period of twenty days, the shades being changed at intervals of five days. The curves in fig. 2 show clearly that when the inner edges of the pairs of leaves were shaded the leaves moved apart, and when the outer edges were shaded they came closer together. The slight irregularities in the curves seem to be due to inequalities in the temperature of the surroundings, which varied from day to day during the course of the experiments. The effect due to this cause was particularly marked during the eighth day of the experiments, when, owing to failure of the heating arrangements in the greenhouse, the temperature fell considerably, with the result that movement of the leaves practically ceased. = oO = my ui > Dinitrodiphenylene Oxide (M.P. 182°C.) (M.P. 245° C.) | va Trinitrodiphenylene Oxide L______-> Tetranitrodiphenylene Oxide (M.P. 223° C.) (M.P. 288°C.) EXPERIMENTAL. 1. Action of Nitrogen Peroxide on Diphenylene Oxide. (a) In the Absence of Solvents—2 g. of diphenylene oxide and some dry liquid nitrogen peroxide were placed in shallow dishes under a bell-jar. After 524 Scientific Proceedings, Royal Dublin Society. remaining at the room temperature for a day the oxide assumed a yellow colour, which deepened slightly as the reaction progressed. The experiment was continued for 5 weeks, the supply of nitrogen peroxide being only once replenished, as its absorption was slow. The yellow product was washed with ligroin, and, after recrystallisation from glacial acetic acid, fine white needles, melting at 245° C., separated in a very good yield. This compound was there- fore evidently identical with the dinitrodiphenylene oxide obtained by Mailhe by the action of nitric acid on the oxide. (b) In Glacial Acetic Acid Solution.—1 g. of diphenylene oxide was dissolved in 20 «ec. of glacial acetic acid, and dry nitrogen peroxide vapour was passed into the solution until the latter was saturated. The reddish-brown solution was allowed to remain in a stoppered bottle at the room temperature, a relatively large quantity of a yellow solid being deposited overnight. After 3 months the product was poured into water; the yellow solid was washed free from acid, and after recrystallisation from glacial acetic acid it consisted of yellow silky needles, melting at 182°C. It was evidently identical with the mononitrodiphenylene oxide obtaimed by Borsche and Bothe (loc. cit.), by the action of fuming nitric acid on a solution of diphenylene oxide in acetic acid. z (c) In Carbon Tetrachloride Solution.—A solution of 2 g. of diphenylene oxide in 40 g. of carbon tetrachloride was saturated with nitrogen peroxide vapour. Some heat was evolved, and on cooling a yellow solid was deposited. The mixture was allowed to remain at the room temperature for 3 months. The solvent was then evaporated, and the solid product was extracted several times with boiling alcohol until only a very small amount of residue was left. From the filtrate a good yield of the above-mentioned mononitro derivative was obtained, while the residue on recrystallisation from glacial acetic acid gave a small quantity of dinitrodiphenylene oxide, melting at 245° C. 2. Action of Nitrous Fumes on Diphenylene Oxide. Nitrous fumes, generated by the action of arsenious oxide on nitric acid, were passed into a solution of 2 g. of diphenylene oxide in 40 g. of glacial acetic acid. After remaining at the room temperature a small quantity of 3 yellow solid was deposited from the dark-green solution, and this increased gradually in amount. After 4 weeks the solution was filtered from the solid residue, which on erystallisation was found to be the mononitro derivative already described. The filtrate on evaporation gave a further yield of this substance in addition to some unchanged oxide. 3. Action of Fuming Nitric Acid on Diphenylene Oxide. (a) In the Absence of Solvents—-2 g. of diphenylene oxide was added eradually to fuming nitric acid (10 e.c.). A vigorous reaction ensued, and a yellow solid was obtained on cooling the mixture. This solid was washed with water, dried, and extracted with boiling ligroin, and finally with boiling aleohol. The undissolved residue consisted of dinitrodiphenylene oxide, melting at 245° C., and from the alcoholic filtrate the mononitro compound, melting at 182° C., was isolated. (b) In Glacial Acetic Acid Solution—To two solutions, each of which contained 1 g. of diphenylene oxide in 100 g. of glacial acid, 1 and 4 molecular amounts of fuming nitric acid were added respectively. The reactions were evidently very slight, even on allowing the solutions to remain at the room temperature for 5 months, since the only substance isolated was, in both eases, the original oxide. tYAN & CuniInane— Action of Oxides of Nitrogen on Diphenylene Oxide. 325 (c) In Carbon Tetrachloride Solution—A solution containing 1 g. of diphenylene oxide and 1 molecular amount of fuming nitric acid in 100 g. of carbon tetrachloride was prepared. This solution was also allowed to remain at the room temperature for 5 months, when, on evaporation of the solvent, a red oil remained. After extracting this with ligroin, the parent substance was obtained from the ligroin extract, and a small quantity of the mononitro derivative previously mentioned was isolated from the undissolved portion A similar experiment was also performed with the employment, however, of 4 molecular amounts of fuming nitric acid. After 5 months the yellow solution was evaporated, leaving a yellowish-white solid. This was washed with ligroin, a small quantity of the original oxide being isolated from the extract. The undissolved residue consisted of mononitrodiphenylene oxide, melting at 182° C. (d) An almost quantitative yield of mononitrodiphenylene oxide was obtained by the following method :—5 g. of the oxide were dissolved in 20 cc. of glacial acetic acid, and 5 ¢.c. of fuming nitric acid were added slowly with shaking. When the initial violent reaction had subsided, the mixture was heated for 5 minutes on the water-bath, and on cooling a yellow solid cake was formed. Jt was washed with water and recrystallised from glacial. acetic acid, and was then found to be pure mononitrodiphenylene oxide. (e) An equally good yield of the mononitro compound was obtained by substituting carbon tetrachloride for acetic acid as the solvent. In this case the solid cake was deposited during the addition of the acid. On reerystallising this from glacial acetic acid it gave the pure mononitro compound. 4. Preparation of the Nitro Derwatives. (a) Dinttrodiphenylene Oxide.—This compound was readily obtained in the following manner:—To a concentrated solution of 5 g. of the mononitro derivative in glacial acetic acid 25 ¢.c. of fuming nitric acid was added slowly. A somewhat vigorous reaction occurred, and, when it had subsided, the mixture was heated for about 20 minutes on the water-bath. The white crystalline deposit formed on cooling was washed with water, and recrystallised from acetic acid. The product consisted of fine white needles of dinitrodiphenylene oxide, melting at 245°C. The yield was 3 g. Dinitrodiphenylene oxide is almost insoluble in ligroin, slightly soluble in aleohol, moderately soluble in glacial acetic acid or xylene, and readily soluble in acetone. (b) Trinitrodiphenylene Oxide-—Fuming nitric acid (10 ¢@c.) was added slowly with shaking to 2 g. of mononitrodiphenylene oxide. A vigorous reaction ensued with evolution of heat. The reddish-yellow solution was allowed to remain at the room temperature for a few minutes, and was then poured into water. The solid product was washed several times with boiling alcohol. The residue was extracted with boiling benzene, from which a small quantity of colourless blunt prisms, melting at 223°C., was isolated. The alcoholic extract contained dinitrodiphenylene oxide. The portion left undissolved by the benzene crystallised from xylene, yielding a small amount of colourless thin plates, melting at 283°C. The compound melting at 223°C. was a trinitrodiphenylene oxide, and it gave on analysis the following results :— 0:1027 g. substance gave 12:3 ¢.c. moist nitrogen at 18° C. and 766 mm. corresponding to N141 C,.H,0,N, requires N 13:9. SCIENT. PROG. R.D.S., VOL. XVII, No. 40. 3P 526 Sctentifie Proceedings, Royal Dublin Society. Trinitrodiphenylene oxide is almost insoluble in ligroin, slightly soluble in alcohol, somewhat soluble in benzene or glacial acetic acid, and readily soluble in acetone. (c) Tetranitrodiphenylene Oxide—2 g. of ‘dinitrodiphenylene oxide was dissolved in a mixture of 10 ¢.c. of fuming nitric acid and 10 ce. of concentrated sulphuric acid. The mixture was heated for several hours on the water-bath, a small quantity of colourless erystals being deposited during the course of the experiment. The product was poured into water, and the white solid deposited was washed with water, and dried. The residue left after washing with hot benzene crystallised from xylene as colourless plates, melting at 283° C., which gave on analysis the following results :— 0:1075 g. substance gave 14-7 ¢.c. moist nitrogen at 16° C. and 770 m.m. corresponding to N 163 C,,H,0,N, requires N 16-1. Tetranitrodiphenylene oxide is very slightly soluble in alcohol, somewhat more soluble in benzene, and readily soluble in hot acetic acid, xylene, or acetone. SUMMARY. 1. Diphenylene oxide interacts with nitrogen peroxide or nitric acid much less readily than diphenylnitrosamine. 2. Nitrogen peroxide vapour converted diphenylene oxide into its dinitro derivative. Nitrogen peroxide in solution in glacial acetic acid, gave the mononitro compound; in carbon tetrachloride it gave the mononitro and the dinitro derivatives. 3. Fuming nitrie acid converted the oxide into its mononitro and its dinitro derivatives. In cold dilute solution in glacial acetic acid the acid had scarcely any action on the oxide, but in carbon tetrachloride the mononitro compound was formed. This body was also produced by the action of fuming nitric acid on a hot solution of the oxide in glacial acetic acid. 4, Dinitrodiphenylene oxide was readily produced by the action of fuming nitrie acid on a solution of mononitrodiphenylene oxide in glacial acetic acid. 5. Trinitro- and tetranitrodiphenylene oxides were both formed by the action of fuming nitric acid on the mononitro compound. The tetranitro derivative was also readily obtained by the action of a mixture of fuming nitric acid and concentrated sulphuric acid on dinitrodiphenylene oxide. In conclusion we wish to state that the above research was undertaken at the request of the Research Section of Nobel’s Explosives Company, and that we are also indebted to the Department of Scientific and Industrial Research for a grant in aid of the investigation. Lo a2} No. 41. THE HABITATS OF LIMNAEA TRUNCATULA AND L. PEREGER IN RELATION TO HYROGEN ION CONCENTRATION. By W. R. G. ATKINS, O.B.E., F.1.C., AND MARI“ V. LEBOUR, D.Sc., Marine Biological Laboratory, Plymouth. (Read January 29. Printed FrBruary 22, 1924.) L. truncatula has long been the accepted intermediate host of the liver fluke Fasciola hepatica, and it has recently been pointed out by Taylor (1922) that sheep are infected with liver fluke in parts of Scotland from which L. truncatula is absent. LZ. pereger is, however, found there, and has been proved to be infected with the cereariae of IF. hepatica. The authors (1923) have already shown that the distribution of many snails is limited by acidity, whereas others are found in acid or almost neutral, but never in alkaline habitats. Attention was drawn by Wallace (1922) to the opinion of Welsh farmers that liver fluke was more prevalent on land which had been limed than on untreated pastures. Since the sheep pastures of Wales are notoriously poor in calcium carbonate the liming must produce a considerable change in the soil reaction, but the reaction is unlikely to be at all as alkaline as in limestone or chalk districts. There appeared, therefore, to be indications that an almost neutral reaction favoured ZL. truncatula. In view of the fact that in certain wet years the ravages of liver fluke among sheep become disastrous, it appeared to be of interest to ascertain what characteristics might be used to define localities in which L. truncatula or L. pereger could develop or could abound. This appears to be the more desirable, inasmuch as it is often the custom to bring sheep down from hill pastures free from these snails to winter in lowlands, which are normally less acid than the hills. The work of Walton (1923), on the destruction of ZL. truncatula by means of copper sulphate, has, however, demonstrated that habitats naturally very favourable may be almost or entirely cleared of this snail. L. truncatula probably attains its greatest abundance in water-logged pastures. Boyeott (1919) does not consider it as a water-snail, though it oceurs in three ‘‘drying’’ ponds out of a total of one hundred and forty-one “‘closed’’ ponds in the parish of Aldenham. It is nowhere recorded by him among the sixty-nine land mollusca of Aldenham and its district (1921). This district, therefore, is apparently not a favourable habitat for Z. truncatula; L. pereger, on the other hand, is truly a water-snail. Boycott (1919) records its occurrence in fifty-one out of one hundred and eight ponds examined by him, as well as in lakes and streams, though not in any of the seventeen ‘‘drying’’ ponds. Stelfox (1911) records the general occurrence of L. pereger in the Clare Island survey, and remarks upon the swarms of a small form SCIENT. PROC. R.D.S., VOL. XVII, NO. 41. 3Q 328 Scientific Proceedings, Royal Dublin Society. of L. truncatula found upon the bare rock-faces of the sandstone cliff of Croaghmore, which faces the Atlantic on the west of the island. The snail is found up to nearly 1,000 ft. Taylor (1920) does not include L. truncatula among the mollusea found in the low-lying district of Audruicq, in Picardy, and comments upon the absence of limestone. Kendall (1921) records the finding of L. truncatula in the Oundle district in river marsh—rush-grown shallows and moist river margins, also in small streams, but not in ‘‘natural marsh’’—an extensive tract of boggy ground with common rush, cotton grass, bog bean, marestail, and peppermint. Through the kindness of Sister Monica Taylor, Mr. ©. L. Walton, Miss K. Carpenter, Mr. T. H. Taylor, and Mr. W. M. Temple the authors have been provided with certain samples and notes concerning the distribution of these snails in Seotland, Wales, and Yorkshire, as well as near Plymouth. The ] 5 values for pH = log a are recorded for these soils or waters, as are also the electrical conductivities of the waters measured at 0°C., at which temperature N/100 potassium chloride has the value 0:00078, and was used to standardise the cell. The data presented in the table opposite are obviously not as numerous as could be desired, but are put on record and discussed now, as the subject is foreign to the authors’ general line of work, and it is not their intention to pursue it further. The crosses given in the table denote a rough attempt at giving an idea of the relative abundance of the two species. It must be explained that pH 69-76, for example, does not mean that the value lay between those limits, but that when examined it was pH 69, and when the water was thoroughly well aerated and the free carbon dioxide was reduced thereby so as to be approximately in equilibrium with the air, the pH value then rose to pH 76. The condition of soft water at pH 69 is therefore very different from that of hard water at the same pH value, for though the reaction is the same the hard water is charged with a great excess of earbon dioxide above the equilibrium value, and is correspondingly poor jn oxygen, whereas a soft water at this pH value may be thoroughly well aerated. It is obvious that the oxygen content of water is of great biological importance. The change in pH value produced by aeration may, therefore, be taken as giving a rough idea of the degree to which the water is saturated with oxygen. The electrical conductivity of the water gives a measure of its salt content, which, when dealing with normal natural waters, may be taken as roughly proportional to its hardness. Hard water has a conductivity of 270-290 < 10°, whereas a very soft water like the Plymouth town supply from Burrator reservoir, on Dartmoor, has a conductivity C = 26-28 X 10°°, with a pH value of 64 to 68 according to season; these values are unaltered by aeration, but a spring water at pH 6:8 and having C = 270 X 10° rises to over pH 8 on aeration. i On inspecting the table it may be seen that, though the records are not numerous, they may be taken as typical of districts of some size. As regards the conductivity a wide range may be noted; this indicates a yet wider range in ‘‘hardness,’’ for even the soft waters of the R. Ystwyth and its neighbours have GC = 39-72 X 10°, the Aberystwyth* town supply being as low as C =19 X 10°. ZL. truncatula has been found in water having conductivity * The authors are indebted to Miss K. Carpenter for these samples. Arkins AnD Lesour—AHabitats of Limnaea Truncatula, §c. 329 LD. trun- Date. Locality. Nature of habitat. Bee L.. pereger. pH. C x 108. Notes. 12/2/’2z3. Dunbarton, Soil by stream. + 0 6°6 — Always infected with basaltic district. F. hepatica. 20/2/’23. Do. Stream. + 0 6-4 — 6°7* 59 Do. 20/3/7238. Dunbarton, Ditch. ++ 0 6:9 — 7°6 186 Never infected, no calciferous sand- sheep. stone district. 20/3/'28. Mugdock reservoir Offset from stream. 0 +++ 68-72 72 L. truncatula never grounds. found. 22/3/?23. Dunbarton, old Pond. + +++ 66-7°3 220 L. pereger highly in- red, sandstone fected with F. quarry, drain- hepatica. age from fields. 14/2/'23. Bodorgan, Stream in shallow + +4++ 68-74 +192 A few ZL. truncatula Anglesey. ditch. along sides. 14/2/’23. Do. Mud from ditch. + +++ 7:2 — — 5/6/'28. Near Abery- Soil from site. + oP 58 = — stwy th. 5/6/?23. Do. Do. 0 + 5°8 —_— = 5/6/’28. Do. Do. + 0 7) = = 22/3/’23. Near Leeds. Stream in flood. 0 ++ T1—7T4 227 Infected with two species of cercariae.T 8/10/23. Allerton Park, Lake with sedges. ++ 0 7-2 — Deer infected with Knaresboro’. fluke. 1/4/?23. Plymouth. Ditch in lane, flow- ++ 0 7°6 — 8:2 294 = ing. 2/4/?23. Do. Stagnant ditch in + a oe 6:9 — 7°8 213 = meadow with duckweed. 2/4/?23. Do. Do., but spirogyra +4 + 77-81 213 = abundant. 8/3/28. Near Malldraeth Pools 2 in.deep in ++++ 0 7:8 — 8:0 290 = Marsh, — Llan- water-logged pas- gefui, Wales. ture. * For explanation of second figure see text. + Neither was /. hepatica. 330 Scientific Proceedings, Royal Dublin Society. ranging from 59-294 10°, and L. pereger from 72-227 X 10°. When taken in conjunction with the failure to find the snails in the R. Yealm, where the water was at C = 32 X 10°, it seems justifiable to conclude that the lower limits recorded have an approximate value as setting a genuine bound to the habitats of these species. The greatest abundance of L. truncatula was, however, in a highly calcareous water near Malldraeth Marsh. On considering the hydrogen ion concentration it is seen that L. truncatula has been found in water between pH 64 and 7°8, and L. pereger from pH 66— 7-7. No differential significance can be attached to these figures, but it seems that L. pereger can endure deeper and less well-aerated water than can L. truncatula, which is amphibious. The records show that snails may be infected even in the most acid of the waters studied. It is, however, noticeable that none of the records of soil are more acid than pH 5:8. Streams draining Dartmoor may be even more acid than the R. Yealm, pH 64, which receives tributaries that lessen its acidity from the Culm Measures. Values such as pH 5:4 above the Culm, on granite, and pH 5:0, for a bog pool, have been observed; and among the China-clay pits pH 62 to 64, with the low value 31 X 10° for conductivity. It appears, therefore, that there must be great stretches of upland pastures in which the water is too acid and too poor in salts for either of these species to exist. The acidity of the soil in these districts is even greater than is that of the water. Wet peat bogs, where cotton grass abounds, may be as acid as pH 41; heather bogs, pH 46; and the siliceous soil of uplands, good permanent pastures with wild white clover, cocksfoot, sheep’s fescue, crested dogstail, and bent, may be as acid as pH 5-4. It is, therefore, not unpleasing to find that such habitats appear to be too acid for L. truncatula to live. It is evident, therefore, that the habitats given in Roebuck’s census (1921) ““T, pereger: fresh and stagnant water generally’’ and ‘‘L. truncatula: ponds, ditches, and wet places generally,’’ require modification in the direction of the exclusion of fresh water of very low conductivity, and also of high acidity in water or soil. It may also be pointed out that ZL. truncatula has been recorded (1900) by one of us as occurring in quantity on the mud between the tidal limits of the upper part of the R. Tweed, where the water was slightly brackish. The nature of the habitats makes it clear that though both species can harbour F. hepatica, yet L. truncatula is the more likely to cause infection. SUMMARY. 1. The habitats of Z. pereger and L. truncatula appear to differ in the fact that whereas the former is truly a water-snail, and can endure even somewhat stagnant water; the latter is amphibious, and can live either in shallow, well- aerated water or on moist land, or even on cliffs in a region of high humidity. 2. The observed ranges for the two species are almost identical as regards acidity and salt content of the water, those for L. pereger being pH 66 to 7:7 and C = 72-227 X 10° at 0°C., and for ZL. truncatula, pH 64-78 and C = 59-294 < 10°. It is noticeable that the records do not include upland waters of very low salt content, with conductivity 20-30 < 10-°, nor regions of high acidity, more acid than pH 6:4 for water or pH 58 for land records. Arxins anp Lesour— Habitats of Limnaea Truncatula, &e. 331 REFERENCES. Arxins, W. R. G., and Lesour, M. V. (1928).—The hydrogen ion concentration of the soil and natural waters in relation to the distribution of snails. Sei. Proe. Roy. Dublin Soce., 17, 233-240. Boycorr, A. E. (1919).—The fresh-water mollusea of the parish of Aldenham. Trans. Hertfordshire Nat. Hist. Soc., 17, 152-200. Boycorr, A. EH. (1921)—The land mollusea of the parish of Aldenham. Loc. cit., 17, 220-245. Kenna, ©. E. Y. (1921).—The mollusca of Oundle. J. Conchology, 16, 240; and 1922, 16, 248. Lesour, M. V. (1900).—Land and fresh-water mollusea in Northumberland. Naturalist, No. 518. Roesuck, W. D. (1921)—Census of the distribution of British land and fresh- water mollusea. J. Conch., 16, 165-212. StetFox, A. W. (1911).—Clare Is. Survey. Land and fresh-water mollusea. Proe. Roy. Irish Acad., 81, No. 23, 1-64. Taytor, JNo. W. (1920).—The land and fresh-water mollusea of Audruicq, Pas-de-Calais. J. Conch., 16, 106. Tayvior, M. (1922)—Water-snails and liver flukes. Nature, 110, 701. Watuace, R. H. (1922).—Water-snails and liver flukes. Nature, 110, 845. Watton, C. L. (1923).—Liver rot of sheep. J. Ministry of Agric. (England), ? SOIENT. PROC. R.D.S,, VOL. XVII, NO. 41. 3R ’ by eX 10. itil. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21, 22. SCIENTIFIC PROCEEDINGS —continued. . On a Phytophthora Parasitic on Apples which has both Amphigynous and Paragynous Antheridia; and on Allied Species which show the same Phenomenon. By H. A. Larrerty and Guoren H. Prruyeripen, Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (Plates I and II.) (June, 1922.) . Some Further Notes on the Distribution of Activity in Radium Therapy. By H. H. Poot, u.a., sc.p., Chief Scientific Officer, Royal Dublin Society. (June, 1922.) . Preliminary Hxperiments on a Chemical Method of Separating the Isotopes of Lead. By I'somas Ditnon, v.sc.; Rosauinn Crarke, p.sc.; and Vicror M. Hinony, s.sc. (Chemical Department, University College, Galway). (July, 1922.) . The Lignite of Washing Bay, Co. 'I'yrone. By 'I'. Jonson, p.sc., F.1.s., Professor of Botany, Royal College of Science for Ireland; and Janz G. Gitmorg, B.sc. (Plate III.) (August, 1922.) . Libocedrus and its Cone in the Ivish Tertiary. By T. Jounson, p.sc., ¥.u.s., Professor of Botany, Royal College of Science for Ireland; and Jang G. Gitmore, B.sc. (Plate IV.) (August, 1922.) - The Electrical Design of A.C. High Tension Transmission Lines. By H.H. Jerroorr. (August, 1922.) The Occurrence of Helium in the Boiling Well at St. Edmundsbury, Lucan. By A. G. G. Lronarp, F.r.c.sc.1., PH.D., F.I.c., and A. M. Ricwarpson, A.R.O.SC.1., A.I.c. (Plate V.) (August, 1922.) [ Nos. 1 to 10, price 9s. ] On the Detonating Action of a Particles. By H. H. Poonm, .a., so.p., Chief Scientific Officer, Royal Dublin Society. (December, 1922.) The Variations of Milk Yield with the Cow’s Age and the Length of the Lactation Period. By Jamms Wison, m.a., B.sc. (December, 1922.) A Note on Growth and the Transport of Organic Substances in Bitter Cassava (Manihot utilissima). By 'l. G. Mason, u.a., B.sc. (December, 1922. [Nos. 11 to 13, price 1s. 6d.] The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylurethane. By Huew Ryan, p.sc., and Anne Donnetnan, m.sc., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on Ethyl-o-Tolyl- urethane. By Huen Ryan, p.sc., and Nicuonas Cunimanr, pPH.p., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on HWthyl-Phenyl- urethane. By Hues Ryan, p.sc., and Anna Connotzy, .sc., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on Phenyl-Benzyl- urethane. By Hue Ryan, p.sc., and James L. O'Donovan, m.sc., University Colleze, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on the Phenylureas. By Hueu Ryan, p.sc, and Prrer K. O’Tooun, u.sc., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on Phenyl-Methyl- urea. By Hues Ryan, p.sc., and Micuarn J. Sweeney, m.sc., University College, Dublin. (February, 1923.) [Nos. 14 to 19, price 4s.] On the Cause of Rolling in Potato Foliage; and on some further Insect Carriers of the Leaf-roll Disease. By Paun A. Murray, sc.D., a.R.c.so.1., Seeds and Plant Disease Division, Depaztment of Agriculture and Technical Instruction for Iveland. (Plate VI.) (May, 1923.) On the Channels of Transport from the Storage Organs of the Seedlings of Lodoicea, Phenix, and Victa. By Henry H. Drxon, sc.p., r.r.s., Professor of Botany in the University of Dublin; and Niert G. Batt, m.a., Assistant to the Professor of Botany in the University of Dublin. (Plates VII-X1.) (June, 1923.) i Irregularities in the Rate of Solution of Oxygen by Water. By H. G. Broker, A.R.C.SC.I., A.1.c., Demonstrator in Chemistry in the College of Science, Dublin; and EH. F. Pearson, a.r.o.so.1., Research Student. (June, 1923.) 28. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 86. 37. 28. 39. 40. 41. SCIENTIFIC PROCEEDING S—continued. The Hydrogen Ion Concentration of the Soil in relation to the Flower Colour of Hydrangea Hortensis W., and the Availability of Iron. By W. BR. G. ATKINS, 0.B.E., SC.D., F.L.C. (June, 1923.) . The Comparative Values of Protein, Fat, and Garboliyarate for the Production of Milk Pat. By HE. J. Sup> > =9.1., B.SC. (HONS.), M.R.1.A., Biochemical Laboratory, D.A.T.I. (Ju, i; [Nos. 2u Su Wiss Ca The Utilisation of Monomethylaniline in the Production of Tetryl. By Tuomas Josrren Nowan, D.sc., F.1.c., and Henry W. Craruam, Nobel Researcli Laboratories, Ardeer. (Communicated by Prof. H. Ryan. ) (July, 1923.) Hvidence of Displacement of Carboniferous Strata, Co. Sligo. By Arraur HW. Crarg, 8.a., Trinity College, Dublin. (Communicated by Mr. L. B. Smytu.) ‘July, 1923.) On a Problematic Structure in the Oldhamia Rocks of Bray Head, County Wicklow. By Louris B. Suyru, u.a., so.zp. (Plate XII.) (July, 1923. ) The Hydrogen lon Concentration of the Soil and of Natural Waters in relation to the Distribution of Snails. By W. R. G. Arxins, 0.8.@., sc.p., F.1.c., and M. V. Lrsour, p.sc. . (July, 1928.) Improved Methods of Evaporation in the Laboratory. By H. G. Brcxnr, A.R.C.SC.I., A.1.0., Demonstrator in Chemistry, College of Science, Dublin. (August, 1923.) A Rapid Gasometric Method of Estimating Dissolved Oxygen and Nitroger in Water. By H. G. Brexrr, a.r.c.sc.1., a.ste., and W. H. Aseorr, A.R.C.SC.1., A.I.C., B.sc. (August, 1923.) Ligneous Zonation and Die-Back in the Lime (Citrus medica, var. Acida) im - the West Indies. By T. G. Mason, m.a., so.p., Botanist, West Indian Agricultural College. (Plates XITI-XVI.) (August, 1923.) E [Nos. 25 to 31, price 6s. 6d. | On the Extraction of Sap from living Leayes by means of Compressed Air. By Henry H. Dixon, sc.d., v.n.s.,-Professor of Botany in the University of Dublin; and Nrexwr G. Baxt, m.A., Assistant to the Professor of Botany, University of Dublin. (December, 1923.) Some Experiments on the Convection of Heat in Vertical Water Columns. By H. H. Poots, sc.v. (December, 1923.) On the supposed Homology of the Golgi Elements of the Mammalian Nerve Cell, and the Nebenkern Batonettes “of the Genital Cells of Invertebrates. By F. W. Rocrers BraMBEtt, B-A., s6.D. (DUBL.), and J. Brontii GaTEnsy, M.A. (DUBL.), D.PHIL. (OXON.), D.sc. (LonD.) (Plate XVII.) (December, 1923.) . Phototropic Movements of Leayes—-The Functions of the Lamina and the Petiole with regard to the Perception of the Stimulus. By Niexn G. Bazt, m.a., Assistant to the Professor of Botany in the University of Dublin. (December, 1928.) The Action of the Oxides and the Oxyacids of Nitrogen on Phenylbenzylether. ‘By Hue Ryan, p.sc., and Jonn Knann, pu.p., University College, Dublin. (February, 1924.) The Action of the Oxides and the Oxyacids of Nitrogen on Hthyl-@-Naphthyl- ‘ether.. By Hueu Ryan, p.sc., and Jonn Kwann, px.p., University College, -Dublin. (February, 1924, ) The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylethylene- ether. By Hucn Ryan, p:sc., and Trrmnor Knnny, u.sc., University College, Dublin. (February, 1924.) The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylether. By Huen Ryan, p.sc., and Peter J. Drumm, u.sc., University College, Dublir. (February, 1924.) The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylene Oxide. By Hueu Ryan, p.sc., and Nrcnonas Cuntinann, pa.d., University. College, Dublin. (March, 1924.) The Habitats of Limnaea truncatula and L. pereger in relation to Hydrogen Ion Concentration. By W. R. G. Arxins, 0.8.u., Fu.c., and Mari V. Lesour, v.sc., Marine Biological Laboratory, Plymouth. (February, 1924. [Nos. 82 to 41, price 5s. 6d. | $$$ DUBLIN > PRINTED Ai’ THE UNIVERSITY PRESS BY PONSONBY AND GIKBS. Nor THE SCIENTIFIC PROCEEDINGS: OF THE .) f 4 209K 2 fr. CCR @ 41929 oe, w J ROVAE DUBIN, SOCIENY. Vol. XVIL, N.S., Nos. 49-47. AUGUST, 1924. 42—_ EXPERIMENTS ON THE POSSIBLE EFFECT OF VITAMINS ON QUANTITY OF MILK AND BUTTER FAT. By E. J. Suesny, F.R.C.Sc.1., B.Sc., M.R.I.A., Bio-Chemical Laboratory, D.A.T.I. 43._A MECHANICAL DEVICE FOR SHALING OFF RADIUM EMANA- TION TUBES. By H. H. Poouz, Sc.D. 44_NOTES ON THE FILTRATION AND OTHER ERRORS IN THE DETERMINATION OF THE HYDROGEN ION CONCENTRATION OF SOILS. By W. R. G. Arxins, Sc.D. 45—VARIATIONS IN THE PERMEABILITY OF LEAF-CELLS. By Henry H. Drxon, Sc.D., F.R.S., Professor of Botany in the University of Dublin. 46—NOTES ON ACARINE OR ISLE OF WIGHT BEE DISHASE. By Lizut.-CoLoneL ©. SamMan, R.A.M.C.; and J. Bronté Gatensy, M.A. (Dust.), D.Pam. (Oxon.), D.Sc. (Lonp.), M.R.1.A., F.R.M.S., Professor of Zoology and Comparative Anatomy in the University of Dublin. (Plates XVIII and XIX.) 47—NOTE ON A PHYSICAL METHOD OF SEPARATING THE FATS IN BUTTER-FAT. By Fenrx E. Hacxert, M.A., Pu.D., Professor of Physics, College of Science, Dublin; and T. A. Crowiey, A.R.C.Sc.I., Assistant-Demonstrator, College of Science, Dublin. [Authors alone are responsible for all opinions expressed in their Communications. ] DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATLH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 2. 1924. Price Four Shillings. Koval Bubltw Soctety. aaa FOUNDED, A.D. 1781. INCORPORATED 1749. OO 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 reoula- tions governing the presentation of communications at Scientific Meetings of the Society, and the publication of papers in the Scientific or Neonomic 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 oadatda as a paper in the Proceedings of the Society shall furnish to the Chief Iixecutive 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 Teonomic Proceedings of the Royal Dublin Society, as directed by the Committee of Science. The copyright of papers so published shall be the property of the Society. ‘« Wifty 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 Hxecutive Officer for Science for further information concerning the preparation of papers and illustrations for press. SCIENTIFIC PROCEEDINGS. VOLUME XVII. 1. Experiments on the Hlectrification produced by Breaking up Water, with Special Application to Simpson’s Theory of the Electricity of Thunder- storms. By Professor J. J. Nouan, .a., D.sc.,and J. ENRIGHT, B.A., M.SO., University College, Dublin. (June, 1922.) 2. Cataphoresis of Air-Bubbles in Various Liquids. By Tuomas A. McLaueuum, M.S0., A.InsT.P., University College, Galway. (June, 1922.) 3. On the Aeration of Quiescent Columns of Distilled Water and of Solutions of Sodium Chloride. By Professor W. EH. ApENrEY, A.R.C.SC.I., D.SC., F.I.C. ; Dr. A. G. G. Lronarp, F.R.0.8sc.1., B.S0., F.I.c.; and A. RicHarpson, A.R.O0.80.1., A.I.c. (June, 1922.) [Continued on p. 3 of cover. [ 333] No. 42. EXPERIMENTS ON THE POSSIBLE EFFECT OF VITAMINS ON QUANTITY OF MILK AND BUTTER FAT. By H. J. SHEEHY, F.R.C.Sce.1., B.Sc., M.R.1.A. [Bio-Chemical Laboratory, D.A.T.I.] (Read FEBRuaRy 26. Printed Aprin 24, 1924.) THAT vitamins, or accessory food factors, are necessary for growth and continued _. metabolic activity of the body, is no longer a matter of controversy. At least three different accessory factors, now known as A, B, and ©, function for different purposes: A supports growth and strengthens resistance to disease, B also supports growth and helps to maintain nervous efficiency, and C prevents skin diseases, vascular, and other disorders. The vegetable kingdom is the ultimate source of the vitamins, which cannot be synthesised de novo by ; animals, but the suckling mother can, to a certain extent, utilise the reserves of her own body to supply the vitamins in her milk, even where her diet is deficient in these materials. When these reserves are exhausted, however, the vitamin content of milk is proportional to that of the food from which the milk is derived.? While a definite relationship between the vitamin content of the food of ~ a lactating animal and its milk is established, there is no information as to the part played by these factors in the processes of milk production. There is evidence to show that the activity of the mammary gland cells is dependent’ on internal secretions from other parts of the body. Developments in the ovary and in, the uterus are succeeded by evolution of the mammary cells,? ~ and there is an immediate response in milk production after the injection of a small dose of pituitrin into the blood of a lactating animal.t Probably there are other internal secretions, as well as some chemical substances of food, whieh act as hormones for the mammary cells, and the possible effect of some or all of the accessory food factors suggested itself. To investigate this problem three goats were selected and fed with a view to exhausting the stores of vitamins A, B, and C in animals I, II, and III, respectively, for some time previous to parturition and subsequent to it. This was done by feeding each goat on a ration complete in all respects except the vitamin. whose effect on the milk yield was afterwards to be investigated. When the animals had milked for several weeks, each was given, without otherwise changing the diet, the particular factor in which the previous food had been lacking, and the effect of the addition noted. The determinations made were total milk, butter fat in milk, and the weight of the animal. The 1The Newer Knowledge of Nutrition. M‘Collum, 1919. Chapter vi. 2Jy, Biol. Chem, xxvii, 33. Ibid. xly, 119. Ibid., 1, 339. Biochem. Jr., xv, 540, Jr. Agric. Se., xiii, 144. : i ‘i ; ° Physiology of "Reproduction. Marshall, 1910, page 580. Proce. Roy. Soc., Ixxxviis, 422.” 4Proe. Roy. Soe., IxxxivB, 16. Jour. Dairy Se., 1, 475. Ibid., iv, 474. Quarterly Jr. Physiol., vi, p. 315. SCIENT. PROG. R.D.S., VOL, XV. No. 42. 8s 304 Seientific Proceedings, Royal Dublin Society. animals were weighed once per week always at the same hour, and the milk fat was tested by the Gerber method in tubes previously checked by the Adams paper coil method. The goats were housed at the Albert Agricultural College, and experimental work was performed partly there and partly at the College of Science. The details of the original scheme are outlined in Table I. TABLE I. | Goat f. | Goav II. Goa III. | a = ne Date of M;: 193 | 30 March, ’25 ¢ i]. HG Parturition. THE Ea yo “2 30 March, "23. 20 April, ’23. | | | Roots, ... Tlb. per day | Fish Meal, 3 lb. per day | Crushed Oats, 13 Ib. per day Dried Yeast, ... 1 lb. Pi | White Wheaten | Fish Meal, + lb. 5 Flour, 6 OES % White Wheaten | Dried Yeast, ... 3 lb. 5 Deficient Flour, 1b 3 | Linseed Oil, ... 1 pint per week Diet. | Bran, + Ib. ” Coco Fat, doo) gy lib.” oop | | Linseed Oil, ... 1 pint per week Linseed Oil, ... 3 pint per week! | | Above lacking in Vitamin A, | Above lacking in Vitamin B. Above lacking in Vitamin C. Similar to above with the | Similar to above plus 6 ounces Similar to above plus 80 c.es. Complete | Linseed Oil replaced by Cod | of marmite per week, i.e. with | of orange juice per day, i.e. with Diet. Liver Oil, i.e. with Vitamin A | Vitamin B added. Vitamin C added. added. Simultaneously it was shown, by tests on rats, that the deficient diet fed to goat I contained but a trace, if any, of vitamin A, and that fed to goat II similarly lacked vitamin B, and, by tests on guinea pigs, that the diet of goat III was likewise deficient in yitamin C. Unfortunately goat III did not give sufficient milk to warrant proceeding with her, and consequently the test with vitamin C was necessarily abandoned. Goats I and II gave a good flow of milk and remained quite normal for a time; but after some weeks of lactation both became constipated, and their weight dropped. This produced a temporary disturbance in the yields of milk and butter fat, and in order to restore normal health the diet was necessarily so altered temporarily that some vitamins were incidentally given to the animals during the convalescent stage. The disturbance in general health was not due, however, to lack of vitamins in the previous diet, because normal health continued on that same diet supplemented by an extra half-pint of linseed oil per week—a material known to contain none of these substances. It became necessary to feed the two animals for another considerable period on diet free from vitamin A and B respectively in order to exhaust any vitamin reserve in the: body before applying a test. In this connexion it 399 Peet of Vitamins on Quantity of Milk and Butler Fat. L, SHEKHY— TABLE IT. Goar ILI. Goar I. Fat over period | Mean °/, °/, Fat in Daily Milk SEABSSMHAADO WAM 10 1D HHH Ha SHH HK SH HH HH SH tH 3119 10 YG HH Hm sH NADPH AM OW PIGS HK SODHABWMwOIN HH HSH SH SH SH SH SH SH St tlt Hc SH rt a1 oD 19 Porro oornN 1D HAD +H SHH HH SH SH SH Daily Milk] °/, Fat in Yield (1b.) | Daily Milk =A) S| aa | | S| ealecalsenlsealtealton) Stsitoltalsteale | italia! bs | HoH SH OSH SHS HSH HH HHH SH HHH SH] SHH HH HHH CO cS CO co CoH HH HHH | HHH HH HH HSS] SS as | Sian |——-—~ sei = a _— = = = | 4 S) e bs g = BIDS DBDODANDHNISOE DDO BWIA HI9 6 HDRMROHAAWMANIOr-DRBOANSD 1D sO O63) OS N A _ A |) saa AAA ANA AN S11 9 5 Shel eas tateitciteaesteaiculcarcutcaren ANA oc om GON oS eae OCD See Se SS | > |) i _ Fak = SS 2) = etl ‘SPAUAUO OT “Sny 5 *6 ysnsny 07 1% ATne ‘g Apap 04 9 oUNEG TOIT JoIp Ul Suryory gq uruteyL 9g A[ng oy ) StuE wory popraoad g urureyt 64 Sg ut 9 Ae OF 9 f Pep A OUT OB ly AUTO BH TAN 96 AINE 07 1 4TnE Pept @ ULB A WO yarp UT SapoRy g urTTEIT A Ot} POPIA eles Aare : : -ord gy UrmeyT A On = orm > Yo) > more re i) = 343 x x ~H Se nw — ———S DS AAD DE 19 69 1D DN rs He E19 DH HID M10 MS O19 69 G2 OUD oH Ho Ho HHH co HH HHH HHH Ht HH IDO SO NS IDI O19 OH co od SHH co HHH co HH HH SH SNS IDO OND DH GM SH SH GD SH clo SH cr SH SH cacentseasealsor|sealeealeenloseetcoleetlsrelseclsosl ten cesl selena tenis calchoal toa eeol eral tena palsealccal sel calsealchoningsabpealstoslteelsHealsHoa|oelstoa|toalts9| Sees coleHonltonl Heel heel shoal toler | cheats et ANANANAAAAANANANNANANAANAAAANTAAN NAAN ANAANAANANAN NANANAANAANAAAGA ales - ———— 2 go a FAA AH IDOM DHOOM ADHIDONMHROUMD Cr OSS SHNMH NON DROM IDOE HRONADH 4 2 are eored toa] ANAS NAAN os ne WAR AA NAINA 5S a! S) n i = = *SPABATO “0g toquraydeg 09 gz Ajne utoay qarp ur SaLyoRy YW ULUTeyT A Bg O79 OROTTEAO120) WOAZ pIpPTAord W ULUTBITA GT 19q0}0Q WOIZ Yorp ULSUTYORl Y UrUreqt A 336 Scientific Proceedings, Royal Dublin Society. has been shown that the body stores up greater quantities of vitamin A?* than of either B or Cj The lactation period was, in consequence of this irregularity, well advanced before the effect of adding the vitamins could be determined. When, eventually, the tests were proceeded with, the yields of both animals were fortunately regular. Table II gives particulars of the results. In the case of goat I the addition of vitamin A was effected by substituting one pint of cod liver oil for one pint of linseed oil; and in the ease of goat II the addition of vitamin B was effected as in the original scheme, i.e., by adding six ounces of marmite per week. The results indicate that the addition of vitamin A to a diet, so composed of ordinary unpurified food as to contain the minimum of that factor, does not affect either the quantity or the fat content of the milk. The same conclusion holds for vitamin B. Apparently neither A nor B is concerned with the activity of the mammary gland cells, or else their presence in extremely small quantity suffices. At any rate it would not appear as if the quantity of either milk or butter fat can be influenced by the feeding of the accessory food factors. It is, however, realised that the tests reported in this paper were so limited in extent that a conclusive statement from these determinations, on the possible effects of the accessory food substances im this connexion, cannot be made. SUMMARY. Neither the quantity of milk nor its richness im fat appears to be influenced by the quantity of either vitamin A or B in the diet of the lactating goat. 1Report on the Present Knowledge Concerning Accessory Food Factors (Vitamins), 1919. Medical Research Committee, chapter ii. Ecker No. 43. A MECHANICAL DEVICE FOR SEALING OFF RADIUM EMANATION TUBES. By H. H. POOLE, Sc.D. (Read Aprin 29. Printed Juty 3, 1924.) In the extraction and purification of radium emanation for therapeutic purposes it is essential that the operator should be guarded against the dangerous effects of excessive exposure to the radiations emitted. The problem which has hitherto presented the greatest difficulty in this respect in the Irish. Radium Institute is that of the protection of the operator’s fingers during the division, by sealing off, of the long capillary tube, containing the emanation, into pieces, each about 1:5 em. long, suitable for insertion in serum needles. It is essential that each seal should be absolutely reliable, and that there should be no enlargement of the tube beyond the diameter of 0:85 mm., which is the maximum allowed by the standard needle. As the process involves the simultaneous rotation and drawing out of the tube, it is difficult to ensure satisfactory results if the parts to be separated are held in forceps. Even the use of rubber gloves, or finger stalls, renders the process so much more difficult that the time required for the sealing off is considerably lengthened. As the division of the tube into short lengths is carried out immediately after the emanation has been drawn into it, so that the activity of the tubes is increasing every instant, this attempt at protection is not of much use, so that the usual practice has hitherto been to hold the tube in the bare fingers, and trust to speedy work. This method has proved satisfactory, in so far as no serious damage has occurred to any operator’s fingers, but the small 8 and y ray activity that soon appears in the tubes is sufficient to cause an appreciable amount of temporary injury, so that the apparatus, photographs of which are shown in figs. 1 and 2, has been constructed to obviate all handling of the tubes during the sub-division. The apparatus has been in continuous use for some weeks, and works extremely well. It is constructed chiefly of meccano parts, and the scale may be judged from the fact that the holes in the metal strips which form the framework are 4 inch apart, between centres. As may be seen from fig. 2, the device may be described as a crude form of miniature lathe, with two collinear chucks, in which the capillary tube may be held, rotated independently at the same speed. Each chuck is allowed about 3 mm. end play in its bearings, so that the space between their near ends may be anything between 75 mm. and 135 mm. They are normally pressed apart by the two pivoted fingers and the wire spring seen _ In the photographs, but may be locked in the ‘‘near’’ position by the cam seen near the bottom of fig. 2. A swinging frame is suspended from the end of the SCIENT. PROC. R.D.S., VOL. XVII, NO. 43. 3T 338 Scientific Proceedings, Koyui Dublin Soctety. fixed cantilever, and carries a small gas jet, for which the nipple of a Primus stove proved very suitable. The position of this jet can be readily adjusted both horizontally and vertically, so that when the arm on the left side of the apparatus is raised the small flame is brought to a position midway between the chucks. The swinging framework is sufficiently flexible to enable the flame to be moved towards either chuck by applying a slight side pressure to the operating arm. The construction of the chucks is best shown in fig. 3, which is drawn to scale, actual size; one of them is shown in section. Hach consists of a piece of steel rod, 14 mm. long by 6:3 mm. (4”) diameter, with a central hole, which for the greater part of its length is 3:2 mm. (#”) in diameter, but at one end is. contracted to 17 mm., thus forming an internal shoulder. Osterhout’s important and pioneer researches (9, 10, 11, 12) have introduced the electrical method for the determination of changes in permeability in homogeneous tissues like those of the great kelps; he and others have shown that the electrical resistance of a tissue is a measure of the permeability of the protoplasmic membranes of its component cells. The methods of making electrical contact with the tissue under experiment, which he found suitable with Laminaria, are, however, not available for leaf-tissues, which are move or less spongy, and are coated with an unwettable cuticle. Consequently it was necessary, if the electrical resistance method was to be used, to find some other means of effecting this end. After a considerable amount of experiment the following method was found satisfactory, and adopted :—A square centimetre is cut with a template from the leaf to be investigated. It is laid across two electrodes (H, fig. 1) formed of two square pieces of platinum foil 0-5 cm. on the side. Hach electrode is given a rectangular fold so as to appear L-shaped in profile. The two horizontal parts face each other and support the leaf-square. To the vertical part of each electrode is soldered a piece of platinum wire (D) which passes through the sealed end of a glass tube. The end of the latter supports the electrode rigidly. The glass tubes are filled with mercury (#) and are fixed in a small frame made of three dises of cork (A, B, C) connected by a glass rod (ZL). The ends of flexible leads dipping into the mereury connect the electrodes with one of the arms of a Kohlrausch bridge. In order to secure good contact between the electrodes and the cut edge of the leaf-square a little sap, pressed from leaves of the experimental plant, is placed upon each electrode. The electrodes themselves are coated with platinum-black to obviate polarization effects. Before cutting out the leaf-square from the leaf, it is advisable to smear the upper surtace of the latter with a thin film of vaseline. This prevents the sap on the electrodes from creeping across and short-cireuiting the tissue. SCIENT. PROG. R.D.S., VOL. Xvit, No. 45, 3x 300 Scientific Proceedings, Royal Dublin Society. The upper surface of the square is turned downwards as it lies on the electrodes. In the other arm of the bridge a high adjustable resistanee (10,000—400,000 chms) is introduced. In practice two pairs of electrodes were fixed in the frame so that two leaf-squares might be experimented upon simultaneously. It was soon found that temperature had a large effect upon the resistance, and it was consequently necessary to have the temperature of the tissue under control. To effect this the frame was fitted into a large test-tube (2, fig. 1) and the whole immersed in a glass vessel containing about two litres of water. The water in this vessel could be cooled by the addition of ice, or heated by the passage of an electric current through a submerged wire of German silver supported on a bent glass rod (R). Adjustment of an external resistance made it possible to maintain the water round the test-tube at any desired temperature between 0° and 50°C. Before putting the frame into the test- tube a few drops of water were intréduced to keep the space surrounding the leaf-square moist, and thus prevent the drying up of the leaf-square during the observations. The resistance of the leaf-squares examined was found to be considerable, amounting to from 200,000 ohms to 600,000 ohms at 0° C. The relation of resistance to temperature in the leaves. of Hedera helix may be seen in figs. 4, 5, 6, and in the leaves of Syringa vulgaris in figs. 2 and 3. Here the ordinates are resistances measured in ten thousands of ohms (thus 12 on the vertical scale indicates 120,000 ohms), and the abscissae are temperatures in degrees centigrade. Fig. 4 contams the records of the behaviour of three leaf-squares of Hedera helix. The curve A shows the change of electrical resistance of a square cut on February 5th from an old last season’s leaf of ivy, fixed on the electrodes as deseribed and raised slowly through the range of 0°-50°. Each reading of resistance was carried out after the water round the test-tube was at the temperature recorded for 15-20 mins. Next day the same leaf-square was heated through the same range, and its resistances are recorded in the curve A’. B and B’ are similar curves traced from the behaviour of a similar square cut from another old leaf. The curve Bd gives the resistances of the same square through the same range after it had been killed by exposure to chloroform vapour for somewhat over an hour. C, 0’, and Cd are similar records for a square cut from a young leat also of the previous season. The idiosyncrasies of the three squares are noticeable. There are consider- able absolute differences in resistance of the three, and differences in the form of their curves. They all show a marked fall of resistance with rise in temperature. In A, however, the rise between 30°-40° is accompanied by a rise in resistance. So far as my experience goes this is a rare occurrence. The reduction after heating is another regular feature. Thus A’, B’, and ©’ are each lower than A, B, and C respectively. Heating appears to initiate a progressive change which continues to reduce the resistance. Thus in the case of the series “of experiments recorded in C and C’ it was found that the resistance of the leaf-square on being cooled to 0°, immediately after being raised to 50°, rose to 43:3 X 10* ohms from 115 X 10* ohms. Next day its resistance at 0° was only 26:0 X 10* ohms. It had been kept during the night at a temperature of about 10°-11°. This pregressive and lasting reduction of resistance caused by exposure Drxon— Variations in the Permeability of Leaf- Cells. bol to high temperatures explains the fact that, generally, the resistances observed in summer are less than those of the same plant found in winter and spring. The amount of the reduction in resistance is by no means regular, as inspection of fig. 4 will show. Fig. 5 also gives a striking example. The squares A and B were cut from the same mature leaf of Hedera helix, and included symmetrically-situated pieces on opposite sides of the mid-rib. Their resistances throughout the first heating were almost identical. On the second heating a marked difference was noted. It is recorded in the curves A’ and B’. Syringa vulgaris 28 A heal es oer OS 2nd 26 NS Ad~ 2 —-e — « | after exposure Resistance. 5 s 10 IS 20 as 30 35 40 Temperature, Fie. 1. Fie. 2. The changes induced by the heating do not always appear to be lethal. This is evidenced by the fact that sometimes the recovery of resistance after the heating may be almost complete, eg. B’, fig. 5. Also a similar change is observed when the heating is carried only as high at 35°. Thus a leaf-square of Hedera helix heated from 0° to 35° on February 28th had a resistance of 41:7 < 10* ohms at 0°, and 10:8 X 10* ohms at 35°. On the next day its resistance at 0° was 30:8 X 10* ohms and 11-8 X 10* hii at 35° The curves show a rapid fall in resistance between 0° and 15°. They are less steep at the higher temperatures, but often show a steeper bend between “—3- -e. Curve Tas ist Rening: July 23rd. 24th \ carbon busulphide Jena Ad wo --0 4s sO Resistance, Ay rittga. vulgayis May /40.,/5h and 16th A and B A’ and B’ 20 SRS OMES'S Temperature. Fig. 3. 40 Curves for fst healing. and 4s 50 Resistances in Ohms x10* 60 Hedera Relix Feb Sth. 20th, $2 A,B enc. C ——_—__—___—__ Curves for fst healing: INS ACL SSeS SS « m. 2nd bs 50 Bd and Ch—-— —-— — Curves e@fler exposure o Chloroform vapour. § 10 bo a Ee ao yea eo CG ee) Temperatures in degrees C., Fie. 4. Tresislance, s 10 1S 20 2s Dixon— Variations in the Permeability of Leuf-Cells. 393 40° and 50°. The initial steepness is most marked in those with the highest resistanee, viz. 40-60 X 10 ohms. The final steepening, which, in some cases at least, seems associated with lethal changes, is, so far as I have observed, most marked in young immature leaves which rapidly change colour about and over 40° C. (see-fig. 6). Over the interval of temperature used the conductivity temperature- coefficient for 10° (i.e. the ratio of the conductivity at T° + 10° to the conductivity at T° C.) varied considerably, viz. from 1:70 to 0:92. The average of the thirty observations plotted in fig. 4 gives 1334 for the coefficient—a figure almost identical with that obtained by Osterhout for Laminaria. This figure is a measure of the variation of permeability with temperature. The resistance of the leaves of Hedera helix appears to grow with age, and this in spite of the fact that during the interval growth in thickness takes place. Hedera helix April Ist. ord. A and B———___———- Curves for fst feating A and B'---—----- “ " Qnd NX Ad and Bd —-—-—.—-~ - @fler exposure Lo Chloroform vapoun Hedera helix May 12m 3B —#—+—+— Old - 12 s. o 5 10 Nall 5 ~ Xs Av ok Bed. SRST Sm a ae re yin ratuce x us 50 o ‘2 re Ore 9 Fic. 5. Fig. 6. TABLE. Age of leat. Date. Temperature. Thickness. - Resistance. \ Young mature .. 975/724 35 0-250 mm. 22-8 < 10* ohms. | Old ie nee Fs 13° 0-375 mm. 28-9 x 10! ,, ( Young p «. 14/4/7724 10-2° 0-238 mm fay 23. WO 5, (old ca a: 35 10-2° 0-277 mm. lo STOR Young immature ... 12/5/724 0° 0-200 mm. 14-49 x 10' ,, Old mature a 0° 0-400 mm. 22-75 X 10* ,, A ——e—o——__ Young leaf of 1924 > I9as = ee foe N A ENG Se R5 30 35 40 45 50 354 Scientific Proceedings, Royal Dublin Society. The probable explanation of this rather surprising observation seems to be that the increase of thickness with age is due to the increase in size of the intercellular system, and does not involve an increase in the cross-section of the conducting cells, or an inerease in the surfaces of contact of these cells. The resistance due to the protoplasm of the constituent cells evidently depends directly on this total surface. Expansion of the individual cells and of the intercellular spaces would tend to reduce this surface, and hence explain the erowth of resistance with age. The contrast between the resistances of the living and of the dead leaf is in every case very marked. That of the living leaf has been found to vary between twenty-two times and six times that of the dead leaf. The form of the curve for the dead oe generally resembles that for the living tissue. This is made plain by fig. 2, in which Ad X 10 is obtained by plotting the temperature-resistance curve for the dead tissue to ten times the vertical scale. The large difference between the dead and the living tissue is in agreement with the view that the electrical resistance of a tissue is controlled by the semi-permeable property of the cells composing it. Figs. 2 and 3 summarize experiments on the leaves of Syringa vulgaris similar to those on Hedera helix already described. The same features, already noticed in the case of Hedera helix, are apparent in the curves for the leaves ot Syringa vulgaris. Preliminary experiments on the effect of light in altermg the permeability have not given conclusive results, inasmuch as the effects observed might be attributed to the heating produced by the illumination. The salient fact brought out by the foregoing experiments is the reduction of resistance or the increase of protoplasmic permeability produced by a rise of temperature. Thus we may expect the permeability of leaf-cells to become about doubled by a rise of temperature from 10° to 30°, and at 20° the permeability of the cells will be 50 per cent. greater than at 10°. A rise in the temperature of the surroundings when the atmosphere round the leaves is saturated will of course produce a rise in the general temperature of the plant including its leaves. Under these conditions all the cells of the plant, if they behave like the leaf-cells, will become more Perreaple, and probably important effects result from this change. In 1905 Brown and Escombe (2 and 3) came to the earnalinetiosa on theoretical grounds that the temperature of a leaf when insolated or exposed to diffuse light did not, in the specific cases considered, differ from that of its surroundings by more than + 1:64° or — 1:84°. In the same year, however, F. F. Blackman and G. L. C. Matthaei (1) measured by thermo-electric means the temperatures of shaded and insolated leaves. It was found that a difference of 16° might be established, while a leaf exposed to diffuse light is often 1°-3° above its- shaded surroundings. The leaves used were of Prunus lawrocerasus, and the thermo-junction was embedded in the mid-rib of the leaf. Recently I have carried out some similar experiments with leaves of Hedera helix and of Syringa vulgaris. The thermo-couple I used was made of constantan and copper. The elements of the couple were in the form of fine silk-covered wires, each end of the copper wire being soldered to a piece of constantan wire. In order to reduce the thermo-electric effect so as to give a convenient deflection with a sensitive galvanometer seven metres of No. 42 s.w.g. (diam. = 0:1 mm.) constantan wire were used. The copper element was formed of 50 em. of No. 36 s.w.g. (diam. = 018 mm.) copper wire. Dixon— Variations in the Permeability of Leaf-Cells. ~ 355 The extreme ends of the elements, stripped of their silk covering, were twisted together and soldered. The soldered junction so formed was reduced to about 0:25 mm. in length by an oblique cut of a sharp scissors. By this means the actual junction was very close to the point formed, so that when the double wire was pushed through even a thin leaf, the junction was still embedded in the tissue, as the point was just emerging. The other ends of the constantan wires were soldered to thicker copper wires, and these latter junctions, together with the support carrying the copper terminals, were immersed in a small: vessel of petroleum. By this means, and by means of a connexion already described, thermo-electric errors were avoided (4, 5, 8). The thermo-couple thus arranged gave a deflection of 26 mm. per 1° difference of temperature of the junctions. Re Spe a egw 80 100 60: 40 3 10 20 30 40 50 60 Fic. 7. In cool weather with intermittent sunshine early in May an insolated patch on a leaf of Hedera helix was found to be 47°-6:2 above an adjacent patch which was shaded by an overhanging leaf. Under similar circumstances a temperature difference of 6:3°-6:6° was found in leaves of Syringa vulgaris. In each case the periods of insolation were less than 5 mins., and the leaves were not normal to the direction of illumination. The leaves were exposed to a brisk breeze and were well supplied with water. Rapid temperature changes were observed, fluctuations of 1°-6° within a minute being not uncommon. Records of four sixty-second observations are shown in fig. 7. Bright diffuse light caused a rise of 1° or 2° above a shaded portion. In more favourable weather, differences as large as those found by Blackman and Matthaei would probably have been obtained. : From these experiments it is evident that during sunshine there are temperature-differences between shaded and insolated leaf-areas of 10° and more. The results connecting resistance and temperature show that these tempera- ture-differences will lead to large fluctuations in permeability, the cells in the heated areas becoming more permeable. It has already been pointed out that such differences in permeability acting in concert with the hydrostatic tension throughout the plant furnish a mechanism for the distribution of dissolved substances through the plant body. 356 Scientific Proceedings, Royal Dublin Society. REFERENCES. 1. Buackmay, F’. F., and Marruart, GasrietiLe L. C.—Experimental Researches in Vegetable Assimilation and Respiration. TV.—A Quantitative Study of Carbon-Dioxide Assimilation and Leaf-Temperature in Natural Illumination. Proce. R. 8., vol. 76B, p. 402. 1905. 2. Brown, H. T.—The Reception and Utilization of Energy by a Green Leaf. Nature. March 30, 1905. 3. Brown, H. T., and Escomps, I'—Researches on some of the Physiological Processes of Green Leaves, with special reference to the Interchange of Energy between the Leaf and its surroundings. Proc. R. S., WO, “Toe, jos ZY, UW: 4, Drxon, Henry H.—Observations on the Temperatures of the Subterranean On Organs of Plants. Trans. Roy. Ivish Academy, vol. xxxiis, p. 145. 1903. A Thermo-electrice Method of Cryoscopy. Proc. Roy. Dublin Society, vol. xii, p. 275. 1910. Notes from the Bot. School of Trin. Coll., Dublin, vol. ii, p. 121. Transport of Organic Substances in Plants. Presidential 1924. Address to Section K—Botany—of the British Association. 1922. Notes from the Bot. School of Trin. Coll., Dublin, vol. iii, p. 207. The Transpiration Stream. London University Press. 8. Drxon, Henry H., and Atkins, W. R. G.—-On Osmotie Pressure in Plants and on a Thermo-electrice Method of Determiming Freezing Points. Proce. Roy. Dubl. Soc, vol. xii, p. 275. 1910. Notes from the Bot. School, Trin. Coll., Dublin, vol. ui, p. 47. 9. OstERHOoUT, W. J. V.—Does the Temperature Coefficient of Permeability il 12. indicate that it is Chemical in Nature? Bot. Gaz., vol. 63, p. 317-320. Oi Conductivity as a Measure of Permeability. Jour. of Biological Chemistry, vol. xxxvi, No. 38. 1918. A Method of Measuring the Electrical Conductivity of Living Tissues. Jour. of Biol. Chem., vol. xxxvi, No. 3. 1918. Injury, Recovery, and Death in relation to Conductivity and Permeability. Lippincott Co., Philadelphia and London. erg Sizer No. 46. NOTES ON ACARINE OR ISLE OF WIGHT BEE DISEASE. By LIKUT.-COLONEL C. SAMMAN, R.A.M.C., AND J. BRONTE GATENBY, M.A. (Dust.), D.Pum. (Oxon.), D.Sc. (Lonp.), M.R.I.A., F.R.M.S., Professor of Zoology and Comparative Anatomy in the University of Dublin. (Puares XVIIL anp XIX.) | (Read Aprin 29. Printed August 21, 1924.) Introduction. In the year 1904 a serious, apparently new, disease of honey-bees broke out in the Isle of Wight, and, spreading rapidly, swept over the British Isles like a plague, completely wiping out the bees in many districts. At least two commissions of investigation failed to find the cause of this disease, and it was only so recently as 1919 that a band of workers in Aberdeen University were able to give the first correct explanation of the cause of Isle of Wight disease in bees. The organism which is the causative agent belongs to the order Acarina. It was first noted by Miss E. J. Harvey, who attached no special importance to its presence in the bee; but it was afterwards seen, and its importance immediately recognized by Bruce White, who has given a description of its effect on the host. The eredit of the discovery of the cause of Isle of Wight bee disease therefore belongs to Bruce White, whose paper on its pathology is the only one up to this date in existence. White’s paper was published simultaneously with other communications from Miss Harvey and Dr. John Rennie. Since Bruce White’s discovery, John Rennie has carried out investigations on the bionomics of this disease, and it is principally to him that we owe what knowledge we have of the habits, pathogenicity, and treatment of this remarkable disease, which is characterized by many peculiarities not hitherto known in parasitology. Our study of this disease in bees has extended about one year. During most of this time one of us has carried out routine diagnostic work for Irish bee-keepers, and it is only more recently, during winter, that a serious attempt has been made to extend our knowledge of the disease: the junior writer had long been interested in the Acarina, and welcomed the opportunity to under- take an investigation of the embryology and anatomy of the causative organism of Isle of Wight disease. Owing to technical difficulties, it is only recently that satisfactory sections of the parasites have been obtained. For nearly a year this work has been carried out with facilities provided solely by the Zoological Department of Trinity College, Dublin. Since the Free State Government kindly granted us fifty pounds for expenses relative to this investigation, we have recently made a little progress. Our work has been SCIENT. PROC, R.D.S., VOL. XVII, NO. 46. BY 358 Scientific Proceedings, Royal Dublin Society. especially hampered by the difficulty in obtaining diseased bees. We must take this opportunity of thanking the Minister of Agriculture, Mr. Hogan, and some of the officials of the Department of Agriculture for the interest they have taken in our work and for the loan of some pamphlets. For the benefit of those not especially acquainted with the subject, the following simple account is given :— In the higher organism, like man, the function of bringing oxygen for respiration to the tissues and cells, as is well known, is carried out by the blood, but among certain of the animal groups, one of which contains the bees, there is found instead a separate system of air-tubes or tracheae which ramify or branch over the tissues carrying oxygen to the cells. Any blockage of these air-tubes brings partial or total asphyxia to the tissues served by the blocked tube, just as stoppage of a blood-vessel going to a part of the body would produce a like change. Biologists have never ceased to wonder at the powers which exist in the thoracic-wing muscles of insects; the rapidity with which these muscles are able to operate, and their endurance are alike remarkable. The air-tubes or tracheae which serve these thoracic muscles with oxygen are the seat of the disease under discussion. The mite or acarid crawls in through the external opening of the tube or stigma, and breeds rapidly within the lumen of the trachea. In later stages the bees are unable to fly, and crawl about the hive or on objects near by: now bees only defecate when in flight, and consequently the ‘‘crawlers’’ retain their foecal matter. This naturally led the earlier investigators to consider that they were dealing with some complaint of the alimentary canal, and no one seems to have thought of a connexion between the state of the gut and a cause of the inability to use the wings. Thus the discovery of the acarids within the air-tubes has proved a remarkable one. Reference to the microphotographs will give a good idea of the disease. In Pl. XVITI, fig. 1, is a photo of the adult male mite; it has four pairs of legs, with the piercing organs, palps, and chelicerae between the first two pairs. These mites are very slow walkers, and are unable to run like many of the free living aearids. In Pl. XIX, fig. 7, is a photo of the upper part of one of the thoracic tracheae; there is a main stem below, which branches to form a Y, from the upper part of which smaller air-tubes branch off, and finally form exceed- ingly fine twigs that carry air to the tissues. In Pl. XVIII, fig. 4, is another tube showing a branching lower down. The end on the left below d connects to the exterior by the air-pore or stigma. In Pl. XVIII, fig. 3, the lowermost. part of the thoracie trachea is shown. The mother mite crawls into the tracheal tube through the stigma, and generally comes to rest just imside. In Pl. XVIII, fig. 2, is a clean tube flattened under the cover-slip and photographed; in fig. 3 is a tube which shows the first stage of infection. The mother mite at ? has already laid two egos, 1’ and 1?; in Pl. XIX, fig. 5, the family has increased, and the mites are working their way up the tube; while in fig. 6 the infection is heavy, and the female mites at ? have been breeding rapidly. Some time after the tube or trachea becomes infected another fact may be noted: in Pl. XVIII, fig. 4, is shown very clearly, at dd, a black smudge on the wall of the tube. This smudge is caused by the mites, and eventually spreads in large patches and areas, and finally the tube becomes blackened, as shown in Pl. XIX, fig. 8. Such bees are doomed, and are unable to fly, or at all events to fly far. The mite is believed to feed by sucking the haemocoel or body cavity (or blood) fluid of the bee which surrounds these tubes, so that the mites, by pricking through the wall with their chelicerae are able to draw fluid for SAMMAN AND GaAtenBY—Acarine or Isle of Wight Bee Disease. 399 thei nutriment. According to John Rennie it is the feeding of the mites that weakens the bee. Rennie’s (1921) paper (jomtly with White), in the Transactions of the Royat Society of Edinburgh, has been superseded by his late memoir (No. 6) published by the North of Scotland College of Agriculture, and entitled ‘‘ Acarine Disease Explained.” In the Transactions of the Royal Society of Edinburgh (as well as in Rennie’s joint papers) is one by Bruce White on the ‘*Pathology of {sle of Wight Disease in Hive Bees.’’ White mentions that female mites may advanee as far as the secondary tracheae before depositing eggs; in the later stages of attack the mites may attain the small tracheae, the thoracic air-saes, and the vessels of the head. Regarding the tracheal system, White states that the change in colour of the tube is accompanied by an increasing hardness and brittleness of the parts, which become rigid. In the early stages of attack there may show, here and there, a few fragments of brownish matter, the faeces of the invading adults. Such granules increase in number, finally forming bands upon the tracheal wall. They are brownish or yellowish in colour, and when densely aggregated appear black. White also states that such foecal matter may become inhaled into the smaller passages, forming emboli in the tracheoles. The muscular system may be visibly affected, the fibres showing atrophis changes; but the number of such showing these signs is small. This author performed the ingenious experiment of blocking up the stigmata, or openings leading to the tubes, in healthy bees, and found that, in some cases, states resembling the symptoms of Isle of Wight disease were produced. White undoubtedly, in this way, has added very strong evidence to the view that the acarid is the causative agent of the disease. Experiments on infection with Acarapis woodi have been made by Miss Elsie J. Harvey, one of Rennie’s assistants. Miss Harvey came to the conclusion that bees were not usually, if at all, infected before emergence from cells. Clean bees placed in a queen cage could be infected if kept in a diseased hive. Miss Harvey showed that experimental infection is not easy to effect, but that the disease is spread by bodily contact between infective and non- infected bees. Rennie, after Bruce White’s and Miss Harvey’s discovery of the causative agent of the disease, attempted to classify the mite. He placed it in Canestrini’s genus ‘‘Tarsonemus’’ and gave it the specific name of ‘‘woodi’’ after A. H. E. Wood, a gentleman much interested in apiculture. Rennie’s excursion into the treacherous grounds of systematic acarology was not altogether successful, for Hirst has removed this form from the genus Tarsonemus and placed it in the genus Acarapis, which is closely allied. Hirst states that there is no nymphal stage either in Acarapis or Tarsonemus, this being entirely suppressed. This disease is spread by adult mites. The Cure or Prevention of Isle of Wight or Acarine Disease. We know that the causative organism of this disease is an acarid—the female of which is tracheate, the male non-tracheate. The healthy bees are infected by peripatetic females, which leave the thoracic tracheae of infected bees, and wander into the tracheae of non-infected bees. The acarids live on the haemocoel fluid of the bees. The problem is to kill the acarids either (a) in the tracheae, by feeding the bees on some substance which might so affect the haemocoel fluid of the host, as to make it toxic to the parasites; or (6) in the tracheae by some gas 0: fumigant which might poison the mites without killing the bees themselves ; 360 Scientific Proceedings, Royal Dublin Society. or (c) on the surtace of the bees, by sprinkling the hive with some substance which would come into contact with the peripatetic gravid females, and thus prevent newly emerged bees from becoming infected; or finally, (d) to find some way of preventing clean stocks of bees from becoming infected. John Rennie and various bee-keepers have tried many methods. The former mentions in one of his papers that many fumigants will kill the mites in experimental chambers: unfortunately Rennie does not appear to have given a list of these substances in any of his published work. In 1923, Wood announced that Rennie had found what might be assumed to be a cure for the disease, and, in his letter to an English bee journal, mentioned the name of an Aberdeen chemist from whom, for the sum of 2s. 6d., bee-keepers might obtain samples of this secret substance. In various bee journals from time to time ‘‘cures’’ for Isle of Wight bee disease have been announced. We ourselves have received visits from various bee-keepers who claimed to have cures for the disease. We have thought it right to publish all our findings, in order that future workers might have the assistance of knowing our methods and our failures. Some, or all, of these methods may have been tried by Rennie—we do not know. We have not had Rennie’s secret treatment analysed, but we have examined lrish bees which were said to have been treated with this cure, and were found still heavily infected. We do not wish to say that Rennie’s treatment is not helpful; we state, however, that it is probably only an alleviative, and not a cure. Some of the ‘‘cures’’ are substances which are slightly toxic to the bees, and by giving the coup de grace to badly diseased and weakened individuals lessen the incidence of infection of the newly-emerged bees. So far as we are aware none of these substances does more than alleviate the disease: this, however, is a step in the right direction, but does not preyent the spread of the disease, and may leave the stock in a weak condition. The danger of such half-measures is that these stocks are a constant menace to clean hives in the neighbourhood. At Trinity College, Dublin, we set up an apiary for diseased bees, and have had six stocks under continual observation. Besides this we have been able to examine and observe at intervals many other stocks, both diseased and healthy. : We arranged our experiments with the object of trying to affect the haemocoel fluid of the bees, and thus to make the conditions unsuitable or toxie for the mites. Two stocks were used as controls, one bemg at Trinity College. The winter months, durmeg which the experiments were carried out, were very favourable for working out percentages of infection, for noting fluctuations if such oeeur, and so on, because no new bees were emerging. Every week we examined: large numbers of bees, and were able to establish the fact that very little, if any, infection of healthy bees occurs during winter within the hives at the winter temperature, and during the special period we observed the hives. This result is in accord with the fact that during winter the mites breed very slowly, and in many eases not at all. On the time taken for bees to become infected. It was important to find out how soon infection of the hive may take place ~ after a parasitized bee has gained access to a clean stock. From our experiments we have been able to gain a clear idea as to this. In April, a frame containing comb with drone brood, capped, was put into SAMMAN AND Garenspy—Aeéarine or Isle of Wight Bee Disease. 361 a White’s queen cage, and the wooden top replaced by a piece of excluder zine allowing free access for diseased bees, but preventing the drones getting out. It was examined on the Sth day, so that drones were seven or less days old. The drones were infected. Hach time the drones were removed and the frame was returned to the White’s cage. The experiment was repeated on Six occasions, ie. on the 7th, 6th, 5th, 4th, 3rd, and 2nd days, and diseased drones were found. In the tracheae of a drone two days old or less, one mite and two eggs were seen (Pl. XVIII, fig. 3). In the trachea of a drone less than twenty-four hours, one mite alone was seen. This seems to prove that the gravid female gains access to the trachea very early and rapidly commences to lay eggs. . Haperiment to show that the brood is free from infection. The Aberdeen workers concluded that the brood is not diseased before emergence. Miss Harvey carried out a number of experiments, and our results confirm her findings. The following is taken from our notes on one of our experiments in this department. (August 8th.) A frame of brood from a native stock infected with Acarapis was put into a ‘‘ White’s’’ queen cage, and a fertile Italian queen placed on the comb; this frame was then put in the hive from which the brood was taken, but no bees were able to get access to the comb. (14th.) Removed the cage. Some young bees had emerged from the cells and the queen had commenced laying. Two more frames of brood now added—one from the same stock, the other from a stock of Acarapis imiected bees, to which a Carniolan queen had been introduced on July 3rd. The young bees from this queen had been examined on August 10th and found infested with Acarapis. (19th.) These three frames were now placed in an upper brood chamber above a strong colony (diseased) with a screen of wire gauze between the chambers. (20th.) Added three more combs of capped brood from diseased colony. (25th.) Put the six frames of brood into a nucleus hive, and transferred them to a fresh locality and allowed young bees to have their first flight. (September 7th.) Transferred frames fo a permanent hive. Queen laying well; brood in all stages. (Novem- ber 22nd.) Examined bees, no trace of Acarapis: shut up for winter. 1923. (February 27th.) Examined bees, no trace of Acarapis. (March 28th.) Examined bees, no trace of Acarapis. (April 11th.) Spring cleaned hive; bees doing well; brood in all stages; plenty of stores. (July 7th.) Stock swarmed (a very large swarm). Removed surplus honey (about 100 lb.); this in a poor honey season. At the present date (April, 1924) this stock is still doing well. Experiments on Diseased Stocks. Four stocks were treated with drugs in food during the winter months of 1923-1924. As is well known, during the colder period of the year the stocks are usually fed on some type of commercial candy. The drugs used by us were mixed with the latter, and by this means the bees were treated in four different ways. 1. Succus Allii of about 15 per cent., in water, sugar added as necessary, according to thickness of food desired. 2. Yadil candy of 23 per cent. 3. Succus Allii of 15 per cent., followed by a treatment of Stockholm tar in candy of 5 per cent. 4. Yadil, as above, followed by Stockholm tar as above. We found that the bees would take candy syrup containing as much as 30 per cent. of Suceus Allii, but not so freely as when the strength of the latter was reduced to 15 per cent. Consequently we worked with 15 per cent. strength. Of the four hives the three treated as in paragraphs 2, 3, and 4, all died 362 Scientific Proceedings, Royal Dublin Society. after some weeks. In the graph in the text figure the Yadil-treated hive is shown in a wavy line. It was 100 per cent. infected at the end of September and the beginning of October. By the middle of October the infection had gone down to about 40 per cent., this probably being due to the hatching out of new bees. Despite the treatment infection steadily crept up, and the stock ultimately died out before the end of the year 1923. Possibly the Yadil was not strong enough, but this was the strength (23 per cent.) sent by the makers. In the case of the two hives treated with Suceus Allii, we had more encouraging results. One stock marked thus ---- in the figure, began at 10 per cent. infection, which remained fairly uniform till December when it rose suddenly, but thereafter fell till all the samples of bees examined were negative. This stock is still negative, and healthy. 1 a pled pul a= Ginccus oth 30-157 Saucers allii ona Gteretrolin Lan. Graph showing percentages of infection of three hives treated respectively with Yadil, Succus Allii, and Succus Alli followed by tar, for four months of the year. The other hive marked -. - - - , gave much the same results, but in this case, Stockholm tar candy was given at the end of December; the stock died out about the end of January, poisoned, we believe, by the Stockholm tar. Never- theless, we consider that if the tar had not been given, there would have been some ground for believing that this stock might have been cured. In the other case where Yadil was given the percentage of infection at the beginning of the experiment was only about 10 per cent., and never rose above 50 per cent., but this stock was the first to die out. These results, meagre as they are, nevertheless lead us to believe that an efficient treatment of Acarine or Isle of Wight disease might be found along these lines. Garlic ean easily be grown by bee-keepers, the bulbs can be dug up, and the juice expressed and mixed up to a strength of 15 per cent. with sugar. SamMan AND Garensy— Acarine or Isle of Wight Bee Disease. 368 This treatment should be tried during the winter months when the bees are clustered. It should be pointed out that our experiments have not been numerous enough to enable us to claim that we have discovered a cure for Acarine disease. But our experience leads us to consider that the method might with benefit be tried by bee-keepers. DESCRIPTION OF PLATES XVIII AND XIX. All figures are from microphotographs. Fig 1, X 400; figs. 2-8, x 120. Fig. 1. Adult male Acarapis woodi. 2, Clear non-infected tube. 3. Early stage of infection, one female (9? ) and two eggs, 1, LP’. 4, Discolouration (d) of the tube caused by mites. 5. Later stage of infection. 6. Multiplication of parasites at its height (summer). 7, 8. Winter tubes showing discolouration and very few mites. ne 2 i] & al Ree Cine i ake ence py PLATE XVIII. -SCIENT. PROC, R. DUBLIN SOC., VOL. XVII. SAMMAN AND GATENBY. PROC. R. DUBLIN SOC., VOL. XVII. SCIENT. SAMMAN AND GATENBY. No. 47. NOTE ON A PHYSICAL METHOD OF SEPARATING THE FATS IN BUTTER-FAT. By FELIX EH. HACKETT, M.A., PxD., Professor of Physics, College of Science, Dublin, AND T. A. CROWLEY, A.R.C.Sc.1., Assistant-Demonstrator, College of Science, Dublin. (Read Aprin 29. Printed AucusT 21, 1924.) Tue texture of butter, an important quality in its marketing value, is controlled primarily by the character of the butter-fat and the churning temperature and also by the conditions incident to the churning, washing, setting, and working of the butter. The most suitable conditions have been arrived at empirically, but our knowledge of the actual constituents of butter-fat and their physical properties is yet insufficient to give a full scientific explanation of the process of butter-making. The work deseribed in this note indicates a method of carrying out a physical analysis of butter and separating to some extent its constituent fats, and so studying their physical properties. As the process is slow and tedious, it has been considered desirable to give this account of some preliminary observations before undertaking a more extended investigation. In the course of an examination of the texture of different butters, it was noticed that those butters which had the firmer texture gave a smaller grease ying on filter-paper. No information on the nature of this oily diffusing substance was found in the extensive literature on the butter industry, though the question of the wrapping-paper for butter is discussed. Apparently the subject has never been investigated. Probably as it was assumed that the fatty acids were present in butter as triglycerides, it was concluded that this substance must be olein M.P. — 4° C., or butyrin M.P.— 76°C. It is impossible, however, to separate butyrin from butter-fat (1) by solution in hot alcohol in the same way as from an artificial mixture of triglycerides. This and other experiments have led to the view that the fatty acids are combined with glycerie in butter mainly as mixed glycerides. Amberger (2) showed by examination ot the portion of the hydrogenated butter-fat soluble in alcohol that the original fat contains butyrodiolein, butyro-palmito-olein, and oleodipalmitin. Of these oleodipalmitin has a melting-point of 38°C., and solidifies about 28° C. The butyro-compounds do not appear to have been isolated. Amberger concluded that the butter examined did not contain more than 3 per cent. of olein. This result is confirmed by the work described in this paper. The oil which diffuses into filter-paper is mainly a substance which solidifies between 10° and 12° C., and liquefies about 19° C. From these constants it cannot be identified with any of the glycerides recorded in Beilstein. : j The initial experiment was made with unpasteurized Irish butter several weeks old. It was distributed in small pats on Whatman filter-papers previously SCIENT. PROC. R.D.S., VOL. XVI, NO. 47. Bz 366 Scientifie Proceedings, Royal Dublin Society. extracted with ether. After twenty-four hours the pats were removed, and the paper scraped clean of adhering butter, and then extracted with ether in a Soxhlet extractor. (It will be convenient to refer to the process by which the butter passes into the filter-paper as diffusion, though obviously diffusion, eapillarity, and viscosity all play a part in it.) 3 grams of a yellowish oily liquid were obtained, which did not solidify when allowed to stand over- night at 15°C., but became solid when placed in a room at 10°C. The solidifying-point was determined by a cooling curve. The extract was heated to 30°C. and placed in a test-tube with a thermometer. The test-tube was separated by an air-space from a boiling tube to ensure more regular cooling. The latter was immersed in a bath containing melting ice constantly stirred. Solidification, or more precisely the clouding of the liquid, began at the bottom at 12°C. and was completed at 10°C. The cooling curve gave a well-marked stationary temperature at 10-2° C., indicating that the extract consists mainly of a single fat or a group of fats having this solidifying- point. A mixture such as butter gives no stationary temperature. The quantity of the extract was too small to make any complete examination by chemical methods. Its iodine value was found to be 41 grams per 100 grams of fat. From this it may be concluded that this extract contains oleic acid in much the same proportion as the original butter. For comparison it may be remarked that a mixed triglyceride such as butyro-palmito-olem would give an iodine value of 38 and butyrediolein of 73. A more systematic investigation was then instituted. Comparison of the diffusion of butter with that of butter-fat showed only a slight difference. The water and the texture of the butter have, therefore, very little influence. As it is desirable to obtain the constituents of butter as far as possible unchanged by subsequent heat-treatment, butter was used throughout these experiments. The diffusion was carried out on Whatman No. 1 filter-papers 15 cm. in diameter, which had previously been extracted with ether. On each filter- paper were placed 15 erams of butter in eighteen small cylinders 1 em. in diameter and about 1 em. high. After varying intervals these butter cylinders were removed, and the paper scraped and extracted with ether in a Soxhlet apparatus. One set of observations was carried out at a temperature of 20°-21° C. which was available in the germinating chamber of the Seed Testing Division of the Department of Agriculture. At this temperature the butter became very soft, and diffusion took place rapidly. An exhaustive experiment to determine the approximate quantity of oil in the butter could, therefore, be carried out in a reasonable time. 75 grams of butter in ninety small cylinders were placed on five sheets of filter-paper, and allowed to remain in the germinating chamber for twenty-four hours. 4 grams of extract were obtained. The residue again placed on fresh filter-papers for twenty-four hours gave 3 grams ot extract. Repetition of the process gave 2 grams. The time was then extended te forty-eight hours, and for the fifth and sixth extract to seventy-two hours, and later to ninety-six hours and one hundred and twenty hours. The increased time was necessary to obtain a workable extract. 20 grams of extract, represent- ing nearly 30 per cent. of the original quantity of butter, were obtained in the course of three weeks. The solidification of this series of extracts, or rather the attainment of a solid-like rigidity with its accompanying opaque- ness, took place between 11° and 12°C. The cooling curve gave no stationary temperature. On standing at 15°C., these samples solidified into a gel-like condition, and the colouring’ matter present became bleached. A progressive change was shown in the iodine values. The following values were obtained :— Hackerr & CrowLey—Method of Separating Fats in Butter-Fat. 367 first fraction, 375; third fraction, 348; sixth fraction, 34:0; butter-fat 32. This approximation towards the value for butter may be due in part to the incomplete removal of the butter from the paper by scraping. Assuming the presence of several substances with different rates of diffusion, it is also possible, though not immediately evident, that the composition of the extract may depend on the time of diffusion. These considerations suggested some modifications in the method of working. The next series of experiments was accordingly arranged to obtain a more uniform extract. The time for diffusion was now reduced to three hours. The paper on which the cylinders rested was cut out of the filter-paper before - extraction. so as to obtain only the substanee which had diffused. This extract was a golden-brown liquid which set to a butter-yellow gel-like solid on standing at 15° C., and gave a stationary temperature on the cooling curve at 11:7° C. The substance was still moist in appearance below this temperature, and when placed im a room at 10° C. it yielded an oil to filter-paper. The solidifying- point of the residue removed from the filter-paper was found to be 116°C. The other extracts obtained between 20° and 21° C. were also allowed to diffuse into filter-paper at 10° C., so as to obtain these extracts as free as possible irom lower melting products previous to analysis. The presence of at least two diffusing oils with different solidifying-points explains the observations made at 16°C. At this temperature 3 grams of extract were obtained in three days. It remained liquid at room temperature. On taking a cooling curve it gave a pasty mass between 12° and 10° C., and solidified completely about 5° C. The two oils were evidently present in nearly equal proportions. No chemical analysis was made on this sample. An experiment on diffusion was also made in a cold room whose temperature varied between the limits 9° and 10°C. The process was very slow, and not more than 1 gram was obtained in three weeks from 74 grams of butter on five filter-papers. To avoid the possibility of extractmg any butter particles, the paper on which the eylinders’ rested was cut away before insertion in the Soxhlet apparatus. A determination of the iodine value gave 82. The iodine value for olein is 86. We may infer, therefore, as one might expect, that the oil diffusing in the cold is olein, but that, im agreement with Amberger’s results, the quantity present is small. The authors are indebted to Mr. Brownlee, Agricultural Analyst of the Department of Agriculture, for an examination of the butter used in these experiments, and of the fat diffusing from it at 20°C. A sufficient quantity of the latter was obtained by combining all the extracts together, and they were partially purified by cooling them to 10° C., and absorbing the liquid present at that temperature in filter-paper. The comparative determinations include the Reichert-Wollny number, indicating the volatile fatty acids soluble in water, and the Avé-Lallemant number, which is calculated from the difference between the insoluble and soluble barium salts of the acids present. Mr. Brownlee (3) has shown that the Reichert-Wollny number has a seasonal variation for Irish butter, reaching its peak value between March and June. As the sample of butter examined was made in March, the figures obtained in the above two determinations are consequently above the average. Butter-fat. Extract. Reichert-Wollny number 31:37 35:03 Avyé-Lallemant number = S7-il7y A GHO Koettstorfer’s saponification value 237-02 235-62 368 Scientific Proceedings, Royal Dublin Society. A more complete examination was made of the extract. It gave 2:74 for the Polenske figure, which estimates the insoluble volatile fatty acids, and 29:9 for the Kirschner value, which is stated to correspond to the butyric acid. The average values for butter of the above Reichert-Wollny number are: Polenske 3-1, and Kirschner 26:0. When the diffusion had gone on for two months there were 27 grams of residue. Accordingly, allowing for 16 per cent. of water, 48 per cent. of the butter diffused into the paper at 20°C., and thus the extract constituted nearly 60 per cent. of the butter-fat. Correspondence between the two analyses is therefore to be expected. The numbers show, how- ever, that the percentage of the butyric acid is higher than in butter, though they do not indicate that all the butyric acid is contained in the extract. In combination with the iodine values already given, they would accord with the view that the diffusing oil consisted in a great part of fats of a constitution similar to the butyro-palmito-olein mentioned by Amberger. The interest of these experiments does not lie so much in the information they yield on the actual fats present in butter, which cannot be in any degree exact, but more in their relation to the churning temperature and consequent texture of butter. According to Hunziker the churning temperature may vary within wide limits, possibly from 5°-24°C. In the northern and central part of the dairy region in the United States the variations are confined within much narrower limits, 9°-12° C. in summer, 13°—21° C. in winter. In Ireland it is found desirable to cool the cream as low as 2:2° C., and begin churning at 6:7° C. The presence in Irish butter of a large quantity of a constituent solidifying only when cooled below 12°C. indicates the importance of low-cooling in Irish ereamery practice to give the fat globules sufficient firmness before churning. Experiments require, however, to be made on a much larger scale, and over a lone period to carry out the separation of the fats in sufficient quantity to determine their physical properties or their chemical composition. SUMMARY. The grease ring formed around butter when placed on paper has been investigated, using fat-free filter-paper. The fat which diffused into the paper was extracted with ether. When the diffusion took place at 20° C., the extract was an oil solidifying between 10° and 12°, and liquefying about 19° C. In comparison with butter it contained a higher percentage of butyric acid, and also of unsaturated fatty acids. Almost 50 per cent. of the content of a sample of Irish butter diffused into paper at this temperature. At 10° C. the diffusion from butter was extremely slow, and the oil obtained had an iodine value corresponding to triolein. The small amount obtaimed and the slowness of the diffusion are in agreement with the work of Amberger, who estimated the percentage of olein present in butter as not more than 3 per cent. REFERENCES. BuiytH and Ropertson.—Proe. Chem. Soe. 1889. Amprrcrr.—Zeitseh. nahr. Genussm. 1918. 35, pp. 313-380. BrowNLEE—Journal Dept. of Agr., Ireland. V. 10. 1910. . Droop-RicumMonp.—Dairy Chemistry. 1914. Hunzixer.—The Butter Industry. 1920. oo 0 NW On INDEX TO VOLUME XVII. tional Muse a Particles, Detonating Action of (POOLE), 93. Asporr (W. EH.) and the late H. G. Brecker. A Rapid Gasometrie Method of Estimating Dissolved! Oxygen and Nitrogen in Water, 249. ADENEY (W. E.), A. G. G. LEONARD, and Miss A. M. RicuarDson. On the Aeration of Quiescent Columns of Dis- tilled Water and of Solutions of Sodium Chloride, 19. Apples, A Phytophthora Parasitic on (LAFFERTY and PETHYBRIDGE), 29. ATKINS (W. R. G.). Notes on the Filtra- tion and other Errors in the Deter- mination of the Hydrogen Ion Con- centration of Soils, 341. The Hydrogen Ion Concentration of the Soil in relation to the Flower Colour of Hydrangea hortensis W. and the Availability of Iron, 201. and Miss M. VY. Lrsour. The Habitats of Limnaea truncatula and L. pereger in relation to Hydrogen Ion Concentration, 327. The Hydrogen Ion Con- centration of the Soil and of Natural Waters in relation to the Distribution of Snails, 233. Bau (N. G.). Phototropic Movements of Leaves—The Functions of the Lamina and the Petiole with regard to the Per- ception of the Stimulus, 281. and H. H. Dixon. On the Channels of Transport from the Storage Organs of the Seedlings of Lodoicea, Phenia, and Vicia, 185. On the Extraction of Sap from Living Leaves by means of Com- pressed Air, 263.- SCIENT. PROC. R.D.S., VOL. XVII, INDEX. Brecker (the late H. G.). Improved Methods of Evaporation in the Labora- tory, 241. and W. E. Assorr. A Rapid Gasometric Method of Estimating Dis- solved Nitrogen and Oxygen in Water, 249. and H. F. Pearson. Irregularities in the Rate of Solution of Oxygen by Water, 197. Bitter Cassava, Growth of, and Transport of Organic Substances in (Mason), 105. BRAMBELL (F. W. R.) and J. B. GATENBY. On the supposed Homology of the Golgi Elements of the Mammalian Nerve Cell and the Nebenkern Batonettes of the Genital Cells of Invertebrates, 275. Butter Fat, Possible Effect of Vitamins on Quantity of (SHEEHY), 333. A Physical Method of Separating the Fats in (HaAcKErT), 365. Cataphoresis of Air Bubbles in various Liquids (McLavcGHuin), 13. Channels of Transport from the Storage Organs of the Seedlings of Lodoicea, Pheniz, and Vicia (DIxXon and BALL), 185. CLiapHamM (H. W.) and T. J. Notan, The Utilisation of Monomethylaniline for the Production of Tetryl, 219. CuarK (A. E.). » Evidence of Displace- ment of Carboniferous Strata, County Sligo, 225. CLARKE (Miss R.), T. DILLON, and V. M. Hincuy. Preliminary Experiments on a Chemical Method of Separating the Isotopes of Lead, 53. ConnotLy (Miss A.) and H. Ryan. The Action of the Oxides and the Oxyacids of Nitrogen on Ethylphenylurethane, 125. 4a 370 Convection of Heat in Water (POOLE), 267, CULLINANE (N.) and H. Ryan. The action of the Oxides and the Oxyacids of Nitrogen on Diphenylene Oxide, 321; on Ethyl-o-tolylurethane, 119. Detonating Action of a Particles (Pootn), 93. DILion (T.), Miss R. CuAREE, and V. M. Hincuy. Preliminary Experiments on a Chemical Method of Separating the Isotopes of Lead, 53. Diphenylene Oxide, Action of Oxides of Nitrogen on (RYAN and CULLINANE), 321. Diphenylether, Action of Oxides of Nitrogen on (RYAN and Drumm), 313. Diphenylethyleneether, Action of Oxides of Nitrogen on (RYAN and Kenny), 305. Diphenylurethane, Action of Oxides of Nitrogen on (RYAN and DONNELLAN) 113. Displacement of Carboniferous Strata, Co. Sligo (CLARK), 225. Distribution of Activity in Therapy (Poour), 45. Drxon (H. H.). Variations in the Per- meability of Leaf Cells, 349. and N. G. Bau. On the Channels of Transport from the Storage Organs of the Secdlings of Lodoicea, Phenix, and Vicia, 185. — On the Extraction of Sap from Living Leaves by mears of Com- pressed Air, 263. DONNELLAN (Miss A.) and H. Ryan. The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylurethane, 113. Drumm (P. J.) and H. Ryan. The action of the Oxides and the Oxyacids of Nitrogen on Diphenylether, 313. eo) Radium Electrical Design of High Tension Trans- mission Lines (JEFFCorT), 71. Enricur (J.) and J. J. Nowan. Experi- ments on the Electrification produced by Breaking up Water, with Special Application to Simpson’s Theory of the Electricity of Thunderstorms, 1. Ethyl-@-naphthylether, Action of Oxides of Nitrogen on (RYAN and Keane), 297. Ethyl-o-tolylurethane, Action of Oxides of Nitrogen on (RYAN and CULLINANE), ably, Index. Ethylphenylurethane, Action of Oxides of Nitrogen on (RYAN and CONNOLLY), 125. Evaporation, Improved Methods of (BECKER), 241. Extraction of Sap from Living Leaves by means of Compressed Air (Dixon and BALL), 263. Laboratory GaATENBY (J. B.) and F. W. R. BRAMBELL, On the supposed Homology of the Golgi Elements of the Mammalian Nerve Cell and the Nebenkern Batonettes of the Genital Cells of In- vertebrates, 275. and C. SAmMMAN. Notes on Acarine or Isle of Wight Bee Disease, 357. GILMORE (Miss J. G.) and T. JoHnsen. Libocedrus and its Cone in the Trish Tertiary, 66. The Lignite of Washing Bay, Co. Tyrone, 59. Golgi Elements, supposed Homology with Nebenkern Batonettes (BRAMBELL and GATENBY), 275. Growth and the Transport of Organic Substances in Bitter Cassava (Mason), 105. Hackett (F. E.). Note on a Physical Method of Separating the Fats in Butter Fat, 365. Helium, Occurrence of, in a Well at Lucan (LEONARD and RICHARDSON), 89. Hincuy (V. M.), T. Dinion, and Miss R. CuaRKE. Preliminary Experiments on a Chemical Method of Separating the Isotopes of Lead, 53. Hydrogen Ion Concentration, Filtration and other Errors in Determination of, for Soils (ATKINS), 341; in relation to the Flower Colour of Hydrangea hor- tensis W. (ATKINS), 201; in relation to the Distribution of Snails (ATKINS and LEBOUR), 233; in relation to the Habi- tats of Limnaea truncatula and L. pereger (ATKINS and LEBOUR), 327. Isle of Wight Bee Disease (SAMMAN and GATENBY), 357. Isotopes of Lead, Preliminary Experi- ments on a Chemical Method of Separating (DILLON, CLARKE, and HINCHY), 53. Jerrcorr (H. H.). The Electrical Design of A.C. High Tension Transmission Lines, 71. Index. 371 Jounson (T.) and Miss J. C. GILMORE. Libocedrus and its Cone in the Irish Tertiary, 66. The Lignite of Washing Bay, Co. Tyrone, 59. Kane (J.) and H. Ryan. The Action of the Oxides and the Oxyacids of Nitro- gen on Ethyl-@-naphthylether, 297; on Phenylbenzylether, 287. Kenny (T.) and H. Ryan. The Action of the Oxides and the Oxyacids of Nitro- gen on Diphenylethyleneether, 305. Larrerty (H. A.) and G. H. Prruy- BRIDGE. On a Phytophthora Pavrasitic on Apples which has both Amphigynous and Paragynous Antheridia; and on Allied Species which show the same Phenomenon, 29. Leaf Roll Disease of Potatoes (MuRPHY), 163. Lresour (Miss M. V.) and W. R. G. ATKINS. The Habitats of Limnaea truncatula and L. pereger in relation to Hydrogen Ion Concentration, 327. The Hydrogen Ion Con- centration of the Soil and of Natural Waters in relation to the Distribution of Snails, 233. LEonArD (A. G. G.), W. E. ADENEY, and ‘Miss A. M. RICHARDSON. On the Aera- tion of Quiescent Columns of Distilled Water and of Solutions of Sodium Chloride, 19. and Miss A. M. RIcHARDSON. The Occurrence of Helium in the Boiling Well at St. Edmundsbury, Lucan, 89. Libocedrus in the Irish Tertiary (JoHNsoN and GILMORE), 66. Ligneous Zonation and Die Back in the Lime (Mason), 255. ‘Lignite of Washing Bay, County Tyrone (JOHNSON and GILMORE), 59. Limnaea truncatula and L. pereger, Habi- tats of, in relation to Hydrogen Ion Concentration (ATKINS and LEBOUR), 327. McLavenuuin (T. A.). Cataphoresis of Air Bubbles in Various Liquids, 13. Mason (T. G.). A Note on Growth and the Transport of Organie Substances in Bitter Cassava (Manihot wutilissima), 105. Mason (T. G.). Ligneous Zonation and Die Back in the Lime (Citrus medica, var. Acida), 255. Milk and Butter Fat, Possible Effect of Vitamins on Quantity of (SHEEHY),, 333. Milk Fat, The Comparative Values of Protein, Fat, and Carbohydrate for the Production of (SHEEHY), 211. Milk Yield, The Variation of, with the Cow’s Age and the Length of the Lactation Period (WILSON), 97. Monomethylaniline, Utilisation of, for Production of Tetryl (Nonan and CLAPHAM), 219. Murpuy (P. A.). On the Cause of Roll- ing in Potato Foliage; and on some further Insect Carriers of the Leaf Roll Disease, 163. Nebenkern Batonettes, supposed Homo- logy with Golgi Elements (BRAMBELL and GATENBY), 275. Nitrogen, Estimation of, Dissolved in Water (BECKER and ABBOTT), 249. Nouan (J. J.) and J. Enricur. Experi- ments on the Hlectrification produced by Breaking up Water, with Special Application to Simpson’s Theory of the Electricity of Thunderstorms, 1. Nouan (T. J.) and H. W. CiapHam. The Utilisation of Monomethylaniline for the Produetion of Tetryl, 219. O’Donovan (J. L.) and H. Ryan. The Action of the Oxides and the Oxyacids of Nitrogen on Phenylbenzylurethane, 131. Oldhamia Rocks, Bray Head, Co. Wick- low, A Problematic Structure in (SmMytH), 229. O’Toote (P. K.) and H. Ryan. The Action of the Oxides and the Oxyacids of Nitrogen on the Phenylureas, 139. Oxides and Oxyacids of Nitrogen, The Action of, on Diphenylene Oxide (Ryan and CuLLINANE), 321; on Di- phenylether (Ryan and Drum™), 313; on Diphenylethyleneether (Ryan and Kenny), 305; on Diphenylurethane (RYAN and DoNNELLAN), 113; on Ethyl- @-naphthylether (RYAN and KEANE), 297; on Ethyl-o-tolylurethane (RYAN and CULLINANE), 119; on Ethylphenyl- urethane (RYAN and CONNOLLY), 125; on Phenylbenzylether (RYAN and Kwang), 287; on Phenylbenzylurethane (RYAN and O’DoNovAN), 131; on Phenyl- methylurea (RYAN and SWEENEY), 157; on the Phenylureas (Ryan and O’TOoLE), 139. O72 Oxygen, Solution of, by Water, Irregu- larities in Rate of (BECKER and PEARSON), 197; Estimation of (BECKER and ABBOTT), 249. Pearson (H. F.) and the late H. G. Brecker. Irregularities in the Rate of Solution of Oxygen by Water, 197. Permeability of Leaf Cells, Variations in (Dixon), 349. Peruypripge (G. H.) and H. A. Larrerty. On a Phytophthora Para- sitic on Apples which has both Am- phigynous and Paragynous Antheridia; and on Allied Species which show the same Phenomenon, 29. Phenylbenzylether, Action of Oxides of Nitrogen on (RYAN and KEANE), 287. Phenylbenzylurethane, Action of Oxides of Nitrogen on (RYAN and O’DoNovay), 131. Phenylmethylurea, Action of Oxides of Nitrogen on (RYAN and SWEENEY), 157. Phenylureas, Action of Oxides of Nitro- gen on the (RyAN and O’TooLs), 139. Phototropic Movements of Leaves (Baul), 281. Phytophthora Parasitic on Apples (LAFFERTY and PETHYBRIDGE), 29. Poourt (H. H.). A Mechanical Device for Sealing off Radium Emanation Tubes, 337. On the Detonating Action of a Particles, 93. Some Experiments on the Con- vection of Heat in Vertical Water Columus, 267. Some Further Notes on the Dis- tribution of Activity in Radium Therapy, 45. RIcHARDSON (Miss A. M.), W. E. ADENEY, and A. G. G. LEonarD. On the Aera- tion of Quiescent Columns of Distilled Water and of Solutions of Sodium Chloride, 19. and A. G. G. Lronarp. The Oc- currence of Helium in the Boiling Well at St. Edmundsbury, Lucan, 89. Ryan (H.) and Miss A. Connotty. The Action of the Oxides and the Oxyacids of Nitrogen on Ethylphenylurethane, 125. Index. RYAN (H.) and N. CuLuinane. The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylene Oxide, 321; on Ethyl-o-tolylurethane, 119. and Miss A. DONNELLAN. The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylurethane, 113. and P. J. Drumm. The Action of the Oxides and the Oxyacids of Nitro- gen on Diphenylether, 313. and J. KEANE. The Action of the Oxides and the Oxyacids of Nitrogen on Ethyl-@-naphthylether, 297; on Phenylbenzylether, 287. and T. Kenny. The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylethyleneether, 305. and J. L. O’Donovan. The Action of the Oxides and the Oxyacids of Nitrogen on Phenylbenzylurethane, 131. and P. K. O’Toonr. The Action of the Oxides and the Oxyacids of Nitrogen on the Phenylureas, 139. and M. J. SWEENEY. The Action. of the Oxides and the Oxyacids of Nitrogen on Phenylmethylurea, 157. SaMMAN (C.) and J. B. GarEnBy. Notes on Acarine or Isle of Wight Bee Disease, 357. SHEEHY (EH. J.). Experiments on the Possible Effect of Vitamins on Quantity of Milk and Butter Fat, 333. The Comparative Values of Pro- tein, Fat, and Carbohydrate for the Produetion of Milk Fat, 211. Simpson’s Theory of the Electricity of Thunderstorms (NOLAN and ENRIGHT), 1. Smytu (L. B.). On a Problematic Struc- ture in the Oldhamia Rocks of Bray Head, Co. Wicklow, 229. Snails, Distribution of, in relation to Hydrogen Ion Concentration (ATKINS and LEBOUR), 233. Solution of Oxygen by Water, Irregu- larities in the Rate of (BECKER and Pearson), 197. Sweeney (M. J.) and H. Ryan. The Action of the Oxides and the Oxyacids of Nitrogen on Phenylmethylurea, 157- Index. Tetryl, Utilisation of Monomethylaniline for Production of (NoLAN and CLAPHAM), 219. Transmission Lines, Design of High Tension (JEFFCOTT), 71. Transport of Organic Substances in, and Growth of Bitter Cassava (Mason), 105. Vitamins, Possible Effect of, on Quantity of Milk and Butter Fat (SHEEHY), 333. Byes) Water, Aeration of (ADENEY, LEONARD, and RIcHARDSON), 19; Convection of Heat in (Pootn), 267; Electrification Produced by Breaking up (Nonan and ENRIGHT), 1; Estimation of Dissolved Oxygen and Nitrogen in (BECKER and Apport), 249; Rate of Solution of Oxygen by (BECKER and PEARSON), 197. Wiuson (J.). The Variations of Milk Yield with the Cow’s Age and _ the Length of the Lactation Period, 97. END OF VOLUME XVII. SCIENT. PROG. R.D.S., VOL. XVII, INDEX. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. SCIENTIFIC PROCEKEDINGS—continued. . On a Phytophthora Parasitic on Apples which has both Amphigynous and Paragynous Antheridia; and on Allied Species which show the same Phenomenon. By H. A. Larrerty and Grorer H. Prraysriper, Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (Plates I and II.) (June, 1922.) . Some Further Notes on the Distribution of Activity in Radium Therapy. By H. H. Poors, m.a., so.p., Chief Scientific Officer, Royal Dublin Society. (June, 1922.) . Preliminary Experiments on a Chemical Method of Separating the Isotopes of Lead. By 'aomas Ditton, p.sc.; Rosaninp Cuarke, d.sc. ; and Victor M. Hrncay, s.sc. (Chemical Department, University College, Galway). (July, 1922.) . The Lignite of Washing Bay, Co. 'yrone. By 'I’. Jonnson, D.sc., F.L.S., Professor of Botany, Royal College of Science for Ireland; and Janu G. Gitmorn, B.sc. (Plate III.) (August, 1922.) . Libocedrus and its Cone in the Irish Tertiary. By T. Jounson, p.so., ¥.1.s., Professor of Botany, Royal College of Science for Ireland; and Janz G. Gitmorn, B.sc. (Plate IV.) (August, 1922.) . The Electrical Design of A.C. High Tension Transmission Lines. By H. H. Jerroorr. (August, 1922.) The Occurrence of Helium in the Boiling Well at St. Edmundsbury, Lucan. By A. G. G. Lonard, F.R.¢.sc.1., PH.D., F.t.c., and A. M. Ricuarpson, A.R.O.SC.1., A.I.d. (Plate V.) (August, 1922.) [Nos. 1 to 10, price 9s. ] On the Detonating Action of a Particles. By H. H. Poonn, m.a., so.p., Chief Scientific Officer, Royal Dublin Society. (December, 1922.) The Variations of Milk Yield with the Cow’s Age and the Length of the Lactation Period. By James Witson, m.a., B.sc. (December, 1922.) A Note on Growth and the Transport of Organic Substances in Bitter Cassava (Manihot utilissima). By I’. G. Mason, u.a., B.sc. (December, 1922. ) [Nos. 11 to 18, price 1s. 6d.] The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylurethane. By Huew Ryan, v.sc., and Annr Donnewnay, .sc., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on Ethyl-o-Tolyl- urethane. By Hueu Ryan, pv.sc., and Niononas CuLuinann, PxH.D., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on Mthyl-Phenyl- urethane. By Hues Ryan, p.sc., and Anna Connouty, u.sc., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on Phenyl-Benzyl- urethane. By Hue Ryan, p.sc., and James L. O’Donovan, m.so., University College, Dublin. (February, 1923.) The Action of the Oxides and the Oxyacids of Nitrogen on the Phenylureas. By Hueu Ryan, p.sc, and Perer K. O’Toonz, m.sc., University College, Dublin. (February, 1928.) The Action of the Oxides and the Oxyacids of Nitrogen on Phenyl-Methy]l- urea. By Huau Ryan, p.sc., and Micnarn J. Sweeney, u.sc., University College, Dublin. (February, 1923.) [Nos. 14 to 19, price 4s.] On the Cause of Rolling in Potato Foliage; and on some further Insect Carriers of the Leaf-roll Disease. By Paun A. Murpay, so.p., a.R.¢.so.1., Seeds and Plant Disease Division, Department of Agriculture and Technical Instruction for Ireland. (Plate VI.) (May, 1923.) : On the Channels of Transport from the Storage Organs of the Seedlings of Lodoicea, Phenix, and Vicia. By Henry H. Dixon, sc.p., F.r.s., Professor of Botany in the University of Dublin; and Nicer G. Batt, m.a., Assistant to the Professor of Botany in the University of Dublin. (Plates VII-XI.) (June, 1923.) £ Irregularities in the Rate of Solution of Oxygen by Water. By H. G. Broxsr, AR.¢.SG.1., A.1.c., Demonstrator in Chemistry in the College of Science, Dublin; and H. F. Prarson, a.R.c.sc.1., Research Student. (June, 1923.) SCIENTIFIC PROCEEDINGS—continued. Bis. “e Cheat . The Hydrogen Ion Concentration of the Soil in relation to the Flower Colour 36. 37. 38. 39. AO. Al. of Hydrangea Hortensis W., and the Availability of Iron. By W. R. G. ATKINS, 0.B.E., SC.D., F.C. (June, 1923.) . The Comparative Values of Protein, Fat, and Carbohydrate for the Production of Milk Fat. By HE. J. SHEEHY, F.R.C.SC.1., B.SC. (HONS.), 1.R.1.A., Biochemical Laboratory, D.A.T.1. (June, 1923.) [ Nos. 20 to 24, price 7s. 6d. | 5. The Utilisation of Monomethylaniline in the Production of Tetryl. By Tomas Josep Nowan, D.sc., F.1.c., and Henry W. Crapaam, Nobel Research Laboratories, Ardeer. (Communicated by Prof. H. Ryan.) (July, 1923.) . Eyidence of Displacement of Carboniferous Strata, Co. Sligo. By Arruur Ti. Cuarx, 8.a., Trinity College, Dublin. (Communicated by Mr. L. B. Smyru.) (July, 1923.) . On a Problematic Structure in the Oldhamia Rocks of Bray Head, County Wicklow. By Louis B. Ssvrn, w.a., scp. (Plate XII.) (July, 1928.) . The Hydrogen Ion Concentration of the Soil and of Natural Waters in relation to the Distribution of Snails. By W. R. G. Avxins, 0.8.5., so.p., ra.c., and M. V. Lrzour, p.sc. (July, 1923.) . Improved Methods of Eyaporation in the Laboratory. By H. G. Brcxrur, A.R.C.SC.1., A.I.c., Demonstrator in Chemistry, College of Science, Dublin. (August, 1923.) . A Rapid Gasometric Method of Estimating Dissolved Oxygen and Nitrogen in Water. By H. G. Buecker, a.p.c.sc.1., aac., and W. H. Assorz, A.R.C.SC.1., A.I.G., B.Sc. (August, 1923.) . Ligneous Zonation and Die-Back in the Lime (Citrus medica, var. Acida) in the West Indies. By T. G. Mason, m.a., so.p., Botanist, West Indian Agricultural College. (Plates XITI-XVI.) (August, 1923.) [Nos. 25 to 31, price 6s. 6d. | . On the Extraction of Sap from living Leaves by means of Compressed Air. By Henry H. Dixon, se.p., F.x.s., Professor of Botany in the University of Dublin; and Nient G. Baur, u.a., Assistant to the Professor of Botany, University of Dublin. (December, 1923.) . Some Experiments on the Convection of Heat in Vertical Water Columns. By H. H. Poous, sc.v. (December, 1923.) . On the supposed Homology of the Golgi Klements of the Mammalian Nerve Cell, and the Nebenkern Batonettes of the Genital Cells of Invertebrates. By F. W. Rocurs Brampxtt, B.a., sc.D. (DuBL.), and J. Bronrii GarEnsy, M.A. (DUBL.), D.PHIL. (OXON.), D.sc. (LonD.) (Plate XVII.) (December, 1923.) 5. Phototropic Movements of Leayes—-The Functions of the Lamina and the Petiole with regard to the Perception of the Stimulus. By Nien G. Batt, m.A., Assistant to the Professor of Botany in the University of Dublin. (December, 1923.) The Action of -the Oxides and the Oxyacids of Nitrogen on Phenylbenzylether. By Hueu Ryan, p.sc., and Joun Kranz, pu.p., University College, Dublin. (February, 1924.) The Action of the Oxides and the Oxyacids of Nitrogen on Hthyl-8-Naphthyl- ether. By Hueu Ryay, p.sc., and Joun Kann, px.p., University College, Dublin. (February, 1924.) The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylethylene- ether. By Huan Ryan, p.sc., and Trrence Kenny, u.sc., University College, Dublin. (Hebruary, 1924.) The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylether. By Huen Ryan, p.sc., and Peter J. Drumm, u.sc., University College, Dublic. (February, 1924.) The Action of the Oxides and the Oxyacids of Nitrogen on Diphenylene Oxide. By Hucu Ryan, p.so., and Nicuonas Cunninann, pu.p., University. College, Dublin. (March, 1924.) The Habitats of Limnaea truncatula and L. pereger in relation to Hydrogen Ion Concentration. By W. R. G. Arnis, 0.8.5., F.1.c., and Marie We ae p.sc., Marine Biological Laboratory, Plymouth. (February, 1924, [Nos. 32 to 41, price 5s. 6d. ] ——— DUBLIN: PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIBBS. » > ))) BD) DDD 2B)» 5 S))> — + Boy » »y>D - >» vy >» BO 2. yy »> Jy A277. ) yy») )D Dy» MD > DY _Y - > yyy yy FW 2Y BY. mat 4 LN aS 1 ‘ ww © I~ a (iw rt we )) ) ‘ ie 2? PP? NVW YD Ww. WP 2D DY DG) DID DD > Dy — Ane Be i> ID? DDR >> 2) WY y # a Sey D> DDN DLY PDP DVD PII )D MD 7 WIP »))) wprDP» YD ) ae YE B2»> 3 2) ! ) p)) DD» »Pp_»» a) >») A » 2» PEP 2» $3< Hf DP 2) Y)_ J) BD) ry > +) ye >» D> DD aw» 2? w .2PLeaw)> EP >) “DD ep)» @D WP) iP Dp») PRE I» PD?) > Wr >»>.) © YOY? DY)? DLP] Po» » “> DIPD ZPD») D2) , Ds stp JP Pee 2 DID) D VPI) OY MVPD) VY > 2) 2D) D> S & yy» > Pwoy Ay 22D? . 2) )') > WIP Wy» 2 WP) By YVQD OD Dd) N » oe > yy) DD) py sy : jy a >» ] > OS 2D D272 0) 4) DB) D> D>) ) a» woes me ve DIL > Pe Bp >» D 22 bp Pye D7 e abe D) DY DOB D Ly Vors OD D2 bo DD D>) D> DD? DD > DD DIQ 2D ? D> D> YD) WDM? D2 272-2 > > PA Pw YM. » WD Vow La e 7D D> 1 DW)>> 727 DSd Or mm >> Dip pw PP? 2D 2d VS «) 12D D> hep p> PP WEP a> YW » can a? 2am ED DD we BS De Eo Dim >> [eRe =a BA» VD wa 75> » DW » YAY? 222 229 222 Pm ») Reece» »\ 02 we ep 2 i.D DP - »D >? 235 ? a2 ) Pia » DY)?» ) we > 1”: Sap Db D> RD) D) MEPP pL» 22. > vaw 2» D> xem dD »).912) 9.9229 ) 2. aw DD.» RD» DLW aI 4D D 4 ep DD Dow DD PD) TWIP DY yop Pee 2 ap 2 2» yy ar5 5 > iia nom). DY TDD Py 92d + yr) Dd DP? 2 MY OW)? 2Y 7 ~. ay EY IY LDP 2) BPM» DDE 22D) DLW MND I) 209 P , oe [pin ID WD 2»? SPA pry DYDD ) yyw “IND py> > > ) LA pew» PDD) De SP» ! 2 > p DY DD_>» yo )_2)-d D>» SIR Pere Vie pps ; ot « X yp) ~ : » ed: » y . Pau? * me) 4 XN we Beow D >) D> 3D) D>) DH» Dd», DID) » ))) aa» DD DD oe, 3D aE >)» D2 B >). DWP> ) > _»») np ») » _») vx yp 2» 2. ) » DD a)? y 5-4 PD) DY» DD») WD -») Wyo } > > Dy». ~ y) »y PP. y ) ») > D» ES. » MY DO D iy » DD) » »- ) » YD Jy?) » wp Jy 4 y)) Ly Dy ¢ OF > »»D DPD»? a SEP LED 2? DP ©) DPD yay 1) DP ADD) D> _Dpy TAD) DD. > ID?) DDI)? > DDD FY Y»>d oye —4 > DD) PZ» IVP HW DD») Wo ) ? 2) DYDD YD DO b)B _ )) Dy» yy >>» yD. >). IY? BD VI? DP ; PPD D2 3 ))y ‘ J) 5 > Y ) yyy’ > yy >) y ) > PD ip) dy» WIP DD Dv) Sx 2 > py d*) ) I)» » Wo soe) WOFD> JID, 45 a ee Sons 7) Yow) Pay y i» 'D ZZ D \ SY ) - ae 2); > ye iS i meee a lt > >> yy)” »> rr »y wm» > 3) . > D2 2 PYM Ip Wwe; 2 ima ay) + »?>D aes 5 > i i» Is o> Bh ib wip p a 2D 9) BD)2 }o Wwe »d. F< v4 »»Dd je yD ») >) y 2 - >? > 7) i a 7 2») D> BY) > DD YH 2>»8B DIA» >. DY? 729) 0D YD D> 2 v72- 22> » > FADD Dw. WD 2a 22 © oes >>) 2 Dd>> 9x29 >) DP >> 27>)» 2D DWV, DID )D VED DD) a 22> snd do> DD av >> DID AST = od Di a y) )) dD D?>.2 2D . ya > Xs » -] Oe 2 Dp 1 Dy»). wp YZ D2”) 222 ee 22) 52 )o» DI? 25D DD ‘eb> 2» > DP», DB VBNGY)22 P22. yy. » ») Des)» ty Rees 25> 2 ID) ») DDD) 9 2.2 2? ) ID). ad a 2%) o re 2 rps Dd? SPD 93 > BD DP») VD DID > 2D.» 2D pe >? DD xD D DL) D>» 9» D?2 oD) 92 D @D_FRPM> Vay. ai) py) a 22 @e> 6] 2d») Fd)» 2D Wlray Y) ) »> p27) ) S45 22. my & 2. 92 DD») 2? pw) 4 7D 10 FP > D DY VQ DY. 92 ray, DIDDY?» DY I) 2) DY BD) } 2. 227) DP ay yy) WD 27A.DY) wr 22> IP) DDN. I) DD 2 DY la@22 » YD DPI?” oe LD xy » ade 2 25S > 7 +)> >. Ap yy wD > >») : > <= yD Pd Ja yy } ~! De >» y Vid 8 1h >. = :. 2.2) 22> D272) BHD 2) 3») DD), 29 ? 2)? 2 ? i) 2D ) 2 »)') aa Ye Zp d ye Ss ee 2 a ee, Be ae DY Pig’ Dd Ml i | See ee se =— a — — = es: = SZ ~~ nee r) w = < « a a z 9 = BB y = cS < z ° 0 x = = 1 oD 3 9088 01303 4244 wo. — SSS at we Sax Zenmee BE oS) Ah Died ELE 7% oye = Ve RES = dwId w AN =, Vys SS SSES AAR fed a vue PULL uve S ve AW Tetaathafe ! S a “ae « « | ill Cac 7 CC ae ie RT hha vile JCUes os AMEN Swweuuyul Wy ple ye Te vee TRITON AE Re PR Rd Ae CV BS yw” we = CEN SSeS ANS werd GIs» [adel Ss wy SA “AN + V3 le E pa OS SS es we LIES { 4 Pig eI \ ui CE eK