Harta watery man Sete aren ner cates ge Pocky wy eae Grate Sree eerie wage yet f 2 Ne ti edt RGU) OG CN ee Wrest ve bayeart, ia Sie DOOSStOa Ser aie ala Sou Whee tnd hs WRN MCR * ‘ cette ty f Giewy . Hawt 3 i “y 4h beh wae ene eat rats “yells ests! SOONG 4a te ‘ a Matyas ie 4, Does ay me Hee ete we 4 a Srey Rew ae xara sea +a Atta lee fet Adie) sista VE aU iat Pris vane redo ", wide FWA eta CAN Atcha aii Sate ch ey vy on Wy uy ape yey Re ey yey At) alee ae Bon vat ists rh oo ' Nites he fatistaptisty Wi yerangy ay braeagh teh SR 5 ia bet ' Aint se iiterteer ede heee ty ; RARE Gat col : Mesto arca ied ws t v IVETE RN He fee VEU eit NS BCR Oe eRe BK relates ¥ rit . Hastie Raat) ; Shieh \ 7 - ‘4 , Ay 9: ey ’ ayy SRL ny . ‘ : Cee CR aN fyb! Pa ty Rue nS aN va Va : - Rat ate Fy Seth te Dey tay), raters vidytat eth ati . wr 79:9 voy i pai ‘ Pit ast a GUS ; ‘ ie ae Pe "i * Sth HF) eSNG Mattel d i Delt 3) i i, : ee fe aes be f ' Ohne a Aes : ahinnc fydiee PMs ee ee ' He reaeghe » 1 ty : . 44k oh F . F ey Nee) ‘ ‘ ' ' \ ‘ ete Ure hy epilation Y + mn whe ton Toate ' i ba Md if of i ' ' ‘ a 3 7 " Ls ; ra Ferrey Shores we ‘ Pansat: near Ms ‘ CODEN ” sahieanast yl be ' “ ” +e ‘f T aj ate Hey wee i \ PUP Sede Cu Nhe a es wane) ieee eC er} ck a ett niet) Py) threes i eulharae Oo ean i 5 Bes ya Saka rach ay + * Ci fran Ra RW - + 4, . ‘ fan Soe an s a We Ae 7 en < } eihy + aoe iat nitesas wpe wate: i Eta teine Saeed Sane ‘ tetera tedswennpres te dep, rte ; nae q itive es : : 7 ar Reabiache a ey : ‘ porubnene Be Nia wre Mr . ae hehe yi ‘ i ft ueits ate : ; CNN 7 aed aa SOOO ki : , Cat tentrr aes ‘ Ai ea eR , bi foe ; ‘lifes i ian aG " 1 ‘ . ba ged ae ty \ a rin tee’ seta wire ' wigs ated aes! Tre atortbeante ’ Met Persea wet IG ates ee as ry . 1 rg rhe . - P mk ' . a ‘ ; Ce hn ’ i : Pee eet a Pia eos Ce dT eet he ‘i aA eisyrittty Pa a ee " tlh : P A escalates : , . PPT HOC eo CR satca tit prc ata ad ret see es i eT erry te ery eS eo Naver bacasyse 68,9 U-alt4 Lara hie Site aati : 4 i Na ‘THE SCIENTIFIC PROCEEDINGS Or THE ROYAL DUBLIN SOCIETY. Defy Devries. VOLUME XII. Gy DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS & NORGATH, 14 HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909-1910, THE Soctrety desires it to be understood that it is not answerable for any opinion, representation of facts, or train of reasoning that may appear in this Volume of its Proceedings. The Authors of the several Memovrs ave alone responsible for their contents, Printed at Tur Untversity Press, Dublin, CONTENTS. VOL. XII. N Page a —The Origin of the Dexter-Kerry Breed of Cattle. By James Wuson, u.a., B.sc. (Plates L-IV.) (January 20,1909.) . 1 Il.—A New British and two New lvish Birds. By Ricnarp M. Barrineton, M.A. (February 2, 1909.) : . o alls) III.—Vitality and the Transmission of Water through the Stems of Plants. By Henry H. Dixon, sc.p., F.R.s. (March 15, 1909. ) Bil IV.—The Absorption of Water by Seeds. By W.R. Grnsron Atkins, g.A. (March 16, 1909.) : : : : 3) V.—Notes on the Pollination of Certain Species of Dendrobium. By A. FE. G. Kerr, up. (Plates V., VI.) (April 8, 1909.) 47 VI.—Production of Ammonia from Atmospheric Nitrogen. By Herman OC. Wourerecr, pu.p. (April 3, 1909.) : . 54 VII.—Note on the Tensile Strength of Water. By Henry H. Dixon, $0.D., F.R«s. (April 5, 1909.) : : : . 60 VIII.—The Colours of Highland Cattle. By James Witson, M.A., B.Sc. (Plate VII.) (May 15, 1909.) : : : OG 1X.—On a Proposed Analytical Machine. a Percy K. Lupearte. (April 28, 1909.) 0 bite a ae X.—The Taxine in Irish Yew. By Ricnarp i ‘Moss, F.LC., F.C.S. (April 30, 1909.) : . : : o | 6B XI.—On Photography by Reflection under Contact. By H. E. Fournier p’ALBE, B.SC., A.R.C.SC., M-R.I.A, (Plate VIII.) (May 10,1909.) 97 XII.—Mechanical Stress and Magnetisation of Iron. Part 1. By Witt1am Brown, 8.sc. (May 28, 1909.) . . . 101 XIII.—The Osmotic Pressures of the Blood and Eggs of Birds. By W. BR. Geusron Argins, 8.4 (May 28, 1909.) ; . 128 XIV.—Chrysophlyctis endobiotica, Schilb. (Potato-Wart or Black Scab), and other Chytridiacee. By T. Jounson, D.sc., F.L.S. (Plates IX.-XI.) (June 16,1909.) . : 131 XV.—The Scandinavian Origin of the Hornless Cattle of the British Isles. By James Winson, M.a., B.SC. (June 19, 1909.) . 145 iv Contents. No. XVI.—Further Observations on ee Potato-Scab, Spongospora subterranea (Wallr.). By ‘I. Jonnson, v.sc., v.u.s. (Plates XIL-XIV.) (July 24,1909.) . XVII.—Mechanical Stress and Magnetisation of Iron. Part 2. By Wittram Brown, B.sc. (June 21, 1909.) XVIII.—Methods of Determining the Amount of Light scattered from Rough Surfaces. By W. F. Barrert, r.r.s. (July 27, 1909.) XIX.—A New Form of Polarimeter for the Measurement of the Refrac- tive Index of cone Bodies. By W.F. Barrert, rr.s. (July 27, 1909.) ‘ ‘ XX.—On Montanin and Montana (Montan) Waxes. By Hueu Ryan, M.A., D.SC., F.R.U.1., and T'Homas Ditton, u.a. (July 17, 1909.) XXI.—The Analysis of Beeswax. By Huan Ryan, m.a., D.Sc., F.R.U.1. (July 17, 1909.) XXIJ.—On the Value of Benzidine for the Detection of Minute Traces of Blood. By EH. J. McWerenry, u.a., mp. (August 14, 1909.) XXIII.—On the Fossil Hare of the Ossiferous Fissures of Ightham, Kent, and on the Recent Hares of the Lepus variabilis Group. By Martin A. C. Hinton. (Plate XV.) (September 8, 1909.) XXIV.—Some Observations of Dew at Kimberley. By J. R. Surron, M.A., SG.D. (January 15, 1910.) ; XXV.—On Osmotic Pressure in Plants; and on a Thermo-Electric Method of Determining Freezing-Points. By Henry H. Dixon, sc.D., F.R.s.. and W. R. Grtston aa M.A. a a 1910.) . : XXVI.—Permanent Steel Magnets. By W. Brown, B.so. (bay 21, 1910.) . : XXVII.—Some Variations in the Sects of the Domestic Horse and their Significance. By Masor F. Hassm, p.s.o. (Plates XVI.—XX.) (March 8, 1910.) ‘ Z 5 4 XXVITI.—The Inheritance of Coat Colour in Horses. a JAMES eee M.A., B.SC. (April 12,1910.) . XXIX.—Chrome Steel Permanent oe ous W. Brown, B.sc. - (Api 21, 1910.) XXX.—On the Distribution of Mean Annual Rainfall and average number of Rain Days per year over an Area including the Counties of Dublin, Wicklow, Kildare, and Meath: A Study in Local Variation of Rainfall. By Wiu1am J. Lyons, B.a., a.R.c.sc. (Lonp.) (Plate XXI.: Map.) (May 19, 1910.) XXXI.—The Vapour-Pressures, Specific Volumes, Heats of Vaporisation, and Critical Constants of Thirty Pure Substances. By Sypnry Youne. p.sc., rn.s. (June 10, 1910.) . Page 225 266 275 312 321 331 349 354 374 Contents. No. XXXII.—A Simple Form of Open-Scale Isothermal Air Barometer. By W. F. Barrett, r.R.s. (May 19, 1910.) 6 : XXXIII.—Agricultural Seeds and their Weed Impurities: A Source of Ireland’s Alien Flora. By T. Jounson, p.sc., F.u.s., and Miss R. Hensman. (Plates XXII., XXIII.) (July 22, 1910.) XXXIV.—Cryoscopic Determination of the Osmotic Pressures of some Plant Organs. By W. R. Geusron Arxixs, m.a. (June 80, 1910.) . 6 : 6 ; peas Cate : XXX V.—The Separate Inheritance of Quantity and Quality in Cow’s Milk. By James Witson, m.a., B.sc. (July 29, 1910.) XXXVI.—Mechanical Stress and Magnetisation of Iron: Part 3. By W. Brown, B.sc. (September 2, 1910.) ; - XXXVII.—Mechanical Stress and Magnetisation of Nickel: Part I. By W. Brown, B.sc. (December 5, 1910.) : 0 InDEx, 5 446 500 519 THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 1. JANUARY, 1909. THE ORIGIN OF THE DEXTER-KERRY BREED (Ot CUMMINS. BY JAMES WILSON, M.A., B.Sc., PROFESSOR OF AGRICULTURE IN THE ROYAL 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 NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price One Shilling. antan nS THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. ———— I. THE ORIGIN OF THE DEXTER-KERRY BREED OF CATTLE. By JAMES WILSON, M.A., B.Sc., Professor of Agriculture in the Royal College of Science, Dublin. [Read November 24; Ordered for Publication DecemBeEr 8, 1908. Published January 20, 1909.] (Pirates I.-IV.) THE theory most widely accepted as to the origin of the Dexter breed of cattle is the one set forth in Professor David Low’s Domesticated Animals of the British Islands, published in 1845. Low writes that the breed ‘“ was formed by the late Mr. Dexter, agent to the late Maude Lord Hawarden. This gentleman is said to have produced his curious breed by selection from the best of the mountain cattle of the district. He communicated to it a remarkable roundness of form and shortness of legs. The steps, however, by which the improvement was effected have not been sufficiently recorded ; and some doubt may exist whether the original was the pure Kerry, or some other breed proper to the central parts of Ireland now unknown, or whether some foreign blood, as the Dutch, was not mixed with the native race.” The facts that there are other theories rivals to Low’s, that Low himself cites no authorities, that he is doubtful whether the method attributed to Dexter was the one he actually employed, and that, by referring to Dexter as “the late Mr. Dexter,” and indicating thereby that he believed the breed SCIENT. PROC, R.D.S., VOL. XII., NO. I. B 2 Scientific Proceedings, Royal Dublin Society. to have been of comparatively recent origin, suggest that his whole statement should be subjected to examination. Two of the rival theories have a strong resemblance to Low’s, in so far as they attribute the origin of the breed to other men called Dexter: one of them a steward to the Knight of Kerry on Valencia Island, the other a coastguard officer resident in Kerry. ‘he former of these theories may be dismissed at once, as the existence of Dexter the steward on Valencia would be very difficult to prove. As will be soen later, this theory arose because there was need for something more likely than the coastguard theory, which itself had arisen because it seemed more reasonable than the one recorded by Low. There is this much in the second theory that the individual concerned was not a myth. He was a retired naval officer who came to Kerry about 1832, and died there about 1858. But, according to his daughter, a lady now residing in Dublin, he never had anything to do with cattle. He used to tell, however, that it was his grandfather who brought the Dexter cattle into existence; and he had the idea that they were brought over from England. i The coastguard officer's grandfather, the original Mr. Dexter, was an Englishman who came over from Gloucester or Somerset about the middle of the eighteenth century to be agent on a nobleman’s estate in Tipperary. He had a son who became a lawyer in Tipperary; and one of this lawyer’s sons entered the navy, from which he retired about 1832, and entered the coastguard service, as stated above. Low identifies the nobleman with whom Dexter took service as Lord Hawarden, afterwards Lord de Montalt, whose property lay just north of Tipperary town; and in his Zour in Ireland, published in 1780, Arthur Young gives us two glimpses of Dexter himself :— “Mr. Dexter of Cullen had a ram .... and great number of ewes sent to him, the breed much improving” ;! and “‘'here have been many English bulls introduced for improving the cattle of the country at a considerable expence, and great exertions in the breed of sheep; some persons, Mr. Dexter chiefly, have brought English rams, which they let out at seventeen guineas a season, and also at 10s. 6d. a ewe, which indicates a spirited attention.’ Unfortunately these glimpses leave us in doubt about Dexter’s position as a breeder of cattle; but had he been noted as such, and especially as a breeder of animals so remarkable as the Dexter, it would hardly have escaped the notice of an observer and inquirer like Arthur Young. 1 Volume ii., page 158. * Tbid., ii., page 261. Witson —The Origin of the Dexter-Kerry Breed of Cattle. 3 We are, therefore, forced to look elsewhere for evidence of Mr. Dexter's connexion with the breed that bears his name; and it must be confessed that the evidence against the theory that Dexter originated the breed, as well as against the method which Low says he was believed to have employed, is very strong. The chief arguments against the Dexter theory of origin are these :— (i.) Dexter’s name is not mentioned in connexion with any breed of cattle by any writer previous to Low. Youatt’s Cattle was published in 1834; Wakefield’s Account of Ireland in. 1812; and the Statistical Surveys of the Irish counties were published in the first years of the nineteenth century; yet, though Kerry cattle and cattle of Dexter type are occasionally mentioned, there is no mention either of Mr. Dexter or of Dexter cattle. Unfortunately, though the Swrvey of Tipperary was begun, it was never finished. (ii) Cattle of Dexter type are recorded not only in Kerry but in other parts of Ireland too early and too distant from Tipperary to be traceable to Mr. Dexter. Youatt? records them in Wicklow in 1834, Sampson® records them in Derry in 1815, Wakefield’ records them in Cork in 1814, and Tighe, in his Survey of Kilkenny,* published in 1802, tells of Kerry cattle coming to Kilkenny, and makes special reference to a bull which is a Dexter rather than a true Kerry :—“ The Kerry cows are often driven into this country for sale; they are preferred in Dairies for their quantity of milk, as well as for the low price they bear even compared with their diminutive size. . . Their size often does not exceed a moderate sucking calf; when put on good grass, they fatten in a greater degree than the cattle of the country... . Their shapes are often very good; a bull of this breed, exhibited at Durrow in July last, was allowed to be nearly perfect in his points.’ A quotation from Arthur Young’ shows that, even in his day (1776), there were at least two types of cattle in Kerry, and suggests that one of them was the Dexter as we now know it :—‘‘ The poor people’s heifers sells at three years old at 30s.; their breed is the little mountain, or Kerry cow, which upon good land gives a good deal of milk. I have remarked, as I travelled through the country, much of the Alderney breed in some of them.” (iii) Wakefield’ not only speaks of two kinds of cattle on the borders of Kerry, but tells how cattle of Dexter type were produced: “In the 1 Cattle: their Breeds, Management, and Diseases, 1834, p. 185. 2 Survey of Londonderry, p. 205. 3 Account of Ireland, vol. i., p. 336. 4 Survey of Kilkenny, 1802, p. 309. 5 4 Tour in Ireland, 1776-79 (1780), vol. ii., 8 Account of Ireland, vol. 1., p. 336. p. 86. B 2 4 Scientific Proceedings, Royal Dublin Society. mountainous parts towards the south-west of this county (Cork), the Kerry breed of cattle is found; but by frequent crossing with the long-horned, they have produced a small breed which has nearly the same character.” It is still possible, of course, that Dexter may have bred or owned cattle of Dexter type; but, if he did so, it is very improbable that they found their way from Tipperary to Kerry, the source of the modern Dexters. In Dexter’s times the movement of Kerry cattle was in the opposite direction, and the tendency was to improve the old Irish cattle by crossing them with English breeds: the Longhorns in great numbers with some Herefords and Devons and a few of the breed we now call the Shorthorn. It must be remembered also that Dexter was an importer of English sheep—Leicesters—and that, if he became concerned in cattle, it was much more likely to have been in one of the English breeds. The arguments against the method attributed to Dexter are still more convineing. He “is said to have produced his curious breed by selection from the best of the mountain cattle of the district. He communicated to it a remarkable roundness of form and shortness of legs.”! The original Kerry breed was light in the body and long in the leg—Low’s statement is itself evidence of that—but to have brought a breed that was light-bodied and long-legged down to another that was stout-bodied and short-legged by simple selection within fifty or even a hundred years is a thing we now know, since the significance of Mendel’s discovery has been realized, to be highly improbable, if not altogether impossible. Before we can clear up the question we must look into the history and character of the Irish cattle themselves and of the cattle that were imported from time to time. The Kerries are all that are now left of the race that at one time inhabited the whole island, but which have been gradually pushed out by imported cattle. The dates of the importations cannot be fixed with accuracy. Among the skulls dug from the Dunshaughlin crannog, the date of which is fixed about the ninth century,? Sir William Wilde identified four different types—viz., a straight-horned, a curved-horned, a short-horned, and a hornless. It may be doubted whether Sir William was justified in separating these first two types from one another. It might be maintained that they were only such variations as might be acquired by individuals of a pure-bred race. In any case, the presence of the hornless type suggests that there was already a mixing of breeds in the neighbourhood of Dunshaughlin. If there were 1 Low’s Domesticated Animals of the British Isles, 1845. 2 Proceedings of the Royal Irish Academy, vol. vii., p. 58. Witson—The Origin of the Dexter-Kerry Breed of Cattle. bo) immigrants, and if they came from Wales, or England, or Scotland, the mixture consisted of branches of the same black Celtic race, since it is very unlikely that Anglo-Saxon cattle reached Wales or the north-west of England early enough to affect the cattle of the Dunshaughlin crannog.. But if Anglo-Saxon cattle were carried to Ireland in these early times, they were of the same race as the great importations of the seventeenth aud eighteenth centuries. If the immigrants were of Norse origin, as the hornless skulls would suggest, then they must have been few in number; for otherwise, since the hornless character in cattle is dominant over the horned, there would have been a far larger proportion of polled cattle in Ireland in later times. The importations that swept out the black race of cattle are of com- paratively recent date, and are to be connected with human immigrations from the other side of the Channel. ‘There have been three great immigra- tions: the first, a result of Strongbow’s invasion ; the second, the English “plantations ” of Elizabethan and subsequent times; and the third, the Jacobean “ plantation ” of Ulster. There are no records to show that the first and last of these immi- grations were followed by importations of cattle ; but it is usual for colonists to be accompanied or followed by their live stock, and there is evidence that this happened in the south of Ireland by reason of the large number of cattle there similar in colour and type to the red Saxon cattle of the south- west of England, the district from which most of Strongbow’s colonists migrated and from which they sailed. It is also probable that the Ulster colonists brought some cattle with them ; but these cattle could have effected very little change, since they were of the same race as the cattle already in Ireland. But there are records to show that the Elizabethan and subsequent plantations were followed by importations from England. In 1611, the Government made the following regulations! :—‘ For 2000 acres, and so rateably the undertaker for the first year may carry? 20 cows, 2 bulls, and 20 young store cattle; 100 ewes and 6 rams; 20 mares, horses, and colts; and as many swine as he will (not exceeding 10).” At the time of Arthur Young’s visit® (1776-79), the bulk of the cattle on the Ivish plain were not only descended from imported English Longhorns, but were Longhorns in type. The native cattle had been crossed so often that most of them had lost their original character, and the others were losing it. A similar change had taken place in the more fertile districts 1 Trish State Papers. 2 That is, may import. 3 See Young’s Zour in Ireland, op. cit. 6 Scientific Proceedings, Royal Dublin Society. branching north and south from the great plain, while, even in the less fertile parts of the country, the transition had well begun. For instance, Lord Massereene had imported Longhorns to Antrim as early as 1735.) Arthur Young reports having seen at the Archbishop of Armagh’s in the North and at Lord Doneraile’s in the South some of the breed we now call the Shorthorn which was soon to take the Longhorn’s place in driving out the native cattle, and eventually to drive out the Longhorn itself. He reports having also seen a few Herefords and Devons, and later writers speak of the increasing importation of Devons, Herefords, and Shorthorns, especially Shorthorns. The preference for the English breeds was so great that the old native race existed pure only in a few districts—indeed, its character had been forgotten in many—and, by the middle of the nineteenth century, it was practically extinct in all but its last strongholds—Kerry and Donegal. What were the characters of the native cattle on the one hand, and of the imported cattle on the other ? Unfortunately we have no statement as to the colour of the original Irish cattle that can be taken with absolute confidence. ‘The earliest state- ment is “in a letter written in 1580 and preserved in the Record Office among the Irish mss. Sir Nicholas White, Master of the Rolls in Iveland, says:—. .. ‘The native cattle were black®.’” But there are no further references to the subject till the beginning of the nineteenth century. The majority of the writers of that date say that the native cattle were black, but some. assert that there were other colours, namely, red, brindle, and also black with white markings on the body, on the head, and along the back. In connexion with this point, however, it must be remembered that the native cattle were pure only in a few places by the beginning of the nineteenth century; that many cattle bearing these other colours and markings had already been brought into the country; and that few of the writers referred to were able to discriminate between the pure native cattle and others that had been crossed for generations. There is, however, other evidence to show that the native Irish cattle were black. The proof is not absolute; but it is such that any other theory is scarcely possible. Let us summarize it :— (a) The native Irish cattle were of the same race as the native Celtic cattle of great Britain. 1 Dubourdieu's Survey of Antrim, 1812. * Houseman and Sinclair's History of the Devon Breed of Cattle, p. 21. Wirtson— The Origin of the Dexter- Kerry Breed of Cattle. i (b) In Ireland and in those parts of Britain occupied till two centuries ago by Celtic cattle, all cattle were referred to as “black cattle” down almost to recent times. This was to distinguish them from horses. (ec) The Celtic cattle in Britain, that is in Wales, the north of England, and Scotland were described by early (sixteenth- and seventeenth- century) writers as black. (d) In parts where the Celtic cattle had been least crossed, they are almost invariably black, e.g., in the north-east of Scotland, Galloway, Wales, and Kerry. (ce) There is very little difficulty in breeding all our black breeds true to colour: the Kerries themselves, for instance. (7) Had the Ivish cattle or any other of the Celtic breeds deviated from black, such descriptions as ‘“ white-faced,” ‘‘ pyed,” “ flecked,” or “‘brindled ” would have been applied to them. g) The appearance of other marks and colours than black can be pp accounted for through the presence of intruding breeds. The imported cattle were of various colours. The Devons were whole- coloured reds, but the three other breeds were combinations of two or more colours. They came from that part of England which was a jumble of breeds and races. Till Roman times the cattle in England were all of the black Celtic race. White cattle were introduced by the Romans. These two races were driven into Wales and the north of England by Anglo-Saxon red cattle. Several centuries ago many red and white cattle, closely akin to the first imported Anglo-Saxon red cattle, were brought over from the Low Countries to the eastern and midland counties, splitting, as it were, the Romano-Celtic cattle in the north from the red Anglo- Saxon cattle in the. south, and uniting with them on both sides. These imported cattle were generally red and white-flecked, like red and white Shorthorns; but some were white-faced, like the Herefords, while others had a white stripe along the back and another along the under-line. The relatives of them all may be seen in Holland and Germany at the present day. Out of this jumble emerged the three large breeds that were imported to Ireland, the Longhorns, the Shorthorns, and the Herefords. The Longhorns were some kind of red roan or brindle, with white streaks above and below, and sometimes with white faces; the Shorthorns were red or red-and-white, 8 Scientific Proceedings, Royal Dublin Society. roan, and white, while the Herefords, which at one time were all red, had begun to acquire their present white markings ere they came to Ireland. These being descriptions of the imported cattle, let us see the kinds of cattle that were then produced when they were crossed with the natives. But, before doing so, let us consider the colours the importations ought to have produced when crossed with black cattle. ‘Two of the four imported breeds are still used for crossing with black cattle in Ireland and Britain, and the colours they produce are as follows :— (i.) Shorthorns.—(a) First crosses with black cattle and red or red-and- white Shorthorns are black. + (>) First crosses with roan Shorthorns are either black or blue-roan. (c) First crosses with white Shorthorns are blue-roan. We need only consider first crosses, because, if these are bred together or with either parent-race, no new colours are likely to emerge, although colours and markings of either race may be transferred to their descendants. (ii.) Herefords.—First crosses with the Hereford are black with white faces. ‘This is a peculiar phenomenon: the black of the black cattle is dominant over the red of the Hereford, while the white face of the Hereford is dominant over the black face of the black cattle. (iii.) Devons.—Devons are not often crossed with black cattle; but as they are the same race as other southern English red breeds, their first crosses with black cattle ought to be black. The few crosses the writer has seen are black. (iv.) Longhorns.—Longhorns are now very few in number, and are seldom, if at all, crossed with black cattle. If they were, their progeny can be indicated by other crosses. Some of the Longhorns were perhaps roans, most were reds and brindles, and nearly all had the white streak down the back. (a) If the Longhorn roans were a similar mixture to the Shorthorn roans, then their first crosses with black cattle would be either black or blue roans. (6) If the roans of the Longhorns were produced by crossing black and red races—it is not asserted that they were—then the first crosses would, in all probability, have been black. Some may have been red. (ec) Among Highland cattle, brindles are crossed with black, and their progeny are blacks, reds, and brindles. ‘The brindle of the Longhorns had a similar origin to that of the Highlanders; and when Longhorns were crossed with black cattle, their first crosses would probably have been blacks, reds, and brindles. Witson—The Origin of the Dexter-Kerry Breed of Cattle. 9) (d) The white streak along the back is now almost extinct, excepting among the Longhorns. At one time it was very common in the western English midlands and in the south-east of Wales. It is still common in some parts of Germany. It was, and still is, a well-defined characteristic ; and from its having been so widespread, and having persisted so long, it may be taken as a Mendelian character, like the white face of the Hereford. There- fore when Longhorns were crossed with black cattle, this black stripe would have come out in the first crosses." Let us now quote some descriptions of the cattle in those parts where the natives were being crossed by imported cattle, but were not yet entirely swept out. 7 Sampson,? in his Survey of Londonderry (1802), writes :—“I observe two varieties of native cows; the one is light in the bone, small in size, extremely active, crooked in the ham, with a good eye and sharp nose, and nice thin neck, a crooked horn, frequently turned upward. ‘This strain is generally black, reddish, or brindled, with some white. There is a coarse-boned, ill- shaped breed also; these have swollen bellies, heavy head, a dew-lap very pendant, a bull-lke aspect— ‘Cui turpe caput, cui plurima cervix, Et crurum tenus a mento palearia pendent, Hit faciem tauro propior.’ Tighe,‘ in his Survey of Kilkenny (1802), writes :—“ The common cattle of this country are a mixture of the Irish breed with some of the longhorned English. A few may be seen of the ancient native stock, or what may be supposed so, whose characters appear to be upright horns, distant, dry, bent somewhat backwards, and tipped with black ; ears rather large; body black, and face white.” Youatt® (1834) writes that “Ivish cattle are evidently composed of two breeds; the middle and the longhorns. The former is plainly an aboriginal breed. ‘They are found on the mountains and rude parts of the country, in almost every district. They are small, light, active, and wild. The head is small, although there are exceptions to this in various parts; and so numerous, indeed, are those exceptions, that some describe the native Irish cattle as having thick heads and necks; the horns are short compared with 1 Since this paper was read the writer has found evidence in the Longhorn Herd-Book that the white back-stripe is a Mendelian character, which is dominant over black at least. 2 Survey of Londonderry, p. 205. 3 This is quoted from Virgil’s Georgics, Book iii., lines 52 and 58, and part of line 68, 4 Survey of Kilkenny, p. 800. 5 Cattle, their Breeds, &c., p. 179. SCIENT. PROC. R.D.S., VOL. XII,, NO. I, 10 Scientific Proceedings, Royal Dublin Society. the other breed, all of them fine, some of them rather upright, and frequently, after projecting forward, then turning backward. Although somewhat deficient in the hind-quarters, they are high-boned, and wide over the hips, yet the bone generally is not heavy. ‘The hair is coarse and long; in some places they are black, in others brindled, with white faces. Some are finer in the bone, and finer in the neck, with a good eye, and sharp muzzle and great activity. ... E “Mr. Anderson! of Shelton, in a letter with which we have been favoured from him at the request of the Harl of Wicklow, describes the old Irish cattle there, as a low, broad, hardy breed, with thick heads and necks, and a thick hide.” Wakefield’ (1812) writes :—“ The native Irish stock were, in my opinion, all black, for, though at present there are very few of that colour, they are universally called ‘black cattle.’ I have seen some which were pointed out to me as the remains of the ancient breed; they were narrow in the loins and thin in the quarters; they had short legs, large bellies, and white faces ; their horns, which turned backwards, were remarkably wide set, and they had large dewlaps; but this breed is now almost extinct.” Low® (1845) describes the Kerry cattle as follows:—“ The Kerry cattle of the mountains are generally black, with a white ridge along the spine, a character agreeing with the account which older writers have given of the Uri of the woods of Poland. They have often also a white streak upon the belly, but they are of various colours, as black, brown, and mixed black and white, or black and brown. ‘Their skins are soft and unctuous, and of a fine orange tone, which is visible about the eyes, the ears, and the muzzle. Their eyes are lively and bright, and, although their size is diminutive, their shape is good.’”* From these quotations it is clear that, where the natives had been crossed by imported cattle, but where the imported type had not yet overwhelmed the native, there are two kinds of cattle, one of which, and sometimes the other, was taken to be the native kind. But when we remember that the imported breeds—the Longhorns, the Herefords, and the Shorthorns, at any rate—were much heavier in the body than the native, it is clear that the stouter cattle, that were sometimes mistaken for the old native race, were the results of crossing the native cattle with these heavier breeds. The colours of both kinds were derived partly from the old native race, partly from the imported breeds. 1 Youatt’s Cattle, p. 185. 2 Account of Ireland, vol. i., page 334. 3 Domesticated Animals, etc., p. 309. A 4 Low is here describing the small Irish cattle. To him they. were all Kerries. Witson—The Origin of the Dexter-Kerry Breed of Cattle. 11 But these stout animals, although they were of Dexter type, were not all Dexters, the only colours among which are whole red and black. Nor is it possible, apart from the fact that they were too large, that the white-faced Hereford, the finch-backed and brindled Longhorn, or the variously coloured Shorthorn could have produced the comstiainily whole-coloured black and red Dexters of the present day. Mendelian researches have shown that, when breeds are crossed and produce what are called intermediate hybrids—that is, hybrids in which the characters of the parent races mix or blend—these hybrids, when bred together, have on the average 50 per cent. of their progeny hybrids like themselves, 25 per cent. like one ancestral race, and 25 per cent. like the other. For instance, when roan Shorthorns are bred together, 50 per cent. of their progeny are roans, 23 per cent. are red, and 25 per cent. are white. But when hybrids in which the characters of one parent race dominate or obscure those of the other are bred together, their progeny split up into some like one parent race, some like the other, and some like one race in some of its characters and like the other race in others. The converse also holds that, when the progeny of a set of animals split up in this way, these animals are hybrids. When Dexters are bred together, their progeny split up into four distinct types, Viz. :— 1. Black-coloured, stout, short-legged’ animals. 2. Red-coloured, stout, short-legged animals. 3. Black-coloured, slender, long-iegged animals. 4, Red-coloured, slender, long-legged animals. In a breed whose records are so scanty, the exact numbers of each type cannot be clearly ascertained ; but it can be said that the first type is by far the most numerous, the second and third types are about equal; and the last is decidedly least numerous. According to Mendelian researches the appearance of these four types suggests that, while the slender black Kerry was one of the races which produced the Dexter, the other was a stout-bodied, short-legged animal, whose colour was red. And that such an animal was imported into Kerry there is sufficient evidence. Wakefield (1812) states :—‘ In the south I met with some persons who had imported Devonshire cattle; Lords Bantry, Shannon, and Doneraile, 1 When the calyes are very young, those that are going to be stout-bodied and short-legged can be told by the length of the leg. See Plate LV. 02 12 Scientific Proceedings, Royal Dublin Society. Mr. Hyde, and others, possess considerable numbers of them. Jord Farnham, in Cavan, has a herd of them, and from what I have seen of this stock in the north of Devonshire, where they are natives of Hxmore, I am inclined to think that they are the best cattle known, and had I anything to do with mountain estates in the south of Ireland, I should strongly recommend them for general use. . . . As I have never seen them to the northward, I should be afraid to introduce them into that part of the kingdom.” Nor is it improbable that these were not the first cattle of the Devonshire breed imported into Ireland. In the seventeenth and eighteenth centuries, many English immigrants came to Kerry and West Cork, and there is strong evidence that they brought over red cattle with them. The immigrants sailed chiefly from Bristol channel ports, and, unless they had driven them enormous distances, could have brought none other than red cattle. The following statement, by the Duke of St. Albans, a part of which has been quoted already, is to the point :—‘ In a letter written in 1580, and preserved in the Record Office among the Irish MSS., Sir Nicholas White, Master of the Rolls in Ireland, says that Dingle harbour, in Kerry, was known as ‘Coon edaf dearg’ which, in Irish, means ‘ red ox haven.’ White says this name was owing to the first settlers who came from Cornwall and brought cattle with them. The native cattle were black.”’ Nor must it be forgotten that Strongbow’s men came from the south- west of England; that they probably imported cattle from that part—the cattle there were red—and that the red race may have found its way west- wards towards Kerry. The probability, therefore, that Dexter cattle are descended from black Kerries and red cattle of Devon type is very high; and if further proof were wanted, it can be found by setting a red Dexter cow side by side with a zed Devon. The only difference between them is that the Devon cow is now slightly larger: a matter that can be accounted for by the Devon having been much better cared for and increased in size during the last hundred years. Wakefield looked upon the Devon as a mountain—and therefore a small—breed; while one Irish writer, Rawson, in his Survey of Kildare, 1807, writes that “the Devons are nothing better than what the mountains of Ireland can produce with any little care.” One of the problems of breeding Dexter cattle with success may be solved by knowing that the Dexter is a hybrid, and that the Kerry is one of the parent races. ‘hat there is a serious problem in the matter was not realized till after the establishment of the Kerry and Dexter Herd Book, by 1 History of the Devon Breed of Cattle, p. 21. Wirson— The Origin of the Dexter-Kerry Breed of Cutlle. 13 the Royal Dublin Society, in 1890, and the separation of the Dexters and the Kerries into two breeds. ‘The effect was, that Dexters could no longer be bred as they had been almost universally bred in Kerry—namely, by crossing the Dexter with the Kerry—and, at the same time, remain eligible for entry in the Herd Book. Dexters must now be bred with Dexters only —a procedure well known in Kerry to be disastrous, since a considerable - number of the calves produced in this way were sure to be so misshapen that they were either dead-born or had to be destroyed. What was formerly known to Kerry men now became known to other breeders who bred Dexters according to the rules of the Herd Book: : with the result that a breed which is not only beautiful and picturesque but exceed- ingly useful is in a languishing condition. This unfortunate state of affairs may be remedied by returning to the ante Herd Book position. The Kerry and the Dexter are one breed, just as much as the red Shorthorn and the roan are one breed. ‘The roan Shorthorn is a hybrid between a red race and a white race; but the hybrid roan and its two parent races are recognized as one breed. The Dexter is a hybrid between a black race and a red, and the hybrid and its black parent race should be recognized as one breed, or, at any rate, as two divisions of the same breed. By doing so, Dexters could be bred with no more than ordinary cattle- breeding risks, and the present handicap against the breed would be removed. The problem might be better understood if it were put concretely and in a Mendelian setting. When red cattle are bred with white, their young are roans. When these roans are bred together, only fifty per cent of their progeny are roans, one-half the remainder being like their white grand- parents and the other half like the red. Mendel formulated a theory to explain such phenomena. He conceived the idea that plants and animals must carry from their very beginning a lot of intangible somethings which determine their future characters: one for colour, one for size, another for the shape of one part, another for the shape of another, and so on. He further conceived the idea that these determinants must be in two halves, and that one-half must come from each parent. Let us see first how this theory works in the case of the red, white, and roan cattle. The red cattle carry red determinants, the white white. Let us represent the colour determinant of a red animal by two black circles § , and the colour determinant of a white animal by two white ones, 9. When a white and a red animal are mated, one-half of a colour determinant from each parent goes to form half of the colour determinant of the progeny, which is therefore a combination of red and white, thus ®. 14 Scientifie Proceedings, Royal Dublin Society. But breed now with the half-bred progeny, and what will happen? It is a question of chances. The following diagram will show that there is one chance of the two red half-determinants uniting, one of the two white, and two of red uniting with white. ASH oie Com Thus will give o@°0o of —-0 That is to say, if a sufficient number of roan short-horns are mated together, their progeny will work out in the proportion of one red to two roans to one white. Another diagram will show that, when roans are mated with whites, the progeny will be one half roan and one half white. OSC e@eo0o Thus gives : eo ©. & © of— 0 Still another will show that when roans are mated with reds, the progeny must be one half red and the other half roan. ec Se Thus »< gives mine oe! o<——>e cant But all colour determinants do not work so simply. Black and red do not blend like the red and the white of the Shorthorns. When black cattle are crossed with red, the progeny are not a mixture of red and black: they are black. The black colour is dominant and gets its way. And when these black hybrids are bred together, their progeny do not come out like the progeny of roan Shortkorns, but in the proportion of three black to one red. Only one, however, of the three blacks is pure: the other two are hybrids, like their parents, in which the black colour dominates the red. Let us put it diagrammatically as before, but using letters instead of circles, with black, the dominant colour, represented by the capital letter B, and red, the subdued? colour, by the small letter 7, A pure black animal is represented by 3 and a pure red by A while a hybrid between them is represented by eB If these hybrids are bred together, their progeny come out thus :— YAS i aS JEM ra Plea biate 1 Mendel used the word ‘ recessive.” Witson—The Origin of the Dexter-Kerry Breed of Outtle. 15 That is, one pure black animal, two that look black, but are really hybrids, and one pure red animal. This is what happens with Dexters in the matter of colour. But they vary in respect of form also, the essential difference being in the length of the leg! A short-legged race has been bred with a tall race; and the short- ness of the one is dominant over the tallness of the other. Let S represent shortness of leg and ¢ tallness. When hybrids are bred together, they come out thus :— Vat ives SoS 8 Way 8 S 8-8 % That is to say, there are one pure short-legged, two that look short: but are hybrids, and one pure tall animal. But it must be remembered that all these animals are involved in the colour variation also. How are we to represent the two-fold variation of form and colour? ‘Take one set of animals, say the short-legged ones. Among every four there must be one pure black one, two hybrid black-reds, and one pure red one. Then these four short-legged animals may be represented as regards both form and colour thus :— SB SB SB iS) S) Si SB Sr The other three sets of animals, : a and a may be similarly represented. If we display all these cases symmetrically, we shall see more clearly the results of breeding Dexter with Dexter. SB SB SB SB SB Sr tB tr Sr . SP Sr Sr SL Sr tB tr tB tB tB tB SB Sr eB t) tr ty th ‘te = SB Sr tB alin 1 Stoutness of body is concomitant with shortness of leg. 16 Scientific Proceedings, Royal Dublin Society. Thus, among every sixteen calves, the progeny of Dexter with Dexter, there ought to be 12 short-legged animals, 9 of them black, and 3 of them red ; 4 tall animals, 3 of them black, and 1 of them red. Of the sixteen animals only four—those underlined—are absolutely pure for both form and colour. Two of these four are new pure types— those underlined twice. ‘The others are all hybrids in respect of either form or colour, or both. An inspection of the diagram will show. That is to say, 16 Dexter cows mated with Dexter bulls ought to produce 9 black Dexters, 3 red Dexters, 3 black Kerries, and J red Kerry. But they do not do so. A number of the Dexters are misshapen and useless.1 The proportion cannot be stated with accuracy. According to some breeders it lies somewhere between 25 and 50 per cent. Assume it to be 33 per cent. ‘Then four of the twelve Dexters referred to above are useless. Thus, by mating his 16 Dexter cows with Dexter bulls the | breeder gets only 4 Kerry calves and 8 Dexter calves—the other 4 being useless—whereas, by mating them with Kerry bulls he would get 8 Kerry calves and 8 Dexters. ‘The same result would be got by mating Kerry cows with Dexter bulls: and in both cases the calves produced would be black. This can also be shown diagramatically. The hybrids or Dexters are represented, so far as size is concerned, by e and the Kerries by A When these two kinds are bred together, their progeny are one half short, one half tall. Thus :— He ives Se eG a, 8 ies (hee Neat Se And since the Kerries are all black, and black is dominant over red, the progeny of pure Kerries and Dexters must all be black. Of course it may still be possible to eliminate the hybrids that have to do with the production of misshapen calves, as it is also possible to find out which of the Dexters are pure for both colour and shape, and then breed pure Dexters from them. To do these things, however, experiment, means, and patience are required. As these misshapen calves may be interesting to students of Mendelism, let us indicate where they may be looked for. A breeder of experience tells us they are always black. The three pure black animals among the Dexters may therefore be suspected. 1 For a description of these calves, see Dr. C. G. Seligmann’s paper on ‘‘ Cretinism in Calves,’’ in the Journal of Pathology and Bacteriology tor March, 1904. Witson—The Origin of the Kerry-Dexter Breed of Cattle. 17 After all, what is the probable origin of the word ‘dexter’? An author, “‘ while travelling in Kerry some years ago, found that the word ‘ dexter’ was used in a generic sense with reference to all diminutive animals, even men, if low-set and bandy-legged; and also that the term was in the first instance applied to short-legged sheep kept by a resident coastguard officer.’”! Murray’s Dictionary gives no indication of the word being used in the above sense. But the original Mr. Dexter was a breeder of Leicester sheep which, in comparison with the native Irish sheep, were short-legged and stout. It is said sheep bred by Mr. Dexter were called “ Dexters.”” Did not the word pass on from Mr. Dexter’s sheep to other stout animals, and even to men? The Secretary of the Royal Agricultural Society of England has very kindly arranged to allow the use of the blocks of the Devon bull and cow in illustrating this paper. 1 Wallace’s Farm Live Stock of Great Britain, 1907, page 205. SCIENT. PROG. R.D.S., VOL. XII., NO, I. D a a fk Bi as i » OF qu patent ie ve SCIENT. PROC. R. DUBL. SOC., N.S., Vor. XII. PLATE I. Fig. 1.—A Kerry Bull. Fig. 2,—A Kerry Cow. ot . eer — ike bye ee ae i ory PP ase foaeMare | Smihiagn SCIENT. PROC. R. DUBL. SOC., N.S., Von. XII. PLATE II. Fig. 2.A Deyon Cow. [These figures are from photographs by Mr. Charles Reid.] ners ano niet by i Po Usstintisyee hast SCIENT. PROC. R. DUBL. SOC., N.S., Von. XII. PLATE III. it a nN SCIENT. PROC. R. DUBL. SOC., N.S., Von. XII. PLATE IV. viii Fig. 2,—A Kerry Calf from Dexter parents. oC 7 SET At, SE 2 Sue AND, a, Pameh etek seit eiel THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII (N.S.), No. 2. FEBRUARY, 1909. A NEW BRITISH AND TWO NEW IRISH BIRDS. BY RICHARD M. BARRINGTON, M.A. [Authors alone are responsib/e for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY ? LEINSTER HOUSE, DUBLIN, WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price Sixpence. Witson—The Origin of the Kerry-Dexter Breed of Cattle. 17 After all, what is the probable origin of the word ‘dexter’? An author, ‘‘ while travelling in Kerry some years ago, found that the word ‘ dexter’ was used in a generic sense with reference to all diminutive animals, even men, if low-set and bandy-legged; and also that the term was in the first instance applied to short-legged sheep kept by a resident coastguard officer.’” Murray’s Dictionary gives no indication of the word being used in the above sense. But the original Mr. Dexter was a breeder of Leicester sheep which, in comparison with the native Irish sheep, were short-legged and_ stout. It is said sheep bred by Mr. Dexter were called “ Dexters.” Did not the word pass on from Mr. Dexter’s sheep to other stout animals, and even to men ? The Secretary of the Royal Agricultural Society of England has very kindly arranged to allow the use of the blocks of the Devon bull and cow in illustrating this paper. 1 Wallace’s Farm Live Stock of Great Britain, 1907, page 205. SCIENT. PROC. R.D.S., VOL. XII., NO, J. D ie A NEW BRITISH AND TWO NEW IRISH BIRDS. By RICHARD M. BARRINGTON, M.A. {Read Novemprr 24.° Ordered for Publication Decrmper 8, 1908. Published Fenruary 2, 1909.] PALLAS’S GRASSHOPPER WARBLER (Locustedlu certhiola, Pall). Tur above species is perhaps the most interesting yet obtained at our Irish Light-stations. It was picked up dead at Rockabill Light-house, on the 28th September last, by Martin Kennedy, assistant Light-keeper. Tf we except the specimen obtained at Heligoland Light-house by Heinrich Gatke, on August 13th, 1856, there is no other European record. Gitke says! that Rudolf Blasius, the eminent German ornithologist, considered it “the jewel of his collection.” It is a native of Eastern Asia, passing through China on migration, wintering in Burma, India, and some islands of the Malay Archipelago. It resembles our own Grasshopper Warbler (Locustella nevia), but is much larger, and the feathers on the back are striped with black. The greyish-white tips to the under-tail feathers first attracted attention, and enabled me to identify it at once by the figure in Dresser’s splendid work on the birds of Europe.? It was in plump condition, and on dissection proved to be a male. ‘This Asiatic species is one of the most remarkable of recent additions to the British avifauna. THE LITTLE BUNTING (Emberisa pusilla, Pall). On October 2nd, 1908, a specimen of the Little Bunting was killed, striking the Rockabill Light-house lantern, and was forwarded by the principal keeper, Mr. B. R. Jeffers, as a Twite (Linota flavirostris). It was like a small specimen of our Common Reed-Bunting except for the absence of the ferruginous colour on the small wing-coverts; but on closer examina- tion it proved to be the above species, and is the first specimen obtained in Ireland. 1“ Heligoland as an ornithological observatory,’’ by Heinrich Gitke, translated by Rudolph Rosenstock. Edinburgh, 1895, p. 126; also ‘‘ Die Vogelwarte Helgoland, von Heinrich Gatke, herausgegeben von Professor Rudolf Blasius. Braunschweig, 1900,” p. 331. 2“ History of the Birds of Europe, including all the species inhabiting the Western palearctic region.’’ London, 1871-96. Barrineron—A new British and two new Irish Birds. 19 It is said to be confiding in its habits, and frequently found in the company of the Twites, and not readily detected. At any rate, up to 1898, only one specimen was recorded from England. Since then there have been several occurrences, chiefly at Fair Island, between the Orkney and Shetland Isles. Its breeding range extends through Russia and Siberia, eastwards from Archangel. When the bird becomes better known, it will probably be recognized as a straggler in small numbers, every autumn, to some portion of the British Isles. On dissection the Irish specimen proved to be a female. THE REED-WARBLER (Acrocephalus streperus, Vieillot). A specimen of this Warbler was recorded! as having been shot at Raheny, near Dublin, December 21st, 1843, by Mr. Montgomery. As the date is an unusual one for a summer migrant, and as no specimen was found in the Montgomery collection, Ivish naturalists have for many years excluded it from our avifauna. It is a regular visitor to England, and fairly common in Wales; and having regard to the marshy character of many parts of Ireland, it is remarkable that no specimen has ever been obtained in this country until the present year. The study of migration, at Light-stations, has added several rare insect-eating birds to the Irish list ; and yet this common English Warbler never occurred before. The specimen now exhibited was killed striking Rockabill Light-house on October 20th, with three Golden-crested Wrens and two Redstarts, and is a female. Tt so closely resembles the Marsh- Warbler, that it is said to be erroneously coloured in Gould’s “‘ Birds of Great Britain”’; also in Lord Lilford’s *‘ Coloured Figures of the BritishB irds,’”’ and to a certain extent in Dresser’s work on the birds of Europe. It differs in its habits and song from the Marsh- Warbler. Mr. A. H. Evans writes that he heard this species singing in a reed- bed on the Shannon, near Portumna, on the 23rd July, 1904. ‘The total number of species new to Ireland forwarded by the Light-keepers is now thirteen, namely :— Lesser Whitethroat (Sylvia curruca), 1890. Yellow-Browed Warbler (Phylloscopus superciliosus), 1890. Melodious Warbler (Hypolais polyglotta), 1905. 1 Zoologist, 1848, p. 2143; also Proce. Dublin Nat. Hist. Soc., 1852, p. 89. 2 “Tilustrated Manual of British Birds by Howard Saunders.’’? London, 1899, p. 81. pd 20 Scientifie Proceedings, Royal Dublin Society. Reed- Warbler (Acrocephalus streperus), 1908. Aquatic Warbler (A. aguaticus), 1903. Pallas’s Grasshopper- Warbler (Locustelia certhiola), 1908. Woodchat Shrike (Lanius pomeranus), 1893). Red-breasted Flycatcher (Muscicapa parva), 1887. American Snowbird (Junco hiemalis), 1905. Greenland Redpole (Linota rostrata), 1889. Little Bunting (Zmberiza pusilla), 1908. Lapland Bunting (Calcarius lapponicus), 1887. Short-toed Lark (Alauda brachydactyla), 1890. ‘The occurrence of the Antarctic Sheathbill (Chionis alba) for the first and only time in Europe, near Carlingford Lough Light-Station, in 1892, may be mentioned as a very remarkable incident not yet satisfactorily explained. SCIENTIFIC PROCEEDINGS ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 3. MARCH, 1909. VITALITY AND THE TRANSMISSION OF WATER THROUGH THE STEMS OF PLANTS. BY JaUGUNUE Ye JEL, IDIONOIN, ScJDs, I oltios, PROFESSOR OF BOTANY IN THK UNIVERSITY OF DUBLIN. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIEYY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETLTA STREET, COVENT GARDEN, LONDON, W.C. WX \SOll!s 1909. [co Price Sixpence. elaine ean ay’ Sr ia Atatan Ub y wy 3 Sen Dezl a 100, VITALITY AND THE TRANSMISSION OF WATER THROUGH THE STEMS OF PLANTS. By HENRY H. DIXON, Sc.D., F.B.S., Professor of Botany in the University of Dublin. [Read Novemperr 24. Ordered for Publication Decrmwsrr 8, 1908. Published Marcu 15, 1909.] Wakn the transpiring cells of the leaves abstract water from the trachese a tensile stress is set up in the water remaining in the trachesw, inasmuch as the latter are normally completely filled with water. This stress is resisted by the walls of the tracheze, to which the water adheres, and is transmitted downwards through the water of the transpiration current permeating and filling a greater or less number of the tracheze composing the wood of the plant. Exception has been taken’ to this view of the mechanism of the ascent of sap on the ground that the resistance of the wood is so great—when the length of stem traversed by the water is considerable—that we cannot imagine either the transpiratory forces to be equal to the task of dragging the water-column upwards, or the column to have the tensile strength to sustain the stress. I have elsewhere shown’ that these objections are based upon an over- estimate of the amount of water transpired, and consequently of the velocity of the transpiration-current, and also on determinations of resistance which are excessive. In order to account for the manner in which the transpiration current is relieved of this supposed resistance, Ewart? has endeavoured to show that the cells in the neighbourhood of the capillaries of the wood exert a lifting force on the ascending water-column. He has invented several mechanical schemes to illustrate how this may be effected, none of which, 1 Ewart, Roy. Soc. Phil. Trans., B., 1905, and Zdid., 1908. 2 Roy. Soc. Proc., B, 1907, p. 42, et seq. 3 Roy. Soc. Phil. Trans., B, 1905, and Ann. of Botany, 1907, p. 443. Roy. Soc. Phil. 'I'rans., B, 1908. SCIENT. PROC., R.D.S., VOL. XII., NO. III. E 22 Scientific Proceedings, Royal Dublin Society. however, are likely to commend themselves, and indeed are scarcely seriously put forward by their author. Ewart believes these lifting forces to be feeble, and has not, as he himself admits, in any case obtained unequivocal evidence for their existence. And yet, according to his own figures, these forces should be easily demonstrable.! According to him, the pressure required to raise water at the transpiration- velocity in an elm tree 12 metres high would be equivalent to a head of 75°6 metres, i.e., about 7°5 atmospheres. Of this he admits’ about 2 atmo- spheres might be supplied by the tension set up by the transpiring leaf-cells, leaving about 5:5 atmospheres to be made good by the lifting forces of the cells in the 12 metres of stem. Therefore the lifting force of the cells of this stem must amount to 0°45 atmosphere per metre of stem, or to a head of water equal to 44 times the length of stem. A lifting force of this magnitude should be easily revealed if the velocity of flow through a branch in the normal direction for a given head were compared with the flow in the reverse direction, or, again, if the amount transmitted downwards in a living stem were compared with that transmitted after death. Asis well known, experiments have not been able to demonstrate a sensible difference in either case. Furthermore, Ewart,’ working very carefully by a different method, has failed to detect the existence of these pumping actions in stems. Consequently it is quite impossible to admit that any large amount of work falls on the cells of the stem in tlie raising of the sap. While these considerations show that forces of any great magnitude are not exerted by the cells in the wood on the transpiration-current, it seemed desirable by some more careful method to test the ‘matter, and see if some much smaller force were not assisting the upward flow of water. In the ordinary methods of testing this question, uncertainties arise from the fact that conditions are not the same before and after the reversal of the current, or before and after the death of the branches. These differences are principally due to changes in temperature, which, as Ewart! has pointed out, entail large differences in viscosity, and to clogging in the experimental stem. In order to eliminate these sources of error, and so be in a position to detect the effect of even a very small force exerted by the stem-cells in lifting water, I carried out some experiments in the following manner :— ‘'wo straight branches (4 and B, fig. 1), about 80 cms. long, without 1 Roy. Soc. Phil. Trans., B, 1904, p. 56. ~ Roy. Soc. Phil. Trans., B, 1908, p. 379. 3 Roy. Soc. Phil. Trans., B, 1905, p. 77. * Roy. Soc. Phil. Trans., B, 1905. Dixon— Vitality and Transmission of Water through Plants. 23 lateral shoots, and as similar to one another as possible,’ are passed through tubulures (@ and 0b) in the bottom of a metal cistern, about 65 cms. deep. The upper ends of the branches projected above, and the lower ends below, the cistern. ‘The joints round the lower ends were rendered water-tight by binding on a rubber tube overlapping the tubulures and the projecting ends of the branches. The cistern may now be filled with water which, if kept in motion, will secure that both branches are almost at the same temperature, and so differ- ences in viscosity in the water passing through the branches will not arise. In order to avoid irregularities in trans- ¥ ~~~ mission, much eare is needed in the prepara- tion of the branches. After selection of the | branches, the upper leafy part is cut away, ! and the cut surface of the lower part still | attached to the tree is moistened by a jet of | water. ‘I'his part is aow cut off under water, | and while still submerged is removed to the | laboratory. Fresh surfaces are now prepared | at each end, and smoothed off by a razor, | under a stream of distilled water. A wide | glass tube about 20 ems. long is attached to the | upper end of each branch. ‘his is kept full | of distilled water, which acts as the supply and | head, driving the water downwards through | the branches. If it is desired to apply picric | acid or some other poison as a killing agent, | the simple glass tube is replaced in each case | by one which is provided with a side tubulure, & bent in a J form, and with two stop-cocks placed as shown in fig. 1. The rate of transmission from above duwn- wards is first observed lor the two branches by weighing the amount of water transmitted in a given time (say, 10 min.). It is evident if Fie. 1. vital actions are at work tending to raise the water in the branches, the rate 1] used Syringa vulyaris, as similar and straight branches of this shrub are readily obtained. E2 24 Scientific Proceedings, Royal Dublin Society. of flow downwards is reduced by this activity; and inasmuch as both branches are under similar conditions, the reduction of flow is the same for both. It may here be noted that an alteration in the head of 10 cms. makes a very sensible difference in the amount transmitted, raising it, to take an example, from 0°450 gram to 0:500 gram per 10 min. If now one branch be killed, the vital lifting-force, if present, will be removed, and we should expect the amount transmitted by the killed branch to increase correspondingly. ‘This increase, even though small, would be easily seen by comparing the flow through the dead and the living branches, when both have been again brought to the same conditions. For killing the branch I used either a jacket of steam or an injection of poison. When it is desired to kill the experimental branch with steam, the water in the cistern is run off through a small side-tubulure (¢), and, when the cistern is empty, steam is passed by the same tubulure into a wide tube (d), now placed round the experimental branch, and fitting tightly into a socket made for its reception in the bottom of the cistern. ‘The space round the top of this branch, and between it and the tube, is packed with cotton-wool. The supply of steam is kept up for 20 min.; the tube (d) is then removed; and the cistern is filled with water through the tubulure(c). After some time, during which the water in the cistern is kept stirred, when it is judged that the experimental and control branches have come to sensibly the same temperature, measurements of the amounts transmitted by each may be resumed. In this way it is easy not only to compare dead and living branches under the same conditions, but also rough manipulation and shaking is avoided. ‘I'hese latter are known to cause irregularities in the amounts of water transmitted by cut branches, probably owing to the displacement of or compacting of some clogging material on their upper surfaces. Observations made in this way showed in each case that the amounts transmitted by a branch before and after killing by steam were sensibly the same ; orif they differed in the experimental branch, the same difference was observed in the amounts transmitted by the control branch at the same time, the observable differences being due to changes in conditions which affected the flow in the living as well as in the killed branch. In Table I. is recorded an example of one of these experiments, which is graphically recorded in fig. 2. Dixon— Vitality and Transmission of Water through Plants. 25 Particulars of Experiment. A and B, two similar branches of Syringa vulgaris, each with five year- rings. A: length, 84:5 em. long. Upper diam. of wood, 0-8 em.; of pith, 0:15 em. Lower diam. of wood, 0'9 em.; of pith, 0°23 em. Head, 9:0 cm. B: length, 83:5 cm. Upper diam. of wood, 0°73 em.; of pith, 0:1 em. Lower diam. of wood, 0°85 em.; of pith, 0‘-l em. Head, 9:3 cm. Tare I. Amount transmitted | Amount transmitted Hour. Temp. of Cistern. per 10 minutes per 10 minutes by A. by B. is I 116°C. 164 gr. | 267 gr. O35 || mie7P 168 | 278 49 | 11°75° 172 272 | li. 24 11°85° 172 272 BO | 11-9° 178 “274 The cistern was emptied, and B surrounded with steam from iii o’clock to ii. 20. When the cistern was refilled and stirred, the measurements were resumed. | Amount transmitted | Amount transmitted Hour. Temp. of Cistern. per 10 minutes per 10 minutes by A. by B. | l il. 44 12:7°C. 203. gr. “216 gr. | he 8 12-8° "185 “230 | the 8B 12-8° “184 “268 iv. 46 12°85° “181 *240 v. 0 12-9° “184 7244 v. 18 peep "186 7253 Immediately after the experimental branch is surrounded with steam, the water transmitted through it becomes coloured, at first amber, changing to 26 Scientific Proceedings, Royal Dublin Society. brown. This change probably indicates the introduction into, or the pro- duction in, the transmitted water of some clogging material; for, instead of an increase in the amount transmitted downwards after the removal of the supposed vital lifting forces, we see from the able that the amount is diminished. According as the downward stream washes out this clogging material, the original rate of transmission is approximated to, but during the experiment not attained. It is evident that if there was any considerable length of branch below the steamed part, this material would accumulate there. The appearance of this material in the water transmitted by a steamed branch is of interest in connexion with some experiments discussed later. The small rise in the amount transmitted by the control branch noted immediately after the observations were renewed is probably to be attributed to a rise in temperature and consequent reduction of viscosity. os 100) Crammes Gonsmilted per (Omir. § 6 us ar 129° Temp. of jacket. Fie. 2. The effect of steaming on the experimental branch and the uniform behaviour of the control is clearly brought out in fig. 2, in which the ordinates are grammes of water transmitted per 10 minutes, and the abscisse are the times of the observations. The full line connects the successive observations on the control branch; the dotted line joins the observations on the experimental branch. The gap indicates the interval during which the cistern was emptied, and steam was applied to the experimental branch. Change taking place in the experimental branch demonstrated by the exit of coloured liquid below after the application of the high temperature renders the experiment somewhat unsatisfactory ; for, although the experiment gives no indication that the removal of vital processes from the stem does away with lifting forces opposing the downward motion of water, yet it is just con- ceivable that, if these forces previously existed and were removed, the clogging Dixon— Vitality and Transmission of Water through Plants. 27 of the branch during the steaming might just compensate for their removal. The possibility of such a coincidence, though very improbable, suggested the use of picric acid as a killing agent in place of steam. The experiment was at first arranged in just the same manner as before; and the initial rates of transmission of the experimental and control branches were determined. Then some dry picric acid was introduced into the water- tube of the experimental branch, and the observations were continued. No change could be detected in the rates of transmission which could be assigned to the removal of vital actions by the picric acid, even when the latter appeared at the lower end of the experimental branch. The gradual killing of the branch and the slow penetration of the picric acid in this method are open to objection, and would tend to render a change due to death less noticeable. To remove this objection, a modification was introduced by means of which the picric acid is quickly forced through the stem under pressure; and, in order to place the control under similar con- ditions, distilled water is simultaneously forced through it. This is arranged by having the tubes containing the water-supplies to each branch provided with a side-tubulure connected with a J-shaped glass tube containing a mercurial column. ‘he J-tubes and the tops of the water-supply tubes are provided with stop-cocks (e, 7, g,and / respectively, fig. 1, p. 23). At first the side-tubulures (e and /) are closed ; and the rate of transmission of distilled water under a low pressure is measured for each branch; then picric acid is introduced into the supply-tube of the experimental branch, and the stop- cocks at the upper ends of the supply-tubes (g and h) of both are closed, and the lateral tubulures (e and /) opened ; so that the picric acid is forced through one, and distilled water is forced through the other. When the pieric acid appears below, by the suitable manipulation of the stop-cocks, the pressure in each is again reduced, and observations are recommenced. Table II. gives the figures of such an experiment, and the results are plotted graphically in fig. 3. Particulars of Experiment. A and B, two similar branches of Syringa rudgaris: A with seven year- rings; length, 80 em. Upper diam. of wood, 0°85 cm.; of pith, 0°14 em. Lower diam. of wood, 0:92 em.; of pith, 0°:20cm. Head, 24 em. of water. B with 4 year-rings ; length, 80° em. Upper diam. of wood, 0°85 em.; of 28 Scientific Proceedings, Royal Dublin Society. pith, 0°16 em. Lower diam. of wood, 0:95 cm.; of pith, 0°20 em. Head, 24 cm. Tasre IT. Amount transmitted | Amount transmitted Hour. Temp. of Cistern. per 10 minutes per 10 minutes by A. by B. xii. 13 eZ 455 gr 470 gr. xii. 26 11:8° | “445 468 xii. 47 11:8° | “450 471 At i. 20, picric acid was supplied to B, and forced through under a head of 41°5 cm. of mercury. At the same time, water was forced through A under a head of 44 cm. of mercury. At i. 45, when the picric acid had appeared in quantity at the lower end of B, the head of 24 em. of water was restored to both. Amount transmitted | Amount transmitted Hour. Temp. of Cistern. per 10 minutes per 10 minutes by A. by B. i. 50 11°8°C. “498 gr. 500 gr. ii. 6 1:92 “497 “501 ii. 19 P60)? 500 503 FR a8 12:0° “501 506 | li. 57 12-0° 503 507 iii. 12 ipl 504 520 li. 27 | ge? 504 508 iii. 46 | IGE}? 506 512 | 9 12°3° “504 515 | iv. 26 12-4° 607 511 It will be seen that the mean rate of transmission before the high pressure was applied was 0-450 g. and 0:469 g. per 10 min. for the control and experimental branches respectively. During this time average tem- perature was 11:8°. After the pressure was applied forcing distilled water through the control, and picric acid through the experimental branch, the rates rose to 0:502 g. and 0:508 g. respectively for an average temperature of 12:0°. That is, the rate of transmission of the killed branch has increased by 8:3 per cent., while that of the living control has increased by 11:5 per cent. Probably the rise in both cases is due chiefly to the flushing-out of the branch Dixon— Vitality and Transmission of Water through Plants. 29 and the washing away of mechanical obstructions by the stream under high pressure, and partly to the small rise in temperature from 11:8° to 12:0°, which would perceptibly diminish the viscosity of the water. Inasmuch as the observed rise is as great in the case of the living branch as in that which is killed during the observations, it follows that there were no vital actions in either retarding the transmission. 500 A Picroe acid. IL * Ss S Gronmes transmitted per [Omin : S Ss uz Ws" 8° 12-42 Temp of jacket Fig. 3. The rapid or ultimate fading of the leaves after the death of the support- ing stem, or some portion of it, has been often used as an argument in favour of the view that vital processes in the stem participate to a large extent in the lifting of the transpiration stream.’ This interpretation of the observation is arbitrary, and, before accepting it, we must assure ourselves that no other change, introduced by the method of killing the stem, is responsible for the fading of the leaves, but that this latter is simply due to the inability of the leaves to obtain water when the cells of the stem cease to assist in raising water to them. A priori it would appear that the observation would receive a more natural explanation if, instead of assuming that the lifting force is reduced, we suppose that the killing of the cells of the stem allows substances to pass into or form in the water-capillaries which may render the walls of the trachez less permeable to water, or which may act injuriously on the cells of the leaves and cause them to fade.? Janse’s* and Ursprung’s‘ observation that the greater the length of the stem killed below the leaves, the more rapid 1 Janse: Pringsheim, Jahrbuch, 1887. Ursprung: Beihefte, Bot. Centralbl., 1904, and Jahrb. f. wiss. Bot., 1906 and 1907. 2 Weber, Ber. d. deutsch. Bot. Ges. 1885. 3 Janse, Joc. cit., 1887. 4 Ursprung, Joc. cit., 1904. SCIENT. PROO, R.D.S., VOL. XII., NO. III. F 30 Scientific Proceedings, Royal Dublin Society. is the fading, seems to me to favour this explanation. Failure to observe plugging of the lumina of the trachez in the killed region is no disproof that the partitions, which the stream normally traverses, are rendered less permeable, and that more complete plugging occurs higher up in the path of the water.' Both these surmises have already received experimental support,” and are further confirmed by some observations to be described later. Furthermore, the possibility that the leaf-cells fade from poisoning must be admitted until evidence to the contrary is forthcoming.® When a length of a branch is jacketed with steam so as to kill it, the fading of the leaves above which supervenes is not identical with that which occurs when leaves are deprived of adequate water-supply. ‘To quote an example: a branch of Populus, about 160 cm. long, carrying about 20 leaves, was jacketed for a length of 85 cm. with steam for 10 min. On the following day the older leaves had lost their fresh lustre; and on the fourth day they had considerably changed. This change consisted in a general dullness of colour over the whole leaf. Later the margin of the leaf became dark ; and this darkness gradually invaded the leaf between the veins, leaving a green border along the veins. The darkened margin subsequently dried and shrivelled, while the small terminal branches, the petioles, and the green parts of the leaf round the veins remained fairly turgescent. As the change proceeded in the mesophyll, the veins became coloured pink, and finally red- brown. This coloration was particularly noticeable when the leaves were viewed by transmitted light. Microscopic examination of the leaves showed that protoplasm of the mesophyll cells in the dark areas had contracted, and the cells were no longer turgescent, and the chlorophyll corpuscles had become discoloured, and bore a dirty brown-green tinge. During even the early stages of fading, sections of the veins revealed a pink coloration in the walls of the trachez, while the contents of the wood parenchyma were uncoloured. Later on the lumina of the trachez became filled with a pinkish brown material which ultimately seemed to choke the tubes. Shrivelling and withering of the leaf, except at the edges, did not occur till after these changes were complete. I have traced changes similar to these in the leaves of Tilia microphylla, Syringa vulgaris, Salix viminalis, and Acer Pseudo-Platanus, when the supporting stems had been steamed. On the other hand, when leaves fade simply from an insufficient water- supply, e.g. on a branch severed from a tree, shrivelling comes on while they are still green ; the smaller branches and the petioles also shrivel. Blackening 1 Weber, Joc. cit., observed obstructions either in the form of a clogging material or of tyloses in the trachew, where the killed region adjoins the living. 2 Dixon, Roy. Dublin Soc. Proc., vol. x., 1905, p. 48. 5 Dixon, loc. cit. Dixon— Vitality and Transmission of Water through Plants. 31 only appears after shrivelling, and occurs in irregular patches. ‘I'he veins do not change colour, and the walls of the tracheee do not appear coloured in transverse sections. The first appearance of colour-change is when the cell- contents of the mesophyll and parenchyma of the veins colour brown after death. This contrast of behaviour appears to me inconsistent with the view that the leaves in both cases fade from precisely the same cause, viz., simply and solely faiiure of water-supply. The fact that the mesophyll of the leaves of the killed branch discolours before shrivelling, while, when the water-supply only is interfered with, shrivelling precedes the discoloration, appears to support Vesque’s conclusion that in the previous case the leaves dry because they die, while in the latter they die because they dry.! But there is some- thing more than this. While the discoloration is proceeding in the mesophyll of the leaf the walls of the conducting tubes of the veins are becoming discoloured, and finally their lumina become choked with a trans- parent coloured material. The experiment described on page 26 indicates the origin of this material. Distilled water urged through a branch jacketed with steam issues as a coloured and more or less viscid fluid. The transpiration current also sweeps this clogging material before it; and, being deposited in the walls and lumina of the upper conducting tubes, it reduces their power of transmitting an adequate water-supply upwards. From this it would appear that the initial stages of the fading were caused by the poisoning of the mesophyll cells, while the final stages, withering and drying, were accelerated by the clogging of the walls, and stoppage of the lumina of the supply-tubes. In a previous paper? I have described how it is possible to cause the leaves of a branch to fade by supplying it with water which has passed through a killed piece of stem, although its direct supply of water from the root is not interfered with. Since then I have confirmed these observations ; and I have found that the leaves on a lateral branch, arising close below the killed region of a stem, may often exhibit the kind of fading which I have just described as characteristic of the leaves on killed branches. As in the previous cases the direct connexion between these leaves and the roots had not been interfered with, but their original water-supply coming up through the uninjured parts of the stem was left to them intact. The only cause I can suspect as being responsible for their succumbing is the passage backwards into their transpiration-stream of some harmful material from the dead region above the base of the lateral branch. It may also be 1 Vesque, Compt. Rend., 1885. 2 Roy. Dublin Soc. Proc., vol. x., 1905, p. 48. 32 Seten tific Proceedings, Royal Dublin Society. that some clogging material is drawn back or chokes the conduits leading to the leaves in question. From these observations I feel compelled to assume that at least part of the fading of leaves of steamed branches is due to the action of deleterious or poisonous substances on their cells. Nor do I regard as a serious objection to this view the fact that the action of these substances is not so rapid nor so complete as that of a solution of copper chloride.! That the acceleration of the shrivelling is in many cases due to clogging of the tracheal walls and to the stoppage of their lumina is shown by direct microscopic observation. It is also indicated by Weber’s results, who found that in some cases the simple reduction of the leaf area above the steamed region was sufficient to restore for a short time turgescence to the remaining leaves. In these cases presumably the water-supply, reduced by coming through the partially clogged branch, together with the water already in the branch, was sufficient to restore turgescence temporarily to the remaining leaves. ‘This supply being quickly exhausted, they withered finally, as the small supply coming through the clogged branch was insufficient to make up for their loss by evaporation. The recovery of the upper leaves when supplied with water above the steamed region is again a demonstration that the final stages of fading are due to the increased resistance opposed to the water-supply by the clogging materials in, or coming from, the steamed region. This increased resistance is also indicated by Weber’s observation’ that, even with a head of 62 cm. of mercury, appreciable amounts of water could not be forced through about 12 cm. of the basal part of the heated branch. The experiments and observations just detailed show, I think, that there ig no reason to believe that vital actions in the stem are needed to assist the transpiration current; the anatomical relations of the cells of the stem to the conducting tubes are also against the view that they can apply any elevating force. The ingenious suggestions of Ewart* show the difficulties a supporter of the vital theory is in on this account; while in Ursprung’s* diagram there seems no reason why the water forced into the conducting tubes should not move downwards rather than upwards. Seeing, then, that there is no adequate reason for supposing the cells of the stem to assist in the raising of the transpiration current, it is of the greatest interest to find out if, to raise the water in the highest trees, the 1 Ursprung, “‘ Ueber die Ursache des Welkens.’’ Beihefte z. Bot. Centralbl., 1905. 2 Weber, Joc. cit. 3 Ewart, Roy. Soc. Phil. Trans., B. 1905 and 1907. . ‘Ursprung, Biolog. Centralbl. Bd. xxvii., 1907, p. 53. Dixon— Vitality and Transmission of Water through Plants. $88 osmotic pressure of the leaf-cells and the tensile strength of the transpiration stream are taxed nearly to their limit. Unfortunately at present there are no determinations as to the pressures in the leaf-cells of these trees during transpiration, nor of the resistance of their conducting tubes to the current. With regard to the tensile strength of water containing air, the highest experimental value obtained for it up to the time of the publication of the tension theory was, we believed, that of Dr. Joly and myself, viz. 74 atmospheres.. But this, from the nature of the method, was necessarily much below the actual value, and only indicated the adhesion of water to a surface it incompletely wets. Berthelot’s? experiments, so far as I am aware, have been always recently quoted as applying only to water which is free from air.’ Not having seen the original papers, Dr. Joly and I were under this impression when we made the experiments just alluded to. As a matter of fact, I now find, on looking up his paper in the Annales de Chimie et de Physique, xxx., 1850, that Berthelot at first experimented on water supersaturated with air. He filled a strong-walled capillary glass tube with water at 28° and allowed it to cool to 18°. Air was drawn into the tube during the contraction of the water as it fell in temperature. The tube was then sealed at its fine-drawn end. After raising the temperature again, the air was forced into solution, and when the water occupied the whole space enclosed by the tube, it was again allowed to cool to 18° or lower. It was then found that the slightest shock caused the air dissolved in the water, up to that moment, to reappear as a bubble. Before rupture relieved the tension, the water was distended by 74> of its volume. The tension required to produce this dilatation Berthelot estimated at 50 atmospheres. Later on, at Regnault’s suggestion, Berthelot performed the same experiment with air-free water, and obtained similar results. The tenacity of a water-film may also give us a minor limit for the tensile strength of water. A film of soapy water in air which is only 12ym, or 12 x 107 cms., thick is stable when stretched in a rigid frame. In this position, it supports the stress of twice the surface-tension (7) of the soap- solution. 7’ = about 25 dynes per cm. Therefore a column of water 1 sq. : ¢ 7 em. across can support a tension of at least — dynes. An atmo- sphere pressure is equivalent to 981 x 10* dynes. So we find according to 1 Roy. Soc. Phil. Trans., B, 1895. 2 « Sur quelques Phénoménes de Dilatation forcée des Liquides.”? Ann. de Chimie et de Physique, xxx., Sér. 2, 1850, p. 232 et seq. 3 Text-book of Physics, Poynting and Thomson, ‘‘ Properties of Matter,’’ 2nd ed., p. 123; and Ewart, Annals of Botany, 1906, p. 444; Worthington, Roy. Soc. Phil. Trans., A., 1892. SCIENT. PROC. R.D.S., VOL. XII., NO. III. G 34 Scientific Proceedings, Royal Dublin Society. this method that the cohesive force or tenacity of water must be at least 42:5 atmospheres. To raise water in one of the tallest trees, say, 100 m. high, would require a force equal to the hydrostatic head, viz., 10 atm., plus the force necessary to overcome the resistance experienced by the transpiration stream. Hwart* estimates this at about 50 atm. in all. This figure he arrives at by assuming the same transpiration velocity and the same resistance as he found in lower trees. As has been pointed out, his determinations of the resistance are excessive; and further, there is every reason to believe that, if the supply is inadequate, the transpiration velocity will be reduced.2 This reduction in velocity has been shown to occur in lower trees. Using my own determina- tion for the resistance,® we obtain a very much lower figure for the tension, which must be supported by the transpiration current of the tallest trees. In the case of a stem with exceptionally high resistance, I found that a head of water equal to the length of the stem traversed was able to move water at the transpiration rate. If this figure holds for the highest trees— and it is likely to be excessive, both because the transpiration rate in the high tree will be slowed down, and because its wood will have a smaller resistance than that of the tree in which the determination was made—we should add to the 10 atm. necessary to support the head another 10 atm. to over- come the resistance of the conducting tubes. ‘The water-columns then in the tree would have to support a tension of only 20 atm., which is well within the minor limit obtained for the tenacity of air-saturated water. With regard to the osmotic pressures in the leaf-cells necessary to keep them turgid against the pull on their sap, we have no information applying to the high trees under consideration. I have found that the cells of the leaves of several low trees, e.g., Laburnum and Tilia, are able to support an external pressure of 26-30 atm. without showing signs indicating that their turgescence was overcome.’ It is certainly not unreasonable, in the absence of direct observation, to assume similar pressures in the leaves of higher trees. 1 Ewart, Roy. Soc. Phil. Trans., B, 1907, p. 391. 2 Dixon, Roy. Soe. Proc., B, 1907. 3 It may be noted that the resistance of the wood of the yew in which my determination was made is exceptionally high, as its wood is exclusively formed of tracheides only about 2 mm. long, and so the number of cross-partitions which must be trayersed-is exceptionally high. +«~—+ +—————}— / ja + +—___+— dk. 100 | j h r aaa FE ! i : ) 80 sags | vil HU ies i | Pe Hip 60 Woy) L eee clea 3) t L 40 U +21 | = = =F ; 7 H eae lh eo be ee ® in LA | | o 20 40 60 80 100 720 140 160 hour» Serres IV.—t=19° C. One living bean, not desiccated, used in each experiment. Time in hours. Increase in weight Remarks. Per cent. 1 it lee 118-2 , a 119-9 | In pure walle The seed in No. 2 2 res 71:8 we been in a damp place pre- 94 ; 95-0 | viously. 450 beans of one sample, dried as described, lost 13:2 °/, of their original weight. 100 beans of another sample lost 11-5 °/,. Adding on this, 12 °/, roughly, to No. 1, Series 1V., gives 182 °/,, which agrees with the results in the other series. 40 Scientific Proceedings, Royal Dublin Society. Serius V.—t = 25° C. One sweet pea; not killed. Remarks. In saturated KNO, solution. changed Time in hours. Increase in weight Per cent. il, UB 0 : : 1:3 3025). : ; 2:7 8503. : . 65:3 874-398 ; 5 laléler 2. 2 ee : : 1:0 TB 6 : : 4:3 1563 ~—Cy : 5 Mle 1 (additional) . 117-4 Removed to pure water ; 23-133 : . 109:2 } several times. 5, 482 ‘ 2 3:0 1GS°23000 ene oI-2 FG ee: 47-72 (additional) 117-2 Removed, &c. 4, Bes q : 5:0 5. 374 : ; : 7:0 ) In almost saturated vapour. 6. 7a te aes ees aa SERIES V. % Increase in wt. t-25°C One diwang pea in saturated ANOs 245 mm pure walter at end. 200 hours subtracted from all the tres of (1) Yiet==t5 a aa 120 = ==} - =x) cee oa AE = = v7 I | H 2) \f | | , 100 Sj + +$-—— ! i le — } ij ! | t 80 iT ! | U | fr 60} +—+ Tia aH is + t | + {I j I 40 + tp - + ! (2. =a h i la alte Vn 20 +t 4A / ve) La 1+ 160 1sollours 200 Arkins— The Absorption of Water by Seeds. 4] Maximum Values of Increase in Weight for Beans. 1, 4. 136°6 (4) t 5. 146:4 (4) Tilo. Bp 120:0 (1) lea 141-0 (1) In pure water. Tee 157-0 (1) Mean (te) = Oe per cent. in- Re 155-0 (1) crease in weight. ur. 1. 126-0 (1) io I, 132:0 (1) Mean, 140-2 (14) ( In eram-molecular KNO,. tin 136:5 (4) Mean (4) = 186:5 per cent. in- crease in weight. m1. 157-0 (1) m. 2. 144-0 (1) . : : 2 4 an ey 140-0 (1) : i In saturated KNO, solution. AEa 146-0 (1) ; Mean (8) = NBDD per cent, in- BO. 169-0 (1) crease in weight. mm. 3. 144:0 (1) Mean 144:6 (10) Comparison of Living Seeds from Series IL. and of Dead Seeds from Series ITI. Livinc. Deap. Per cent. incr. Per cent. incr. Per cent. incr. Per cent. incr. in water. in saturated KNOs; sol. in water. in saturated KNOs3 sol. 120 144 126 146 = 157 — 169 — 140 — 142 These results make it clear that, while there are considerable variations between individual seeds, there is no distinguishable diversity in the behaviour of living and dead seeds. ‘They also show that the presence of potassium nitrate has no effect upon the final weight except in so far as the density of the solution is altered; also the salt may be washed out of the seed, its final weight then being what it would have been if placed direct in pure water On carefully drying the surfaces of seeds taken from the nitrate solutions, and allowing them to dry internally by evaporation, a forest of crystals appeared all over the surfaces, growing to a height of several millimetres. This is a further proof that the salt really entered the seeds in quantity. 42 Scientifie Proceedings, Royal Dublin. Society. A study of the time-weight curves shows that there is no difference in the rate of absorption of seeds, living or dead, in pure water or in salt solutions. Hence one is forced to the conclusion that there is nothing functioning as a semipermeable membrane in the seeds, either living or dead; for if there were, water only would pass in, whereas in reality potassium nitrate solution passes in just as freely. These results are in complete agreement with those obtained by A. J. Brown; for it is only in the special outer covering of Hordeum and some other Graminee that he finds a semipermeable membrane. ‘Thus it appears that these seeds do not take up water by the agency of osmotic forces, but solely by capillarity and imbibition, so that the material of the dried seed before absorption of water is complete must be in quite a different condition from that which it is in when germination begins and the cells resume their active life. Density Determinations.—It was noticed that beans placed in potassium nitrate solution sank, but rose to the surface subsequently, and then, after an hour or more, sank. At first it was suspected that this was due to the | expansion of the seed caused by the absorption of pure or less saline water, its density being thus brought below that of the salt solution. ‘This was completely disproved by a determination of the density and weight of a seed, apparent and real, in pure water. Time Weigh my Weight in water. Density. | b= 227590. | Hrs. Mins. | ah ine 0 2 | 0-684 | 0:129 (100 °/,)| 1:282 Air-dried bean used. O 32 | 0-694 | 0-116 | 1:201 | Marked crinkles at one end. 1 8 | 0-788 | 0-115 | 1-186 | Crinkles at both ends. 1 26 0-753 | 0:113 Mona 1 52 | 0-795 | 0-111(86°/,) | 1:162 (94 °/,) | 5 20 | 1267 | 0-164 1-148 | Gaaties etimos) ome | 20 45 | 1-883 | 0-179 | 1-148 | 25 0 1:398 | 0-178 (188 °/,) | 1:146 | Experiment stopped; ab- sorption not complete. Thus it is clear that the rising to the surface of the bean in the salt solution was due to the apparent weight falling to 86 per cent., and the sinking due to its rising to 1388 per cent., whereas the density decreased throughout. The crinkling of the coat changes the apparent weight. Beans in a moist place may become so crinkled as to float in pure water. Another air-dried bean used had the initial density of 1:287, Its apparent weight decreased to Arxins—The Absorption of Water by Seeds. 43 89-5 per cent. and rose to 125°5 per cent. A desiccated bean was found to have a density of 1:180 initially, and after 310 hours in water it had fallen to 1119. The density of desiccated bean substance was determined by placing the chopped-up material in a specific-gravity bottle with mercury. The value thus obtained, 1:523, enables us to calculate the size of the internal hollow in the seed—about 33 per cent. of the whole volume. Titration Method. In order to detect the presence of a small area of semipermeable membrane, which might have escaped notice by the weighing method, owing to the difference between seed and seed, A. J. Brown’s device was adopted. Quantities of bean seeds, as bought and after desiccation, were placed in measured volumes of normal sulphuric acid, 49 grms. per litre, decinormal iodine in potassium iodide, 12°7 grms. of iodine per litre, and decinormal sodium chloride, 5°85 grms, per litre. They were left in these solutions for forty-three hours; then 25 c.c. was drawn off from each and titrated, the chlorides were evaporated to dryness and ignited to destroy organic substances which interfered with the silver nitrate and ammonium In each case one hundred beans were used, weighing 62 grms. as bought. Of this sample 98 per cent. germinated. The hundred desiccated beans of experiments 4 and 10 weighed 47 grms. In this sample 100 per cent. germinated before drying. thiocyanate methods of titration. SCIENT. PROG. R.D.S., VOL. No. Condition. Solution added. Se ee eee 1 As bought. 100 c.c. H,80,(N) 82:2 c.c. HpSO,. 2 % ‘5 9p 81°6 ¢.c. 3 » . on 80°8 c.c. 4 Desiccated. 5 “s 85:6 c.c. 5 As bought. 100 ¢.c. water. 0:3 c.c. alkali. 6 i 100 c.c. waci (7): 99-2 ¢.c. NaCl. T z 100 ¢.c. water. Traces of chlorides. 8 » 100 c.c. water, after rinsing. | Traces of chlorides. 9 59 100 ¢.:, Iodine (3) Decoiorized. 10 Desiccated. % 55 Decolorized. 11 After germination.| 100 c.c. NaCl la): 92 c.c. XII., NO. Iv. 44 Scientific Proceedings, Royal Dublin Society. From these results it may be seen that there is no concentration of the solutions, as would be the case if a semipermeable membrane existed. There is, however, a weakening of the solutions due to various causes. In the acid experiments it is due to the alkalinity of the cell substance, as was noticed by A. J. Brown in Hordeum. Experiment 5, however, shows that, on crushing the beans and extracting with hot water, only an amount of alkali equal to 0:3 ¢.c. normal NaOH was found derived from one hundred beans. The extract was boiled to expel CO, before titrating. In the case of the undesiccated seeds, some of the weakening of the acid is accounted for by the moisture of the seeds diluting the solution. ‘This is evidence in favour of the permeability of the seeds. The diluting effect does not, of course, occur in experiment 4 with desiccated beans. To ascertain whether the seeds weakened the acid by unextractable-alkali, or by con- centrating the solution in themselves, the seeds used in experiment 3 were drained, leaving 48 ec. of weak acid, corresponding to an absorption of 52 ce. of acid solution. 37 c.c. of normal acid was found in the beans, though they had absorbed 52 cc. of normal acid, besides weakening the remaining 48 c.c., so that it only contained 38°8 c.c. of normal acid. This gives 75°8 c.c. of normal H,SO, remaining; 24:2 c.c. must therefore have been neutralized by the bean-cell substances, or, allowing for the dilution effect as a maximum of 10 e.c., 14:2 c.c. was neutralized, as in experiment 4. This gives 0°142 c.c. of normal H,SQ, per bean, viz. 0:007 grm., or 1°47 per cent. of the weight of desiccated bean-substance. The sodium chloride was practically unchanged, the dilution only being slight, as traces of chloride were found in experiments 7 and 8, having been washed out of the seeds. The chlorides initially present prevent the dilution being noticeable. In the iodine experiments, microscopic examination showed that all parts of the seed were penetrated by the iodine; the solution was decolorized by the formation of large quantities of starch iodide. Thus none of the facts ascertained give any reason to believe in the presence of a semipermeable membrane in dry seeds absorbing water in their initial stages. In experiment 11 the seeds used had all taken up about 140 per cent. of their dry weight of water, and had just put forth radicles. They were dried in a warm room, about 15° C. for three hours, and then placed in the salt solution. A dilution effect was noticed amounting to 0°92 of the original con- centration. But as the hundred beans contained about 65 c.c. of water, if they were completely permeable, the dilution should have been 122 = 0°61. So here semipermeability is met with, though it is not absolute. Evidently the protoplasm in the growing living cells is the semipermeable substance. Avxins—The Absorption of Water by Seeds. 45 Note on the Colouring Matter of Beans. It was observed in titrating that the colouring matter of the beans originally in acid turned purple with alkali, forming an accurate end-mark as tested with phenolphthalein. The purple colour is due to a gelatinous precipitate, which is dark-brown when seen on filter-paper, the filtrate being a deep red when alkaline, almost colourless when acid. The precipitated colouring matter dissolves in dilute acid, giving an orange-yellow solution. Experiments on the first Appearance of Respiration in Seeds. It was thought that it might be legitimate to use the production of CO, by seeds as a test for the beginning of active cell-metabolism, and consequently the end of the period of latent life. Accordingly, about two hundred beans, as bought, were sterilized outside by placing in a strong solution of mercuric chloride for several minutes, and then well rinsed with distilled water. This procedure does not injure the seeds; for 98 per cent. subsequently germinated. They were then placed in a conical flask similarly sterilized, completely filled with distilled water. ‘This had a rubber cork, and a long and short glass tube. A current of air was passed by a force-pump through a Geissler potash bulb, and through a second bulb with barium-hydroxide into the conical flask, which was tilted so that the water trickled out slowly as the CO, free air entered. ‘his was continued till only a little water remained— just enough for the beans to imbibe. The tube through which the water had flowed was then connected to a barium-hydroxide bulb joined to another Geissler potash bulb. Two clips closed the tubes leading to the flask. A number of such experiments were performed; and the time that elapsed before CO, could be detected on removing the clips and passing in purified air was gradually narrowed down till it was detected with certainty two hours after the moistening of the seeds with mercuric chloride solution, though a slight turbidity appeared in the barium hydroxide bulb on the exit an hour and a quarter after moistening. The great difficulty of obtaining a perfectly clear solution of this hydroxide is the cause of the uncertainty. Identical experiments with a saturated solution of chloroform in water as the moistening liquid gave a copious precipitate of BaCO; after eight hours, though none of the chloroformed seeds germinated. This gives us reason to believe that the evolution of CO, cannot be taken as an indication of the renewal of living cell-metabolism; and thus the only ready chemical test that can be applied is useless. I subsequently found that F. Stoward! has shown that this evolution of CO, may, partly at any rate, 1 Ann. Bot., xxii., 1908, p. 410. 46 Scientific Proceedings, Royal Dublin Society. be due to a respiratory enzyme being active after the cells have been poisoned by chloroform or toluene, but not after formaldehyde poisoning. This evolution of CO, was also found in the case of desiccated beans heated to afew degrees above 100°C., when moistened as described. A copious precipitate was observed after six hours, but no experiments were tried over a shorter period. All these oxidation experiments were carried out at about 12°C. Summary. 1. Bean seeds, living or dead, take up the same quantity of water in their initial stages. 2. The final weight reached, whether living or dead, is independent of the presence of potassium nitrate, except in so far as the salt present alters the density of the water. 3. The rate at which distilled water is taken up is not greater than the rate at which the salt solution is absorbed, any difference which may exist being masked completely by the variations between apparently similar seeds under similar conditions. 4. Seeds placed in potassium nitrate solution, and then in pure water, lose weight ; their final weight is then what it would have been had they been placed direct in pure water. 5. Titrations show that bean seeds placed in normal H.SOQ,, decinormal iodine, and decinormal sodium chloride produce no concentration of these solutions. ‘he chloride remains almost unchanged. ‘The iodine is de- colorized. The H,SO, is to a small extent diluted; a very small amount is neutialized by extractable alkali, while the tissues of desiccated beans neutralize 1-47 per cent. of their weight of H.SO,. 6. These facts prove that there is no semipermeable membrane in bean seeds till germination begins and the cell-protoplasm acts as such, and that there is no difference in absorption between living and dead seeds until after germination. The forces concerned are those of capillarity and imbibition in the initial stages, but of osmosis after germination. 7, The evolution of CO, may be detected less than two hours after the air-dried seeds have been first moistened. CO, may be detected with living seeds, and with those killed by chloroform. I wish to thank Dr. H. H. Dixon for his constant advice and for every facility in obtaining apparatus and material. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 5. APRIL, 1909. NOTES ON THE POLLINATION OF CERTAIN SPRKCIES OF DENDROBIUM. BY A. WL Gs INBIRIR NLD. (PLATES V.,, VI.) [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price Sixpence. tee Ja! Museum useun® 7 he nua ls ead V. NOTES ON THE POLLINATION OF CERTAIN SPECIES OF DENDROBIUM. By A. F. G. KERR, M. D. [ COMMUNICATED BY PROFESSOR H. H. DIXON, SC.D., F.R.S. | Read Decrmprr 22, 1908. Ordered for Publication January 12. Published Aprit 3, 1909. Prares V. ann VI. Tue following notes were made in northern Siam, either on orchids growing naturally, or on specimens removed from the neighbouring jungles to my garden for convenience in observation. As I was able to work out most fully the mechanism of pollination in Dendrobium Dathousieanum Paxt., and as its mechanism is typical of that of many other species of Dendrobium, I will give my observations on that species first. Dendrobium Dathousieanum has large flowers with a spreading perianth. The lip is large and concave, with a patch of fimbria on its tip; posteriorly it is produced into a short spur, which contains the nectary. A passage, the roof of which is formed by the column, the sides by the sides of the lip, and the floor by the disc of the lip, leads to the nectary. Hereafter I shall refer to this portion of the flower simply as the passage. The direction of the passage is upwards and backwards. Owing to a thickening of the dise, the floor of the passage rises posteriorly till it meets the roof, thus blocking the entrance to the nectary. ‘There are three longitudinal ridges on the disc, which end abruptly before the column is reached. A slight weight on the lip depresses it sufficiently to allow of the entrance of a proboscis into the nectary. A reference to Plate VI., fig. 1, which represents a longitudinal section, slightly to one side of the median line, through the column and base of the lip, will explain the structure of the flower more clearly. ‘The anther has a thin free edge pointing backwards and slightly overlapping the SOIENT. PROC. R.D.S., VOL. XII., NO. V. K 48 Scientific Proceedings, Royal Dublin Society. rostellum. ‘The filament is attached just behind the apex of the anther, and broadens towards its base, where there is a well-marked fold as though it were being kept forcibly bent backwards; a great deal of the elasticity of the filament is no doubt due to this fold. The rostellum is broad; its anterior margin is thin and recurved; its posterior margin is thickened ; it is jointed about its middle to the prolonged anterior border of the stigma. Immediately behind the rostellum is the large, slightly concave stigma which occupies almost the whole breadth of the column, and forms a great part of the roof of the passage. This Dendrobium is pollinated, as far as my observations go, entirely by one species of bee. This bee Lieut.-Col. C. T. Bingham has very kindly identified for me, from a very imperfect specimen, as Lithurgus atratus, a species about the size of our honey-bee. When the bee enters an undisturbed flower, it forces its way along the passage towards the nectary, till its head rests in the angle formed by the thickened disc of the lip and the column, its weight depressing the lip sufficiently to admit of the insertion of its proboscis into the nectary. So far, the position of neither the anther nor the rostellum has been altered. On looking at the semi-diagrammatic representation of a section through the flower on Plate VI., fig. 1, it will be seen that the retreat of the bee is impeded by the posterior margin of the rostellum and the free margin of the anther, both of which point downwards and backwards. As the rostellum is jointed on to the upper border of the stigma, and the anther is jointed to the tipof the filament, the bee, as it retreats, tilts both these structures. ‘he rostellum is tilted upwards till it comes in contact with the pollinia, and deposits on them some of its sticky secretion ; the anther is continuously tilted till the pollinia, now smeared with the sticky secretion of the rostellum, come in contact with the thorax of the bee, to which they firmly adhere. As the bee further retreats, the pollinia, now adherent to the bee, are drawn completely out of their cells. By this time, the anther has been tilted till it is completely free of the clinandrium, and when, by the further retreat of the bee, it has been completely released, the elasticity of the filament comes into play, and the empty anther is jerked downwards till it lies in front of the entrance to the passage (Plate VI, fig. 2). When free from the flower, the bee flies away with the pollinia adherent to its thorax; on visiting another undisturbed flower, it makes its way into the passage of that flower, where the pollinia on its thorax come into contact with the stigma, and adhere toit. As the bee retreats, it leaves the pollinia of the first flower on the stigma of this second flower, and withdraws the pollinia of this flower from their cells, leaving the anther blocking the entrance to Krrr—The Pollination of certain Species of Dendrobium. 49 the passage, as in the case of the first flower. If, however, a bee attempts to enter a flower which has received a previous visit from that or any other bee, it finds the anther blocking the entrance to the passage; and if it attempts to force its way in, it only pushes the anther backwards, till the latter comes in contact with the stigma (Plate VL, fig. 3), to which it adheres and which it completely covers, thus preventing the deposit of any more pollen on the stigma, or, if no pollen has already been deposited, of any pollen at all. When the flowers are being frequently visited by bees, many may be found with their empty anthers pushed backwards on to the stigmas. This mechanism prevents self-pollination, and allows of only a single visit to the nectary of each flower. In the case of the first flower visited by a bee, poliination cannot take place, as the bee enters without pollinia, and on its retreat leaves the anther blocking the way to the stigmas. I found that when a flower of Dendrobium Dalhousieanum had its pollinia removed, and placed on its own stigma by artificial means, the flower soon withered away ; but when the pollinia were placed on the stigma of the flower of another plant, that flower formed a capsule. Thusif self-pollination could take place by natural means, it would probably be of no benefit to the flower. I artificially pollinated seven flowers of D. Dalhousieanum ; of these six were pollinated with their own pollen, and all withered without showing any swelling of the ovary ; the seventh was pollinated with pollen from the flower of another plant, and formed a good capsule. I did not experiment on a sufficient number of flowers to enable me to say definitely that a self-pollinated flower would never produce a capsule; but from these experiments, and from the mechanism of the flower, it may be concluded that a capsule is never naturally produced by self-pollination in this species. Most of the Dendrobiums belonging to the section Hudendrobium show the same mechanism as D. Dalhousieanum. In all the species belonging to this section examined by me the filament was elastic, and the anther, after dislocation, was jerked downwards in front of the entrance to the passage. In the following list I give the results of artificial pollination in various species of this section. In many cases the number of flowers experimented with is too small to enable definite conclusions to be drawn with regard to that species; but the results, taken all together, tend, with a few exceptions, to support one another. By the term “ cross-pollinated,’ I mean that a flower has been pollinated with the pollen from a flower on another plant. Under the heading “Self-pollinated” I have included eleven flowers of D. capillipes, which were pollinated with the pollen from other flowers on the K 2 50 Scientific Proceedings, Royal Dublin Society. same raceme, none of which produced capsules. In all other cases under this head the flowers were pollinated with their own pollen :— Nember (orl tcelfpollinetiont Cross-pollination. ee spain mented on. | Successful. Vereen See, sae 1. D. Falconert Hook, é 1 0 1 0 0 2. D. capillipes Reich. f., . 28 2 24 2 0 3. D. Dalhousieanwm Paxt., 7 (0) 1 0 4. D. senile Parish, . ; 2 2, 0 0 0 5. D. aggregatum Roxb., 28 0 26 2 0 6. D. gratiosissimum Reich. f., 8 0 6 2 0 7. D. tortile Lindl., 3 2 0 1 0 8. D. Pierardi Roxb., . 4 0 4 0 0 9. D. primulinum Lindl., 5 0 3 2 0 10. D dixanthum Reich. f., . 3 0 3 0 0 11. D.crepidatum Lindl. « Paxt. 2 2 0 0 0 12. D. thyrsiflorum Reich. f., 4 0 4 0 0 13. D. chrysotozwm Lindl., 4 0 3 il 0 14. D. crystallin Reich. f., . 4 0 4 0 0 15. D. Calceolaria Carey, 5 0 4 1 0 16. D. fimbriatum Hook, 4 0 4 0 0 ToraLs, 0 5 112 8 92 12 0 Taking the totals of the above list, we find that 100 per cent. of the cross- pollinated flowers formed capsules, while only 8 per cent. of the self- pollinated flowers formed capsules. If we except the three species, D. senile, D. tortile, and D. crepidatum, which, I will show below, have certain peculiar characteristics, then less than 3 per cent. of the self-pollinated flowers formed capsules. ; The following are a few notes on some of the above-mentioned species :— Dendrobium capillipes—Of the two flowers which are given above as forming capsules after self-pollination, one dropped its capsule while still green, so that really only one self-pollinated flower produced mature seeds. The fertility of these seeds was not examined. Kerr The Pollination of certain Species of Dendobrium. d1 I frequently tried in this and other species to see if the pollinia would rebound from the lip on to the stigma, when the anther was suddenly dislocated, as C. Darwin has described in the case of D. chrysanthum; in D. capillipes, after many trials, I obtained this result once. In all my trials the flower was held in its natural position. If such self-pollination occurs in nature in D. capillipes, it must be of very rare occurrence, and most probably would not be followed by the formation of a capsule. I have had no opportunity of examining D. chrysanthum. Dendrobium senile.—In this orchid the filament is elastic; and the anther, on dislocation, falls downwards and forwards. Still access to the stigma is not completely prevented, as the filament is relatively short, and does not allow the anther to fall down sufficiently to entirely block the passage. Hach shoot of this species only produces one or two flowers. Frequently a whole aggregation of plants growing naturally may be observed with only one or two flowers ; so it may often happen that a bee, after visiting a flower, returns to that same flower, and pollinates it with its own pollen, having meanwhile made a fruitless search for another flower. Such a flower would then prob- ably produce a good capsule, as in both the flowers which I experimentally pollinated with their own pollen, good capsules were produced. ‘The flowers of this orchid remain fresh for a much longer period than those of any other Dendrobium which I have observed. One of the flowers was self-pollinated after having been in bloom for a month. Dendrobium tortile—The success of self-pollination in this species is somewhat remarkable. The anther is rather small, but still when pushed backwards it adheres to and protects the surface of the stigma. Dendrobium crepidatum.—In this species the filament, though elastic, is very slender, and the anther is readily knocked off, in which case free access to the stigma is allowed after the flower has been already visited. Not only did self-pollination succeed in the two flowers experimented with, but in another plant of this species cleistogamy occurred, three buds which did not expand forming capsules, In this case I believe the pollen tubes penetrated the rostellum which normally, before the flower opens, comes to lie in close proximity to the pollinia. I have examined the flowers of a few species of Dendrobium not belonging to the section Eudendrobium. The following are my results :— Dendrobium lasioglossum Reichb. f. (?), D. draconis Reichb. f., D. for- mosum HRoxb.—In these three species the filament is not elastic. When 52 Scientific Proceedings, Royal Dublin Society. dislocated the anther does not fall forwards and block the entrance to the passage; yet in none of these species was self-pollination effectual. In D. lasioglossum (?) one flower was self-pollinated without result; three flowers of D. draconis were self-pollinated also without result, while two flowers of the same species which were cross-pollinated produced good capsules ; two flowers of D. formosum were self-pollinated without result. Dendrobium parcum Reichb. f.—In these species the flowers are small ; the lip stands out horizontally much below the top of the column; no passage is formed between the lip and the column; the column is not horizontal, but nearly erect. ‘The filament is very elastic, so much so that when the anther is dislocated, it jerks the latter strongly downwards on to the stigma. The anther then remains adherent to the stigma, preventing further access to that organ. Five flowers of this species were self-pollinated without the formation of a capsule; one flower which was cross-pollinated formed a good capsule. Dendrobium secundum Lindl.—In this orchid the lip is small, and closely parallel to the column; the sides of the lip are turned up, and partially embrace the column, forming a narrow passage (see Pl. VI., fig. 4). On the roof of this passage is the rather deeply concave stigma. The anther has a short non-elastie filament; when dislocated, it is prevented by the up-turned sides of the lip from falling down and blocking the entrance to the passage. The filament is so slender at its insertion into the anther that the latter is very easily knocked off. The pollinia, which are nearly black, adhere more firmly to the object, by means of which they are removed, than in the case of any other species of Dendrobium examined by me. The posterior border of the rostellum forms a stiff forked plate, projecting backwards and down- wards. The whole rostellum is fixed, and not hinged on the upper border of the stigma, as is the case in D. Dathousieanum. The flowers grow in a dense, short raceme. On examining one of these racemes, which is in full bloom, and which is being frequently visited by insects, it may be seen that the older flowers towards the base of the raceme have lost their anthers, the most recently opened flowers alone retaining their anthers. If a young flower be taken, its pollinia removed, and then an attempt made to insert the pollinia into its own stigma, it will be found quite impossible to do so owing to the narrowness of the entrance to the stigma. On examining an older flower, it will be seen that the stigma has widened considerably, so that the pollinia can enter with ease. If the pollinia be removed with a fine style from a young flower, and introduced into one of the older flowers with a wide stigmatic cavity, it will be found that, on withdrawing the style, the Krrr—The Pollination of certain Species of Dendrobium. 53 pollinia are caught by the stiff posterior border of the rostellum, which scrapes them off the style and retains them in the stigmatic cavity. In this species, then, self-pollination is prevented, as the pollinia are removed from a flower before its stigma has opened out sufficiently to admit them. A reference to Pl. VI., figs. 4, 5, and 6, will better explain the structure of the flower in this species. Dendrobium stuposum Lindl.—The filament is not very elastic; but the anther on dislocation falls over the entrance to the passage. Three out of five self-pollinated flowers produced capsules. Dendrobium incurvum Lindl. (?).—On dislocation the anther falls forward ; but the filament is not long enough to allow of the anther being pushed over the stigma; however, the dislocated anther would probably prevent access to the stigma unless it was knocked off the filament. Five self-pollinated flowers produced no capsules. Dendrobium bellatulum Rolfe.—The filament is slightly elastic; but the dislocated anther is with difficulty shoved on to the stigma. The passage is very large, and only very partially blocked by the dislocated anther. Two self-pollinated flowers produced capsules. Dendrobium ciliatum Parish.—The filament is very short. ‘lhe anther, on dislocation, does not block the passage; any pressure from the front only pushes it back into its original position. Although in this species there is no mechanical contrivance to prevent a flower being pollinated with its own pollen, my experiments show that self-pollination is not effective. Sixteen flowers which were self-pollinated produced no capsules, two cross- pollinated flowers producing good capsules. Bariee i) =o es EXPLANATION OF PLATE V. PLATE Y. Reproductions of photographs of the flower of Dendrobiwm Dalhousieanum Paxt., with the petal and sepal on one side removed, also half of the lip and of the dorsal sepal. The nectary has not been opened. For three of these photographs I used a dead specimen of the common English honey-bee, which is approxi- mately the size of Lithwrgus atratus, the bee which fertilizes this orchid in N. Siam. The photographs are slightly smaller than life size. The flowers were — very kindly supplied by Messrs. Sander, of St. Albans. Fig. 1—Normal position of anther. Fig. 2.—A bee, after entering the passage, has just commenced to retreat. The anther is very slightly tilted. Fig. 8.—The bee has retreated clear of the passage, but is not yet free of the anther, which is now considerably tilted. The pollinia have been drawn from their cells, and can be seen adherent to the thorax of the bee. Fig. 4.—A bee attempting to enter after the dislocation of the anther. » SCIENT. PROC. R. DUBL. SOC., N.S., Vor. XII. PLATE V. Cun ree. | ayer ten Saree ; ; a ae Se it UA sles her arches teen SN em 7 a aren Pre ‘ EXPLANATION OF PLATE VI. PLATE VI. Dendrobium Dathousieanum Paxt.—Three semi-diagrammatic sections through the column and base of lip, slightly to one side of the median line. Fig. 1.—Normal position of parts: a, vascular bundle of column; 8, base of dorsal sepal removed ; ¢, fold of filament; d, filament ; e, anther ; f, pollen-cell ; g, rostellum ; h, stigma ; 7, passage ; k, ridge on lip ; J, callus on lip; m, base of sepal ; », nectary ; o, stigmatic canal. Fig. 2.—Anther after dislocation, thrown forwards and downwards, by the elasticity of the filament. The anther should be a little lower. Fig. 8.—Position into which the anther is shoved if a bee attempts to enter a flower for a second time. The anther is adhering to the stigma. Dendrobium secundum Lindl. Fig. 4.—Longitudinal median section through the flower: p, petal. The rest of the lettering as in Fig. 1. Fig. 5.—Column and portion of sac of a young flower showing narrow stigma. Fig. 6.—Column and portion of sac of an older flower with the stigma widened out; the anther has gone. SCIENT. PROC. R.DUBL. SOC.,N.S., Vou. XII. PLATE VI. West, Newman lith. i Ties ran Waris Hak ts Tee oe aie aren) ote Sapte THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 6. APRIL, 1909. PRODUCTION OF AMMONIA FROM ANTE EO SIP Tel I TR ILC) WY AD Les) (Es DN BY HERMAN C. WOLTERECK, Pu.D. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN : PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price Sixpence. =) SNE K ~ / 7) \ OCh AiG. 3 ational om 7 Viusev becattka tn I hea Kerr—The Pollination of certain Species of Dendrobium. 53 pollinia are caught by the stiff posterior border of the rostellum, which scrapes them off the style and retains them in the stigmatic cavity. In this species, then, self-pollination is prevented, as the pollinia are removed from a flower before its stigma has opened out sufficiently to admit them. A reference to Pl. VI., figs. 4, 5, and 6, will better explain the structure of the flower in this species. Dendrobium stuposum Lindl.—The filament is not very elastic; but the anther on dislocation falls over the entrance to the passage. Three out of five self-pollinated flowers produced capsules. Dendrobium incurvum Lindl. (?).—On dislocation the anther falls forward ; but the filament is not long enough to allow of the anther being pushed over the stigma; however, the dislocated anther would probably prevent access to the stigma unless it was knocked off the filament. Five self-pollinated flowers produced no capsules. Dendrobium bellatulum Rolfe.—The filament is slightly elastic; but the dislocated anther is with difficulty shoved on to the stigma. The passage is very large, and only very partially blocked by the dislocated anther. Two self-pollinated flowers produced capsules. Dendrobium citiatum Parish.—The filament is very short. he anther, on dislocation, does not block the passage; any pressure from the front only pushes it back into its original position. Although in this species there is no mechanical contrivance to prevent a flower being pollinated with its own pollen, my experiments show that self-pollination is not effective. Sixteen flowers which were self-pollinated produced no capsules, two cross- pollinated flowers producing good capsules. SOIENT. PROG. R.D.S., VOL. XII., NO. Y. L ae i) VI. PRODUCTION OF AMMONIA FROM ATMOSPHERIC NITROGEN. By HERMAN C. WOLTERECK, Pu.D. [COMMUNICATED BY PROFESSOR HUGH RYAN, M.A., D.Sc. | Read Decemper 22, 1908. Ordered for Publication Feprvary 15. Published Apri 3, 1909. TuIs process was developed by the investigation of certain observations made in the course of experiments for the synthetical production of hydrocyanic acid, which were carried out by Professor Eschweiler and myself in 1900 at the Technische Hochschule, Hanover. In the course of our observations it was found that ammonia is always formed when a dry mixture of hydrogen and nitrogen is passed over so-called reduced iron suspended on asbestos fibre, and heated to a temperature approaching dark-red heat (about 500° C.). This forms an excellent lecture experiment, as it requires little preparation and time, and gives a very strong reaction with nesslerized water a few minutes after starting the experiment. Sir William Ramsay, in October, 1901, was kind enough to verify this observation, and produced 11 mg. of ammonia in his laboratory by this method. However, this reaction is fugitive, and the formation of ammonia stops after a certain time. On modifying these experiments, and starting with iron oxide, over which a mixture of air and coal-gas was passed, larger quantities of ammonia were obtained. his led to a series of experiments with different metallic oxides. It was found that under the above conditions ammonia was produced by the oxides of the following metals :—nickel, cobalt, copper, cadmium, silver, lead, bismuth, chromium, and iron—the last three giving the best results. The experiments were generally arranged in such a manner that the oxide was supported by asbestos fibre in a combustion-tube which was heated to dull red heat. Illuminating gas and air in about equal quantities, after having been passed through dilute sulphuric acid previously tested for nitrogen compounds, were passed over the oxide, which in most cases became incandescent. Water was condensed and collected in a U tube arranged behind the combustion tube, and gave a distinct alkaline reaction after several hours’ duration of the experiment. The use of glass combustion-tubes proved very inconvenient on account of their frequent breakage by the condensed Wo.rereck—Production of Ammonia from Atmospheric Nitrogen. 55 water, and a 32-inch iron barrel was substituted therefor. It was found, after many experiments, that the presence of an increasing quantity of water- vapour was an advantage, and the gases were passed through distilled water at about 80° C. before entering the tube. I have since been able to repeat and confirm these experiments with a quartz tube. From the experiments carried out the following are typical—A ?3-inch iron barrel filled with crystalline iron oxide was heated to a dull red heat; the quantity of gases passed through was—air 1:81 cubic feet, and illuminat- ing gas 0-92 cubic feet (proportion 2:1); the total ammonia produced was 544 mg. Oxidized wire was next substituted for iron oxide and the quantity of illuminating gas reduced step by step until 5:2 cubic feet of air and 0:066 (proportion 78:1) of illuminating gas produced 1122mg. ammonia. This led to experiments in which illuminating gas or free hydrogen was omitted altogether, an experiment being made under the following conditions :— A 13-inch iron barrel with a T piece fitted into it for insertion of a pyro- meter, so that the exact temperature could be determined, was filled with freshly reduced iron wire gauze; and over this air, carefully freed from impurities and mixed with steam, was passed. A more exact description of the apparatus used in the following quanti- tative experiments described will show that no precaution was omitted to obtain indisputable results. The general arrangement of the apparatus in all the quantitative experi- ments referred to was as follows :— ‘ The air was taken through an experimental wet meter (either pressure or suction may be used) giving readings in v4 cubic foot per revolution, and having four dials to allow readings to be taken up to 1000 cubic feet to allow for continuous work. The water filling the meter was carefully tested for ammonia. The gases from the meter passed a precision-valve, by means of which the flow of the same could be exactly regulated to prevent any change in the amount of air passing the apparatus in any given time after having once been regulated. The air passed next through a wash-bottle containing concentrated sulphuric acid which had been proved to be free from any nitrogen compounds by careful tests. From this wash-bottle the air passed through a flask con- taining a measured quantity of distilled water, free from nitrogen compounds, kept in a continuous level water-bath at a temperature of 80°C. From this flask, the air, now saturated with water-vapour, passed into the iron barrel placed in a combustion furnace, and containing the rolls of iron gauze or other 56 Scientific Proceedings, Royal Dublin Society. material referred to later on. The iron barrel, of a diameter of 14 inch, con- sisted of two six-inch lengths connected in the middle by a T piece, which allowed the insertion of a thermometer or pyrometer, the bulb of which could be placed in the exact centre of the iron barrel. This iron gauze or other material (peat, &c.) was arranged in equal quantities on either side of the thermometer bulb. The thermometer was held in position in the T piece by a special preparation of asbestos fibre and plaster-of-Paris, which had also been carefully tested, and proved to contain no nitrogen compounds. The iron barrel was provided at both ends with reducing pieces to 4 inch, allowing a short length of }-inch iron tube to be screwed into the end admitting the air from the flask The gases were passed through a Liebig condenser, and through dilute sulphuric acid in a special absorption apparatus. Allresults were duplicated for comparison and determined by distillation with sodium carbonate, absorption in 745 n. HCl, and titration with ,4> n. Na,CO3. To determine the most favourable temperature a series of experiments was made under the following conditions :— A 13-inch iron barrel was filled with freshly reduced iron wire gauze, and air saturated with steam at 80° C. was passed over it at a uniform rate and under identical conditions, except for variations in temperature. The duration of each experiment was 43 hours; and the amount of air passed was 43 cubic feet, the contact material being reduced after each test for several hours by means of carbon monoxide. The yields in milligrams at the different temperatures were as follows :— 250-300° a0 80°3 mg. 400-450° ee 73:4 me. 300-350° .. 204:0 mg. 450-550° Ae 41-7 mg. 350-400° oo JETS) 0) says, 550-650° an 23°6 mg. By comparing the results of this series of experiments, it was found that under the various tested conditions the most favourable temperature was between 300 and 350° C. To exclude ammonia, which might be obtained from the decomposition of iron nitride, the iron tubes used in these experiments were treated for several days with steam and carbon monoxide alternately, until the quantity of ammonia formed per hour did not exceed 0:01 mg. Experiments were then made to determine whether it would be possible to lengthen the zone of contact; and an iron barrel 6 feet long and 13 inches in diameter was filled with wire gauze, and placed in a lead bath at a temperature of 350° C., air and steam being passed through it under the same conditions as in the former series. Wo ttereck— Production of Ammonia from Atmospherie Nitrogen. 57 The yields of ammonia obtained in six separate experiments compared with those at the corresponding temperature of the former series with a 6-inch zone of wire gauze were only between one-seventh and one-tenth of the former; and therefore it is to be concluded that ammonia is redecomposed by the prolonged contact with theiron. An observation which was made in the course of the series of experiments described showed that, as the oxidation of the iron proceeded, the formation of ammonia decreased, and that it was necessary to reduce the iron from time to time by passing carbon monoxide or hydrogen over it at a higher temperature. Thisrather complicated matters ; and experi- ments were made to discover a material the oxidation of which should produce favourable results, while being so cheap as not to require a subsequent reduction. Such materials were found in coke, charcoal, brown coal, peat— in fact, any carbonaceous material. Experiments were made by passing air and steam over coke, 80 grams of which yielded in 47 hours 17948 gram of ammonia, equal to 8:9 per cent. of sulphate of ammonia on the quantity of coke consumed: the amount of air passed over was 31 cubic feet. On repeating the experiment at a slightly higher temperature with 78 grams of coke, 213 cubic feet of air produced in 25 hours 1:443 gram of ammonia, being 7°3 per cent. of ammonium sulphate on the coke consumed. ‘This result shows that, with the higher temperature in a shorter time and with less air, very little less ammonia is produced than with the lower temperature, in spite of the fact that the heating was, in the first place, carried on for a longer time. The coke experiments closely approach the field so thoroughly investigated by Beilby and Young, and by Dr. Ludwig Mond ; but, however nearly these investigators approached the actual synthesis of ammonia from the nitrogen of the atmosphere, they did not achieve it, since the principal object aimed at by them was the production of gas suitable for heating or for power purposes, and ammonia was only considered as a by-product. The temperature employed by Beilby and Young was about 1000°C. and the average composition of the gas obtained by them was :— Per Cent. CON: : : : 8-1 CH,, . P 6 : 2°3 | 39 per cent. combustible gases. H, : : : . 286 CO,, 6 . 5 . 16:6 N, : ; : . 44-4 100:0 58 Scientific Proceedings, Royal Dublin Society. Dr. Mond obtained his results at a slightly lower temperature, stated by him asa dull red heat, and corresponding to about 800°C. The average composition of the gases obtained by him he stated to be :— Per Cent. OO; , : ; 10 H, j ‘ : ‘ 23 |} 36 per cent. combustible gases. Hydrocarbons, . : 3 Owing to the low temperature employed in my process, no combustible gases can be produced, since the reaction between carbon and steam only begins at and above 550° C.1 On examining the gases leaving the apparatus in my laboratory experi- ments, the entire absence of hydrogen and carbon monoxide could be shown; and on experimenting on the largest scale only maximum quantities under 4 per cent. of hydrogen and carbon monoxide could be detected. These were entirely due to local superheating, since the temperature of reaction was always kept below that at which steam may be decomposed by carbon (550° C.) or carbon dioxide reduced to carbon monoxide (600°C.). The gases produced contained generally about 18 per cent. of carbon dioxide. The time required in the coke experiments already mentioned appeared very long; but by substituting peat for coke it was found that the pro- duction of considerable quantities of ammonia was possible in a very much shorter time. ; 40 grams of peat containing 26:2 per cent. of moisture and 1°54 per cent. nitrogen, calculated on absolutely dry peat, were treated with air and steam, and left 11 grams of ash. ‘The quantity of ammonia pro- duced was 370 mg., equal to 8 per cent. of sulphate of ammonia on the peat substance consumed. A great many experiments were then made with a horizontal iron retort, such as is employed in coal-testing; and it was found that average yields of 10 per cent. of sulphate of ammonia were obtained on the quantity of peat consumed. As it was found that in most cases the ammonia produced greatly exceeded the quantity that could be accounted for by the nitrogen in the peat, the conclusion was of necessity arrived at that part at least of the ammonia must have been produced synthe- tically, induced by the oxidation of the carbon obtained from the peat. To determine with absolute certainty the question whether the oxidation of carbon under these conditions, as was the case with iron, would induce the formation of ammonia, pure sugar carbon was used, thus excluding the possibility of any nitrogen being present in its combined state. ‘ Dammer, ‘‘ Handbuch der Anorganischen Chemie,’’ 1892, vol. 1, p. 365, line 19. Woxtereck—Production of Ammonia from Atmospheric Nitrogen. 59 Quantity of air per hour, C.c. of water condensed, Duration of experiment in hours, . " Z Quantity of carbon con- sumed, : Ammonia obtained in mg. Percentage of ammonia calculated on consumed carbon, ‘ 2. 3. 4. 5. 6. TEMPERATURE, 700-720°C. | 600-610°C. | 550°C. | 380-392°C. | 450°. 450° 50 L. 50 L 50 L 50 L. 40 L 18 L 520 410 510 540 700 300 42 42 42 4h 6 6 39 g. 34 2, 29 g. 62g. 115 g 8 ¢. 15°5 18-0 27°65 40-0 102°5 1850 [cent. [cent.| [cent. “04 per cent. | 0°5 percent. | -09 per | -66 percent. | ‘9 per | 2°3 per The above results clearly prove that nitrogen is capable of chemical reaction if present during the moist oxidation of certain metals, or of carbon within certain limits of temperature. x att at THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XID. (N.S.), No. 7. APRIL, 1909. NOTE ON THE TENSILE STRENGTH OF WATER. BY HENRY H. DIXON, Sc.D. F.BS., PROFESSOR OF BOTANY IN THE UNIVERSITY OF DUBLIN. [Authors alone are responsib/e for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. “KS MSUty pe . . am <\\ f Price Stxpence./ ES eon 4: 1911 “tional Museu™ Doe me Ate nk Worrrreck— Production of Ammonia from Atmospheric Nitrogen. 59 1. 2. 3. a: 5. 6. TEMPERATURE, 700—720°C. | 600-610°C. | 550°C. | 380-392°C. | 450°. 450°. Quantity of air per hour, 50 L. 50 L. 50 L. 60 L. 40 L. 18 L. C.c. of water condensed, 520 410 510 540 700 300 Duration of experiment in hours, . c : 43. 4} 4} 43 6 6 Quantity of carbon con- TTC Hes aa ae ile 39 g. 34 g. 29 g. 6g. 1l'5g.| 8¢. Ammonia obtained in mg. 15:5 18-0 27°5 40:0 102°5 | 185-0 Percentage of ammonia calculated on consumed [cent. [cent.) [cent. carbon, : 5 - | °04 per cent. | 0°5 percent. | -09 per | ‘66 percent. | -9 per | 2°3 per The above results clearly prove that nitrogen is capable of chemical reaction if present during the moist oxidation of certain metals, or of carbon within certain limits of temperature. SCIENT. PROC. R.D.S., VOL. XII., NO, VI, M (CGO eay WUT NOTE ON THE TENSILE STRENGTH OF WATER. By HENRY H. DIXON, &c.D., F.B.S., Professor of Botany in the University of Dublin. Read January 26. Ordered for Publication Frpruary 9. Published Apri 5, 1909. Donny,! in 1846, found that it was possible for a column of sulphuric acid 1:255 m. long to hang in a vertical glass tube sealed at the upper end, even when the atmospheric pressure was removed from the lower open end. He explained this phenomenon as due to the adhesion of the sulphuric acid to the . glass and to the cohesion of the liquid itself, He compares this behaviour of the sulphuric acid to the well-known experience that the mercury in a barometer is retained above the actual barometric height, if the tube, completely filled by inclining it, is gradually raised to a vertical position. He further points out that this phenomenon has been explained by Laplace, as due in a similar manner to the adhesion of the mercury to the glass and to its own cohesion. Donny points out that when one withdraws a plane dise from contact with a surface of water, the cohesion of the latter does not come into play ; but the column of water connecting the disc with the liquid below at first grows gradually thinner until, at a moment when the dise has been raised to a certain height above the general level of the lower liquid, the column spon- taneously draws in from the edges of the disc, and when its diameter becomes extremely small, breaks in two. He also shows that, in a tensile liquid column a bubble, sufficiently small to have surface-tension forces capable of supporting the hydrostatic head of the liquid below, will not destroy the tensile state. He, however, failed to demonstrate the cohesion of water by the same method which had been successful in the case of sulphuric acid. But he points out that the cohesion of water may be observed in another way. A straight glass tube sealed at both ends, partly filled with water, and enclosing some air, is supported vertically in one hand of the experi- menter, while he vigorously strikes the lower end with the palm of the other F. Donny, Sur la cohésion des liquides, et sur leur adhérence aux corps solides. Ann. de Chimie et de Physique, sér. 3, tome xvi., 1846, pp. 167 ef seg. Drxon—Wote on the Tensile Strength of Water. 61 hand. At the moment of the shock, bubbles open in the water and close immediately with a metallic click. When a tube, exactly similar to the last, but exhausted of air, is treated in the same way, no bubbles are developed, nor is the click heard. Donny explains that, in the first case, the blow causes minute bubbles in the liquid to be opened against their surface-tension forces and the air-pressure in the tube, and that therefore, in the second case, where the bubbles are not formed, the cohesion must be greater than these two forces taken together. Donny believed that even a very little air dissolved in the water suffices to reduce the tenacity of water to a vanishingly small figure. ; He also points out that the boiling of liquids is retarded when air is removed from them, owing to their increased cohesion ; and it is the cohesion being suddenly overcome which causes explosive boiling. Berthelot,! a few years afterwards, succeeded in showing very simply that water has a very considerable cohesion, and, under proper conditions, can sustain a very great tensile stress. His experiment has been so often misquoted that it may be well to quote his own description of his method :— “Si Von remplit d’eau a la température de 28 ou 30 degrés un tube capillaire un peu fort, fermé par un bout et terminé de l’autre par une pointe effilée ; si Von refroidit ce tube jusqu’a 18 degrés, de fagon a y faire rentrer une certaine quantité d’air par la pointe ouverte; si alors on le ferme et qu’on chauffe de nouveau jusqu’a 28 degrés, et graduellement au-dessus, au bout de certain temps lair se dissout complétement. Si l’on refroidit 4 18 degrés température initiale a laquelle le tube renfermait a la fois du gaz et du liquide, on remarque que l’eau continue a occuper la totalité de la capacité intérieure, et conserve ainsi une densité invariable de 28 a 18 degrés.... “La variation de densité ainsi produite est énorme: pour l’eau elle est égale a 75> deson volume a 18 degrés ... Unsemblable effet, pour se produire en sens contraire, exigerait une pression d’environ 50 atmosphéres pour l'eau.” Dr. Joly and the author? also carried out some experiments on the behaviour of stressed water in presence of air. These experiments were designed to find out if water adhered to the conducting tubes of plants suffi- ciently vigorously to transmit the stress needed to raise the transpiration stream. Also we were misled by erroneous quotations from Berthelot’s paper into believing that this investigator had only experimented with air-free water. As was anticipated, we found that water containing large quantities of air in 1M. Berthelot, Sur quelques phénoménes de dilatation forcée des liquides. Ann. de Chimie et de Physique, xxx., 1850, pp. 232 et seq. cane 2 Dixon and Joly, On the Ascent of Sap. Phil. Trans., Roy. Soc., vol. 186 (1895), B. 62 Scientific Proceedings, Royal Dublin Society. solution and in contact with wet wood, was stable while sustaining consider- able stresses. In our experiments the tension developed as in Berthelot’s by the con- traction of the cooling liquid was measured by the deformation of the containing vessel, which was consequently made sufficiently large, and with walls of sufficient elastic yield, to render its collapse appreciable. The change in volume due to the pressure of one atmosphere was determined ; and so the tensions producing the observed change could be estimated. We found that the adhesion to the glass and to the conducting tracts of plants and the water’s own cohesion must be greater than 7°5 atm. More recently the author! showed that the water in a cell, distended by osmotic pressure, could also sustain a considerable tension. Having had occasion recently to look up Berthelot’s paper, I repeated his experiment on water enclosed in thick capillary tubes; and as some of my results give a much higher minor limit for tensile strength of water than his, I have thought it of interest to record them here. The dimensions of the tubes used were as follows :— No. Length. Bore. Thickness of Wall. I. 22 em. 1:0 mm. 2-0 mm. II. 22 cm. 1:0 m. 2:0 mm. Iii. 19 cm. 1:0 mm. 3:0 mm. IV. 19 cm. 1:0 mm. 3:0 mm. Wo 14°5 cm. 1:0 mm. 3:0 mm. VI. 15 cm. 1:0 mm. 3:0 mm. VII. 15°5 cm. 1:0 mm. 3:0 mm. VIII. 17°5 cm. 1:0 mm. 3:0 mm. In each case the tubes were first cleaned with a solution of caustic potash, which was afterwards removed by repeated rinsing with boiled, distilled water. “ General View of the Agriculture, &c., of the Islands on the Coast of Normandy.” SCIENT. PROC. R.D.S., VOL. XII., NO. VIII. oO 74 Scientific Proceedings, Royal Dublin Society. it should do so is not surprising, since whole white produces hybrids— roans—with at least two of these colours, red and black. The conclusions we arrive at are these :— (1) There are four colours forming the basis of present-day Highland colours, viz., black, blackish-brown or donn, red, and light dun. One other colour, white, and other “markings” have been absorbed from time to time ; but these have been almost entirely bred out. The reds may be of several shades, but there are not sufficient data to separate them. (2) Black is the dominant of red. (8) Black produces dun hybrids—registered “dun,” “dark dun,” &¢.— when mated with light dun. (4) Donn or blackish-brown produces brindle hybrids when mated with black, red, and light dun. (5) Red produces yellow hybrids when mated with light dun. The above conclusions can be shown by the following diagram. The fundamental colours are underlined. The hybrids have arrows leading towards them from their parent colours :— ea. aS Ge ORO J ee brindle brindle. BROWN BLACK RED LIGHT DUN black ‘See dite As already stated, these conclusions are not put forward as absolute certainties, but as conclusions for which some further confirmation is not undesirable. It is hoped that breeders of Highlanders, Longhorns, and Jerseys may give closer attention to shades and markings when registering their stock, and so we shall acquire more accurate data. Meantime, for any who may wish to criticize our conclusions or look into the subject more closely, we place in parallel columns the results to be expected from the different matings according to our conclusions and the results as recorded in the Herd Book. As already explained, discrepancies must be looked for chiefly in connexion with brindles and reds, although other colours are not altogether free from them. Witson—The Colours of Highland Cattle. 75 Taste [V.—Colours expected and recorded in volume 11. - CoLours oF CALVES | Contours or Cavers EXPECTED. RECORDED. Contours or PareEnvs. 4 : ¢ E : 4 “ | E : = | FLY |S ES ee as Dhak Bed, ERE x BHA 2 ollse| ses Oia Blowh w Bh Rd, Ree xR 2 alse lisa is 101 101 | 18 | 10 | 9 Black x Brindle, Bag x Bex, - -[x|x1x|x|x/ 81/53/40] 4] 3 Black x Yellow, fee XP. -|x|x|x|x|x| 20/12] 2 22/18 Black X Dun, Tue x ue : a || $< S21 SK SS BE OH a yh at | By Red xX Red, B x a | x 4 {118 2/11 1 Rel sein sc BER, lls | x|xlx 7 (126 | 43/11] 1 Red x Yellow, B yx ® x x 6| 56/14] 43] 2 Red x Dun, B x ue x ae | - NO | 22 © |) BR | Brindle X Brindle, BOF x 7B © ol x|xix|x|x| 10] 28} 40] 2] 0 Brindle x Yellow, B8F xP | |x) x|x|x| x] 3] 24] 25] 54] 5 Brindle x Dun, 888 yxBU | x |x| x |x| x | 16] 8 {19 | 16 | 28 | Yellow x Yellow, B x H 6 : x | x1 >< | 2 We 0 | 46] 11 Yellow X Dun, B 4 LB A 5) SK | x|~x Wis i) By OO) |) 883 Dun XX Dun, Te x Te i : x 1 0 1 | 14 1 There are browns or donns registered as blacks. 76 Scientifie Proceedings, Royal Dublin Society. Tasix V.—Colours expected and recorded in volumes Xxill., XIv., and Xv. Cotours oF CALVES Contours or CALVES EXPECTED. RECORDED. GORCOSS or Parents. - E 2 | b 2 a hs E : 5 & =I | = 8 | 3 Black x Black, 1 >< || 4) al ti} Bi @} a 0 Black x Red, “x1x 1x 16/10} 8] 6 0| 0 Black < Brindle, S< 1 Seb Se SS SK AN | Gy | Nes Bi) al |) oO Black < Yellow, SS Sea SS te Seal Se 18 5 8 | 19 | 16 0 Black < Dun, < Sel] SSP Se 0} 0 0 1 3] 0 Black x Light Dun, x |X| xX QO! 21) OF 2} 8} oO Red x Red, |x 0}80/14} 9} 0} O Red x Brindle, x |X|] XX) Xx 6 151] 85|}15| 0) 0 Red x Yellow, x ~< AN ad |) Qt | G2) 2] i Red x Dun, x x i] Si a] ai) aio Red Xx Light Dun, x ©} al} Ol} i} of @ Brindle x Brindle, x || oS OS SKK PK 5 | 36 | 60 3 4 0 Brindle x Yellow, So SS 4 SIP Se Se li Se DN Sy HB) i 7 5 0 Brindle x Dun, x x | SS) SK 2) 5 5] 8 5 0 Brindle x Light Dun, << SS 1 OS I SS ISS Se 0 1 0} 138 5 0 Yellow X< Yellow, x x x 1 15 5 | 40) 5] 8 Yellow x Dun, x oS [E2S I OS 1 0 0 4 2 3 Yellow x Light Dun, x << OO} O}] O} 2) 2] 2 Dun x Dun, x S< || Se 0 0 0 il 0 0 Dun x Light Dun, S< || Se 0 0; 0 0; 0} 0 Light Dun X Light Dun, x OF OF] OW} LU} O| Oo | | The writer of this paper is indebted to Mr. Duncan Shaw, W.S., Inver- ness, Secretary of the Highland Cattle Society, for the loan of a set of the “ Highland Herd-Book”’; to Mr. D. G. Lillie, of St. John’s College, Cambridge, for having suggested that there might be two kinds of blacks ; and specially, to Mr. Donald Fletcher, so long in charge of the Kinnaird herd, for much valuable information, and for having supplied the tufts of hair reproduced in the accompanying plate. PLATE VIL. SCIENT. PROC.R.DUBL.SOC,,N_.S., Vou. XII. West, Newman chr. DuN. SILVER DUN. LIGHT DUN. sete ih Paar eat ¢ aa ‘ Pay Aes THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 9. APRIL, 1909. ON A PROPOSED ANALYTICAL MACHINE. BY IPBIKOW 18, JOIWIDG AMS. [Authors alone are responsible forall opinions expressed in their Communications. | DUBLIN : PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSLER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. AE ee aes nan insti;, ~ Jo Xw°w“s0 Stity Ry) UF 2 aX (Olea Ze eve } Price Sixpence. No. “onal Musew 7 Sowa wan VAP Low J IDS ON A PROPOSED ANALYTICAL MACHINE. By PERCY EH. LUDGATE. (COMMUNICATEL BY PROFESSOR A. W. CONWAY, M.A.) {Read Frpruary 23. Ordered for Publication Marcu 9. Published Aprin 28, 1909.] I purpose to give in this paper a short account of the result of about six years’ work, undertaken by me with the object of designing machinery capable of performing calculations, however intricate or laborious, without the immediate guidance of the human intellect. In the first place I desire to record my indebtedness to Professor CO. V. Boys, F.x.s., for the assistance which I owe to his kindness in entering into correspondence with me on the matter to which this paper is devoted. It would be difficult and very inadvisable to write on the present subject without referring to the remarkable work of Charles Babbage, who, having first invented two Difference Engines, subsequently (about eighty years ago) designed an Analytical Engine, which was shown to be at least a theoretical possibility ; but unfortunately its construction had not proceeded far when its inventor died. Since Babbage’s time his Analytical Engine seems to have been almost forgotten; and it is probable that no living person understands the details of its projected mechanism. My own knowledge of Babbage’s Engines is slight, and for the most part limited to that of their mathematical principles. The following definitions of an Analytical Engine, written by Babbage’s contemporaries, describe its essential functions as viewed from different standpoints :— “A machine to give us the same control over the executive which we have hitherto only possessed over the legislative department of mathematics.” “‘The material expression of any indefinite function of any degree of generality and complexity, such as, for instance :—F (e, y, 2, log x, sin y, &c.), which is, it will be observed, a function of all other possible functions of any number of quantities.” 1C, Babbage: ‘‘ Passages from the Life of a Philosopher,’’ p. 129. 2 R. Taylor’s ‘Scientific Memoirs,’’ 1848, vol. iil., p. 691. SCIENT, PROG., R.D.S., VOL. XII., NO. Ix. P 18 Scientific Proceedings, Royal Dublin Society. “An embodying of the science of operations constructed with peculiar reference to abstract number as the subject of those operations.” «‘ A machine for weaving algebraical patterns.’ These four statements show clearly that an Analytical Machine ‘“ does not occupy common ground with mere ‘calculating machines.’ It holds a position wholly its own.” In order to prevent misconception, I must state that my work was not based on Babbage’s results—indeed, until after the completion of the first design of my machine, I had no knowledge of his prior efforts in the same direction. On the other hand, I have since been greatly assisted in the more advanced stages of the problem by, and have received valuable suggestions from, the writings of that accomplished scholar. There is in some respects a great resemblance between Babbage’s Analytical Engine and the machine which I have designed—a resemblance which is not, in my opinion, due wholly to chance, but in a great measure to the nature of the investigations, © which tend to lead to those conclusions on which the resemblance depends. This resemblance is almost entirely confined to the more general, abstract, or mathematical side of the question; while the contrast between the proposed structure of the two projected machines could scarcely be more marked. It is unnecessary for me to prove the possibility of designing a machine capable of automatically solving all problems which can be solved by numbers. ‘The principles on which an Analytical Machine may rest “have been examined, admitted, recorded, and demonstrated.’”* J would refer those who desire information thereon to the Countess of Lovelace’s translation of an article on Babbage’s Engine, which, together with copious notes by the translator, appears in R. Taylor's “Scientific Memoirs,” vol. iii.; to Babbage’s own work, “ Passages from the Life of a Philosopher”; and to the Report of the British Association for the year 1878, p. 92. These papers furnish a complete demonstration that the whole of the developments and operations of analysis are capable of being executed by machinery. Notwithstanding the complete and masterly treatment of the question to be found in the papers mentioned, it will be necessary for me briefly to outline the principles on which an Analytical Machine is based, in order that my subsequent remarks may be understood. An Analytical Machine must have some means of storing the numerical data of the problem to be solved, and the figures produced at each successive TR. Taylor's ‘‘ Scientific Memoirs,’’ 1848, vol. iii., p. 694. * loc. cit., p. 696. 5 C. Babbage: ‘‘ Passages from the Life of a Philosopher,”’ p. 450. LupeatE— On a Proposed Analytical Machine. 79 step of the work (together with the proper algebraical signs); and, lastly, a means of recording the result or results. It must be capable of submitting any two of the numbers stored to the arithmetical operation of addition, subtraction, multiplication, or division. It must also be able to select from the numbers it contains the proper numbers to be operated on; to determine the nature of the operation to which they are to be submitted ; and to dispose of the result of the operation, so that such result can be recalled by the machine and further operated on, should the terms of the problem require it. The sequence of operations, the numbers (considered as abstract quantities only) submitted to those operations, and the disposition of the result of each operation, depend upon the algebraical statement of the caleulation on which the machine is engaged; while the magnitude of the numbers inyolved in the work varies with the numerical data of that particular case of the general formula which is in process of solution. ‘The question therefore naturally arises as to how a machine can be made to follow a particular law of development as expressed by an algebraic formula. An eminently satisfactory answer to that question (and one utilized by both Babbage and myself) is suggested by the Jacquard loom, in which interesting invention a system of perforated cards is used to direct the movements of the warp and weft threads, so as to produce in the woven material the pattern intended by the designer. It is not difficult to imagine that a similar arrangement of cards could be used in a mathematical machine to direct the weaving of numbers, as it were, into algebraic patterns, in which case the cards in question would constitute a kind of mathematical notation. It must be distinctly understood that, if a set of such cards were once prepared in accordance with a specified formula, it would possess all the generality of algebra, and include an infinite number of particular cases. I have prepared many drawings of the machine and its parts; but it is not possible in a short paper to go into any detail as to the mechanism by means of which elaborate formule can be evaluated, as the subject is necessarily extensive and somewhat complicated; and I must, therefore, confine myself to a superficial description, touching only points of particular interest or importance. Babbage’s Jacquard-system and mine differ considerably; for, while Babbage designed two sets of cards—one set to govern the operations, and the other set to select the numbers to be operated on—I use one sheet or roll of perforated paper (which, in principle, exactly corresponds to a set of Jacquard-cards) to perform both these functions in the order and manner necessary to solve the formula to which the particular paper is assigned. To 80 Scientific Proceedings, Royal Dublin Society. such a paper I apply the term formula-paper. Each row of perforations across the formula-paper directs the machine in some definite step in the process of calculation—such as, for instance, a complete multiplication, including the selection of the numbers to be multiplied together. Of course a single formula-paper can be used for an indefinite number of calculations, provided that they are all of one type or kind (7.e. algebraically identical). In referring to the numbers stored in the machine, the difficulty arises as to whether we refer to them as mere numbers in the restricted arithmetical sense, or as quantities, which, though always expressed in numerals, are capable of practically infinite variation. In the latter case they may be regarded as true mathematical variables. It was Babbage’s custom (and one which I shall adopt) when referring to them in this sense to use the term “ Variable” (spelt with capital V), while applying the usual meanings to the words “number” and “ variable.” In my machine each Variable is stored in a separate shuttle, the individual figures of the Variable being represented by the relative positions of protruding metal rods or “ type,’ which each shuttle carries. There is one of these rods for every figure of the Variable, and one to indicate the sign of the Variable. Hach rod protrudes a distance of from 1 to 10 units, according to the figure or sign which it is at the time representing. ‘The shuttles are stored in two co-axial cylindrical shuttle- boxes, which are divided for the purpose into compartments parallel to their axis. The present design of the machine provides for the storage of 192 Variables of twenty figures each ; but both the number of Variables and the number of figures in each Variable may, if desired, be greatly increased. It may be observed, too, that the shuttles are quite inde- pendent of the machine, so that new shuttles, representing new Variables, can be introduced at any time. When two Variables are to be multiplied together, the corresponding shuttles are brought to a certain system of slides called the index, by means of which the machine computes the product. It is impossible precisely to describe the mechanism of the index without drawings; but it may be compared to a slide-rule on which the usual markings are replaced by movable blades. The index is arranged so as to give several readings simultaneously. The numerical values of the readings are indicated by periodic displacements of the blades mentioned, the duration of which displacements are recorded in units measured by the driving shaft on a train of wheels called. the mi//, which performs the carrying of tens, and indicates the final product. The product can be transferred from thence to any shuttle, or to two shuttles simultaneously, provided that they do not LupeatE— On a Proposed Analytical Machine. 81 belong to the same shuttle-box. The act of inscribing a new value in a shuttle automatically cancels any previous value that the shuttle may have contained. ‘T'he fundamental action of the machine may be said to be the multiplying together of the numbers contained in any two shuttles, and the inscribing of the product in one or two shuttles. It may be mentioned here that the fundamental process of Babbage’s Engine was not multiplication but addition. Though the index is analogous to the slide-rule, it is not divided logarithmically, but in accordance with certain idee numbers, which, after some difficulty, I have arranged for the purpose. I originally intended to use the logarithmic method, but found that some of the resulting intervals were too large; while the fact that a logarithm of zero does not exist is, for my purpose, an additional disadvantage. The index numbers (which I believe to be the smallest whole numbers that will give the required results) are contained in the following tables :— [TABLEs, 82 Scientific Proceedings, Royal Dublin Society. Tasie |. Unit. Simple Index No. Ordinal. 0 50 9 1 0 0 2 1 1 3 7 4 4 2 2 5 23 7 6 8 5 7 398 8 8 3 3 9 14 6 TABLE 2. Partial Comp. Partial Comp. Partial Comp. product. Index No. | product. Index No. product. Index No. 1 0 16 30 36 16 2 1 16 4 40 26 3 7 18 15 42 41 4 2 20 25 45 37 5 23 21 40 48 11 6 8 | 24 10 49 66 U 33 | 25 46 54 22 8 3 27 21 56 36 9 14 28 35 63 47 10 x i BD 31 64 6 12 9 32 5 72 We 14 34 35 56 81 28 I Comp. index numbers of zero :—50, 51, 52, 53, 57, 58, 64, 73, 83, 100. LupGatE—On a Proposed Analytical Machine. 83 Comp. Index No. Partial product. Comp. ndex No. Partial product. cs ou ma NOOK WwW WD HS OS wNowwp wv WwW bY NY RB Be eS Se eS ee NO rR wWNRe SO OTM oH # & bt = 28 qo ww co c bw wonwnre Oo © ~ | 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 56 56 57 58 59 60 oe 14 an 28 Col. 1 of Table 1 contains zero and the nine digits, and col. 2 of the same Table the corresponding simple index numbers. all partial products (a term applied to the product of any two units), while Col. 1 of Table 2 sets forth 84 Scientific Proceedings, Royal Dublin Society. col. 2 contains the corresponding compound index numbers. The relation between the index numbers is such that the sum of the simple index numbers of any two units is equal to the compound index number of their product. Table 3 is really a re-arrangement of Table 2, the numbers 0 to 66 (representing 67 divisions on the index) being placed in col. 1, and in col. 2, opposite to each number in col. 1 which is a compound index number, is placed the corresponding simple product. Now, to take a very simple example, suppose the machine is supplied with a formula-paper designed to cause it to evaluate x for given values of a, 6, c, and d, in the equation ab + cd = x, and suppose we wish to find the value of w in the particular case where «a = 9247, b = 8182, e = 21893, and d = 823. The four given numbers are first transferred to the machine by the key-board hereafter mentioned; and the formula-paper causes them to be inscribed in four shuttles. As the shuttles of the inner and outer co-axial — shuttle-boxes are numbered consecutively, we may suppose the given values of a and ¢ to be inscribed in the first and second shuttles respectively of the inner box, and of 6 and d in the first and second shuttles respectively of the outer box; but it is important to remember that it is a function of the formula-paper to select the shuttles to receive the Variables, as well as the shuttles to be operated on, so that (except under certain special circumstances, which arise only in more complicated formule) any given formula-paper always selects the same shuttles in the same sequence and manner, whatever be the values of the Variables. ‘The magnitude of a Variable only effects the type carried by its shuttle, and in no way influences the movements of the shuttle as a whole. The machine, guided by the formula-paper, now causes the shuttle- boxes to rotate until the first shuttles of both inner and outer boxes come opposite to a shuttle-race. The two shuttles are then drawn along the race to a position near the index; and certain slides are released, which move forward until stopped by striking the type carried by the outer shuttle. The slides in question will then have moved distances corresponding to the simple index numbers of the corresponding digits of the Variables 0. In the particular case under consideration, the first four slides will therefore move 3, 0, 7, and 1 units respectively, the remainder of the slides indicating zero by moving 50 units (see Table 1). Another slide moves in the opposite direction until stopped by the first type of the inner shuttle, making a movement proportional to the simple index number of the first digit of the multiplier a—in this case 14. As the index is attached to the last- mentioned slide, and partakes of its motion, the relative displacements of Lupeate—On a Proposed Analytical Machine. 85 the index and each of the four slides are respectively 3 + 14, 0 + 14, 7+ 14, and 1 + 14 units (that is 17, 14, 21, and 15 units), so that pointers attached to the four slides, which normally point to zero on the index, will now point respectively to the 17th, 14th, 21st, and 15th divisions of the index. Consulting Table 3, we find that these divisions correspond to the partial products 72, 9,27, and 18. In the index the partial products are expressed mechanically by movable blades placed at the intervals shown in column 2 of the third table. Now, the duration of the first movement of any blade is as the unit figure of the partial product which it represents, so that the movements of the blades concerned in the present case will be as the numbers 2, 9, 7, and 8, which movements are conveyed by the pointers to the mill, causing it to register the number 2978. A carriage near the index now moves one step to effect multiplication by 10, and then the blades partake of a second movement, this time transferring the tens’ figures of the partial products (i.e. 7, 0, 2, and 1) to the mill, which completes the addition of the units’ and tens’ figures thus— 2978 7021 73188 —the result being the product of the multiplicand 6 by the first digit of the multiplier a. After this the index makes a rapid reciprocating movement, bringing its slide into contact with the second type of the inner shuttle (which represents the figure 2 in the quantity @), and the process just described is repeated for this and the subsequent figures of the multiplier a until the whole product ab is found. The shuttles are afterwards replaced in the shuttle-boxes, the latter being then rotated until the second shuttles of both boxes are opposite to the shuttle-race. These shuttles are brought to the index, as in the former case, and the product of their Variables (21893 x 828) is obtained, which, being added to the previous product (that product having been purposely retained in the mill), gives the required value of z It may be mentioned that the position of the decimal point in a product is determined by special mechanism which is independent of both mill and index. Most, of the movements mentioned above, as well as many others, are derived froma set of cams placed on a common shaft parallel to the driving-shaft ; and all movements so derived are under the control of the formula-paper. - The ordinals in Table 1 are not mathematically important, but refer to SCIENT. PROC., R.D.S, VOL. XII., NO. IX. Q 86 Scientific Proceedings, Royal Dublin Society. special mechanism which cannot be described in this paper, and are included in the tables merely to render them complete. The sum of two products is obtained by retaining the first product in the mill until the second product is found—the mill will then indicate their sum. By reversing the direction of rotation of the mill before the second product is obtained, the difference of the products results. Conse- quently, by making the multiplier unity in each case, simple addition and subtraction may be performed. Tn designing a calculating machine it is a matter of peculiar difficulty and of great importance to provide for the expeditious carrying of tens. In most machines the carryings are performed in rapid succession; but Babbage invented an apparatus (of which I have been unable to ascertain the details) by means of which the machine could “‘ foresee” the carryings and act on the foresight. After several years’ work on the problem, I have devised a method in which the carrying is practically in complete mechanical independence of the adding process, so that the two movements proceed simultaneously. By my method the sum of m numbers of » figures would take 9m +n units of time. In finding the product of two numbers of twenty figures each, forty additions are required (the units’ and tens’ figures of the partial products being added separately). Substituting the values 40 and 20 for m and », we get 9 x 40 + 20 = 380, or 93 time-units for each addition—the time-unit being the period required to move a figure-wheel through {5 revolution. With Variables of 20 figures each the quantity » has a constant value of 20, which is the number of units of time required by the machine to execute any carrying which has not been performed at the conclusion of an indefinite number of additions. Now, if the carryings were performed in succession, the time required could not be less than 9 + 2, or 29 units for each addition, and is, in practice, considerably greater.’ In ordinary calculating machines division is accomplished by repeated subtractions of the divisor from the dividend. The divisor is subtracted from the figures of the dividend representing the higher powers of ten until the remainder is less than the divisor. ‘he divisor is then moved one place to the right, and the subtraction proceeds as before. ‘The number of subtractions performed in each case denotes the corresponding figure of the quotient. This is a very simple and convenient method for ordinary calculating machines ; but it scarcely meets the requirements of an Analytical Machine. At the same time, it must be observed that Babbage used this method, but found it gave rise to many mechanical complications. 1 For further notes on the problem of the carrying of tens, see C, Babbage: ‘‘ Passages from the Life of a Philosopher,’’ p. 114, &e, Lupeate—On a Proposed Analytical Machine. 87 My method of dividing is based on quite different principles, and to explain it I must assume that the machine can multiply, add, or subtract any of its Variables ; or, in other words, that a formula-paper can be prepared which could direct the machine to evaluate any specified function (which does not contain the sign of division or its equivalent) for given values of its variables. Suppose, then, we wish to find the value of P for particular values of p and g which have been communicated to the machine. Let the first three figures of ¢ be represented by /, and let A be the reciprocal of f, where A is expressed as a decimal of 20 figures. Multiplying the numerator and : A; : denominator of the fraction by A, we have Te where Aq must give anumber of the form 100... because Ag = 200 placing the decimal point after the f Pp unit, we have unity plus a small decimal. Represent this decimal by z: then— a foal or Ap(1 +a) Expanding by the binomial theorem— (1) a Ap (1 -«#+ 2-2 + wt — a + &.), or (2) Fa Ap (1-2) (1 + 2”) (1 +24) (1 + 2°), &e. The series (1) converges rapidly, and by finding the sum as far as «*° we obtain the correct result to at least twenty figures; whilst the expression (2 gives the result correctly to at least thirty figures. The position of the decimal point in the quotient is determined independently of these formule. As the quantity A must be the reciprocal of one of the numbers 100 to 999, it has 900 possible values. ‘The machine must, therefore, have the power of selecting the proper value for the quantity A, and of applying that value in accordance with the formula. For this purpose the 900 values of A are stored in a cylinder—the individual figures being indicated by holes of from one to nine units deep in its periphery. When division is to be performed, this cylinder is rotated, by a simple device, until the number A (represented on the cylinder by a row of holes), which is the reciprocal of the first three figures of the divisor, comes opposite to a set of rods. These rods then transfer that number to the proper shuttle, whence it becomes an ordinary Variable, and is used in accordance with the formula. It is not necessary that every time the process of division is required the dividing formula 88 Scientific Proceedings, Royal Dublin Society. should be worked out in detail in the formula-paper. To obviate the necessity of so doing the machine is provided with a special permanent dividing cylinder, on which this formula is represented in the proper notation of per- forations. When the arrangement of perforations on the formula-paper indicates that division is to be performed, and the Variables which are to constitute divisor and dividend, the formula-paper then allows the dividing cylinder to usurp its functions until that cylinder has caused the machine to complete the division. It will be observed that, in order to carry out the process of division, the machine is provided with a small table of numbers (the numbers A) which it is able to consult and apply in the proper way. I have extended thiy system to the logarithmic series, in order to give to that series a considerabiv convergency; and I have also introduced a logarithmic cylinder which has the power of working out the logarithmic formula, just as the dividing cylinder directs the dividing process. This system of auxiliary cylinders and tables for special formule may be indefinitely extended. The machine prints all results, and, if required, the data, and any note- worthy values which may transpire during the calculation. It may be mentioned, too, that the machine may be caused to calculate and print, quite automatically, a table of values—-such, for instance, as a table of logs, sines, squares, &c. It has also the power of recording its results by a system of per- forations on a sheet of paper, so that when such a number-paper (as it may be called) is replaced in the machine, the latter can “read” the numbers indicated thereon, and inscribe them in the shuttles reserved for the purpose. Among other powers with which the machine is endowed is that of changing from one formula to another as desired, or in accordance with a given mathematical law. It follows that the machine need never be idle ; for it can be set to tabulate successive values of any function, while the work of the tabulation can be suspended at any time to allow of the determination by it of one or more results of greater importance or urgency. It can also “feel” for particular events in the progress of its work——such, for instance, as a change of sign in the value of a function, or its approach to zero or infinity ; and it can make any pre-arranged change in its procedure, when any such event occurs. Babbage dwells on these and similar points, and explains their bearing on the automatic solution (by approximation) of an equation of the nt» degree ;! but I have not been able to ascertain whether his way of attaining these results has or has not any resemblance to my method of so doing. 1 (©. Babbage: ‘‘ Passages from the Life of a Philosopher,” p. 131. Lupeate—On a Proposed Analytical Machine. 89 The Analytical Machine is under the control of two key-boards, and in this respect differs from Babbage’s Engine. The upper key-board has ten keys (numbered 0 to 9), and is a means by which numbers are communicated to the machine. It can therefore undertake the work of the number-paper previously mentioned. The lower key-board can be used to control the working of the machine, in which case it performs the work of a formula- paper. The key-boards are intended for use when the nature of the calculation does not warrant the preparation of a formula-paper or a number-paper, or when their use is not convenient. An interesting illus- tration of the use of the lower key-board is furnished by a case in which a person is desirous of solving a number of triangles (say) of which he knows the dimensions of the sides, but has not the requisite formula- paper for the purpose. His best plan is to put a plain sheet of paper in the controlling apparatus, and on communicating to the machine the known dimensions of one of the triangles by means of the upper key-board, to guide the machine by means of the lower key-board to solve the triangle in accordance with the usual rule. The manipulations of the lower key-board will be recorded on the paper, which can then be used as a formula-paper to cause the machine automatically to solve the remaining triangles. He can communicate to the machine the dimensions of these triangles individually by means of the upper key-board ; or he may, if he prefers so doing, tabulate the dimensions in a number- paper, from which the machine will read them of its own accord. The machine is therefore able to “remember,” as it were, a mathematical rule ; and having once been shown how to perform a certain calculation, it can perform any similar calculation automatically so long as the same paper remains in the machine. It must be clearly understood that the machine is designed to be quite automatic in its action, so that a person almost entirely ignorant of mathematics could use it, in some respects, as successfully as the ablest mathematician. Suppose such a person desired to calculate the cosine of an angle, he obtains the correct result by inserting the formula-paper bearing the correct label, depressing the proper number-keys in succession to indicate the magnitude of the angle, and starting the machine, though he may be quite unaware of the definition, nature, or properties of a cosine. While the machine is in use its central shaft must be maintained at an approximately uniform rate of rotation—a small motor might be used for this purpose. It is calculated that a velocity of three revolutions per second would be safe ; and such a velocity would ensure the multiplication of any 90 Scientific Proceedings, Royal Dublin Society. two Variables of twenty figures each in about 10 seconds, and their addition or subtraction in about 3 seconds. The time taken to divide one Variable by another depends on the degree of convergency of the series derived from the divisor, but 1} minutes may be taken as the probable maximum. When constructing a formula-paper, due regard should there- fore be had to the relatively long time required to accomplish the routine of division ; and it will, no doubt, be found advisable to use this process as sparingly as possible. The determination of the logarithm of any number would take 2 minutes, while the evaluation of a” (for any value of n) by the expotential theorem, should not require more than 1} minutes longer— all results being of twenty figures.’ The machine, as at present designed, would be about 26 inches long, 24 inches broad, and 20 inches high; and it would therefore be of a portable size. Of the exact dimensions of Babbage’s Engine I have no information; but evidently it was to have been a ponderous piece of machinery, measuring many feet in each direction. The relatively large size of this engine is doubtless due partly to its being designed to accommodate the large number of one thousand Variables of fifty figures each, but more especially to the fact that the Variables were to have been stored on columns of wheels, which, while of considerable bulk in themselves, necessitated somewhat intricate gearing arrangements to control their move- ments. Again, Babbage’s method of multiplying by repeated additions, and of dividing by repeated subtractions, though from a mathematical point of view very simple, gave rise to very many mechanical compli- cations.” To explain the power and scope of an Analytical Machine or Engine, I cannot do better than quote the words of the Countess of Lovelace: “There is no finite line of demarcation which limits the powers of the Analytical Engine. These powers are coextensive with the knowledge of the laws of analysis itself, and need be bounded only by our acquaintance with the latter. Indeed, we may consider the engine as the material and mechanical representative of analysis, and that our actual working powers in this department of human study will be enabled more effectually than heretofore to keep pace with our theoretical knowledge of its principles and laws, through the complete control which the engine gives us over the executive manipulations of algebraical and numerical symbols.’* A Committee of the British Association which was appointed to report 1 The times given include that required for the selection of the Variables to be operated on. * See Report Brit. Assoc., 1878, p. 100. 3 R. Taylor's ‘‘ Scientific Memous,’’ 1843, vol. iii., p. 696. Lupeatr—On a Proposed Analytical Machine. 91 on Babbage’s Engine stated that, ‘‘ apart from the question of its saving labour in operations now possible, we think the existence of such an instrument would place within reach much which, if not actually impossible, has been too close to the limits of human skill and endurance to be practically available.” In conclusion, I would observe that of the very numerous branches of pure and applied science which are dependent for their development, record, or application on the dominant science of mathematics, there is not one of which the progress would not be accelerated, and the pursuit would not be facilitated, by the complete command over the numerical interpretation of abstract mathematical expressions, and the relief from the time-consuming drudgery of computation, which the scientist would secure through the existence of machinery capable of performing the most tedious and complex calculations with expedition, automatism, and precision. 1 Report Brit. Assoc., 1878, p. 101. Ft id We ee BUPA CU ORT EP Rieke Seer THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 10. APRIL, 1909. THE TAXINE IN IRISH YEW. BY RICHARD J. MOSS, F.I.C., F.C.S. [Authors alone are responsib/e for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. oan NSU? (63 anor {if Price Sixpence. 1 A soli Perens C ? f NS VV. : “onal Museu 7 ian ae feelevehnpa Lupeatre-—On a Proposed Analytical Machine. 91 on Babbage’s Engine stated that, “ apart from the question of its saving labour in operations now possible, we think the existence of such an instrument would place within reach much which, if not actually impossible, has been too close to the limits of human skill and endurance to be practically available.’ In conclusion, I would observe that of the very numerous branches of pure and applied science which are dependent for their development, record, or application on the dominant science of mathematics, there is not one of which the progress would not be accelerated, and the pursuit would not be facilitated, by the complete command over the numerical interpretation of abstract mathematical expressions, and the relief from the time-consuming drudgery of computation, which the scientist would secure through the existence of machinery capable of performing the most tedious and complex calculations with expedition, automatism, and precision. 1 Report Brit. Assoc., 1878, p. 101. SOIENT. PROO., R.D.S. VOL. XII., NO. IX. R Cod X. THE TAXINE IN IRISH YEW, TAXUS BACCATA var. FASTIGIATA. By RICHARD J. MOSS, F.LC., F.C.S. Read Frsruary 23. Ordered for Publication Marcu 9. Published Aprin 380, 1909. Tuoucu the poisonous nature of Yew has been known for many centuries, it is only quite recently that the extraction of the supposed poisonous principle taxine has been carefully studied. In a paper on taxine, by IT’. E. Thorpe, c.s., £.R.s., and George Stubbs," the literature of the subject is reviewed ; and a simple method for the extraction of taxine from the leaves is described. The authors show that taxine readily undergoes change when warmed on the water-bath with dilute acid. All the earlier investigations involving the evaporation of acid-solutions in the water-bath must have yielded erroneous results. In the method of Thorpe and Stubbs the air-dried leaves are powdered, and then extracted at the ordinary temperature with 1 per cent. sulphuric acid. The extract is rendered alkaline with ammonia, and shaken with ether. The ethereal solution is then shaken with dilute hydrochloric acid; and this operation is repeated until a colourless ethereal solution is obtained. From this the taxine is finally extracted with dilute hydrochloric acid, and precipitated by ammonia. ‘I'he taxine is next dissolved in ether, the ether is evaporated, and the residue dried over sulphuric acid and weighed. The quantity of alkaloid obtained in this way corresponded to 0°18 per cent. of the green leaves of the male tree. Other experiments led to the separa- tion of 0-12 per cent. of the green leaves of the female tree. As no particular variety of Yew is mentioned, I presume the kind used was the common Yew. I have applied the method to the leaves and fruit of the Ivish Yew, known to botanists as the Florence Court Yew, Taxus baccata var. fastigiata; and, as I obtained a much larger yield of taxine, I thought the results worth placing on record. ‘Journal of the Chemical Society, Transactions, 1902, p. 874. Moss—The Taxine in Trish Veuw. 93 T'axine when precipitated from its solution in dilute acid is at first in a stale of extremely fine division. In a few minutes, if the quantity present is not too small, the particles aggregate, and flocculent masses are formed, which are very easily filtered and washed. In my experiments I washed this precipitate on a weighed filter until the washings, when acidulated with nitric acid, gave either no turbidity, or only a faint turbidity with silver nitrate. ‘he precipitate was then dried in vacuo over sulphuric acid, until the weight was constant. In this respect the method I employed differed from that of Thorpe and Stubbs. A quantity of leaves taken on November 29th, 1908, from a female tree growing at Ballybrack, County Dublin, weighed, in the fresh state and detached from the twigs, 790 grammes. The leaves were dried at room- temperature, powdered, and extracted as above. The precipitate weighed 4-7093 grammes; this corresponds to 0°596 per cent. of the original leaves. Thinking that more accurate results could be obtained with a much smaller quantity of leaves, I gathered 150 grammes of leaves from the same tree on January 11th, 1909; and taking every precaution to ensure complete extraction, I obtained taxine weighing 09340 gramme, corresponding to 0-623 per cent. Seeds from the fruit of this tree treated in the same way, except that they were ground to a fine pulp without drying, gave from 45:75 grammes of seeds, taxine weighing 0:0363 gramme, or 0:079 per cent. The leaves of another female tree gathered on December 6th gave taxine corresponding to 0323 percent. This tree had produced a very abundant crop of berries. I found in the seeds 0-082 per cent. of taxine—very nearly the same proportion as in the other tree, while the leaves contained only about half the quantity of taxine. ‘he arillus or fleshy part of the berries is quite free from taxine. I failed to obtain any indication of the alkaloidal substance, operating upon 3800 grammes of the fleshy part. This is what one might expect, considering the sweet taste of this part of the fruit, and the complete absence of the bitter taste so characteristic of the leaves, bark, and seeds. Ala The residual pulp of the seeds, from which the taxine had been extracted by dilute sulphuric acid, was dried and extracted with ether in a Soxlet extractor. On evaporating the ether a greenish-coloured oil was left, corre- sponding to 11°25 per cent. of the original seeds. ‘This oil has a barely perceptible bitter taste, probably due to a trace of taxine. The leaves of a male specimen of common Yew growing in the same place, but in a position much shaded in summer by large trees, gave a much smaller quantity of taxine than the two specimens of Irish Yew. ‘The precipitate 94 Scientific Proceedings, Royal Dublin Society. from 150 grammes of leaves gathered on January 17th, 1909, weighed only 0:1236 gramme, corresponding to 0-082 per cent. This is less than half the quantity found by Thorpe and Stubbs in the leaves they examined. I believe these estimations are fairly comparable with one another, and that they may be accepted as showing a wide variation in the quantity of alkaloidal substance in the leaves of different trees. Carnevin' points out that the young leaves of a light green colour are not nearly so poisonous as the older dark green leaves; accordingly I took some trouble to ensure that the leaves gathered fairly represented the foliage of each tree, and did not include an undue proportion of either old or young leaves. ‘The number of determinations is too small to warrant the final conclusion that the variety known as Irish Yew contains much more taxine than common Yew, but that is certainly the direction in which the results point. This is quite in accor- dance with an opinion which I find has long been entertained. For example, in the Pharmaceutical Journal of 1877,? there is a reference to letters in the Field (1 have been unable to find the originals in that journal), in one of which a correspondent states that in the case of pheasants poisoned by Yew he has observed that the leaves found in the crops and gizzards of the birds are always Irish Yew; and he believes they may eat common Yew with impunity. The very marked variation in the quantity of alkaloidal substance in the leaves of the plants also explains in some degree the widely different opinions expressed from time to time as to the toxic properties of Yew. I have not had an opportunity of examining the seeds of common Yew ; but from the similarity of the results obtained with the seeds of different specimens of Irish Yew it seems probable that the alkaloidal substance is more constant in quantity in the seeds than in the leaves. Many of the recorded cases of poisoning by the seeds of Yew are presumably attributable to the Irish variety, as the trees are referred to as growing in cemeteries, and it has long been the custom to plant the Irish Yew in such places. I have assumed that the alkaloidal substance referred to is the toxic principle of the Yew; but it is right to point out that this is by no means certain. Experimental work in this connexion is very desirable. Taxine and its compounds are non-crystalline; and they exhibit no distinctive reaction. It is therefore not easy to ensure that the alkaloidal substance obtained from different sources is the same thing. The melting-point of taxine is not clearly marked. I find that when the solid obtained by evaporating the ethereal solution of taxine, and drying in 1 «« Des Plantes Véneéneuses et des Empoisonnements qu’elles déterminent,”’ p. 48. 2 Ser. 3, vol. vii., p. 697. Moss—The Taxine in Irish Yew. 95 vacuo over sulphuric acid, is placed between microscope cover-glasses, held together by a clip, it is possible, under a low magnifying power, to detect indications of softening, at a temperature of 60°, twenty-two degrees below the reputed melting-point. Thorpe and Stubbs obtained a chloride, a sulphate, and a compound with gold chloride, having each a composition closely corresponding to that indicated by theory. I find that taxine dissolved in ether may be titrated with = hydrochloric acid in the manner sometimes used for the titration of alkaloids, using a minute quantity of methyl orange as indicator.’ T'axine weighing 0:2790 gramme took 0:0159 gramme hydrochloric acid, instead of 00151 indicated by theory on the assumption that the hydrochloride is CsH;.0.N, HCl. Thorpe and Stubbs point out that the alkaloid is decomposed when warmed in the water-bath with dilute hydrochloric acid, one of the products being a brown substance. I find that, even without warming, some decompo- sition takes place. When 0°6026 gramme of taxine was dissolved in 0°3 c.c. of hydrochloric acid diluted with 2c. of water, and placed in a vacuum over caustic potash, the residue was of a dark-brown colour. On prolonged drying over sulphuric acid, it was found that the increase in the weight of the taxine was 0:0400 gramme, instead of 0:0328 required theoretically for the formation of the hydrochloride. On treating the solid obtained in this way with water, most of it dissolved, leaving a brown humus-like substance. On adding ammonia to the solution, a brown precipitate is thrown down; and on shaking with ether, and allowing the liquids to separate, they are both coloured brown. When the ethereal solution was shaken with dilute hydro- chloric acid, and ammonia added to the acid liquid, the precipitate was again brown ; and on standing in the liquid for a few days, it aggregated to oily- looking brownish spheres about a millimetre in diameter. It is clear that under the conditions described hydrochloric acid has a marked action upon taxine at ordinary temperatures. It is doubtful whether the use of a mineral acid, even highly dilute, is admissible in the extraction of the alka- loid. In an experiment on the extraction of the alkaloid, by means of a 1 per cent. solution of oxalic acid, instead of dilute sulphuric acid, acting on the leaves; and using dilute oxalic acid, instead of dilute hydrochloric acid, to remove taxine from its ethereal solution, I obtained a final precipitate corresponding to 0:402 per cent. of taxine, against 0°323 per cent. when mineral acids were employed. 1 «¢ Commercial Organic Analysis,’’? A. H. Allen, vol. iii., part ii., page 131, foot-note. 96 Scientific Proceedings, Royal Dublin Society. As I have pointed out, taxine when first precipitated is in an extremely finely divided state. The flocculation that takes place in the precipitate is due to the growth of minute spherical masses. These spherical masses grow as if they were crystals; in the course of two months, at ordinary room tempera- ture, some of the spheres attained a diameter of 0:01 mm. Examined in polarized light these spheres show no indication of double refraction. When taxine is suspended in water, and carbon dioxide passed through it, some taxine is dissolved. On filtering and allowing the clear liquid to stand, exposed to the air, a pellicle forms on the surface and a precipitate separates. No trace of crystalline structure can be detected in either the pellicle or the precipitate. When a solution of taxine in ether is cooled by means of liquid air, a white cloud forms consisting of very finely divided matter, and the cloud increases as the temperature falls; but at no time is there any appearance of crystallization. Several other attempts to induce crystallization by methods in common use were equally unsuccessful. THE SCIENTIFIC PROCEEDINGS Or THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 11. MAY, 1909. ON PHOTOGRAPHY BY REFLECTION UNDER CONTACT. BY K. EH. FOURNIER pv’ ALBE, B.Sc., A.R.C.Sc., M.R.IA. (PLATE VIII.) {Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price Sixpence. os OCT 4 1911 N 5, ae, onal Musev- A sssonlall tj? Auiise Nal ate \ GON _ Ae he or eae es ar. = hy eee Mo ON PHOTOGRAPHY BY REFLECTION UNDER CONTACT. By E. EK, FOURNIER v’ALBE, B.Sc., A.R.C.Sc., M.R.T.A. [Published May 10, 1909.] (Prats VIII.) IN the usual methods of contact photography, a copy is taken of the original or negative by allowing light to pass through the latter on to a sensitive surface. The resulting picture is due to differences in opacity in the various points of the original or negative. The new method to be described consists in transmitting the light in the reverse direction, and producing a picture, not by differences of opacity, but by differences of reflecting power in the original. The obvious objection to such a method is that the sensitive film, being exposed to a uniform incident illumination coming through the back of the plate, will be uniformly “fogged”; and the resulting positive will he marred by a brightness which invades and partly obliterates all the dark portions. If this difficulty can be overcome, we obtain a method of copying any flat picture or design without a camera; and we avoid the difficulties of distortion, curvature of field, chromatic and spherical aberration, flare, astigmatism, and lack of uniformity of illumination, which beset all but the best lenses, and which cannot, in practice, be simultaneously reduced to a negligible amount. When the original to be copied has no half-tones, it is possible, by suitable exposure and development, to eliminate the fog entirely. The general principle is to employ exposures and developers which in ordinary photography “suppress the detail in the shadows,” or, in other words, confine the developed image to those portions which have received the maximum illumination. In copying a line-drawing, a page of print, or similar full- 1 Such a reflection process was devised in 1897 by J. H. Player (see ‘“‘The Photographic Journal,” 1897, p. 222). He used bromide paper, and transmitted the light through a green glass. He ‘‘could not succeed with plates.”” It is said (bid.) that positive copies of maps were for some time made in the French army by contact with Ilford process plates, but no details are given.—E. E. F. 98 Scientific Proceedings, Royal Dublin Society. toned matter in this manner, the areas of maximum illumination are those which are illuminated by the incident light p/us the light reflected by the white paper. The areas in contact with the black portions of the original are only illuminated by the incident light plus the small proportion reflected by the black ink. ‘he incident light greatly exceeds the reflected light in amount; and the difference in illumination relied upon to produce the necessary contrast does not, in most cases, exceed 5 per cent. With this smaJl margin of additional illumination it is found possible, however, to obtain a contrast in the negative amounting to as much as 40 or 50 per cent. Once this is obtained, the negative may in turn be used to produce a transparency in which the contrast is further increased by a second application of the same treatment. Workable lantern-slides are obtained in this manner (PI. VIIL,, fig. 5); but much better results, comparable to the best lantern-slides obtained with the camera, are secured by two additional reversals, to which the same principle is applied (PI. VIII., figs. 4 and 6). Instead of making a number of successive reversals, the fog may be eliminated by reduction and subsequent intensification. Howard Farmer’s reducer (potassium ferricyanide and hyposulphite of soda) dissolves away the fog more than the full-tone if sufficiently concentrated. The negative is intensified with mercuric chloride and silver nitrate. The best results are obtained with slow plates of the “ photomechanical ”’ class, and the developer used was the following :— No. | solution. Hydroquinone, j : . 80 grains. Potassium metabisulphite, . 120 grains. Potassium bromide, o . 10 grains. Water, &., .. : : . 10 ounces. No. 2 solution. Caustic potash, : : . 200 grains. Water, &, . : : . 10 ounces. Equal parts of both solutions are mixed. The best fixing agent is potassium cyanide, on account of its solving action on thinly deposited silver. But I used the ordinary ‘“ hypo” bath. Other Applications of the Process.—Excellent paper negatives are obtained with rapid bromide papers, and also with gas-light papers... On printing positives from them in the ordinary way, the grain of the paper negative 1 See foot-note, p. 97. Fournter bp’ ALBE— On Photography by Reflection under Contact. 99 disappears, as it is automatically compensated by the grain within the paper which gave rise to it. Printing-out papers also give negatives which can be used for printing positives on bromide paper without previous toning and fixing. But the exposure has to be very long. Bichromated gelatine plates give a good relief by this method, and, unlike the usual printing with bichromate, the more insoluble portion of the gelatine is nearest the support, whether glass or paper. Direct positives may be obtained either by over-exposure (fifty times), or, better, treating ordinary plates or papers with a 10 per cent solution of potassium bichromate, and exposing them for the period usual in bichromate printing. ‘I'he action in this case is as follows:—the bichromated gelatine which receives the reflected light is rendered more insoluble than that which adjoins the black areas. Only the latter, therefore, absorb the developer; and the whole film being ‘‘ fogged,” a direct image of the black portions results. The Active Agent.—The very striking results obtained with bromide paper are attributable to the double absorption undergone by the incident light in penetrating the developed film and returning to the eye of the observer. There appears to be in reality nothing but a purely optical effect. The suggestion was somewhat obvious that some radiation from the pigment of the original, usually, perhaps, intercepted by a slight thickness of air, might inhibit the reduction of the silver salts. This, however, is rendered very improbable by the following observations :— 1. On impregnating paper with sugar, glycerine, mercuric chloride, uranium nitrate, and various other substances which do not produce a visible effect, it is found that paper so treated has the same photographic action as the original paper. 2. The gradation of colours is found to be the same as that obtained with the camera. 3. Red has no special, inhibiting action. On applying this process to a deep red paper printed with black lettering, a faint negative was obtained, in which the black printed whiter than the red. Had red light exerted any inhibiting action, the fog should have been reduced, and the red should have appeared whiter in the negative. In applying the process described to black and white originals, certain advantages are gained over the ordinary methods withthe camera : 1. The reproduction is of the exact size of the original ; 2. The sharpness of definition is only limited by the size of the silver grain in the plate ; 100 Scientific Proceedings, Royal Dublin Society. 3. All differences in the angle of reflection of light by the original are avoided, all the effective light emerging at right angles to the surface. This last circumstance places at our disposal a delicate and accurate method of comparing the reflecting powers of surfaces for vertical incidence. It consists in mounting the surfaces in the same plane side by side, and exposing a plate withits film in contact with both. If the surfaces are large enough, a slight gap may be left between the sensitive film and the surfaces to be tested, so as to eliminate photographically active radiations or emanations from the surfaces. SCIENT. PROC. RB, DUBL. SOC., N.S., Vou. XIL. PLATE VIII. Fic. 2.—Second Negative. Itc. 4.—Second Positive. ” Ay Intso Book Ornament, ‘ Tue Enp,’’ py Facan. Fie. 5.—‘* The Seven Grades of Bards.’’ Fic. 6.—An Mlustration from ‘“ Andersen’s Fairy Tales.’’ Reproduced from the Book of Ballymote. By Helen Stratton. Cy. <<) 0 acrarar Cy) st ias Sime eraeeed THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 12. MAY, 1909. MECHANICAL STRESS AND MAGNETISATION OF IRON. Part I. BY WILLIAM BROWN, B.Sc. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price One Shilling. o Ly /9, y . OCT 4 1911 ) Vee. Po om, og ie 3 os [ tol 4 XII. MECHANICAL STRESS AND MAGNETISATION OF IRON. Part I. BY WILLIAM BROWN, B.Sc. [Read Fesruary 23. Ordered for Publication Manco 9. Published May 28, 1909.] Since the year 1837, when Professor G. Wiedemann observed the effects of torsion on the magnetisation of iron, a good deal of work has been done on the effects of mechanical stress on the magnetism of the magnetic metals, especially iron and nickel. An excellent report of the work done up to the year 1900, with references to the original papers, has been prepared by Professor H. Nagaoka,! in which the theoretical and experimental results so far obtained are summed up in two parallel columns containing sixteen statements. One of the columns is headed—* Deformations produced by magnetisation,” and the other—‘ The influence of mechanical stress on magnetisation.”’ Of the sixteen separate statements of results, the following six bear on the subject-matter of the present paper. 1. A wire magnetised longitudinally is twisted by circular magnetism. la. When a longitudinally magnetised wire is twisted, it becomes circularly magnetised. 2. A wire circularly magnetised is twisted when magnetised longi- tudinally. 2a. A circularly magnetised wire becomes longitudinally magnetised when twisted. 3. The torsion produced by longitudinally magnetising a wire of iron or nickel when magnetised circularly, attains a maximum in feeble magnetic fields. 3a. The transitory current due to the torsion of a wire of iron or nickel magnetised longitudinally, attains a maximum in a weak magnetic field. 1 Rapports du Congrés International de Physique, Paris, 1900, vol. ii., pp. 586-556. SCIENT. PROC. R.D.S. VOL. XI., NO. XII. i 102 Scientific Proceedings, Royal Dublin Society. These general statements are necessarily qualitative, and when taken in pairs as marked they are complementary to each other; for instance, in the first or 1, by the application of magnetic forces a mechanical result is obtained, and in la, a mechanical twist is applied with magnetic effects. The same holds with the other two pairs of statements. In what follows an attempt has been made to measure quantitatively the effects produced by varying the quantities above mentioned. Now it is well known that the transitory currents set up in a longi- tudinally magnetised iron wire are due to the circular magnetisations produced by twisting the wire; and Professor Ewing’ in 1883 showed how circular magnetisation exhibits hysteresis with regard to the angle of twist. It occurred to me some time ago that this might be a means of finding, among other things, a relation between the longitudinal stress per unit area on the wire and the circular magnetisation produced by the twist. As far as I know, this has not been directly touched upon in any of the papers on this subject, to the number of seventy or so, that have been pub- — lished before or since 1900. I have therefore in this paper brought together some results obtained with iron wires, showing the effects of keeping :— 1. The longitudinal magnetisation and the cross-sectional area of the wire constant, and varying the longitudinal stress per unit area. 2. The longitudinal magnetisation and the stress per unit area constant, and varying the cross-sectional area of the wire. 3. The longitudinal magnetisation and cross-sectional area of the wire constant, and varying the circular magnetisation and the longi- tudinal stress per unit area. 4. The longitudinal magnetisation and the stress per unit area constant, and varying the cross-sectional area of the wire and the circular magnetisation. The arrangement of the apparatus used is shown in fig. 1. BY is an iron wire suspended in the vertical component of the HKarth’s magnetic field. The top end B is fixed to an insulated support on a cross- beam in the ceiling of the room, and the lower end, about 9 feet below, dips into a mercury cup OC. On the lower end of the wire is fixed a non-magnetic cylinder or weight made of brass and lead, which has attached to it two light handles or pointers P, by means of which the required twist could be given to the vertical wire. 1 Proc. Roy. Soc., London, 1883, also ‘‘ Magnetic Induction in Iron and other Metals,” 3rd ed., p. 237. Brown—Mechunical Stress and Magnetisation of Iron. 103 The top of the wooden guide-frame F has a scale marked on it which is divided into spaces of 10 degrees, from 0 to 180 round each half. The wire BV under test could be put in cireuit—through the plug key K, —with a sensitive moving coil ballistic mirror galvanometer BG of the Ayrton and Mather type, with a coil of resistance 325 ohms. it | | Fig. 1. An earth-inductor £ could also be put in circuit with the galvanometer and wire, by opening the key XK, and closing K,; this was for the purpose of testing or checking the constant of the galvanometer during the progress of the work. At the beginning and end of every set of observations on the t 2 104 Scientific Proceedings, Royal Dublin Society. wire, K, was opened and XK, closed, and the earth-inductor turned through an angle of 90° in the horizontal component of the Harth’s magnetic field. The distance of the scale and light spot from the mirror of the galvanometer was 101°3 cms. throughout the experiments, and the constant of the galvanometer was measured in the usual way by means of the earth- inductor, and found to be practically 70 x 107° coulombs per scale-division, and 700 divisions on the scale equalled 44 cms. Before coming to the subject proper of this paper, I shall explain how an interesting experiment can be made with this simple arrangement, to show the presence of a transitory electric current due to circular magnetisation produced by twisting an iron wire, which is hanging or placed in a magnetic field. A No. 16 ordinary soft iron wire was suspended in the earth’s vertical field about 0°45 c.g.s. unit, and when the weight on the end of the wire was turned or twisted round the central axis of the wire, a momentary electric current was produced, giving a throw on the ballistic galvanometer. The transitory current thus obtained flowed down the wire when the weight was twisted from left to right with respect to an observer standing beside the wire. The galvanometer was temporarily taken out of circuit, and it was found by calculation and experiment that a length of 227 cms. of the wire, with a vibrator of about 6000 grammes at the lower end of it, had the same period of torsional vibration as the period of vibration of the galvanometer coil, viz., 3°69 secs. The galvanometer was now put in circuit, and a series of observations taken thus:—The vibrator was turned through an angle of, say, 45°, and allowed to vibrate freely round the axis of the wire. The current produced in the wire caused the galvanometer coil and the spot of light to swing in unison with the vibrator; then the 4th and 5th swings of the light spot were read off in every case, and the mean of the two taken. ‘lhe vibrator was now started from 90°, and so on up to two whole turns. The results obtained are shown in Table I. and as a curve in fig. 2. Deflections on Scale. Brown—Mechanical Stress and Magnetisation of Iron. Tasre I. Deflections on Galvanometer Scale. Guanes) Left. Right. Mean. 22-5° 4:8 4:2 4:5 45:0 13:0 13:0 13:0 90:0 52°5 52°5 52°5 135°0 86:0 86:0 | 86:0 180:0 101-0 101:0 101-0 225-0 112-0 113:0 112°5 270:0 117-0 118:°0 117°5 315:0 118°5 119°5 | 119-0 360°0 120°0 121:0 120°5 540:0 122:0 123-0 122°5 720:0 122°0 123°0 122°5 105 90 180 270 360 450 540 630 Amplitude of Vibrator in Degrees. Fie. 2. 106 Scientific Proceedings, Royal Dublin Society. An initial twist of 180° was just within the elastic limit of the wire, but beyond that angle the wire was permanently deformed as indicated by the curve becoming flat. Nagaoka’ has described an interesting experiment designed to show the Wiedemann effect to a large audience by means of capillary ripples on the surface of mercury; and important results have been obtained by Gray and Wood? on the effect of a magnetic field on the rate of subsidence of torsional oscillations in iron. In order to carry out experiments to test the effect of stress on the material as indicated in the first statement on page 102, another iron wire of the same gauge, No. 16, was taken and carefully prepared and annealed. One end of the wire was fixed to a support and stretched horizontally with a string over a pulley having a weight of about 7 kilos. attached. The flame of a good bunsen burner was then passed slowly over the wire, raising it to a bright red heat; by this means all the kinks were taken out of the wire. It was then hung loosely in a horizontal direction and the bunsen again passed slowly over it twice, which left the wire annealed or as soft as the nature of the material would permit. The wire was then cleaned with emery cloth and placed in position in the Earth’s vertical field, 0°45 ¢.g.s. units, as indicated in fig. 1; a vibrator or weight of a certain definite amount was attached, and the galvanometer put in circuit. Then by means of the handles P the wire was ultimately put through a complete cycle in steps of 20° each; at every 20° mark in the scale on the top of the guide-frame F’ a small hole was made to hold a stout pin which was moved along at each step and served as a stop or guide for the handle P. Thus the pin being in the 20° mark the handle P was moved from 0° to 20° and the throw on the galvanometer read; then the pin was moved into the 40° mark and the handle moved from 20° to 40°, and the throw on the galvanometer scale again read off, and so on right round the complete cycle. The throws or deflections on the galvanometer scale were read off at each of the nine steps from 0° to 180°, and from this point the wire was put through four complete cycles before reading the scale further—in order to bring the wire to a proper cyclic state. hat being done, the throws on the scale were again observed at each step from 180 to 0, and on the other side from 0 to 180, then back from 180 to 6, and finally from 0 to 180. ‘Then summing up the galvanometer throws or transient currents, we have a measure 1 Nature, vol. lxix., p. 487. 2 Proc. Roy. Soc., vol. Lxxiii., p. 286. Brown—WMechanical Stress and Magnetisation of Iron. 107 of the circular magnetisation produced by the torsion of the wire in a magnetic field. Table II. is an example of one of the sets of observations taken for a complete cycle when the load on the wire was 9:9 x 10* grammes per Sq.cm. TABLE II. Angle. Deflection. Sum. | Angle. Deflection. Sum. jal o | 0 1O 5:5 39°5 20 — 4:5 — 4:5 120 3:5 43-0 40 12-0 16°5 140 2°5 45:5 60 11:5 28-0 160 11 46-6 80 80 | 36-0 180 11 47-7 100 6-0 42:0 160 0:8 48-5 120 2-0 44:0 140 0-9 49-4 140 2-5 46°5 120 0:8 50-2 160 0-5 47-0 100 0 50°2 180 0-5 47-5 80 — 06 49-6 160 1:0 48-5. 60 1:8 47°8 140 0-7 49:2 40 2-6 45-2 120 0:8 50-0 20 3:6 41:6 100 0-2 50:2 0 80 33:6 80 +1:2 49-0 20 16°6 17-0 60 lek) POS 40 24:8 = 78 40 2-6 45-0 60 16°8 24°6 20 3:0 42-0 80 9-6 34-2 0 9-0 33:0 100 5:8 40:0 20 15°5 17-5 120 2:8 42°8 AQ. | OBO || a Gol 140 3:0 45°8 60 17-0 24:5 160 0-4 46:2 80 10-0 34:0 180 1:0 47-2 This was the best and most symmetrical cycle obtained in the whole series of ninety or so. In most of the other cases the axis of abscisse had to be shifted down from two to eight seale-divisions, so as to make the curve 108 Scientific Proceedings, Royal Dublin Society. symmetrical round both axes; this was no doubt due to the iron wire not being quite free from torsion at the beginning of the cycle. The load on the wire was then increased by putting on extra lead discs, and a complete cycle again obtained ; and this was repeated for eight different values of the load. The cyclic curves thus got were plotted on millimetre paper, and the complete area of each determined in sq. cms. by the method of counting the squares. These curves were plotted on a fairly large scale : on the axis of abscissa 1 cm. represented 20° twist, and on the axis of ordinates 4 cms. represented ten divisions on the galvanometer scale; the latter might have been made to represent coulombs, but for the purposes of comparison scale-divisions were used. 3 2 s S R HH Oo 8 z 5 b = s z ——4 s io) =| ° n g fo} & i 5 2 o c= @o (=) JO a oO 40 0 Angle of ‘I'wist. Fie. 3. In the sample half-curve given in fig. 3, however, on the axis of abscissee lem. still represents 20° of twist ; but on the axis of ordinates 2 cms. represent ten divisions on the galvanometer scale. This ratio was taken because it shows out the proportions of the curve better than the other scale. The results obtained for the eight different loads are given in Table III., and are also shown in the curve, fig. 4, in which the stress in grammes per sq. em, are abscissee, and the areas of the complete cyclic curves are ordinates, Brown—WMechanical Stress and Magnetisation of Iron. 109 Tase IIT. Stress in grammes Total areas of cyclic curyes per sq. cm. in sq. cms. 9-9 x 10! 132:0 14:3 39 114°3 175 29 106°4 20°6 99 99-0 OCT as 84:6 26°9 99 72°6 31:9 $s 67:0 37:2 99 65:1 Area of Cyclic Curve in sq. cms. Stress in Grammes per sq. cm. Fie. 4. This curve indicates that, as the load increases, the circular magnetisation in the wire decreases at a rapid and uniform rate up to a stress of 2:7 x 10° grammes per sq. cm., which is about the 4 part of 65 x 10° grammes per sq. SOIENT. PROC. R.D.S,, VOL. XII., NO. XII, U 110 Scientific Proceedings, Royal Dublin Society. em., the tenacity or breaking weight of iron. Beyond this point the rate of diminution is slower; and if the curve were continued, it would meet the axis in the figure approximately at the point 7°5 x 10°. It was found by a separate experiment that the elastic limit of this wire was reached with a load of 16°5 kilos, or about 8 x 10° grammes per sq.cm. When the curve is plotted on a larger scale, with both the axes starting from zero, a tangent drawn from that point on the axis of abscissee representing the tenacity of iron (i.e. 65 x 10°) just meets the curve at the point of flexure; whilst the straight part of the curve produced meets the axis of ordinates at the point 165. From this we might infer that maximum magnetisation would be obtained with a minimum weight, and vice versa; the straight part of the curve, however, is the only part we are certain about, and this shows that, as the stress per unit area increases from 10° to 2°5 x 10° grammes per sq. cm., or 260 per cent., the area of the cyclic curve or the circular magnetisation in the iron wire decreases from 130 to 80, or about 40 per cent. Referring again to fig.8: the whole eight cyclic curves obtained were plotted to the same scale as here used, and the distances on the curve marked mn, ob, and oa were measured. ‘The corresponding distances on an ordinary H-B cyclic curve for iron are what determine its shape and area, so here we get some interesting relations between these values on the circular magnetisa- tion curves when they are plotted against the load on the wire. These values are given in able [V., and shown in curves in fig. 5. Taste LV. Sinera fin prema Distances on curye, Fig. 3. eL aa oat mn Teen. bas Cpe)s3 IO 47-4 31°8 32°6 43: 18 43°5 28:0 315 17°5 A 40°5 25°8 29°8 2.026). 38°0 23°0 28°9 PSP ap 35:0 20:0 28:0 PPE) 0 32°5 17:0 26'8 BG) op 27°8 13°'8 25'8 BUH 9 26:0 12°3 26°5 The numbers in columns marked mn and od are deflections on the galvanometer scale where 2 cms. represent 10 divisions, and the numbers Brown—WMechamceal Stress and Magnetisation of Iron. 111 in column marked oa are angles of twist where 1 cm. represents 20°, as in fie. 3; but in the curves, fig. 5, four ems. are taken to represent 20° of twist, so as to avoid a second set of numbers on the axis of ordinates. ‘That is, the numbers on the axis of ordinates represent deflections on the galvanometer scale for the curves mn and 0b, and represent angles of twist in degrees for the curve oa. The mn and od curves are practically parallel up to a load of about 3 x 10° grammes persq. cm.; and as the load increases from about 10° grammes per sq. cm., these quantities decrease at a uniform rate—i.e. they diminish by about 36 per cent. when the load is increased three times. In the curve marked Stress Grammes per sq. cm. Fie. 5. oa, for the same increase in the load, the rate of decrease of the quantity represented is smaller, and the diminution about 20 per cent. In order to test the effect of increasing the cross-sectional area of the wire, and keeping the longitudinal magnetisation and the stress per unit area constant, six new sets of wires were obtained of standard gauge, Nos. 12, 14, 16, 18, 20, and 22. These wires were of the best Swedish charcoal iron, and were delivered in perfectly straight lengths of about 9 feet, and all in the same physical state with respect to heat treatment. An extensive series of cycles were taken with these wires, as in the u2 112 Scientific Proceedings, Royal Dublin Society. previous experiments, with a constant load on of 10° grammes per sq. em., the J:th part of the tenacity of iron, and, before testing, they were, as in the first series, all passed twice through a bunsen flame and carefully annealed. To find, however, the effect of annealing on the material, one of the wires—a No. 16—was tested in the condition in which it came from the manufacturer; it was then annealed and again tested, with the result shown in fig. 6, where the area of the cyclic curve is increased about six times by the annealing. CCE ae Heats. | Ls Deflections on Galyanometer Scale 40 0 TTee 720 160 200 heres of Twist. Fie. 6. These iron wires were very much softer and better magnetically than those used in the first set of experiments, as may be seen by comparing the areas of the cyclic curves in the two cases when the same size of wire—No. 16—was used, and the same load on, and when in the same magnetic field of 0°45 ¢.g.s. units. When plotted to the same scale, the area given by the new wire is about two and a half times that given by the other. The curves in fig. 6, and all the curves drawn to obtain the values for the curve in fig. 7, were plotted to a different scale from the previous Brown—Mechanical Stress and Magnetisation of Tron. 113 ones; on the axis of abscissee 1 cm. still represents 20° of twist, but on the axis of ordinates 1 cm. represents 10 divisions on the galvanometer scale instead of 4 cms., as in the previous cases. In curving both sets of observations a scale was chosen, so as to get a minimum of error in counting the squares for finding the areas of the cyclic curves. As already stated, each wire before being tested was annealed with a bunsen flame, then carefully cleaned with emery-cloth, and hung up in the Karth’s vertical field of 0°45 c.g.s. units as indicated in fig. 1. The throws on the galvanometer scale were observed, as before, for the first part of the cycle by going steps of 20° each from 0° to 180°; the galvanometer was then cut out of circuit, and the wire put through four complete cycles; the galvanometer was put in circuit again, and the cycle completed in the usual way. All the six sets of wires were done in this manner with a longitudinal stress or load on of 10° grammes per sq. cm., and in a magnetic field of 0:45 ¢.g.s. unit. The results are here given in Table V., and shown in the curve fig. 7. TaBLe V. Cross-sectional area of the wires Complete area of the cycle in sq. cms. curves in sq. cms. 3°9 x 10% 17-7 6°5 Pe 31:7 11:7 AG 56:7 20°6 55 87-2 32°3 o 91°6 54:7 5 94°6 Since the area of the cyclic curves represents in arbitrary units the amount of circular magnetism produced by the twist, we find that for wires up to about 15 x 10% sq. cms. cross-sectional area, the circular magnetism is very nearly proportional to the sectional area of the wire, and from 24 x 10° to 54 x 10% sq. ems. the change is very small. ‘he latter is due to the fact that these two wires were permanently deformed when the angle of twist was 180°, ic. at the top part of the cycle. This was observed, as when the cycles for these two wires were finished, and the vibrator left 114 Scientific Proceedings, Royal Dublin Society. free, it did not return to zero as in the other four wires, but was from 5° to 10° out. The positions therefore of these points on the curve is what one =r Tins SS 1 als LE { Bee Area of Cyclic Curve in sq. ems. a SS a 15 ; Rae 45 605 Cross-sectional Area of Wire in sq. cms. x 10-%. Fie. 7. would expect, for if the outer layers on the wires get a permanent shear, or are taken beyond the elastic limit, the molecules of the iron in these layers would contribute little or nothing to the circular magnetism or area of the cyclic curve. In order now to find the effect of varying the circular magnetism on a given wire with different loads, the ballistic galvanometer was taken out and the circuit re-arranged. In series with the wire under test were put a storage-cell, ammeter, variable resistance, reversing key, and plug key. A No. 16 iron wire was taken, and carefully annealed, and hung in the Harth’s vertical magnetic field with a given weight on the free end as in fig. 1. On the lower end of the wire a small plane mirror was fixed, and a lamp and scale suitably placed; by this means the twist of the end of the wire could be measured. Now seeing that when we twist an iron wire which is placed in a magnetic field a momentary current is produced, it follows that if we send an electric current through the wire the free end of it will twist. ‘This is the Wiedemann effect. The circuit being arranged as indicated above, currents were sent through the wire in eight different steps, from zero to 2°06 amperes, and the twist or deflections on the scale read off in each case. The current was then diminished by the same steps to 0°12 amperes, and the circuit broken, the Brown—WMechanical Stress and Magnetisation of Tron. 115 eurrent was reversed and the same steps taken to 2:06 amperes and down to zero, when it was again reversed, and by the steps increased up to two amperes again, thus finishing the complete cycle. ‘The twist on the wire lags on the current going through it, and when the currents in amperes through the wire are plotted against the twist or deflections on the scale a very symmetrical cyclic curve is obtained. The maximum current of 2°06 amperes used through this No. 16 wire was at a current density of 100 amperes per sq. cm., and was not large enough to appreciably change the physical condition of the wire. A cycle was taken when the wire was suspended in the vertical component of the Harth’s magnetic field 0°45 c.g.s. unit, and with a load on the wire of 10° grammes per sq. cm. The observations were then plotted on millimetre paper, where 2 cms. represented one ampere on the axis of abscissee, and 1 cm. on the axis of ordinates represented ten divisions of twist on the scale, the distance from the scale to the mirror being 114 cms. and 700 divisions on the scale being 44 cms. The area of the complete cyclic curve so obtained was found to be 28°5 sq. ems, and the maximum twist for 2:06 amperes was 36°8 scale divisions. When the load on the wire was doubled and another cycle taken in the same field, and with the same maximum current, the area of the cyclic curve was increased by 3:5 sq. cms., and the maximum twist increased by about 5 scale-divisions. (H. Gerdien! has shown that the cyclic variation in the torsion of an iron wire produces a cyclic variation in the longitudinal magnetic moment.) In order to test the effect of a stronger magnetic field, a solenoid was now put round the wire when stili in a vertical position. This solenoid was 236 cms. long, and 2 cms. internal diameter, and consisted of 7707 turns in 4 layers of number 20 d.c.c. copper wire wound uniformly over the whole length, the resistance of the coil bemg 17 ohms at 12°C, and the magnetic field inside the solenoid being about 41 c.g.s. units per ampere. When the solenoid was in position, the wire projected from the ends by about 2 ems. at the top, and 8 cms. at the bottom. This method of getting a magnetic field does not allow the wire—as here arranged—to be in a uniform magnetic field throughout its entire length, so that results got by its means are not quite comparable with those obtained when the wire is in a uniform magnetic field such as the vertical component of the Karth’s magnetic force. The solenoid was, however, tested experimentally by means of an exploring coil, and the internal magnetic field was found to be perfectly uniform along the whole length to within 5 cms. of each end, so that in the 1Ann. d. Physik, 1904, pp. 51-86, 116 Scientific Proceedings, Royal Dublin Society. following experiments we may take it that about 226 cms. of the wire was in a magnetic field of uniform strength. The circuit was now arranged with the No. 16 iron wire hanging vertically in the centre of the solenoid, and a longitudinal stress, or load, put on the end of the wire of 10° grammes per sq.cm. A current was sent round the solenoid in such a direction as to give, along with the Harth’s vertical force, a total magnetic field of 7 c. g.s. units. The wire was then put through a complete cycle, by sending currents through it, increasing by steps up to 2°06 amperes as maximum, and the deflec- tions on the scale or twist read at each step as before. The load on the wire was then increased, and acycle again taken, and so on for seven different values of the load in all. The cyclic curves were all drawn to the same scale, viz.,2 cms. representing one ampere on the axis of abscisse, and 1 cm. representing ten scale divisions, or twist, on the axis of ordinates. The area in sq. cms. of the complete curve was measured in each case, and also the total twist for the highest current of 2°06 amperes observed. The results for a total magnetic field of 7 c. g.s. units are here given in Table VI. Taste VI. Load on wire in Max. twist or deflec- | Total area of cyclic grammes per sq. cm. | tion in scale divisions. curves in sq. cms. 1:00 x 10° 56:0 26°4 TEAS} a 52'0 24°8 1:48) is 48°5 23°0 ile7fik | So 47:0 22:0 2,0 Gaus 46:0 21:0 Ql Sia tess 465 21:0 240 eee 47:0 21-7 If we plot the numbers in the first column of Table VI. as abscissee, and the numbers in the second and third columns as ordinates, we obtain two curves which fall to a minimum towards the axis of abscissa, and then rise slightly. It will, therefore, be found that when the load on the wire is increased about 70 per cent. the maximum twist is decreased about 16 per cent. ; and fora load of 2:1 x 10° grammes per sq. cm., or about the 5th part of the tenacity of iron-wire, the twist reaches a minimum; and when the load Brown— Mechanical Stress and Magnetisation of Iron. 117 is further increased by about 16 per cent., the twist is slightly increased, or about 2 per cent. It is difficult to see why this increase should take place. It might, however, be due to a slight error in reading the value of the current through the wire. Tt will also be seen that the area of the cyclic curve which measures the circular magnetisation is decreased by about 9 per cent. for an increase of 70 per cent. in the load, and reaches a minimum at a load of 2'1 x 10° grammes per sq.cm.; and then there is a slight increase in the cyclic area fora further increase of 16 per cent. on the load. The same wire, No. 16, was now annealed by passing the flame of a bunsen burner slowly over it; and was again tested for two different loads in magnetic fields of different values. That is, it was tested after annealing with a load on of 10° grammes per sq. cm., when placed in magnetic fields of strengths from 4 to 13°5 ¢ g.s. units. Itwas then re-annealed, and again tested for a load of 2-43 x 10° grammes per sq. cm. when placed in the same fields. The results thus obtained are given in ‘lable VII. and fig. 8. Taste VII. Load on wire, Magnetic Field, Maximum Twist, Total area of cyclic grammes per sq. cm. c. g. 8. Scale divisions. curve in sq. cms. 0°45 36°8 23°5 4 64:0 33°0 1 x 10° | 7 56:0 26:4 10 49-0 22-0 | 13°65 | 38°0 16°5 — 4 55:0 29°5 7 47-0 22°5 | ee 10 39-0 17-0 13-5 28-0 12-0 Tf we plot the values of the magnetic fields as abscissee and the two last columns in Table VII., as ordinates, we obtain the curves shown in fig. 8, where the first curve at the top is that of maximum twist and magnetic field for a load on the wire of of 10° grammes per sq. cm., and the second curve is that for a load of 2°43 x 10° grammes per sq. em. SCIENT. PROG. R.D.S., VOL. XII., NO, XII. x 118 Scientific Proceedings, Royal Dublin Society. The upper of the two lower curves is the curve obtained by plotting the area of the cyclic curves against the magnetic fields when a load of 10° ——, a a 5 2 - 4 2 g H DR! 70 a=! i 2 97] ay) f g a ‘S| ] BI c H a a JO} a = Pl eabeinill ia i G ) H 8 Bind: FO" _# Hf | Std H “ i 30% yi X Area of Cyclic Cnrve in sq. ems. ox S Magnetic Field in C. 8. G. Units. Fie. 8. grammes per sq. cm. is on the wire; and the last curve is that for a load of 2°43 x 10° grammes per sq. cm. The two upper curves in fig. 8 are practically parallel for the two different loads on the wire, and show that as the magnetic field round the wire is increased from 4 to 10 ¢.g.s. units, the maximum twist is decreased about 23 per cent. for the smaller load, and about 30 per cent. for the higher load, i.e., when the load is increased 2°43 times the diminution of the maximum twist is further magnified by about 7 per cent. As the magnetic field is further increased from 10 to 13°5 units, the maximum twist decreases about 22 per cent. for the light load, and 28 per cent. for the heavier load. This drop in the curves for a field of 13°5 units is what one would expect from the curves given in Nagaoka’s' report, where the maximum twist is 1 Nagaoka, Report, loc. cit., p. 550, Brown—WMechanical Stress and Magnetisation of Iron. 119 shown to be obtained in a field of about 10 c. g. s. units, which depends, however, on the value of the current in the wire. From the two lower curves in fig. 8 we see that as the field is increased from 4 to 13°5 units, or about three times, the circular magnetism or area of a eyclic curve is decreased about 50 per cent. for the smaller load, and 60 per cent. for a load 2°43 times greater. From some experiments which are now in progress, but just started, with the wire in a uniform magnetic field throughout its entire length, it has been found that the maximum transitory ; current produced by twisting through 100° a wire of the same dimensions, and with the same load on, as the one used above, was obtained in a field of 2:5 ¢g.s. units; so that for fields from 0-45 to 4 units the first part of the curve, fig. 8, would probably be as shown by the dotted lines.’ The whole six wires were re-annealed by means of the bunsen flame, and the maximum twist or deflection on the scale measured. In each case the longitudinal stress or load was 10° grammes per sq. cm., and the current density through each wire 100 amperes per sq. cm. as before. The tests were made when the wires were in the Harth’s vertical magnetic field, and also when they were inside the solenoid with a total magnetic field of 7 ¢.g.s. units. The results are here given in Table VIII., and shown in curves in fig. 9, where the cross-sectional areas of the wires in sq. ems. are abscissee, and the twist or maximum deflections on the scale are ordinates. TABLE VIII. Maximum Twist for Cross-sectional area of wires in sq. cms. e104 om Hae 3°9 x 10-8 4 32°5 6°5 56 10 36°0 07 a 19 43°5 20°6 " 36°8 56:0 32°3 5 35 53°0 54:7 we 28 51°0 1 Subsequent work has shown that this is so: see figs. 2 and 3 in Part II. of this paper, which was read before the Royal Dublin Society on April 20th, 1909. x2 120 Scientific Proceedings, Royal Dublin Society. The lower curve in fig. 9 is that due to a field of 0:45 units, and the upper curve that due to 7 units. From this we see that in both cases SA Maximum Twist in Scale-divisions, 0 20 40 6d Cross-sectional Area of Wire in sq. cms. x 103, Fie. 9. the maximum twist is obtained in these magnetic fields with a wire of cross- sectional area 20°6 x 10° sq. cms., and that when the cross-section of the wire is increased about 54} times, i.e., from No. 22 to 16 standard gauge, the twist is ¢rereased about 9 times in a field of 0:45 ¢. g.s, units, and about 13 times in a field of 7 ¢. g.s. units. Also for a further increase in the cross- sectional area of the wire of 2°65 times the twist is decreased about 7 per cent. in the lower field, and about 2:4 per cent. in a field of 7 c. g.s. units. The decrease in the maximum twist with the two thicker wires is some- what difficult to account for'; a similar phenomenon occurs when the cross- sectional area of the wire and the current through it are kept constant, and the longitudinal magnetic field round the wire varied. In the latter case the twist increases up to a certain value of the longitudinal magnetism, and ‘ Since the above was written the author has observed that the magnitude of the longitudinal magnetic field required to give maximum twist is different for different-sized wires, i.e. about 2°3.c. g. s. units for a No. 12 wire, and 2°8 units for a No. 14 wire, when the load on each is 10° grammes per sq. cm. Brown— Mechanical Stress and Magnetisation of Tron. 121 a then diminishes as the surrounding field increases, and even reverses if the field is sufficiently increased, as was shown by Villari.1 In the former case it would seem from these experiments that when the longitudinal magnetic field and the current density in the wire are kept constant, and the cross- sectional area of the wire varied, the maximum twist attains its highest value with a wire of cross-sectional area 20°6 x 10 sq. cms. Conclusions. The following conclusions may be arrived at from these experiments :— 1. he momentary electric current produced by an axial twisting in a magnetic field of an iron wire which ig fixed at the top and in a vertical position with a vibrator on its lower end, is very approximately proportional to the amplitude of oscillation of the vibrator within the elastic limit of the wire. 2. When the longitudinal magnetism and the cross-sectional area of an iron wire are kept constant, the circular magnetism produced by twisting the wire is diminished as the load is increased up to the 34th part of the tenacity of the wire. When the load is increased 33 times, the circular magnetism is decreased about 40 per cent. 3. When the load and the longitudinal magnetism on an iron wire are kept constant, the circular magnetism produced by twisting is proportional to the cross-sectional area of the wire within the elastic limit of the wire. 4. When the longitudinal magnetism is kept constant round an iron wire, which is carrying an electric current of density 100 amperes per sq. cm., the maximum twist of the free end of the wire is decreased 16 per cent. for an increase of 70 per cent. in the load; and for the same change in the load the circular magnetism is decreased 9 per cent. 5. When the load is kept constant on an iron wire which carries a current of density 100 amperes per sq. cm., the maximum twist of the free end of the wire decreases as the longitudinal magnetism dnereases. Kor a load of 10° grammes per sq. cm. the twist is decreased 23 per cent. when the magnetic field is increased 2°5 times; and when the load on the wire is increused 2°43 times, the maximum twist is decreased 30 per cent. for the same variation of the longitudinal magnetism. 6. When the load on an iron wire is kept constant at 10° grammes per sq. cm., and the longitudinal magnetism increased from 4 to 13°5 units, or about three times, the circular magnetism is decreased about 50 per cent. ; 1 Poge. Ann., 1868. 122 Scientific Proceedings, Royal Dublin Society. and when the load on the wire is increased 2°48 times, the decrease of circular magnetism is about 60 per cent. for the same range of field variation. 7. When the load per unit area, the longitudinal magnetism, and the current density in iron wires are kept constant, the maximum twist of the free end of the wire increases with the cross-sectional area of the wire up to about 20°6 x 10° sq. cms. In a magnetic field of 0°45 c.g. s. units, when the cross-sectional area of the wire is increased 5} times, the maximum twist on the wire is increased 9 times; and in a field of 7 c.g. s. units the maximum twist is increased about 1} times for the same change in the wire’s cross-sectional area. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 13. MAY, 1909. THE OSMOTIC PRESSURES OF THE BLOOD AND EGGS OF BIRDS. BY W. R. GELSTON ATKINS, B.A., UNIVERSITY CHEMICAL LABORATORY, TRINITY COLLEGE, DUBLIN. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN : PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. - WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price Stapence. OUUE THH OSMOTIC PRESSURES OF THE BLOOD AND EGGS OF BIRDS. By W. R. GELSTON ATKINS, B.A., University Chemical Laboratory, Trinity College, Dublin. [ COMMUNICATED BY PROFESSOR H. H. DIXON, SC.D., F.R.S. | (Read Maren 28. Received for Publication Apuin 6. Published May 28, 1909.] WHILE the relation between the osmotic pressures of the blood or body- fluids of marine and fresh-water vertebrates and invertebrates and that of the surrounding medium has been studied by many workers, notably by Bottazzi, Garrey, Rodier, Greene, and Dakin; and while the blood of mammals has been examined by Dresser, Koranyi, Hamburger, and others, the blood of birds has been quite neglected, as far as I am aware, except for three determinations of the osmotic pressure of the blood of the common fowl, Gallus bankiva, by Hamburger, and three by Grijins. ‘This may be accounted for by the fact that the study of the blood of birds apparently affords no question of interest comparable to the equilibrium maintained in water-dwelling organisms. The attractiveness of the study of the osmotic forces in land animals lies in the equilibrium between the various body fluids and between the same fluids at various times. The work of Koranyi, Bousquet, Kummel, and the great mass of literature on the normal constancy and pathological variations in human blood, and the relation between the osmotic pressures of the blood and urine, bear witness to the great importance of this branch of the subject. It seemed likely, however, that the relation between the osmotic pressures of the blood and eggs of birds might be of interest ; so the study was begun, but difficulty in obtaining the requisite materials delayed it, and still prevents its completion. ‘The method adopted was the usual one—the determination of the freezing-points of the various liquids. Technique of freezing-point Determinations. The apparatus used was the ordinary Beckmann with thermometer graduated in hundredths of a degree; tenths of a scale-division were SCIENT, PROC. R.D.S. VOL. XU., NO. XII, Y 124 Scientific Proceedings, Royal Dublin Society. estimated, using a pocket-lens, care being taken to avoid parallax. The liquid, placed as usual in the inner tube, was slightly supercooled, and then caused to freeze by addition of a minute speck of ice. The highest point reached by the mercury column on rising after supercooling was carefully observed. As the separation of pure solvent in the solid state increases the concentration of the remaining liquid, the freezing-point as thus found is theoretically too low, viz. further removed from that of pure water than it ought to be. If the supercooling has not exceeded half a degree, the error will be very slight. Bearing these facts in mind, the routine adopted was to place the tube with the thermometer, after noting the reading, in a beaker of water till the temperature rose to about 2° C. It was then free from flakes of ice; but minute fragments still remained. On replacing the tube in the cooling-space the temperature fell, but remained stationary at the true freezing-point ; for solidification begins immediately the freezing-point is reached, if even a minute particle of the solid phase is present. After a while the mercury column will begin to fall; this is due to an appreciable alteration in concentration occasioned by the separation of ice. If there was a slight difference between the readings obtained by these two methods, the last was always a smaller depression than the first. In such cases the last determination, and not the mean, was taken, as it is less likely to have any error due to supercooling. But if the supercooling was slight, there was no measurable difference between the results got by the two methods. The zero of the thermometer was redetermined from time to time. It was found to fall from 3:64-3'33 in about two years, the change being quicker in summer than in winter. This change in the zero, referred to by most experimenters, is due, I believe, not to changes in the glass, which are relatively unimportant even with a large, thin bulb, but to the distillation of the mercury from the curved surface of the hanging drop in the top of the reservoir of the Beckmann thermometer, to the flat surface at the lower end. The drop is caused by the expansion of the mercury, on the scale at 0° C., when warmed up to room temperature. Now, although the pressure of mercury vapour at about 15° C. is very slight, yet that of the curved surface always exceeds that of the flat surface. Thus, as both are at the same temperature, slow distillation must go on from the curved to the flat surface. The result is that when cooling to 0° C. causes the drop to withdraw into the narrow capillary stem, it does not stand as high as it did formerly, owing to loss of mercury. The drop is more curved in summer, as more mercury leaves the bulb; hence there is a greater difference in vapour-pressure, and consequently the distillation is more rapid, Arkins— Osmotic Pressures of the Blood and Eggs of Birds. 125 Vigorous and continued stirring is necessary in all work with such viscous fluids as blood and eggs. Materials used. The blood examined was obtained from the neck of the bird, and drained into a clean dry bottle, which was then corked, and sent to the laboratory. The whole blood was used, both the serum and the contents of the clot. ‘The specimens were examined as soon as possible; but standing for two days in a cold room in winter was found to produce no alteration. Putrefaction, however, alters the freezing-point, increasing the number of molecules by the breakdown of complex into simpler ones, and by disintegration of the red corpuscles. The blood was always in the arterial condition when examined, being saturated with oxygen by the constant stirring. The eggs used were perfectly fresh, and were broken and drained direct into the inner tube. No difference was found in the freezing-point of the white and yolk of the same egg, and a mixture of white and yolk gave the same depression. Accordingly, in all the experiments recorded, the white of the egg was employed, as it was much easier to stir than the yolk. he results obtained are tabulated below. Common Fown (Gallus bankiva). Blood A. Beg ix 1. 0:610 0:470 2. 0°605 0:480 3. 0:600 0-480 4. 0°935 0°450 5. 0°597 0:445 6. 0-610 0-482 7. 0°603 0-449 8. 0:662 0-482 9. 0-617 0-442 10. 0:60 | 0:427 11. 0:62 /;Grijins. 0:457 12. 0:62 J (Pfltiger’s Arch. 63. 1896. S. 86). 0-430 18. 0°591 14, 0605| Hamburger. Mean, 0°454 15. | (Osmotischer Drucke und 16. 0-600) Tonenlehre). Mean (15) 0-607 (omitting No. 4). 126 Scientific Proceedings, Royal Dublin Society. a See S } Mean, COIAPD NFR O pr oS Mean, em Oo De Mean, Turkey (Meleagris gallopavo, L.). Blood A. °C 0:620 0°610 0°642 0:607 0:627 0:682 0°612 0:621 Duck (Anas). Blood a. “0, 0635) 0-660 0:562 (mean of two birds). 0:580 0-587 0°572 0559 0:580 0:585 0:587 0-574 (omitting 1 and 2). Mean ‘somewhat decomposed. Gooss (Anser). Rhea Americana. Blood a. 0°:662 C. Arxins— Osmotic Pressures of the Blood ani Eggs of Birds. 127 Binp. A OsmorTic Pressure. Blood. Keg. Blood. Egg. Common Fowl, 0:607° C. 0:454°C. 7:3latm. 5:47 atm. Turkey, . . 0°621° 7:48 Duck, . 5 Ops 0:452° 6:92 5:45 Goose, . 5 Opi 6°65 Rhea Americana, 0°662° 7:97 The osmotic pressures were calculated from the formula [I = 12:06 A — 0:021 A’, It is noticeable that all the samples of blood examined froze between — 053°C. and — 0°66°C., with one exception, probably pathological, though each class has a fairly constant freezing-point. In Gallus bankiva the variations are very small—0:031°C. between the extremes, neglecting one remarkable value, A = 0:9385°C. The latter was repeated three times; the blood was perfectly fresh. It was the only sample received from a farm; and no particulars as to how it was collected were obtainable. It seems probable that it was got by killing a diseased bird. In Meleagris gallopavo a practically identical range was found, 0:035°C. ; while in Anser also the range was 0:035°C. In Anas the difference between the extremes is 0°028°C., neglecting the results of the two de- composing specimens. It is to be noted that the values of A for the fowl, A = 0°607°C., and turkey, A = 0°621°C., are higher than those of the duck, A = 0:574°C., and the goose, A = 0°352°C., both of which are aquatic. Hamburger states’ that the osmotic pressure of the blood of birds of prey is higher than that of other birds, being isotomic with a solution of sodium chloride considerably above IT percent. He gives 0:598°C. as the value of A obtained by him for a 1 per cent. solution of this salt. The value of A 0°662°C. given by the blood of the American ostrich, Rhea, is the highest found for any bird.” It is impossible to draw definite conclusions from one experiment, especially as the blood was obtained from a bird which had died from cold. It was, however, in perfect condition, and the blood was quite fresh ; and itis worthy of note that Rhea is a Ratite bird, all the others being Carinate. Turning now to the egg, it is found that the mean value of A for the ege of Gallus is identical with that for Anas, the difference 0:002° C. being quite negligible, while the value of A given by the blood of these two differs ! Osmotischer Drucke und Ionenlehre, vol. i., p. 458. 2? Thave to thank Dr. A. F. Dixon and the Royal Zoological Society of Ireland for this and several other specimens. 128 Scientific Proceedings, Royal Dublin Society. by 0:035° C. The range of variation in A for the egg of Gallus is 0°053° C.; while in Anas, differences amounting to 0 086° C. were found. ‘These fluctu- ations are considerably greater than those found in the blood ; and the value of A is in every case much less than for the blood. The egg-white is a secretion ; and as secretions often differ widely from the blood in concentration, there is nothing surprising in this. It is, however, remarkable that the ovum proper, which was in equilibrium with the blood and body-fluids before becoming detached, is found in a nutrient medium of much smaller osmotic pressure. The cause of this reduction in osmotic pressure has now to be sought. Is it due to a diminution in the inorganic salts, of which sodium chloride is by far the most important, and the others may be assumed to vary with it; or is it due to a decrease in the organic crystalloids present ? ‘To decide this question an analysis was made of the chlorine in the white of duck-eggs and in the plasma of duck-blood, from which the red corpuscles had been separated by a centrifuge. A few determinations of the chlorine in sera of other birds were also made ; but these were not quite free from red corpuscles. ‘The analysis was carried out by drying and igniting the egg or serum in a covered erucible, and then dissolving the ash in dilute nitric acid and titrating the chlorides by Volhard’s method. As the quantities of chlorine found amounted only to from 0:03 — 0:02 grm., the accuracy attained is not very great. Per cent. chlorine in impure serum. Fowls, . : : 0:273 96 : 6 6 0:277 Turkey, . 0 : 0:290 Goose, . : ; 0:2938 50 : . : 0:317 Duck. Plasma Egg-white per cent. chlorine. per cent. chlorine. 0:278 0:080 0:276 0-088 0:312 0:088 0:104 Mean, 0:287 Mean, 0-090 As NaCl, 0°473 per cent. As NaCl, 0-148 per cent. These results make it clear that there is a great loss of chlorides in the egg, the amount present being only about one-third of that in the blood plasma. Consider now the réle played by the chlorides in lowering the freezing-point. The amount of chloride in the serum, calculated as the Arkxins—Osmotic Pressures of the Blood and Eggs of Birds. 129 sodium salt, viz. 0°473 per cent., would give a value A = 0:282°C., leaving 0:292° C. as the depression produced by the organic solids and the remainder of the inorganic. In the egg the sodium chloride amounts to 0148 per cent., which produces a depression A = 0:091°C., leaving 0°361°C. as due to the organic and remaining salts, so the egg is richer than the blood in organic crystalloids. ‘These values of A were got by interpolation from the results of Loomis ;! as his observations are very close together, interpolation introduces no error. Considering, now, the difference between the values of A for serum and ege A, — A, =0:122°C.; the difference in the sodium chloride depressions gives Asnac! — Acnaci = 0°191° C. Thus the loss of chlorides is more than sufficient to account for the difference in the values of A for blood and eggs. This is using the mean values in the calculations. But taking the maximum values of A for serum and the minimum value for egg, A; — A. = 0:171°C.; while the minimum amount of sodium chloride in the serum and the maximum amount in the egg give Asnaci — Acnaci = 0'158°C., thus reversing the result. On the whole, it appears probable that there is in the egg a reduction in the inorganic salts beyond what is required to account for the difference in A given by serum and egg, the excess depletion being compensated by an increase in the organic crystal- loids. I hope, however, to investigate this point further. This reduction in pressure accounted for above may take place either— (a) in the ovary, (0) during the passage of the egg down the oviduct, or (c) in both stages. Consider now the first case—(a), The ovum is being enlarged by the deposition of nutrient substances, resulting in a giant cell composed almost entirely of metaplasm. It is hard to see why there should be any reduction in pressure under these conditions; so it seems probable that the osmotic pressure of the ovum in the ovary is the same as that of the surrounding body-cells. (b) When the ovum begins to travel down the oviduct, it is coated with various layers of albumen secreted by the glands of the duct. Now these secretions may have a low osmotic pressure, like human saliva, which freezes at — 0°10° C., while human blood freezes at - 0°56° C. Diffusion taking place between the yolk and the albuminous coating would soon equalize the concen- tration, so there would be a slow fall in the freezing-point of the yolk, from that of the blood to that of the egg as determined when laid. This 1 Wieden. Annal. 51, 1894, s, 515, 1380 Scientific Proceedings, Royal Dublin Society. fall would, according to the above view, take place as the yolk travels down the oviduct. (c) It is possible that both the yolk as formed in the ovary and the albuminous secretions of the oviduct are isotonic, both bemg deficient in chlorides. In any case it is found that, when the two layers of membrane and the shell have been formed around the egg, the osmotic pressures of the yolk and white are the same. The two membranes together are semi- permeable; but it was not ascertained which of them prevents the passage of salts, or whether both act. The deposition of the shell, which is nearly pure calcium carbonate, must be a great drain upon the supply of caleium in the blood feeding the shell-glands. At any rate the amount of calcium found in the interior of an egg is small, only 0:04 per cent., or 0:023 grm., in an egg weighing 52°84 erm. without the shell, which weighed 6:18 grm. It has recently been proved (En. Carpiaux, Bull. Acad. Roy Belg., 1908, pp. 283-295) that during incubation some of the shell is dissolved by the carbonic acid formed by the respiration of the chick, and the calcium thus obtained is utilized in growth. The supply of phosphates for bone-formation is obtained from the lecithin of the yolk. Why the germ-cell is exposed to such a change of osmotic pressure at all is not clear; but it has been experimentally proved that dilute solutions are more favourable to cell-division than concentrated ones. It is also possible that this reduction in pressure is a provision for the increase of pressure arising from the metabolism of the chick in the egg, with no possibility of eliminating waste except into the cavity of the allantois; for the breakdown of the complex proteids of the egg into simpler molecules must greatly increase the osmotic pressure; and this increase might be injuriously great were it not for the previous reduction in pressure. I hope to study the relation further, and to determine the osmotic pressure of the egg during various stages of incubation. Summary. The blood and eggs of birds are not isotonic, the osmotic pressure of the egg being considerably the lower. ‘The blood of each kind of bird has an almost constant freezing-point, the fluctuations being of the same order as those met with in mammals. The differeuce in the osmotic pressures of the blood and egg is slightly more than accounted for by the diminution in the inorganic salts of the egg as compared with the blood. I wish to thank Dr. Sydney Young, r.r.s., and Dr. H. H. Dixon, r.r.s., for their advice and criticisms throughout the work, THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XIT. (N.S.), No. 14. JUNE, 1909. CHRYSOPHLYCTIS ENDOBIOTICA, Schilb. (Poratro-wart or Brack Scas), AND OTHER CHYTRIDIACEZ. BY T. JOHNSON, D.Sc., F.L.S., PROFESSOR OF BOTANY IN THE ROYAL COLLEGE OF SCIENCE, DUBLIN. (PLATES IX.—XI.) [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATEH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price One Shilling. . ian Instit,,.. te eum tty U bse) \ NOV A 1s09 fel 4 XIV. CHRYSOPHLYCTIS ENDOBIOTICA, Schilb. (Poraro-warr or Brack Scas), AND OTHER CHYTRIDIACEA. By T. JOHNSON, D.Sc., F.LS., Professor of Botany in the Royal College of Science, Dublin. (Prares TX.—X1.) {Read Marcu 23, Ordered for Publication Aprit 6. Published Junx 16, 1909. ] My attention was first called to the warty disease of potatoes by an article in the Journal of the Board of Agriculture (England), entitled, “ A new Potato Disease, Chrysophlyctis endobiotica, Schilberszky,” from the pen of Professor M. C. Potter (1), who, in the beginning of 1908, sent me some spore material. Later in the year, through the late Dr. Maxwell Masters of the Gardeners’ Chronicle, I got a quantity of warty tubers from Cheshire, from Mr. New- stead, the Curator of the Grosvenor Museum, Chester. I naturally watched with interest the spread of the disease through Great Britain, and its continued non-appearance in Ireland. Attention in Ireland had been called in 1907 to the dangerous character of the disease, by the publication of an illustrated leaflet by the Department of Agriculture and Technical Instruction for Ireland (2), and early in 1908 by the issue of an Order in Council, giving power to penalize anyone found importing or concealing cases of the disease. Hundreds of samples of scabby potatoes were in consequence sent in for report; but they proved to be either the ordinary scab or Spongospora scab. In October, 1908, however, by the discovery of three cases of warty tubers in Co. Down, Ireland’s clean record was lost. Steps were immediately taken by the Department to stamp out the disease, it need hardly be said. At the end of October, I received from it some of the warty tubers (PI. TX., fig. 1). Mxamination of the haulms or “potato- tops” of the diseased plants showed that the trouble was not confined to the tubers, but affected also the branching underground stem (the stolons), and even the fibrous roots. This is in harmony with the observations of Borthwick (8) in Scotland, and of Salmon (4) in Kent, who give excellent SCIENT. PROC., R.D.S., VOL. XII., NO. XIy. Z 132 Scientific Proceedings, Royal Dublin Society. illustrations of the warty excrescences on the haulms at or near the collar. Hence it may be concluded that there is no part of the potato-plant not liable to injury by the disease. This being the case, it follows that no part of the plant should be left undestroyed in attempts to eradicate the disease from an affected centre. It is known that diseased tubers leave spores in the ground, and the diseased shoots no doubt do the same. A diseased tuber is readily recognizable by the coral- or brain-like wart, tumour, or carbuncle on it at one or more points, but especially at the toe or crown end. ‘The wart is a wrinkled proliferation or corrugation of the flesh of the tuber (Pl. IX., fig. 2), due to excessive cell-division caused by the stimulating presence of the fungal parasite. It is apparently an attempt on the part of the host to rid itself of the enemy. Microscopic examination of one of the convolutions of the wart shows the abundance of the resting spores of tho fungus (Chrysophlyctis endobiotica, Schilb.), (Pl. IX., fig. 3). To the naked eye the first signs of attack of a tuber are observable in the abnormal appearance or malformation of a sprouting “eye.” The grey surface of the swollen eye is dotted over with golden-yellow rings, as seen with a pocket lens. ‘he microscope shows that these rings are optical sections of the walls of the resting spores. When ripe these walls become dark-brown, and the whole wart then looks more or less black, giving the term of “ Black Scab,” or, better, “‘ Black Wart,’ to the disease. It is unfortunate that Schilberszky’s (5) first and only account of the trouble in 1896, good as it is, is brief, incomplete, and unillustrated. His reference to the crater-like excavations of the potato flesh, arising in the later stages of the disease, suggests that the black scab attack he had to deal with was complicated by some other disease such as the deep form of Spongospora scab. Further, except for the knowledge due to Potter, that the trouble, as would be expected, is perpetuated by the resting spores, his account was not added to until November, 1908, when I was able to announce in Nature that I had succeeded in making the recalcitrant resting-spores germinate. Chrysophlyctis is a member of the Chytridiaceze, a low group of fungi having many points of affinity with the Slime fungi or Myxomycetes, and being, in the opinion of some botanists, the originating group of fungi in general. ‘They are classified partly by the character of their plant- or vegetative body, partly by their mode of reproduction. In the lowest forms the plant-body is a naked plasmodium, and propagation is mostly vegetative : in the highest there are hyphee, suggestive of a mycelium; and reproduction is sexual as well as vegetative. Chrysophlyctis is a member of the lowest group, the Olpidiacese. Its plasmodium is distinguishable in the host-cell by being denser, homogeneous, and finely granular (Pl. XI, fig. 1). It may Jounson—Chrysophlyctis endobiotica and other Chytridiacce. 138 be seen abutting on the host-protoplasm, and disputing with it, as it devours it, occupation of the enlarging cell-cayity. ‘I'he host-nucleus is first affected, turning brown. The protoplasm follows, and then the cell-wall. This, though brown, does not, like the protoplasm and nucleus, disappear. The starch-grains are the last attacked, and remain white and uninjured for some time in an invaded cell. ‘The parasitic plasmodium passes from cell to cell by boring its passage through the host cell-wall. No doubt the parasite secretes the cellulose-dissolving ferment cytase, as well as a toxin to kill the host protoplasm before absorbing it. It is in this stage that it stimulates to active cell-division the surrounding host-cells, and produces the gall or wart. Propacation.—A. Swarm or Summer Zoosporangia. The parasite not only penetrates through its host from cell to eell by its plasmodium ; it also spreads from tuber to tuber, and from plant to plant, by the formation during the growing season of zoosporangia. ‘I'hese are elliptical bodies, with smooth yellowish walls, and numerous zoospores which escape through a hole in the wall, and attack healthy potato-tissue. Schilberszky saw the escape of these spores, and states that they can bore a passage for themselves into neighbouring host-cells, when their discharge is internal. The spores escape readily when the sporangia are placed in water or in macerations, says Schilberszky. B. Winter or Resting Sporangia or “ Spores.” As the tuber ripens, the parasite replaces the summer sporangia by resting ones which carry the disease through the winter, and serve to propagate it in the spring. Schilberszky mentions these resting spores, but contents himself by noting their thick, dark-brown walls, described by him as smooth. He leaves to the future an account of their origin, structure, and fate. The resting sporangia, 30-70 mw in diameter, are very numerous in diseased tubers, and are easily recognizable with a pocket lens (Pl. IX.., fig. 3). Under the microscope the wall is seen to be not smooth, but ridged or angular (Pl. XI., fig. 8). These brown ridges or bands form part of a kind of epispore which arises as the sporangium ripens, and seems to be formed from the residual contents of the host-cell when not also from its cell-wall as well. I find W. B. Grove (Gardeners’ Chronicle, 1905) supports this view in part, in that he says that “though the resting spores are smooth outside, they are sometimes closely invested with the brown angular remnants of the host-cell in which they had formed.” The epispore is thus deposited from without, as a third layer on the thickening wall of the sporangium. If this more or less artificial epispore is ignored, then one may speak of the spore-wall as smooth. Zz 2 134 Scientific Proceedings, Royal Dublin Society. Its presence is, however, enough to prevent one from confusing the resting spores of Chrysophlyetis with those of Urophlyctis leproides—the cause of beet- tumour, a mistake which has unfortunately crept into mycological literature. As a rule there is only one resting sporangium in a host-cell; occasionally there are two. Through the dark wall of the sporangium one can see the spongy or granular protoplasmic contents enclosing in the centre a bright body which reagents show to be a large reserve fat globule. The origin of these resting spores can be traced in microtome preparations (Pl. X., figs. 3, 4,5). The naked dense plasmodium soon becomes wholly converted into a sporangium which continues to enlarge and to occupy more and more of the enlarging host-cell. Chrysophlyctis is thus holocarpic, and differs from the eucarpic Urophlyctis, in which the whole of the vegetative body is not used up in the production of the propagating organs, as is the case in Chrysophlyctis. I followed the resting spore from its foundation to complete maturity, but must leave to a future occasion an account of the nuclear and other changes in the developing sporangium, as my results are not yet complete. I may say there are not wanting signs that the small host-cells surrounding the invaded enlarged one may lose their cell-walls in part, and form a sort of symplast round the developing sporangium. GERMINATION OF RestinG SporANneGiuM (or “ spore”’). Judging from analogy and from the examination of preserved sporangia, T concluded the “ resting spore’ must be a composite body or resting sporan- gium (Pl. XI., fig. 4). I proceeded to try and rouse it from its winter rest by stimulating solutions of one kind or another, e.g., by :—A, 1 per cent. sugar, B, 2 per cent. sugar, C, A or B with a trace of the proteid ferment papayotin, D, potato juice alone ; avd Z, D with papayotin. Cultures in these and in water were kept, as well as macerations, in darkness, in light, at room temperature, and in the incubator at a constant temperature of 20° C., and examined from day to day. I constantly found empty spore-cases with ruptured walls, but for some time failed to see any signs of escaping zoo- spores. At last the potato juice, exercising possibly a chemotactic influence, gave success ; and sporangia with split walls and escaping zoospores were found (Pl. XI., fig. 5). These showed the same sluggish movements observed in the sporangia of certain other Chytridians, disturbed during their resting period. Hach sporangium contains hundreds of more or less pear-shaped uniciliate zoospores. As the zoospores are being formed, the large central fat globule disappears, being broken up into innumerable minute globules which are Jounson—Chrysophlyctis endobiotica and other Chytridiacee. 135 absorbed by and reappear in part in the individual developing zoospores. When looking for sprouting sporangia, I had an experience one day (20th Nov., 1908) of sufficiently general interest to deserve a detailed account. I saw several zoospores rushing very rapidly backwards and forwards through the field of the microscope. I found next an elliptical sporangium with a small hole in its wall at one end. ‘The sporangium was practically full of zoospores swarming most vigorously and crowded round the opening in the wall in what looked like a general wild rush to be amongst the first to escape into the fresh air dissolved in the surrounding water. Within a minute the sporangium was emptied. ‘The motion of the zoospores through the field of the microscope was comparable to that of a mouse in a room; and the rough analogy is not lessened by the fact that the pear-shaped zoospore moved with its single cilium behind. Some of the zoospores were caught between the glass-slide and cover-slip and became stationary. ‘he body of the zoospore (1°5-2°4 « in diameter) was actively amoeboid, going through changes of form too rapid for me to sketch in succession (Pl. XI., fig. 6). The cilium remained either straight or became curved like a carriage-whip, and usually moved passively with the body of the zoospore, though it also showed slow sweeping movements backwards and forwards. Suddenly the body would be seized with a violent spasmodic fit, shake itself free (or not) from its temporary imprisonment, and swim away. ‘The amabidity of the body of the zoospore and comparative passivity of the cilium were striking. My letter to Nature in November, 1908, announcing the observation of the multisporous character and the germination of the resting sporangia, ealled forth a letter from KE. F'. Weiss recording his observations of germinat- ing spores in August. The time of the year suggests that these must have been the summer ones whose germination Schilberszky had already described. The resting ones do not, in my experience, germinate with ease in potato-juice, as Weiss found his sporangia to do. If they had done so, I should not have spent three months in examining them day by day under various conditions. Marchal had the same difficulty with the resting spores of the Chytridian Asterocystis radicis, He saw the escape of the zoospores from the summer sporangia readily, but failed to make the resting ones germinate, though he found them from time to time in his cultures ruptured and empty. This is equally true of the resting spores of Olpidium brassicae, the cause of ‘“ black leg” in young cabbages. It looks as if, judging from Chrysophlyctis, rest- ing spores must be cultivated in a decoction of or in immediate contact with the host-plant, that its chemotactic influence may be exercised on them to arouse the zoospores to activity. Summer sporangia sprout without diffi- culty in water material in most Chytridians. 156 Scientific Proceedings, Royal Dublin Society. Prnerration into Hosr. As our knowledge stands at present we are left to hypothesis to account for the entrance of the parasite into the potato-plant. The general assump- tion is that the zoospore, whether it comes from the summer or the resting sporangium, on reaching the potato-tuber, bores its way into the young epidermal cells of the tuber at one of the eyes, and there begins its destructive work. Massee (7) holds that the parasite is an epidermal one only, and that when the resting spores appear deep-seated, it is due to the depression of the infected epidermal cells by the division aroused in the surrounding cells. It is, I think, necessary to distinguish between the resting eyes of a ripe tuber in the autumn and those of one sprouting in the spring. The autumn eye is small, dormant, and protected; the spring one exposes its delicate tissue to attack when sprouting, and would more easily fall a victim to marauding zoospores. In my winter material I found tubers showing quite small, slightly warty eyes. [xamination of these showed them to be full of the parasite, not only of its well-developed resting spores, but of the plasmodium also (Pl. IX., figs. 5,6). Some of the leaves in the shoot forming the eye are, as the figure shows, riddled by the resting spores, and practically destroyed. Hxa- mination of the cortical cells of the tuber in the region of the eye, and beneath the protecting cork layer, shows that these cells are occupied by the parasitic plasmodium, which is, as Schilberszky states, thus subperidermal. The condition of affairs is such that one cannot imagine the parasite can have obtained its thorough hold of the eye by simple penetration of zoospores from without. It is more reasonable to conclude that (as I have already shown in the case of the Spongospora scab (8) of the potato) the black scab, or warty disease of the potato, is propagated, not only by the contact of healthy tubers with spores in the soil, but by the internal passage of the plasmodium from diseased tubers into the potato-tubers formed from them as seed, the next year. On the 6th November, 1908, I inoculated tubers, and slices of tubers, with decoctions and macerations of spores. The tubers were kept in darkness in sand at room temperature. The eyes have sprouted, and though the shoots now (11th March, 1909) show discolouration, it is too early to say if infection has taken place. I also placed a number of warty tubers under somewhat similar conditions, exclusive of the sand. In their case the eyes have sprouted a little ; and the new material in them shows an almost continuous “coating ”’ of the resting spores, which could scarcely have arisen except from within. The vascular tissue of the tuber, in connexion with the eye, was slightly yellow, as if discoloured. It is this last-mentioned mode of propagation, i.e. Jounson—Chrysophlyctis endobiotica and other Chytridiacece. 137 by warty seed-tubers, which, I think, accounts for the outbreak of black wart in Co. Down. Kilkeel, one of the disease-spots, is an important fishing-village on the coast, with considerable intercourse with the English and Scotch coasts, from which slightly diseased tubers could easily be imported, unnoticed by the buyer. The warty disease was reported from Germany in the autumn of 1908, and is now known from various districts in Westphalia and the Rhine Provinces. It would appear from reports in the Deutsche Landwirtschaftliche Presse (80th Sept., and 17th Oct., 1908), and the Zustrirte Landwirtschaftliche Zeitung (2nd Dec., 1908), copies of which I have received from Dr. Riehm, that, as in Cheshire so in Germany, the disease was left unnoticed and unreported for several years before the attention of botanists was called to it as an increasingly troublesome pest. ‘lhe tendency in Germany, judging from the articles, is to regard the disease as not very serious, since it is confined, it is stated, to small garden-plots, where potatoes are grown year after year in the same ground. Germany would make a serious mistake if it treated the outbreak of the disease with indifference. ‘This, I know, its mycologists will not do. It needs only a very casual acquaintance with the facts of the case in the British Isles from the time of the discovery of the trouble by Potter in 1902 to the present time to warn one of the necessity of taking all possible steps to stamp out a disease which may become as serious as ordinary leaf-blight, and less amenable to treatment.t ‘The soil, manure, and weather are not considered in Germany as predisposing causes. The disease, most marked in the variety Magnum Bonum, called locally Stiipmoll, is attributed solely to want of rotation in the crops. It is called locally canker,’ and is believed to induce cancer in eaters of diseased tubers. ‘There is no evidence of this connexion. AFFINITIES. Chrysophlyctis endobiotica is a member of the group Olpidiacese, and is allied to Asterocystis radicis, Massee regards it as a Synchytrium, basing his opinion on :— (a) The epidermal nature of the parasite. (4) The presence of an enveloping membrane round the protruding contents of the sporangium. I have already shown that, though the parasite may enter the potato, in some cases, at the epidermis, it is not its only seat, and that it may occur in a deep-seated position, even below the cork. 1 The only redeeming feature in connexion with the disease is that its germs do not, so far as we know, spread aérially in the growing crop. 2 Tn some parts of England it is called ‘‘ cauliflower.”’ 138 Scientific Proceedings, Royal Dublin Society. T have never seen the enveloping extruded membrane mentioned, in any fresh, ruptured sporangium. The wall ruptures; and the zoospores escape without any sign of the extrusion of a membrane. This is not, however, the case in artificially preserved material. I hardened a rotting piece of potato- wart, kept for some time under cultivation in potato-juice, and full of resting sporangia, in Flemming’s solution, and then stained it by. the three-colour method. On squeezing under the microscope a sporangium isolated from the general mass, its outer wall burst, and the contents as separate zoospores were extruded, enclosed by the inner wall as a hyaline membrane (Pl. IX., fig. 7). The persistence and protrusion of the inner wall as a membrane thus seem to be due to the artificial preservation of the same by reagents. If the existence of this protruding membrane in nature could be accepted, it would not indicate Synchytrium affinities. The membrane in Synchytrium is an external one, and envelops a sorus of 150 or more sporangia. It is quite different from the inner wall of the single sporangium. This no more protrudes in Synchy- trium than in Chrysophlyctis. My figure of the sporangium is, in all essentials, like that of Synchytrium taraaact (fig. 8e), figured by A. Fischer (9). His figure 8c (p. 46, 0.c.) shows the membrane enveloping the sorus of sporangia in S. Mercurialis. PREVENTION OF THE Disnase. It is stated that the disease first appeared in 1896 in Birkenhead district, and that it was brought there in Continental cattle-boats. It is now found in many different districts in England, Wales, and Scotland. it is often so pronounced as to destroy the whole crop, and itis not confined to garden-plots. Warty tubers are naturally poorer in food-matter than healthy ones, and when not destroyed in the field do not keep well in store. They ought to be destroyed as soon as found, and on no account saved for seed. The most that can be done with them is to boil and feed them to pigs. The haulms should be gathered and burnt. Infected land should be kept free from potatoes for possibly seven years. ‘The application of kainit or other potash-manure and of phosphates, the close inspection of seed-tubers, and the avoidance of low-lying or undrained land should prove beneficial in preventing a recurrence of attack. In keeping with the known aquatic habits of the Chytridiaces as a group, a wet undrained potato field would be a predisposing cause, favouring the spread of potato-wart. This is true of flax-yellowing due to Asterocystis radicis Wilde., of lucerne-canker in South America, due to Urophlyctis alfalfe, and of cabbage black-leg due to Olpidium brassice in North America, all diseases caused by Chytridiacezs. Curiously enough, I have seen only one reference to the bearing of the condition of the soil on cases of potato-wart, and have Jounson— Chrysophlyctis endobiotica and other Ohytridiacece. 189 not myself visited the Irish centres of disease. The west of Ireland, which has potato-troubles enough, would, I am afraid, prove ideal ground from the parasite’s standpoint for the spread of the potato-wart. I am trying the effects of soaking warty tubers in such fungicides as cor- rosive sublimate and formalin, as well as Bordeaux mixture; but I think it would be better to destroy the tubers outright when the tumour is obvious. In some cases the tumour could be broken off by its narrow neck of attachment and the potato eaten, though even this might prove false economy. Gras-lime applied in May or June, and sulphur mixed with the soil, in probably non- paying quantity, destroy the pest in the soil. Brutr-rumour, Urophlyctis leproides. Another Chytridian of considerable biological interest causes tumours on the bulb or root of the beet’ or mangel, Beta vulgaris var, rapacea, in Algiers, where it was discovered in 1894 by Trabut (10), and was called by him Entyloma leproides. Saccardo regarded it as a Smut, and placed it in the Ustilaginacese as Gidomyces leproides—a view supported by Prillieux and others. Magnus, however, in 1897, in the “Annals of Botany,” records it as a Chytridian under the name of Urophlyctis leproides (11); and, curiously enough, falls himself later into the error of ascribing potato-wart to it, as do Massee and others. Urophlyctis is an interesting genus which causes aérial galls on some plants, such as clovers, Umbelliferes, and Chenopodiaces, and root-galls on others, e.g., lucerne and beet. The gall is in some a much-enlarged epidermal cell, almost surrounded by many layers of other cells. In others the cavity containing the spores is added to by the ramification of the original cell through the gall-tissue, accompanied or not by sieve-like perforations or by complete absorption of the walls of the adjoining attacked cells. Ac- cording to Prillieux (12), in Urophlyctis leproides, the cavities containing the spores are formed by the hollowing out of the thin-walled parenchyma, forming the fleshy tumour in which the spores accumulate. It is assumed, I think naturally, that Prilliewx means to imply that the hollowing out is due to the breaking down of the walls of the surrounding cells. Magnus does not accept this view of the origin of the cavities in the flesh of the tumour. He considers that the cavities or cysts in a swelling or lobe of a tumour are all connected and derived from one original cavity or host-cell which sends out, when invaded, processes. These ramify through the substance of the swellings, remain in connexion with the parent cell, and 1 The beet-tumour I examined from Co. Wexford showed no spores, and was not, as Lat one time thought (Jour. of Dept. of Agric., i., 1902), caused by U. leproides. The statement got into print before I could suppress it. SCIENT, PROC. R.D.S., VOL, XII., NO. XIV. 2a 140 Scientific Proceedings, Royal Dublin Society. form numerous other spore-containing cavities by enlargements of the processes. One must imagine a beet-cell when attacked sends out living diverticula, which make a passage for themselves between the other surrounding host-cells (stimulated thereby to division), and sooner or later dilate to form the cavities or cysts, the whole being thus one malformed host-cell. ‘Through the kindness of M. Trabut, I have been able to examine some of the original material, and find myself unwillingly compelled to differ from Magnus. Serial sections show the layer or layers of cells between adjoining cavities broken or in the act of breaking down. Thus the cavities containing the spores are due, I take it, at any rate in part, to the disappearance by absorption of the surrounding host-cells. ‘They are therefore lysigenetic in origin, as is the case, to a certain extent, in Urophlyctis pulposa (Wallr.) Schroet., and U. major Schroet., and still more so in Urophilyctis Riubsaameni Magn. (13). This is a point of con- siderable interest. In 1895 von Lagerheim reported the occurrence of galls on the roots of Lucerne (Medicago sativa) in Keuador. These galls were caused, he stated, by the same fungus as that responsible for the beet tumour. In the ‘Berichte der deutschen botanischen Gesellschaft” (14) for 1902, Magnus describes the same root-galls on lucerne material received from Behrens from Colmar in Alsace, and finds certain points in which the lucerne gall differs from the beet one, sufficiently important to justify him in describing the lucerne fungus as a distinct species under the name of Urophilyctis alfaife (von Lagerheim olim) Magnus. The chief distinction is that in U. alfalfe the walls between the cavities in the gall are broken down, thus putting adjoining originally independent cavities into connexion ; while in U. Jeproides, as already stated, no such dissolution of septa, according to Magnus, occurs. My material, however, shows such collapse of the separating partitions, and thus U. alfalfe as a distinct species must rest, if I am right, on other grounds. Interesting in this connexion is the illustrated account by 8. Kusano (15) of a cyto-pathological study of Synchytrium puerarie. The parasite is chemotactically attracted to the deeper tissues of the host, avoiding, though a Synchytrium, the epidermis and the green cortical cells. Its plasmodium develops in the host-cell. Both increase much in size. The surrounding host-cells lose their walls and form a symplast, enveloping the developing Synchytrium sorus, which, at first intracellular, in the end lies in a lysigenetic intercellular space. Oriain oF Restine Spores In Urnoputycris. The resting spore of Urophlyctis is a smooth, thick-walled body, com- parable to a plano-convex lens in section. From the centre of the flattened side a small papilla may project. There are two distinct views as to the origin of these spores. Magnus has described in U. leproides and other Jounson —Chrysophiyctis endobiotica and other Chytridiacee. 141 species the presence of two distinct mycelia—one giving small male organs or antheridia, the other, larger female receptive organs or oogonia. The antheridium becomes connected by a narrow canal with the oogonium, and fertilization occurs. The resting spore is thus an oosperm or zygote. Fischer, on the other hand, says in Urophlyctis there is no act of fertilization. The resting spore is the terminal cell of a hypha which also forms, just below it, an accessory or collecting, more or less swollen cell, which becomes less conspicuous as the resting spore develops. Magnus’s views on U. leproides are based, he states himself, on the spirit material he received. I have not seen any illustration of the female mycelium he describes beyond the indistinet structure he shows (op. cit., figs. 29 and 33, U. leproides) and calls “the remains of the carrying threads [or mycelium] of the receptive cells.” In his paper on U. Riibsaameni Magnus considers ‘as the remains of the host-cells’ tufts of short appendages described as protoplasmic threads or appendage-threads by Schroeter and Biisgen. Theso occur often at the same point on the convex side of the resting spore in JU. Rubsaameni, as his remains of the female mycelium in JU. leproides. If I am correct in my observations that in U. leproides the host-cells become disintegrated as the tumour enlarges, may not the U. leproides remains also be those of its host-cells? In the examination of the spirit-material of U. leproides at my disposal, I have seen nothing to support the view of the sexual origin of the resting spores in Urophilyctis. FLax- YELLOWING. The flax crop, almost wherever it is cultivated, is liable to be troubled not only by the yellow rust Melampsora lini, but by a more serious form of yellowing, called in Belgium / few or la bridure, due to a member of the Chytridiacess, Asterocystis radicis Wild., the life-history of which, so far as it is known, was made out by E. Marchal (16) in 1892. I have received flax-yellowing due to it for report from various parts of Ulster. The disease appears in wet, low-lying parts of the field, and is indicated by the premature yellowing of the seed- and other lower leaves of the young plant. The stem loses its firmness, and its tip topples over towards the ground. The roots have a peculiar glassy, flabby appearance, and are easily pulled out of the ground. The parasite is confined to the roots, attacks the root-hairs and the roots generally, leaving only the central cylinder untouched. ‘Thus the yellowing of the foliage is caused by the choking up and destruction of the absorbing tissues of the roots, i.e., to starvation of the plant by the interference of the parasite. ‘his occurs as a plasmodium in the host-cells, and passes, in a way at present conjectured only, from cell to cell in the root, reproducing itself by swarm-sporangia like those of other Chytridiaces. These sprout 242 142 Scientific Proceedings, Royal Dublin Society. readily in water, and form the usual uniciliate zoospores, which in wet soil infect neighbouring roots, probably through the root-hairs. Marchal found, e.g., that the zoospores spread the disease a distance of 20 cms. (8 inches) in all directions in 24 hours. Resting sporangia are common in older plants, and have been shown to be the means of propagating the disease from season to season, though their actual germination has not yet been observed. Flax-yellowing is a disease of youth. The critical period of attack varies between the thirteenth and twenty-fifth day after the time of sowing, and is partly dependent on the weather, especially the temperature. Asterocystis occurs also on the roots of grasses, and of many weeds found in a flax-field. Its eradication depends mainly on crop-rotation (in which flax comes once in seven or even ten years), helped by the pulling up and burning of the diseased plants. Flax soil, says Marchal, should not receive excess of nitrogenous manure, but a sufficiency of phosphoric acid. The extensive manurial experiments carried out by the Irish Department of Agriculture each year since 1901 demonstrate “that as regards artificial manures potash should be the dominant, if not the sole, manurial ingredient in those used for flax.” Muriate of potash, at the rate of 1-114 cwt. per statute acre, mixed with sand, fine soil, or sawdust, is recommended for application (17). The Department finds that this potash manure prevents flax-yellowing in Ireland, and, as a result, for the last two or three years very few cases have come in to me for report. Marchal, however, found the zoospores of Asterocystis bore easily a dose of 0:02 per cent. (1 in 5,000 parts) of potash in their food solutions or in soil. Marine CHY?RIDIACER. T have collected from various parts of the coast of Ireland alge infested with Chytridians. One of these, on a species of Ectocarpus, was described by Dr. I. P. Wright (18) as Rhisophidium Dicksonii. In his monograph of the group, A. Fischer says of this species the sporangia are of very abnormal form; and the relation of the parasite to its host seems to deserve fresh investigation, as its sporangium appears, from the figures, to lie partly within the host-cell. Rbizophidium Schenk shows an external sporangium and an internal rhizoidal or haustorial mycelium. My intention to act on Fischer’s suggestion by an examination of the collected material was anticipated by P. Magnus, who gives the results of his examination in Hedwigia in 1904, N. Wille having already in 1899 decided that the plant was not a Rhizophidium, as it possesses no signs of a mycelium. He called it Olpidium Dicksonti. Magnus, it seems, had the plant before him in 1875 (two years before Wright’s account appeared), when describing the Marine Olpidess, but left it unnoticed for the time being. He now regards it as Jounson—Chrysophlyetis endobiotica and other Chytridiacee. 143 the representative of a new genus which he calls Hurychasma (Eurus broad, kasma a slit). It lives entirely in the interior of the host-cell attacked, does not produce rhizoids (i.e., it has no intramatricular mycelium), and its whole vegetative body becomes converted into a zoosporangium. It is thus holocarpic. Resting spores are unknown. It differs from the other Monolpidiacess in which Wille placed it by the mode of dehiscence of its sporangium. ‘This splits open and forces apart the cell-wall of the host-cell, projects through the opening made, sends out one or two slight prominences, the ends of which open to allow its zoospores to escape. Hurychasma Dicksonii (Wright) Magnus is the only species. It is very widely distributed, and has been recorded from the coasts of Ireland, Scotland, Norway, and from the Adriatic. It grows on many species of Ectocarpus, on Py/aiella littoralis, and on Striaria attenuata. WHauck found it in localities in the Adriatie where the sea-water was not pure. Fischer states that the Chytridians prefer pure water, and that when, e.g., cultures of decaying algeo which Chytridians are destroying begin to smell badly, the Chytridians themselves are soon destroyed. Macerations are, however, recommended by several observers for the germination of the resting spores. Olpidium sphacellarum Kuy is another marine form not uncommon on species of Sphacelaria from various parts of the Irish coast, and hitherto unrecorded. It occurs in the apical cells of the main and side shoots of Sphacelaria and Cladostephus. Its sporangia, one to nine in each cell, are rounded or flattened by pressure, have a smooth wall and a quite short projecting emptying-neck. Resting spores are unknown. The host-cells attacked cease dividing, become club- or pear-shaped, and their contents, when not absorbed, turn brown or blackish. No group of plants has more frequently led to error than the Chytridians which simulate the reproductive organs of the marine alge. I have to confess myself at fault (19) in describing a Chytridian, possibly this Olpidium, as the sporangia of Desmarestia ligulata, which thus still awaits the discovery of its reproductive organs. The following list of Chytridiaces on marine alge, given by A Fischer (op. cit., p. 26), may be of use to Irish students :— Olpidium bryopsidis on Bryopsis. i aggregatum in Cladophora. sphacellarum Kny. Ks entosphericum in Bangia and Hormidium. ss tumefaciens in Ceramium. » plumule in Antithamnion plumula. 144 Scientific Proceedings, Royal Dublin Society. 10, 11. 14. 15. 16. Wo 18. 19) List oF AUTHORITIES QUOTED IN THE Text. M. C. Porrer: A new Potato Disease (Chrysophilyctis endobiotica), Journ. Board of Agric., ix., Dec., 1902. . Department oF AGRICULTURE AND ‘TEcHNICAL INSTRUCTION FOR Trevanp, Leaflet No. 91, ‘ Black Scab in Potatoes.” . A. W. Borruwick : Warty Disease of Potato. Notes from the Royal Botanic Gardens, Edinburgh, No. xviil., pl. xxii., 1907. . H. 8. Satmon: Black Scab or Warty Disease of Potatoes. (Leaflet, 8. E. Agricul. College, Wye.) . K. Scui.Berszky: Hin neuer Schorfparasit der Kartoffelknollen. Ber. d. deutsch. bot. Ges. xiv., 1896, p. 36. . I. Jonnson: Letter to Nature (vol. Ixxix., Nov., 1908) announcing success in making resting spores of Black Scab germinate. . G. Massee: Linnean Society (Dec. 17th, 1908). Proceedings of Meeting: Exhibit of Black Scab, with notes. . T. Jonnson: Spongospora solani, Branch. Heonom. Proc. Roy. Dublin Soe., 1908. . A. Fiscuer: Phycomycetes, 1892, in Rabenhorst’s Kryptogamen-Flora, Abt. iv., Bd. 1. Traut: Gdomyces leproides (Trab.) Sace. Rev. Gén. de Bot., vi., 1894. P. Macenus: On some species of the genus Urophlyctis, with plates vii. and viii., Annals of Botany, xi., 1897, p. 87. 12. Priniievx: Maladies des Plantes agric., tom. i., 1895, p. 198. . P. Macnus: Ueber eine neue unterirdisch lebende Art der Gattung Urophlyctis, mit Tafel xxvii. Ber. d. deutsch. bot. Ges., xix., s. (145), 1901. P. Maenus: Ueber d. in d. knolligen Wurzelauswiichsen der Luzerne lebende Urophlyctis, mit Tafel xv. Ber. d. deutsch. bot. Ges. Bd. xx., s. 291, 1902. S. Kusano: On the Cytology of Synchytrium. Centralb. f. Bakt. Abt. ii., Bd. 19, 1907. HE. Marcuat: Iecherches biologiques sur une Chytridinée parasite du lin., Bull. de Vagric. Belg., 1901. DeEraRtMENT oF AGRICULTURE AND ‘TkcHNIcAL InsrrucTION FoR Jretanp. Flax experiments, Leaflet No. 52, 1906. i. P. Wriewr: On Chytridia parasitic on Ectocarpus. Trans. Roy. Trish Acad., xxvi., Pl. vi., 1877. ‘I’. Jounson: Observations on the Pheeozoosporese. Annals of Bot., v., 1890. EXPLANATION OF PLATE IX. Fig. Fig. Fig. PLATE IX. . Potato-tuber showing pronounced wart at toe or crown. The malformed sprouting eye and roots may be seen in the midst of the distorted tissue. (Natural size.) . Section through the convoluted surface of the wart showing numerous ‘resting spores.’ (Magnified.) . A macerating piece of the wart showing numerous ‘resting spores’ as black dots. (Slightly magnified.) . Microphotograph of a single resting sporangium, showing the angular wall and the separate zoospores. . Vertical section through a diseased eye . The same more highly magnified. . An artificially ruptured sporangium, from Flemming-fixed material The individual zoospores are clearly recognizable. The mass of zoospores is enclosed by the inner wall as a membrane, and to the left the dark edge of the ruptured outer wall is just visible. PILeveS IDX. Scient. Proc. R.Dubl. Soc, N.S, Vol. X01. Bemrose Lt4, Derby. HXPLANATION OF PLATE X. PLATE X. Figs. 1 and 2. Microphotographs of the same sporangium at two different levels. In fig. 1 the zoospores are just recognizable as distinct. In fig. 2 the slit-like rupture of the wall of the sporangium is seen. Figs. 8 and 4. Microphotographs of vertical sections of a young wart convolution showing early stages of infection and of intracellular sporangium. n is the host-cell nucleus. p is the parasite. Fig. 5. Microphotograph of two resting sporangia, showing their central vacuoles normally occupied by one or more fat globules. In the sporangium to the right the multinucleate condition is clear. Scient. Proc. R. Dubl. Soc.,N.S., Vol. XI. Plate X. Bemrose L'4 Derby EXPLANATION OF PLATE XI. Fig. Fig. Fig. Fig. Fig. PLATE XI. . a, b, ¢. Three different host-cells, showing their nuclei and the intra- cellular parasite. . A resting sporangium in development. Wall and fat globules omitted. (x 2000.) . A ripening resting sporangium, showing the zoospores forming, and three fat globules. . A resting sporangium cultivated in potato juice. The individual zoospores are shown. . Aruptured sporangium from potato-juice culture. A few laggard zoospores are shown. . Escaped zoospores, showing amcboid changes and the single cilium. (x 2000.) Goospores seen in the resting sporanginm. Flemming-fixed and stained material. PLATE XI. SCIENT. PROC. R.DUBL.SOC. N.S., Vol. XII. West, Newman lith. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 15. JUNE, 1909. THE SCANDINAVIAN ORIGIN OF THE HORNLESS CATTLE OF THE BRITISH ISLES. BY JAMES WILSON, M.A., B.Sc., PROFESSOR OF AGRICULTURE IN THE ROYAL COLLEGE OF SCIENCE, DUBLIN. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PU SEED BYE ROYAL. DUBIN SOC avy, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price One Shilling. Av conan | AStrt ¥\ A. ike Ue a Utip, n, h UY \\¥ Oe THE SCANDINAVIAN ORIGIN OF THE HORNLESS CATTLE OF THE BRITISH ISLES. By JAMES WILSON, M.A., B. Sc., Professor of Agriculture in the Royal College of Science, Dublin. [Read Marcu 23. Ordered for Publication Aprit 6. Published Junr 19, 1909.1 THE prevailing opinion among those who have dealt with the subject is that most of the British hornless breeds of cattle have originated inde- pendently. Those who think that cattle were hornless before they were horned hold that the hornless breeds have originated in reversions to the earlier hornless condition; while those who think that cattle were horned before they were hornless hold that the hornless breeds have originated in independent “spontaneous variations,” as Darwin called them. Darwin’s view, which is chiefly held, may be gathered from the following quotations :— “The aboriginal species from which our domesticated cattle and sheep are descended no doubt possessed horns, but several hornless breeds are now well established.” “Tt is possible that some breeds, such as the semi-monstrous niata cattle, and some peculiarities, such as being hornless, &c., have appeared suddenly from what we may call a spontaneous variation ; but even in this case a rude kind of selection is necessary, and the animals thus characterized must be at least partially separated from others.” “No one can give any explanation—although no doubt there must be a cause—of the loss of horns, any more than of the loss of hair, both losses strongly tending to be inherited. It is, I think, possible that the loss of horns has occurred often since cattle were domesticated, though I can call to mind only a case in Paraguay about a century ago.’ 1«¢ The Variation of Animals and Plants under Domestication,’’ 1868, vol. ii., p. 29. 2 Tbid., vol. i., p. 92. 3 Letter to Messrs. Macdonald and Sinclair, published “in their ‘‘ History of Polled Aberdeen or Angus Cattle,”” 1882, p. 12. SCIENT. PROC. R,D.S., VOL. XII,, NO. XV. 2B 146 Scientifie Proceedings, Royal Dublin Society. Against both views there are several considerations, viz.: — (a) It is unlikely that precisely the same variation—or reversion—should have occurred in at least a dozen separate districts in the British Islands. (b) It is likely that, if it occurred so frequently in Britain, it should also have occurred with similar frequency on the neighbouring continent, especially in the ow Countries, where the.cattle are closely allied to those in Britain. (c) It is unlikely that a phenomenon believed to have occurred so frequently in times gone by should now be unheard of. A hornless calf from pure-bred parents of any horned breed is never seen. Many are got by the crossing of horned with hornless cattle; and, because of the hornless condition being dominant to the horned, it isa very simple matter to make a horned breed hornless by persistent selection of the hornless descendants of such a cross for breeding purposes. A horned breed has hornless calves in no other way. It would be very difficult indeed to prove that the niata cattle referred to by Darwin and the hornless breeds in some countries alleged to be descended from pure-bred horned ancestors have been abso- lutely free from a hornless cross. It is the purpose of this paper to show that the existing and extinct British breeds of hornless cattle may all be traced back to a Scandinavian origin. Breeds of hornless cattle are reported to have existed in the eighteenth century in the following centres :—Suffolk, Yorkshire, Forfarshire, Aberdeen- shire, Sutherland, Skye, Galloway, Somerset, Devon, and parts of Ireland. A few hornless cattle were reported in other districts; but the smallness of their numbers suggests that they were migrants from the above-named centres. OF the herds of so-called “wild”? white cattle at least seven were hornless at one time or another—viz.: at Cadzow in Lanarkshire, Ardrossan in Ayrshire, Middleton and Whalley in Lancashire, Somerford in Cheshire, Wollaton in Nottingham, and Gisburne in Yorkshire. Leaving aside the “ wild” herds, which will receive separate consideration later, it will be noticed that the districts in which hornless cattle were recorded are all maritime and situated directly upon the Norsemen’s tracks—cireum- stances which at once suggest that these hornless cattle are of Scandinavian origin. To make this theory good, it will have to be shown— (1) That these hornless breeds were originally of similar character, and therefore presumably of common origin, Witson—Seandinavian Origin of Hornless Cattle of British Isles. 147 (2) That their arrival in the British Islands coincided with that of the Norsemen. (3) That cattle of similar character were taken to other parts of Hurope with which the Norsemen were associated. (4) That at least traces of cattle of similar character are to be found in Scandinavia. Let us begin by collecting the available information as to the first of these considerations :— Suffolk.—The Suffolk cattle used to be called “the Suffolk Duns.” Their original shade of colour, as may be inferred from a letter from Sir Thomas Beevor in the Bath Society’s “ Letters and Papers,’’! was what is now called light dun among Highland cattle, a kind of steely or slaty whity-grey : ‘The cows you saw were bred from the polled or horn-less Suffolk dun-coloured? cows . . . by a Derbyshire black and white bull, given me by my friend Lord Townshend. This mixture produced their uncommon colour of mouse and white ”’—a colour which is the intermediate hybrid of light dun and black, and whose only source is the crossing of these colours.? In John Kirby’s “Suffolk Traveller,” published in 1735, the Suffolk Dun is described as having ‘“‘a clean throat, with little dewlap, a snake head, thin and short legs, the ribs springing well from the centre of the back, the carcase large, the belly heavy, the back-bone ridged, the chine thin and hollow, the loin narrow, the udder square, large, and loose, and creased when empty, the milk veins remarkably large and rising in knotted puffs ; and this so general, that I scarcely ever saw a famous milker that did not possess this point, a general habit of leanness, hip-bones high and ill covered, and scarcely any part of the carcase so formed and covered as to please an eye that is accus- tomed to fat beasts of the finer breeds.” Youatt (1884) writes :—“The Suffolk Dun used to be celebrated in almost every part of the kingdom, on account of the extraordinary quantity of milk that she yielded. The dun colour is now, however, . . . rarely seen in Suffolk. . .. The prevailing and the best colours are red, red and white, brindled, and a yellowish cream colour. ‘he bull is valued if he is of a pure and unmingled red colour.’’ The Suffolk Duns were crossed with other cattle, but chiefly with the Norfolk 1 Vol. iii., second ed., 1788, page 280. 2 Since the above was written the following statement has been found in Culley’s ‘‘ Observations upon Live Stock,” second ed., 1794, p. 66: ‘*'The Suffolks are almost all light dun.” * See Roy. Dublin Soc. Proc., vol. xit., No. vir1., 1909: ‘‘'The Colours of Highland Cattle.’’ * Quoted from Youatt’s ‘ Cattle,’’ 1834, p. 174. 5 Ibid., p. 175. 282 148 Scientific Proceedings, Royal Dublin Society. red cattle which were horned. From this cross yellow' coloured polled cattle were born, and from subsequent crosses red polled cattle. By persistently breeding from the red polled stock thus produced, the Suffolk breeders carried their cattle through a yellow stage to the final red of the present day. At the same time, by crossing with the hornless Suffolk Duns, the Norfolk horned eattle lost their horns. The two breeds are now one. Yorkshire—The “Northern or Yorkshire Polled Cattle’ never seem to have occupied an extensive territory. ‘luke, in his “Agriculture of the North Riding of Yorkshire,” published in 1800, says that “Henry Peirse, Esq., of Bedale, has a breed of very large hornless cattle,” and he publishes a drawing of a “polled Teeswater cow belonging to Richard Raisen, Bishop- thorpe’; but the centre of the breed seems to have been in the Hast Riding, most likely in Holderness. John Lawrence describes them in 1805?:— “These have the same qualities as the short-horned cattle, carrying vast substance, and some I have seen lately are of great size, although in that particular they are most conveniently various. In my. opinion, they are a most excellent breed, and well merit improvement, with the view of labour, by a selection of the finest-boned and most active individuals. From the shape of these polled cattle, they hold a strict affinity in all respects with the short-horned’ amongst which they are found; and it seems that various breeds of cattle are attended with hornless but pefectly congenial varieties. The above, for example, and the polled galloways of Scotland, of similar shape and quality with the long horns, also the Devon natts, or polled cattle on the coast.” In Strickland’s “Agriculture of the East Riding of York- shire,” published in 1812, we are told that the “ Original Holderness breed . is distinctly marked by its colour, being variously blotched with large well-defined patches of deep red or clear black, in some families of dun or mouse-colour on a clean white ground; they are never of a brindled or mixed, and rarely of one uniform colour.”” Here again, as with Sir Thomas Beevor’s Suffolks, the dun colour is evidence of a light dun ancestry. Another matter of importance, as will be seen when Aberdeenshire cattle are dealt with, is Lawrence’s statement that these Yorkshire cattle were “most conveniently various” in size. er At the end of his chapter on polled cattle, and immediately after referring to those of Yorkshire and Devon, Youatt! draws attention to another point of some importance: ‘“‘ Many breeders pay particular attention ' See Roy. Dublin Soc. Proc., vol. xu., No. vitt., 1909: ‘‘ The Colours of Highland Cattle.’’ 2 « General Treatise on Cattle,” &c., p. 71. 3 i.e. of 1805. 4«< Cattle,” 1834, p. 179. Witson—Seandinavian Origin of Hornless Cattle of British Isles. 149 to the shape of the head in these polled cattle, and, to a certain extent, also, in the horned ones, If the crown of the head is fine like that of a doe, and drawn almost to a point on the top, the breed is supposed to be good.” Durham.—The legend goes that the monks who were carrying St. Cuth- bert’s body to its last resting-place set it down and couid not raise it again. ‘‘ Whereupon they fasted and prayed three Days with great Devotion, to know by Revelation from God, what to do with the holy Body, which was soon granted to them, it being revealed to Hadmer, a virtuous man, that he was to be carried to Dunholme, where he was to be received to a Place of Rest. ‘They were again in great Distress, in not knowing where Dun- holme lay; but as they proceeded, a Woman wanting her Cow, called aloud to her Companion, to know if she had seen her? Who answered, She was in Dunholme.” The position of Dunholme being thus revealed, St. Cuthbert’s body was buried there; and a figure of the cow was “ carved on the north-west corner turret of the Nine Altars or eastern chapel of the cathedral about the year 1300.’? Our present interest in this cow is that she was reputed to have been dun :—- “The dun cow’s milk Makes the prebend’s wite go in silk.” Repute, however, may not be fact, especially in connexion with an animal dead so many hundred years. But if the words of the rhyme are not evidence that the original cow was dun, they are evidence that, when they came into vogue two or three centuries ago, there were dun cattle in Durham. There were both dun and yellow cattle in Durham in the eighteenth century. About 1777, Mr. Hutchinson of Smeaton “possessed himself of a large yellow cow with some white.” Mr. Thomas Hutchinson wrote of her in 1821: “ She might have been descended (for anything I know to the contrary) from the old woman’s propitious dun cow.... ‘hey were nearly all of the same colour.’” And there is some presumption that the original figure of 1300 was horn- less. Having become broken and effaced, it was “restored” about 1778. ‘The restorers saw the original figure hornless; but, believing that the horns had 1 Quoted in Bygate’s ‘‘ Cathedral Church of Durham,”’ 1899, from Sanderson’s ‘‘ Antiquities of Durham.” __ 3 : P 2 Bates’s “ Thomas Bates and the Kirklevington Shorthorns,”’ 1897, p. 25. 3 Tbid., p. 46. H j : E 150 Scientific Proceedings, Royal Dublin Society. been worn off by time, “the horns were made this time of lead, lest she should ever again be reduced to the condition of a polled beast.” drawing of the original figure shows a hornless cow.’ Grimm ’s The Durham Cow. On the other hand, Dean Kitching informs us that “ there is at the British Museum a map of Durham City, a.p. 1594 [probably]. In the corner stands the Dun Cow—with a pair of fine horns!” The North-east of Scotland —Although the evidence is fullest as regards Forfarshire and Aberdeenshire, because these two counties took the chief part in evolving the present Aberdeen-Angus breed, there is little doubt but that hornless cattle have existed for centuries round the north- east coast from Forfarshire to Morayshire, probably farther. The earliest evidence of hornless cattle in the north-east of Scotland has not hitherto been seriously noticed; but, in view of the other evidence to be brought out in this paper, it may now be taken with confidence. It consists of about a dozen stone slabs, bearing chiselled-out figures of bulls, dug up on the shores of the Moray Firth, chiefly at Burghead in Morayshire, which at one timo was a Norse or Danish stronghold. ‘The surfaces of these stones are so worn down by time that the lines of the figures cannot be followed easily by the naked eye, but they can still be brought out by careful rubbings on paper. he difficulty has been to determine whether the lines rising from the bulls’ heads were meant to represent horns or not; but a careful examination of a number of these rubbings shows that 1 Bates’s ‘‘ Thomas Bates and the Kirkleyington Shorthorns,”’ 1897, p. 45. 2 See Hutchinson’s ‘‘ History of Durham,”’ 1785, vol. ii., p. 226. Witson—Scandinavian Origin of Hornless Cuttle of British Isles. 151 some of the bulls were hornless. From the following illustration’ of Burg- head bulls, it will be seen at a glance that one of the bulls is hornless, and the other is horned. © Q = G ‘Si = ‘A iB SS Burghead Bulls. 4 aE = From the following illustration of a polled Angus ox copied from Youatt’s book,’ it will be seen how difficult it would be to determine whether the lines projecting from the ox’s head are ears or horns, were they carved in stone rather than drawn in pen-and-ink, Pine PE OT fe “ ( SM , Angus Ox. The evidence next in chronological order is to be found in a legal document dated 1523, “Instrumentum sasine in favorem Johannis Cumying,” in which it is recorded that the lands of Culter in Aberdeenshire passed from one man’s possession to another by the new owner receiving not the usual token, namely a handful of earth and a stone, but “unum bovem nigrum 1 Copied from “the Sculptured Stones of Scotland,’ plate xxiii., vol. ii. 2 “ Cattle,’? 1834, p. 168. 152 Scientific Proceedings, Royal Dublin Society. hommyll appretiatum ad quadragintas solidos et octo denarios monete Scotie”* —a black hummle—i.e. humble or hornless—ox, valued at 40s. 8d. Scots. There is another gap till the middle of the eighteenth century, from which time onwards the references to hornless cattle in the north-east of Scotland occur with increasing frequency; but as these are collected in Macdonald and Sinclair’s “ History of Polled Aberdeen or Angus Cattle,” all that need be said here is that these cattle were confined originally to the lands near the coast, but as time went on, and especially after the opening up of the southern trade in the eighteenth century and the increased demand for hornless cattle, they crept farther and farther inland. Youatt has several references to the colours of these North-Hastern cattle during the time when the inland ones were passing from the horned to the hornless condition, viz. :—Of the Forfarshire horned cattle: “ The prevailing colour is black, but with more admixture of other tints: some have white spots on the forehead, and white on the flanks and belly. There are more brindled cattle than in Aberdeen ; some are dark red, and others of a silver yellow or dun. A few are black with white hairs intermixed; and occasionally a beast is seen that is altogether white, with the exception of a few black hairs about the head.” Of the Forfarshire hornless cattle: “The greater part of them are black or with a few white spots. ‘lhe next general colour is yellow, comprehending the brindled, dark red, and silver-coloured yellow.’® The colour of the Aberdeenshire horned cattle ‘“‘is usually black, but sometimes brindled.”* Many were dun within the memory of living men. Of the ‘‘ Buchan cattle,” that is the cattle on the Aberdeenshire coast, “the best sort used to be polled, and some of them that do not begin to have the Ayrshire blood in them, are so still, and are of a dark or brown colour.”® Macdonald and Sinclair write: ‘Formerly, both in Angus and Aberdeen, the breed embraced a variety of colours as well as difference in size. Black, with some white spots on the underline, was the prevailing colour. Some were brindled—dark-red and black stripes alternately ; others were red; others brown; and a few what Youatt called ‘silver-coloured yellow.’”® The following statement, apparently about a century old, is made with regard to the cattle in the parish of Rathven in Banffshire :—‘‘ Dealers who purchase the cattle for the south are somewhat particular with regard to 1 See ‘The Spalding Club’s Collections for a History of the Shires of Aberdeen and Banff,” vol. iii., p. 644. 2 « Cattle,’ 1834 ed., p. 114. 5 Ibid., p. 167. 4 Thid., p. 106. © Ibid., p. 107, 6 « History of Polled Aberdeen or Angus Cattle,’’ 1882, p. 76, Witson—Sceandinavian Origin of Hornless Cattle of British Isles. 153 what they call points of form and colour. These points are short legs, a fair proportioned round body, straight along the back, and, in their third year, a long, slender white horn tipped towards the point with black. The favourite colour is pure black. The brindled ranks next in esteem, and the dun is not disliked. Pure white or streaked are counted inferior,”? Note the presence of the two colours yellow and dun among these north- eastern cattle as among those of Suffolk and Yorkshire. Here again we have cattle descended from light dun ancestors. From a statement printed in Messrs. Macdonald and Sinclair’s book,? and communicated to them by Mr. William Forbes, an Hast Aberdeenshire farmer, we may gather some further particulars as to the early character of the cattle on the east coast of Aberdeenshire; and Mendelian students may see how size is affected when a small breed is crossed by a large. There is a system of cattle-breeding well known in America at the present day, and widely practised at one time in the British Islands, called grading. For instance, the Americans put Shorthorn bulls to what are called “native” eattle again and again in succeeding generations, and, by this process, the native cattle eventually become Shorthorns. By the continued use of Long- horn bulls the old native cattle of Ireland were made Longhorns in the eighteenth century. In the same way the cattle of the north-east of Scotland were changed in some of their characters. The process was begun in the eighteenth century and continued well into the nineteenth: the breeds used having been Longhorns and other large English cattle, Fifeshires, and, latterly, Shorthorns. Mr. Forbes’s statement is as follows :—“ The cattle in Buchan about half a century ago and earlier might be said to have consisted of horned and polled black cattle in about equal proportions. The polled cattle were of two classes, one large and another small. I knew the smaller kind well. They were rather puny creatures, always thin in flesh, and very badly used. ‘They were pre-eminently the crofter’s cow, as they were able to live through the winter on the straw of oats and bere, if necessary. Ofthe larger portion of the cattle, about one-half were jet black, excepting the udder, which was usually white, and often the whole underline was white. They could not stand starvation so well as the small polls, but with better treatment they gave a heavier yield of milk, When creamed, however, their milk was thinner than that from the small cows. A considerable portion of the cattle were large-sized, well-fleshed, brindled polls; and these were the finest-looking animals in Buchan. When well fed, they had a 1 See an article on ‘‘Aberdeenshire Horned Cattle,” by Mr. James R. Barclay, in the “Transactions of the Highland and Agricultural Society of Scotland,” for 1906, p. 204. * « History of Polled Aberdeen or Angus Cattle,” 1882, p. 72. SCIENT, PROC. R.D.S,, VOL, XII., NO. XV. 2c 154 Scientific Proceedings, Royal Dublin Society. short, glossy coat of hair; some were good milkers, but some went to flesh and fat instead of milk. A few were of a dull red colour, but they were not so high in favour as the brindled cattle. ‘The polled cattle were the dairy stock. ‘The butter they produced was fine in summer and autumn, but hard and white in winter. ‘The establishing of a beef trade with England, and the introduction of Shorthorn bulls and turnip husbandry, opened up a new era for Buchan. ‘The native cattle fattened well, and money was made by doing so. Shorthorn bulls were introduced and put to all kinds of cows. Often when a Shorthorn bull was mated with a small polled cow, the produce was a black poll of the finest character—immensely superior to either of the parents. When a heifer of this stamp was again put to a good Shorthorn bull, the result was quite as fine a black poll, of still larger size. If the produce were also a heifer, and mated with a pure Shorthorn bull, the produce was still a poll, yet larger in size, but bluish-grey in colour. If a heifer again, and put to a Shorthorn bull, the produce was once more a grey poll, probably lighter in colour. When this form of crossing was continued further, Shorthorn colours appeared, sometimes with scurs, but oftener with the regular short horns of the male parent. I observed this experiment tried in several cases, with exactly the same result. With the larger polls with white underlines, the horns and colour of the shorthorn bull were earlior transmitted to the produce, generally at the second or third crosses. I therefore look upon the small polls without white spots as the pure original Buchan Humlie.” The points of immediate importance in Mr. Forbes’s statement are that the original Hast Aberdeenshire hornless cattle were small, puny, thin-fleshed dairy cattle, whose milk, so far as the fat was concerned, was apparently similar in character to that of the modern Jersey. These cattle were crossed by larger cattle from the south, and eventually lost nearly every character they possessed excepting their hornlessness; or, it might be put the other way: that the intruding cattle retained the characters they brought with them excepting their horns, which they lost by crossing with the Buchan Hunlies. Sutherlandshire.—It is only from almost a chance remark of Pennant’s that we know of the existence of hornless cattle in the northern counties. He states’ that ‘Sutherland is a county abounding in cattle, and sends out annually about 2500 head, which sold about this time (lean) from 2/. 10s. to 3/. per head. ‘These are very frequently without horns, and both they and the horses are very small.” Yet the probability is that there were hornless cattle round the northern 1 « Tour in Scotland,”’ 3rd edition, 1774, p. 170. Wirtson—Seandinavian Origin of Hornless Cattle of British Isles. 155 coasts just as in Aberdeenshire, and that they occupied even a considerable area in Caithness. What were called the “native ” cattle there bear a strong resemblance to those of the Aberdeenshire coast. “Sir John Sinclair had a large property in Caithness: he observed and lamented, and materially suffered by, this wretched state of the cattle, and thought of many plans for their improvement. He first tried what he could do by crossing the native breed. ‘he chest was small, and the ribs flat, and the back thin; there was not room for the heart to beat, nor the lungs to play.” Again: ‘“‘ Oxen are yet used in Caithness for husbandry work. The native breed has neither sufficient substance nor spirit; the Galloways are heavier but slow, and do not thrive well in Caithness, and, on the whole, the Highlanders are the best working oxen.’? ‘The native breed of Sutherland is much smaller than that of Caithness.’® In Ross and Cromarty, “The cattle which are kept in the lowlands are principally for the dairy, and they are a mixed breed. There are many pure West Highlanders, but not so small as the common breed of cattle in the counties further north, but there are more of the native cattle, with various degrees of crossing; and others have the Fife and the Moray, and crosses of every kind with them.”* Thus, in the northern counties, just as in Aberdeenshire, the small cattle were being “graded” by cattle from the south; and although we have no information as to their horns, excepting in Sutherlandshire, the descriptions just quoted bear a strong resemblance to those of the cattle on the Aberdeen- shire coast. As to their colour, only one definite statement can be found, viz., the Ross-shire breed “are of all colours, but black and brindle predominate, and are the favourites, as indicating most constitution.”° Skye.—For the information that there were hornless cattle in Skye, we are indebted to no less a person than Dr. Samuel Johnson : “ The cattle of Sky are not so small as is commonly believed. ... Of their black cattle, some are without horns, called by the Scots humble cows, as we call a bee an humble bee that wants a sting. Whether this difference be specifick, or accidental, though we inquired with great diligence, we could not be informed. We are not very sure that the bull is ever without horns; though we have been told that such bulls there are. What is produced by putting a horned and unhorned male and female together, no man has ever tried that thought the result worthy of observation.”® There is no record in which the colours of the older cattle of Skye are specially mentioned ; but they may be inferred from those of the cattle on 1 Youatt, p. 88. 2 [bid., p. 90. 3 Ibid., p. 93. 4 Tbid., p. 95. 5 Ibid., p. 97. This statement was made to Youatt by Mr. MacKenzie of Millbank near Dingwall. 6 <¢'The Works of Samuel Johnson,’’ Dublin, 1798, vol. iv., p. 479. a) c 156 Scientific Proceedings, Royal Dublin Society. the mainland opposite and on the neighbouring islands. Low writes that the colour of the West Highland cattle “is various, often black, sometimes brown, or a mixture of brown or black, and often mouse-dun; and Macgillivray, in his “Report on the Present State of the Outer Hebrides,” published in 1831, writes that “the most common colours are black, red, brown or brandered, that is, a mixture of red and brown in stripes.) A whitish dun colour is also pretty frequently seen.’”? An examination of the “ Highland Herd-Book” shows that there is an unusually high proportion of dun-coloured cattle among the early entries from the smaller islands lying to the south of Skye. Galloway.—Excepting that they have lost more territory to the Shorthorn --to the Ayrshire branch—the cattle of the south-eastern counties of Scotland—Dumfries, Kirkcudbright, Wigtown, and Ayr—have run a course closely parallel to that run by the cattle of the north-east. They have been longer hornless, however, by reason of their feeling and responding sooner to the English demand. They were hornless even in Culley’s time, although he notes the presence of scurs—a phenomenon not yet entirely eliminated from hornless breeds—and erroneously calls them horns: “Their most essential difference from every other breed is in having no horns at all; some few indeed (in every other respect polls) have two little unmeaning horns, from two to four inches long, hanging down loose from the same parts that other cattle’s horns grow, and are joined to the head by a little loose skin and flesh.”* But in the middle of the eighteenth century, according to Youatt, “the greater part of the Galloway cattle were horned—they were middle horns; but some of them were polled.” Half a century ago there were several herds of polled Ayrshires. The Galloways of to-day are nearly all black, but a few are dun. Youatt wrote that “the prevailing and fashionable colour is black—a few are of a dark brindled brown, and still fewer are speckled with white spots; and some of them are of a dun or drab colour, perhaps acquired from a cross with the Suffolk breed of cattle.”® Here again we have proof of a light dun ancestry. Somerset and Devon.—The polled cattle of these counties are long extinct and little is known of them. Low writes of “the Sheeted Breed of Somerset. It has existed in the same parts of England from time immemorial. The red colour of the hair has a slight yellow tinge, and the white colour passes like a sheet over the body. The individuals are sometimes horned, but more 1 « Domesticated Animals,’’ 1845, p. 300. 2 “Prize Essays and Transactions of the Highland Society of Scotland,’’ 1831, p. 263. 5“ Observations on Live Stock,’’ second ed., 1794, p. 60. 4 Youatt’s ‘‘Cattle,”’ p. 154. 5 Ibid., p. 157. \ Witson—Seandinavian Origin of Hornless Cattle of British Isles. 157 frequently they are hornless.”' The Devonshire hornless breed was found about Barnstaple, at one time a Danish settlement, on the northern coast. Youatt writes that “the Devonshire Natts, or polled cattle, now rapidly decreasing in number, possess the general figure and most of the good qualities of the horned beasts of that district,’? which were the North Devon breed. Lawrence also calls them ‘ the Devon natts, or polled cattle on the coast,” and writes, ‘‘ They were described to me as coloured, middle- sized, thick-set, and apt to make fat, but coarser than the true-bred Devons.’® Their colour is not recorded ; but in a letter on the cattle of South Devon in the “Animals of Agriculture,” 1792,‘ Paul Treby Treby mentions both yellow and hornless cattle. At the same time he and other writers mention the importation of cattle from Normandy and the Channel Islands; and it is not altogether impossible that both the yellow colour and the hornlessness may have been introduced by these cattle. A quotation from Treby might be made :—‘“‘ There are also some of a yellow colour ; these are going out fast, being apt to steaé [a provincial word for diarrhoea]: therefore are much less sought after, and sell at a less price . . . The late Lord Boringdon brought into this parish [Plympton St. Mary] a great variety of bulls: some turned out well, others as bad; nevertheless, with Guernsey and Jersey cows, and the ugly breed of blacks that (I believe) were here originally, they have contributed to produce the most motley herd, with and without horns, any country can boast of—I should be ashamed of.” Lreland.—Ireland has also a hornless breed of cattle—the Maoiles, Moyles, Mullines, and so on, which, unless some fortuitous change of taste intervene, seems destined to become extinct Only afew beasts are found here and there in the south-west, the midlands, and in some parts of the north, and there are only one or two small herds of animals picked up by owners who wish to keep the breed alive. They are usually full-sized cattle; and Major Fox, of Harmony Hall, near Athlone, who has a herd of them, writes that “the Mulline cattle are never black, always yellow, or what one should call ight bay; but some of my cows are yellow and white-piebald . . .. Ihave seen brindled Mullines ; and I had a steel-grey Mulline myself; but I regarded her as a hybrid—more of a roan Shorthorn than a Mulline.”’ Others who have known these horn- less cattle mention red, yellow, dun, and brindle as their common colours. They are generally good milkers, and frequently short-legged and sickle- hocked. Low writes that the polled Irish breed “ has existed in Ireland for an unknown period, and appears to have been once widely diffused. It is 1“ Tomesticated Animals,”’ p. 350. 2“ Cattle,” p. 179. 3“ General Treatise on Cattle, &c.,” 1805, p. 45. 4 Vol. xvil., p. 304. > In letters to the writer of this paper. 158 ~ Setentifie Proceedings, Royal Dublin Society. now scattered throughout the country, but is found only in some numbers in the vale of the Shannon. ‘he cattle are of a light brownish colour, and destitute of horns, on which account they have been supposed to resemble the Suffolk Duns. But they are superior in size to the Suffolk duns, equalling, in this respect, the larger class of shorthorns.”! From the description quoted it will be seen that all these cattle, living chiefly in isolated pockets round the coasts of Britain till, at any rate, the eighteenth century, resembled each other in at least three important characters, viz. :— (a) They were horuless. (b) ‘They were either light dun or yellow or dun, which showed that their more remote ancestors had been light dun: yellow having been got by crossing with red cattle, and dun by crossing with black. (c) They were small, puny, short-legged, sickle-hocked, narrow-chined, thin-fleshed, long-headed cattle, which were usually esteemed for the dairy. Those in Yorkshire and Ireland were the only excep- tions in size; but the cattle in both those districts had long been crossed by larger and fleshier breeds of cattle. And from these circumstances, as well as from the fact that they differed entirely in the first two, if not also in the third, from other British cattle, we can scarcely conclude otherwise than that they were all of the one race. The date of the arrival of these hornless cattie in Britain can be fixed with approximate accuracy. The absence of hornless remains of Roman and Saxon age? in the districts inhabited by hornless cattle, and the maritime position of the hornless cattle, show that, if they were brought in by the Anglo-Saxons, it must have been at the close of their invasion ; while the absence of any record or sign of the importation of hornless cattle since the Norman Conquest places their arrival in England before 1066. It is possible they may have been brought to Scotland at a later date. The Burghead carvings suggest for their arrival in Morayshire a date falling within the period of the Danish and Norse invasions; and by the discovery of a number of hurnless skulls in a crannoge near Dunshaughlin, about seventeen miles north-west from Dublin—also a Danish centre—their arrival in Ireland can be shown to have coincided with the Norse invasions. Horuless cattle could not have been imported to Ireland from Hngland till very late in Anglo-Saxon times, for there were none to import; and, since Sir William 1“ Domesticated Animals,’’ p. 327. 2 See McKenny Hughes’s Paper, ‘‘On the more important Breeds of Cattle which have been recognized in the British Isles in successive periods,’’ published in Ar ch@ologia, vol. ly., pp. 125- 158, 1896. Witson—Scandinavian Origin of Hornless Cattle of British Isles. 159 Wilde was able to “ fix the range of date of that [Dunshaughlin] crannoge and its osseous contents, viz., from A.p. 848 to a.D. 933,’ their arrival in Treland may be fixed approximately at the ninth century. The Danes made their first appearance in Ireland at Lambay Island, about fifteen miles north-east from Dublin and twenty miles east from Dunshaughlin, in the year 795.2 ‘The Battle of Clontarf was fought in 1014. We have next to show that cattle like those round the British coasts have been found in other countries where the Norsemen settled. We are handicapped by an imperfect knowledge of the cattle of those other countries; but, nevertheless, the likeness of a few can be shown. Perhaps the most interesting case is the discovery of skulls like those found at Dunshaughlin in some earthen mounds [terpen] in Friesland and Groningen in North Holland. ‘The likeness between the Dutch and Ivish skulls will be seen by a glance at the accompanying illustrations. Trish hornless skull. Dutch hornless skull. In dealing with the Dutch skulls in an article in Cultura for 1908, the magazine of the old students of Wageningen in Holland, Professor Broekema points out that some Scandinavian bracelets and cloak-pins were found in the same mounds. ‘The first of a number of attacks on the Frisian coast was made by the Norsemen in 799.1 Professor Broekema, who also points out that hornless cattle occur very rarely in Holland at the present day, inclines to the view that they belong to the hornless race of “‘ Scandinavia and the adjacent parts of Northern Kurope.” 1 Roy. Ivish Acad. Proc., vol. vii., 1862, p. 68. * Keary’s ‘‘ Vikings in Western Christendom,”’ 1891, p. 486, 160 Scientific Proceedings, Royal Dublin Society. The other places in which Norsemen settled, and in which cattle bearing a resemblance to the hornless cattle of the British coasts live or lived till recently, are Normandy and the Channel Islands, Orkney, Shetland, and Iceland. ‘The Norman and Channel Islands cattle are identified by their shape and by the presence of the two colours silver grey and yellow; those ‘in Orkney and Shetland by shape and the dun colour; while those in Iceland were identified by shape and the absence of horns. A former student! under the writer of this paper connected with the Orkneys has acquired the information that in the middle of last century there were many dun cattle in Orkney and Shetland, but especially in Shetland; and the writer saw dun Orkney or Shetland cattle in Aberdeen not many years ago. Low is quite clear that the cattle of Shetland are of Norwegian origin; but unfortunately he gives them horns. They “are distinctly Norwegian in their characters, and a similar race extends to Iceland. ‘They are small, but of very good form when pure, and fatten with great quickness, when carried to superior pastures. Their horns are short, their skin is soft, and their flesh is equal to that of any cattle produced in the British Islands. They are of various colours, generally parti-coloured, and tending more to the lighter shades than the cattle of the Highlands. . . . The cows are tolerably good milkers, in which respect they agree with the cattle of Norway, and differ from those of the Highlands; and in this respect, too, they agree with the cattle of Jersey and the islands of the Channel, which are likewise believed to be of Norwegian origin. ‘These cattle are smaller than those of Norway, which is to be ascribed partly to the absence of shelter, and partly to the want of artificial food.’” With regard to these old Orkney and Shetland cattle, it is safe to say, although there is no other evidence, that their shape and size, together with the presence of dun, show them to have been of the same race as the cattle of the British coasts, and therefore at one time hornless. With regard to Iceland, Uno von Troil writes in 1780: “ Their beeves are not large, but very fat and good. It has been reported by some, though without foundation, that there are none without horns: it is true, however, that they seldom have any.’ ‘The Norsemen settled in Iceland about the time they settled in Britain. “Are Frode, born 1068, ... expressly says, in the first chapter of the book (Landvama Bok) that Iceland was settled by the Norwegians in the time of Alfred, King of England, and of Edward, his son.’ 1 Mr. R. J. Anderson, of Messrs. Reith and Anderson, Aberdeen. 2 «© Tomesticated Animals,” p. 297. % “ Letters on Iceland,” 1780, p. 132. 4 Thid., p. 60. Witson—Scandinavian Origin of Hornless Cattle of British Isles. 161 There are small bunches of hornless cattle here and there in other parts of Europe, but, according to Wilckens, ‘‘ they are found chiefly in Northern Hurope, in North Russia, Finnland, Lappland, Sweden, Jemtland, Norway, and Iceland, settled by the Norsemen in 874 a.p.”! Wilckens adds, erroneously, that ‘‘in these places they are invariably white.” Some of them may be so: the Fyjall race of Sweden, for instance, seems now to be white; but the quotations which follow show that others are not white. It is impossible to give a general description of these hornless cattle of Northern Europe; but afew quotations from letters kindly sent by correspondents in Scandinavia will show that many of the cattle in that part of the world are extraordinarily like those that have already been described as existing now or formerly in the British Islands. Professor Isaachsen, of Aas, Norway :—“ As to our cattle up to the year 1600, we know very little. But in those days, like in ours, there were several distinct breeds in our country, and probably they have not changed their characteristics very much. Especially in the western and south-western parts of Norway, the so-called “ Vestland,” from which part of the country the first settlers are supposed to have come to your country, the breed is partly horned, partly polled, about half of the animals being polled, I think. ‘The colour of the breed is either black, dun, red, or grey, whole- coloured, or with small or large white marks and spots. “Tn the south-eastern parts of Norway, especially in Akershus and Smaalenene (two “amter’’-shires), the indigenous breed is constantly ved and polled (Det réde pollede Ostlandstae—-the red polled breed). “Tn Hsterdalen and Gudbrandsdalen, the two large eastern vaileys of the country, the native breed is black or dun, in some cases red, most of the individuals being horned: only a few are polled. “‘The breed in the western parts of Norway we suppose to be the most ancient or one of the most ancient in our country.” Professor Maar, of Copenhagen: “ Mr. Mérkeberg? thinks that most of the Danish cattle in early times were horned, but that hornless cattle may have been found, not in Jutland, however. I shall add the remark that small dun-coloured cattle are still—though seldom—found in the Danish isles, and sometimes are supposed to be the original Danish cattle. I suppose that by dun-coloured you mean cattle with black or grey hairs or black and grey hairs mixed thoroughly up with white hairs.” Professor Redlund, of Stockholm, says that the Swedish hornless cattle— the mountain or fell race [ Fjiill rasen ]—are undoubtedly the original Swedish Grundztige der Naturgeschichte der Haustiere, 1905, p. 308. ?j.e. Staats Consulent Morkeberg. SCIENT. PROC. R.D.S., VOL. XII., NO. XV. 2D 162 Scientific Proceedings, Royal Dublin Society. cattle, and that this is proved by the finding of hornless skulls in Scania, and by the occurrence of a hornless, whity-grey “‘ Risinge-race”’ in Ostergotland [Hast Gothland] in the eighteenth century. This breed has a lengthened hornless head and short crook-hocked legs. It is now white, with small black specks. The following description and illustration of a cow of the Fjall race are taken from Sundbiirg’s ‘Sweden: its Population and its Industries,” published in 1904:—“ The history of cattle in our country presents a good many Swedish Fjall Cow. vicissitudes. The Law of Uppland, a.p. 1296, describes Swedish cattle as being small, hornless, white or whitish-grey, often with dark spots. The Alpine breed in Northern Sweden is so still—a race we have every reason to consider as being the oldest in the country. But at an early day there came into the country—probably from the east—a larger horned race of cattle, reddish-yellow in colour, which towards the north more and more invaded the districts of the older race. This race has by degrees been crossed with and in many places replaced by purely foreign breeds; but it long survived typically in the forest districts of Smaland, and it is still found in the Island of Gothland.” In collecting data for this paper, several authors were found who, more or less tentatively, suggested a racial connexion between the hornless cattle of Britain and those of the Continent. Middendorff maintains that the hornless eattle of the ancient Scythians, referred to by Herodotus, migrated north- wards with their owners through Russia to Finland, from thence to Witson—Scandinavian Origin of Hornless Cattle of British Isles. 163 Scandinavia, and then to Britain. He presumes their arrival in Britain at a date, however, which is far too early: “In Great Britain, the hornless cattle were driven to the extreme north among the Scottish mountains in company with the Kelts fleeing before the Romans and the Anglo-Saxons.”" Mr. Henry F. Huren, who has written a history of the English red polled breed, suggests that the old Suffolk cattle were descended from Scythian stock. It has not been possible to see a copy of Arenander’s paper on the hornless cattle of Northern Kurope. There are several other questions connected with the present subject that might be referred to shortly. It was pointed out early in this Paper that several of the “ wild’ white herds in Britain are or were hornless. Are these of the same race as the hornless cattle of Scandinavian origin? For the present that question will have toremain unanswered. If they are of thesame race, then their colour has been changed. ‘This could have happened in only a few generations by crossing with white horned cattle, while, at the same time, the hornless character could have been retained. But it would be somewhat extraordinary if all the hornless herds had changed their colour. It is also possible that the “wild” white hornless herds were once horned and lost their horns by crossing with hornless cattle. A statement by Professor Cossar Ewart, in the article “Cattle” in the “Standard Cyclopeedia of Agriculture,” to the effect that he has found hornless cattle-skulls in the Roman fort at Newstead in Berwickshire, suggests that the Romans may have brought hornless as well as horned cattle to Britain: both being of the South Huropean white race. This would mean that two sets of hornless cattle came to Britain: one, the light-dun set by the northern, and the other, the white set, by the southern route. That suggests the further question: Where did these two sets of hornless cattle come from originally, and where did they split partnership? Did they split somewhere in the south-east of Hurope? and, tracing it backwards, does their track lead through Asia Minor, down through Syria, and across the Isthmus of Suez into Egypt? Or were the Egyptian hornless cattle merely another branch of a race having its origin in Asia? and should the track strike eastwards from Asia Minor? According to Keller,’ from whom the follow- ing illustration is copied, there were hornless cattle in Egypt as early as the fourth and fifth dynasties. 1 « Landwirtschaftsliche Jahrbiicher,’’ vol. xyii., 1888, pp. 299 and 300. * Wallace’s ‘‘ Live Stock.”’ 3“ Naturgeschichte der Haustiere,’’ 1905, p. 115. 164 Scientifie Proceedings, Royal Dublin Society. The presence of scurs among hornless cattle has already been referred to. ‘These are mere loose. epidermal growths, sometimes an inch or two Old Egyptian hornless Cattle. long attached to the skin where the horns ought to be. But, though the skull is often full and square in the animal with scurs, there is no developed core or bony outgrowth beneath the scurs. Scurs are fairly common among the Irish Maoiles, but they have been almost entirely bred out among the other hornless breeds. In a mixed herd of horned and hornless cattle where scurs are common this bony outgrowth is used to tell whether the calves are going to have horns or not. ‘Those with fixed immovable outgrowths will have horns; those with slack movable scurs will have none. ‘hese phenomena raise a series of questions which of course we cannot decide: What are scurs? What are their functions? ‘They occur independently of horns. Have they any bearing upon the origin of horns or upon the origin of hornlessness? Did the growth of the bone of the horn at one time cause a sympathetic, perhaps concomitant, growth of epidermal horn? or are the scurs merely a legacy left by the horned upon the hornless cattle they crossed, and do they appear because the epidermis is willing to perform its part while the bone below refuses ? THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 16. JULY, 1909. FURTHER OBSERVATIONS ON POWDERY POTATO-SCAB, SPONGOSPORA SUBTERRANEA (Wallv.). BY T. JOHNSON, D.Sc., F.LS., PROFESSOR OF BOTANY IN THE ROYAL COLLEGE OF SCIENCE FOR IRELAND. (PLATES XII.—XIV.) [Authors alone are responsib/e for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATEH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price One Shilling. DB iets oe . ne el UO +. XVI. FURTHER OBSERVATIONS ON POWDERY! POTATO-SCAB, SPONGOSPORA SUBTERRANEA (Wallr.). By T. JOHNSON, D.Sc., F.LS., Professor of Botany in the Royal College of Science, Dublin. (Pirates XII.-XIV.) [Read Arrit 20. Ordered for Publication May 11. Published Jury 24, 1909.] In February, 1908, I gave an account of a scab (1) of the potato-tuber which I had found to be very prevalent in the west of Ireland. I have had tubers similarly diseased from other parts of Ireland, as well as from Hngland and Scotland. ‘The scab is due toa slime-fungus, described in part and named by Brunchorst (2) in 1886 as Spongospora Solani. In the paper mentioned, and an earlier one, (3) I sought to fill in certain gaps in the life-history of the fungus, more especially in reference to its relation to the host, its plasmodium, the structure and germination of the spores. I sent samples of scabby tubers to Kew for the Museum, and, in acknowledging them, Colonel Prain, the Director of the Gardens, replied :— “ Kew, Pebruary 17th, 1908. “Very many thanks for the specimen of Spongospora Solani of Brunchorst. We are inclined to think that it is the same fungus as was described in the first instance by Berkeley as Zuburcinia Scabies, and afterwards transferred to Sorosporiwm as S. Scabies by Fischer de Waldheim ; Fischer’s name has been taken up by Saccardo and by Massee. I am sending you a fragment of the type which is in this Herbarium. As you have been at work on the Spongospora, you will be able to say straight away whether the two are the same or are different.” Massee had already (Gardeners’ Chronicle) stated that the new scab was nothing but the scab described by Berkeley fifty years ago, and in 1904, the year in which I first saw and described the scab, he gave in the Journal of the Royal Horticultural Soctety (4) a short account of the scab as the smut i Wound-cork is such a frequent and marked feature of ordinary scab that I prefer the term “powdery ”’ to ‘‘ corky”’ scab, the name I suggested in my earlier paper. SCIENT. PROC, R.D.S,, VOL. XII., NO. XVI. 25 166 Scientific Proceedings, Royal Dublin Society. Sorosporium Scabies. Fearing that I had made a serious mistake, and identified as a slime-fungus an organism which was regarded by Massee and earlier observers as a smut, I restored the scrap of type-material and examined it, to find that it agreed in all respects with my material, that it showed no trace of any Ustilagineous characters, and that it was identical with the slime-fungus Spongospora Solant. I wrote to Kew to this effect, and inserted the following paragraph at the end of my paper :— “Through the kindness of Colonel Prain, F.r.s., the Director of the Royal Botanic Gardens, Kew, I have been able to compare type-herbarium material of Sorosporium Scabies (Berk.), Fischer de Wald., with Spongospora Solani, Brunch. I can see no difference in size or structure of the spore- balls of the two, and believe microscopic examination of restored material will show that Sorosporium Scabies, with its ‘glomerulis 1-2 lacunosis,’ should be removed from the Ustilaginese— that it is really Spongospora Solani.”’ Massee evidently accepted my view, since a few months later, under the heading “English Potato Scab,” it is stated in the Jowrnal of the Board of Agriculture (England), vol. xv., p. 509, that “in 1886 Brunchorst described a destructive organism on potatoes as Spongospora Solani, Brunch., and this species was in reality the pest previously named by Berkeley, although not recognized as such by Brunchorst.” My recognition of the identity is ignored, and the name proposed is Spongospora Scabies, Massee. Under this name the ‘“‘Corky-scab,” as I had named it, is described in detail in the November number (5) of the Journal (vol. xv., p. 592) with a plate of illustrations. To this paper I shall return. Berkeley’s (6) first account of the scab is illustrated by two small figures of the spore-ball of the fungus, each with a short stalk of attachment. Was there anything to account for these stalks? The spore-balls appear to me to lie free in the cavities of the cells, and to show no trace of a stalk. The shrivelled remains of the host-cell cling to the ball sometimes, and might be mistaken for a stalk. It is clear that Berkeley gave little attention to this scab, as he was fully occupied with the ‘‘murrain.” Thus, in the article on the murrain, he writes (p. 38): ‘“ Amongst the diseases noticed by Martius is one which he considers as depending on a species of Protomyces. As I have seen this in various stages of growth, and attached to its flocci, I have thought it worth figuring. It appears to me to belong to the genus Tuburcinia, Fr. The spores have usually one or more cavities in the surface communicating with the interior cavity. They may, perhaps, therefore be considered rather compound bodies, consisting of a quantity of cells arranged in the form of a hollow ball. This view of their structure requires more attention than I am able to give to it at present ....” (p. 10). ‘Another Jounson—Surther Observations on Powdery Potato-Scab. 167 disease, arising from a very different fungus, is frequent, especially in calcareous districts, and is known commonly by the name of the scab, the surface of the potato being covered with pustules, which at length become cup- shaped, and are powdered within with an olive-yellow meal, consisting of the spores of afungus. This also has been partially investigated by Martius, who has illustrated his observations with some characteristic figures.” In another article, entitled “On a form of Scab in Potatoes” (King’s Cliff, Nov. 1847), Berkeley writes, “There are two very different diseases known commonly under the name of scab.” (His article deals mainly with what we still call, without understanding its cause, “common scab.”) “The first, of which it is not now my intention to treat, was described and figured by Martius (Die Kartoffel-Epidemie, p. 23, tab. 2, figs. 9-13; tab. 3, figs. 36-38), and is characterized by the presence of an olive-green or brownish pulverulent Hyphomycete (Zuburcinia Scabies, Berk., Journal Royal Horticultwral Society, London, 1846, vol. i., p. 33, tab. 4, figs. 30, 31), which gives a very peculiar appearance to the pustules, and to which indeed it is not confined, but occasionally forms a stratum a line or more in thickness beneath the greater portion of the cuticle. A few scattered tubers occur now and then affected by this disease, but it is very rarely so prevalent as to draw much attention. The potato crops, however, suffered greatly from its ravages in the Scilly Islands and in Cornwall during the present summer, where it appeared under a very destructive form. Mature specimens were forwarded to me, with the promise at some future period of a supply of tubers in every stage of the disease. I was, however, disappointed in my hope of being enabled to investigate its nature more closely, possibly because the malady, as Martius reports, is several weeks in going through its phases.” Again, in the Annals and Magazine of Natural History (7), under “ Notices of British Fungi” by Berkeley and Broome, our scab appears as Tubercinia Scabies, Berk.; “Sporis globosis cavis hic illic lacunosis olivaceis” (Berk., Journ. Hort. Soc., vol.i., t. 4, figs. 30,31). Rhizosporium Solani, Rab., No. 900. On potatoes: very common: often confounded with the true potato scab. “The spores of this species are very curious; they are composed of minute cells, forming together a hollow globe with one or more lacune communicating with the external air. A hollow shell with one or two apertures will give a notion of their form. ‘hey are generally attached laterally by a delicate thread.” This interpretation of the scab is repeated by M. C. Cooke in “Micro- scopic Fungi” (p. 231), where the scab is called “ Potato Smut.”” Plowright includes the scab in the Ustilaginess in his work on “Uredinez and Ustila- gine,” but states that he has never seen the organism, and can find no trace of it in his specimens from Cooke’s Evsiccati. 2u2 168 Scientific Proceedings, Royal Dublin Society. Writing in 1877, A. Fischer de Waldheim (8) had, without assigning reasons in detail, placed 7’. Scabies in the genus Sorosporium as 8. Scabies, retaining it amongst the Ustilaginee. Saccardo adopts this view, as did Massee. It is, I think, evident that mycologists generally accepted the views expressed by Berkeley, who had, as the quotations I have made show, not found time to give much attention to the disease, and relied mainly on Martius’s work, though he clearly regarded the fungus as a Hyphomycete. Martius (9) wrote in 1842 his article, Die Kartoffel-Epidemie. The article deals in detail with two potato troubles—“ Stockfaule,’ due to Fusisporium Solani, and suggestive of Fusarium dry-rot of to-day, and “ Kartoffel-riude’’ or scab, the subject of this paper. Martius gives an excellent account of the early stages and external appearance of the scab, and of the injurious effect of the disease on the tubers. The tubers, he states, become unsightly, in pronounced cases have a disagreeable taste, often keep badly in store, and as seed-tubers may fail to sprout, or give shoots which are weak and in the end fall off. Had his valuable article been translated, digested, and more generally acted on, the benefit to the potato-crop of Britain would have been great. His explanation of the origin of the spore- balls is vitiated by his belief in spontaneous generation. He fully appreciates the important influence of predisposing causes, but states that, owing to unfavourable conditions, the organic juice or sap of the tuber undergoes degeneration, and this so altered juice becomes converted into an “Urpilz”— his Protomyces—from which the “ grains” (spore-balls) arise. He is, he says, further gradually coming to the conclusion that the view that disease is transmitted by means of a myasmic effluvium from a diseased to a healthy organism, is untenable, that a solid body is necessary for infecting purposes, though in such a case as scab the low organism responsible can, he thinks, arise de novo. ‘The first account of the scab fungus is, however, given by Wallroth (10) in Linnea (vol. xvi., 1842). Wallroth states he has long known the disease which was then so frequently mentioned in agricultural papers, and he gives the following diagnosis :— “ Hrysibe subterranea, a. Tuberum Solani tuberosi. Sporis subrotundis maximis obscure cellulosis tenuissimis, primum flavicantibus dein fusco- virescentibus sub summa tuberum subterranearum vegetorum epidermide livescente maculari dein colliculosa lacero-fissa grumulos ovato-subrotundos hemispheericos immersos polysporos iisque effoetis scrobiculos superficiales nudos preestantibus.”’ Martius saw Wallroth’s preparations and description, and agreed with him as to the nature of the organism. Martius transfers the fungus, however, to Protomyces Luberum Solant Tuberosi, and gives an amended diagnosis of it in Jounson—Further Observations on Powdery Potato-Scab. 169 which the spore-balls are described as ‘‘globulis,” the equivalent of Wallroth’s “polysporis,” not “pseudosporis,” as quoted by him. In both the diseases described by Martius he sees a similar matrix or degenerated cell-sap. In the scab this becomes transformed into a powder, but in the other disease it gives rise to an “after organization,” the mould or Hyphomycete Fusisporium. It is not surprising that at this date, four years before Berkeley’s account of Phytophthora infestans appeared, disease organisms should have been confounded together. In the present case it is, I think, evident that Pl. IIL, fig. 20, of Martius (Pl. XII., fig. 2, of this paper) represents the same organism as Brunchorst’s figure, here reproduced as Pl. XII., fig. 1. Beyond the one reference to the insignificant stalk occasionally seen attaching the “grains” or spore-balls to the cell-wall, there is no suggestion in Martius’s account of the presence of hyphe in the scab organism. Berkeley, however, without comment places it in the Ustilaginee. Wallroth in a few lines accompanying his diagnosis calls the fungus a smut. Was this Berkeley’s reason for placing it in the smuts? Were hyphe observed ? Frank (11) examined the organism and accepted it as a slime-fungus, but considered it to be a saprophyte, not the injurious parasite it is. The general character of the Spongospora scab is evident from Pl. XIL., fig. 3, showing a tuber with several scabby or wart-like patches. ‘The wart is a proliferation of the tissue of the tuber, and its disintegrating cells are full of spore-balls. The plasmodial stage of the fungus may be seen in the deeper-seated cells passing on to the ordinary starch-cells. In Pl. XII., fig. 4, the plasmodium-occupied cells appear dark-coloured, and lie between the cells with spore-balls and those with starch in them. Pl. XII, fig. 5, taken from a scabby tuber, shows the plasmodial stage only, beneath the cork which is recognizable to the left of the figure. Pl. XIL., fig. 6, shows a little of this attacked tissue more highly magnified. In some of the host-cells the starch and other cell-contents have been entirely absorbed by the parasitic plasmodium. In others it is in course of absorp- tion, The wall of the host-cell is mostly intact and unattacked. There are no signs of fungal hyphe. In Pl. XIII., fig. 6, the vacuolated plasmodium is shown, just before spore-formation begins. There is abundant evidence that the Spongospora parasite is an intra- cellular organism, and that it occurs both in its plasmodial and fruiting stages within the host-cells. Plate XIII., fig. 5, shows a spore-ball lying in the host-cell. At x the nucleus of this cell is still observable. The light spots at s are unabsorbed starch-grains. The walls of the host-cells are clearly present, and there are no fungal hyphe. In Pl. XIII, fig. 8, a host-cell is seen half oceupied by a spore-ball. In the other half are 170 Scientific Proceedings, Royal Dublin Society. protoplasmic remains and two impoverished nuclei suggestive of parasitic stimulation. PJ]. XIII., fig. 7, is taken from a preparation of restored type material of Z. Scabies, and shows a cell occupied apparently by the plasmodium, judging from comparison with the appearance of the cells in Pl. XIIL, fig. 6. Tue Spore-Batts. Martius describes the spore-balls as opaque grains with a shagreen-like surface, and formed of several larger and smaller globules. Together they form the black or dark brown powder, each grain being about twice the size of a starch-grain and “8, — =(%, mm.” in diameter. Berkeley describes the spore-balls as hollow spheres with one or two apertures communicating with the interior—an error in description repeated by several subsequent writers. Brunchorst, who made microtome preparations of the balls, saw and described their true structure. Hach spore-ball, ovate-oblong to more or less spherical in shape, consists of hundreds of angular spores 3°5 wu in diameter, firmly bound together to form a sponge-like body. ‘The spore- ball is not hollow but honeycombed by cavities or passages in communication with one another and with the exterior. ‘The spores are thus arranged like trabecule or strands traversing and enclosing the cavities. Untortunately Massee repeats the earlier erroneous description, even in the description of the origin of the spore-ball from the plasmodium. ‘The figures here given, Pl. XIII., fig. 2 (from Berkeley’s own material), and Pl. XIII., fig. 1, show that the spore-ball is not a hollow sphere, but a sponge-like body, the cavities or plasmogenetic intercellular spaces, as I have called them, being the vacuoles of the sporogenous plasmodium. Massee speaks of the appearance of the ripening spore-ball, as seen “‘in optical section,” and his mistake in description may be due to want of use of microtome preparations. The inaccurate description of the spore-ball’s structure made me doubt the reliability of the observation by Massee of the earliest stage of the invasion of the host-cells by the parasite. Massee says: ‘‘The earliest condition observed in a cell of the host consists of a very few irregularly _ globose protoplasmic bodies aggregated round the nucleus of the cell. [ His fig. 4, here reproduced in Pl. XIIL,, fig. 4]... . When fixed and stained these amoeboid bodies are seen to possess a single nucleus. . . . How these amesboid bodies gain an entrance into the cell of the host has not yet been observed; but such invaded cells are always the most internal of the cells affected by the parasite, and are always immediately adjoining other cells Jounson—Further Observations on Powdery Potato-Scab. 171 containing the organism in a more advanced stage of development... . It appears highly probable that the cells are invaded by the parasite for some time before the amceboid bodies surrounding the nucleus can be seen ; for when they are present, the starch has, as a rule, disappeared from the cells.” Sections stained with congo red are described in support of the view that these bodies are the myxamoebe of other slime-fungi. I have spent what has sometimes seemed to me a disproportionate amount of time in the microscopic examination of Spongospora without observing any signs of such amoeboid bodies as those just described. I suspected them to be, in fact, an accumulation of young starch-grains around the host nucleus. Such grains are spherical, with a centric hilum, which takes up a stain readily and might pass for a nucleus. I have placed side by side with Massee’s figure of the myxamoebe a micro-photograph of a Flemming-fixed and stained healthy potato-cell (Pl. XIII, fig. 3). The similarity is, I think, striking. Massee himself states that the starch-grains have disappeared when the amoeboid bodies appear. Replace “amoeboid bodies” by young “starch-grains” in the above-quoted statement, and everything becomes clear. JI am in accord with Massee in the view that the parasite gnaws into the flesh of the tuber by its penetrating plasmodium—the more easily the moister the soil. The drier the soil the more readily is the parasite kept in check by the formation of protecting wound cork. The plasmodium, too, carries the disease from the seed-tuber through the stoloniferous branches over into the new tubers, making them scabby, as I have already shown. I must refer readers to my earlier paper for a general account of the changes which I consider lead to the maturity and germination of the spores. Massee sees in each ripe spore a single nucleus. At germination the uninucleate spore-contents escape, as in slime-fungi generally, as a single body, which shows for some time sluggish ameeboid movements, and then becomes stationary. Its nucleus next divides into two, followed by complete fission of the body. By repeated fission numerous 1-nucleated bodies are formed, which ultimately coalesce to form a plasmodium. I regard the spore as much more complicated, and comparable to the spore of Ceratiomyxa, as recently described by Jali (12), and Olive (13). Pl. XIV., fig. 1, is a microphotograph of a spore-ball in culture, in section. The details of the contents of the different spores in the ball deserve examination by pocket- lens. In the spore marked ¢ there are apparently four nuclei present, arranged tetrahedrally, highly suggestive of a stage observed in Ceratiomyza. The spore is so small that I cannot yet give a satisfactory connected comparative account of the nuclear changes. In my earlier paper I have 172 Scientific Proceedings, Royal Dublin Society. suggested what appears to be the line of development followed. In the interpretation of the phenomena in the larger spores of Ceratiomyxa, Jahn and Olive differ fundamentally; and I may well hesitate in the much more difficult case of Spongospora to interpret the appearances presented. Pl. XII., fig. 7, shows some of the stages drawn from preparations under the highest magnifications. The possibility that there was some justification for the statement made by Berkeley and others, that the spore-balls were attached by a side-stalk, and that the fungus was a smut, early disturbed me, and I was quite prepared to find in the type material of Sorosporium Scabies hyphe in the scab spots. In one section I found cells filled with fungal hyphe as figured (Pl. XIV., fig. 4), suggestive of the sclerotium of Rhizoctonia (or Corticium). In some Scotch material I examined, hyphee were clearly present in the scab areas outside the cellular tissue of the tuber; and though some could be accounted for as the mycelial hyphe of Rhizoctonia scab, there were others not so explicable (Pl. XIV., fig. 5). The hyphz are swollen, septate, and branching ; their contents abundant and granular. In some cases chains or masses of spores may be seen arising from the protoplasmic contents of the hyphe (Pl. XIV., fig. 2). Spongy spore-balls, very like those of Spongo- spora, arise; and ultimately the enclosing walls of the hyphe disappear and leave the balls lying free. The more external balls lose their compactness and break up into single spores or small groups of spores (Pl. XIV., fig. 5), so that they form a finer powder than Spongospora, whose spore-balls remain intact to the end. ‘The mode of origin of the spores just described is not unlike that met with in the Usti/agince, so far as they have been studied; and it thus appears as if (as I wrote in my working notes as long ago as May, 1908) there are two kinds of potato scab characterized by powdery spore- balls, the one with a plasmodium—Spongospora subterranea Wallr.—the other a Hyphomycete, possibly one of the Ustilaginacee. In 1892, Lagerheim (14) published a short note in the Journal of Mycology, in. which he stated that the scab observed by Brunchorst was well known to the natives in 8. America, though due, they thought, to the gnawing of worms. Lagerheim states that Brunchorst wholly misunderstood and misinterpreted the organism, and that the wart-like tissue is not potato- flesh, but a pseud-parenchyma of fungal hyphe, in which the characteristic spore-balls arise. The hyphze are filled with a colourless protoplasmic substance, which is very often full of vacuoles. This is perhaps, says Lagerheim, what Brunchorst mistook for plasmodia. In the warts, containing mature spore-balls, the hyphz are usually empty, and the spore-balls are not free, but fastened to the surrounding hyphe of the pseud-parenchyma. JoHnson—Further Observations on Powdery Potato-Seab. 173 Lagerheim did not see any early stages of spore-formation, but presumes that, according to their mature structure, “neighbouring hyphe develop upon the pseud-parenchyma, and divide up into small cells, which cling firmly to and partly surround the sporogenous hyphe.’’ The spores could not be made to germinate, and Lagerheim’s statement is not illustrated, but the reconciliation of his observations with those of Brunchorst is to be found possibly in the account I have given of the smut-like organism. It is possible, judging from some preparations I have, that a third organism may produce yet another form of powdery scab. Lagerheim concludes that his fungus is the same as that called Spongospora Solani by Brunchorst, and that it has been long known and described by various authors as Erystbe sub- terranea Wallroth. “Unfortunately,” he says, “the publications of Wallroth, Martius, and Berkeley are not accessible to me, which renders it impossible for me to decide this question. If my supposition be correct, the fungus should be called Spongospora subterranea (Wallr.)”’—a name I had decided to adopt long before my attention was called to Lagerheim’s note. (A somewhat parallel case is mentioned by Sorauer in his Handbuch :— The coral-like swellings on the roots of the alder, bog myrtle, &., were at one time described as due to a filamentous fungus Schinaia Ani. Moller then attributed them to a plasmodium to which he referred the Schinzia. _Woronin saw the two forms existing side by side. Brunchorst, it is interesting to note, next saw no signs of a plasmodium, but a filamentous fungus, which he distinguished from Schinzia as Frankia subtilis. Finally, Bjorkenheim concludes the fungus is a filamentous one, with swollen hyphz, which form bladder-like swellings, mistaken for spores by earlier observers.) PREVENTION OF THE DISEASE. In 1909 I made a number of experiments by pot cultures to test the possibility of preventing Spongospora scab. In all cases the soil was good garden loam, and healthy seed in it gave a healthy crop. 1. Seed-tubers of the varieties Red Skerries and Champions suffering from scab gave a scabby crop. 2. Similar seed first treated for eighteen or twenty-four hours, with a 2 per cent. solution of Bordeaux mixture, gave tubers free from scab. 3. Such seed treated with sulphur gave a scabby crop. 4, | planted healthy tubers of Sutton’s Superlative, into which I had previously grafted a wedge of a scabby tuber, from which all the spore-balls, as far as 1 could make sure, had been scraped off. ‘lhe tubers formed were scabby. Thus it appears that the resting plasmodium can communicate infection to healthy tubers. SCIENT. PROC. R.D.S., VOL. XII., NO. XVI. 2F 174 Scientific Proceedings, Royal Dublin Society. 5. I soaked uncut tubers of Sutton’s Superlative in a decoction of spore- ball material. The resulting tubers were healthy. ‘Thus, as far as it goes, this experiment indicates that whole seed, even in scab- contaminated soil, may resist attack and give a healthy crop. The potato crop in Ireland would be much more satisfactory if whole tubers, 34 oz. in weight, and previously soaked in one or other fungicide, were planted. Poor seed tubers, often badly diseased, cut up into three, four, or five sets, placed with exposed surface in direct contact with the manure, frequently of inferior quality, cannot but give a diseased crop. All the predisposing causes of the production and propagation of disease are allowed to act, and a poor crop must result. Sponcospora ScaB LrreratTure. 1. Jounson, T.: Spongospora Solani, Brunch. (Corky Scab). Econ. Proce. Roy. Dublin Soe., 1, 1908, Part xii. 2. Bruncuorst, J.: Ueber eine verbreitete Krankheit der Kartoffel- knollen. (Bergens Museums Aarsberetning, 1886). 3. JoHnson, T.: Der Kartoffelschorf: Spongospora Solani, Brunch. Jahresb. d. Vereinigung d. Vertreter d. angew. Bot., iv., 1906. 4, Masszxz, G.: Some Potato Diseases. Journ. Roy. Hortic. Soe., xxix., 1904. 5. Masszx, G.: English Potato Scab. Journ. Board of Agriculture, xv., No. 7, 1908. 6. Berxeiey, Rev. M. J.: “ Observations, Botanical and Physiological, on the Potato Murrain.” Journ. Hortic. Soc., i., 1846. 7. BerkeLey AND Broome: Notices of British Fungi. Anu. Mag. Nat. Hist., v. (2nd Series), 1850, p. 464. 8. Fiscour pe WaLpueim, A.: Apercu Syst. sur les Ustilaginées, 1877. 9, Marrius, C. F. P. von: Die Kartoffel-Epidemie, 1842. 10. Watirora: Lrysibe subterranea. Jinnea, 1842. 11. Franx, H. B.: Kampfbuch g.d. Schadlinge uns. Feldfriichte. 1897, p- 177. 12. Jaun, E.: Myxomycetenstudien. Ber. d. Deutsch. Botan. Ges., xxvi., 1908. 13. Oxtve, E. W.: Cytological studies on Ceratiomyxa. Trans. Wisconsin Acad., xv., 1907. 14, Laceruem, G. pe: Remarks on the Fungus of a Potato Scab (Spongospora Solani, Brunch.). Journ. of Mycology, vii., 1892. 15. SorsueR: Handb. d. Pflanzenkrankh., 1908, p. 15. HXPLANATION OF PLATE XII. Fie. Fig. Fig. PLATE XII. . Spongospora-infected tuber sprouting feebly. (May, 1909.) . Reproduction of a figure of Protomyces from Martius’s ‘‘ Die Kartoffel- EKpidemie.” . Reproduction of a figure from Brunchorst’s article on Spongospora Solant, showing the plasmodia in potato-cells. . Microphotograph of plasmodia p, and spore-balls. . Microphotograph of plasmodial stage of scab. At ¢ cork cells are visible. . A little of fig. 5 more highly magnified. Host-cells are seen in various stages of invasion by the plasmodium, e.g., in a, host-nucleus, and starch-grains are still present. . Sketches of spore in spore-balls in different stages of development or of germination. (Mostly x 2000.) SCIENT. PROC. R. DUBL. SOC., N.S., Vou. XII. PLATE XII. EXPLANATION OF PLATE XIII. Fig. Fig. Fig. Fig. Fig. 8 PLATE XIII. . Section through a group of free spore-balls of Spongospora. . Three spore-balls in section, from type herbarium material. . A healthy potato-cell, showing nucleus surrounded by small starch-erains. . Copy of Massee’s figure of potato-cell to show Spongospora myxamebe, surrounding host-nucleus. . Several intra-cellular spore-balls : n host-nucleus in impoverished condition ; S, S, S, Starch grains. . A single host-cell occupied by vacuolated plasmodium (p) (x 2000). . A potato-cell of the type herbarium material, occupied by Spongospora plasmodium. . A host-cell, one half occupied by a spore-ball, the other half by the protoplasmic and starch contents. Two feebly-staining host-nuclei are present in this cell (7). SCIENT. PROC. R. DUBL. SOC., N.S., Vou. XII. PLATE XII. PLATE XIV. Fig. 1. A spore-ball in culture, showing the spore in various stages, e.g., at ¢, the tetrahedral arrangement of nuclei. Figs. | Section through Hyphomycete scab, showing chains and balls of spores, 2, 3, 5.) both arising and formed. Fig. 4. Host-cells filled with fungal hyphe, from type herbarium material of Tuburcuma scabies. Fig. 5. The spore-groups of the Hyphomycete in a much more powdery condition than in Spongospora. ‘The sporogenous lyphez are clearly seen in the figures 2, 3, 5. SCIENT. PROC. R. DUBL. SOC., N.S., Vou. XII. PLATE XIV. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 17. JUNE, 1909. MECHANICAL STRESS AND MAGNETISATION OF IRON. Animale BY WILLIAM BROWN, B.Sc. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price One Shilling. fe 7 XVII. MECHANICAL STRESS AND MAGNETISATION OF IRON, Parr II. By WILLIAM BROWN, B.8c. [Read Aprin 20. Ordered for Publication May 11, Published June 21, 1909.] A PREVIOUS communication brought by me before this Society! contained some results obtained in investigating the relations between mechanical stress and magnetisation with iron wires, which wires in most of the experiments were not in a uniform longitudinal magnetic field throughout their entire lengths. The results which I now submit were all obtained with the iron wires in a perfectly uniform magnetic field. The arrangements for the first part of the experiments were the same as those described in the former paper, where the wire under test was suspended in a vertical direction, and joined in series with a moving-coil ballistic galvanometer, having a resistance of 325 ohms. The galvanometer-scale was placed at a distance of 101°3 cms. from the galvanometer-mirror, and was divided into 700 divisions in a length of 44 ems. An earth-inductor was so arranged that it could be inserted in the circuit, and the constant of the galvanometer tested at any part of the experiment.” In order to get the uniform longitudinal magnetic fields stronger than the vertical component of the Harth’s magnetism, a long solenoid was used, which had 7707 turns, in four layers, and total length of 2386 cms. The internal magnetic field of this solenoid was perfectly uniform in strength for a distance of 226 ems., or to within 5 ems. of each end. ‘The wire under test was suspended in the middle of this solenoid, in a magnetic field of 1 Scient. Proc. Roy. Dublin Soc., 1909, vol. xii., p. 101. 2 Ibid., Fig. 1, p. 108. SCIENT., PROC. R.D.S., VOL. XII., NO. XVII. 26 176 Scientific Proceedings, Royal Dublin Society. uniform strength, by means of two three-jaw clutches at a, a, fig. 1. These clutches were made in the ends of brass rods, half an inch in diameter, and gripped the wire firmly; the top rod was fixed to a heavy wooden beam in the ceiling of the room, and to the lower rod a brass vibrator was fixed, the weight of which could be altered by means of discs of lead. From the lower end of this vibrator an iron wire dipped into a mercury-cup, so as to complete the circuit through the galvanometer. ‘The brass rods and clutches were both covered with rubber-tubing insulation, so as to prevent them from coming in contact with the inner tube of the solenoid, and thereby avoid any leakage Fig. 1. of the transitory electric current, which is produced when the lower end of the iron wire is twisted. A secondary battery, an ammeter, a reversing key, and suitable rheostat, were put in series with the solenoid, so as to produce the desired longitudinal magnetic field round the wire under test. That the transitory electric current produced when an iron wire is twisted in a magnetic field is entirely due to the presence of the field was shown by the following experiment :—An annealed, No. 16 iron wire, being suspended 3rown— Mechanical Stress and Magnetisation of Iron. 177 in the middle of the solenoid, the circuit was completed through the galvano- meter. A current was then sent round the solenoid in such a direction as to annul the vertical component of the Harth’s magnetic field; and when the lower end of the wire was twisted through an angle of 100° no trace of current was produced through the galvanometer; but, when the circuit of the solenoid was open, and the Earth’s vertical field of 0°45 c. g. s. units allowed to go round the wire, a twist of 100° on the end of it gave a deflection of 30 divisions on the galvanometer scale. In all the experiments mentioned in the first section of this communica- tion, the results are given for a maximum twist on the wire of 100°, so that when the thicker wires were under test this angle of twist would be within their elastic limits. As in the previous experiments the cyclic curves were obtained by twisting the lower free end of each wire under test through a series of steps of 20° from 0° to + 100°, then from + 100° to = 100°, and again from — 100° to + 100°, thus completing the cycle. All the cycles so obtained were plotted on millimetre paper, where on the axis of abscisse one centimetre represented 20° of twist, and on the axis of ordinates one centimetre represented 10 divisions on the galvanometer scale; the total areas of all the curves in sq. cms. were then measured. In order to find the effect of varying the longitudinal stress per unit area on the wire when it was placed in a magnetic field greater than that of the Harth’s vertical force of 0-45 ¢.g.s. units, a preliminary experiment was made with a No. 16 iron wire; and it was found that the maximum transitory current and the maximum area of the cyclic curve were obtained when the wire was in a magnetic field of between 2 and 3c. g. s. units: a field of 2°5 units was therefore used. A No. 16 iron wire of cross-sectional area, 20°6 x 10 sq. cms., was carefully annealed and suspended in the middle of the solenoid in which was a magnetic field of 2°5 units; a load of 0:3 x 10° grammes per sq. cm. was put on the end of the wire, and a complete cycle taken with a maximum twist of 100°. The load was then increased, and a cycle again taken, and so on for six different values of the load up to 2 x 10° grammes per sq.cm. The cyclic curves were all drawn to the same scale, and the areas in sq. ems. found ; the largest transitory current for 100° twist was also observed. ‘I'he results are here given in Table I. [Tasuy I, 262 178 Scientific Proceedings, Royal Dublin Society. Taste I. Load on the wire, Total area of cyclic curves Max. trans. current in grammes per sq. cm. in sq. cms. galv. scale-divisions. 0°3 x 10° 55'5 97 O:5, 54°5 94 OF pp 53°0 91 ILO) 5 51°5 87 13) 7 55 47°5 81 PAN) ge 44:0 74 If the numbers in the first column of this table be plotted as abscissee and those in the second column as ordinates, it will be found that the points all lie very approximately in a straight line; and it will also be seen that the area of the cyclic curve or circular magnetism is decreased about 20 per cent. when the longitudinal stress or load on the wire is increased about seven times, i.e. when the wire is in a magnetic field of 2°5 units. In the former paper it was shown that when the same size of wire was placed in the Harth’s vertical field of 0°45 units, and when the longitudinal stress per unit area was increased 2°5 times, the circular magnetisation was decreased 40 per cent. This is what one would expect, for on the curve of magnetic field and twist in fig. 8 of the previous paper! the lower field is on a part of the curve where the slope is rapidly changing, and the higher field is on the peak of the curve, so that any changes in the load and circular magnetism would be more pronounced in the weak field. When the values in the third column of Table I. are plotted as ordinates against the load as abscissa, the points will also be found to lie in a straight line practically parallel with the other line, and for the same increase of the longitudinal stress on the wire, viz. about seven times, the maximum transitory current is diminished about 23 per cent. In order to test the effects of various longitudinal magnetic fields on the different wires when each was under a constant longitudinal stress or load of 10° grammes per sq. em., the five wires, Nos. 12, 14, 16, 18, and 20, were all carefully annealed and successively tested when placed in magnetic fields ranging in value from 0:45 to 14 ¢.g.s. units. ‘he complete areas of the cyclic curves or the circular magnetism were measured, and the maximum transitory current observed in each case; the results obtained are here given in Table II., and shown in curves in fig. 2 and fig. 3. 1 Scient. Proc. Roy. Dublin Soc., 1909, vol. xii., p. 118. Brown—Mechanical Stress and Magnetisation of Tron. Tasie IL. 179 Cross-sectional Magnetic field Total area of cyclic Maximum transitory area of wire inc. g. s. units. curve in sq. cms. Cuurenvam ealyanoueter in sq. ems. scale-diyisions. 0°45 58°95 112 1:0 88°8 168 2°0 89-4 207 2°5 88:0 207 d4:7 x 10° 4:0 80:0 198 6:0 70:2 189 9-0 62°5 175 12:0 58-0 165 14:0 54:8 158 0:45 30°2 68 1:0 46°0 96 2°0 62°8 124 2°5 65:0 127 382°3 x 10° 4:0 61:0 126 6:0 52-4 120 9-0 41°6 102 12-0 33°0 89 14:0 29°4 80 0:45 24:0 30 1-0 32°0 59 2°0 47-0 83 2°5 51:5 87 20°6 x 10-3 4:0 50:0 85 6-0 42-0 79 9°0 32°5 70 12-0 26°5 64 14:0 24°5 61 0:45 1071 14 ~ 1s) 18°5 22 2°0 32-1 45 2°65 30°0 53 11-7 x 10° 4-0 34:6 55 6-0 28°5 47 9-0 23:0 39 12:0 18:1 36 14:0 16°6 ot 0:45 — — 1-0 — — 2:0 9-2 10°5 2°5 11:0 13-0 6°5 x 10° 4-0 12°8 16:0 6:0 12:0 14°8 9-0 9-1 11:0 12-0 6-0 88 14:0 571 8:0 180 Scientific Proceedings, Royal Dublin Society. In fig. 2 are plotted the values of the magnetic field or longitudinal magnetisation as absciss®, and the total area of the cyclic curve or the circular magnetisation as ordinates. The top curve is that obtained with the thickest wire of cross-sectional area, 54:7 x 10° sq. ems., and the lowest curve that got with the wire of 6:5 x 10 sq. ems. cross-sectional area. From these curves we see that the circular magnetism rises to a maximum in a fairly weak magnetic field, and then diminishes as the field strength is increased; and also that the maximum circular magnetism in each of the several wires is obtained for different values of the longitudinal magnetic field. Neglecting the lowest curve, since its peak or highest point is somewhat indefinite, if we pick off from the first four curves the value of the longi- tudinal magnetic field corresponding to the highest point on each curve, and : ce | | | PB HRLEE HH ren [ one ee } — | dae ere ee a E nity 5 70 75 Magnetic field in ¢. g.s. units. Fie. 2. & S Dp SS —_— — Area of Cyclic Curve in sq. cms. plot these magnetic fields as ordinates against the corresponding values of the cross-sectional area of the wires as abscissa, it will be found that the four points le practically in a straight line. This shows that when the cross-sectional area of the wire is increased from 11-7 x 10% sq. ems. to 54:7 x 10° cms., or about 4:7 times, the longi- tudinal magnetic field which produces the maximum circular magnetism decreases trom 3:3 to 1:5 c. g.s. units, or about 54 per cent. Again, if from the same curves in fig. 2 we take the maximum circular magnetisation, or the highest points attained in each of the first four curves, and plot the values as Brown—Mechanical Stress and Magnetisation of Iron. 181 ordinates against the corresponding values of the cross-sectional area of the wire as abscisse, these four points will also be found to lie practically on a straight line; and, as before, when the cross-sectional area of the wire is increased about 4:7 times, the maximum circular magnetism is énereased 2°5 times. In fig. 3 are plotted the longitudinal magnetic fields as abscissee, and as ordinates the maximum transitory current obtained by twisting one end of the wire through an angle of 100°. As one would expect, these curves are of very much the same form as those in fig. 2: the value of the maximum transitory current increases with the increase of the magnetic field round the wire up to a certain value, and then diminishes as the field is further increased. Max. Transitory Current. _ -Magnetic fieldin c. gs. units. Fi. 3. As in fig. 2, the top curve in fig. 3 is that obtained for the thickest wire, and the lowest curve that for the wire of smallest cross-section : i.e., taking the curves from the top downwards, they were obtained with wires Nos. 12, 14, 16, 18, and 20 respectively. If we pick off from the first four curves the values of the magnetic fields which give the highest points on the curves, and plot these values as ordinates against the corresponding values of the cross-sectional areas of the wires as abscissee, the four points will be found to lie in a straight line. This shows 182 Scientific Proceedings, Royal Dublin Society. that as the cross-sectional area of the wire is increased about 4°7 times, the longitudinal magnetic field required to give the highest value of the maximum transitory current is decreased from 3°2 to 2°3 c. g.s. units, or about 28 per cent. Again, if the highest values of the maximum transitory current be picked off from the first four curves in fig. 3, and plotted as ordinates against the cross- sectional area of the wires as abscisse, the four points will be found to he practically in a straight line. ‘This shows that as the cross-sectional area of the wire is ‘nereased 4:7 times, the highest value of the maximum transitory current increases from 56 to 207 galvanometer scale-divisions, or about 3°7 times. The arrangement of the apparatus was now changed so as to obtain the circular magnetisation, and to measure the amount of twist in an iron wire when carrying an electric current and when placed in a uniform longitudinal magnetic field throughout its entire length. The vibrator or weight at the end of the wire was modified in such a way as to allow the reflecting surface of a small concave mirror to be coin- cident with the axis of the wire under test, the wire being at the same time in a uniform magnetic field. ‘he stem of the vibrator which was used in the experiments just mentioned was cut, and a brass tube 6 ems. long and 3 ems. diameter inserted, in which were arranged the mirror and its means of adjust- ment as shown in fig. 4. At the upper end a is a three-jaw clutch for clamping the wire inside the solenoid, and at 6 is a torsion-head with some- what stiff motion, by means of which the lower part of the vibrator can be turned round the axis of the wire with respect to the upper part a b held stationary. By this means the spot of light from the small mirror m could be brought on to the scale when the wire under test was in position inside the solenoid. The mirror 7 is fixed on a small hinged plane surface which can be given two motions, viz., a motion at right angles to the vertical wire by means of a small slide-rest and the screw d, and avery slight motion round the hinge or pivot of the mirror by means of the screw c. ‘The first movement brings the reflecting surface of the mirror in the vertical plane through the axis of the wire, and the second gives a means of getting the spot of light on to the scale. ‘I'he scale used was divided into 700 divisions in a length of 44 cms., and was placed at a distance of 109 cms. from the mirror. In the following tests the longitudinal magnetisation round the wire was obtained by means of the long solenoid, and the circular magnetization by sending a certain definite electric current through the wire; each of these circuits had its own independent accessories in the form of secondary battery, variable resistance, ammeter, reversing key, and plug-key. The Brown—WMechanical Stress and Magnetisation of Tron. 183 maximum current density employed in all the wires was at the rate of 100 amperes per sq. cm., and to obtain a measure of the circular magnetisation the results were all plotted on millimetre paper, so that two centimetres represented one ampere on the axis of abscisse, and on the axis of ordinates one centimetre represented ten divisions of deflection or twist on the scale. TOO Oy Ss giammmmaniiis. VSLSSTMLETO EH hd, —— » "i —— Fia. 4. As was shown in Part I. of this paper,! this method of stating the results gives a very symmetrical cyclic curve; and one would expect that an iron 1 Scient. Proc. Roy. Dublin Soc., 1909, vol. xii., p. 101. SCIENT, PROC. R.D.S., VOL, XII,, NO. XVII. 24 184 Scientific Proceedings, Royal Dublin Society. wire tested in this way would exhibit loops of hysteresis in the same way as in the ordinary B-H curve for iron! In order to test this, a carefully annealed No. 16 iron wire of cross-sectional area, 20°6 x 107? sq. cms., was suspended inside the solenoid, which was arranged to give an internal longi- tudinal magnetic field of 2°5 c.g.s. units; the longitudinal stress or load on the end of the wire being that of the vibrator only, viz, 0:34 x 10° grammes per sq.cm. A complete cycle was taken in which the increments of current through the wire were small—so as to get numerous points on the curve—the maximum value being 2°06 amperes. During the cycle two hysteresis-loops were obtained ; and the values for the cycle and loops are given in Table III., and shown as curves in fig. 5. Tasre IIT. Amperes. Deflection. | Amperes. Deflection. Amperes. Deflection. 2:06 580+ || 0-3 25-0 14 43-0 1'8 575 06 10:0 1-7 46-0 1-6 565 0-9 5-0 - 1-9 520 1-4 55:5 1 17-0 206 | 58-0 1-2 54°5 1:3 29:0 1-5 56-0 0:9 525 15 38-0 iO | 690 0-7 50:0 1:7 46-0 0-7 50:0 0:5 47-5 1-5 45°5 OS | aes 0-3 A) if ee 45-0 O8 | 440 O65 I) ep 11 44:0 0-0 | 360 OY | a%e 0:9 43:0 || 0-3 26-0 ot | apo 0-6 Oy || OS | iro 1:2 47-6 05 39:6 0-7 6:0 1-5 50-0 0:3 37-0 0-9 6-5 + 12 49°5 05 37:4 1-1 18-0 1-0 49-0 06 aes i} 1B | S00 oy | aes 0-9 39:0 || 1-6 43-0 0°5 46-0 11 40:5 1:8 51-0 0:3 44:0 1-2 41-2 || 2-06 58-0 0-0 36-0 | ' Ewing’s ‘‘ Magnetic Induction in Iron and other Metals,’’ 3rd ed., p. 95. Brown—WMechanical Stress and Magnetisation of Iron. 185 Experiments were now made to test the effect of varying the longitudinal stress on the wire when it was placed ina uniform magnetic field throughout its whole length. An annealed No. 16 iron wire was fixed in position inside the solenoid, which was arranged to give a longitudinal magnetic field of —— —— = = re) t “Nw 2-5 ag.s. units. A cycle was taken when a known load was on the wire, the maximum current through the wire being 2°06 amperes; the load on the wire was then increased and a cycle again taken, and so on for eight different values of the load, the longitudinal magnetic field round the wire being 2° 2u2 186 Scientific Proceedings, Royal Dublin Society. units in each experiment. The total area in sq. ems. of each of the cyclic curves was then measured, and the maximum twist also observed in each case ; the results so obtained are given in ‘lable LV. Tasie LV. | Load on wire, Area of cyclic curve Maximum twist grammes per sq. cm. in sq. cms. in scale-divisions. | 0°34 x 10° 33°C | 56:0 OPS) 55 32°2 55°95 Osis 31-4 55:0 I@ 30:0 536 18} 54 29:0 52°5 eG} 55 27-6 52°0 Sie. 26°4 50°5 8) 5 23°0 48-0 If the numbers in the first column of Table IV. be plotted as abscissa, and those in the second column as ordinates, the points will all be found to lie practically in a straight line, and will show that when the longitudinal stress is increased about seven times, the area of the cyclic curve or circular magnetisation is decreased about 30 per cent. If the numbers in the third column of the table be plotted as ordinates against the same values as before on the axis of abscissa, the points of maximum twist will also be found to lie in a straight line, and will show that for the same dncrease in the longitudinal stress (about seven times) the maximum twist is decreased about 14 per cent. These values were obtained in a magnetic field of 2°5 c.g:s. units, or at the peak of the curve, and are therefore less pronounced than if they had been taken in a magnetic field of higher or lower value.’ 1 This same No. 16 wire when re-annealed was subsequently tested magnetometrically for the B-H cyclic curves, when the wire was subjected to two different constant longitudinal loads. The Maximum magnetic force applied was 24°6 c.g.s. units, obtained with the long solenoid, which allowed the wire under test to be 1400 diameters long. ‘The effect of the heavier load was to harden temporarily the wire under test, as is shown by the following results :— | Load in grammes Maximum Loss in ergs per | per sq. cm. Induction. RG Naar. | || Cucnenie Howe, cycle per c.c. | 0°34 x 10° 15060 10700 2-0 12120 2°72 x 105 16810 13300 2°5 13680 These results show that, when the longitudinal stress per unit area on the wire is increased ? 8 i eight times, the maximum induction is increased about 11°6 per cent.; the retentivity and coercive force are each increased 265 per cent., and the loss in ergs about 13 per cent. Brown— Wechanical Stress and Magnetisation of Tron. 187 With a constant longitudinal stress on each wire of 10° grammes per sq. cm., the effects of varying the cross-sectional area of the wire, the longi- tudinal magnetisation, and the circular magnetisation were now tested; and to do this four different sizes of wires were employed, viz., Nos. 18, 16, 14, and 12 standard gauge. ‘These wires were properly annealed, and succes- sively tested when placed in magnetic fields varying from 0°45 to 14 c.g.s, units, the maximum current density through each wire being at the rate of 100 amperes per sq. cm. Lach of the wires was tested in nine different longitudinal maguetic fields and the total area of the cyclic curve for each set of observations measured: the results obtained are given in Table V., and shown in curves in fig, 6. Tasie V. Total area of cyclic curves in sq. cms. Magnetic field meal No.18. | No. 16. No. 14. No. 12. 0°45 5°86 18°66 24°9 30°0 1 10:00 25°04 34°5 43°5 | 2 16°20 29°50 44:0 54°5 3 18:20 31-00 45°5 oa'4 4 18°20 30:00 44-2 52°5 6 14:94 25°02 37°6 46:0 9 10°50 19°12 29°5 38°1 12 8:10 14°80 22°6 32°5 14 7:00 15°00 20:0 29°5 The top curve is that obtained with the No. 12 wire of cross-sectional area, 54°7 x 10~* sq. cms., and the lowest curve with the No. 18 wire of cross-sectional area x 11-7 10°* sq. cms. ‘These curves in fig. 6 are of much the same shape and form as those in fig. 2, which were obtained by twisting the wire in a magnetic field, and measuring the transitory current produced. From the curves in fig. 6 we see that the circular magnetism rises to a maximum in a weak magnetic field, and then diminishes as the field is increased. It will also be seen that the maximum circular magnetisa- tion in each of the several wires is reached for different values of the longitudinal magnetic field. If we pick off from these curves the value of the longitudinal magnetic field corresponding to the highest point om each curve, and plot them as 188 Scientifie Proceedings, Royal Dublin Society. ordinates against the corresponding values of the cross-sectional area of the wires as abscissa, it will be found that these four points lie practically in a straight line. It will be seen from this that when the cross-sectional area of the wire is incieased about 4:7 times, the longitudinal magnetic field which produces the maximum circular magnetisation decreases from 3°d to 25 «g.s. units, or about 28 per cent. If from the same curves in fig. 6 we take the maximum circular magnetisation in each of the four curves, and plot these values as ordinates against the corresponding values of the cross-sectional areas of the wires as abscissa, the points will also be found to lie very Total area of Cyclic Curves in sq. cms. 5 OF 70 15 Magnetic field in ¢. g.s. units. Fic. 6. approximately in a straight line, showing that when the cross-sectional area of the wire is increased about 4:7 times, the maximum circular magnetism is increased about three times. The following conclusions may be arrived at from. these experiments :— A. When a mechanical twist is given to the wire and the transitory electric current measured :-— 1. In an iron wire under constant longitudinal stress and magnetised longitudinally, the circular magnetisation rises to a maximum in a weak magnetic field; and the value of the longitudinal Brown— Mechanical Stress and Magnetisation of Iron. 189 magnetic field which produces maximum circular magnetisation is different with different-sized wires. 2. When the longitudinal load is 10° grammes per sq. em. and the cross- sectional area of the wire is ¢ncreased about 47 times :— (a) The longitudinal magnetic field which produces the maxi- mum circular magnetism is decreased about 54 per cent. (6) The maximum circular magnetisation is increased 2°5 times. (c) The longitudinal magnetic field required to give the highest value of the maximum transitory current is decreased 28 per cent. (d) The highest value of the maximum transitory current is increased 3:7 times. B. When an electric current is sent through the wire and the mechanical twist measured :— 1. The change in the mechanical twist lags behind the change in the circular magnetisation, a cyclic curve and hysteresis loops are obtained. 2. When an iron wire of cross-sectional area, 20:6 x 10° sq. ecms., is placed in a uniform magnetic field of 2°5 c.g.s. units, and the longitudinal stress per unit area on the wire is increased about 7 times, the circular magnetism is decreased about 30 per cent., and the maximum twist of the free end of the wire is decreased about 14 per cent. 38. The value of the longitudinal magnetic field which produces maximum circular magnetisation is different with different-sized wires. 4. When the longitudinal load is 10° grammes per sq. cm. and the cross- sectional area of the wire is increased about 4:7 times :— (a) The longitudinal magnetic field which produces maximum circular magnetisation is decreased about 28 per cent. (6) The maximum circular magnetisation is increased about three times. I am indebted to Mr. F. W. Warwick, B.a., B.E., of the Physics Laboratory, Royal College of Science, for assistance in reading the gal- vanometer scale during some of the experiments for both this and the previous paper, THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 18. JULY, 1909. METHODS OF DETERMINING THE AMOUNT OF LIGHT SCATTERED FROM ROUGH SURFACES. BY W. F. BARRETT, F.R., PROFESSOR OF EXPERIMENTAL PHYSICS, ROYAL COLLEGE OF SCIENCE FOR IRELAND. | Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN : PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price Sixpence. a ET iy Fa Ss A sonian nstityg, . OV & 809 Brown— Mechanical Stress and Magnetisation of Iron. 189 magnetic field which produces maximum circular magnetisation is different with different-sized wires. 2. When the longitudinal load is 10° grammes per sq. cm. and the cross- sectional area of the wire is increased about 47 times :— (a) The longitudinal magnetic field which produces the maxi- mum circular magnetism is decreased about 54 per cent. (6) The maximum circular magnetisation is inereased 2°5 times. (ce) The longitudinal magnetic field required to give the highest value of the maximum transitory current is decreased 28 per cent. (d) The highest value of the maximum transitory current is increased 3°7 times. B. When an electric current is sent through the wire and the mechanical twist measured :— 1. The change in the mechanical twist lags behind the change in the circular magnetisation, a cyclic curve and hysteresis loops are obtained. 2. When an iron wire of cross-sectional area, 20°6 x 10 sq. cms., is placed in a uniform magnetic field of 2°5 c.g.s. units, and the longitudinal stress per unit area on the wire is increased about 7 times, the circular magnetism is decreased about 30 per cent., and the maximum twist of the free end of the wire is decreased about 14 per cent. 3. The value of the longitudinal magnetic field which produces maximum circular magnetisation is different with different-sized wires. 4. When the longitudinal load is 10° grammes per sq. cm. and the cross- sectional area of the wire is increased about 4:7 times :— (a) The longitudinal magnetic field which produces maximum circular magnetisation is decreased about 28 per cent. (6) The maximum circular magnetisation is increased about three times. I am indebted to Mr. F. W. Warwick, B.A., B.E., of the Physics Laboratory, Royal College of Science, for assistance in reading the gal- vanometer scale during some of the experiments for both this and the previous paper. SCIENT. PROC. R.D.S., VOL. XII., NO. XVII. 21 F 100 9 XVIII. METHODS OF DETERMINING THE AMOUNT OF LIGHT SCATTERED FROM ROUGH SURFACES. By W. F. BARRETT, F.RS., Professor of Experimental Physics, Royal College of Science for Ireland. [Read Aprit 20. Ordered for Publication May 11. Published Jury 27, 1909.] Wuen light falls on a rough or unpolished surface, such as a sheet of ground glass, the reflected beam is scattered or diffused in every direction owing to the irregular nature of the surface, and each point of the surface thus becomes a source of light.!. From whatever direction the surface is viewed, it is thus rendered visible ; and the scattered beam is white or coloured according as some wave-lengths are, or are not, absorbed. Even the whitest surface absorbs some of the incident light, but the proportion of light reflected from a more or less smooth surface is increased as the angle of incidence becomes greater, so that with very oblique incidence a fairly smooth surface, such as a sheet of white paper, can give an indistinct image of a luminous body. It is obvious that the amount of light thus scattered by large surfaces, such as a building or white-washed wall, is of great practical importance, more especially in legal disputes where a case of “ ancient lights” is concerned. It was such a case upon which I was consulted, that first drew my attention to the need of accurate information on the relative amounts of light scattered 1 The term ‘scattered light’ is usually and properly restricted to those cases where the particles of the body are small in comparison with the wave-length of light. Under such eircumstances the ordinary laws of reflection are not obeyed, for only single wavelets are formed by the scattering of the incident or primary wave. Lord Rayleigh has shown that for a given amplitude of the primary wave, the amplitude of the scattered wave varies inversely as the square of the wave-length, and the intensity of the scattered light varies inversely as the fourth power of the wave-length. Thisis the case when light is scattered from minute particles of smoke, or milk and water, the scattered light being bluish, since the intensity varies as above stated. In the present paper I use the word ‘scattered ’ to mean ‘ irregularly reflected,’ the particles of the reflecting surface being comparatively coarse, and thus the reflected wave is formed by a reinforcement of wayelets generated at neighbouring points of the surface. In the case of opalescent surfaces, the particles are small, and Lord Rayleigh’s law applies. Barrett—Amount of Light Scattered from Rough Surfaces. 191 by different surfaces. The estimates given were variable, and appeared untrust- worthy ; nor could I find by what method of observation they were arrived at. One difficulty has arisen from the fact that the ordinary photometric method, which involves the law of inverse squares, is inapplicable to large Fig. 1. reflecting surfaces. A modification of that method is possible by receiving the scattered light upon a small screen of ground glass or tracing-paper, and using this as the source of light which can be balanced against a standard 212 192 Scientifie Proceedings, Royal Dublin Society. light. Blackened tubes are necessary to prevent the intrusion of extraneous light into the photometer, which should be of considerable length to diminish the error arising from the magnitude of the translucent screen. But the results so obtained are not very satisfactory, owing to the unavoidable reduction and consequent feebleness of the source of light thus measured. It may therefore be of interest to bring before the Society two other methods of comparing the scattered light with some standard source of light, and reducing the illuminating power of the stronger in some other way than by varying the distance. Method A consists of a rapidly revolving opaque dise with a transparent sector that can be altered in size, and its angular magnitude measured. ‘The apparatus shown in fig. 1 can be driven by hand; a simple speed-gear is all that is necessary. It is placed at a given distance between the reflecting surface, which is illuminated by the sun or strong artificial light, and the photometer. The width of the sector is altered until equality of illumination between the reflecting surface and a standard source of light is obtained as shown by some transmission photometer, such as Bunsen’s, Joly’s, or Lummer and Brodhun’s. If the scattered light is coloured, as from a brick building, in front of the standard light, a wedge of suitably coloured glass, or a coloured gelatine film of increasing thickness, is gradually interposed until a similar tint is obtained. This method of diminishing the intensity of light by a revolving sector of variable size is well known, and has been used for other purposes by Swan in 1849, and Abney in 1890; but, so far as I know, it has not hitherto been employed for the purpose I have described. There is another application of the revolving sector which is very useful for teaching-purposes, and possibly has been so used, but I have not met with it. ‘I'wo equal sources of light are balanced against each other by means of, say, a Joly photometer. ‘he revolving sector is placed before one, and the opening adjusted in successive experiments to, say, 180°, 120°, 90° and 60°, so that on rapid revolution one-half, one-third, one-fourth, and one- sixth of the light on that side falls on the photometer. The distance the other light, or the photometer, has to be moved to restore equality of illumination in each case, demonstrates with ease and accuracy the law of inverse squares. ‘Thus, if the sector be set to 90°, one-fourth of the light is transmitted by the revolving disc; and the photometer has to be placed twice as far from one candle as the other in order to obtain equality of illumination. This experiment is so useful and interesting for young students that I have added it to the regular course of introductory practical physics in my laboratory. Barretr—Amount of Light Scattered from Rough Surfaces. 193 I now pass to Method B, which consists in reducing the intensity of the stronger light by an absorbing medium. For this purpose the following arrangement is adopted; it isin principle somewhat similar to the “ colori- Fie. 2. meter ” often used in chemical analysis, and is an adaptation of the method I recently patented for determining the “light-threshold” of the eye.! The absorbing medium is a liquid of neutral tint, best formed from fine China or 1 See Patent No. 24162, November, 1907, 194 Scientific Proceedings, Royal Dublin Society. Indian ink mixed with water, and allowed to stand for forty-eight hours; the coarser particles are then deposited, and the supernatant liquid is employed. ‘The apparatus is shown in fig. 2 and in section in fig. 3. A variable depth of the liquid is obtained by the movement of a plunger with glass bottom H, fig. 3, which can be gradually immersed within a cylinder or cistern I, also with a glass bottom, containing the absorbing liquid. Light is reflected upwards Fie. 3. through the cylinder by means of a mirror, M, at 45°. The amount of light scattered from various large surfaces can thus be very easily compared by the relative depth of liquid required to produce extinction. The pillar L is graduated, and the cistern I raised or lowered by the rack and pinion K. In Barrert—Amount of Light Scattered from Rough Surfaces. 195 order to exclude extraneous light, the observer rests his forehead in a shaped head-rest A, and a black cloth covers the head. After one minute the eye attains a fairly steady state, and either eye can be used at pleasure by sliding the head-rest to and fro. On the glass bottom of H is a minute photograph of the graduated test type used by oculists. This is viewed through a small lens F, adjustable at , until a sharp image is seen by the observer. When the cistern I is raised until the glass bottom of H and I touch, the scale reading on L then indicates zero. The depth of the liquid (as indicated on the scale) required to produce complete extinction of the light measures the dntrinsic brightness of the source. Or with a constant source of light the depth measures the “light threshold,” or the sensibility of the observer’s eye to light. This sensibility rapidly rises during the first minute of observation, and becomes nearly constant after two or three minutes. The form sense, or “ visual acuity,” of the eye is measured by the depth of liquid required to obscure and produce illegibility of the test type, and this also measures the ‘J/uminating power of the source of light. The illuminating power of the source may be reduced to any given fraction by means of the adjustable and rapidly revolving sector, fig. 1, or by other means; and it will be found that the depth of liquid required to produce extinction of the /ight is practically the same, even when the illumi- nation from the source is reduced to a very minute amount: in other words, the intrinsic brightness remains the same. On the other hand, the legi- bility of the test type varies with the amount of illumination; and it is this we require to measure in the case of light irregularly reflected from rough surfaces. Hence this arrangement affords an accurate method of testing the illuminating power of any surface that scatters light, whether large or small. It is only necessary to use a steady source of artificial light, and note the depth of immersion of the plunger H which is required to produce illegibility when a silvered mirror is employed; then replace or cover the mirror by a similar-sized piece of the reflecting surface to be tested, and note the depth now required for extinction, the distance and intensity of the source of light remaining unchanged. The following table gives the result of a few observations with this arrangement. Calling the depth of liquid 100 when a silvered reflector is employed, the percentages indicate the depth of the liquid required to extinguish the legibility of a given line of test type, when light from the same source is scattered from the various surfaces named. I have to 196 Scientific Proceedings, Royal Dublin Society. thank my assistant, Mr. W. Warwick, n.., for repeating these observations, and also for other assistance in this paper :— Silvered mirror, oe a .. 100 Plane glass surface, .. & ee Oo) Ground glass, e a oo aS White card, Grey card, .. = me oO Dark grey card, ee ae a eiage i Smooth black paper, .. ae so | OA) Black cotton cloth, .. ee ah 16 Dull black woollen cloth, .. ae 5 As light is scattered at all angles, the reflecting surface can be more or less inclined, and the relative reflecting powers at different incidences thus determined. When the scattered light is coloured, the exact amount scattered can be determined by interposing a suitably coloured glass screen when the silvered mirror is used, and comparing the depth of liquid then required for extinction with that required when the light is scattered from the coloured surface. If, instead of finding the depth of the neutral tinted liquid required to obscure the reading of the test type, we measure the depth required to extinguish the light scattered from various surfaces, the remarkable fact is revealed by these preliminary experiments, that even with dark surfaces, such as black paper or cloth, the intrinsic brightness is not reduced much more than 40 per cent., though the illuminating power is reduced 80 to 90 per cent. This subject requires, and I hope shortly to give it, further investigation. ‘lhe apparatus also affords a means of measuring the relative sensibility of the fovea and of the retina immediately surrounding it to (i) the light sense, (ii) the colour sense, and (iii) the form sense or visual acuity. Furthermore, this method of absorption enables one to determine with accuracy the somewhat difficult problem of the relative value of different systems of lighthouse illumination ; for, as already shown, it measures the perceptive power of the eye for any given light, and this depends on the brightness, as distinguished from the illuminating power, of different sources of light. The theoretical law which expresses the absorption of light by a homo- geneous medium of varying thickness is well known. In the particular case of the foregoing instrument, as every successive increment in the depth of the Barraerr—Amount of Light Scattered from Rough Surfaces. 197 liquid through which the light passes reduces the intensity of the transmitted light by a given ratio, it follows that as the depth of the absorbing liquid increases in arithmetical progression, the intensity of the light is diminished in geometrical progression, and accordingly the law is expressed by an exponential function. This law, however, only holds true if no light were lost by internal reflection and scattering within the liquid, and if the light is of a given wave-length. By employing the revolving sector, fig. 1, which can be adjusted to reduce by definite amounts the incident light, the coefficient of transmission of white, or of monochromatic, light through a neutral tinted or coloured liquid can be experimentally determined ; and thus the actual law of absorption in any given case can be found. To this I hope to return in a subsequent paper. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 19. JULY, 1909. A NEW FORM OF POLARIMETER FOR THE MEASUREMENT OF THE REFRACTIVE INDEX OF OPAQUE BODIES. BY VG ID, Jevaededdlshy Te cleSe, PROFESSOR OF EXPERIMENTAL PHYSICS, ROYAL COLLEGE OF SCIENCE FOR IRELAND. | Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price Sixpence. ot sonian Instity~ me a Ya NOV 4 Po = Barrett—Amount of Light Scattered from Rough Surfaces. 197 liquid through which the light passes reduces the intensity of the transmitted light by a given ratio, it follows that as the depth of the absorbing liquid increases in arithmetical progression, the intensity of the light is diminished in geometrical progression, and accordingly the law is expressed by an exponential function. This law, however, only holds true if no light were lost by internal reflection and scattering within the liquid, and if the light is of a given wave-length. By employing the revolving sector, fig. 1, which can be adjusted to reduce by definite amounts the incident light, the coefficient of transmission of white, or of monochromatic, light through a neutral tinted or coloured liquid can be experimentally determined ; and thus the actual law of absorption in any given case can be found. To this I hope to return in a subsequent paper. SCIENT. PROC. R.D.S., VOL. XII., NO XVII [ 18 4 XIX. A NEW FORM OF POLARIMETER FOR THE MEASUREMENT OF THE REFRACTIVE INDEX OF OPAQUE BODIES. By W. F. BARRETT, F.RS., Professor of Experimental Physics, Royal College of Science for Ireland. [Read Aprit 20. Ordered for Publication May 11. Published Juny 27, 1909.] In 1814 Sir David Brewster, after an extensive series of measurements of the angle of maximum polarization of various reflecting surfaces, was led to the well-known and important law which bears his name, viz., that the index of refraction of any substance is the tangent of the angle of maximum polarization for that substance. Hence, when a ray of light incident on a transparent body is polarized by reflection, the reflected ray forms a right angle with the refracted ray. By means of Brewster's Law the indices of refraction of various opaque non-metallic reflecting surfaces have been obtained. As every different colour has a different index of refraction, the law shows that the polarizing angle correspondingly varies with the different rays of the spectrum, being, for a given substance, smallest in the red and largest in the violet. In bodies of low dispersive power the angle of maximum polarization is nearly the same for all colours, and white light can be used as the source. In other cases monochromatic light must be employed—either a sodium flame or suitably coloured glass in front of the source described below. The amount of light reflected from some opaque bodies is small ; and hence the determination of the polarizing angle is difficult, unless we can always keep the analyser placed in the reflected beam at the same angle as the ray incident on the opaque surface under examination. ‘To secure this I have devised the following instrument (fig. 1), whereby with a rack- work and simple link-motion, the collimator, which renders the incident rays parallel, and the telescope, in which is placed the analysing Nicol’s prism, are simultaneously moved through equal angles. ‘The opaque body is placed on a movable table with levelling screws, which are adjusted until the reflecting surface is level, and at the centre of the graduated circle, round which travel the telescope and collimator. A small but brilliant source of light is employed, such as a Nernst or a 10-volt electric glow-lamp; a small lens throws a brilliant image of the light on to the adjustable slit of the collimator. This Barrert—A New Form of Polarimeter. 199 latter contains a lens in a draw-tube, so that a parallel beam falls on the opaque reflecting surface; and a sharp image of the slit is obtained by the lens in the telescope, which also contains a small Nicol’s prism. Upon turning the rack-work handle, the source of light and collimator move Fig. 1. together, and through the same angle as the telescope. The observer now turns the polarizing plane of the Nicol at right angles to the plane of the reflected polarized ray, and watches the gradual extinction of the light as the polarizing angle is approached. At the angle of maximum polarization the 200 Scientific Proceedings, Royal Dublin Society. light is extinguished (reappearing as the angle is passed), the clamping- screw is then turned, and by means of the vernier the angle is read off in degrees and minutes. As the polarizing angle of nearly all substances lies between 48° and 68°, the circle need not be finely graduated for more than 20°. An enlarged view of scale and vernier is shown in fig. 2. It is, of course, essential that the light incident on the opaque surface be strictly parallel; and careful adjustment of the collimator must be made beforehand for this purpose. Provision is also made in the instrument for two other adjustments, namely (1) the coincidence of the axis of collimation and the Fic. 2. zero of the scale, and (2) of the reflecting surface to be tested with the centre of the graduated circle. This latter is accomplished by a small projecting arm with a platinum point, shown in fig. 2. Liquids are placed in a little glass capsule on the levelling-table, which is adjusted until the platinum point, indicating the centre of the circle, just touches the liquid surface. Bodies having an irregular, granular, or crystalline surface, if fusible, are melted. This is accomplished by placing them in a small capsule of metal or porcelain, which is heated by a current of steam or an electric current traversing a platinum wire coiled round the capsule. Barrettr—A New Form of Polarimeter. 201 In practice a difficulty occurs in determining the precise angle of maximum polarization ; for the extinction of the reflected ray seems to spread over a narrow region rather than occur at a definite point. This error can be lessened by careful attention to the parallelism of the incident rays and the homogeneity and intensity of the light. I have employed a small direct vision-prism in the collimator and obtained a sharp spectrum, using, of course, in this case a very brilliant source of white light. In this way I hoped to obtain the angle of extinction for a definite colour, and thus see a dark band pass across the spectrum, as the polarizing angle for each colour was reached. But the result, as might be expected, was not very successful in bodies of low dispersive powers; a faint, broad shadow is observed to move across the spectrum, the best position to read being when the shadow is in the green or greenish-blue: results can then be obtained within 20’ to 30’, even with this preliminary apparatus. In the case of bodies of very high dispersion, like the nitroso-dimethyl aniline, as pointed out by Prof. Wood in his excellent work on Optics, the determination of the polarizing angle is easy, as the dark band is then sharp and well-defined. If it were the rotation of the plane of the polarized light, and not the amount of polarization, any of the methods used in saccharimetry could be employed for accurately determining the polarizing angle. Possibly an effect similar to the half-shadow polarimeter might be obtained by using a divergent incident beam and a split lens, with inclined axes, in the telescope. What is needed is a comparative effect ; and for this purpose a double-image prism in the telescope, instead of a Nicol, would be worth trying. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XID. (N.8.), No. 20. JULY, 1909. ON MONTANIN AND MONTANA (MONTAN) WAXKS. BY HUGH RYAN, M.A., D.Sc, F.R.U.L, AND THOMAS DILLON, M.A. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETLA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price Stxpence. tan instet or Instit riqny A LSU eae Wo, x ’ Barrutt—A New Form of Polarimeter. 201 In practice a difficulty occurs in determining the precise angle of maximum polarization ; for the extinction of the reflected ray seems to spread over a narrow region rather than occur at a definite point. his error can be lessened by careful attention to the parallelism of the incident rays and the homogeneity and intensity of the light. I have employed a small direct vision-prism in the collimator and obtained a sharp spectrum, using, of course, in this case a very brilliant source of white light. In this way I hoped to obtain the angle of extinction for a definite colour, and thus see a dark band pass across the spectrum, as the polarizing angle for each colour was reached. But the result, as might be expected, was not very successful in bodies of low dispersive powers; a faint, broad shadow is observed to move across the spectrum, the best position to read being when the shadow is in the green or greenish-blue: results can then be obtained within 20’ to 30’, even with this preliminary apparatus. In the case of bodies of very high dispersion, like the nitroso-dimethy] aniline, as pointed out by Prof. Wood in his excellent work on Optics, the determination of the polarizing angle is easy, as the dark band is then sharp and well-defined. If it were the rotation of the plane of the polarized light, and not the amount of polarization, any of the methods used in saccharimetry could be employed for accurately determining the polarizing angle. Possibly an effect similar to the half-shadow polarimeter might be obtained by using a divergent incident beam and a split lens, with inclined axes, in the telescope. What is needed is a comparative effect ; and for this purpose a double-image prism in the telescope, instead of a Nicol, would be worth trying. SCIENT. PROG, R.D.S., VOL, XII., NO. XIX. 2h [ ROe | ON MONTANIN AND MONTANA (MONTAN) WAXES. By HUGH RYAN, MAS DSc FRU, University College, Dublin, and THOMAS DILLON, M.A. [Read May 25. Ordered for Publication Junn 8. Published Juny 17, 1909. ] Tx connexion with a paper on the Irish Peat Industries, which one of us read before the Royal Dublin Society in March, 1908, we obtained a specimen of a commercial wax called in this country Montana Wax. As the substance was said to have been obtained from peat, we thought an examination of it might be of some interest. We found that the wax consisted of a “free” acid (M.P. 83°C.), and an unsaponifiable portion (M.P. 58-59° C.), and was therefore of a very different composition from that recently found for Peat Wax by R. Zaloziecki and J. Hausmann'. By saponifying Peat Wax the latter chemists obtained an alcohol (C2H400,) M.P. 124-130°C., an acid (C,,H,;0;) M.P. 184°C., and another acid (C.,H,;0;) which did not melt below 260° C. Tn connexion with their theory of the origin of petroleum from diatoms, G. Kriimer and A. Spilker have examined* the wax extracted by means of hot benzene from pyropissite and the peat of low-land bogs. By saponifying the wax they obtained an alcohol M.P. 76-78° C., and a erude acid of mean formula C,.H,,O., which they found to be probably a mixture of arachidic, behenic, and lignoceric acids. Montana Wax, therefore, differs markedly from Peat Wax; but, on the other hand, it seems to be identical with the Montan Wax prepared according to the process of Hi. von Boyen* from Saxo-Thuringian lignite. By distillation of lignite in a current of superheated steam, or by extraction of the dried substance with benzene, a bituminous body was obtained, which, when distilled with steam heated to 250° C., afforded a yellow waxy solid. 1 Zeitschr. fiir angewandte Chemie, xx. (1907), p. 1141. 2 Ber. xxxil. (1899), p. 2940, and Ber. xxxy. (1902), p. 1212. 3 Patentbl. xe. (1899), p. 97. Ryan & Ditton—On Montanin and Montana (Montan) Wazes. 203 Montan Wax is described by von Boyen' as a white high-melting wax, of greater value for the manufacture of candles than paraffin or stearine, After purification with benzene it consists, according to von Boyen, of an acid O,,H;O, (M.P. 84-85° C.), and another substance, probably an alcohol (M.P. 60°C.) the two of which are charred by concentrated sul- phuric acid. The bitumen was regarded as the non-crystallized ester of the “alcohol” and acid, and was supposed to be hydrolysed during the steam distillation. The acid in Montan Wax was named Montanic Acid by von Boyen, and Geocerinie Acid by Hell.? Ulzer and Pastrovich® examined a specimen of Montan Wax, and found that it had an Acid Number 100°88, and contained only a trace of ester (saponification number, 101:37). J. Marcusson‘ found that crude Montanic Acid had an Acid Number 143 and consisted of insoluble hydroxy-acids, melting above 100°C., and a waxy acid melting below 100°C. Cholesterin was present in the crude, but absent from the purified, unsaponifiable matter. Both the unsaponifiable and the saponifiable constituents were dextro-rotatory. Montana War.—Montana Wax is a yellowish, crystalline, waxy solid, which is scarcely soluble in alcohol, ether, chloroform, or petroleum-ether in the cold, but easily soluble in the warm, solvents. It has a faint odour of petroleum, and melts at 76°C. The Acid Number of the wax was determined in the usual way, by titrating its solution in hot alcohol with N/2 alcoholic potash, using phenol- phthalein as indicator. Owing to the sparing solubility of the potassium salt of the acid in alcohol, the solution must be kept near its boiling-point during the titration ; and to avoid hydrolysis of the potassium-salt, addition of water in the preparation of the alcoholic potash should be avoided. The Acid Number was 73:3, and cnly a trace of ester was found (the saponification number being 73:9). The Lodine Number of the wax was determined by allowing 0°5 grm. wax, dissolved in chloroform, to interact with 20 c.c.’s of the Hiibl- Waller Iodine reagent for about 12 hours. The lodine Number was 16-0, indicating the presence of an unsaturated compound in the wax. A small quantity of the wax was tested for resin, by boiling with nitric 1 Zeitschr. fiir angewandte Chemie, xiy. (1901), p. 1110. 2 Thid., xiii. (1900), p. 556. 3 Chem. Rey., Fett- und Harz-Industrie, x. (1903), p. 277. 4Chem. Rey., Fett- und Harz-Industrie, xy. (1908), p. 193. 242 204 Scientific Proceedings, Royal Dublin Society. acid, cooling, diluting with water, and adding excess of ammonia; the .non-development of a reddish-brown colour showed that the substance did not contain resin. By the action of sulphuric acid on the cold saturated solution of the wax in acetic anhydride, the play of colours, characteristic of cholesterin, was not obtained. The percentage by weight of the acid in the wax was determined by neutralization with alcoholic potash, addition of pure dry sand, evaporation to dryness, and extraction of the unsaponifiable portion by warm, low- boiling petroleum-ether in a Soxhlet apparatus. A. 2 grammes Montana Wax gave 0°9410 grm. of unsaponifiable matter, equivalent to 47°05 per cent. B. 2 grammes Montana Wax gave 0:9382 grm. unsaponifiable matter, equivalent to 46-91 per cent. The mean (47) of the two experiments is taken as the percentage of unsaponifiable matter in Montana Wax, and the percentage of acid is therefore 53. The unsaponifiable portion, extracted as described above, was found to be free from potassium salts. The Acid Number of the wax being 73:3, and the percentage of “ free acid” in it being 53, it follows that the Acid Number of the “free acid”’ in the wax must be 138-3, and its molecular weight 406, assuming the acid to be monobasic. Following von Boyen, we shall call the crude acid of Montana Wax, Montanic Acid. Montanic Acid.—As it seemed probable that the crude acid constituent of Montana Wax was a mixture of two or more acids, inasmuch as the molecular weight, 406, would correspond to a formula—C,,;H;,0,—containing an uneven number of carbon atoms in the molecule, whereas the higher-saturated fatty acids, which occur naturally, such as Arachidic, Behenic, Lignoceric, Cerotie, and Melissic Acids, all contain even numbers of carbon atoms in their molecules, we separated the potassium-salts of the acids from the sand by extracting with hot water, and decomposed the resulting soap solution by addition of a slight excess of dilute hydrochloric acid. The separated fatty acid was neutralized with potash, evaporated to dryness, and exhausted with petroleum-ether. ‘The regenerated acid was dissolved in absolute alcohol (hot), decolorized with animal charcoal, and filtered. The Montanic Acid, which crystallized on cooling the alcoholic solution, was filtered, washed with absolute aleohol and anhydrous ether, and then dried in a steam-oven. It melted sharply at 83° C., had an Acid Number of 131°6, corresponding to a Ryan & Ditton—On Montanin and Montana (Montan) Waxes. 205 molecular weight of 426 (C.,H;,O. requires 424°45), and gave on analysis the following results :— 0:1675 grm. subst. gave 04860 grm. CO, and 0°1985 gym. H.0. C79:13 H 1817. C.sH;,O2 requires C 79:24 H 13:21. “ Montanic” Acid, whose mean Acid Number is 138°3, consists therefore mainly of an acid whose molecular weight is 424, with a smaller quantity of acids of lower molecular weight. The purified Montanic Acid is nearly insoluble in alcohol or ether, in the cold, but readily soluble on warming with these solvents. It is slightly soluble in cold, easily in hot chloroform, benzene, or petroleum-ether, from which it crystallizes in curved needles. Unsaponifiable portion.—The unsaponifiable portion of Montana Wax crystallizes from hot benzene in glistening scaly masses of fine curved needles, having a specific gravity of 0°92 and M.P. 58-59°C. (uncorr.). It is scarcely soluble in cold alcohol or ether, is slightly soluble in chloroform or petroleum-ether, and readily soluble in the hot solvents. Unlike the saturated hydrocarbons, it dissolves easily in hot absolute alcohol. Unlike the aleohols, it was insoluble in hot acetic anhydride, with which it did not react to form an acetyl derivative. Two analyses of the unsaponifiable portion were made with the following results :— A. 0:1330 grm. subst. gave 04080 grm. CO., and 0°1650 grm. H,O corresponding to C 83°66, H 13°79, O 2°5 per cent. B. 0:1250 grm. subst. gave 0°3830 grm. CO,, and 0:1575 grm. H,O corresponding to C 83:56, H 14-0, O 2:44 per cent. The sum of the percentages of carbon and hydrogen being less than 100 forced us to conclude that the substance is not a pure hydrocarbon, and left the question of its alcoholic nature still open. The fact that acetic anhydride does not react with the unsaponifiable portion to form an acetyl derivative showed that the wax cannot contain appreciable quantities of primary or secondary alcohols. The absence of primary alcohols was confirmed by us by another method :— When one of the higher primary alcohols is heated with potash-lime for some hours to a temperature of 250° C., it is oxidized to the potassium-salt of the corresponding acid. Thus myricyl alcohol gives potassium melissate and hydrogen. C,,H;,CH,OH + KOH = C,,H;,COOK + 2H.. 206 Scientific Proceedings, Royal Dublin Society. The product of the action of potash-lime on the higher primary alcohols is therefore insoluble in petroleum-ether; while, on the other hand, when the potash-lime is heated with hydrocarbons, no reaction takes place, and the latter can be extracted from the residue by means of hot petroleum- ether. We have already seen that Montana Wax contains 53 per cent. of acid, and 47 per cent. of unsaponifiable matter. If the latter be an alcohol (primary), the wax, when heated with potash-lime, should become completely insoluble in petroleum-ether. 3°217 grms. of the unsaponifiable matter were heated for two hours to 250°C. with a mixture of potash and soda-lime ; and on extracting the residue with petroleum-ether, 2°75 grms. of extract were obtained, corresponding to 85°48 per cent. of hydrocarbons or other substances not fixable by potash under the conditions of experiment. The Hibl-Waller Iodine Number of the unsaponifiable matter was 31:13; that for the crude acid was 4°9; and assuming that the wax contained 58 per cent. acid and 47 per cent. unsaponifiable matter, its Lodine Number should be 17.2. Direct experiment gave 16:0 (v. supra). Similar results to the above were obtained with another specimen of Montana Wax examined by us. While engaged in the examination of Montana or Montan Wax, we obtained a sample of a similar wax, called Montanin Wax, which was apparently of a similar chemical composition to the former, but, nevertheless, had very different physical properties. We can find no references to Montanin Wax in the literature. Montanin War.—Montanin Wax is a white, hard wax, capable of being powdered in a mortar, of a high melting-point (95-97° C.), and having a specific gravity 0°980 at 15° C. Its Acid Number was 56:9, Ester Number 1:0, and Saponification Number 57-9. Like the Montana Wax, it contains only a trace of ester. The percentage of unsaponifiable matter in the wax was determined by the method described above. 1-4 grm. of wax gave 0:4866 gim. unsaponifiable matter, corresponding to 34°8 per cent. The unsaponifiable portion melted at 58-59° C., had a specific gravity of ~ 0:92, did not react with acetic anhydride to form an acetate, and gave on analysis the following results :— 0:192 grm. sbst. gave 0°5895 grm. CO, and 0°2430 grm. H,0, corre- sponding to C 83°73, H 14:06, O 2°21. Ryan & Dinton— On Montanin and Montana (Montan) Wares. 207 The appearance, melting-point, specific gravity, lodine Number (30:4), and solubilities prove that it is identical with the unsaponifiable matter of Montana or Montan Wax. We isolated the acid from Montanin Wax, found that it melted at 83°C., and that in appearance and solubilities it was identical with Montanic Acid. The Acid Number of the purified acid was 131°9 (C,,H;,O. requires 182°3). The acid and unsaponifiable constituents of Montana and Montanin Waxes having been thus shown to be identical with the acid and unsaponifiable constitituents of Montana Wax, it was not at once evident why a decrease of 12 per cent. in the amount of unsaponifiable matter present in the wax should raise the melting-point from 76 to 95° C., and alter so completely the appearance of the wax. The Acid Number of Montanin Wax (56'9) being lower than that (73:3) of Montana Wax, it was apparent that the former must contain a smaller percentage of free acid than the latter; and as the Montana Wax also contains a higher percentage of unsaponifiable matter than the former, the Montanin Wax must contain a substance, insoluble in petroleum-ether, which is not present in Montana Wax. After many experiments we found that a portion of Montanin Wax is insoluble in chloroform, and that the wax, when burned, left an appreciable quantity of ash, which consisted of sodium carbonate, with a trace of free alkali. 5 grms. of Montanin Wax gave 0:1386 grm. ash, corresponding to 2772 per cent. ash, which on titration with N/2 EICl. was equivalent to 2'83 grms. Na,CO3. As it was apparent that the sodium was present in the Montanin Wax as sodium montanate, we determined the percentage of the latter in the wax by the direct method (extraction of non-soapy matter by petroleum- ether). 1:2500 grm. wax gave 0:9516 grm. extract, corresponding to 23°87 per cent. of sodium salt. If we assume that the mean molecular weight of the acid in Montanin Wax is the same as that (406) already found above for the acid in Montana Wax, and calculate from the ash determina- tion the percentage of sodium montanate in the wax, we find 23 per cent., which agrees approximately with the percentage (23°87) got by direct determination. With the same assumption we find from the Acid Number of Montanin Wax (56:9) that the percentage of ‘free acid’ in the wax is 41:14, while by subtracting the sum (58°67) of the percentages of unsaponifiable matter (34:8) 208 Scientific Proceedings, Royal Dublin Society. and sodium montanate (23:87) from 100, we find for the percentage of ‘free acid’ in Montana Wax 41°33. The Hibl- Waller Iodine Number of Montanin Wax was found to be 10:5, while that caleulated from the Iodine Number (30°4) and percentage of unsaponifiable matter in the wax is 10:6. The direct and indirect determinations agree within the limits of experimental error in fixing the composition of Montanin Wax as— Montanic acid .. ave 41°38 Sodium Montanate me 23:87 Unsaponifiable .. ie 34°80 100-00 We next boiled Montanin Wax with water, and set free the acid from the soap solution which formed under the wax. The Montanin Wax thus freed from sodium salt melted at 76-78° C., had a higher Acid Number than Montana Wax, but was otherwise similar in appearance and properties to Montana Wax. Montanin Wax seems, therefore, to be obtained from Montana Wax containing 65 per cent. of ‘ free’ acid by neutralizing a portion of the ‘ free’ acid by sodium hydroxide; its dense, hard structure and its high melting- point are due to the presence in it of sodium montanate. The unsaponifiable portions of Montana and Montanin Waxes are unsaturated, while the acids belong to the saturated series. Trish Lignite Wax.—To compare the wax contained in Irish lignite with Montana Wax, we obtained, through the kindness of Mr. H. J. Seymour, a sample of lignite from Straid, Co. Antrim. The lignite was of a dark, earthy nature, had a specific gravity of 1:55, contained 9°45 per cent. of water, and 43:7 per cent. of ash (in dried lignite). 27°145 grms. of the dry powdered lignite, extracted with petroleum-ether in a Soxhlet apparatus, gave 0:064 grm. of wax, corresponding to 0.24 per cent. wax in dry lignite. On drying, powdering, and extracting 1134 grms. of the lignite, we obtained 2°37 grms. of a brownish waxy solid, which had an Acid Number 70°28, and contained only a trace of ester. The amount of wax at our disposal being too small for a complete examination, we merely determined its specific gravity (0°989), melting- point (72° C.), Acid Number (70°28), the melting-point of its free acid (80°C.) Ryan & Ditton—On Montanin and Montana (Montan) Waxes. 209 and of its unsaponifiable portion (58-59° C.)—all of which agree fairly closely with the corresponding constants for Montana Wax. In conclusion, we wish to draw attention to the fact that Montana and Montanin Waxes contain acids whose mean molecular weight is higher than that of Cerotic Acid, and, therefore, that processes for the analysis of Beeswax, which depend mainly on the estimation of the molecular weight of the ‘free acid, cannot give correct results in the presence of Montana or Montanin Waxes. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XID. (N.S.), No. 21. JULY, 1909. THE ANALYSIS OF BEESWAX. BY HUGH RYAN, M.A., D.Sc., F.R.U.L, UNIVERSITY COLLEGE, DUBLIN. [| Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price Sixpence. - pad Re mh a \sonian NSE; Zio “ap { ‘S £ an ~ Resse “iy ok Ryan & Ditton—On Montanin and Montana (Montan) Waves. 209 and ofits unsaponifiable portion (58-59° C.)—all of which agree fairly closely with the corresponding constants for Montana Wax. In conclusion, we wish to draw attention to the fact that Montana and Montanin Waxes contain acids whose mean molecular weight is higher than that of Cerotic Acid, and, therefore, that processes for the analysis of Beeswax, which depend mainly on the estimation of the molecular weight of the ‘free acid,’ cannot give correct results in the presence of Montana or Montanin Waxes. SOCIENT. PROO. R.D.S., VOL, XII., NO. XX. L ZO J XXI. THE ANALYSIS OF BEESWAX. By HUGH RYAN, M.A., D.Sc, F.R.U.L, University College, Dublin. [Read May 25. Ordered for Publication Junz 8. Published Jury 17, 1909.] Ir we regard beeswax as a secretion of the honey bee (Apis mellifica), we should expect that the composition of the wax would be more or less constant, and that it would to some extent depend on the variety of the bee by which it was secreted. The determination of constants, such as the melting-point, the specific eravity, Acid Number, Ester Number, Ratio Number, and Iodine Number, showed that the composition of genuine beeswax varies only between comparatively narrow limits, while the composition of other waxes, such as Indian Ghedda wax, secreted by Apis dorsata, florea, and indica, and bumble-bee wax, secreted by Bombus terrestris, may differ widely from that of true beeswax. ‘The following table illustrates the difference between true beeswax and Indian Ghedda wax :— Melting | Specific Acid | Ester Ratio Iodine Point. | Gravity. Number. Number. Number. Number. | Beeswax, 62-65°C. | 0:96-0:97 18-22 70-78 3°5-3'8 2-12 Ghedda wax,!| 62-66°C. —_— 6-10 76-112 11-18 6-10 The more commonly employed criteria of the purity of a beeswax sample are the melting-point, the specific gravity, the Acid and Hster Numbers. It is, however, possible to prepare so-called “composition” waxes, containing no beeswax, which will have Acid and Hster Numbers lying between the limits given in the above table for beeswax. Thus a mixture of 10:15 parts of stearic acid, 38°6 parts of beef tallow, and 51:25 parts of paraffin will have Acid Number 20, Ester Number 75, and contain 51:25 per cent. of unsaponifiable matter; but to determine the percentage of such a com- 1G. Buchner. Chem. Ztg. xxx. (1906), p. 528. Ryan—The Analysis of Beeswaw. 211 position wax in adulterated beeswax, it is simply necessary to find the percentages of glycerides and hydrocarbons in the wax. If, however, the glycerides (tallow or Japan wax) be replaced by spermaceti or carnauba wax, the solution of the problem becomes more difficult. For example, a sample of wax containing 10:15 per cent. of stearic acid, 58:12 per cent. of spermaceti, and 31°78 per cent. of paraffin, will have Acid, Ester, and Iodine Numbers equal to the corresponding Numbers for beeswax, and in addition will, like beeswax, be free from glycerides. The percentage (31°73) of hydrocarbons in the “ composition ”’ wax is, however, much higher than that (13) found in beeswax by Buisine’s method. In the presence of carnauba wax the difference in the percentages of hydrocarbons in the genuine wax and in the composition wax may be made very much less. Such a composition wax will, however, contain a smaller percentage of “free” acids than that which is present in genuine beeswax. In fact, while the Acid Number of the “free” acids in genuine beeswax is about 141°d, that of the “free’’ acid in the composition wax is usually about 197, so that, in addition to determining the Acid and Ester Number of the wax, it will be necessary to find the percentage by weight of the “free” acid contained in it, in order to be able to calculate the Acid Number of the “free” acid and the relative amounts of cerotic and stearic acids present, from which the percentage of beeswax (if any) in the mixture can be determined. In a paper read before the Society of Public Analysts, in January, 1909, Otto Hehner proposed a method for the determination of the percentage of beeswax in complex wax mixtures, which is based on the assumption that the molecular weight of cerotic acid is higher than that of any other “ free” acid that may occur in composition waxes intended for the manufacture of candles. : Simultaneously with, and independently of, Hehner, the principle mentioned above, which is theoretically identical with that adopted by the latter chemist, was used by me for the analysis of beeswax candles. The method proposed by Hehner is, however, somewhat different in detail from that adopted here. Hehner determines the Acid Number of the wax, dissolves the soap formed in the requisite quantity of warm water, removes the cake of “ester”’ which separates on cooling, and extracts the last traces of ‘“ ester’’ from the soap solution by means of warm ether. ‘The insoluble acids are then set free from the soap, washed, and dried. The mean molecular weight of the insoluble acids is determined by titration with alcoholic alkali; and from 2m 2 212 Scientific Proceedings, Royal Dublin Society. this, by assuming that the mean molecular weight of the “ cerotic” acid of bleached beeswax is 407, and that of the “stearic” acid of commercial “stearine”’ is 272, the relative quantities of “cerotic” and “stearic” acids in the insoluble “free” acids can be calculated. From the mean molecular weight of the “ free” acids in the wax and the Acid Number of the wax, the percentage by weight of the “free’’ acids, and therefore the percentage of cerotic acid, can be deduced. Taking the percentage of cerotic acid in bleached beeswax as 15:8 (a number higher than that—13-32—which I have found for unbleached beeswax by the method described further on), then by multiplying the percentage of cerotic acid found in the wax by 6°3 the percentage of beeswax in the mixture will be obtained. In the method used by me the Acid Number of the wax is first found by heating four grammes of the wax with sixty cubic centimetres of neutral absolute alcohol in an Erlenmeyer flask of Jena glass until the alcohol begins to boil; a few drops of phenol-phthalein are added, and the acidity of the wax determined by neutralization with N/2 alcoholic potash. ‘To the warm mixture about 40 grammes of clean, dry sand are added, and the contents of the flask are evaporated to dryness in a steam oven. ‘The residue is transferred as completely as possible to a Soxhlet filter-cone, which has been previously dried in a steam oven; the flask is washed out with some _ warm dry sand and then with hot low-boiling petroleum-ether (the sand and petroleum-ether being poured into the filter-cone, which has meanwhile been placed in a Soxhlet extractor, to which a clean, dry, tared flask has been attached). ‘The “ester” is extracted for about four hours by low-boiling petroleum-ether into the tared flask. The petroleum-ether is distilled off, and the flask is transferred to a steam oven, where the last traces of petroleum-ether are removed by addition of a little anhydrous ether and reheating until the weight becomes constant. The percentage of “ ester”’ (petroleum-ether extract) is calculated; and by subtracting from 100, the percentage by weight of the free acids in the wax is obtained. From the latter number and the Acid Number of the wax, the Acid Number of the free acids is calculated. Assuming that the only free acids in the wax are crude cerotic and commercial stearic and palmitic acids, the percentage by weight of the crude cerotic acid can be calculated on the assumption that its mean Acid Number for unbleached beeswax by the method above described is 146, while the corresponding number for commercial stearic acid is 206. By multiplying the percentage by weight of the crude cerotic acid by 7:5, the percentage of unbleached beeswax in the sample will be found. In many cases it will be desirable to find the percentage by weight of the unsaponifiable matter in the wax, the number of cubic centimetres of Ryan—The Analysis of Beeswax. 213 hydrogen at normal temperature and pressure liberated by heating one gramme of the wax with potash-lime to 250°C., the percentage of hydro- carbons in it, and the mean molecular weight of the combined acids, in addition to the Acid, Ester, and Iodine Numbers, and the percentage of glycerine obtainable from it. The following results were got with samples of beeswax' from widely different portions of the Harth’s surface, and show that genuine beeswax is fairly constant in composition :— Wax FROM— Treland. Chili. Sierra Leone. | Madagascar. Mean. Specific Gravity (15°C.), 0:963 0:963 0:966 0:967 0:965 Melting Point (°C.), 9 62-63 63-64 63-64 63-64 62-64 Acid Number, : : 19:06 19°6 19°6 19°6 19°465 Ester Number, . : 75°04 73°52 75°04 75:7 74:825 Saponification Number,.| 94:10 93:12 | 9464 95:3 94:29 Iodine Number, . : 11:2 12 8-2 9-2 9:95 ; eth (i Acids, . 13°62 13°32 13°00 13°34 13°32 & £ 2 ee Meee 66:38 86°68 87-00 86°66 | 86-68 ° & \ Unsaponifiable, | 51:8 55°5 | 58-7 53°06 53°51 Having found, in conjunction with Mr. ‘T. Dillon (see previous communication), that purified montanic acid, which occurs in the free state in Montana (Montan) wax, has a molecular weight higher than that of cerotic acid, it was obvious that by addition of stearic acid, whose molecular weight is lower than that of cerotic acid, to Montana wax the mean molecular weight of the free acids could be made equal to that of cerotic acid, and that by addition of the required quantities of spermaceti, carnauba wax, and ceresine, a composition wax could be prepared having Acid, Ester, Ratio, and Iodine Numbers identical with those of beeswax, and, moreover, containing a mixture of free acids having a mean molecular weight equal to that of cerotic acid. Such a mixture was made by melting together 2 grammes of Montana wax, 0°3 gramme of stearic acid (pure), 5 grammes of spermaceti, 1°5 1The author is indebted to Mr. J. G. Rathborne for seyeral of the samples of waxes described in this and the previous communication. 214 Scientific Proceedings, Royal Dublin Society. gramme of carnauba wax, and 1:2 gramme of ceresine. On examination of the composition wax thus formed, by the methods mentioned above, I obtained results very similar to those got from genuine beeswax :— Composition Wax. Beeswax. Specific Gravity, . ; : 0-960 0:96-0:97 Melting-Point, C., . : : 46-67°C. 62-65°C. Acid Number, : j : 21:28 18-22 Ester Number, 6 : : 177 70-78 Saponification Number, . : 98:98 90-98 Iodine Number, . : ; 6:66 2-12 Free Acids, Percentage, . : 14-21 13-14 Kster, Percentage, . : f 85:79 86-87 Unsaponifiable, Percentage, . 56:09 51-56 The composition wax having an Acid Number and a percentage of free acids approximately the same as beeswax, it would, if the variation in melting-points were neglected, be returned as pure beeswax if the analyst followed the method proposed by Hehner, or that proposed here for the partial examination of complex wax mixtures. The above wax composition contains 12 per cent. of ceresine, and 9:4 per cent. of so-called hydrocarbons from Montana wax, making a total of 21:4 per cent. of ‘ hydrocarbons,” whereas beeswax contains only 11-14 per cent. Again, by hydrolysing the “ester” obtained in the gravimetric estima- tion of the free acids, and separating the insoluble acids from the soap formed, it will be found that the molecular weight of the combined insoluble acids is greater than the mean molecular weight of the combined acids in myricin. Assuming that the only acids present in the esters are palmitic and cerotic acids, the percentage by weight of the cerotic acid (combined) can be found; and from this the percentage of carnauba wax, or of Chinese wax, can be calculated. In the results given above it will be noticed that the mean number of milligrammes of potash neutralized by 1 gramme of the crude cerotic acid contained in unbleached beeswax is 146, whereas 1 gramme of pure cerotic acid neutralizes only 141°6 milligrammes of potash. Crude cerotic acid, therefore, contains in addition to cerotic and melissic acids, other acids of — lower molecular weight than cerotic. Again, from the percentage by weight of the ester in the different samples of beeswax examined, and the Ester Numbers of the waxes, it follows that the mean molecular weight of the ester, assuming that the Ryan—The Analysis of Beeswaz. 915 petroleum-ether extract consists solely of ester, is 650, whereas that of myricin (myricyl palmitate) is 676. If myricin has a molecular weight—676—agreeing with formula C,;H,, COOC;,H.4, cerotic acid a molecular weight 396, and if the average Acid and Ester Numbers for beeswax be Acid Number 20, Ester Number 75, then 100 grammes of beeswax should contain 14:11 per cent. of cerotic acid, and 90°36 per cent. of myricin, making a total of 104:47 per cent.; so that it is obvious, as the gravimetric estimation indicated, that the molecular weight of myricin (beeswax ester) must be less than 676. If in addition beeswax contains about 13 per cent. of hydrocarbons, and 13:32 per cent. of free acids, then the percentage of ester in the wax will not be more than 73°68 ; and as the mean Ester Number is 74°825, the molecular weight of the ester will be only 554. If the extract obtained by Buisine’s method consists of hydrocarbons, and the combined acid of beeswax be palmitic, the molecular weight of “myricyl” alcohol is only 316, corresponding to the formula CH,,OH. On the other hand, the oxidation of myricyl alcohol to melissic acid shows that the formula for the alcohol is C;,.H,.0 (molecular weight 438). Further experiments will be made with Buisine’s apparatus, to determine the cause of the apparent disagreement, which is probably due to the presence of alcohols (other than primary) which are not converted into salts of the corresponding acids when heated to 250°C. with potash-lime. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 22. AUGUST, 1909. ON THE VALUE OF BENZIDINE FOR THE DETECTION OF MINUTE TRACES OF BLOOD. BY i. J. McWHENEY, M.A., M.D. (R.U.L), D213 RCP. PROFIFSSOR OF PATHOLOGY AND BACTERIOLOGY IN THE CATHOLIC UNIVERSITY SCHOOL OF MEDICINE, DUBLIN. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price Sixpence. abla ST : -CoHs:—O; Hu 0 , which is the mother-substance of cerulignon (2), C;H.,,(OCH,),—O,;H.(OCH:)2 | —O———_—_-O ceerulignon, a violet-blue powder, which is obtained in the purification of pure wood- vinegar, the relation being somewhat analogous to that which holds between aniline and its oxidation-product, quinone. oe XO) 8he INTE CHaly. 18 [ae acme li, .C,H,. H, (ase aniline. : quinone, McWrrnry— Value of Benzidine for Minute Traces of Blood. 217 Another allied substance is guaiacol C,H, an (ortho), which is the active body in the well-known test with guaiacum resin and “ozonised ether,” “oxydized turpentine,” or peroxide of hydrogen solution, associated with the names of Van Deen and Schoenbein.' Oxidation of guaiacol produces a blue base, probably of the nature of a di-pheno-quinone (cf. the formation of ceorulignon, supra). The colour formed in the guaiacum test differs from the benzidine colour- base in the same way as a phenol differs from an amine (aniline) or an aurine dye from a rosaniline dye. The first writers to draw attention to the behaviour of this class of organic compound with blood were Oskar and Rudolf Adler (3), who found that oxidation of many substances, such as aniline, diphenyl-amine, p. tolui- dine, xylidine, 0.m. and p. phenylene-diamine, phenol, 0. m. and p. cresol, thymol, pyrocatechin, guaiacol, resorcin, hydroquinone, orcine, pyrogallol, phloroglucinol, salicylic, protocatechuic and gallie acids, bensidine, tolidine, aand 3 naphthol, and naphthylamine, by H,0,, in presence of blood, took place with formation of brilliantly coloured products. ‘They found that the most satisfactory results were obtained from leuco-malachite-green and benzidine. It has been found that ferrous and cuprous salts, sulphocyanides and oxidative ferments gave similar results. My attention was first drawn to benzidine by a paper (4) in the Deutsche med. Wochenschrift, 1906, by Schlesinger and Holst, of Prof. Strauss’s Policlinic in Berlin, who found benzidine a most delicate means of demon- strating minute traces of blood in fecal matter—much more delicate than either of the tests hitherto employed for that purpose in clinical medicine— aloin and guaiacum. It occurred to me that it might prove a valuable means of discriminating blood-stains on clothing, &c.; and having procured asample from Merck, and tried numerous experiments with it, I shall now proceed to describe the technique of its use, and the results which I have obtained. 1. echnique-—An approximately saturated solution of benzidine is made with glacial acetic acid by putting a knife-point full of the crystals into a test-tube, followed by 3 or 4 cc. of glacial acetic acid, and shaking till solution is effected. Of the clear brownish liquid so produced about } cc. is transferred to a fresh test-tube, and about 2 cc. H,O2 solution (the 20 vol. solution diluted with an equal part of water) added. ‘I'he whitish opaque mixture is the reagent; and to it is added the material which it is desired to examine for blood, even minute traces of which produce a fine blue colour. The mixture of benzidine and hydrogen peroxide has a strong tendency to 1 According to Doebner and Liicker, the blueing of guaiacum tincture is due to the oxidation of guaiaconic acid C2oH40;. [Archiv f. Pharmak., ccvrxiv., 1896, pp. 234, 590.] 218 Scientific Proceedings, Royal Dublin Society. turn blue spontaneously, the change usually coming on after about a minute. In judging the result it is therefore advisable to have regard only to decided tinges of blue coming on at once on addition of the suspected fluid or substance, and, in case of doubt, to make up double the above specified amount of reagent, and dividing it between two test-tubes, observe the reaction in one, keeping the other as control. The test-tubes employed should be absolutely clean, as the minutest traces of oxidizing agents (bleaching-lime, plumbic oxide, permanganate of potash) suffice to bring about the colour change. Later on I shall have more to say regarding the fallacies of the test and the mode of guarding against them. 2. Delicacy.—In a series of experiments, with accurately made-up decimal dilutions of sheep’s blood, I found that a specimen of benzidine from Kahlbaum gave a distinct bluish-green reaction in a few seconds, with so great a dilution as 1 in 500,000, whilst a control, put up with tap-water instead of the blood-solution, did not begin to turn blue till two minutes after admixture. At the end of five minutes there was hardly any difference between the two tubes, both being a decided blue. I find, however, as the result of a number of tests made with different preparations of benzidine, and different test-tubes, that one cannot always be sure of avoiding pseudo- reactions at so great a dilution, and I am in agreement with Schumm and Westphal (5) that 1 in 200,000 is the greatest dilution, the recognition of which by this test can be safely depended on. By comparative work with freshly prepared guaiacum tincture and hydrogen peroxide, I found, using the same solutions of blood, that the guaiacum ceased to react at 1 in 20,000. Moreover, I find that the most distinct reaction with extremely dilute blood- solutions can be had by superposition of the suspected solution on the surface of the reagent. The blue zone which forms on the line of contact is quite unmistakable. 3. Use in medico-legal investigations.—But the real value of benzidine in medico-legal work is its power of reacting, not with very dilute solutions of blood, but with minute traces of blood in particulate form; and this is a point on which I desire to lay special emphasis, as I have not seen it referred to in the literature. The material submitted in medico-legal work consists practically always of suspected stains that have long since become quite dry. The matter of which the stain is composed can usually be obtained in particulate form by a process of scraping, and if the tine dust so produced be allowed to fall on the surface of benzidine sol. + H,O, freshly - mixed in a clean watch-glass, each particle, if composed of or containing blood, at once betrays its nature by turning a vivid blue, and, as it goes to solution, yielding a blue streak when the fluid is gently stirred. Should McWeunny— Value of Benzidine for Minute Traces of Blood. 219 it be found impracticable to obtain a scraping, the fabric can be teased in a drop of normal saline, the stained fibres removed to another slide, and treated between slide and coverglass with a drop of the reagent, when the presence of blood at once reveals itself by the brilliant blue coloration of the affected fibre. This is the most delicate test possible for minute quantities of dried blood, and groups of a dozen or fewer shrivelled and deformed red corpuscles can readily be picked out by its means. Should there be available a little more of the stained material, the positive benzidine test may be most readily corroborated by treating between slide and coyerglass with 32 per cent. solution caustic potash or soda. ‘his renders visible the outlines of the individual red cells, and causes them to separate from each other without producing distortion, so that they can be readily recognized and measured. Moreover, the strong alkali makes the little mass translucent, brings out its reddish colour, and transforms the blood-pigment into heemochromogen, which can be easily identified as such by means of the micro-spectroscope. Should all the suspicious matter go to solution during the initial microscopic examination, the best way to proceed is to take it up in a fine glass tube, and then allow some of the benzidine reagent to run up by capillarity, whereupon the blue colour, developing at the line of junction between the fluids, at once reveals the presence of blood. 4, Sources of evvor.—We have first of all to guard against pseudo- reactions due to the presence of oxydizing substances in the test-tubes and other glass-ware used. I need hardly say that test-tubes that have been used for urine-testing and have traces of copper solutions adherent to their walls are quite unsuitable. Even the use of perfectly new test-tubes does not always get rid of this source of error. I have occasionally met with new test-tubes in which the benzidine + acetic acid at once struck a most brilliant blue without addition of hydrogen peroxide—a phenomenon due, no doubt, to the presence of plumbic oxide in or on the glass. Of course such tubes must not be used. In general I think a few preliminary experi- ments with the batch of test-tubes about to be used are advisable, and in doubtful cases one-half of the quantity of reagent (acetic benzidine + H,O. kept over, without addition, for comparison. Coming now to the reactions given by benzidine with the substances most likely to occur in the examination of specimens for forensic purposes, I find that saliva does not give a typical reaction either when directly added to the reagent or dried on filter-paper. The same remark applies to nasa/ mucus and mucoid sputum, if free from blood. I have on several occasions tested spots of nasal mucus dried on a pocket-handkerchief without obtaining more than a dirty greenish-blue tinge remaining localized, and without that 220 Scientific Proceedings, Royal Dublin Society. tendency to diffuse, which is characteristic of the reaction as produced by blood. Seminal stains yielded a pale blue to the naked eye; but on teasing the fabric in a drop of normal saline, and testing the resultant fluid under the microscope, the spermatozoa failed to react. Pus strikes a deep blue colour, somewhat slowly. I am not sure to what extent this is due to the leucocytes or to the traces of blood which are generally present. Urine I have hitherto tested once only, and that was a normal specimen. It did not yield a blue reaction within the specified period. Mik | have tested on three occasions, and I find that bluish streaks and areas appear almost at once in the coagulum produced by the acetic acid present in the reagent. They showed no tendency, however, to diffuse, and never became anything like so blue as those due to blood. Boiling the milk did not seem to make any difference. Dried on filter-paper, and the stain tested by imbibition, milk yields a dirty green tinge quite unlike that of even very dilute blood-solutions similarly treated. Sweat I have not tested as such, but have frequently observed the effect of benzidine on extracts of under-garments and socks that had been soaked with perspiration. ‘The reaction was always negative. Fecal matter I have not yet tested ; but, if derived from an ordinary meat diet, it would, owing to traces of unaltered muscle-heemoglobin, undoubtedly give the reaction. It takes two days on a diet free from meat before the feeces are free from hemoglobin (Schlesinger and Holst, /oc. cit.). In addition to these body-fluids I have also tested from time to time most of the substances likely to occur as substrata to blood-stains. Metallic iron en masse, or coarsely divided (iron filings), does not yield a blue colour, but after a few seconds gives a diffuse bright brown coloration. Rust gives the same result. Such specimens of woollen, linen, and cotton as I have tested do not yield the blue reaction, nor do wood, plaster, stone, or leather. These are the substances most likely to be met with in medico-legal practice; and I cannot say that I have found any one of them to give a reaction which, with reasonable care, technique, and experience, could be mistaken for that of blood. I come now to a class of objects which, with benzidine, yield colour reactions of a most intense character, quite similar and indeed exceeding those given by blood—I refer to fresh vegetables and fruits. If on a freshly cut surface of raw potato, white turnip, tomato, or onion, some acetic solution of benzidine be dropped, and after a moment followed by a drop or two of hydrogen peroxide, a most brilliant blue colour is, within a few seconds, developed, beginning in the fibro-vascular bundles. ‘These appear distinctly McWrrnny— Value of Benzidine for Minute Traces of Blood. 221 mapped out as blue dots or streaks according as they are cut crosswise or in their length. The subcortical layer is in most cases stained, though more slowly, an intense blue. After a few moments the colour spreads to the whole of the cut surface and the pattern is obliterated. Subjoined is a list of the vegetables and fruits which I have so far tested, with the results obtained :— Potato: diffuse deep blue at once. White Turnip: beantifully distinct patterns of the fibro-vascular bundles picked out in deep blue on a white ground. Tomato: irregularly distributed f.-v. bundles clearly picked out, sub-cortical and pericarpal tissues deep blue. Onion: remarkably distinct pattern of concentric circles and dots corre- sponding with the unopened leaves and f.-v. bundles of the bulb. Carrot: very slight reaction. Pea: beautifully distinct mapping-out of the in-folded cotyledons and f.-v. bundles of the embryo. Radish: distinct pattern. Apple: reaction slow to come on; violet tinge. Banana: distinct though slow mapping out of the fibro-vascular tissues. Strawberry : Isolated f.-v. bundles picked out. Orange: very little change. Cherry : change also slight. This effect is evidently due to powerful oxydases present in the fibro- vascular bundles. The layer of test-fluid covering the cut surface is at first locally tinged, corresponding to the bundles, and then diffusely, whereupon the whole cut surface becomes uniformly stained. Such fresh vegetable matters do not occur in medico-legal practice, and are not, I think, liable to cause error. Still the question must be solved whether and to what extent the reaction would be yielded by fragments of such vegetable parenchyma when dried. The only experiment I have hitherto made was with a slice of potato dried at 37° in the incubator and tested next day. The scrapings struck a faint greenish blue with the benzidine reagent—quite unlike the colour given by blood. I have found that boiling the vegetables in water for a few minutes destroys their power of affecting the change of colour, owing, no doubt, to the destruction of the oxydase. Hven weak solutions of blood (1: 1000), however, yield the reaction after boiling for ten minutes quite as well as before. The proposal has been made by Hinhorn of New York (6) to employ 222 Scientifie Proceedings, Royal Dublin Society. benzidine-paper as a test for blood in clinical work. He uses strips of filter- paper that have been soaked in the acetic solution of benzidine and dried. The slip is moistened with the fluid to be tested, and a few drops of hydrogen peroxide dropped on. I have not as yet enough experience of this method to enable me to speak with certainty as to its merits. But such experiments as I have tried—about a dozen—with benzidine-paper have not been encouraging. I find that when H,0,, is dropped on to benzidine paper, without addition of any suspected fluid, greenish to bluish tinges make their appearance quite regularly as the paper dries, and give rise to doubt. Whatever be the value of the method for clinical purposes, I consider that for medico-legal work it is far inferior to the direct use of solution in ordinary or capillary tubes. Hinhorn records an observation which I thought it well to repeat. In testing the effect of various food-stuffs on benzidine he observed that porridge (Grwtze) boiled with water or milk at once yielded the blue reaction with benzidine-paper. JI have tried the effect of porridge made with “flake oatmeal” on the benzidine reagent in test-tubes, both with and without milk, and have obtained a practically negative result. Another method of using benzidine is in alcoholic solution. This is the way recommended by Ascarelli of the Institute for Legal Medicine at Rome (7). He mixes about 2 ¢.c. of the saturated solution in absolute alcohol and H,O,, and adds a few drops of glacial acetic acid. With this reagent he obtains reliable positive results with blood-solutions as weak as 1: 250,000. I can only say that my experience of alcoholic solutions of benzidine has not been satisfactory. In my hands they have proved capricious and occasionally insensitive. An objection urged against the test by Schumm (8) is that samples of the substance obtained from the most reliable sources vary considerably in their delicacy. Of fourteen samples obtained from Merck he found that three were sensitive up to 1: 200,000, whilst two did not reach beyond 1 : 50,000, and nine gave only a dirty red coloration with 1: 10,000. Of two pre- parations from Kahlbaum one was sensitive up to 1 : 2,000,000 (!), the other to 50,000. My own experience on this point is limited to three preparations, two from Merck and one from Kahlbaum. I found that one from the first- named source was very sensitive at first; but after standing in a corked bottle in the laboratory for over a year it had turned a faint pink, and was insensitive at 1: 10,000. The other Merck preparation was, when fresh, sensitive up to 1: 50,000, not beyond. The Kahlbaum product reacted up to 1 : 500,000 (controlled by addition of pure water to the other half of the test-solution). McWEENEY Value of Benzidine for Minute Traces of Blood. 228 I will now formulate the conc/usions at which I have arrived as the result of my three years’ experience of benzidine :— 1. In this substance we have the most delicate reagent for blood hitherto discovered. 2. On contact with benzidine in acetic acid solution + H.O,, dried blood in particulate form from stains, &e., at once assumes a deep blue colour which is quite distinctive. 3. None of the secretions and excretions hitherto tested, and none of the substances more frequently occurring in medico-legal practice, have been found to behave towards benzidine in the same way as blood. 4. Precisely similar reactions are given by many freshly eut vegetables and fruits; but these are eliminated by boiling, whereas the reaction given by blood-solutions is unimpaired by boiling for ten minutes. 5. When testing very dilute solutions of blood (weaker than 1 : 50,000), the test fluid should be divided and one-half kept as a control. No regard should be paid to colour-changes coming on after the lapse of one minute from the time of admixture. 6. Whilst emphasizing the fact that I have hitherto met with no substance reacting with benzidine in the same way as blood, I do not suggest (as does Ascarelli [/oc. cit. sub fin.], that positive results with benzidine should be looked upon as absolutely conclusive proof of the presence of blood. Negative results may, however, safely be taken as proving its absence. For the present, until our experience has matured, I am of opinion that positive results with benzidine require confirmation by other methods (e.g., microscopic demonstration of the corpuscles, heemin crystals, the spectroscope). 7. Benzidine is much superior to guaiacum as a preliminary test, and forms an important and most useful addition to the armamentarium of the medical jurist. I desire, in conclusion, to gratefully acknowledge the assistance I have received from Prof. H. Ryan, p.sc., F.r.u.1., on the chemical side of this paper. : [During and after the reading of the paper, the effect of the benzidine test on dilute solutions of blood, as well as on fresh vegetables and fruits, was demonstrated. | SCIENT. PROC, R.D.S., VOL. XII., NO. XXII, 20 224 Scientific Proceedings, Royal Dublin Society. Last or AUTHORITIES QUOTED. (1) J. Wotrr: Eine Farbenreaction zur Erkennung von Benzidin, &c. [Ann. Chem. anal. Appl., Bd. iv. (1899), p. 263. ] (2) Witisriprur und Kais: Ueber chinoide Derivate des Di-phenyls. [ Ber. d. deutschen chem. Gesellsch. Bd. xxxvii. (1904), p. 8761: Bad. xxxvill. (1905), p. 1232.] (3) O. und R. Apter: Ueber das Verhalten gewisser organischen Verbin- 8 g dungen gegentiber Blut, mit besonderer Beriicksichtigung des Nachweises von Blut. [Ztschr. f. physiol. Chemie, Bd. xli. (1904), p- 09. | (4) Scuiesincer und Horsrv: Vergleichende Untersuchungen uber den Nachweis von Minimalblutungen in den Feces, nebst einer neuen Modifikation der Benzidinprobe. [Deutsche med. Woch., 1906, No. 36.] (5) Scoumm und WusrpHaLt: Ueber den Nachweis von Blutfarbstoff mit Hilfe der Adlerschen Benzidinprobe. [Ztsch. f. physiol. Chem., Bd. xlvi. (1905), p. 510.] (6) Max Einuorn: Ueber eine neue Blutprobe. [Deutsche med. Woch., 1907, No. 27.] (7) AscareLir: Der Nachweis von Blutspuren mittels der Benzidinprobe in forensischer Beziehung. [Deutsche med. Woch., 1908, No. 53.] (8) O. Scoumm: Zur Kenntnis der Benzidinblutprobe. [/did., 1907, No. 42.] Not referred to in the above Communication. Arz: Ueb. d. Verw. v. Benzidine zum forensischen Blutnachweis. [Chem. Ztg. Bd. xxxi. (1907), p. 787.] THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. SEPTEMBER, 1909. Vol. XII. (W.S.), No. 28. ON THE FOSSIL HARE OF THE OSSIFEROUS FISSURES OF IGHTHAM, KENT, AND ON THE RECENT HARES OF THE LHPUS VARI ABILIS GROUP. BY MARTIN A. C. HINTON. [ COMMUNICATED BY R. F. SCHARFF, PH.D., B.SC. | (PLATE XV.) [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATEH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1909. Price One Shilling and Sixpence. apsonian last Nov 4 1909 \ i PUssH Ss pn u ‘ 4, XXIII. ON THE FOSSIL HARE OF THE OSSIFEROUS FISSURES OF IGHTHAM, KENT, AND ON THE RECENT HARES OF THE LEPUS VARIABILIS GROUP. By MARTIN A. C. HINTON. [COMMUNICATED BY R. F. SCHARFF, PH.D., B.SC. | (Prate XV.) [Read Junz 22, Ordered for Publication Jury 13. Published Szpremuer 8, 1909.] In 1894 Mr. E. T. Newton published an account of the numerous bones of the Hare which had been obtained by Mr. W. J. Lewis Abbott from the Pleistocene deposit filling the fissures in the Kentish Rag at Ightham, Kent ; and he referred these remains to Lepus europeus, Pallas.1 In a later paper Newton, with the advantage of further material illustrating the dentition, came to a different conclusion, and referred many of the fossil bones from Ightham, notwithstanding their robust proportions, to Z. variabilis, Pallas.* After the publication of Newton’s second paper, the work of collecting from the Ightham fissures was continued by Mr. Lewis Abbott, and by Messrs. Corner and Kennard. The former geologist obtained, besides a large number of additional Hare bones and skull fragments, a very fine skull with the lower jaw complete; whilst his successors were so fortunate as to receive a nearly perfect skull, both mandibular rami, and a complete set of the more important limb-bones, together with some ribs and vertebra, all belonging to one individual. This skeleton was found in the “ Wolf” fissure at Ightham at a depth of nearly eighty feet from the surface. My best thanks are due to the three gentlemen above-named for entrusting me with this magnificent material, and for permitting me to describe it here. I have also to thank Dr. C. I. Forsyth Major, F.x.s., Mr. Gerrit S. Miller, Jun., of the United States National Museum, Mr. Oldfield Thomas, F.R.s., and Professor Keith for much kind help. I should state at the outset that I prefer to follow the lead of Winge and Newton by retaining the name of L. variabilis, Pallas, for the Snow or Vary- 1 Newton, Quart. Journ. Geol. Soc., vol. 1., p. 198. 2 Tbid., vol. ly., p. 421. SCIENT. PROC. R.D.S., VOL. XII., NO, XXIII, 2p 226 Scientific Proceedings, Royal Dublin Society. ing Hare. The use of the name “ Z. timidus,’ Linné, for this species after its long if erroneous application to L. ewropeus, Pallas, can only serve to confuse matters. It would indeed be well if the systematists could be induced to realize that science is concerned with things rather than with their names, and that the ‘‘ balance of convenience” is as worthy of consideration in natural as it is in legal science. Fortunately I need not trouble to deal with all the previous discoveries of fossil remains of the Hare in the Pleistocene deposits of Britain ; because Forsyth Major, in preparing his catalogue of the fossil Rodentia in the British Museum, has already had occasion to examine them ; and he has stated that, as the result of his researches, all the Pleistocene Hare-remains from this country hitherto discovered, and determinable with certainty, are referable to LL. variabilis, Pallas, and that LZ. ewropeus, Pall., is probably to be regarded as a much more recent introduction.+ Tar Sxuru. For the purposes of comparison, the fine series of skulls in the British Museum and those in the Museum of the Royal College of Surgeons have been used. Besides the measurements recorded in the tables, many others were taken; but only those useful in connexion with the fossil material before me are here preserved. Many different characters have been cited by different authors’ as dis- tinguishing the skull of Z. variabilis from that of L. europeus; but after carefully considering them all, I have come to the conclusion that as regards the form, only those mentioned by Winge* and Lonnberg* are at all diagnostic. In ZL. variabilis the frontals behind the superciliary processes are both absolutely and relatively broader than in JZ. europeus. This feature appears to be very constant, for, in the large series of variabilis crania 1 Forsyru Mayor, Geol. Mag., N.S., dec. v., vol. cxliii., p. 10. 2 For references see Bibliography given by Herscurter, ‘‘ Die Tierreste im Kesslerloch bei Thaiingen.”” Neue Denkschr. d. allgem. Schweizer, Gesell. f. d. Gesamt Natur., Band xliii, p. 152, and in Hirzurmer, ‘ Die Hasenarten Huropas,’’? Jahreshefte des Vereins fir vaterl. Naturkunde in Wirtt., 1908, p. 418. 3 Winer, ‘‘ Grénlands Pattedyr,” Meddelelser om Grénland, xxi., pp. 857, 875-882, 1902; and “¢ Pattedyr '’ (Danmarks Fauna), 1908, pp. 56-62. 4Lénnnere, Proc. Zool. Soc., 1905, vol. 1., p. 280. Naruusrus has given excellent figures of two Hare skulls in his paper, ‘‘ Ueber die sogenannten Leporiden.’’ ‘he subject of fig. 1 of his plates is rightly referred to L. ewropeus [L. “‘ timidus’’]; but the subject of fig. 2, also referred to the ‘‘ Feldhase,’’ is in reality referable to L. variabilis, as is evident from the form and measure- ments (see Table III., p. 236). ° LONNBERG, op. cit., p. 281, Hinton— The Fossil Hare of the Ossiferous Fissures of Ightham. 207 examined by me, in only one instance, viz., a skull of ZL. var. hibernicus (B. M. 3. 1. 28. 1), is an approach towards the constriction of the temporal region of LZ. ewropeus seen. A still more important character is that afforded by the anterior part of the zygoma. The malar is frequently deeper in ZL. variabilis than in L. europeus ; and the groove on its outer surface lodging the deep part of the masseter lateralis extends further forwards in the former species than it does in the latter. If one measures the least distance between the anterior edge of this groove and the front vertical border of the zygoma, one finds that it is considerably less than the vertical height of the zygoma measured in the same region in LZ. variabilis, whereas the converse condition is found in L. europeus.! Hilzheimer* has stated this to be inconstant ; but he has, in my opinion, confused skulls of £. variabilis with those of L. ewropeus; at any rate, in the material before me the distinction is quite clear. Winge® and Lonnberg! have given some further characters relating to the length, breadth, and form of the nasals in the two species, and these appear to me to be borne out by the material. Another important character, first pointed out by Woldrich,’ and more recently by Winge,° is that, while in Z. europeus the anterior upper incisor is quite confined to the premaxillary, in Z. variabilis this tooth extends its alveolus backwards, so as to impinge upon or just enter the maxillary bone; and its termination is marked by a palatal swelling which is absent in ZL. ewropeus. After the ‘ables I. and II. had been drawn up, I received from Dr. Hilzheimer a copy of his recent paper. ‘The “ Basilarlange”’ of his table of measurements’ is a different measurement from the “basal length ” as here given. In my tables the measurement expresses the distance between the anterior margin of the foramen magnum and the most anterior point of the premaxillary. Certain of Hilzheimer’s measurements do correspond with mine; and in order to compare them, the measurements in question were reduced to values of the “extreme length” taken as 100. The results of this comparison are summarized in Table III.; and attention is here called to this summary, as it serves to bring out certain differences of proportion which characterize the skulls of the two species with great clearness. It would appear from the table that the zygomatic breadth and least frontal 1 LONNBERG, op. cit., p. 280. 2 HILZHEIMER, op. cit., p. 406. 3 Wines, ‘‘ Grénlands Pattedyr,”’ p. 858 ; ‘‘ Pattedyr,” pp. 58, 61. 4 LONNBERG, op. cit., p. 280. ® Woxpricu, Sitzb. d. k. akad. Wien. math.-nat. Cl., xxxiv., Bd. 1, Abt., p. 220. 5 Winee, ‘‘ Gronlands Pattedyr,” p. 358; ‘‘ Pattedyr,” pp. 58, 61. 7 HILZHEIMER, op. cit., pp. 416, 417. 2P2 228 Scientific Proceedings, Royal Dublin Society. width are almost constantly, and the width of the brain-case very frequently, greater proportionally in Z. variabilis typicus, as compared with the. corre- sponding measurements in ZL. ewropeus, and that the diasteme is usually proportionally shorter, and the molar series proportionally longer, in the former species than in the latter. Turning now to the fossil skulls from Ightham, that in the collection of Messrs. Corner and Kennard is associated, as already stated, with the greater part of the skeleton. Except for the fact that the nasals are wanting, it is perfect, and has all the teeth in place. The general form, the little constricted temporal region (Pl. XV., figs. 1-4), the form of the anterior part of the zygoma (Pl. XV., fig. 4), the measurements recorded in the Tables 1-3, as well as the characters yielded by the dentition and lower jaw, to be discussed later, all agree in proving this skull to be referable to a form of L. variabilis. The skull in the collection of Mr. Lewis Abbott is a little larger than that just described ; and it comes from a somewhat older animal, the sutures being more tightly closed. ‘The zygomatic arches are broken off just behind their maxillary roots; and the upper rim of the right orbit is broken away. The face has been crushed in, and not very skilfully restored, which may account for some slight differences in the proportional measurements. In form this skull agrees perfectly with the fossil already described; and it is referable to a form of Z. variabilis. The nasals, which are here fortunately preserved, enable us to complete the comparison. Their greatest length is about equal to the length of the sagittal suture of the frontals, and their upper surfaces are sharply bent upon the planes of the outer surfaces (Pl. XV., fig. 5). The foremost upper incisors distinctly extend backwards into the maxillaries. Sanford long ago described a fossil skull from one of the Somerset caves; referring it to Z. diluvianus, Pictet.2 He pointed out that the zygoma and superciliary processes were much as in the Polar Hare (ZL. var. arcticus) ; and it is evident from his figure that the skull in question is to be referred to the variabilis group. A comparison of the Ightham skulls with Sanford’s figure shows that all three skulls are to be referred to one and the same form, although the measurements recorded by Sanford, when reduced, are somewhat different from those of the Ightham specimens—differences which are, perhaps, best accounted for by the differences in the methods of measurement employed. The large size of these fossil skulls is equalled only by that of the skulls of the sub-species L. var. grenlandicus, L. var. ainu, and L. var. tschuktschorum. 1SanrorpD, Quart. Jowrn. Geol. Soc., vol. xxvi, p. 126, Pl. viii., fig. 5. * Prorur, Lraité Elementaire de Paléontologie, 1844, t. i., p. 207- Hinron—The Fossil Hare of the Ossiferous Fissures of Ightham. 229 The skull of Z. var. grenlandicus! is at once distinguished by the form of the rostrum and the curiously straightened upper incisors, which project far in front of the nasals. The incisors of the Ightham skulls, and those from Yuzlawitz (Pleistocene), according to Woldrich,*? when compared with the Alpine, Scotch, and Irish forms of Z. variabilis, are seen to be straighter ; but this straightening is very much less than is seen in Z. var. grenlandicus. L. var. ainu® has the brain-cage proportionally very narrow; and the extreme length, compared with the basal length, is very short. LZ. va. tschuktschorum! is characterized by its great zygomatic breadth. In his description of the Somerset skull, Sanford says’ :—“The frontals differ in form from those of every hare with which we have compared them, with the exception of the much smaller form, Z. altaicus, of the British Museum Catalogue; but in the skulls of this animal the post-orbital processes are much less developed, and do not extend so far back as the parietals, with which these processes coalesce in the fossil.” In the smaller of the two skulls, labelled Z. “ adéaicus” in the British Museum (Zool. Dept., No. 50. 5. 28. 3), which is the one, I think, that Sanford had in mind, the constricted temporal part of the frontals is short antero-posteriorly, so that the brain-case appears rather abruptly truncated in front; and the fossil skulls from Ightham are similar in this respect. But one finds the same frontal form together with the yery massive superciliary processes (which characterize not only the Somerset but the Ightham skulls) frequently in the skulls of LZ. variabilis scoticus, and occasionally in those of LZ. var. hibernicus. When viewed from the front, the superciliary processes of the Ightham skulls rise more gently from the frontal surface, and the median convexity of the frontal is less bold than in LZ. var. scoticus, so that the superior transverse contour appears flatter in the fossil skulls; and in this respect they agree exactly with L. var. hibernicus (Pl. XY., fig. 3). | As regards size, the skulls of Z. var. scoticus are smaller than those of L. var. hibernicus ; and these again are smaller than the fossil skulls. If the 1 Lyon, ‘‘ Smithsonian Miscellaneous Collections,’’ vol. xlv., pls. 80 and 81, fig. 7. ‘This is the name by which Rhoads proposes to distinguish the Greenland Hares from those inhabiting the western coast of Davis Straits. Winge, however, says that there is no specific distinction, and that Rhoads has only seen the extreme form of the Greenland Hare (Grénlands Patiedyr, p. 375). Judging from the material before me, both the highly specialized form which Rhoads calls L. grwnlandicus, and the more normal form, for which he uses Ross’s name of L. arcticus (= L. glacialis Leach), and which he only knows from the regions west of Davis Straits, inhabit Greenland. I agree with Winge that there is no specisic distinction between them; but as the two forms are very different in skull, I think it better, at all events for the present, to keep them apart as sub-species of L. variabilis. 2 Woupricu, Sitzb. d. k. Akad. Wien. math.-nat. Cl., ixxxii. Bd., ii. Abt., p. 15. Winge, ‘¢Gronlands Pattedyr,’’ pp. 358 and 376-378. 3 Barrett-Hamitton, Proc. Zool. Soc., 1900, p. 90. 4 Lyon, Joe. cit., fig. 3. > SanForD, Quart. Jowrn. Geol. Soe., vol. xxvi., p. 127. 230 Scientific Proceedings, Royal Dublin Society. proportional measurements (Table II.) be compared, it will be seen that the Ightham skulls most closely correspond with those of the Irish Hare. The skulls here referred to, LZ. var. arcticus, Ross (= L. glacialis, Leach), principally differ in their somewhat greater zygomatic breadth. LZ. var. varronis' also comes near to the fossils as regards the temporal part of the frontals and the form of the brain-case; but it is distinguished by its much smaller size and by the much gentler way in which the sides of the post-palatine vacuity fall away from the inner alveolar border. As what I have to say about Hilzheimen’s? classification of the European Hares depends principally upon the skulls, it will be convenient to deal with it here before passing on to the consideration of the lower jaw. Hilzheimer distinguishes three species of Hare as inhabiting the European mainland and Britain, viz., LZ. variabilis, Pall (= timidus, Linn.), L. ewropeus, Pall, and LL. medius, Nilsson. Blasius long ago gave a good account of ZL. medius, Nilss.,? and of L. aquilonius, Blas., and L. caspicus, Ehrenberg (the two latter forms being regarded by Hilzheimer as subspecies of Z. medius, Nilss.) ; and he showed conclusively that these three forms were nearly allied to each other, that cranially they were, according to the then prevalent idea of the nature of species, indistinguishable from Z. ewropeus, and that in fact they were merely climatic pelage phases characteristic of the northern and eastern parts of the range of LZ. ewropeus. That Blasius’ opinion as to the cranial identity of “TL. medius” with L. europeus is correct is proved by a reference to the analysis of Hilzheimer’s measurements of the skulls from Sarepta,‘ which he regards as probably referable to LZ. caspicus, Ehrenberg, given in Table III. of the present paper. Hxcept that the molar series is a trifle shorter, there is nothing to distinguish ZL. “medius” caspicus from L. europeus typicus. Further, it may be mentioned here that the two skulls from Sarepta mentioned by Hilzheimer’ as possibly referable to “ L. timidus” [= variabilis | (““? hyemalis”), are unquestionably to my mind referable to a form of LL. europeus, and probably are identical with LZ. e. caspicus, the proportional values for these specimens being, in the order of ‘lable III. :—100 (100), 48:1 (45°5), 83°4 (82°38), 18°25 (14:15), 31-4 (81-4), and 17-2 (17:2). Hilzheimer does not stop, however, at recognizing L. ‘ medius” and its two allies as forming a distinct species. He extends Z. “medius” to include 1 Lyon, loc. cit., fig. 8. * Hitzurmer, Jahreshefte des Vereins fiir vaterl. Naturk. in Wiirtt., 1908, p. 384. 3 Buastus, ‘‘ Sawgethiere Deutschlands,’’ 1857, p. 414. 4 HiLzHEIMER, op. cit., p. 416. > HILZzHEIMER, op. cit., pp. 409-417. Hinton—The Fossil Hare of the Ossiferous Fissures of Ightham. 231 what in-no view of the value of that “species” it can include, viz., forms really belonging to the variabilis group. The forms which he thus regards as sub-species of Z. “ medius” are L. var. varronis, Miller, L. var. breviauritus, Hilz., and LZ. var. scoticus, Hilz.1| With regard to the first two I think a glance at Table III. will show that they are referable to the variabilis group. The case of Z. var, scoticus demands rather more notice here, because it is of importance to our present object. Hilzheimer bases his description on specimens from the north of Scotland collected in 1855, and the account of the pelage agrees very well with the summer and winter coats of the Scotch mountain hare. In the “ Anmerkung” he proceeds :— Barrett-Hamilton stellt diesen Hasen zu LZ. timidus [i.e., variabilis], da er keinen Unterschied finden kann. Bei beiden schottischen Exemplaren der Strassburger Sammlung, die das Sommerkleid haben, ist der Schwanz oberseits schwarz, wahrend er bei den drei entsprechenden Stockholmer Stiicken weiss ist. Auch werden die Ohren der Schotten im Winter nicht ganz weiss.”? All turns here upon what is to be understood as a ‘‘ black” upper side to the tail. The tail of the summer coat of Z. var. scoticus is certainly darker above than below, but it is not black, as is the tail of Z. ewropeus or “ L. medius,” and in no way can it be ealled bicoloured. Further, it is curious to note that Hilzheimer leaves L. var. hibernicus, which is darker still, and which becomes still less white in winter, as a subspecies of L. variabilis (= timidus, Linn.).3 It may be, however, that Hilzheimer has described the “black” tail from a Scotch skin of L. ewropeus, because a little later he states erroneously that L. europeus does not occur in Scotland.* Be this as it may, there can be no doubt that the Scotch mountain Hare is referable to L. variabilis, as is proved by the skull, although it has certain features peculiar to itself which are deserving of subspecific recognition. Hilzheimer’s name of “ scoticus” is therefore retained here for the subspecies. 1 HILZHEIME Rk, op. cit., pp. 389-392. 3 Op. cit., p. 387. * HILZHEIMER, op, cit., p. 389. ~ £ Op. cit., p. 398. [ TaBLes. 232 Scientific Proceedings, Royal Dublin Society. TapLE I.—MEASUREMENTS OF 3 ae ; IGHTHAM 5 L. var. hibernicus. =) Fossiu. a British Museum. 3 = 3 a oO ey Sn — . oO s A ea he _ | ga) 3 S| 3 g a 2 ef 2 « PS + | n I i? eo. EME coleelee ls lac | == eS) “4 +m . 8 | 3 | 8 |as/ea| S |eais 1. Basallength, .. 83 85°8| 87 | 79:1 | 76°6 | 75:9 | 74-4 | 72:0 2. Extreme length, 1032 | 1038 | 110 98:5 | 95:0 | 94:0 | 92:0 | 89:3 3. Zygomatic breadth, .. 50°6| — | 50:8) 48:2 | 45:8 | 45:7 | 45:1 | 45-4 | 4. Width of braincase midway between ear and zygoma, 301) — — | 33:4 | 32:4 | 32-1 | 33:0 | 31-9} 5. Distance between anterior border of foramen magnum 32°2 | — — | 29-8 | 99-5 | 28-8 | 29-0 | 27-9 | and plane of posterior edges of m. 8. | 6. Least width of frontals behind superciliary processes, | 16°5| 17-8] — | 13°9 | 16-5 15-2 | 16°6 | 17°8 | . 7, Extreme width of frontals across superciliary pro- | 87:5] 41:6] 33°3| 34 32°6 | 33°38 | — | 3d1_ cesses. Ca Ca | | 8. Width of nasals at point where frontal, nasal, and 20:3] 22:5| 22°8| 21:0 | 20-8 | 18-6 | 18-5 | 19:8 | premaxillary come in contact. 9, Vertical diameter of malar, 10:0} — 10:0) 7:8 9-7 8:7 8:0 79 10. Length of Diasteme, .. 29-6 | 28:2] 32:3] 28:0 | 26-0 | 27:0 | 25-0 | 247 11. Alveolar length of molar series, 90 ae 19-1| 19-9| 19:5] 18-6 | 18-4-| 18:1 | 17-5 | 17:6 | 12. Width between inner edges of m. 8, .. 17:4] 18°3)) — | 16:0 || 116-4 | 15:9) | 15:6 Milage § | | 13. Width between inner edges of p. 3, .. 14:7] 14:8) — | 18-4 | 12-0 | 13:3 | 12:8 | 12a | 14. Width between outer edges of alveolar walls between 99:0| 29:5) — | 27-5 | 28-4 | 26-9 | 27-0 | 26:0 m. land p. 1, | | | | | | Hinron—The Fossil Hare of the Ossiferous Fissures of Ightham. 283 SKULLS (ABSOLUTE VALUES). tp ; ‘ : & L. var. areticus. L. v. grenlandicus. | L. var. ainu.| 2 LL. europaeus. 0 TER SEDATE a British Museum. British Museum. Brit. Mus. 3 Roy. Coll. St $ Surg. = 56 : ig 5 ay . 9 Bo me re g 0 es Pease a 6 3 a iB Lincoln Bay.| , Si ee. | 83 | 3 | & be elie fies eee vila Bo | |S Sei a 8 | 4 a |e g © | 3+ =r elles +3 Sa ra SS : . So ae | ay |e | ss a || 4 Bee a S ia | co | SS a. | 22h 2c 8 c : “We 3 Sia Q Be os || 2a § |Sa| 8 | 4 les |s2] ea] € | 2S | Be ; | S32 3 | 5 5 | Bs || 3 a : J Ses il Se “2 | BEES | SE Se) Sa) eae on fee — =< a ~ 3 5 3 5 1a Rn Ses ie Ss q A 68°5 | 73°1 | 78-0 | 74:8 | 72-4 | 80-4 | 76°9 | 79:7 87°5| 86:6} 84:4) 82:0 | 82-5 87:0 || 78:0 | 76-5 86°8 | 90°4 | 89°6 | 91:4 | 90:0 | 98-8 | 96:4 | 96:0 | 109-6 | 107-2 | 105-0} 97-7 | 98-8 | 106-4 || 96-7 | 95-0 41°38 | 46-1 | 46-7 | 45:5 | 47-7 | 49-1 | 49:8 —_— 52:3] 52:7] 50-3] 48-0 | 48-2 56°9 || 47:2 | 45:9 | j 80°7 | 31-0 | 31°6 | 31:6 | 32:6 | 33-8 | 35:2 | 33:2 36°8| 35°7] 35:0} 30°7 | 32-6 37°0 || 32°2 | 33-2 | — 28°6 | 27:7 | 30:5 | 27-0 | 29-7 | 29-1 | 31:8 81:9} 30°7| 30°6] 38:4 | 32°3 34:7 — _— '16°6 | 15°8 | 16°3 | 15-9 | 17-8 | 17-6 | 16:2 | 15°6 16°9} 15°6} 19:4] 15:4 | 15-4 16°8 || 12°8 | 11:9 81:8 | 33°9 | 34:8 | 34:0 | 36:0 | 39°5 | 39:1 | 34°8 40°1] 41:9} 40:0} 35°6 | 35:0 — 35°7 | 33°1 Ca 17-7 | 19:0 | 18°5 | 19:7 | 19°8 | 21-2 | 20°5 | 21-4 22°8| 21:0] 21°8| 24:5 | 22°6 22:0 || 21:2 | 22-4 7:0 7:9 8:0 8-7 10-1 | 11°6 9°6 77 10:0 9-5] 10:0) 9:5 8:8 11:3 81 7:0 /24°6 | 25°65 | 2571 25:2 | 25:3 | 29-2 | 26-7 | 27-4 33:6] 31:1] 31-1} 27:1 | 26:0 31:0 || 28:5 | 26:8 UG 2) || 17:2 |) 17-2 17-0 | 16:7 | 18-7 | 18:5 | 18-7 19:4} 20:2} 19:1] 19:4 | 20-7 20°5 || 18°0 | 18°5 1 15:0 | 15:0 | 16-0 | 15-5 | 16:0 | 16-0 | 17:0 | 17-5 18°6] 16:9] 18:3} 16°8 | 16:8 18°8 || 16°6 | 16-1 el2*2 | 11:9 | 12°6 | 13:0 13-2 | 18:7 | 13°5 | 13:3 15:0} 14°38] 14:4] 12°6 | 12°6 151 || 13°8 | 13:0 W255 | 26-8 | 26:7 | 26:1 | 27:5 | 28:6 | 27-2 | 28-3 29°75) — 28:4 | 30°6 | 30°6 80:1 || 28-1 | 27:5 SCIENT. PROC. R.D.S,, VOL. XII., NO, XXIII, 2a 234 ; og IcHTHamM & L. var. hibernicus. 3 Fossiu. a=] British Museum. 2 wn iz ¥ Z ie 3 mn : 2 : 4 : E Ss | a : Se | 2 | 8 \ oes eee | ao Oo oe . OG sc Fie $3 a Be || SSI 5 wee xt to 3 ® Sy | ag © ee | So BO | EES ee 1S 1 ee | S@ a) < ea) Fo | eH 19 rH iy } 1. Basal length, .. bo a0 20 60 60 100 100 100 | 100 | 100 100 | 100 | 100 | 2. Extreme length, 124 121 | 126-5 | 124:5| 124 124 | 123-5] 124 3. Zygomatic breadth, 61:0} — 58:5 | 61:0} 59°9] 60°3| 60-6} 63-0 |, di 4. Width of braincase midway between ear and zygoma, | 42°3| — — 42:2} 42°3) 42°3] 44:4] 44-0 5. Distance between anterior border of foramen magnum | 38:8 | — — 37°7| 38:5} 38:0} 39:0] 88-7 || and plane of posterior edges of m. 3. | 6. Least width of frontals behind superciliary processes, 19:9} 20-7} — 17°6| 21:6] 20:0] 22:3) 924-7]) 7. Extreme width of frontals across superciliary pro- | 45:2} 48°5| 88°3| 48:0] 42:6] 43:9} — | 45-9 cesses, 8. Width of nasals at point where frontal, nasal, and | 24:5] 26:2) 26-2) 26:6] 27:2] 24:5] 24-8] 97:5 premaxillary come in contact. 9. Vertical diameter of malar, 12:05] — | 11-5] 9:88] 12-7] 11-6] 10-75) 11-0} 10. Length of diasteme, 35°7| 32:9] 87-1] 35:4] 34:0] 35-6] 33:6] 34-4 11. Alveolar length of malar series, 23°0 | 23:2]: 22-4) 28°5) 24:0) 23:9) 23:5) 24°51 | 12. Width between inner edges of m. 8, 21-0] 21:4) — | 20°25] 21-4} 21:0] 21:0) 21-0 13. Width between inner edges of p. 8, ee | 7=3)|) —— 16-9 5s ee dloe7illeledeon | eelerieod el Ges 14. Width between outer edges of alveolar walls between 34:9] 34-4) — 34:8] 37-1] 85:4! 36-3] 36:1] | m. 1 and p. 1. Scientific Proceedings, Royal Dublin Society. TABLE I].—MEASUREMENTS OF SKULLS Ye) ae) ea = Pe : q — oH Co) ca oo RX 8s O8ZE “ON = * 5 es | = S © S oq a = ae = BoE hte oes mea ee eye ae eee Meio tes eS S 2 9 Ss am > 2 a 1° ea O . . So sH (0) aN s Qa 6168 “ON Ss a Ss 4 | fe) 12 & = = cs oO SS S o os“ eal = Dal a = oO N a = 0 a Ss OTTO = SS : L°1 4°09 “Wf S © ~o GA) és S anqazjoyy “wnLoyasrynyas, = A 1D a a Ss : > > © Ra) = 3 a] = Ce) + ie) Go) io) 20 = > ee = A co o9 a a Si % Ss Re ~% N “P eceF1°9 ° a si ‘S a > 12 x 12 12 o S aS . S a eo} Cy oO wt Gr} cl . 8 Oplexyx[O FL = =i 16 for) (eo) i) i < an NS : oF oO = 3 = Ss 1D So 5 & S : = Se & 8 8 soe ao) a ‘GGL F 'FR S Ne} Ne} i Ce} 19 i oS = 5 q ‘ Ss 3 ee) = ~ : J > 1 for} ives} Ss UMOUyUN VOT “I FEC S Ss a & os 3 3 nn) Ht 19 S = S 5 O a = ~ oi I aa A & as Ge) os (fe) 1D Ss Kae] Nn N a oD iS rd 83 Ve) co el Oo se P) $ 2 eyse[y “2 Fee Ss % a, eS = z 2 5 = eo o = = = eS 4 R=} Sl S I Se oY i Go) a 10 a cx aa x nN Sy ry ise x isal a 5 p3 oO ea) So Oo Oo Oo a ay = & 2 See! puepuesry) 54 FES Ss a bl a o 4 és Bs 2 » 2 2 S 10 | 4 or) a =H > 19 Ny S| Ce ical ic) a = ma oD fe FQ 4 Sa es eee | ‘purlmeety Be FeG 2 oF & ie = Go) > roy ° = a a S| iS; a oS se = x or ~ co = 0 a & a A IT Ss aq S a S ise ae 2 bs f a a a ito} sH H a =H al oO a S S or jan [o) a re oO a = i Be i=) “ZG PRZL 86 So H o a = = a g 5 S Gr) > 12 H We} for) Gr) co = ‘snl Jig ‘e1A S os a Gr oO a = 1D a a g 5 ce ag ie Se C0 ee Ray ogee st) Ses oO =) = =) i= S ‘LPG "86 S = So H = = =A = = = il : “sn, Ig 691K Ss a oD nN fs a = oo oO Gel = : otA4[V — =I oO H rt eo) Ro) S tH . or) ~ S Yo) =] = a a a = ay ce Ss cS) © yy TErqooT “G8Ge “ON *suoasINg TIO) “IL 100 126°3 60-2 44°8 24°2, 46°5 26°8 10:2 35°8 23°7 21°9 17'8 37-2 Scientific Proceedings, Royal Dublin Society. 236 “SSUIMBIp eq} Jo UoTjoodsurt ue Aq yno oULOg UOTsN[oU0D v—snwdowna “T 0} JOU pUL sy2gneuwa “7 OF o[GBAOJoI AT[CaA St UOT}senb Ut ][NYs oy} yey} epnpouod 0} OW pea] SeINSy oseyy, “F-61 : 9-8G : 9-91 : GLE : 1G suorjiodord Surmoypjoy oy} svy ‘(soyepq sty Jo Z By) ,, weptioderyT uojuvuuesos orp reqey,, aoded stq ur ‘suisnyjey Aq pornSy [ys puooss oy, 001I—(S}UoMIeINsvoUE ;snisnyjyeN Wory peonped) Jromreqz [tH Aq painsveu s[[nys 1 8-LT G-9T 6-18 LL 6-71 8-11 L-G§ 1-36 G-8F GSP 00 : ‘snardspa snwdoimea “'T soured z Aq painsvout s[[nys g ¥-81 0-LT 8-3E 9-66 6-F1 6-11 LSE 1-6& LSP 9-FF 00T : : ‘snodh} snadosna sndaT 6:06 6-61 8°16 £-96 8-G1 9-GT 0-86 G1 G:6F 8-8F oor = a2 “heey sungurime “"T @-61 G-6G 8-GT 8-FE F-8G 00T re “UnLoyaszynNyas, SYQurewa “T 6:81 LLY 1-08 0-62 e.81 o¢-FL 9-€€ €-8& G-6F 8-LP 00r RS ‘* ‘snorpunquob syrgnriima “'T G-61 9-81 $-66 LLG 8-61 6-91 G.98 GFE &¢ L-6F 00T = ‘snaijodm syrquewwa “TT 9-81 8-L1 9-06 9-16 FLT 1-91 8-46 FE-PE 0-06 6-8F 00r oy vs ‘S1UOLIDA SYIQVIADA *'T 6-61 1-81 ¥-86 0-86 G-61 LT ¥-GE &-F§ G-3S 9:LF oor Ye i ‘snovjoas syrgpruma “'T ¢-81 1:86 0-91 ig 0-6F 00T si ‘noyuryy ‘snaubun syrquema T L-61 9-16 6-61 G-GS 8-0¢ 00T oe ne ‘sugosagn) siiqniiwa “T $-61 6-81 1-86 6-16 I-81 L-F1 6-C¢ 6-86 6F G:8F Oot or He ‘snaqusagry syrgurime “7 OCALA Aq painsvetr s]pnys 9 4-61 I-81 | ¢-08 83 | 1-81 GL-FIT | 9-88 1-58 1¢ 67 (COlen ts " “‘snaadhig syqvieva sunday “xP ‘ur | *Xe1T ‘uN | “XPT UL XPT “UITT “xe “Ur *SOLIOS Iv[OUr *9ua}SeIp *s[eqUOL osvourelq “yypraaq “Wy Sue] JO WSUS] Ie[OOATY jo W)8ueT JO Y}PIA ysvary JO PIA oryemosiz Eluichabd | ‘OOT = HIONG]T ANGULXY],, OL SNOILOAGAY— TI] FIVE Hinron—The Fossil Hare of the Ossiferous Fissures of Ightham. 237 Lower Jaw. The lower jaws of both the fossil skulls described above from the Ightham Fissure are preserved ; and they yield important evidence confirming the reference of these remains to the JZ. variabilis group. (Pl. XV., figs. 6 and 7.) Winge has pointed out that the molar teeth are somewhat more hypsodont in ZL. variabilis than in L. europeus; and a reference to the proportional measurements recorded in Table IV. will show that correlated with this greater hypsodonty there is an increase of all dimensions of height in the lower jaw of the former group as compared with that of the latter. The measurement ‘‘5”, which expresses the distance between the hinder edge of the alveolus of the last molar and the hindermost point of the condylar process, appears to be of some importance. With the exception of the highly modified LZ. var. greniandicus, this dimension appears to be constantly greater in L. variabilis than in L. ewropeus, and its value increases in young indi- viduals. The jaws from Ightham, which are adult, behave in this respect as the recent young jaws. Again excepting L. var. greniandicus, the diasteme appears constantly if minutely shorter in L. variabilis than in L. ewropeus. With regard to form, Woldrich’ has stated that in Z. variabilis the anterior margin of the ascending ramus rises somewhat more steeply from the alveolar border than in Z. ewropeus; and Hilzheimer® has added that in the latter species the inferior margin of the ramus is usually straighter, and not so convex as in the former. These characters appear to me to be fairly constant, but the most diagnostic are those furnished by the proportions and by the dentition described below. 1 Winas, “ Groénlands Pattedyr,”’ p. 358; and Danmarks Fauna, ‘‘ Pattedyr,” p. 58. 2 Woupricu, Sitzwngsb d. k. Akad. d. Wien. math. nat. Cl., \xxxiy. Bd., 1 Abth., p. 222. The measurements of the jaw from Zuzlawitz included in Table IV. are given in Woldrich’s first paper : ib., lxxxii., Bd., 11 Abth., p. 16: 3 HitzHEIMER, op. cit., pp. 408-9. [Taste IV. 238 (jo) SE GD Scientific Proceedings, Royal Dublin Society. TABLE L[V.—MEASUREMENTS OF 3 IGHTHAM s S L. var. hibernicus. Fossin. 5 ) British Museum. n | = ‘C = ole S) & | . o Rn a |e [8 |e , | ic) N —4 AQ rt oS o Rs) iS a8 as ey Bs a S 2 ce) A a [2 \ 8 | 20) S83 ee ae a 4 Dn: D S| 3 | 24 | 8 |a@5/S£] 8 (ae . Length from antero-internal point of incisive | 78°5| 80 | 70°2 | 75:0 | 74:6] 73-3) 72:1) 70-6 alveolus to hindermost point of condyle, Ditto to hindermost point of angle, .. o. | T4:0) 7374) — 70 70°0| 69°3| 67:3] 66°6 Ditto to anterior edge of alveolus of p. 2, 24°3| 28:8] 22-1 | 23 23°0 | 22°5| 22°5) 21-1 . Alveolar length of molar series, ae 20:2) 20°5] 19-1 | 20 21:0] 19°6| 19:6} 19:0 . Distance between hinder edge of alveolus of m. 3 and | 36:2) 37:0} — 34 82:6 | 32:4] 31-7) 31°9 hindermost point of condyle. i . Height of horizontal ramus beneath anterior edge | 14°5| 14:8| — 13:5 | 12:4) 136) 13:4] — of p. 2. Dittoatm.1, .. O10 dD 60 po iP uO. alse) uliier — 1670} 15:8) 15°53} — Ditto at m. 3, .. be ao ce co |) WSO) EE |) 17°5 | 17-0) 17-4) 17-4) 16°6 . Height from deepest part of ramus to highest point | 47°7| 51:0} 47°3 | 46:0 | 46°3|) 46:0] 44:9] 441 ot condyle (oblique), . Width of incisor, : 2-9 Sol || 3:0 3:0 2:9 3:0 2°6 Reduction to 1 = 100, .. 100 100 100 100 100 100 100 | 100 2= +e oo || Ged!) Gilkey 93°38 | 94:0] 94:4] 93°56) 94-4 3 31:0] 29°7] 31-4 | 380-7 | 380°8| 30-7] 381-2) 29:9], 4s 25°7| 25:6) 27°2 | 26°7 | 28:2) 26:7] 27-2) 26-9 a= 46-2) 46:2) — 45°3 | 43°7|) 44:1] 438°9} 46:2 B= S25) els sou ee 18:0 | 16°6} 18:5) 186) — eS 21:7] 23:1} 22:3 _ 21:5] 21:5) 21°35) — 8= 22:9))| 23:4) — 23°38 | 22:8] 23°7| 24:2] 23°6 9= 60:9| 63:8] 67°3 | 61:3 | 62:1] 62°7] 62-3] 62:5 10 . 0 3°7| 3:88) — 4:0 | 4:03} 3°95] 4:16] 3:68 9 = 100 ve 100 100 = 100 100 100 | 100 100 2= .e . 2 ee . | 155 | 144 _— 152 151 152 150 151 Hinton— The Fossil Hare of the Ossiferous Fissures of Ightham. 239 LOWER Jaws. g. d oO L , . A . 4. EUTOPRUS. 5 . S L. var. areticus. L. v. grenlandicus. | L. v. aint. 2 p = L.. var. scoticus. 2 ie Sie - 2 Roy. Coll. 2 | British Museum. British Museum. Brit. Mus. 2 § a 3 s urg. — oOo > Peel .3 Soni Z Bee wo ga | 3 3 3 "S E Lincoln Bay. | | sn || 3 28 = as = = Si 3 g a Ss g | 1¢-b |88-F | 82-4 | ~~ ne < = 4 €1-8 |GL-8 | F-S | 10-8 | 68-2 ar — |411-8 |G9¢-8 | €L-8 Oo}, om se Sy GL-G |GP-G | 86-9 |G6-G | 80-¢ = —| | 66-9 6-9 | TPG | 49-9 | a ae S8 GFF |OL-F | 9F-F | 88-6 | 89-€ | L9-b | L8-F | 68-7 Fb | GL-F | 69-F | °° me 6 =6 OOL | OOT | OOT OOL | OOT OOT | OOT oOT | OOT OE |) MOE || > = = “00T = 1 worjonpary, 9-8 |9-8 | 8-2 I-L | 9:8 = = Hi} ee Ta |] hs ‘GOIyB[NOT}AV [eISIP JO “WIP esIOASUUL, *9 9G |L-¢ |F-F |9-% |T-¢ — — 9:7) ‘rare \Kgncl lick¢ +" ToTje[NaTyIe [vISIp FO ‘ULVIp [BOI}AEA “G 1G Ve V3 NEB- | OB — — |8-6 |¢-01 | 0-01 | 6-6 ‘aonypnoize -xoid Jo ‘WvIp eSIOASUVL], “F 0-9 LG GG -G 8-¢ — —= |i Le) T-L 6-9 ¥-9 ‘+ ‘qorTyepnojre *xord Jo ‘MIP [BOWIE “e *(eaoqe) 601-66 ‘({EsSoF) 2IMUIZNZ |9-F |0-G |F-F |0-F |GF |6-F |F9 |FG [6-9 |F-G | FG YJCYS JO LoyaMVIp OsTOASUBI] TINUIIUT °% GOL ‘(yuooex) Smqz[eg | €-F01 | G-GOT | 1-86 | 8-GO1 | &-FIT | G-SOL IIL | 6-I11 | $-O0GT | 9-F1T | G-S11T] ~~ ae a ** qysu9] OMAR *T wa leale) elel sl] = 5 | 2 | & P| 2 | bs S g e, xe oo oo co a oo ¥ Sx, 20 SS) 8 SSS | sl @ 4 a | a | eel Be Ta k > © P| eel Plas SS Em| Se TOrpTOM Aq B Os ; Pe | Se “Tay poptodad syzqve0ma *T JO eS = 8 "oi || Iiavy Tipe JO syySue, omexXy oa eS Bi © a EB S iS é ‘Smg:jop:hoy| -Bmg “To9 “Loy a 5 prvuuey pur ToUI0\ snodowna -'T *9n91j098 "a “TT “TIssOWy NKVHIHOT *SATaVIT 2s SCIEN’. PROG. R.D.S., VOL. XII.) NO. XXII, vw Or S Scientific Proceedings, Royal Dublin Society. PELVIS. The only difference of proportion seen between the os innominatum of L. europeus and that of £. variabilis is that, in the former, the ilium is relatively broader than in the latter. Judging from the proportional measurements, given in the table, of the pelvis of the young LZ. variabilis (No. 3283 R.C.S.), the relative breadth of the ilium becomes greater as age advances. The outer surface of the ilium in JZ. variabilis and L. v. arcticus is more boldly sculptured than in LZ. ewropeus, the median ridge, for the gluteal muscles, and the fosse above and below it being higher and deeper respectively. The fossil examples from Ightham agree in these respects with L. variabilis, and are distinguished from the recent specimens only by their large size. 251 of Ighthamn. SSULES O @ Hinton—The Fossil Hare of the Ossiferous Fi 9-88 = 8-18 9-8& 0-48 0-48 8-98 0-0F G.Gg 8-18 ; : : 3 : Se P-GP = &-1F G-6€ 8-GF (So) 7 €-GP §-6F 6:€F L-GF ; s ES 2 Si G-OF a 0-97 o-LY G-9F 0-LF 1-97 LaF €-SF &-9F oe : at G-GE &-L6 G-86 €-96 0-86 1:96 8-86 0-86 F-1E 8-66 : =. : : =6 Oot 00 001 00T 00T 00T 00 00 0OT oor i - a ss 001 = 1 wononpayy ssisfydurds orqnd jo pue zo1eyue 08 — €-98 0-8 F-ge 6-18 8-98 4-68 e.12 $-68 pur unjnqejeoe Jo ursavur torradns usaajoq eourqsiqy “¢ TUNIS Taqny G-8e — 1-68 6-22 L-GF 0-68 PF 8-1F G.CF PF 0} MnNNqvjeov Jo ulsivm toT1a}sod woz eouvyst(T “F 9-1F = GFF 9-68 8-oF G-FF 0-87 G:F 9-LF G-8F ‘TANNGL}99v JO ULFIVUT LOLLOJUY MoI NTE Fo WSU -¢ 0-66 1X6 F-LG 1-16 GLE GPG 0€ L-L && 0-18 Sy a oy “* “(Quo4y Ut) TINIE Fo Yypeorg °Z 0-06 66 6-96 ¥-€8 ¥-86 0-46 G-F0I 66 1-01 I-01 = 4 ra j : ‘qjSu9] ewexXE “| Z| leis) Gis | et el ss : ~ Wo DAs te D D les} e o tae) 2.9 S = eo 8 a: = KO oo see. Ss ze) g bo 3 =. 8 bo iS) >= + Q O Sei ieee| ee | ee | © fe] io | * |v | ee 23 Ee E: = & = 2 IS = SI “NOAGVNINONNI SQ) : © 0g =i aS) ot 3 nm @ Bp ae Sy ee |g a 8 5 @E | -smg yoo -Loy = pieuuey pur rau10—y a Q *snovgoos “aT “TISSOY WVYHLHOT ‘SIATEG 282 252 Scientific Proceedings, Royal Dublin Society. Femur. The only constant difference between the femur of ZL. variabilis and that of LZ. ewropeus is that, in the former, the shaft is proportionally more slender than in the latter. Besides the femora of the fossil skeleton in the Corner and Kennard collection, I have many others before me from Ightham, including the one figured by Newton.1 Although they vary a good deal in form, they all show the slender shaft of Z. variabilis. The femur of ZL. v. hibernicus is a little stouter, relatively, than that of Z. v. scoticus, and, as the measurements show, the fossil femora agree in this with the former sub-species. ‘They are distinguished only by their larger size. 1 Newton, Quart. Journ. Geol. Soc., vol. 1., pl. xi., fig. 4. Hivron—The Fossil Hare of the Ossiferous Fissures of Ightham. FEMUR. 200 cd “TOTITISOFL a . g re [Issoy Wasuneyy, a *S271QUrlDe * a LA: IE S) ap oD co oo o 4 co oD . bE x = x é 5 oo = 2 9 7 a) a a 2 a * . 5 3 72 °O Sc = a =) = Rio O8cé “ON i oO a 4 Ss co a a = S,- Ss Sd a2 & = Sl > rae ~ 19 fo) co ye “6172 “ON a = = ea) a S oS 1D a ey E a Cry a a S me} a a 4 3 H 2 i) ~ 0 reg ‘unosnyy Ys Scie sterance viene Crys ea eet tence WON i) Goals at “BysVly ‘sn9070.00 °a “7 Sy Se) cn q = rt a a = 2 9 = = So) tS) & SI ‘9 Q “Tere + a (2) na Mo} ot 5 ay 28 °O ao oS 7 o oO g P8GE “ON = Cag CN ee Or SCAN Nec g 2 sy = Q ) Ss mw 3 SH oo SF “Sunox Z S = Ge) Ns} oa oo 12 ae a ~ ~ H ~ i 19 : 1D zl "E838 “0 = Qe SS eI es, eS 38 NT =) = AQ a H Ss Ns BS S + 7) + oe} Ce 4 H 4 “Z9ze *0 4 a a a a So — 1D x t fat s86E “ON 63 a) a =| S mK = cs XS Toh 6: Lon: I N a .) ~ ao a =H f 91 “WN a a S a KL S ~ a co + snaruiagry °a'T ich Se) a o Ss a a = ] | “moTZO9TTON 930qq = a 2 eA Co) =H ~ Lal ~H On ee Coes 0 S - = a) S a ; FSy CS Te Sr) J PRO UA Ee DI ies ee Bee hn pao Goro Gee) rey 0 4 S > nN i) q 114 5S) D £0 = SS CAE ER ee BS in PNR, SR a NERD NN 3 a a co ~ Gey 19 ~ @ | roa} a S Ge) a) S S ~ 1 g ‘ oo = cml 1S 2 So) o oo a a S a a = Bt | Be & |e a eS ee g2) ce 8 Fass 5 po ey ae —~ BS 5 : o> i=} (s) D =| S ot. ee = ° eS Bia Ga - 5 gs. es o ~. eas we 55 S ey = TOO 0qq Vy [10D preauey puv 1ou10p oO “Bing ‘Top “oy | “Sing [109 “Aoy | x “snodouna “YK *S70092098 *0 “T m “TISSOW NVHLHOT ‘VIAL, Hinron—The Fossil Hare of the Ossiferous Fissures of Ightham. 255 TIBIA. T am unable to point out any difference between the tibia of LZ. variabilis and that of Z. ewropeus. In the Scotch form of the former species it has the shaft proportionally a little more slender, but in JZ. v. hibernicus and the fossil Hare from Ightham the tibia has the stout proportions of L. ewropeus. The great size of the fossil tibize is conspicuous. METATARSALS. — | Mt. II. | Mt. III. | Mt. IV. | Mt. V. Extreme length, max., 30 o0 58:7 59°7 57:1 50°5 Skeleton (Corner and Kennard Coll.), 56:0 55°9 — 47-6 Min., .. 00 00 00 50 54:1 55:2 54-7 47°5 Metatarsal 11.—Of this bone I have nine examples from the Ightham Fissures before me. In one or two I think I am able to see traces of the small facette for the precuneiform described by Forsyth Major,’ but it is not very distinct in any. Metatarsai 11J.—Nine examples examined. Metatarsal 1V.—Six examples examined. Metatarsal V—Nine examples examined. These bones call for no further comment here. Nathusius? has described the differences which exist between the structure of the limbs, and particularly of the fore-limb, of the Hare (ZL. ewropeus) and those of the Rabbit Caprolagus (Oryctolagus) cuniculus. Forsyth Major° has more recently considered this subject from a far wider standpoint, and has conclusively shown that the fore-limb of the Hare has been specialized for fleetness, and that that of the Rabbit remains, as regards the antebrachium, practically in the primitive condition, instead of having been modified to meet the requirements of a presumed fossorial existence, as Nathusius supposed. And he further shows that of the Rabbits only one, viz. C. hispidus, possesses a limb-skeleton which betrays any trace of a fossorial specialization. 1 Forsyru Masor, ‘‘ Recent and Fossil Lagomorpha,” Trans. Linn. Soe., Ser. 2, Zool., vol. Vil., p. 505. ? Naruusius, ‘‘ Ueber die sogenannten Leporiden,’’ 1876, p. 31. 3 Forsytu Maso, ‘ Fossil and Recent Lagomorpha,’”’ Trans. Linn. Soc., Ser. 2, Zool., vol. vii., pp- 487-493. 256 Scientific Proceedings, Royal Dublin Sociely. It is necessary to glance at the differences which Nathusius describes as existing between the Hare and the Rabbit. He gives the following comparison! :— Hare. Rapstr. Ulna weaker than and situated Ulna stronger than and situated behind the radius. beside the radius. Compared with the basal length of the skull and the length of the vertebral column the fore and hind limbs in their entirety and in their individual parts are relatively :— Longer in the Hare. Shorter in the Rabbit. Humerus Jonger than antebra- Humerus and antebrachium of chium. approximately equal length. Antebrachium shorter than tibia Antebrachium shorter than tibia by about one-fourth of the length of by about one-half of the length of the latter. the latter. Nathusius says elsewhere that “the humerus of the Hare is longer rela- tively to the antebrachium than that of the Rabbit.”? This is just the reverse of the truth; for Nathusius’ own measurements show that the humerus is shorter than the antebrachium in the Hare and approximately equal to it in the Rabbit. It will be of interest to compare the relations subsisting between the principal bones of the limb-skeleton and between the limbs and the skull in the fossil Hare from Ightham, the recent forms of Z. variabilis and in L. ewropeus. The investigation of the material before me had already been completed when I received from Dr. Herluf Winge a copy of his valuable paper on the mammalia of Greenland.‘ In this paper he gives a table of measurements of the skulls and limb-bones of a suite of specimens of Z. v. glacialis (= arcticus and greniandicus of the present paper), ZL. variabilis (Sweden and Norway), and L, europwus. I had only one imperfect skeleton of L. v. arcticus before me; and therefore the data furnished by Winge are the more welcome. In order to facilitate the comparison, the table at pp. 258-259 has been drawn up; and it contains not only the results of my own measurements, but similar calculations based on Winge’s figures. : Natuustius, op. cit., p. 67. 3 Naruustivs, op. cit., pp. 44, 45. * Ibid., op. cit., p. 46. 4 Wines, ‘‘ Grénlands Pattedyr,” pp. 377, 378. Hinton—The Fossil Hare of the Ossiferous Fissures of Ightham. 257 Winge takes the condylo-basal length of the skull as his standard, while I took the basal length. The average relation between the basal length and condylo-basal length in the Hares appears to be as 100:110; and in section 4 of the table my results are reduced in this proportion so as to be comparable with those obtained by Winge. Although not quite accurate, the figures in section 4 relating to my material are accurate enough for the present purpose. The first fact brought out by the table is that in the southern forms of L. variabilis the limbs, in proportion to the skull length, are considerably longer than in Z. ewropeus. But in L. v. hibernicus and the Ightham fossil Hare the limbs are shorter relatively than in the Scotch form (represented by the skeletons Nos. 3282 and 3283 Roy. Coll. Surg.). The Norwegian form measured by Winge apparently agrees with Z. v. scoticus in this respect, while, on the other hand, the Swedish L. variabilis of Winge’s table agrees in this with Z. v. hibernicus. The high northern forms have shorter limbs; and the hind limb especially shows a tendency to return to the short-limbed condition of ZL. ewropeus. This shortening is here probably a secondary specializatioi. induced by the frigid habitat, and not a primitive condition as in LZ. ewropeus. The second point worthy of notice concerns the relation of the radius to the humerus. The southern forms of Z. variabilis, including the fossil Hare (in which it is marked), have the radius approximately equal to or a little longer than the humerus, i.e. they retain the primitive relationship of these bones found in the Rabbit. In Z. ewropeus the radius becomes considerably longer than the humerus; and this specialization appears also in some of the boreal forms of LZ. variabilis. I suspect that eventually it will be found to be the normal condition of ZL. v. grenlandicus. [ TABLE. SCIENT. PROC. R.D.S., VOL. XII., NO. XXII. 27 258 Scientifie Proceedings, Royal Dublin Society. L. v. scoticus. IGuTHAM FossiL. Roy. Coll. Sur & =: We 0 S: 8 . =. x = i cS) "3 g 3 oa 3 aoe — gs | ss Es Cre] ae Be | He Sa a | oe eS 25 ag oH A. Si 2 2.8 2 am Be am aoe a % &,° % ae) ‘S R Se 3 ame | 3 2 S S y A A A >) 1. 1. Basal length of Skull, Bb 100 _— — 100 100 100 — 100 2. Length of Humerus + Radius, 277 — — 278 292 282 — — 3. 99 Femur + Tibia, 349 = — 847 363 366 _— — 4, 9% Humerus, xe 138°5 oa — 136°5 143 137 — 137 6. a0 Femur, 163 _— _— 164 170 167°5 —_ 157 2. 1, Length of Humerus, 100 100 100 100 100 100 100 — 2. 9p Radius, 10071 99 102 103°5 103°5 105°5 101 — 3. 1. Tength of Femur, Oo 100 —_ 100 100 100 100 100 100 2. » Tibia, is 113 = = 111°5 | 113 118 = a a 3. 50 Humerus 85 _— 84 83°5 84 82 82 87:2 4. 1. Condylo-basal length of Skull, 100 — — 100 100 100 = 100 2. Length of Humerus + Radius, 252 — — 258 265 256 — — 3. $5 Femur + Tibia, 318 — — 316 330 333 — — 4. 59 Humerus, 126 —_— — 124 130 125 — 124 bop ea Memur, 148 = as 149 155 152 = | tag | Hinton—The Hossil Hare of the Ossiferous Fissures of Ightham. 259 4 ino ? LT. europeus. (Calculated from Winge’s data.) | Roy. Coll. Surg. LT. v. arcticus and grenlandicus. L. variabilis. | | ; = rd px 5 I 5 | ote : 9 a 5 A 3s = 3 d= < 4 | 2 2 Ss 8 23 ae aq 3 = & ; $4 a a S PRES en recs al meSoae ee ; 5 3 =3 3 5 Ny “a : "e 8 12 5 5 2 & 4a A S ey E 2 5 a Z 7, & s | 100 100 100 = — 261 262 193 — i 334 329 266 _ — —_ — — — — = pat 126 125 91 —_— _ — — = = 157 155 118-5 —_ —_ — — —_— = = = ie 100 100 100 100 100 100 100 100 100 100 100 100 107 109 99°5 109°5 108 102 106 104°5 105 102 104°5 108 100 100 100 100 100 100 100 100 100 100 100 100 114 112 116-5 118-5 117 115°5 117 114°5 116°5 120 119°5 116 80:0 81 77 84 85 88 85:8 87-8 82°7 84°5 86°2 83 100 100 —_ 100 100 100 100 100 —_ 100 100 100 237 238 — 254 244 242, 242, 247 —_ 260 253 240 304 299 —_ 314 300 294 297 296 —_— 335 316 300 116 114 — 121 117 120 117 121 —_— 129 124 115 143 141 _ 143 138 137 137 138 —_— 152 144 139 272 260 Scientific Proceedings, Royal Dublin Society. In the preceding pages I have not dealt with the Irish fossil remains of Hares, because I have no material, and such descriptions as have been published are insufficient for my purpose. In his first account of the mammalian remains from the caves of county Clare, Scharff! says :— “Compared with a modern skeleton of the Irish Hare, the Cave Hare was a considerably larger and more powerfully built animal, yet the distinctive characters of the Mountain Hare are retained. . . . It would seem, therefore, as if the Irish Hare has deteriorated in size and strength since it arrived in this country.” In a later paper, on the bones from the “Catacombs,” he states that all the humeri and femora were longer than those of the recent skeleton ; while, on the other hand, ‘of ten complete tibize measured, four were smaller than the recent tibia.””’ Further:—*“ This fact seems to indicate that the upper portions of both front and hind limbs have become shortened in the Irish Hare in course of time, possibly owing to change of habit.” He gives the following measurements of some limb-bones from Newhall and Barntick Caves :— a Et I Si Ta Nc Dr oa re REDUCTION. Minimum. Maximum. Minimum. Maximum. Femora, 117 131 100 100 Tibie, 129 149°5 110 114 Humetri, . 100 109 85:5 83 These figures suffice to show that the fossil hare of Ireland had already become differentiated from Z. var. anglicus, and had already developed into the comparatively small island-form, L. var. hibernicus. ConcLusIon. The variabilis group of hares has a circumpolar distribution ; and whilst it ranges continuously throughout the more northern regions, isolated colonies of it are found stranded far to the south in the mountains of central and southern Europe. In Britain it inhabits the more mountainous parts of Scotland; and it lives throughout Ireland, from which latter country, as is well known, Lepus ewropeus is absent. 1 Scuarrr, Trans. Royal Irish Acad., yol. xxxil., Sect. B, p. 199. * Ibid., vol. xxxiii., Sect. B, p. 37. Hinton—The Fossil Hare of the Ossiferous Fissures of Ightham. 261 There can be no doubt that the hare which lived in southern England during late Pleistocene times was a member of the ZL. variabilis group. So close is the affinity between this fossil form and the living ZL. var. hibernicus that we may regard the latter as the direct descendant of the former. The fossil English Hare is distinguished from the Irish Hare by its larger size, and some other characters already dealt with, which go to show that the fossil form was a little less specialized than its living representative. Further, the English Pleistocene Hare has no closer affinity with the fossil forms described by Woldrich and Hescheler from Zuzlawitz and Thaiingen than has L. var. hibernicus with L. var. scoticus, L. var. varronis, or any of the other subspecies which are recognized to-day. ‘he fossil remains ascribed to L. “diluvianus” by the older French paleontologists are too imperfectly known to be drawn into the comparison; and in any case their nearest living relatives are probably to be sought among some of the continental subspecies of Z. variabilis, or even perhaps of L. ewropeus. In these cireum- stances it appears to be desirable to regard the English Pleistocene form of L. variabilis as a distinct subspecies, for which I venture to suggest the name of L. var. anglicus. Lepus ewropeus does not appear to have arrived in England until after the Pleistocene period; and it may also be that during this period, when the plains of western continental Europe were inhabited by members of the variabilis group, this species was absent from the north-western part of Hurope generally. According to the prevalent view the variabilis group is of boreal origin ; and the extension of its range southwards is usually explained by reference to the glacial phenomena of Pleistocene times, the temperature during this period having been, it is supposed, so lowered as to enable the group to live in the plains. With the return of temperate conditions, the Snow-Hare either fled northwards or retreated to those mountain fastnesses in which we now find it living. ‘The separation of Ireland from Britain, which is supposed to have happened before the passing of the cold conditions, prevented L. variabilis from leaving that country when the climatic conditions were no longer favourable to it; and accordingly it had to adapt itself to the subsequent changes. Stated in this way the theory explains most points in the geological history and geographical distribution of the group ; and conse- quently we find ZL. variabilis figuring in geological literature as one of the witnesses in favour of the former existence of extreme climatic conditions in the present temperate parts of western Hurope. The introduction of the trinomial nomenclature into zoological classifica- tion, with the necessity for increased exactness of observation which its 262 Scientific Proceedings, Royal Dublin Society. application involves, has conferred a great boon upon those investigating late Tertiary history, since it supplies us with a very powerful means of testing such generalizations as that outlined in the preceding paragraph. It has long been recognized that, as any given mammalian group moves outward from its centre of dispersion, it suffers more or less profound modifications of structure, so that the most highly specialized forms are in general precisely those which exist, or have existed, furthest from the original centre of dispersion. Good instances of this principle are afforded by the cervine ruminants, where we find such transformations of the primitive type as the Giraffe, the Elk, and the Reindeer as peripheral representatives.' I hope to show that exactly the same principle asserts itself in the case of the subspecies of the ZL. cariabilis group. The Hares present us with a curious combination of ancient and modern characters. The dental formula is very primitive compared with that of other rodents; but the individual cheek-teeth have undergone the most profound specialization. The organs of sight and hearing are greatly enlarged ; and the limb-skeleton is well on its way to the reduced condition of the swift-footed ungulates. Of the two groups of Hares with which we have been concerned in this paper, viz., the variabilis and europeus groups, the former is on the whole the more specialized. It is necessary for our present purpose to look at the various members of the variabilis group a little in detail. Dealing first with the skull, Winge has shown us how to interpret the changes found in proceeding from one member of the group to another. We will confine our attention to the relations subsisting between the eye and the ear, because the evidence of the fossil form and of Z. v. hibernicus on this point also has a geological bearing. Both organs are abnormally large in all Hares. With regard to the variabilis group, as we proceed northwards to colder climes, the peripheral parts tend to diminish in size, not because they are less useful to the animal, but because of the cold. In accordance with this, the outer ear becomes shorter in the northern forms of the group; but, as Winge points out, the loss in this respect is compensated by an increase in the size of the eye; and this increase has its due effect upon the skull, for as the eye becomes larger the superciliary processes of the frontal tend to elevate themselves more and more in proportion.? In perfect harmony with this we find that in Z. v. hibernicus and in L. v. anglicus the superciliary processes are less elevated and the interorbital region flatter than in any other forms of 1 Rurimeryer, Beitrdge zu der Geschichte der Hirschfamilie 1. Schidelbau. Verhand. der Naturforsch. Gesell. in Basel, vii. Theil, 1. Heft, 1882, pp. 45-67. * Winer, ‘‘ Gronlands Pattedyr,”’ p. 358, and ‘‘ Pattedyr,’”’ p. 58. Hinton—The Fossil Hare of the Ossiferous Fissures of Tghtham. 263 the group. We may conclude from this that with regard to the organ of sight, L. v. anglicus and L. v. hibernicus are the least specialized members of the variabilis group ; and we may infer further that both the forms in question are forms which live or have lived under temperate conditions. The limb-skeleton, as we have seen above, gives similar evidence; for in these two forms of the variabilis group, we have the primitive combination of short limbs, and equality of the brachial and ante-brachial lengths. It is true that in the arctic forms of the group short limbs are characteristic, but that this is a secondary specialization or retrogression here is shown by the fact that the radius has become considerably longer than the humerus. The intermediate forms from Scotland and Scandinavia show longer limbs and have rather longer radii than those of the southern forms. In these respects, therefore, and in every other point dealt with in the preceding pages, L. v. anglicus and L. v. hibernicus distinguish themselves as the two least specialized members of the LZ. variabilis group, and, on the other hand, the arctic forms are as clearly the most specialized. It is consequently highly improbable that the group can have had a boreal origin. Looking at the whole subject broadly, it appears to me to be more likely that the group had its origin somewhere in Central Asia. From this region it later spread eastwards into North America and westwards into Central Europe. The colonization of the high north by the group may have taken place by three distinct routes, viz., (1) directly from Central to Arctic Asia; (2) through North America; and (3) from Central Europe along the western seaboard through Scotland and Scandinavia. The group has become specialized and differentiated into subspecies in the course of its travels, and particularly as it has proceeded northwards. Therefore the southern forms, living and extinct, remain nearer to the primitive Z. variabilis than do the northern ones. But if this view brings the distribution of the group into harmony with the structural peculiarities of its members, there yet remains to be explained the singular fact that, with the exception of Ireland, the Z. variabilis group no longer inhabits the plains of temperate Kurope. In a former paper I have argued that to explain such changes in distribu- tion it is not necessary to invoke great changes of climate.! In the present case the structure of Z. v. anglicus itself tends to disprove the idea that the climate of the south of Kngland was any less mild at the time of this Hare’s existence there than it is at present. Moreover its nearest ally lives under ? Hinton, ‘On the Existence of the Alpine Vole in Britain during Pleistocene Times,”’ Proc. Geol. Assoc., vol. xx., p. 54, 264 Scientific Proceedings, Royal Dublin Society. the temperate conditions which at present obtain in Ireland. In my view the reason why Z. variabilis continues to live in the Irish plains is that L. europeus has been unable to reach Ireland, and the reason why L. variabilis has been driven out of the plains of Britain and of Western Europe is that there Z. ewropeus has proved too strong a competitor. In the paper above cited, this question of the reaction of one species upon another has been dealt with; but perhaps it may be mentioned here that the views there advocated have received lately a rather striking confirmation. Crespon’? long ago described a vole from the neighbourhood of Nimes in the department of Gard in France under the name of Arvicola (Microtus) lebrunti. Blasius, in 1857,? said that, so far as he could gather from the somewhat unsatisfactory description, this animal should be placed with MW. nivalis. There the matter rested until some little time ago, when Mr. Charles Mottaz visited the locality in the interests of the British Museum; and he was fortunate enough to obtain specimens of I. lebruniz. These were examined by Mr. Gerrit S. Miller, who found that I. lebrunii undoubtedly does belong to the M. nivalis group, and indeed to what was previously known as the /ewcurus section‘—a section which probably existed in Britain during Pleistocene times. But UM. /ebrunii differs from all its living allies in the fact that it lives at so low an altitude as 550 feet above sea-level, and this “in the midst of the vine and olive region of the Mediterranean coast,’ whilst the other living nivaloid voles exhibit a marked preference for the snow-line. Here surely is the clearest proof that mere temperature has nothing whatever to do with the distribution of the nivalis group, just as the tenancy of the Ivish plains by L. variabilis proves that the more generalized and primitive forms of the arctic or mountain Hare do not require cold. 1 Sonarrr, ‘‘ History of the European Fauna,”’ 1899, pp. 136-8, 148, 315-6. “‘ European Animals,’’ 1907, pp. 89, 97, 189-140. In these extremely valuable works Dr. Scharff arrives at similar conclusions with regard to the relations subsisting between L. variabilis and L. europeus, and to the bearing of the former species upon the question of Pleistocene climate, to those herein expressed. The standpoint from which I have approached this subject is, however, a quite different one from that of Dr. Scharff. He assumes an Arctic origin for the Z. variabilis group, and also thinks it likely that it reached western Europe in Pliocene times. I attempt to prove that the group had its origin in temperate latitudes ; and I do not believe that LZ. variabilis came into existence until the Pleistocene period ; and as regards England, for reasons to be stated elsewhere, until a quite late moment in Pleistocene time. So far as Dr. Scharff’s main contention is concerned, viz. that the climate of the ‘‘ Glacial Period ’’ was not so severe as the preyalent theory supposes, I unreservedly agree with him, not merely upon the grounds stated here and in former papers, but for many other reasons with which I hope to deal in the near future. 2 Cresvon, ‘‘ Faune Méridionale,’’ i., p. 77, 1844. 3 Buastus, ‘¢ Saéugethiere Deutschlands,”’ p. 362. * Mitier, Ann. and Mag. Nat, Hist., Ser. 8., vol. i., p. 101, Hinton—The Fossil Hare of the Ossiferous Fissures of Ightham. 265 The moral which I wish to draw is perhaps pointed by the fact that, although many attempts have been made to introduce LZ. ewropeus into Treland, they have all proved more or less unsuccessful.1 One may be tempted to ascribe the failure to some climatic difference, although it is difficult to quite imagine what such a difference may be. The early separation of Ireland from Britain, which has kept Ireland free from L. europeus, may also have kept out of the latter country some other organism whose presence is essential to the prosperity of the common Hare or inimical to that of the Irish Hare. If one could find out exactly what this factor is, and succeed in introducing it into Ireland, one might then succeed in introducing LZ. ewropeus. In the latter case L. variabilis would, judging from what has happened in England and elsewhere, rapidly lose ground. The moral itself is that, however tempting the idea of great changes of climate may be as affording a ready explanation of the changes which have taken place in the distribution of animals and plants since the Pleistocene period, or since a still more remote time, we should not adopt it without considering whether or no the organisms themselves, by their own interaction and interdependence, can have produced the modifications which they have suffered both in their habitats and in their structure. It might be put even higher, namely, that the paleeontologist should not dream of changes of climate and the like until he is satisfied that the organized part of nature is incapable of working out its own destiny. There can be no doubt that the relations existing between the elements of a fauna and flora are of the most eomplex kind, the right appreciation of which can only come to us as the reward of the most patient investigation in the future; and therefore, having regard to our present ignorance, it does seem a little arrogant to say that this or that case transcends the capabilities of organic nature, and that consequently we must attempt to explain it by burdening the equation with other powers still more difficult to comprehend and still more far-reaching in their effect. 1 Barrerr-Hamiiron, Irish Naturalist, vii., 1898, p. 69. SCIENT. PROC. R.D.S., VOL. XII., NO. XXIII. 2u EXPLANATION OF PLATE XV. PLATE XY. Fig. 1. Lepus variabilis anglicus, Hinton. Skull of the type skeleton : seen from above. From a Pleistocene rock-fissure at Ightham, Kent. (In the collection of Dr. Frank Corner, F.G.S., and Mr. A. §. Kennard, F.G.8.) Fig. 2. The same: palatal aspect. Fig. 8. The same: anterior view. Fig. 4. The same: right profile. Fig. 5. L. variabilis anglicus. Anterior view of nasal bones and premaxillaries of a skull from a Pleistocene rock-fissure at Ightham. (In the’ collection of Mr. W. J. Lewis Abbott, F'.G.8.) Fig. 6. L. variabilis anglicus. Left mandibular ramus of the type skeleton. Inner view. Fig. 7. The same: outer view. (All the figures are of natural size.) PLATE XY. R. DUBL. SOC., N.S., Vou. XII. SCIENT. PROC. Watford Engr. Co. ma ON a ct 7 With the Author‘s Compliment: THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XIL (N.8.), No. 24. JANUARY, 1910. SOME OBSERVATIONS OF DEW AT KIMBERLEY. BY J. R. SUTTON, M.A., Sc.D. |comMMUNICATED BY PROFESSOR W. BROWN, B.SC. | { Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN : pe PUBLISHED BY THE ROYAL DUBLIN sporty,‘ LEINSTER HOUSE, DUBLIN. ‘ MAR 14.3919 WILLIAMS AND NORGATE, ear AU 14, HENRIETTA STREET, COVENT GARDEN, LONDON, Wee Ona) Muses j 5 —_ TOMO: Price Sixpence. Weis cz een 5 ae 1 [ 26 4 XXIV. SOME OBSERVATIONS OF DEW AT KIMBERLEY. By J. R. SUTTON, M.A., Sc.D. [COMMUNICATED BY PROFESSOR W. BROWN, B.SC.] [Read Novemper 23. Ordered for Publication Drcrmper 7, 1909. Published January 15, 1910.] Britisu meteorologists of late years, in discussing the theory of dew and its formation, have shown some disposition to abandon the teachings of Wells in favour of those of Aristotle. A few quotations (out of many that might be made) from the writings of well-known authors will serve to illustrate this backsliding :— 1. “‘It is observed that dew is never copiously deposited in situations much screened from the open sky, and not at all in a cloudy night; but if the clouds withdraw even for a few minutes, and leave a clear opening, a deposition of dew presently begins.”’ 2. Dew “is never deposited in cloudy weather; and so strict is its connexion with a clear sky that its deposition is immediately suspended whenever any considerable cloud passes the zenith of the place of obser- vation.” * 3. “ Dew is deposited over the Harth’s surface on comparatively clear and calm nights. . . . Dew is not deposited in cloudy weather, because clouds obstruct the escape of heat by radiation.” * 4. “Tt is known to everyone that dew does not appear on a cloudy night.” 4 5. “ Dew may be defined as moisture deposited when visible cloud is absent.’’ ° Statements such as these are the more remarkable in view of the fact that on the soil of Great Britain have been made the observations which, more than any other, have put the theory of dew upon a rational basis. It is not 1J. Herschel, ‘‘ Preliminary Discourse on the Study of Natural Philosophy,’’ 1831, p. 162. 2J. Herschel, “ Meteorology,’’ 1862, Art.91. The italics are Herschel’s in both quotations. ’ A. Buchan, in Art. ‘¢ Meteorology,’’ Ency. Brit., 1878. 4R. H. Scott, ‘‘ Elementary Meteorology,” 1893, p. 117. 5W. Allingham, ‘‘ A Manual of Marine Meteorology,” 1900, p. 149. Surron—Some Observations of Dew at Kimberley. 267 a strictly true that a clear sky is essential to the condensation of moisture in the form of dew. Dew may appear under a clouded sky, as plenty of English authors since the time of Wells—notably Harvey! and Tyndall*—to say nothing of foreign ones, have pointed out. A clear sky is only favour- able to dew-making when other conditions remain the same. ‘I'hat is to say, let the temperature of the air and of the dew-point at sunset be respectively the same on any two evenings; then if one of the evenings be clear aud the other cloudy, the former will show, to begin with, the more rapid rate of dew-formation. But the point to be taken into account — is that in nine cases out of ten the dew-point is higher on the cloudy night than it is on the clear night, and hence that the temperature of the air has not so far to fall to the saturation-point, so that, although the clouds do considerably check the radiation of heat from the Harth’s surface, yet, on the other hand, no great intensity of radiation is required in order that the lower air may cool sufficiently to allow its excess of moisture to be condensed.* This fact is very well exemplified by the observations of dew made at Kimberley. It is not in the clear, bright, calm nights of the Kimberley winter that the most dew (or frost) is deposited, but rather in the relatively more clouded autumn. My own observations go to show that on a clear, damp night a great deal of dew is made in a short time, but that the energy of the dew-making process soon diminishes, largely, perhaps, because the high and inereasing relative humidity of tie layer of air in contact with the surface hinders a continuous rapid fall of temperature, and partly, perhaps, because of the high specific heat of water. Ona night when there are clouds, however, the rate of condensation may be less rapid; yet there are times when as much dew is deposited in the long run, in spite of the clouds, as in the former case when the sky is clear. Ihave not yet been fortunate enough to observe the rapid alternations of condensation and evaporation, as clear sky has alternated with clouds, which seem to have been observed by others. Nor have I ever seen any pronounced rise in the reading of a radiation thermometer lying on the grass (and concomitant evaporation of dew) which could be ascribed solely 1G. Harvey, in Art. ‘‘ Meteorology,”’ Hncy. Met., 1845. 2J. Tyndall, ‘‘ Heat a Mode of Motion,’’ 1880, p. 496. 3 It is a curious circumstance that a phenomenon so much within the province of meteorologists as dew is should haye been so often wrongly explained by them, whereas the physicists seem somehow to have got the explanation nearly always right :—e.g., W. H. Besant, ‘‘ A Treatise on Hydromechanics,” 1877, p. 116; also Poynting and Thomson, ‘‘ Heat,’’ 1904, p. 218. 4 These are mainly observations of dew proper, and not the ‘‘ dew” which proceeds directly from growing plants. 2u2 268 Scientific Proceedings, Royal Dublin Society. to the influence of a passing cloud. In my experience there is only one meteorological factor competent to produce a great and sudden rise of temperature at night, and that is a gust of wind, disturbing an evening calm. The gust almost invariably checks the formation of dew, and, if it continue long enough, evaporates what is already there. I have never seen a cloud alone do this. Table I gives a summary of the number of times dew was observed in the four years 1905-8, at 8a.m., 8 p.m., and 11 p.m. The designations “0. “<1,” “2.” at the head of each column, indicate roughly the quantity of dew observed at the respective hours; ‘‘0” indicating a slight dew; “2,” that everything is streaming wet; “1,” an intermediate state of affairs. We see from this table that dew is of infrequent occurrence in the last quarter of the year, the minimum of frequency falling in November, while the maximum isin April. It must be borne in mind, however, that there are occasional nights of dew in the summer, which do not appear in the record because it has been evaporated by the sun’s heat before 8 am. For that reason it is not possible to compare the state of the sky with the presence or absence of dew at 8 a.m. in the summer months. When frost is taken into account with the dew, the months of maximum and minimum are not changed. June is the month of most frost, almost all of which is observed at 8 a.m., i.e. the temperature rarely falls to the point at which frost can be deposited until after 11 p.m. And here it may be noted that just as there is no definite freezing-point of water (the thawing- point is of course quite definite), so there is no definite temperature which determines whether the air’s excess of ioisture shall be deposited in the shape of dew or frost. On astill night a water-surface may fall considerably below 32° F. without congelation, while solid bodies at the same low temperatures may remain covered with dew. 1 For an interesting case in point, see J. R. Sutton, ‘‘ A Low Freezing-Point,”’ Symons’s Met. Mag., July, 1905. Surron—Some Observations of Dew at Kimberley. 269 Taking dew and frost together at 11 p.m., we have the following monthly summary of the number of times either has appeared in four years, compared with the mean cloudiness of the sky :— Dew. Cloud. January : : 5 times. ; : 34 per cent. February . 5 Oe i ; 35 - March . F , N82 5. : ; 32 5 April . ; : BH é . 18 e May . : 5 PS : : 18 ‘{requie0aq 0} 19quieydag : Gsnsny 0} Avy - Gridy 0} Avnuepe “uta (7) “Nosvag ‘pnopo Aq poanosqo Ays Jo osvyuso10d oy} 07 Surpxoooe poSuvare ‘urd TT ye (g) pue “urd g qv (g) “meg ye (T) ‘g-Go6T ‘saved anoy oy} UI pearesqo Ueeq Avy (4s01f-1v0Oy PUL) Mop SOUIT} Jo tequINU oy} SuIMoyY TI Gav, R.D.S., VOL. XII., NO. XXIV. SCIENT. PROC. 274 evening sky. Tase III. Showing the number of times dew or hoar-frost has been observed in the four years 1905-8, at 11 p.m., arranged according to the state of the Scientific Proceedings, Royal Dublin Society. Sky, 30°|, Clear, xx. Clear, xx. Cloudy, xx. | Cloudy, xx. or more, Clear, xxiii. | Cloudy, xxiii.| Clear, xxiii. | Cloudy, xxiii. Clouded at | xx. and xxiil January, — — 2 | 3 1 February, . — — | 6 3 March, 4 4 1 | 9 6 April, 16 3 3 10 3 May, 16 1 4 8 Zz June, 12 — 5 8 _ July, 1 = os 3 = August, 2 — — = — September, 2 1 1 1 1 October, = — 2 3 2 November, 1 — 1 3 = December, — — = 7 5 YEAR, 54 9 22 56 23 THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 20. FEBRUARY, 1910. ON OSMOTIC PRESSURES IN PLANTS; AND ON A THERMO-ELECTRIC METHOD OF DETER- MINING FREEZING-POINTS. BY HENRY H. DIXON, Sc.D., UNIVERSITY PROFESSOR OF BOTANY IN TRINITY COLLEGE, DUBLIN. AND W. R. GELSTON ATKINS, M.A., DEMONSTRATOR IN BOTANY, TRINITY COLLEGE, DUBLIN. [Authors alone are responsib/e for all opinions expressed in their Communications. } DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1910. é ian Insti, NSO AST/f, os Uy i 19) Price One Shilling and Sixpence. JUN 43 1910 No, ional luseo™ Roval Dublir Society. NO Oe Oe Oe FOUNDED, A.D. 17381. INCORPORATED, 1749. OEE EVENING SCIENTIFIC MEETINGS. Tue Scientific Meetings of the Society are held alternately at 4.30 p.m. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Editor. XXV. ON OSMOTIC PRESSURE IN PLANTS; AND ON A THERMO- ELECTRIC METHOD OF DETERMINING FREEZING- POINTS. By HENRY H. DIXON, Sc.D., F.R.S., University Professor of Botany in Trinity College, Dublin ; AND W. R. GELSTON ATKINS, M.A., Demonstrator in Botany, Trinity College, Dublin. [Read November 23. Ordered for Publication Drcemper 7, 1909. Published Frpruary 28, 1910.] Summary. Tue following paper contains A. A description of a thermo-electric method of cryoscopy suitable for the determination of the freezing-point of small quantities of liquids with considerable accuracy. B. Records of a number (101 in all) of freezing-points of the saps of plants determined by this method, together with the osmotic pressures calculated from the freezing-points. These observa- tions apply to the— Leaves of : Name. noe : Name. a e 1. Catalpa bignonioides, . : : 2 11. Ivis germanica, . : ; : 101 2. >» Speciosa, 3 5 : 6 B 12. Magnolia acuminata, . 5) G5 75 UD; 20 3. Cerasus Laurocerasus, . ; ~ 29, 30 13. Pinus Laricio, . : : . 85, 86 4. Chamaerops humilis, . : 100 14. Populus balsamifera, . : : 1 5. Cordyline australis, . ° ‘ 99 | 15. Pteris aquilina, é s ; 96 6. Equisetum Telmateia, . : : 95 16. Syringa yulgaris, : 4, 31-74 7. Mucalyptus globulus, . : - 87-938 17. Ulmus campestris, : 3 . 21-24 8. Fraxinus excelsior, cae . 11-16 18. Vitis Veitchii, . : ; . 25-28 9. 29 oxyphylla, . ‘ 2 We Ue 19. Wistaria sinensis, : 2 . 80-84 10. Helianthus multiflorus, . . 97, 98 Roots of : 20. Eucalyptus globulus, . 6 . 94 | 21. Syringa vulgaris, : 5 . 75-79 Fruits of : 22. Ribes grossularia, : : o &O | 28. Ribesrubrum, . ; : x) tO Fifty of these observations were carried out on the sap of Syringa vulgaris with a view to tracing the changes in osmotic pressure with changes of external conditions. SCIENT. PROG, R.D.S., VOL. XIT., NO. XXY. 2y 276 Scientific Proceedings, Royal Dublin Society. C. Records of determinations of the mean molecular weights of the dis- solved substances in the saps upon which the experiments were made. Since the epoch-making work of Pfeffer and De Vries the osmotic pressures of plants have been investigated by many from various points of view. The present investigation was undertaken primarily to determine if the osmotic pressures in the leaf-cells of plants are sufficient to resist the tensions in the sap indicated by the considerations on which the Cohesion Theory of the Ascent of Sap is based. It was also proposed to determine if any relation existed between the osmotic pressure in the leaves and the height above the roots, or between it and the resistance of the stem connecting the water- supply and the leaves. Other problems also presented themselves during the work which have not as yet been solved, and indeed afford a large field for further investigation. Determinations of osmotic pressures in plants have usually been carried out by the plasmolytic method; and a number of determinations by this method are available. These observations have been mostly made on the tissues of stems, and, as a rule, have not been applied to transpiring organs. Quite recently H. Pringsheim? has applied the plasmolytic method to determining the turgor of leaves, dealing mostly with the leaves of succulents. He states that the turgor of the plants he examined was greater in the apical leaves than in the basal leaves, except in those of Phaseolus vulgaris. In winter turgor is less than in summer. Ewart® attempted to establish a rise in osmotic pressure according to the height above ground. In one case he found that the leaves of an elm at a level of 1350 cm. above the ground had an osmotic pressure two to three atmospheres greater than those at a level of 250 em. He found high pressures, e.g., equivalent to the osmotic pressures of a solution of potassium nitrate having a concentration of 6 per cent. or more, and records that the “plasmolytie concentrations” for leaf-cells of Acacia, Eucalyptus, and Grevillea vary in the same plant and at the same level between wide limits. At the same time he comments strongly on the want of precision inherent in the plasmolytic method. 1H. H. Dixon, Transpiration and the Ascent of Sap, Progressus Rei Botanic, Bd. iii. Hit. 1, 1909, p. 60 e¢ seq. 2, Pringsheim, Wasserbewegung und Turgorregulation in welkenden Pflanzen. Jahrb. f. wiss. Bot. Bd. xliii. 1896. 3A. T. Ewart, On the Ascent of Water in Trees (First Paper). Phil. Trans. Roy. Soc. yol. exevili, 1895, B, p. 77, and idem (Second Paper), yol. excix, 1908, B. p. 341. Dixon anp AtrKiIns—On Osmotic Pressure in Plants, Ge. 200 a In 1907 E. and H. Drabble! investigated the osmotic pressure of the cells of many plants, mostly herbaceons, by the plasmolytic method. They obtained results varying from 4.45-19.68 atm. In addition to many interesting observations, they claim that their results show in general that physiological drought is the principal determining factor in the osmotic pressure of plants, and they hold that the plasmolytic method is, with certain precautions, reliable. By quite a different method one of us carried out a number of observa- tions on the osmotic pressures in leaves. The method consisted in balancing the internal osmotic pressure against high external gas-pressures.? ‘he danger attending the observations render the method unsuitable for a detailed investigation of a large number of leaves under varied conditions. The results, however, which were obtained by this method were consistent, and indicated that the osmotic pressures in the leaves are at least as great as the tensions developed in the sap. On the whole, the results obtained at that time were a good deal higher than the generally accepted values for leaves which were based on plasmolytic observations ; but, as will appear later, these gas-pressure results have received a general confirmation in the present research. Sutherst,* by squeezing the juice out of certain leaves, showed that the sap so obtained has a freezing-point considerably below 0°C. ‘This author’s work was designed to explain why leaves do not freeze when exposed to temperatures even below 0°C. Sutherst’s results were obtained with easily expressed saps—e.g., celery, cabbage, and carrot, etc., amounting to seven in all. Beyond stating that he used a small quantity—viz., 5 c.c.—and a freezing-mixture of Glauber’s salts and hydrochloric acid, Sutherst did not describe what precautions he took in making these determinations: conse- quently his results must be looked upon as of a preliminary nature. The determination of the freezing-point affords a most convenient method of measuring the osmotic pressure, which is constantly made use of in all kinds of physical and physiological work. Livingston’ used Sutherst’s determinations, and calculated the osmotic pressures they indicated. With suitable apparatus and proper precautions the freezing-points of various solutions may be determined with great accuracy. From this the osmotic pressure of the solution may be deduced, inasmuch as the depres- 1H. and H. Drabble, The Relation between the Osmotic Strength of Cell Sap in Plants and their Physical Environment. Biochemical Journal, 1907, vol. ii., p. 117 et seq. 2H. H. Dixon, On the Osmotic Pressure in the Cells of Leaves. Proc. Roy. Irish Acad., vol. iy., ser. 3, 1896, p. 61, and On the Physics of the Transpiration Current, Notes trom the Botanical School, Trinity College, Dublin, No. 2, 1897, pp. 81 and 82. 3°W. F. Sutherst, Chemical News, 1901, p. 234. +B. E. Livingston, The Réle of Diffusion and Osmotic Pressure in Plants. Chicago, 1903. 2x2 ES. Scientific Proceedings, Royul Dublin Society. sion of the freezing-point of a solution and its osmotic pressure are directly dependent on its concentration." Beckmann’s apparatus, which is usually employed in these determina- tions, involves the use of a thermometer with a large bulb which must be immersed in the solution whose freezing-point is to be determined. ‘The size of the bulb necessitates the use of a considerable quantity of the solution, viz., 12 to 15e.c.as a minimum. Such large requirements seemed to us to preclude the application of the method to the determination of the osmotic pressures of the sap of transpiring organs, of which but small quantities can be conveniently obtained. ‘This objection applies all the more strongly to the other more elaborate methods of determining the freezing-points of solutions as described in Hamburger’s Osmotischer Druck und Ionenlehre. In order to circumvent this difficulty, it occurred to us to replace the mercurial thermometer by a thermocouple, and to compare directly the freezing-point of water with that of the solution. Past experience* had shown that with a suitable galvanometer and a single element of nickel and copper, it is easy to obtain a motion of the spot of light on the galvanometer scale of one millimetre for a difference in temperature at the junctions of 0:01° C. As the variations in the freezing-point of the sap of various plants and under various conditions are large, this is ample delicacy. Apparatus. The couple for observing the temperature was made of a piece of silk- insulated nickel wire 0°15 mm. in diameter, and 30 cm. long. Tach end of this was stripped for a few millimetres, and soldered to a fine insulated copper lead. he nickel wire was bent in a V-form; and each lead was twisted on the arm of the V to which it was attached. From the angle of the V the two leads were twisted together till they again diverged from one another to make connexion with the terminals of the galvanometer. Hach arm of the V was bound on a rigid rod to give it the necessary stiffness. ‘his support was either a piece of drawn pinewood injected with paraffin or a goose-quill. The silk-insulation of the wires and the junctions was smeared over with several coats of rubber-solution to water-proof it and 1902. [While this paper was in press we received through the kindness of Professor I’. Cayara two papers containing important results obtained by this method, viz. F. Nicolosi-Roncati, Ricerche su la Conduttivita Elettrica e la Pressione Osmotica nei Vegetali. Bull. dell’ Orto Botanico della. R. Uniy. di Napoli, T. ii., Fase. 2°, 1907; and G. Trinchieri, Su le Variazioni della Pressione Osmotica negli Organi della Salpichroa rhomboidea, Miers., ibid., 'T. ii., Fasc. 4°, 1909.] * Henry H. Dixon, Observations on the Temperature of the Subterranean Organs of Plants, Trans. Roy. Ivish Acad., vol. xxxil., sect. B., part ili., p. 145 e¢ seg. Dixon anp ArKINs—On Osmotic Pressure in Plants, &c. 279 prevent short circuiting and the chemical action of the solutions on the junctions. The solution whose freezing-point was to be determined and the distilled water which was to act as the standard were put into two small test-tubes about 10 cm. long and 1 cm. in diameter. ‘These were supported in a large eork bung which fitted into the neck of a cylindrical glass vessel. The latter could be immersed up to the neck in a larger vessel which contained a freezing-mixture of ice and salt. The galvanometer employed is an Ayrton-Mather instrument manufac- {ured by the Cambridge Scientific Instrument Co. Its resistance is 20°7 w. The deflection at one metre for one micro-volt is equal to 10 mm., and for one micro-ampere 206 mm. ‘The translucent screen was set at 94 em. distance from the mirror of the galvanometer, while a Nernst lamp illuminated the mirror. With this form of apparatus the freezing-point of 2°5 e.c. to 5 cc. of a liquid could readily be determined with considerable accuracy to 0:01° C.; and a number of determinations of the freezing-points of saps extracted from various leaves were obtained. It will be noted that in this arrangement continuous leads connected the junctions with the terminals of the galvanometer; and at the outset it was thought that, by thus avoiding all discontinuities in the circuit, and by having the continuous leads themselves forming the junctions in the couple, all extraneous thermo-electric effects, except those at the terminals of the galvanometer, would be eliminated. ‘his hope seemed to be confirmed when, on placing the two junctions close together in a vessel of water, which was vigorously stirred, the spot of light from the galvanometer-mirror came to rest at the scale zero. Greatly to our surprise, however, when the two junctions were put in melting ice, the zero was not attained if any part of the V-shaped couple remained outside the ice and at a higher temperature. This failure to reach the zero was not apparently due to heat conducted more freely to one junction than to the other, as it occurred even when both were immersed precisely to the same depth in the ice, and consequently would receive sensibly the same quantity of heat conducted in from the outside. As soon as the whole of the V was immersed in a vessel at 0° C., this zero-error disappeared. It seems that for such a sensitive galvanometer the bend acts as a junction, and the two halves of the nickel, not being absolutely identical, act as elements of a couple. The difference of temperature of the air and the melting ice is able, with this unlooked-for couple, to produce a deflection of the spot of light on the scale amounting to 4-7 mm. 280 Scientific Proceedings, Royal Dublin Sociery. In order to avoid this zero-error, deeper vessels (m and e) were used for containing the freezing mixture and for enclosing the test-tubes (see fig. 1). The latter (@ and 6) were as before carried in a perforated cork bung (Cc), which, however, was made to fit loosely down into the inner vessel, and was supported there by a stout brass wire (w) passing down through an upper cork (d@) which fitted the neck of the inner vessel (e) closely. In this way the cold air-space en- closed not only the tubes containing the freezing liquids and the junc- tions, but also included the whole of the couple, which was thus throughout its whole length brought approximately to the temperature of to Gabranometer the junctions. The rods (7 and 7’) supporting the arms of the couple (J and J’) were led through holes in the upper cork vertically over the mouths of the tubes, so that the fluids in the tubes could be kept agitated during an experiment. In order to equalize the effects of this agitation, provision was made for the connexion. of the two upper ends of the supporting rods by means of the screw-clamp (s). This clamp served as a handle which simultaneously stirred both tubes. ‘The small heat- ing effects of stirring which have been noted by other investigators were in this way cut out. It has been elsewhere noted! that Fie. 1, the shifting of the zero of such a galvanometer as must be used in these determinations is a cause of trouble, and, unless specially guarded against, will give rise to inaccuracy. ‘I'wo kinds of shifting of the zero may be distinguished: one which occurs in the galvanometer itself, even when the circuit is open, and the other which is noticeable only when the circuit is completed. The first is due to the fatigue of the metal-suspension, and may be observed whenever the galvanometer is 1 Trans. Roy. Irish Acad., vol. xxii, sect. B, pt. m1, p. 149. Dixon anp ArKins—On Osmotic Pressure in Plants, &e. 281 set up. It is a very slow change, and may amount to 2-3 mm. in a day. The second is only apparent when the circuit is closed, and seems to be due to the difference in temperature of the terminals of the galvanometer. When the galvanometer is carefully screened from rapid fluctuations in temperature, this second error is also insignificant during the time of an observation. In cases of excessive fluctuation of temperature it may amount to 10 mm. in the day. In the later observations it was, however, completely eliminated by the use of a reversing key in the circuit, by means of which the deflection due to a difference in temperature of the junctions might be instantaneously changed from side to side of the zero. The mean of the deflections will then give the true deflection, irrespective of a zero-error due to the difference of temperature of the terminals. It is evident that the reversing key used with this object must be of special construction, so that, it shall not itself introduce thermo-electric effects into the circuit. With this object a modification of a key, used already successfully with thermo-couples in the investigation of the temperature of plant organs, was adopted. This key consists of a wooden-spring clip, carrying the exposed ends of the leads coming from the galvanometer on the inside of its jaws. The leads from the couple terminate on each side of a glass support. When the clip is closed upon this support, the leads coming from the galvanometer lie at right angles across those coming from the couple. Reversing the clip from side to side of the support reverses the current. Even when the leads coming from the galvanometer and from the couple are made from the same piece of wire, differences of temperature on the opposite sides of the support and the clip will be found to give rise to thermo-electric effects in the circuit. To equalize the temperature of these parts, it was found convenient to immerse the clip and the support in a vessel of petroleum which was kept stirred during the observation. With this device the zero-error was satisfactorily eliminated, and no thermo-electrice error introduced. The calibration of the apparatus was carried out by one of two methods. At first the incompletely jacketed couple was used. While one junction was kept in melting ice, the other junction, along with a Beckmann thermometer, was placed in succession in solutions of sodium chloride of increasing concentration. The deflection registered by the spot of light, and the simultaneous reading of the thermometer at the freezing-point of the solution, were recorded. By a series of these observations a satisfactory calibration was effected. The following are the observations for the calibration of the thermo- couple No. 1 by this method. With the Beckmann thermometer used, the scale-reading 4:140 282 Scientific Proceedings, Royal Dublin Society. corresponded to the temperature of melting ice. When the two couples were at the same temperature, i.e., were close together in a vessel of water vigorously stirred, the cross-line on the spot of light stood at 250 mm. on the galvanometer scale, its middle point. Calibration. Scale Reading of | Scale Reading of | Deflection from Depression in | Galvanometer. Thermometer. | 0 in mm. Of, Sol. I. 310°7 3°571 60:7 0-569 Sol. IT. 362°5 3071 112°5 1:069 Sol. ITT. 483°6 | 1:905 233°6 2-235 By subtraction we get— (1) 112°5 mm. — 60°7 mm. = 51:8 mm. = 1:069° — 0:569° = 0:500° .. 1mm. = 0:00965° C. (2) 233°6 mm. — 60:7 mm. = 172:9 mm. = 2:235° — 0:569° = 1:666° a Lemime — 020 096421C% (3) 283°6 mm. — J12°5 mm. = 121:1 mm. = 223°5° — 1:069° -. 1mm. = 0:00920° C. 1:166° No. 5 gives a larger deflection per 0:01°C., inasmuch as for the greater deflections the spot of light is approaching the limit of the scale. This calibration agrees well with the calibrations by a different method given later on. Subsequently, when it was found desirable to surround the couple completely in the cooling-bath, the calibration was made by having distilled water in one of the test-tubes in the apparatus already described, and a solution of sugar having a known concentration in the other. From Raoult’s results the freezing-points of sugar solutions of various concentrations are known with great accuracy ; and consequently the value in degrees centigrade could be assigned to the deflection registered by the depression of freezing-point for any solution. In fig. 2 the line & shows Raoult’s results! for the depression of the freezing-point corresponding to different concentrations, the abscissae being the temperatures and the ordinates the number of grams of sugar per 1 Raoult’s results are quoted from H. J. Hamburger. (smotischer Druck und Ionenlehre. Dixon anp Arkins— On Osmotie Pressure in Plants, §e. 283 Ey aa js | A - te | 3 N| aan ois SHES | | L St ‘eG i ears | 5 : ‘ : it ‘ 7 ras | 14 ih | S| S 2 — 3 - a FL © I< N S | ck N | iS $ 8 s a aS {| —__}_ |. | 1 AY — X N me ‘ : 5 2 é LOTOM SO SULOLP Qoy fod LDBUS JO SULPLD SCIENT. PROC. R.D.S., VOL. XII., NO. XXV, 22 224 Scientific Proceedings, Royal Dublin Society. 100 grams of water. The curve in the figure marked J is that traced by the deflections of the galvanometer in connexion with couple No. 1, corresponding to the freezing-points of the same solutions. As before, the ordinates represent the concentration of the solution. But the abscissee for this curve represent the deflection of the spot of light in millimetres. The two curves lie close to one another, because each scale-division, viz. 1 mm., nearly corresponds to a deflection due to a difference of temperature of the junctions of 0:01°C. Curve JZ is the plotting of the readings given by a similar couple constructed as like the first as possible. Curve III gives the curve of couple No. 8, determined by means of the improved apparatus in which the couple is completely immersed in the cold chamber ; and the zero- error has been eliminated by the introduction of the petroleum reversing key. It will be seen that the accuracy of the individual observations is considerably increased, so that all the readings lie on a uniform curve. The general agreement of the three calibrations must be regarded as satisfactory. Extraction of the Sap. At the beginning it was feared that to obtain even 2°5 c.c. of sap from leaves of most plants would be a matter of trouble and difficulty. Our first method was to break up a weighed quantity of leaves to a nearly uniform pulp in a mortar, and then to add a measured quantity of water to the pulp. This water and pulp were intimately mixed together, and then as much of the fluid as possible was pressed from the pulp and used for the determination of the freezing-point. A separate observation gave the percentage of water in the uninjured leaves. Knowing this, and the weight of the pulp and added water, the dilution of the solutions caused by the addition of water to the pulp was calculated; and by suitable allowance the freezing-point of the more concentrated original sap was deduced. The dilution does not introduce more than a slight error. Our first seven observations were made by this method of extraction. They are recorded in Nos. 1-7 of the Table of Results. Later on, to our surprise, we found that the pressing of the sap from the leaves was quite an easy matter. To effect it, a few leaves, their midribs being removed, are taken, and crumpled up into a tight pellet. The pellet is wrapped in a double coat of fine lmen and placed between the jaws of a powerful vice. It was found convenient to have two stout silver discs, which were not appreciably attacked by the sap, between which the pellet was introduced into the vice. ‘There were marked differences in the behaviour of leaves under pressure. Some gave up their juice readily when first the jaws Drxon anp ArKtns—On Osmotic Pressure in Plants, &c. 285 pressed them; others required to be somewhat broken up before the sap would issue from them. Then it was necessary after the first crushing to open the vice, take out the crushed pellet, and loosen and crumple anew the compressed mass of ‘leaves within. On the second or third readjustment of this sort the sap usually flowed freely from the crushed leaves. In this manner sap was expressed more or less readily from all the leaves tried, with the following exceptions :— (1) The leaves of Zilia microphylla.—Only one sample of this tree was tried, but from it could be obtained no more sap than would just moisten the linen. (2) Shoots of Thuja plicata also yielded extremely little sap, which, owing to its very viscid nature, would not drop from the linen. (3) The buds of Syringa vulgaris, taken at the end of September, also refused to yield any appreciable amount of sap. At that stage we were unable to devote the time to see if the difficulties which prevented sap being obtained in these three cases could be surmounted ; and it remains to be ascertained whether, by more painstaking work, these, too, can be got to yield sufficient sap, or whether the list of organs which cannot be treated in this manner will have to be extended. When pressing the sap one cannot fail to be struck with the difference in colour of the sap issuing from various leaves; and one is astonished to find intense coloration in the pressed sap from leaves apparently of the purest green. For example, the green leaves of Syringa vulgaris, Fraxinus excelsior, F, oxyphylla, Catalpa bignonioides, Magnolia acuminata, all yield _a sap more or less intensely brown. In some cases~e.g. Syringa vulgaris—it is of such a deep colour that it appears almost opaque when held up to bright day-light in a small test-tube about 1 cm. in diameter. This coloration does not appear to be due to oxidation, as it is observed in the drops as they issue from the leaf. In some cases it intensifies on exposure to the air; but this change is a slow one. From other leaves the sap is of a greenish-grey hue—e.g. from the leaves of Vitis (Ampelopsis) Veitchii, Pteris aquilina. This coloured sap is also found in the green shoots of Hquisetum Telmateia. Pale-green sap is found in Wistaria sinensis, Eucalyptus globulus, Pinus Lavricio, Cordyline australis, Chamaerops humilis, Iris germanica. In the case of Vitis, Wistaria, and Eucalyptus, leaves which were evidently coloured by anthocyan yielded a sap tinged with pink. Helianthus muiltiflorus was remarkable in yielding an almost black sap. The sap of the roots examined was of a paler hue in the case of Syringa vulgaris than that of the leaves. The roots of Eucalyptus yieid an ochre-coloured sap. 27 2 286 Scientific Proceedings, Royal Dublin Society. Experimental. Having set up the galvanometer and scale, and connected the leads of the spring-clip to the terminals of the galvanometer, freshly boiled distilled water is introduced into one of the test-tubes in the frame, and about 3 ce. of the pressed sap into the other, and the junctions of the thermo-couple are immersed in them. ‘he freezing-jacket, consisting of ice and brine, is adjusted to a temperature 0:5°— 1:0° C. below the suspected freezing-point of the sap. The two test-tubes are cooled in a freezing-mixture, and a lining of ice is formed in that one which contains the distilled water. The junction stands in the liquid surrounded by a shell of ice. The sap is by this time probably cooled below its freezing-point, but will only begin to freeze after its “inoculation”? with an ice-crystal or the local application of intense cold. When by either of these means crystallization has been started, the _ frame carrying the test-tubes is immersed in the freezing-chamber; and connexion is made between the couple and galvanometer by setting the clip on the terminals of the former. During this time the junctions are kept stirred about in their test-tubes by moving the clamp holding the supports. Immediately on connexion being made, the spot of light moves from zero, and assumes a position due to the depression of the freezing-point of the sap below 0°C. The first observation made in this way usually gives too great a depression, because the sap has probably been supercooled; and consequently a great deal of ice separates out, leaving the remaining liquid rauch concentrated. This ice should now be very gradually melted. When the test-tubes are again introduced into the cooling-chamber, after a very small amount of supercooling, crystallization usually supervenes, as minute ice-crystals romain over after the first freezing, unless the sap has been warmed considerably above 0° 0. ‘This leads to a slight rise in temperature, which is recorded by a small movement of the spot of light towards the zero. When the spot of light has come to rest—in the meantime stirring is continued— a reading is made, and immediately by means of the clip the current is reversed. The spot of light now travels to the other side of the zero; and when it is again motionless, its position on the scale is recorded. The mean of the two deflections (by referring to the deflection curves and Raoult’s temperature curve given in fig. 2) affords the measure of the depression of the freezing-point. After some practice and experience reliable readings are obtained with certainty. In the following tabular statement of our results, the depression of the freezing-point of the sap is found in the column under A. ‘he osmotic pressure given under P has been calculated from the known relation between Drxon anp ArKins—On Osmotic Pressure in Plants, Siero it and the freezing-point. This relation has been determined experimentally and also deduced theoretically. Nernst gives the equation! :— A x 12:03 = P in atmospheres. In ‘column M is found the calculated mean molecular weight of the dissolved substances causing the depression of freezing-point. To obtain this if is necessary to ascertain the weight of the solvent and the dissolved substances in the solution. These data are obtained by filtering and weighing the sap after its freezing-point has been determined, by evaporating to dryness in a steam-oven and weighing the residue. J is then obtained by the formula? :~— In this s = weight of solute, and Z weight of solvent, viz. the weight of the sap minus the weight of the dissolved substances. Great accuracy is not to be expected from this method of determining the molecular weight ; but it is hoped that by its means an indication of some value has been obtained of the composition of the sap. Where the mean molecular weight is much above 200, it seems reasonable to suppose a disaccharide such as saccharose (mol. wt., 542) or maltose is a large con- stituent of the dissolved substances, while, when the molecular weight is found to be about 150, these bodies cannot be present in any large proportion. 1 W. Nernst, Theoretical Chemistry. English ‘I'ranslation of Fourth Edition. Macmillan, London, 1904, p. 144. * James Walker, Introduction to Physical Chemistry. Macmillan, London, 1899, p. 186. [Taste oF Ruesuirs 288 Scientific Proceedings, Royal Dublin Society. TABLE or ResuLts. ou Yj Bey nee Description of Sample. A 1p M : E 5 | ee 1 | Populus balsumifera leaves gathered 6 ft. level, 10 a.m., | Aug. 30, crushed, diluted and examined, 1-639 | 19°72 | 58-2 2 Catalpa bignonioides leayes gathered 10 ft. ‘evel, 4, 30 p.m. | | Aug. 80, crushed next day, diluted and examined, . | 1-905 | 22-92 | 72:3 3 | Catalpa speciosa leaves at 10 ft. level gathered 4. 30 p-m. | | Aug. 30, crushed next day, diluted and examined, 1-724 | 20°73 | | G12 4 | Same sap a after standing 24 hours in dark, | 1:748 | 20-96 61:2 5 Syringa vulgaris leaves at 4 ft. level gathered 10 a.m. 09 Aug. 81; crushed, diluted and examined Sept. 1. 2-234 | 26°87 | | 63°3 6 Magnolia acuminata leaves at 38 ft. level g gathered 11 a.m., | : crushed, diluted and examined 2.30 p.m., Sept. 1, 1-628 | 19°58 | 67:7 ra Magnolia nouminata leaves at 4 ft. level gathered 11 a.m, | crushed, diluted and examined 3.30 p.m., Sou il 1°858 | 22°34 | 69:1 8 Ribes grossularia fruits, first sample, 1-063 | 12°79 | | 9 », second sample, 1-120 | 18°47 | 10 Ribes rubrum (white var.) fruits, 1-410 | 16°96 11 Fraxinus excelsior leaves at 20 ft. level gathered a: nee | overcast morning (some rain, no sun day before), pressed | 12, and examined Sept. 2, . ‘ 5 . | 2:097 | 25:22 | 12 Fraxinus excelsior leaves at 8 ft. level, shaded, gathered, | and pressed as 11, : a - | 1:020 | 12-27 | 13 Fraxinus excelsior leaves at 43 ft. level gathered 8.30, pressed and examined noon Sept. 4. Littlesun day before, | 1°380 | 16°60 61:7 14 Fraxinus excelsior leaves at 2 ft. level, shaded, garnered | 8.30, pressed and examined noon Sept. 4, 1-000 | 12°03 68:0 15 Frasinus excelsion leaves 48 {t. level gathered 8. 30 a.m. op I | | pressed, and examined 3 p.m. Sept. 4, 1-094 | 18°16 | 16 | Fraxinus eacelsior leaves 2 ft. level gathered 8.30 a.m. mae | pressed and examined 3 p.m., Sept. 4 | 0-936 | 11-26 | 17 Fraxinus oxyphylla leaves 49 ft. level gather ed 8. 30 a.m. ; | pressed and examined 11 a.m., Sept. 11, overcast morning, | 9 hours’ sunshine on 10th, . | 1:280] lo-15 | 204 | 18 Fraxinus oxyphylla leaves 49 ft. level, same sample as in “| 17, left in diffuse light till beginning to wilt, | 2-003 | 24:09 | 228 19 Magnolia acwninata leaves 38 ft. level gathered 9 a.m., examined 11 a.m. Sept. 3. 2:7 hours’ sunshine on 2nd, 1:373 | 16°51 20 Magnolia acuminata leaves 4 ft. level garhered and ex- amined as 19, 7 0 1:142 | 13°74 | Ulmnus campestris (town specimen) : leaves. 21 Gathered noon, Sept. 6, bright, showery morning, from shaded short branches on top of arched branch at 18 ft. level, 0°888 | 10°68 152 22 Gatherednoon, Sept. 6, bright, showery morning, from shaded short shoots at base ‘of trunk at 1 ft. level, 5 0°763} 9:18] 148 | 23 Gathered noon, Sept. 6, bright, showery morning, from shaded short shoots at outer end of arch at 10 ft. level, 1-030} 12°39) 165 24 Gathered noon, Sept. 6, bright, showery morning, exposed short branches from trunk at 10 ft. level, 1550/1864) 155 | | Vitis (Ampelopsis) Veitchii (town specimen), W.exposure: leaves. 25 | Gathered 3.30 p.m., bright, 7-4 hours’ sunshine day before, young and old leaves at 1 ft. level, 0°S16 | 9°81 26 Gathered 3.30 p.m., bright, 7°4 hours’ sunshine day before, young leaves at 24 ft. "level, 0°653 | 7°85) 177 27 Gathered 3.30 p.m., bright, 7°4 hours’ sunshine day before, old leaves at 24 ft. level, 0°783 | 9°34 28 Gathered 3.30 p.m., bright, 7-4 hours’ sunshine day before, young leaves at 6 ft. level, 0°519 | 6:24) 154 Dixon anp Arkins— On Osmotic Pressure in Plants, &¢. TABLE OF ReEsuLrs—continued. 289 Expt Description of Sample. A Cerasus Laurocerasus leaves. 29 From 4 ft. level on south side, gathered 10 a.m., Sept. 10, sunny morning, 9 hours’ sun day before, 0:600 30 From 4 ft. level on north side, gathered 10 a.m., Sept. 10, sunny morning, 9 hours’ sun day before, 5 . | 0568 Syringa vulgaris (country specimen), 8.W. exposure: leaves. 31 Gathered 6.30 p.m., Sept. 8, after showery, bright day, 8 hours’ sunshine, examined llam., Sept.9, ~ 1580 32 Gathered 5 a.m., Sept. 9, after fine night, examined 11. 30, Sept. 9, 1:344 33 Gathered 6 p.m., Sept. ®), after dark day, no sunshine, | examined 11 a.m., Sept. 1 1-352 34 Gathered 6 p.m., Sept. 9, aie dark day, no sunshine, exposed to diffuse light till beginning to wilt, then examined, 2-002 35 Gathered 6 p.m., Sept. ®; “after dark day, no sunshine, exposed to diffuse light, supplied with water, 1:586 36 Gathered 6.30 p.m., Sept. 10, after bright, breezy aay, 9 hours’ sunshine, examined 1 a.m., Sept. iil, 1°862 37 Shaded leaves in dark bag from 3 p.m., Sept. 11, till 10 am., Sept. 13, then gathered, examined noon, Sept. 13, 0°962 38 Exposed leaves gathered 10 a.m., examined lla.m. Sept. 13, control of 37, some sunshine each day, 1-423 39 Same as previous sample, exposed with petioles in w ater, 12-1.30 p.m., to sun and wind, . 2-135 40 Same as previous sample, kept with petioles in water in the dark, noon Sept. 13—noon, Sept. 15, 1/188 41 Gathered 6.30 p.m. after bright day, 7 hours? sunshine, pressed 9.30 p.m., Sept. 13, examined 11 a.m., Sept. 14, | 1-696 42 Gathered 6.30 p.m. after bright day, 7 hours’ sunshine, pressed 11.30 a.m., examined noon Sept. 14, | 1°810 Syringa vulgaris (country specimen), E. exposure: leaves. 43 Gathered 10.30 a.m. after sunny morning; and examined same day 11.30 a.m., Sept. 14, 1:947 44 Gathered 10.30 a.m., sunny morning, in bag ‘on tree Sept. 15-17 [5:4 + 7°5 hours’ sunshine], examined noon, Sept. 17, 1:798 45 Gathered 10.30 a. m., control of previous, not enclosed in bag, otherwise similarly treated, 27048 46 Gathered 10 a.m., sunny morning, in bag on tree, Sept. 15-18 (544+ 7547 “6 hours’ euehine ls examined 11.30 a.m., Sept. 18, 1582 47 Coal of previous, “not enclosed in bag, otherwise treated similarly, 1°663 48 In_ bag, 10 a.m. 18th, gathered 10.30 a. m., Sept. 24 [65 + 34+ 7:44 0- 7+0°6 + 4-2 hours’ sun], examined ll a.m., Sept. 24, 1-263 49 Control of previous, similarly treated but not i in bag, 1-470 50 In bag 10 a.m. 15th, gathered 10.30 a.m. Sept. 2, dull morning, examined i, 30 am., . 1:328 51 Control of previous, similarly treated but not in bag, 1-589 52 Exposed leaves gathered wet morning, 10.30 a.m., Bert: 27, examined 11.30 a.m., 1-440 ad Ton lor) ano a oOo wa 17°32 M | Per cent. of water in sample. 69°16 63°5 71s 66° a bow 290 Scientific Proceedings, Royal Dublin Society. Taste or Resutrs—continued. No.of 258 Expt Description of Sample. A Pp M og g | ire Syringa vulgaris (country specimen), E. exposure: leaves. 53 Shaded leaves gathered 10.30 a.m., wet morning, exam- ined noon Sept. 27, 25th, 2:9 hours? » 26th, 0 hours’ sun, . | 1°171 | 14:09) 172 54 Part of previous sample kepti in close jar in dark, examined 11.30 a.m., Sept. 30, . 1-080 | 13:00] 179 55 In bag 18th, gathered 10.30 a.m., ‘Sept. 30, fine morning after 2 days? wet, 1:010| 12:15 | 136 56 Control of previous, similarly treated but not in bag, . | 1°608 | 19:34] 250 al Part of sample 55 kept in dark in closed jar 24 hours and examined, 1°157 | 13°92 | 217 58 In bag Sept. 18, gathered 10. 30 a.m., Oct. 9, 21 days in bag, ” examined 11.30 a.m., Oct. 9, 0-963 | 11°58 | 249 59 Control of previous, similarly treated but not in bag, some- what loose and inclined to fall, but still green, . 1504 | 18-10] 256 60 Exposed leaves, yellowish, ready to fall, parliored wet morning, 10 a.m., Oct. 28, 1:215 | 14°61 61 Exposed leaves, green, fairly firmly attached, 1:345 | 16°18 Syringa vulgaris (town specimen), S. exposure: leaves. 62 Gathered 2.45 p.m., morning uy, Sept. 14, examined immediately, branch (a), 1-306 | 15-79 63 Gathered 2.45 p.m., morning sunny, Sept. 14, examined immediately, branch (B), 1:315 | 15°81 64 Gathered 2.45 p.m., morning sunny, Sept. 14, examined immediately, branch (7), - 1:310 | 15:76 65 Halves of the leaves of branch (y /) kept in closed jar in dark, examined next day, 1:433 | 17:24] 193 66 Gathered 2.30 p-m., Oct. 25, after sunny morning, examined immediately, 1-215 | 14:61 | 246 | 71°8 67 Halves of leaves used in previous, pressed immediately, sap kept and examined next day, 1°220 | 14°67) 231 68 Halves of leaves used in 66, kept in dark till next day, then pressed and examined, 1:227) 14:76 | 252 | 73-7 69 Gathered 2.80 p.m., morning dark, Oct. 26, pressed and filtered immediately, sap kept and examined next day, . | 1-445 | 17°38 70 Gathered 2.30 p.m., morning dark, Ovt. 26, pressed and not filtered immediately, sap kept and examined next day, . | 1-456 | 17-52 71 Gathered 10.30a.m., Oct. 28, half leaves pressed and exam- ined immediately, 1-424] 17:18) 211 72 Gathered 10.30 a.m., Oct. 28, half leaves kept 24 hours in dark., pressed and onamincal Oct. 29, . 1-424 17:13) 217 73 Gathered 10.30 a.m., Oct. 28, half leaves pressed ard filtered immediately, sap kept and examined next day, . 1540 | 18°52] 210 74 Gathered 10.30 a.m., Oct. 28, half leaves pressed, not filtered. Sap kept 24 hrs., “examined Oct. 29, 1:505 | 18°10} 216 Syringa vulgaris (country specimen) : roots. 75 Gathered 10 a.m., Sept. 25, examined noon same day, - |0°460] 5:58! 178 | 78:0 76 Part of last pamiple) kept in closed jar in dark 2 days examined noon Sept. 27, . 0-456) 5:48) 167 Ud Gathered 10 a.m., wet morning, examined L p-m., Oct. 23,.|0°357| 4:29) 254 78 Gathered 2.30 p.m., examined immediately Oct. "98, 0-490} 5:90] 208 79 Gathered 9 a.m., examined 11 a.m., Oct. 80, 0-425] 5:11) 184 TaBLE OF ReEsunrs—continued. Dixon anp Arxins—On Osmotic Pressure in Plants, Sc. 9) a 91 53 Bn Description of Sample. A 12 M 8 g e Be. Wistaria sinensis (country specimen): leaves. 80 From short basal shoot 3 ft. level, gathered 9.30, sunny morning, examined 11.30 a.m. , Sept. 8, 0°412| 4:95] 149 St From distal end of horizontal branch 65 ft. long, sft. level, gathered 9.30 a.m., examined 11.30 a.m., Sept. 8, 0-437 | 5:25) 169 82 From 27 ft. level, exposed, gathered 9.30, dull morning, examined 11.30 a.m., Sept. 9, 8 hours sun, Sept. 8, . |0°550) 6:61} 162 83 From 3 ft. level, shaded, gathered 9.30, dull morning, examined 11.30 a.m., Sept. 9, 8 hours’ sun, Sept. 8, 0°443 | 5°53] 169 84 Part of sample 82, kept in dark without water supply and not shielded from evaporation, examined 1la.m., Sept. 10, when some leayes were showing signs of wilting, - | 07946) 11:38] 194 Pinus Laricio: leaves. 85 From current year’s growth at 14 ft. level, gathered 10 a.m., examined noon, Oct. 9, 07910} 10°95) 115 86 From last year’s growth at 14 ft. level, gathered 10a. m., examined noon, Oct. 9, < 5 . | 1:045 | 12-57) 192 Eucalyptus globulus. 87 Buds and small leaves up to 3 cm. long, gathered Oct. 5 from tree A, bright showery morning, examined noon, 0-703} 8-45} 199 88 Young stems from last sample, 0:702| 8:44} 120 89 Leaves about 10 cm. long from same branches as in two foregoing, 0°673} 8:09} 198 90 Leaves about 15- 20 cm. long from same branches as in three foregoing, 0°683'| 8°21} 208 91 Mature horizontal leaves gathered from tree B, 10 a.m. examined 11.380 a.m., Oct. 7, 0:507} 6:10} 200 92 Mature vertical leaves gathered from tree B, 10 Aig examined 11.30 a.m., “Oct. 7, off same branch as 91, |0°526| 6°33] 169 93 Mature horizontal leaves of previous year, reddish ; treated as in last, off same branch as 91, 0 . | 0°666] 8-01] 280 94 Roots gathered 10 a.m., wet morning, examined 12.45 p.m, Oct. 23, : 0°433 | 5°33 MIscELLANEOUS DETERMINATIONS. 95 Equisetum Telmateia, green lateral branches, gathered 9.30 a.m., examined 11.30 a.m., Sept. 12, 9 hours’ sunshine day before, 0:946 | 11°38] 122 96 Pieris aquilina, pinnae, overshadowed, gathered 10. 30a. m., examined 4.30 p.m., Sept. 23, . 0619) 7544) — 97 Helianthus multifiorus in sunshine, not insolated in mor ning gathered 12.30 p.m., examined immediately, Sept. 15, 0605} 7:28) 169 98 Helianthus multiflorus in shade, insolated earlier, 12.30 p.m. examined immediately, Sept. 15, 0°649 | 7:63} 160 99 Cordyline australis, top leaves 13 ft. high, gathered 10.30 a.m., sunny, examined 3.50 p.m., Sept. 2 0°543 | 6:53 | — 100 Chamaerops humilis, 6 ft. level, gathered To a.m., dull, examined 11.30 a.m., Sept. 24, 0°315| 3:79) 323 101 Tris germanica, gathered 2.30 p.m. » bright day, examined 3 p.m., Oct. 3, . 0:600| 7:22] 138 BA SCIENT. PROC. R.D.S., VOL. XII., NO, XXV, 292 Setentifie Proceedings, Royal Dublin Society. Discussion of Results. At the outset it was necessary to find out if there really existed any regularity in the osmotic pressures of leaves, and to examine the possibility that the exact pressures were of no consequence, and that measurements of them might show them to vary in quite an erratic manner. Experiments 62, 63, 64, 65 were made with a view to test this possibility. Tase II. Syringa vulgaris : Leaves. No. of apse Expt. | Description. A. 1D, (i) INE 62 Gathered 2.45 after sunny morning, Sept. 14, examined imme- diately, branch (a), : 1:306 | 15°70 | — 63 Gathered 2.45 after sunny morning, Sept. 14, examined imme- diately, branch (8), - : : A : . | 173815] 15°81) — 64 Gathered 2.45 after sunny morning, Sept. 14, examined imme- diately, branch (¥), 6 : 1°310 | 15-76 | — 65 Halves of leaves on branch (y) kept in closed jar in dark, examined next day, 5 5 ‘i e o |} 19448383 |) Uzfomde || iB} Three samples of leaves were gathered from three similar adjacent branches, which, so far as could be seen, were, and had been previously, under precisely similar conditions of transpiration, assimilation, &c. The freezing- point of the sap pressed from these three samples was observed ; and the three readings in the first column show how closely the concentrations of the sap in the leaves of the three branches agree. ‘his result is what would d priori be expected ; and the concordance of the experiments may be taken as a measure of the accuracy of the results obtained by this method. It is to be noted that these figures were obtained by the first arrangement before the complete jacketing and the reversing-key were introduced. It appears that, even with the less perfect arrangement, we may hope that the errors do not amount to more than 1 per cent. A series of experiments was designed to test the possibility that the osmotic pressure of the sap of the leaves at any region in the branches is defined by the resistance which has to be overcome in drawing the trans- piration current from the roots to that part. wart! had previously looked for such a difference by means of the plasmolytic method, but seems to have LA. J. Ewart, Ascent of Water in Trees. Phil. Trans. Roy. Soc., B., vol., excvill. (1905) and idem, vol. excix. (1908). Dixon and ArKins—On Osmotic Pressure in Plants, Se. 293 encountered difficulties and left the question undecided. In our experiments sap was pressed from leaves taken at a considerable height above the ground ; and its freezing-point was compared with that of leaves from near the ground- level. The following pairs of experiments bear on this question. Tasxe III. | Tee Description of Sample. A. P. i | | 6 } Magnolia acuminata, leaves from 38 ft. level, | 1°628 19°58 | v$ 5) 0 5 ay ANE 55 é : - | 1858 | 22°34 Pate » » 99 9p BBE pp : : 6 1:373 | 16:51 20 | » » dp. || bg) Seat Bp : : : 1142 | 18-74 11 Fraxinus eacelsior, leaves from 20 ft. level, 0 : > |) BOY 25°22 12 | 99 3 shaded leaves from 3 ft. level, : : | 1-020 12-27 13 35 33 CEO! 55 py GBI, 5p % 5 1°380 16°60 14 | 59 90 Shaded se sueneammorttsnmnrs 1-000 | 12-03 15 99 te exposed ,, ,, 43ft. ,, ‘ ; 1-094 13°16 16 x sp MEL RK op : _ | 0-936 | 11-26 25 Vitis Veitchii, leaves from 1 ft. level, “ 0 5 0°816 9°81 26 » ” 99 » 24 ft. ,, c G 0 r 0°653 7°85 27 99 9% ety ee edettsn issr .e 5 : . | 0-783 9°34 28 - as Pee GHEE A. foes Hr ‘ F . | 0-519 6-24 It here appears that, on the whole, taking the experiments in pairs, the leaves at the lower level contained sap with a lower (sometimes considerably lower) osmotic pressure than that of higher leaves. But the experiments are far from satisfactorily bearing out this view; for it will be noted that the osmotic pressure of the sap from leaves at the same level, but at different times and under different conditions, by no means corresponds in each case, although it is often higher than that of leaves at a lower level. ‘I'he reverse, however, is sometimes found, as in expt. 6 and 7, where the pressure in the lower is much greater than in the higher leaves. The possibility that these discrepancies might be due to resistance in the conducting tracts apart from that offered by the hydrostatic head had to be examined, and experiments 80, 81, 82, 83 on Wistaria sinensis and 21, 22, 23, 24 on Ulmus campestris were carried out. Baz 294 Scientific Proceedings, Royal Dublin Society. Taste LV. Wistaria sinensis : leaves. Expt Description of Sample. A. 1, M. 80 | Shaded from 8 ft. level on basal shoot, i x . | 0:°412] 4:95 | 149 81 ) Exposed from 3 ft. level at distal end of horizontal branch 65 feet long, 0 ¢ 0 ' : - | 0-437) 5:25 | 169 82 | Exposed from 27 ft. level, 3 . é 3 - |0°550) 6°61 | 162 83 }| Shaded from 3 ft. level, : =. . [0-443] 5-53 | 169 Hxperiments 80 and 81 were made on sap from the leaves of an old Wistaria trained on a low wall. One sample of leaves was gathered from short branches near the base of the main stem. The leaves were about three feet over the ground. The second sample of leaves was taken from the terminal branches of a stem running 65 feet approximately horizontally along the wall at a level of about 3 feet. Here again we find a slight difference in pressure in favour of the distal leaves. Tasie V. Uimus campestris: leaves. se Description of Sample. A. P. M. 21 From short shoots on tp of arched branch in shady position at . 18 ft. level, . ° ° 6 0:888 | 10°68 152 22 From short shoots at base of trunk in anady Pore at 1 ft. | - level, : ; ; : 0°763 | 9°18) 148 23 From short shoots at outer end of arched branch in shady position at 10 ft. level, 0 7 6 3 - | 1:030 | 12°39} 165 | 24 From short shoots on trunk in sunny position at 10 ft. level, . | 1°550 18°64} 155 | The above experiments show the real meaning of the results which apparently go to show that the level is the controlling factor in determining the osmotic pressure. If the hydrostatic head defined the pressure of the leaves, it is evident that the pressure of expt. 21 should be the greatest; if the resistance of the water-tracts were the controlling factor, expt. 23 should have the maximum pressure, which should be much greater than 22 and 24. The actual order is 24, 28, 21, 22. From this it is clear that the resistance Dixon anp Arkins—On Osmotic Pressure in Piants, Se. 295 of the water-tracts was not the controlling factor of the pressure; accordingly some other cause for its variation must be sought. ‘I'his cause seems to be principally the fluctuations in the sugar-content of the leaves due to difference in illumination. In expt. 21 and 22, the leaves of which are from shaded positions, smaller pressures are found than in expt. 24, which was performed on sap from leaves in a sunny position, while expt. 23 on leaves coming from the outside of the crown facing a clear north sky, and being consequently better illuminated than the other two samples, 21 and 22, has a higher pressure than they, though considerably lower than in 24. The effect of exposure to the sunlight on the osmotic pressure is illustrated by a great number of experiments. Nos. 2 and 30 show the difference between the pressure in leaves on the north and south side ofa shrub of Cerasus Laurocerasus. Taste VI. Cerasus Laurocerasus : leaves. ‘T | Tae Description of Sample. ey || Bs |) 29 Gathered from south side on sunny morning, 9 hours’ sunshine on previous day, : C 5 : , . |0°600| 7:22 | 256 30 Gathered from north side on sunny morning, 9 hours’ sunshine on previous day, . 6 : 0 - |0°568 | 6°85 | 250 Possibly here the high mean molecular weight may in part be due to the presence of small quantities of the glucoside amygdalin. Aiter a sunny day the pressure of leaves rises. It falls at night, as may be seen below in the case of Syringa vulgaris. | Tass VII. “Syringa vulgaris : leaves. ee | Description of Sample. A. | IP | M. ] - 31 Gathered at 6°30 p.m. after showery, bright day, : . |1:580| 19-01] 225 | Gathered at 5 a.m next morning, ; : . | 1:844) 16°17] 214 Both samples of leaves were taken from the same position on the same shrub, and so, if they had been taken at the same time, should have had equal pressures, as was shown earlier. 296 Scientific Proceedings, Royal Dublin Society. Similarly the osmotie pressure in leaves gathered after a dull day is lower than after a bright one, as appears from the following observations. Tasie VIII. Syringa vulgaris: leaves. | No. of Bung Expt. Description of Sample. A. P. M. 33 Gathered after dark day, no sunshine, . 5 5 . | 1:352] 16°26 |- 206 86 ae », bright day, 9 hours’ sunshine, . 0 . | 1°862 | 22°40} 227 41 y p ro Se OUTS aera 4 O - |1°696 | 20°40) — As might be expected from the foregoing, and as has already been seen in the case of Ulmus, leaves taken from a position naturally overshadowed are found to have a lower pressure than those taken from an exposed position. Tasie IX. Syringa vulgaris : leaves. |W ne Description of Sample. A. 12 M. | 53 Shaded, gathered wet morning after day with no sunshine, . | 1:171/14:09| 172 52 Exposed, gathered at same time, 0 0 0 . | 1:°440) 17°32] 239 The effect of assimilation may also be easily observed, by comparing the freezing-points of the sap of exposed leaves with that of leaves in a similar position, but shaded by an opaque cover. We obtained many results illustrating this. In the following table the leaves giving the (a) samples were surrounded by an opaque paper bag at 10 a.m., September 15; at intervals samples were taken out and examined. ‘The ((3) samples were furnished by neighbouring leaves under similar conditions, but not covered. Drxon anp AtTKiIns—On Osmotic Pressure in Plants, Sc. 297 TABLE X. Syringa vulgaris: leaves. Ne i | Water me Description of Sample. Do P. M. per cent. 44 (a) Covered 2 days, gathered sunny morning, Sept. 17, sun- | | shine on previous days, 5-4 hours and 7:5 hours, . | 1:798 | 21°63 | 244 | 59:9 | 45 (8) Exposed 2 days, gathered sunny morning, Sept. 17, sun- | shine on previous days, 5°4 hours and 7°5 hours, . | 2°048 | 24°57 | 259 | 61:0 | 46 (a) Covered 3 days, gathered sunny morning, Sept. 18, sun- | shine on 17th, 7°6 hours, . 5 : . | 1-582; 19°08 | 273 | 63-5 47 (8) Exposed 3 days, gathered sunny morning. S Sept: 18, sun- shine on 17th, 7-6 hours, 1-663 |,20:00} 234 | 65-1 50 (a) Covered 7 days, gathered dull morning, Sept. 22, sun- shine on previous days, 6°5 hours, 3°4 hours, 7-4 hours, and 0-7 hours, ‘ : 3 . | 1328 | 15:97] 253) — | 51 (8) Exposed 7 days, gathered dull morning, Sept. 22, sun- shine on previous days, 6°5 hours, 34 Jone v4 hours, and 0°7 hours, : 5 . | 1°589 | 19:12) 234) — During the period of this series of observations, the pressure of the exposed leaves fell from 24:5 to 19:0 atmospheres, while the leaves in the bag were reduced in pressure by 3 atm. below the exposed ones, and this difference is recorded at the end of the series. Strangely enough, in the two cases observed, the percentage of water in the exposed leaves is greater than that in the covered leaves. This fact may be explained by the greater distension of the cells of the leaves which contained larger amounts of dissolved substances. ‘This naturally raises the proportion of the liquid to the solid constituents of the leaves, the latter remaining approximately constant during exposure, except for the formationjof starch. The high mean molecular weight of the solutes maintained in the shaded samples is surprising. The highest mean molecular weight in the series is recorded for a sample which had been cut off from the light for three days. ‘This seems probably due to the hydrolysis of starch and the formation of maltose, also possibly to the transport of this sugar, or to the formation of the salts of organic acids during respiration, 298 Scientific Proceedings, Royal Dublin Society. A similar series is exhibited in the following table :-—— TasLe XI. Syringa vulgaris : leaves. No. of Water Sa Description of Sample. is P. M. | per anne cent. 48 (a) Covered 6 days, gathered dull morning, Sept. 24, pre- vious days’ sunshine, 6°4 hours, 3:4 hours, 7:4 hours, 0:7 hours, 0°6 hours, and 4-2 hours, 5 jf Seas 1F WeSeRA) WRI) | 7ile® 49 (8) Exposed 6 days, gathered dull morning, Sept. 24, pre- vious days’ sunshine, 6:4 hours, é*4 hours, 7°4 hours, 0:7 hours, 0°6 hours, and 4:2 hours: 3 . | 1:470'| 17:68] 226 | 66-2 or on (a) Covered 12 days, gathered pet, 30, bright morning. previous 2 days wet, - | 1:010) 12°15) 156 | — 56 (8) Exposed 12 days, gathered Sept. 30, pene z morning eos 2 days wet, . - | 1°608 | 19°34) 250 | — 58 (a) Covered 21 days, gathered Oct. 9, bright morning, pre- vious days’ sunshine, 1-9 hours, 44 hours, 5°5 hours, 7°3 hours, 3°8 hours, and 5°2 hours, : - |0°953] 11-58) 249 | — 59 (8) Exposed 21 days, gathered Oct. 9, bright morning, pre- vious days’ sunshine, 1-9 hours, 44 hours, 5:5 hours, 7:3 hours, 8°8 hours, and 5:2 hours, 5 . | 1-505) 18-10) 256 | — In the first case the shading reduced the osmotic pressure by 2°5 atmospheres, while the mean molecular weight remained sensibly the same at the end of six days. If monosaccharides were used principally in respiration, no fall in the mean molecular weight should be expected to accompany the fall in osmotic pressure. The second pair of observations offer in this respect asharp contrast. During two days of continuous rain the osmotic pressure of the darkened sample has been greatly reduced, probably in a small degree owing to dilution, and principally owing to the fact that little or no carbo- hydrates have been brought into solution, or have been translocated from the exposed leaves. ‘The failure of the supply from the exposed leaves is to be attributed to the unfavourable conditions for carbon assimilation on the two days previous to the determination, which probably did not permit more carbohydrates to be formed than were utilized in the leaves where they were produced. The startling reduction in the mean molecular weight of the solutes in the shaded leaves is also to be explained by the exhaustion of the dissolved carbohydrates in those leaves, and their consequent percentage reduction in the solutes. The ((3) sample, on the other hand, has a high osmotic pressure and mean molecular weight, because the assimilation on the bright morning when the leaves were gathered has more than compensated for the reduction of assimilation of the two previous days, Dixon anp Arkins—On Osmotic Pressure in Plants, §e. 299 From the third pair of observations it appears that translocation, or some other process, increased the percentage of disaccharides, or possibly salts of organic acids, among the solutes to more than restore the original mean molecular weight, while the actual concentration of the solution has further diminished, so that the comparatively low osmotic pressure of 11:58 atmospheres is attained. These experiments are interesting, showing that the external conditions can bring about a great change in the osmotic pressure of leaves, viz., from 24:57 to 11:58 atmospheres. TaBLe XII. Syringa vulgaris: leaves. No. of ees Expt. Description of Sample. A | P M. = SNES LS so 2 zs ese eee eect aay 37 Shaded leaves in dark bag, 3 p.m., Pept 11, till 10a.m., pert: 13, | some sunshine each day, . - | 0°962] 11°57} 160 38 Exposed leaves in sunny position, gathered Sept. 13, . - | 1-428) 17-12] 202 | In experiments 37 and 88 a large difference is shown, both in the osmotic pressure and in the mean molecular weight; but the sample of leaves enclosed in the opaque bag grew in an overshadowed position, attached to branches rising from the base of the stem. Consequently these had a low osmotic pressure to start with, and possibly also a low mean molecular weight. In any case the isolation of the leaves in question would greatly delay the transport of substances with high molecular weight which could make up for the loss of carbohydrates by respiration. It is curious that the two lowest determinations of the osmotic pressure 37 and 55, viz., 11:57 and 11°58 atmospheres respectively—one obtained by short covering of normally overshadowed leaves, and the other by a prolonged covering of normally exposed leaves—should so closely correspond. Possibly we have here an irreducible minimum of healthy Syringa leaves. It has also been easy to show the marked effect of assimilation on the osmotic pressure in detached leaves. For this experiment leaves gathered from the same region of a tree were divided into three samples—a, 6, and ec. Sample a was examined immediately. ‘The leaves of 6 were set with their petioles in water, and exposed to conditions favouring assimilation, viz., sunshine and breeze, for an hour and a half, and were then pressed and examined. Sample c was kept for two days in the dark, the petioles being in water. SCIENT, PROG. R,D.S. VOL. XI., NO. XXV, oB 300 Scientific Proceedings, Royal Dublin Society. Taste XIII, Syringa vulgaris: leaves. No. of é Water Oe © Description of Sample. A. 1D. M. | per Expt cent. 38 | (a) Gathered, and examined immediately, : - | 1:423] 17:12] 202 | 69-16 39 (2) Gathered, supplied with water, exposed to sun and breeze for 14 hours, then examined, : . | 2°185 | 25°68] 202 | 63°50 40 (c) Gathered, supplied with water, pert in dark for two days, then examined, : » | 1183) 14:23) 211) — Here the rise, owing to insolation, is very marked, viz., from 17:12 to 25°68 atmospheres. Some of this, however, must be attributed to concentra- tion by loss of water, as, at the end of the experiment, some of the leaves had begun to wilt. Desiccation showed that this concentration could by no means account for the total rise in osmotic pressure. ‘he observed change, viz., from 69:16 per cent. to 63°5 per cent. of water, would account for a change of depression of freezing-point from -1:423°C. to -1:551°C., or a raising of the osmotic pressure from 17:12 to 18-66 atmospheres, leaving the remainder to be accounted for by the formation of carbohydrates. Experiments 33, 34, 85 form another series, illustrating the effects of assimilation and transpiration on the osmotic pressure. The leaves in this case were gathered at 6 p.m., after a dull day, and kept in dark till 11 a.m. next day. They were then divided into three lots—a, >, and ¢. ‘The sap of a was examined immediately as a control; 6 was exposed to diffuse light for four hours without water-supply ; ¢ was exposed to diffuse light for four hours, but the petioles were supplied with water. At the end of the four hours the saps were examined. Taste XIV. Syringa vulgaris: leaves. No. of wea : > Expt. Description of Sample. A. 1P, M. | = 2 tree a a i} 33 (a) Control, - . : : $ : . | 1:352| 16-26 | 206 34 (2) Exposed to light, without water-supply, for four hours, | 2:002 | 24-07 | 199 | 35 (c) * with VBR aaah FS ( . | 1-586 | 19-08 | 2¢1 Dixon anp Arkins—On Osmotic Pressure in Plants, §e. 301 At the end of the experiment, some of the leaves in sample 6 had begun to wilt. It is evident that in this case the concentration due to loss of water was responsible for the greater part, but not all, of the rise in osmotic pressure. In the figures given above (see Table VII.) experiments Nos. 31 and 82, leaves gathered in the evening were set aside, kept from light, and pro- tected from evaporation, and were then compared with others gathered next morning. On examination the evening sample gave a greater depression of freezing-point. It is evident that some or all of this increase might be due to some change taking place during the night in the gathered leaves. Experiments 41 and 42 bear upon this point. In these it is shown that the keeping, although introducing a change, will not account for the total difference observed. The leaves gathered, after seven hours’ insolation, were divided into two lots—a and $6. From a the sap was pressed immediately, and kept over till next morning for investigation. Lot b was set aside uninjured, protected from light and evaporation till next morning. Its sap was then pressed and inyestigated. The figures are repeated here. TasLte XY. Syringa vulgaris: leaves. l No. of ar Expt. Description of Sample. A. P. 41 (a) Gathered 6.30 p.m. after seven hours’ sunshine, rap pressed 9.30 | p-m. Sept. 13, examined 11 a.m. Sept. 14, - | 1-696 | 20-40 42 (4) Gathered 6.30 p.m. after seven hours’ sunshine, Sept. 13, and stored, sap pressed 11.30 a.m., and examined noon, Sept. 14, . - | 1:810) 21-77 During the night changes have gone on in the sap of the intact leaves which did not progress so rapidly in the expressed sap. Observations 64 and 65 afford a similar instance. In these experiments the leaves were gathered on the afternoon of the 14th. Of these two lots, a and 6, were formed by dividing each leaf down the midrib, which was always rejected. The sap of lot a@ was extracted and examined immediately. Lot 4 was kept over till the afternoon of the 15th. Its sap was then expressed and examined. ‘I'he results substantiate the former observation, and, as the difference in both cases is similar, indicate that considerable changes in osmotic pressure did not occur in the pressed sap of No. 41. 3B2 302 Scientifie Proceedings, Royal Dublin Society. TasLe XVI. Syringa vulgaris: leaves. Noiof Description of Sample. A. P. Expt P Y 64 (a2) Gathered 2.45 p.m. after eunny, morning, Sept. 14, examined immediately, y 5 ‘4 : 1°310 15°76 65 (8) Gathered 2.45 p.m. after sunny morning, Sept. 14, kept in dark, examined Sept. 15, . : : é 1:433 17-24 Here two reasons may be adduced as principally responsible for the rise in pressure, viz., (1) the hydrolysis of starch and the production of maltose; and (2) the inversion of cane-sugar into dextrose and levulose. Unfortunately no determination of the mean molecular weight of the (a) sample was made, and consequently we are unable to decide between these two alternatives. A similar rise in pressure on keeping has also been observed in leaves which have not been exposed to light for twelve days. Taste XVII. Syringa vulyaris: leaves twelve days in dark bag. tet Description of Sample. A. iB M. 55 (a) Gathered, pressed and examined Sept. 30, . 0 . | 1:010 | 12°15} 156 57 (8) Gathered Sept. 30, pressed and examined Oct.1, . | 1-157 | 13:92'| 217 i In this case the rise of the mean molecular weight indicates that the rise in the osmotic pressure is due to the formation in the leaves of some disaccharide, probably maltose from the hydrolysis of starch. Observations 53 and 54, Table XVIII., favour this view. These were made on leaves taken from an overshadowed position, and which consequently had not the opportunity of storing appreciable quantities of carbohydrates. As was expected, the overshadowed sample did not give the usual rise in pressure on keeping. Presumably they contained neither starch to hydrolyse nor cane- sugar to invert. The consumption of dextrose in respiration was probably accountable for the slight fall in pressure observed. Dixon anp Arxins—On Osmotic Pressure in Plants, §c. 308 Tass XVIII. Syringa vulgaris:: leaves. ee Description of Sample. A 12 M. i | 53 Overshadowed, gathered 10.30 wet morning, examined noon | same day, Sept. 27, | 1-171) 14-09} 172 | 54 | Part of same sample, eet in dark in jar, examined 11.50 a.m., | | Sent 80, ; : : : - | 1-080] 13-00] 179 | Quite comparable with the observations on the overshadowed leaves were the results obtained with roots :— Taste XIX. Syringa vulgaris: roots. ee Description of Sample. A. IP M. 75 Gathered 10 a.m. Sept. 25, examined same day, 0 - | 0°460| 5°53 | 178 76 Gathered 10 a.m. Sept. 25, Bent in closed j bjar in dark, examined Sept. 27, 9 0-456 | 5°48 167 Unfortunately in both these cases (viz., experiments 53, 54, and 75, 76) the test was not made at the end of the first day; and the possibility remains that, if it had been carried out then, a rise might have been registered, as in the other cases tested at.the end of the first day, instead of the fall actually observed at the end of the second and third days. Later on in the season —in point of fact, towards the latter end of October— we attempted to return to this problem, and we endeavoured to ascertain if the rise of osmotic pressure took place to any extent in the sap after having been extracted from the cells, and, if so, whether it would take place in the sap freed from suspended matter by filtration. The results obtained on this reinvestigation showed, as a rule, much smaller differences produced on keeping the leaves and the sap, whether filtered or unfiltered. The more stable condition in these cases may be attributed to the different condition of 304 Scientific Proceedings, Royal Dublin Society. the leaves later in the season, or possibly the lower temperature at which the gathered leaves and saps were kept. These later results are summarized in the following tables :— TABLE XX. Syringa vulgaris: leaves. No. of Samet a | | Expt. Description of Sample. A. P. M. 66 Gathered 2.30 p.m. after sunny morning, Oct.25, examined imme- diately, 9 9 0 ° ° 0 - | 1°216 | 14°61 246 67 | Halves of leaves used in No. 66, pressed immediately, pep aoe | and examined next day, 0 : 1°220 |) 14°67} 231 68 | Halves of leaves used in No. 66 neu in dark, pressed and examined next day, . 5 A (i227 el As Gil eo 2) Here the change in pressure is very small; but the sap in the cells has increased in osmotic pressure somewhat more than the pressed sap in the test-tube. We may explain the smaller rise in No. 67—the pressed sap—as almost entirely due to the inversion of cane-sugar, which, at the same time, caused a depression of mean molecular weight. In No. 68 the stored starch in the uninjured cells furnished a supply of maltose, which not only augmented the increase of osmotic pressure, but, at the same time, raised the mean molecular weight. Taste XXI. Syringa vulgaris: leaves. Lear ae Description of Sample. A. 12, 69 Gathered 2.30 p.m. after dark morning, Oct. 26, brsssed and filtered immediately , kept and examined next day, . | 1445 17°38 70 Part of sample 69. Pressed immediately, not filtered, kept and examined next day, ° 2 0 : > |) 1-456 17-52 Here the unfiltered sap, presumably sap containing starch, altered more than the filtered sap; and the observations thus fall into line with those in Table XX. But the obvious explanation just given requires modification before it will apply to experiments set out in Table XXII. In these, two lots were made from opposite halves of leaves of one sample. From lot @ Dixon anp Arkins—On Osmotic Pressure in Plants, &¢. 305 some 10 c.c. of sap were pressed immediately, and divided into three parts. The first was examined immediately. The second was filtered, stored till next day, and then tested. The third was stored unfiltered, and tested next day. Lot 6 was stored in the dark till next day, when it was pressed and examined. Taste XXII. Syringa vulgaris: leaves. | fee Description of Sample. A. Pp. M. | | i) | 71 | (a) Half leaves (1), sap pressed and examined immediately, | 1:424]17-13} 211 | 73 (2), sap pressed and filtered, kept and examined | next day, 9 0 0 . | 1-640} 18°52) 210 | | yee (3), sap pressed, not filtered, kept and examined | | next day, o : 5 . | 1505) 18-10} 216 | 72 (6) Half leaves kept in dark, examined next day, 5 . | 1:424] 17-18] 217 The rise in mean molecular weight, in Nos. 74 and 72 is small, and if of any significance, again indicates the production of maltose from starch, which must have been present in the uninjured cells and in the unfiltered sap. At the same time respiration in the uninjured cells, we may assume, reduced the amount of dextrose, and so prevented a rise in osmotic pressure. In No 73—the filtered sap—the pressure rose owing possibly to the inversion of cane-sugar to dextrose and levulose. ‘The same process would cause a slight fall in molecular weight. ‘The solid matter having been removed, no maltose could be formed, and so no rise of mean molecular weight from this cause was to be expected. In 74 there is the same cause for a rise in osmotic pressure as in 73, and the same reason for increased mean molecular weight as for 72. Whatever is the cause of the alteration in the osmotic pressure of the sap on keeping, its amount, which in no case exceeded 1°7 atmospheres, is quite unable to obscure the effects of assimilation and evaporation, which have been observed to cause changes of pressure amounting to as much as eight atmospheres. It is interesting to find that these results are quite justified by Brown and Morris's analyses' of the sugars of leaves, and investigation of the change in them on keeping. ‘These authors give, in their memoir on the 1 Brown and Morris, Chemistry and Physiology of Foliage Leayes: Journ. Chemical Soc., May, 1893. 306 Scientific Proceedings, Royal Dublin Society. Physiology of Foliage Leaves, the amounts of the various sugars found immediately on gathering leaves of TZropwolum majus, and the amounts determined after the leaves had been kept twenty-four hours in darkness. First ExpErimMent. SrconpD EXprrimeEnr. Percentage on dry weight. Percentage on dry weight. When gathered. | After 24 hours dark. | When gathered. | After 24 hours dark. Cane-sugar, 9-98 3°49 7-33 3-35 Dextrose, . 0-00 0°58 0-00 1:34 Leyulose, . 1:41 3°46 | Dall 3°76 | Maltose, 2-25 | 1-86 271 1-28 Allowing 342 for the molecular weight of cane-sugar and maltose, and 180 for that of dextrose and levulose, and calculating the number of gram- molecules of these sugars present as a measure of osmotic pressure, we find in the first experiment the number changed from 0°0436 gram-molecules when gathered to 0:0381 gram-molecules after twenty-four hours in dark. This change would evidently cause a fall in the osmotic pressure. In Brown and Morris’s second experiment the ratio of the number of gram-molecules is 0:0411 to 0:0419 after twenty-four hours. ‘This would, of course, pro- duce a rise of pressure, and would quite parallel our pair, 66 and 68, where the osmotic pressures before and after twenty-four hours’ darkness were 14:61 and 14:76 atmospheres. Considered from the point of view of the osmotic pressures, Brown and Morris’s analyses are extremely interesting, as showing that it is possible for the osmotic pressure to remain steady when the total percentage of the sugars varies considerably, and vice versa. Taking a general survey of the experiments on Syringa vulgaris, we see that the highest pressure observed in the leaves freshly gathered was, at the end of August, 26°87 atmospheres (experiment 5). ‘This observation was made by means of the early arrangements, in which the sap was diluted, and possibly is not so reliable as the determination obtained on the pressed sap. ‘The result, however, is supported by experiment 45, in which undiluted sap was used. ‘The pressure estimated in the latter experiment was 24°57 atmospheres. ‘Chat both these pressures might, under conditions particularly Dixon any Arkins—On Osmotic Pressure in Plants Se. 307 favourable to assimilation and transpiration, be considerably raised appears probable from experiments 38 and 39. ‘The former shows that the pressure of the leaves when gathered was 17°12 atmospheres, while the latter shows that, by exposing some of the same sample of leaves to very favourable conditions, this pressure was caused to rise to 25°68 atmospheres. A. similar rise is shown in experiments 33 and 34, where assimilation and transpiration have caused a rise from 16°26 to 24:07 atmospheres. These observations would lead us to believe that the osmotic pressure of the sap of leaves of Syringa vulgaris might easily rise to 80 atmospheres or more. Turning now to the determinations of the mean molecular weights of the dissolved substances, they are seen to vary from 156 to 273. The great majority lie about, or over, 200. It is remarkable that, while screening the leaves from light always reduces the osmotic pressure, it often has little or no effect on the mean molecular weight. In fact, the highest mean molecular weight observed in the series on Syringa was in the sap of leaves which had been cut off from light for three days (experiment 46). On the other hand, the lowest mean molecular weights are also found after the leaves have been eut off from light (ep. experiments 37, 53, and 55). As was pointed out before, we must probably look to the fluctuation of the sugar-content of the cells as the chief cause of variation in the mean molecular weight of the dissolved substances. When the percentage of the disaccharides rises, the molecular weight will more nearly approach 342, the molecular weight of saccharose and maltose. A rise in the concentration of the monosaccharides, dextrose, and levulose, having a molecular weight of 180, will not necessarily tend to bring up the mean molecular weight of the solutes above that figure. The formation of organic acids will not tend to bring the mean molecular weight above 200. The fact that saccharose is known to be formed during assimilation is sufficient explanation of the observations showing that insolated and well-illuminated leaves have a sap with a high mean molecular weight. The general high mean molecular weight of the solutes in the sap of leaves screened from light must be explained by assuming that some disaccharide is produced in them or is transported to them. From Brown and Morris’s elaborate work it is certain that maltose is developed by the action of diastase on stored starch. They show reason to believe that carbohydrates are transported in the form of levulose. ‘This would seem to cut out translocation as the direct means of maintaining the high molecular weight, and would seem to show that the hydrolysis of starch explains the high molecular weight found in these cases. The very low values exhibited in some of the experiments on screened leaves are also readily explained by the exhaustion of stored carbohydrate, SCIENT. PROC. R.D.S., VOL. XII., NO. XXV. 3c 308 Scientific Proceedings, Royal Dublin Society. viz. starch. The mean molecular weight, 160, observed in experiment 37, is to be explained in this way. The leaves under observation in this experi- ment grew in a naturally overshadowed position, so that they had little opportunity of storing starch. Screening them completely from light pre- vented them from forming saccharose, and consequently the mean molecular weight of their sap was run down in three days to the low figure of 160. In marked contrast to this is the behaviour of most of the other screened samples which, before they were artificially screened from light, were in an exposed position. In them presumably, as has been pointed out, the stored carbohydrates present were able by their conversion into maltose to keep up the mean molecular weight. The low mean molecular weight of experiment 53, viz. 172, was likewise found in leaves which were naturally in an over- shadowed position, after a continuance of cold and wet. In this case, too, there is reason to assume an absence of starch and consequently, no source. of maltose, while the dark weather and shaded position prevented an active formation of saccharose. In experiment 55, we have the record of the lowest molecular weight for Syringa. ‘The leaves had been cut off from light for twelve days, so that their supply of starch must have been approaching exhaustion. ‘That it was not quite exhausted is shown by the fact that a rise in osmotic pressure and molecular weight was observed when some of the same sample of leaves were kept in the dark after gathering for twenty-four hours. This observation rendered it probable that the low mean molecular weight was due to retardation of the hydrolysis of starch in the dark weather, and not to its exhaustion. This conclusion was rendered all the more probable by a subsequent observation on the remaining leaves of the darkened shoots. From this it appeared, although cut off from light as before, the mean molecular weight rose subsequently on the return of the favourable conditions. Thus in experiment 58, made nine days later on these leaves—that is, after they had been in almost complete darkness for twenty-one days—the mean molecular weight had risen to 249. The general supply of carbohydrate, however, being much reduced by the long sojourn in the dark, the osmotic pressure had fallen to 11°58 atmospheres. Leaves from the same branch, as was shown before, seem in each case investigated to have approximately the same osmotic pressure when exposed to the same conditions. The experiments on Eucalyptus globulus (Nos. 89 and 90, 91 and 92) and Pinus Laricio (Nos. 85 and 86) further illustrate this. In both these cases, however, the leaves of the previous year showed a dis- tinetly higher figure for the osmotic pressure of the sap than those of the current year. Dixon anp ArKins—On Osmotic Pressure in Plants, &c. 309 Another point which stands out clearly from the observations is the contrast offered by the sap of the roots to that of the leaves. In Syringa the pressure of the sap of exposed leaves was found to vary from 14 to 24 atmospheres, while that of the roots lay between 4 and 6 atmospheres. In Eucalyptus the osmotic pressure of the leaves ranged between 6:1 and 8-4 atmospheres; that of the roots was 5:3 atmospheres. Large differences were observed in the mean molecular weight of the sap of the roots of Syringa, viz. 167 to 254. The former of these was obtained at the end of September, and the latter towards the end of October. Possibly the lower figure represents the normal mean molecular weight of the solutes of the root-sap, while the higher figure is attained only when large quantities of carbohydrates are being transported from the leaves into the roots, e.g., previous to the shedding of the leaves. In succeeding observations the dissolved substances in the sap of the root showed signs of gradually regaining their original low molecular weight. The second change may have been brought about by the deposition of starch in the roots, and the consequent reduction in the amount of dissolved sugars of high molecular weight. It may be noted that the lowest osmotic pressure observed so far by this method was in the sap of the leaves of Chamaerops humilis, viz. 3°79 atmo- spheres. This sap is also remarkable in containing dissolved substances having the highest mean molecular weight observed, viz. 323. The sap of the three monocotyledons examined was remarkably clear, being almost colourless, and strikingly different from the dark sap of many of the other plants. Only two cryptogams have been examined up to the present. The osmotic pressure of their sap is in no way remarkable. ‘The mean molecular weight of the dissolved substances of Equisetum is very low, viz. 122. This is the more remarkable, seeing that the sap was obtained from insolated plants, and there had been nine hours’ sunshine on the previous day. Possibly the observation may indicate that a sugar of low molecular weight, such as glucose, is formed in the early stages of the assimilatory process. Glancing over the list of determinations, it will be seen that in several instances the osmotic pressure of the sap of the leaves investigated was very high, viz. :— Catalpa bignonioides, . . 22:9 atmospheres. Syringa vulgaris, : 5 26:8 i Magnolia acuminata, . 0 BS D Fraxinus execelsior, . a EPR 3 Fraxinus oxyphylla, . eZ 4-0) el 310 Scientific Proceedings, Royal Dubhin Society. These determinations agree well with the highest determinations obtained by the gas-pressure method. It is also possible that, being observed late in the summer, when the conditions for assimilation and transpiration were not optimal, these pressures are not maxima, but might be higher under more favourable conditions; and in several cases it has been found possible to considerably raise the pressure by evaporation before wilting ensues. In any case the observations show that the osmotic pressures in the cells of the leaves are more than ample to resist the tensions developed in the transpira- tion-current. ConcLusions. 1. Osmotic pressures are variable with the species and individual. 2. Leaves of the same individual under similar conditions have the same osmotic pressure. 3. In the same individual considerable variations are found under varying conditions; for example, in Syringa vulgaris, the pressure in the leaves of one plant was found to vary from 24:58 atmospheres to 11:58 atmospheres. 4, Variation in pressure is not defined by the height of the leaves above the ground, nor by the resistance of the conducting tracts supplying the leaves. In each case the osmotic pressure was much greater than the tension of the water-supply could have been. 5. The variations in the osmotic pressure observed are probably due principally to fluctuations in the carbohydrate- contents of the cells. Assimilation leads to a rise in the osmotic pressure, and in the mean mole- cular weight of the solutes. 6. A similar, but smaller, rise in osmotic pressure, and a similar rise in mean molecular weight, may be observed in plucked leaves stored in the dark. These changes are probably due largely to the hydrolysis of saccha- rose and of starch. 7. No such rise was observed in starved leaves or in roots kept in the dark. 8. The osmotic pressure of leaves still attached to a plant may be reduced greatiy (e.g. from 18°10 to 11°58 atmospheres) by shielding from light. 9. Other things being equal, mature leaves showed a higher osmotic pressure than developing leaves. 10. The roots examined had comparatively low osmotic pressures, viz., 4 to 6 atmospheres. » Dixon anp ArKins—On Osmotic Pressure in Plants, §c. 311 11. The greatest depression of freezing-point was recorded by the sap of Syringa vulgaris, viz. — 2°234° C., corresponding to an osmotic pressure of 26°87 atmospheres. ‘The smallest depression was found in the sap of Chamaerops humilis, and amounted to only — 0:315° C., equivalent to an osmotic pressure of 3°79 atmospheres. 12. It is not probable that 26°87 atmospheres is the maximum osmotic pressure for Syringa vulgaris. It was found that assimilation and evapora- tion could raise the osmotic pressure of leaves from 17:12 atmospheres to 25°68 atmospheres before wilting supervened. In summer, when the leaf- cells are loaded with greater quantities of sugars, the wilting concentration will be higher. In this case pressures from 30 to 40 atmospheres in Syringa vulgaris are not improbable. . sCluwt., PROC. R.D.S., VOL. XII.) NO. XXYV. BD THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 26. FEBRUARY, 1910. PERMANENT STEEL MAGNETS. BY W. BROWN, B.Sc., PROFESSOR OF APPLIED PHYSICS, ROYAL COLLEGE OF SOIENCE FOR IRELAND. [Authors alone are responsible forall opinions expressed in their Communications. | DUBLIN : PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, ae 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1910. YUN 143 19/0 Price Sixpence. errs gansenian instig/ oN ty, " “ional Museu" ve Roval Bublt Society. ea aaa a aa FOUNDED, A.D. 1781. INCORPORATED, 1749. EVENING SCIENTIFIC MEETINGS. Tue Scientific Meetings of the Society are held alternately at 4.30 p-m. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the [ditor. Dixon ano Arxins— On Osmotic Pressure in Plan 5, Cpe BULL 11. The greatest depression of freezing-point was recorded by the sap of Syringa vulgaris, viz. — 2°234° C., corresponding to an osmotic pressure of 26°87 atmospheres. The smallest depression was found in the sap of Chamaerops humilis, and amounted to only — 0°315° C., equivalent to an osmotic pressure of 3°79 atmospheres. 12. It is not probable that 26°87 atmospheres is the maximum osmotic pressure for Syringa vulgaris. It was found that assimilation and evapora- tion could raise the osmotic pressure of leaves from 17-12 atmospheres to 25°68 atmospheres before wilting supervened. In summer, when the leaf- cells are loaded with greater quantities of sugars, the wilting concentration will be higher. In this case pressures from 30 to 40 atmospheres in Syringa vulgaris are not improbable, SCIENT, PROC. R.D.S., VOL. XII.) NO. XXV, 8D [ ae XXVI. PERMANENT STEEL MAGNETS. By W. BROWN, B.Sc., Professor of Applied Physics, Royal College of Science for Ireland. [Read Decemprr 21, 1909. Ordered for Publication January 11. Published Frpruany 21, 1910.] Secrion I. Tue first important experiments on permanent steel magnets—since the . publication of Gauss’s paper on Terrestrial Magnetism in 1832—were made by TT. Gray! in 1878, who gave the results in absolute c.g.s. measure. The magnets he used were 5 em. long, and 0:1 cm. diameter, or about 50 diameters long; and the magnetic moment per gramme varied from 62 to 79. In 1882 L. M. Cheesman® published the results of some important experiments on magnets made of English silver steel, which were hardened by stretching; he found that the best retentivity was obtained with a magnet of dimension ratio* (length divided by diameter) 41; the diameter being 0°128 cm. In 1885 T. Gray‘ published the results of some determinations of the horizontal component of the earth’s magnetic field, and at the end of his paper a table of values is given for glass-hard steel magnets. By plotting from this table the dimension ratio of the magnets against the magnetic moment per gramme, we find for the kind of steel he used, that the best length of magnet would be 10 cm., or dimension ratio 30, and magnetic moment per gramme 40. In 1887, the author’ published the results of some tests of steel magnets, where, among other things, an attempt was made to obtain the most effective dimension ratio for a given diameter of magnet. 1Phil. Mag., vol. vi., 1878, p. 321. 2 Wied. Annal. Bd. xv., 1882, p. 204. 3In the Phil. Mag., vol. xvii., 1909, p. 733, Professor S. P. Thompson and E. W. Moss define dimension-ratio as the length of the magnet divided by the square root of its cross-sectional arva. By this method, magnets whose cross-sections are round, square, or oblong can be compared. 4Phil. Mag., vol. xx., 1885, p. 497. 5 Phil Mag., vol. xxiii., 1887, pp. 293 and 420; vol. xxyii., 1889, p. 270, Brown— Permanent Steel Magnets. 313 Some results of experiments, published by H. Frank' in 1900, on English silver-steel confirm the results obtained by T. Gray; the author also gives some interesting cyclic curves obtained by relating magnetic moment and temperature. The first section of the present short paper gives the result of some experiments in which an attempt has been made to find the best length of magnet to use for a given diameter, the same kind of steel being used throughout, and in the same physical condition. By the best length is meant that obtained from the dimension ratio which gives the highest magnetic moment per gramme, for a minimum weight of material. For this purpose Stubs’ tool steel was employed, a material the chemical composition of which is not made public; but it can be obtained of cylindrical section in lengths of 383 cm. or so, and of diameter up to one centimetre or more, of practically the same cross-sectional area throughout each length. Twelve different sizes of this steel were obtained, varying in diameter from 0:16 cm. to 1:0 cm., and six magnets of each sample were made varying in length from 15 em. to 4 cm. The magnets were all submitted to the same heat-treatment, that is, they were raised toa bright red heat in a gas muffle furnace, and then dropped end on into 3 feet of water at 16°C., they were then cleaned up, measured, and weighed, and magnetized in a magnetic field of 2,000 c.g.s. units. The magnetic moment per gramme, or specific magnetization of each, was then carefully determined by means of the usual magnetometric method. The results for each set of magnets were then plotted in a curve, when the abscissee represented the dimension ratio, and the ordinates the maguetic moment per gramme. The values for one set of experiments are here given in Table I., and shown as a curve in fig. 1. Taste I. Length and diameter Dimension ratio Weight in Magnetic moment of magnets in cms. length/diameter. grammes. per gramme. 15 x 0°16 | 94 2:27 50-0 2; 75 1-82 49°5 10, 62-5 1-49 48-0 gas 50 1-21 42-0 6 375 0-90 304 pp 25 0°60 12:0 1 Ann. d. Physik., Bd. i., 1900, p. 190, 638. 3Dd2 314 Scientific Proceedings, Royal Dublin Society. From the curve in fig. 1 we see that as the dimension ratio increases the magnetic moment per gramme increases up to a certain point, and then the curve becomes flat, showing that there is one value of the dimension ratio, or one value of the length of magnet for a given diameter that gives the best result. That is, in this curve the dimension ratio 80, or a magnet of length 12°8 ems., is the best length to make with this steel of 0:16 cm. diameter. ’ The same operations were gone through for other eleven samples of the same kind of steel, i.e. six magnets were made of each, tempered, cleaned, weighed, and tested for magnetic moment per gramme. gramme b o c) (Zig ‘ Magnetic Monet A curve for each set was drawn like fig. 1, and the most effective value of the dimension ratio taken from the curve, and the best length of magnet ascertained. The results are shown in Table II. ; and if we were to plot the values in column 2 of the Table as abscissae, and the corresponding values in column 3 as ordinates, the curve obtained would show that from a diameter of 0:16 cm. to about 0°5 cm. the points lie very approximately in a straight line, whereas for diameters larger than 0°5 cm. or up tol 0 cm. the curve droops slightly towards the origin. Brown—Permanent Steel Magnets. TaseE II. Diameter of magnet Dimension ratio Best length of magnet (cms.) length/diameter. (cms.) 0-160 50 12°8 0-166 (i) 12-5 0-200 61 7, 0°285 38 10-4 0-300 34 10-2 0°322 32 10°71 0°333 30 10-0 0-467 20 9:3 0-600 14-7 8-8 0-700 12°3 8-6 0-800 10°5 8-4 1-000 10-0 8-1 315 The results obtained are, perhaps, better shown by plotting in a curve the values in column 1 of Table II. as abscissae, and the corresponding values in column 38 as ordinates. We are thus enabled to fix on the most effective length to make the magnet when we know the diameter, as shown here in fig. 2. > N Q Fie. 2. 316 Scientific Proceedings, Royal Dublin Society. In order now to test what would be the best shape of magnet to make with this same Stubs’ steel, four magnets were made, each 10 ems. long, and 0-4 cm. diameter. Magnet A was a cylinder with its end-surfaces at right angles to its length; B had each of its ends rounded off into a hemisphere ; C’ had its end-surfaces at right angles to its length, but was turned down to 0:3 cm. diameter at each end, the middle cross-section being 0:4 cm. diameter ; and magnet D was the same shape as C, with each of its ends turned down to 0:2 cm. diameter, the middle part also being 0-4 cm. diameter. Magnets C and D were approximate elongated ellipsoids. These four magnets were hardened, cleaned, weighed, magnetized, and tested as in the previous cases. ‘The magnetic moment per gramme of each was carefully determined, and also the percentage loss in the magnetism of each, when they were let fall end on (with the ¢we north or souwth-seeking pole downwards) jive times through a height of 100 ems. on to a block of glass. ‘The results obtained are shown in Table III. Tasxe III. Veig: i o ic Mark. Weight ne llegpeie ENTAAGN Percentage loss. gramumies. per gramme. | A 8:83 30°5 5-71 B 8°46 31:0 1°65 C 8:03 31:2 3°80 D 7°04 3814 8°52 Assuming that the magnets are all in the same physical condition, or glass-hard, this shows that the magnet B, with the rounded ends, was the least affected by percussion. Section II. At intervals during the five years from 1900 to 1904 communications" were brought before this Society on the magnetic properties of alloys of iron, which were in the form of long rods. Most of the rods had been returned to the manufacturer for the purpose of being tested mechanically ; there were still available short pieces of nineteen of these rods, and eleven more samples had been obtained from the makers at the Hecla Steel Works, 1 Barrett, Brown, and Hadfield, Scient. Trans. Roy. Dublin Soc., vol. vii., 1900, p. 67; vol. viii, 1902, p. 1; vol. viii., 1904, p. 109. Brown—Permanent Steel Magnets. 317 Sheffield. The writer endeavoured to find out whether these thirty specimens were suitable for permanent magnets. After a good deal of preliminary work only fourteen of them were found to give anything like satisfactory results, and for the purpose of testing, these were all turned down to 0:3 em. diameter, and from 8 to 10 cms. long according to the material at command. In order to get the magnets made as permanent as possible, the method suggested by Barus and Stronhal' was adopted, viz.:—IWMuke the magnets glass-hard, then place them in steam at 100° C. for 20 hours. Magnetize them as fully as possible, and again heat them in steam for 5 hours. It was not convenient to steam the magnets for twenty hours continuously but for three successive days they had 7 hours’ steam and 14 hours’ rest, or 21 hours’ steam altogether. They were then magnetized in a magnetic field of 2000 c.g.s, units, and again steamed for 5 hours, and when cooled they were tested for the first time. The magnets were prepared, hardened, and tested in exactly the same way as already mentioned with the Stubs’ steel magnets. he first tests were made on June 17th, about an hour after the magnets had been steamed for the final 5 hours, and the other tests took place on the same date of the succeeding months up to December. In the intervals between the tests, the magnets were kept in a vertical position, away from all iron or other disturbances. Table IV. gives the chemical composition of the magnets, and the results of the tests for the first, second, and sixth months respectively, as well as the percentage loss in magnetic moment per gramme for six months. 1 Bulletin No. 14, 1885; Dept. of the Interior, U.S.A. Geological Survey. [Tasre IV. 318 Scientific Proceedings, Royal Dublin Society. Tasie LV. : Magnetic moment | @ » Chemical Composition.? en a Gama Oa ae No. | Mark. ratio UE sai eege U) o.4 C Mn Si Ni W Cr A June | July | Dec. é P 1 | S.C.T. | 0-028 0-07 | 27 | 20-2) 20-0 | 12-2 | 40-0 2 1392 L | 1-09 0°32 | 0:06 | | by 26°7 | 28:6 | 21:4 | 20:0 3 | 14204 | 0-75 | 1-00 | ane 33°5 | 35:3 | 31-2 | 6-8 4 1420 B | 0:50 1:00 eel. Om | 90 30°6 | 36:2 | 27-5 | 10-1 5 1430 1°30 3°09 | 8-92 99 30°7 | 33:1 | 26:4 | 14:0 6 S.C I. | 0-028 0:07 | 33 20°7 | 20°0 | 12-0 | 42-0 7 1397 A | 0:22 0°18 | 0°44 | on 28:1 | 29:6 | 25-0 | 11-0. 8 1397 B | 0:26 0-718 | 0°33 | 0°48 a 29°7 | 32-0 27°5 74 9 1294 H | 0-28 0:28 3°5 99 24°0 | 25°5 | 22-4 6°6 10 | 1294L | 0-76 | 0-28 15°5 0 39°6 | 41°56 | 38:5 | 2-7 11 687 0-40 2°25 | 3°25 AB 33°5 | 34°3 | 29°6 | 11:6 12 1189 B | 0-25 2-00} 0°75 99 34°9 | 37:3 | 30°0 | 14-0 13 1177 I_| 0°48 3°25 90 35°9 | 38:0 | 34:5 3°6 14 1177N | 1:09 9°50 90 80°7 | 383°0 | 27-5 | 10-4 15 1286 A | 0°25 2°76 0°75 99 Wedel Ss Tale lbs0N ellos? 16 1286 C | 0-31 2°50 1°75 Ag 33°4 | 385°3 | 30:2 9°6 The results are arranged under the dimension ratios 27 and 33 respectively, and for the purpose of contrast a magnet 8.C.I., made of practically pure iron, is placed at the head of each group. Comparing Nos. 3 and 4, which have the same quantity of manganese, the one per cent. of nickel in No. 4 does not seem to have much effect on the moment, as the reduction of the magnetic moment by nearly 9 per cent. can be accounted for by the difference of 0°25 per cent. of carbon in No. 4; the nickel, however, may account for the decrease of nearly 50 per cent. in the retentivity. In Nos. 7 and 8 the small percentage of nickel has increased the magnetic moment by nearly 6 per cent, and improved the retentivity some 48 per cent.; the other chemical constituents being approximately the same, a small percentage of nickel does improve the retentivity.? In the tungsten steels, Nos. 9 and 10, the 0°5 per cent of carbon, and the 1 The chemicul analyses of the specimens were made in the chemical laboratory of the Hecla Steel Works, Shettield, the amount of iron being estimated by difference. 2 Scient. Trans. Roy. Dublin Soc., vol. vii., Plate iv., Jan. 1900, p. 67. Brown— Permanent Steel Magnets. 319 12 per cent of tungsten has made an increase of 65 per cent. in the magnetic moment, and increased the retentivity nearly three times. In the chrome steels Nos. 13 and 14, the lower percentage of chromium seems to have a better effect on the moment than the higher, as is also indicated in No. 5, though when nickel is present, as in Nos. 15 and 16, it appears to mask the effect of chromium if the amount of the latter is small. Abt? has published some results on the comparison of the magnetic moments of magnets, which have the same dimensions, and magnetized in the same magnetic field, but of different composition, that is, crucible, diamond, and tungsten steels. The magnetic moments of these magnets were 33°7, 39°9, and 62°15, and at the end of the four months the percentage Fie: 3: losses were 1-1, 25:9, and 26°5 respectively. The last value is very high for tungsten steel; the chemical composition of the steels, however, is not given in the paper. Mme. Curie,’ in an important paper on the magnetic properties of tempered steels, where most of the specimens tested were in the form of bars of square section 20 cms. long and one centimetre in the side, gives the results of some tests on tungsten steels. If we take the steels with the highest and lowest amounts of tungsten mentioned in the paper, and assume the density of the material to be 8, we find the magnetic moment per gramme to be 63 for the steel with W = 2:7, and 45 for that with W =7°7. The results for the four best specimens in Table IV., namely, the two tungsten steels (Nos. 9 and 10), and the two chrome steels (Nos. 13 and 14), are shown in curves in fig. 3, which exhibits the behaviour of each specimen 1 Wied. Ann. Ixvi. 1, pp. 116-120, 1898. 2 Bull. de la Société d’ Encouragement, pp. 36-76, 1898. SCIENT. PROC. R.D.S., VOL. XI., NO. XXVI. 35 320 Scientifie Proceedings, Royal Dublin Society. during a period of six months. In every case, as will be seen from Table IV., the magnetic moment is increased after the month’s rest, due no doubt to the material coming back to its state of molecular equilibrium after the heat treatment; there is then a gradual decrease in the moment till each of the four magnets becomes practically permanent. It is well known that the value of the dimension ratio of a magnet makes a great difference in its permanent magnetism.! The B—-H cyclic curves of the four specimens in fig. 3 were determined some ten years ago,’ their dimension ratio being about 200; and by taking the residual induction then obtained, and dividing by 4m times the density of the material, we get the specific magnetization or magnetic moment per gramme as shown here in Table V. Taste V. Mark. Residual reduction. | Coercive force. Density.? tik See in 1294 H 12500 5:5 80645 128 1294 L 9500 13°8 8°7720 86 11771 8480 12°3 7°7653 87 1177 N 10560 8:3 77032 109 There is no evident relation between the numbers in the last column of this table and the corresponding numbers in the tenth column of Table IV. ‘The values in Table V., however, are of the same order as those obtained by Hopkinson for chrome and tungsten steels, whose method of testing was a closed circuit one. ‘he values of the specific magnetizations are 97 for the specimen containing W =4:65; C=1°36, and 68 for that containing W = 2-35; C = 0°86. 1 The Electromagnet, 8. P. Thompson, p. 383. * Scient. Trans. Royal Dublin Soc., vol. vii., 1900, p. 67. 5 Scient. Trans. Royal Dublin Soc., vol. ix., 1907, p. 59. 4 Phil. Trans. Roy. Soc., vol. clxxvi., part 2, p. 468, 1885. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.8.), No. 27. MARCH, 1910. SOME VARIATIONS IN THE SKELETON OF THE DOMESTIC HORSE AND THEIR SIGNIFICANCE. BY MAJOR F. EASSIE, D.S.O. [COMMUNICATED BY DR. R. F. SCHAREF, PH.D., B.SC.] (PLATES XVI.—XX.) [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1910. eel Pag et 4 oansenlan Insti fg Stitupe~ ) % Price One Shilling. pea H \ YUN £6 1910 \ &y ANE 7 Hf Pal Musew™ Roval Bublir Society. ADIT FOUNDED, A.D. 1731. INCORPORATED, 1749. wae, EVENING SCIENTIFIC MEETINGS. Tur Scientific Meetings of the Society are held alternately at 4.30 p.m. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee.' The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Editor. [ 8 J XXVIT. SOME VARIATIONS IN THE SKELETON OF THE DOMESTIC HORSE AND THEIR SIGNIFICANCE. By MAJOR F. EASSIBE, D.S.O. |COMMUNICATED BY DR. R. F, SCHARFF, PH.D., B.SC. | (Prarrs XVI.-XX.) [Read Decemprr 21, 1909. Ordered for Publication, January 11; Published, Marcu 8, 1910.] My first object will be to show that the skeleton of the domesticated horse frequently gives proof of deterioration from the type of the skeleton of the wild horse. In order to make this clear, I shall first consider the latter, and show how this deterioration manifests itself. A group of qualities essential to the horse are affected by the deterioration of the skeleton, such as strength, vitality, nervous energy, and stamina; all of which are normally attributes of the wild animal. Certain variations in the skeleton will then be discussed; and I shall endeavour to show that these are apparently transmitted as independent hereditary characters. Three variations in the skeleton may be considered :— 1. The relative length of the spine, including the head. 2. The relative length of the arm. 3. The relative length of the thigh. One of my arguments is that the short spine is a racial character in the Arab; and that it is transmitted in the first cross of the Arab, I shall also attempt to show that the relative shortness respectively of the arm and of the thigh are ultimate phases in evolution ; that they determine qualities in the horse of the utmost economic value, namely, balance of the body on the supporting limbs, and speed; and that both are apparently separate dominant characters as regards heredity. Finally “action” in the horse will be considered, and how it is dependent on the foregoing characters of the skeleton. Before entering on the discussion of the subject, I ought briefly to state the cireumstances under which the ideas contained in this paper have been formed. I was first led to study the subject by some remarks upon degeneration in the horse made to me by an officer of my own service’ about ! Major, now Colonel Blenkinsop, D.S.O. SCIENT. PROC, R.D.S., VOL. XII., NO. XXVII, 3F 322 Scientific Proceedings, Royal Dublin Society. nine years ago. At that time I was managing a Remount Depot in South Africa during the war, where there were seldom less than two thousand horses at a time. Over forty thousand horses passed through this depot in the two and a half years of my management. ‘Ihese were, moreover, from different countries of origin. Owing to the system on which this depot was conducted, it lent itself peculiarly to the observation of horses. ‘The horses were kept loose in large enclosures provided with shelters. They were exercised by letting them into a central track, about two hundred at a time, where they were allowed to trot and canter round at their own will. They were watered by diverting them into enclosures where water troughs had been established, and returned to their enclosures to be fed after watering, the food having been put out for them during their absence. By seeing horses manceuvred in this way, it was possible quickly to get to know them individually, because they all had necessarily to pass a point in review many times a day at the morning and evening exercise. In the interval of time between these exercises and the watering and feeding, I had every day to pass many hundreds through an arrangement in connexion with the track in which, in a narrow gangway, they could be handled and caught, or let go, as required. It was here that I had an opportunity of studying the extent of degeneration in different breeds, and the variations in the skeleton which I deal with in this paper. Digs Te Fra. 2. Fic. 1.—Limb Bones. Fig. 2.—H. J. Hip Joints. S.J. Shoulder Joints. E.J. Elbow Joints. S. Stifles. A study of the skeleton of existing mammals in natural history collec- tions, from the largest to the smallest, and of remains of extinct ones, will make it clear that the bones of the limbs are roughly cylindrical, the upper and lower articular facets having the same direction (fig. 1). Since the bones are also straight (fig. 1), it follows that the pairs of limbs lie in parallel planes (fig. 2), the pairs of joints being equally wide apart, and the joints of each limb in line (fig. 8@). The only variation apparently occurs in animals of unusual strength and power. In these, the arms and thighs are Eassts—Some Variations in the Skeleton of the Domestic Horse. 328 bent outwards, the elbows and stifles being wider apart than the shoulders and hip-joints (fig. 36). In this arrangement there is the advantage of a very wide base of support, and consequently of greater stability. The parallelism of the limbs, and the forward direction of the bones and joints is well illustrated in the skeleton of the giraffe in the Irish National Museum, and in that of a young zebra in the same collection.’ The latter illustrates the points alluded to in an existing wild species of the same genus as the horse. ELBOW---* Fic. 3. I have now to show that. in the domesticated horse this type of the skeleton is sometimes widely departed from. ‘To begin with, the limbs fall inwards under the body, the elbows and stifles being then closer together than the shoulder and hip joints (fig. 3c). In addition, the bones of the limbs subside in a spiral manner, so as to turn the joints outwards in an increasing degree as far down as the knees and hocks. If this were continued, the feet would be turned entirely outwards. ‘There is, however, a rectifying twist in the bones below the knees and hocks, which partially restores the forward direction of the limbs (fig. 4, next page). It need hardly be mentioned that in the horse showing this condition of the limbs, the alteration of the bone is not confined to the limbs. The subsidence of bone is found also in the rest of the bones of the skeleton, which together form the two great cavities—the neural cavity and the visceral cavity of the skeleton. The essential organs contained in these restricted cavities are on the same inferior scale; and function consequently is everywhere inferior. This deterioration in the skeleton of the horse is easily seen in the living animal. It is noticeable, for example, in the turning inwards of the femur, especially during movement. Viewed from behind, the horse free from 1 T am indebted to the courtesy of Dr. Scharff for an examination of these specimens. 352 324 Scientific Proceedings, Royal Dublin Society. deterioration gives the impression always of a square figure formed by tlie hip joints and the stifles (fig. 5a). On the contrary, in the horse in which skeletal deterioration exists, the impression received is that of a wedge-shaped figure (fig. 5 0). In such a ‘wedged horse,” if we may call it so, the whole sequence of defects of the skeleton certainly exist. Many of the small breeds of horses exhibit this deterioration of the skeleton, such as the Iceland, the Burmese, and the Kaffir ponies of South Africa. On the other hand, some small races are appa- rently quite free from it; for example, the Basuto and the Connemara ponies apparently show no deterioration of the skeleton. It would seem that locality determines this con- dition to a great extent. It comes about apparently by the distribution of horses in localities unsuitable to them, and into which in the natural course of things they would not have found their way. ‘To some extent it is present in almost all races of horses in which any size has been attained. This is appa- rently due to the attempt to force size at the expense of the substance of bone. It must be due also, no doubt, to artificial conditions of life. Perhaps more than from any other cause, it is perpetuated by breeding from animals in which the condition has been inherited. I found all the Hungarian horses imported into South Africa, that passed through my hands (and there were hundreds of them), deteriorated in a marked degree. Only very few of a large number of Australian, North American, and Canadian horses were found to be so affected. About fifty per cent. of the artillery horses were degenerate in the sense indicated, but only to a slight degree. About fifty per cent. of the English cavalry horses were deteriorated. On the other hand, many thou- sands of American mules from New Orleans were quite free from any trace of deterioration ; and very IP. eh few instances of degeneracy were noticed among the Italian mules imported into South Africa. Eassts—Some Variations in the Skeleton of the Domestic Horse. 325 Plate XVI. represents an extreme case of deterioration of the skeleton in a Kaffir pony. In Plate XVII. are shown a number of South African ponies all deteriorated except the one on extreme right. Cases of deteriora- tion are also shown in Plate XVIII. In fig. 1, the horse on the left is an English thoroughbred by Ingomar—dam by Stirling, which was imported for stud purposes. Another example is that of an English thoroughbred by Bon Frére—Lightfoot, by Blair Athol (fig. 2). It is significant that out of five thoroughbred horses captured from the Boers and sent to me to take care of, four were markedly “ wedged.” I found also, in South Africa, one hackney stallion, and several Arabs imported into South Africa to improve the Basuto breed, all markedly “ wedged.” As further illustrating the distribution of deterioration of the skeleton in the horse, another picture of this condition in an Indian country-bred is seen in Plate XIX., fig. 1. It is curious that this evidence of deterioration of the skeleton of the domesticated horse from that of the normal wild horse is not mentioned in any work, that I can find, by English or Continental authorities, including the works of Sanson, Chauveau, Goubaux, and Barrier, the standard authorities on the zootechnic of the horse. a ) Fig. 5.—P. II. Prominence of Hip. H. J. Hip Joint. S. Stifle. I shall now refer to some important variations in the skeleton of the domesticated horse quite distinct from those alluded to. In the Arab the spine is short, because the segments or vertebree are short. ‘The head is short because of the shortness of the bones forming the skull. According to Sanson, the African variety of the Arab has one lumbar segment, and the English thoroughbred has one, or even two, less than other races of horses. Neither this author, nor Goubaux, nor Chauveau, nor any other writer that I can find, makes any reference to relative shortness of the segments. How the shortness of the spine affects beauty in the horse will be realized 326 Scientific Proceedings, Royal Dublin Society. by a comparison of the short and the loug spines which I have endeavoured to show diagramatically in fig. 6. It will be evident that the short, straight neck of the Arab determines the carriage of the head at an obtuse angle with the neck. It will be manifest also that the shortness of the spine gives compactness to the body generally, and brings the pelvis closer to the ribs. It will be evident, finally, that the straight spine determines the high carriage of the tail and the long, straight quarters, the pelvis necessarily following the line of the straight spine. On the other hand, it will be seen that with the long spine there are vertical curvatures which, although they are deformities, may be regarded as rectifying. When any one of these is absent—as is sometimes the case—the defect of the long spine is even more apparent. cay re SESS SS Sees = / ANN if FT Nae i Fre. 6. Now, it is said that the Arab stamps his stock. I find that what he stamps especially is the short spine. It is for this reason, principally, that all the progeny of the first cross resemble him. It is probable that the short head and spine is the primitive type. ‘The character of the skull of a horse found in the Pleistocene formation in the valley of the Nerbada, of which a cast is in the Irish National Museum, resembles that of the horses in the Greek sculptures and of the Arab horses of the present day. It is only when we get away from the influence of the eastern horse, that we find the long head and the long spine. Another important variation in the skeleton of the horse is to be Easstr—Some Variations in the Skeleton of the Domestic Horse. 327 found in the relative length of the arm. In the fossil ancestors of the horse, from the five-fingered and five-toed animal of the lowest Hocene formations, through the whole series to the horse of recent formations, the humerus has gradually been shortened in relation to the bones below it. In the various races of horses of the present day, there is a variability in the length of the humerus, apparently due to the fact that some are more advanced than others as regards this evolutionary shortening. It seems to me that the ultimate shortening determines perfection of function of the limb of the horse. In the work by Goubaux and Barrier on “‘ The Exterior of the Horse,”! it is shown that a large scapula for the attach- ment of the muscles, and a short humerus, are of advantage as regards speed. I have attempted to show this in the diagram, fig. 7. It will be evident that the long humerus gets as far as (a) while the short humerus reaches (0), a point far in advance of («), describing only the same length of are. It is beyond question therefore that a short humerus is requisite for speed. But a short humerus determines something even more «~~ : : : 0 ono Fic. 7. important. ‘l'o begin with, I find no appreciable variation : in the angle of the scapula with the humerus. It is always a little more SUPPORTING __ POWER, ~---- CENTRE OF GRAviTY. Vi I'rc. S$. than a right angle (fig. 8). ‘here is, however, admittedly, a wide difference in the angle at the elbow. When the humerus is long, it is recumbent ; ! Translated from the Second I'rench Edition by S. J. J. Harger, Philadelphia, 1892, 328 Scientific Proceedings, Royal Dublin Soctety. when it is short, it is erect. With the recumbent humerus, the scapula is erect; with the erect humerus, the scapula is thrown backwards into an oblique position. Now this determines balance in the horse, because the body is mainly suspended by muscles coming from the summits of the scapula bones. When the scapula lies far back on the thorax, the body rides forwards and the weight is nearly under the supporting power. When, on the contrary, the scapula is erect and far forward on the thorax, the body necessarily hangs back a long way behind the supporting power. It will be realized how badly handicapped as regards the carriage of the body is the horse with the upright scapula. Nature, therefore, in shortening the humerus, confers on the horse the quality of balance and to some extent the attribute of speed. I have now to show that there is a variation in the hind limb which affects speed more directly, since the hind limbs are the propelling limbs of the body. Just as speed is determined in the fore limbs by the short humerus, so it is determined in the hind limbs by the shortness of the femur. The long femur, like the long humerus, is recumbent; the short femur, like the short humerus, is erect. The criterion of speed in both the limbs is the open angle. In the fore limb it is the angle of the elbow; in the hind limb that of the stifle. We recognize the advance that has been made in the perfection of the horse as regards both speed and balance by the wide variation of these angles. ‘That as regards balance at least, the limit of perfection has not been reached in the horse, will be learnt by studying the skeleton of what I assume to have been a fast animal, such as the Irish elk, adapted to carry heavy antlers (fig. 9). As may be seen in the specimens in the National Museum, Dublin, and elsewhere, the scapula is relatively very large, and it is thrown back on the thorax by ashort erect humerus. ‘That the Irish elk was fast is proved also by the short femur and the very open angle of the stifle. Such glimpses as this into the aims and attainments of nature, while exciting wonder and reverence, will make regret the more keen that whole races of the horse have deteriorated under domestication from the type of the natural animal, and that we should allow to pass, as of little consequence, the primitive excellence of the short head and spine, without realizing that under so simple a character lies so much that is beautiful and useful in the horse. How do the above-mentioned variations in the skeleton of the horse affect “action”? It will be obvious that true direction of movement is only possible with the primitive arrangement of the limbs in parallel planes, Easstr— Some Variations in the Skeleton of the Domestic Horse. 329 It will be evident also that elegance of carriage is affected adversely by the very long spine and the excessive vertical curvatures. It will be manifest, moreover, that the carriage of the body to the best advantage is necessary to the freedom of movement of the fore limbs. Finally, only with open angles in both the fore and hind limbs can there be due length and rapidity of stride and co-ordination between the limbs. The figures (Plates XVI. to XX.) of different races of horses, exhibiting striking variability of the length of the spine and of the fore and hind limbs respectively, illustrate my remarks on this subject. Fic. 9. Plate XIX., fig. 2, represents the first cross of an Arab with a South African pony, showing the short head and spine of the Arab, the balance of the body, and the racial defect of this breed in the bent hind legs. That the Arab is a slow pony is notorious. I contend that the reason is to be found in this defect. In this case the same fault existed undoubtedly in the dam, and remains therefore uncorrected in the progeny. Plate XX., fig. 1, represents an English thoroughbred horse by St. Simon —Rowena, showing the short head and spine of the Arab. In some strains of the English thoroughbred this character is apparently beginning to be lost; and in this breed the short humerus is a fixed character. It is apparently also a dominant character. As balance is the first quality in the saddle-horse, the importance of direct descent on one side from the English SCIENT. PROC. R.D.S., VOL. XII., NO. XXVII. 3G 330 Scientific Proceedings, Royal Dublin Society. thoroughbred will be understood. ‘To select a sire from any other race but one in which this character is fixed seems to me a very doubtful experiment. Hven the harness-horse is the better for being a balance-horse, since otherwise he carries his own weight at a disadvantage. Above every other character this is one of most importance in the troop-horse. The figure above quoted shows very plainly the short thigh and the open angle of the stifle. It will be evident that the English thoroughbred has not inherited this character from his Arab ancestors. It is an historical fact that the Arabs imported into England were themselves slow, whereas their immediate progeny were fast. The inference is that the English mares had this character. It is, however, not absolutely invariable in the English thorough- bred. It is apparently a dominant character. Plate XX., fig. 2, shows a typical trooper supplied from Canada to South Africa for the war, showing the long spine and the vertical curvatures, also the defect of the fore limb. In it, however, there is the character of the straight hind leg. Such a race as this, influenced by the Arab, would probably lose the defects of both sides. Plate XX., fig. 3, represents a typical Australian horse, showing the short fore and hind limbs and the defect of the long spine with the marked vertical curvatures. Might not the primitive character of the short spine and skeleton, as opposed to deterioration from this type, be Mendelian dominant features? It appears to me also that the latest phases in the evolutionary shortening of the arm and thigh probably descend in the same way. Whether this is so, and whether also the descent of these characters is limited by sex influence, only the test of breeding can determine. I desire, before concluding this paper, to acknowledge my great indebtedness to Dr. Scharff, of the National Museum, Dublin, for the help he has given me in its preparation. yy TAX FLVId TIX “TOA “SN “OOS “IGN “Y ‘00%d “LNAIOS PLATE XVII. SCIENT. PROC. R. DUBL. SOC., N.S., Vou." XII. PLATE XVIII. SCIENT. PROC. R. DUBL. SOC., N.S., Vor. XII. tty y / PLATE XIX. SCIENT, PROC. R. DUBL. SOC., N.S., Vou. XII. hoes SCIENT. PROC. R. DUBL. SOC., N.S., Vou. XII. PLATE XX. ot oy THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XIE (N.S.), No. 28. APRIL, 1910. THE INHERITANCE OF COAT COLOUR IN HORSES. BY JAMES WILSON, M.A., B.Sc., PROFESSOR OF AGRICULTURE IN THE ROYAL 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 NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1910. re =, ae Zs NMG Price One Shilling. 3 ”, JUN 46 ly LU Roval Dublin Society. eee FOUNDED, A.D. 1731. INCORPORATED, 1749. EVENING SCIENTIFIC MEETINGS. Tue Scientific Meetings of the Society are held alternately at 4.30 pm. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Cominittee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Editor. : Cee XXVIII. THE INHERITANCE OF COAT COLOUR IN HORSES. By JAMES WILSON, M.A., B.Sc., Professor of Agriculture in the Royal College of Science, Dublin. [Read January 25: Ordered for Publication Fesruary 8: Published Apri 12, 1910.] Tue data which this or any similar paper must rely upon are unfortunately not of the highest accuracy, and reasonably reliable results can be obtained only when the inaccuracies are appraised and fairly allowed for. The chief source of inaccuracy lies in the indefinite and varying notions of horse- breeders as to the colours of their stock. Many Thoroughbreds are described as “bay or brown,” others as “ brown or black”; and grey and roan are some- times confused: the breeder’s eye seeing perhaps what the mind would like it to see. Clydesdale breeders have tried to be more accurate with regard to bays and browns, for they enter “light” bays and “dark” bays, and “licht,” “dark,” and even “very dark” browns; but it is doubtful whether this very attempt has not placed many bays among browns and browns among bays. At the same time they have an aversion to calling a roan a roan. It may be long before the distinction between the rufous coat of the bay and the darker coat of the brown is clearly and generally known; but the other common colours should present no great difficulty. Unless when white “socks” or ‘“‘stockings’’ have intruded, the legs of bays and browns are black towards the ground, but the body-colour of chestnuts, blacks, and greys is also the colour of the legs, excepting that in chestnuts and greys the legs are usually darker than the body. Most bays and all browns have a lighter- coloured patch at the nose—ifrequently sandy in the bay and tan in the brown —hbut blacks are black right down to the muzzle. The tan muzzle is the readiest means of distinguishing a very dark brown horse from a black. So far as can be made out at present roan is the intrusion of white hairs through a coat of any of the above-mentioned colours: the legs at the same time being unaffected, or nearly so. That is to say: a bay roan has black legs, a chestnut roan chestnut legs, and so on. In order to make the conclusions of this paper clear, we shall follow the course of inquiry by which they were arrived at. SCIENT. PROC, R.D.S., VOL. XII., NO. XXVIII. : 3H 332 Scientific Proceedings, Royal Dublin Society. Little or no assistance can be got from history, for the reason that the migrations of the horse and the interactions of race upon race have not yet been clearly traced. Local and breed histories are very unsatisfactory. Large effects are attributed to very small beginnings, and importations of horses from parts of Europe to places in Britain and Ireland are presumed upon the flimsiest evidence. If a writer cannot find some foreign prince to send a present to his friend the King of England or Scotland, or some English or Scots nobleman to import half-a-dozen stallions from Holland or Flanders, he has always the Spanish Armada to fall back upon; and “‘t és said” has been the foundation-stone of many an essay in equine history. Professor Ridgeway’s ‘“ Origin and Influence of the Thoroughbred Horse” is the outstanding exception ; but some of its conclusions are weakened by the assumption that when two races of different colour unite a new colour may be produced. For the present inquiry a jumping-off point had already been fixed by Mr. C. C. Hurst, whose paper “On the Inheritance of Coat Colour in Horses”’ was read before the Royal Society in December, 1905. Mr. Hurst showed chestnut to be recessive to bay and brown: bay and brown being taken together as one colour. A further hint as to the relationship subsisting between grey, bay, and black was afforded by the following communication published in The Breeders’ Gazette, an American journal, on the 9th of June, 1909 :—* In an experience of over thirty years in using Clydesdale stallions to all sorts of mares I have never known a mare drop a chestnut colt to one of them. One of my stallions sired some gray foals from gray mares, but these gray fillies, when mated with black Percherons, often had bay foals with white feet and stripes.” The Clydesdale stallions referred to above were probably bays or browns, for few of any other colour have been imported to America. Thus grey seemed dominant to bay and brown; and, as these came in between grey and black, it was further probable that black was recessive to bay and brown. The absence of chestnuts, against the expectation raised by Mr. Hurst’s conclusion, is not astonishing, since chestnuts have been unpopular among Clydesdales for many years. To test these tentatively-formed theories, an appeal was made to the Shire stud-book, because it contains a fair proportion of all the common colours. ‘The Thoroughbred stud-book, for instance, contains plenty of chestnuts, but it is deficient in blacks, roans, and greys. The first inquiry was confined to a tabulation of the colours of the foals entered in the first four volumes and of their sires and dams. After some working it became apparent, by observing how the colours did or did not “contain”? each other, that our tentative theory was approximately correct. Witson— The Inheritance of Coat Colour in Horses. 333 It also became apparent that there was some possibility of separating the bays and the browns; and in all tables these are put in separate columns. There was, however, considerable difficulty about the roans, in regard to which we started with no definite theory. Several were tried, but none was found to answer. ‘The one expected to fit best was that the roans were a hybrid between grey and one or more of the other colours, but it had to be given up. Eventually it was found that roans stand by themselves; and, for the present, they may be left out of consideration. We have thus five colours left to deal with: chestnut, black, bay, brown and grey. The dominance of one colour over another is shown by— (a) the dominant mated with the recessive producing sometimes the dominant sometimes the recessive ; (6) the recessive mated with itself producing always the recessive. For example, Mr. Hurst showed that bays mated with chestnuts produced both bays and chestnuts, while chestnuts mated with chestnuts produced chestnuts only. Aceording to this criterion our figures from the first four volumes of the Shire stud-book indicate, although the numbers are small, that grey is dominant to the other four colours, and that black, in addition to chestnut, is also recessive to bay or brown. ‘The relative positions of black and chestnut on the one hand and of bay and brown on the other are not clear. Black may be dominant to chestnut and brown to bay, but there are discrepancies in both cases. MATINGS IN VOLS. I. TO IV. OF THE SHIRE STUD-BOOK. es Chestnut. Black. | Bay. Brown. Grey. Ch. Bl. By. Br. Gr.| Ch. Bl. By. Br Gr.} Ch. Bl. By. Br. Gr.| Ch. Bl. By. Br. Gr.| Ch. Bl. By. Br. Gr. Chestnut, .| 27 -— 1 1 = 23 22:12 4 1 48 23 93 8 — 13 13 24 an = | ¢ 8 8 Til Black, . - - -=--- 217 -—- = 11] 19 34 76 42 - 2 23 23 8} = BG BB i Bay, = = = Se ==> = = =} 21 11 248 23 - 11 18 158 77 — 6 6 386 6 46 Brown, Se = SF] eo SS Seles Ss S= 1 8 19 387 -/|] — 616 12 20 Grey, = = ==s=s == = == = S5 = = = = = S£ = =i dos § Colours of foals. Another set of figures was collected, but this time foals of the more It was thought thus to include more valuable and And the search was continued 3H2 eminent sires were taken. therefore more carefully registered animals. 334 Scientific Proceedings, Royal Dublin Society. to the tenth volume, in the hope that errors incidental to the earlier volumes of stud-books might be eliminated. The results are as follows :— SELECTED MATINGS IN VOLS. I TO x. OF HE SHIRE STUD-BOOK. Colours of Chestnut. | Black. Bay. Brown. Grey. | Parents. | | Ch. Bl. By. Br. Gr.| Ch. Bl. By. Br. Gr. | Ch. Bl. By. Br. Gr. | Ch. Bl. By. Br. Gr. | Ch. Bl. By. Br. Gr. | Chestnut, .|| 44 1 5 1 — | 1926 27 7 — | 6719113 7 1 | 1415 43 91 — | 5 13 ea) a8 | eBlackeae Ss 52 =] 98 9 2 31 19 oo Rg = | 299 TO 28 =| 2 6 O18 o le Bay, ss SsSe}s 5 = = 28 13.287 18 — | 5 23 133 56 1 | 4 6 6011 46|)8) Broun eee oe ise Sa oi Reece od Ewer nemo es cwon call 2 Grey, Oe Ss It will be seen that the figures in this Table are practically a re-echo of those in the former. The relations of black to chestnut and brown to bay are no clearer. Remembering the errors of description to which stud-books are liable, black, brown, and bay, when bred to their own colours, seem to breed with regularity; but black is disturbed by the intrusion of chestnut, bay by black, and brown by black and chestnut. A similar and larger collection of data from the Clydesdale stud-book, which was next resorted to, gave almost identical results and left us no farther forward. But a deeper search into this book showed a possible cause of the disturbances referred to a few lines back. Among Clydesdales, breeding stock are usually entered in the stud-book twice: first as foals and after- wards when they themselves have foals to be entered; and their colours are usually mentioned both times. It was found that the colours of a good many animals were not the same at the second as at the first entering. The changes were not only from bay to brown or from brown to bay; but almost as many were from brown to black or black to brown. There was also a considerable number of changes involving chestnut and brown and chestnut and bay. We were thus driven to the Thoroughbred stud-book, which had been avoided so far because of the large amount of tedious labour involved in ascertaining sires’ and dams’ colours. Thoroughbred colts are not entered a second time unless to make corrections; and the colours of fillies when they re-appear with their progeny have been given only in the last three volumes. For our purpose, Thoroughbreds have an advantage over Clydesdales and Wi.son—The Inheritance of Coat Colour in Horses. 339 Shires in that their colours are not usually stated till they are beyond foal age, and thus some possible errors are avoided. In dealing with Thoroughbreds a number of the more important sires was selected, and the colours of the dams with which they were mated and of the foals produced noted down. It was hoped that in this way the gametic composition of the sires might be ascertained and assistance gained there- from. One drawback was that, while there were plenty of chestnut, bay, and brown sires, only two grey sires (Grey Friars and Grey Leg) and one black (Desmond) having a fairly large number of foals in the latest volumes could be found. Another drawback was the large number of stock entered as “bay or brown” and ‘ brown or black.” These doubtful colours were omitted, although it was found necessary to consider them afterwards. The following are the figures got in this way from vols. x1x. and xx. of the Thoroughbred stud-book :— MATINGS IN VOLS. XIX. AND xXx. OF THE THOROUGHBRED STUD-BOOK. | Cotouxrs or Dams. (Clos ae |] Laas este glee ba Sires. Chestnut. | Black. Bay. Brown. Grey. Ch. BI. By. Br. Gr.| Ch. Bl. By. Br. Gr.| Ch. Bl. By. Ir. Gr.| Ch. Bl. By. Br. Gr.| Ch. Bl. By. Br. Chestnut, .|508 - 8 - — Qe il iil S287 2 A Ge) aloe We 72 Ss) des i = a Black, . . eee Oe rir |e 8 6-) = 1 2 6 =| == = = = Bay, . -|469 — 586 48 —/ 14 1 383 27 = 1/270 1 1295 125 —| 71 3359151 -| - - 2 — 6 _Brown,. .| 99 10 278 100 — = 8 12 20 —| 627 385 205 —| 11 6 78 114 -| —- —- 1 — E | CH a tof Wa Be a ay at) ese hl I ey OS an ey) PA so) she eye 2) oe ea eal S These figures are, if anything, more confusing than those got from Shires and Clydesdales ; but a very pertinent discovery was made during the course of their compilation. It became apparent that the only black sire of the lot, Desmond, was breeding in a very peculiar manner for a black. It was against all experience that a sire should reproduce his own colour only three times in thirty-seven. Besides, the colours of his foals suggested that he was not a. black at all, but a brown. Desmond stands in ireland, and several gentlemen likely to know him maintained, when appealed to, that he was a brown. When the question was put to one of these, “Is Desmond a black or a brown?” he replied, “He is brown: he has a tan muzzle.” Finally his owner’s Lord Dunraven’s secretary wrote “ Desmond is a brown horse.” Desmond was entered “ brown or black” as a foal, but this designation was subsequently changed to black, 336 Scientific Proceedings, Royal Dublin Society. This discovery with regard to Desmond suggested that inquiries should also be made about the black mares involved in the Thoroughbred table. It will be noticed that they breed very like Desmond when mated with chestnut, bay, and brown stallions. Presumably therefore they are also browns. Inquiry was made about a number of these mares, especially those having several foals, and in every case where a definite reply was received the mare turned out to be a brown. This led to inquiries further afield—to horses not concerned in our tables—and the conclusion arrived at is that most, if not all, throughbred “black” horses are really browns.' The authorities of the Trakehnen stud-farm in Germany desired at one time a black thoroughbred stallion to cross with their black mares, and they only found such an animal, or, at any rate, what they took to be such, with the greatest difficulty. There being thus a lack of blacks among Thoroughbreds, it became necessary to return again for such animals to the Clydesdales, among which, in their early days and in recent years, a fair number can be found. By avoiding sires whose blackness is doubtful, the following statistics were collected :— PROGENY OF TWENTY-TWO BLACK CLYDESDALE SIRES. Oy Eb. DB. Be Ge, From Chestnut Mares, . : ; ce, 5 3 6 - » Black Mares, - 36 - Pp} - » Bay Mares, 62 29 900 44 - » Brown Mares, i Hl | BO - » Grey Mares, 2 2D} se Siler 3 5 The bays and browns from chestnut mares, the browns from black mares, and the browns from bay mares were disturbing; but their numbers did not seem outside the error that might be expected in view of the proportion of brown and black mares and foals wrongly described by their breeders. A corresponding set of figures was compiled from the Shire Stud Book :— PROGENY OF TEN BLACK SHIRE HORSES. Ch. BI. By. Br. Gr. From Chestnut Mares, : } oe un I) STA 1 = » Black Mares, . 3 : A= Qo 189) = = = » Bay Mares, : 5 ! 5 Wey I) Be IS - » Brown Mares, 3 2a) id = 2 = xrey Mares, : : 3 Shel 1 2 2 9 1 Since this paper was read diligent search has been made for undoubted black Thoroughbreds, but without success, Witson— The Inheritance of Coat Colour in Horses. 337 Assuming the discrepancies referred to above are the result of confusion among blacks and browns, the appearance of so many chestnut foals from undoubted black sires makes it obvious that black contains chestnut; that is that chestnut is recessive to black, which, as we have already shown, is recessive to bay and brown. We should next have had to inquire closely into the true colours of the mares and foals involved in the last two tables—an inquiry that could not have been satisfactory because of many of the mares and foals being now dead and forgotten—but we were saved this by the discovery of a note on the point at issue in the English translation of Herr von Mittingen’s recently published work on “Horse Breeding.” Herr yon (ittingen is director of the royal stud at Trakehnen in Germany. He tells us (p. 329) that for over a hundred years blacks have been bred in one special stud (Gurdszen, 90 to 100 brood mares), browns in another (Dranzkehmen, 70 to 80 brood mares), and chestnuts in another (Jonasthal, 50 to 60 brood mares) ; while ‘in two studs (in T'rakehnen 80 to 100 brood mares, and in Bajohrgallen 60 to 70 brood mares) all colours are represented and mixed with each other.” Then having pointed out that at T'rakehnen they have thus “plenty of material at hand from which to construct laws as to the transmission of coat colour,” he proceeds :— ‘There exists a distinct regularity with grays, chestnuts, and blacks as regards transmission. ‘This regularity is as follows: grays and chestnuts mated only to their own colour, produce either chestnuts or grays, and black with black about 8 per cent. chestnuts (often dark chestnuts), the rest always blacks, never black-brown or dark brown.” This statement confirms our conclusion that black is dominant to chestnut, and also the other that it is recessive to brown. It also confirms the opinion already expressed that most if not all “ black” Thoroughbreds are browns. Herr von Gitmann’s observation that greys with greys breed only greys and chestnuts is no doubt correct ; but in two mixed studs containing in all from 140 to 170 mares there could not have been enough grey mares and sires from which to draw a general conclusion. The relative positions of bay and brown remain to be settled; and although there is evidence in favour of brown being dominant to bay, this conclusion is not clearly established. It must be remembered these are the colours breeders have the greatest difficulty in discriminating; and errors affect sires and dams and foals. In regard to sires it has been possible to correct the registered colours in several cases; and while every correction has increased the evidence in favour of brown being dominant, it is still 338 Scientific Proceedings, Royal Dublin Society. possible there may be some other explanation, as, for instance, that bay is a diluted brown. Unfortunately we cannot experiment with horses as we could with smaller animals. Extracting them from the three tables already given, the data in reference to the bay and brown matings are as follows, greys being neglected :— Colours of — Chestnut. Black. | Bay. Brown. Parents. | GhaBl By. bre Cheb lb yb ©hvals] eb yam brale Ghee ls Shires vols 1-0 ive, © = = =|) Bay, 9. =) 4823) 89oN Si 1934. 76) 425) 21 248523) leet sialss Sams, WOR Is WO)2o, 6 a 5 oll op 5 a) Oye Me) ay) aI) ae) eS) TB EY 5 23 183 Thoroughbreds, vols. xrx. and Eval 40 a ofS OHS TOF 4) 1! WW ZSL SBI KO) al key alsy |) 1123 il) 7/212 Shires, vols. 1. to1v., . . . .|Brown,. .| 13138 44 11) 2 28 28 35] 1118 168 77| 1 8 SRS, Wl ie (OS65 6 6 6 6 a > oll MAI 43 Bi 2 89 19 3B | i 283 1188 66 Mf Thoroughbreds, vols. x1x. and xx., | An 5 6 PAO ly? za Le) 2 = 8 12 20 | 128 10 744 356) 11 6 78 With regard to this table, it must be remembered that probably all the black Thoroughbreds are brown, and also that some black Shires are probably brown and some browns black. When thoroughbred bays are mated with bays there are 270 chestnut foals, 1 black, 1295 bays, and 125 browns. Assuming no misdescriptions, this suggests that bay is dominant to both chestnut and brown. This would place brown between bay and chestnut. But in that case, in a mixed population of browns, bays, and chestnuts, there ought to be a larger number of brown foals than 125. The same remark applies to the Shire figures. Again, when thoroughbred browns are mated with browns there are 11 chestnut foals, 6 blacks, 78 bays, and 114 browns. This, on the other hand, suggests that brown is dominant to both bay and chestnut. But one of these suggestions must be wrong. The latter has the greater semblance of correctness: that is that brown is dominant. Apart from the suggestion that a diluting or saturating factor may connect bay and brown, the figures in the table suggest the possibility that some bays may be hybrids between chestnuts and browns: a suggestion which is upheld by the breeding of brown with chestnut, viz., 205 chestnut, 17 black, 452 bay, and 172 brown foals among Thoroughbreds. But again, if this be the explanation, the bays when bred together ought to produce more than 125 browns, even assuming that a larger proportion of them are pure and not impure bays. Witson— The Inheritance of Coat Colour in Horses. 339 The explanation seems rather to lie in the error of misdescription ; and the point will be brought out by an examination of the table on the three following pages, which shows the results of the matings of a number of individual sires with mares of their own and the other four colours. The Thoroughbreds are taken from volume xrx. of the stud-book, the Shires from volumes 1. to x., and the Clydesdales from volumes 1. to xxx1. Among the Thoroughbreds the foals entered as ‘“‘ bay or brown” are given under the bays and enclosed within brackets, those entered as “ black or brown ” under the browns. The probability is that many bays are browns and browns bays—the wish to have them one or other occasioning some doubt—and most, if not all, of the thoroughbred “blacks or browns” are really browns. The registered colour of every sire is given, and, where it has been obtained, the corrected colour is noted below. At the same time the probable gametic composition of the sire is indicated. On looking at the cases of Florizel II., Baron’s Pride, and Baron of Buchlyvie, whose colours, according to authorities, are not those given in the stud-books, the effect of a misdescription in weighting one colour at the expense of another will be seen. ['TaBLEs. SCIENT. PROC. R.D.S., VOL. XII., NO. XXVIII. 3B1 Scientific Proceedings, Royal Dublin Society. 340 ‘sXeq [|B 010M oTIQnP PUOWLIG, pur UOMMUISIEg Ss1oyjOIg 044 Sly pue “TT joztIo[q shes Ayeq some -1Yy z ‘ayepsopAT toyz (Q) ‘oaryg 10F (g) ‘parqySnox0yy, 10F spuvys (7) , rae eta a) Se ee ee lea - g1 {aq “4q * (L) SIT PHT eg IE a eee ele I = 3 Nolet “q : (I) ‘weaouog eee eng ees as o Basle te he Sats es han} Aq > (I) Sg Soysng ee ee re es er ee 2 = t|=6 6 = «| {oe} “£q * (D) ‘epuy prog SN ES een ae oy eee ee eer hele “£q * (q) ‘10;Suray Sea | Seta) nee eo ee eal el 14 * (9) Gsexeag arg soe = =fS8l= oe 6 =|] 8 0 Gs |ot = 6 eles 4 < & at 14 "+ (9) foronjeg See eer epee (om OS i aE ay Pore ae ete es eS ea 14 * (9) ‘aoqstepy ou, [SSeS] 1.8 Glo e oe PIs = = @ Tle s = So] 1 19 * + (g) ‘quedg ge = be Or Ts eae ese pe et aH "yo | * (8) 48eA4 049 Fo topu0 Se See © ° eel ee ere tee is foe 4 Hah "yo > (L) ‘PregTaTA eh es eg " Teas Be Ee aaa ae ee ne Kash “yo * (1) PIA 0101, oS ee at Bee So ee ee ea ees ae) (I) “IT eur poy eg | SUA 8 ee es ae a aes Hoh "yp * (L) ‘Apost ay e407 - - - - -|-¢ ¢ - e|- i, ee eu tS aie ee ene, oe oo Neat “yo * + ({) ‘uo,8uysy Sere: || = fe Gere Conon ee a ee oe ee 0g tata “yo "+ (q) ejaureg SSeS SSS 5m e- fle =i SS = Ss SSeS 7 Ss a yeas “qo * + (q) ‘overs ef ss sl= 2 7) Pf B= 6 OG = Wil] =- 6 = gles = = wf ee “qo * (1) ‘uorydmy a) 4d Aq ‘ld Wd} 49 1g 4a 1a 4 | 49 3a “Aa Id YO | 2D 3G “Sq lag 40 |49 3a ‘4a ‘la ‘40 -hory “mMoIg | hog “yorrg “qnuysoyQ Srortecdnien Be : oyemeyg | samojog _ IO SUNOTOD GrUxaLSIOAy AU ee “SWV( “sso 341 Horses. Yr Un Wi.tson— The Inheritance of Coat Colou ,¢ Seq yep oq 03 pourpout ‘Aeq pifos pood B,, st es10y sty} sAus oSvpleN OVW ‘IIL < «UMoIq 10 Lvq Yep ,, B st estoy siqy shes ‘Ayo100g osi0y ofepsepATO 94} Jo Arejzor00g oy} ‘oSuploN Oey “ATL 1 “SVT 40 SUNOLO/) GaNaLSIoay -sIdoy == = 2 ele g 4 = S20 v - 1 |- wg eer eens ene “Aq ]* + (a) Sysnayy, pessrpy SS oS elo Rie = =! 2 ms = l=r 6 ft Sls = Gt = t oat fy foo. (in) Sepa == 6 = 60M 8 elo @ & 8 6 i)i- = 6 b wiles f 1] eg PY] £4 [+ (@) tavyonateg yo conn Fa ee SG CI Gite 6 = oral eS ger) eG real cachet Great ee eee = {-R} eq | * (0) “opr s,uoreg = = 2/203 6 8 =|] 6 8 = |e @ = = slew = - (ae rig | + (9) Sordqyong go uowg ee a CC ee ee OG ea z ont ‘kq |+ (9) ‘uoovig, Jo oprrg SS = = sie g 3 === 9 @=2)= i 6 = sle>5 iW Or Hone ‘hq |> * (q) ‘hoptodiey, ee = = oi Gg : Pe aaa AU SS eae lee ea! fof ‘hq | * — (q) opaay ag S92 = = ols 6 gy - 9/- - @& -% |-% - - -|-1 # ol {a} ‘hq fs + (q) Aemy qy8ny PAI aa i a = T= ti = ¢ |= t @ s-sl= @ P rat oa “hq |* (1) ‘Avpqgarg s,uoang eRe a Ge i ale 8 le fl - ot ele hq fs (q) ‘wommmsreg (3S ee Se ¢ n - t|- ¢ te peeps ba tr - #1 edt hy je 8 8 (in) GxIe@ @-- - -|-¢ pt e[-- w-6 |-- - - -|-1 6 - ew] {if} ‘hq fo + (q) ‘ouoawry | 1G ny Sp ee ee ee Kotat hq |* ++ (q) ‘sse[surey Sp eres ee - 9 e Slog f= l= @ a MI edad ‘hq |* + (4) ‘very Surdozreg 4D tq hE IEG | 49 wa Sa ta YO |39 sa Sa Ta “HO | 40 “Aa “Aa Ig “WO |19 “Ja “Aa Id “49 “hoary UMOLG keg “oul “ynuyse) “uortsodwm0g ae —]| onemey | sanojog — eiqeqorg | po10y 3812 Scientific Proceedings, Royal Dublin Society. 342 = ees 36) B = slg wes lo 2 t tf sie t 2 = = et “38 eae ae Bet Pts cal Poles Oe= Cra Ge gases mls | SG meena 0 at “13 eae 2 a — gw TE “yo = I Z =|8 © 8 I Se = aM ae oy : 13 (1) () () (@) al = - = - =f § fT @l= oO BeBe l= = 7 T el] 6 aes Norell “aq (3) (1) (#) Cd) “iq ess 6 sl69 2 = Sle @ eg seloe ft = sfoo m= sy fou “1q (2) (1) (¢) (6) Eq s£2 6 clomp t Slee wesl=@ = = slen we si heal | on (2) (2) (8) () “q So = SG 6 f s[20 @ ese = = =]25 8 er te | ae Se St = Ore Ure ft 2 We See! BSS ip a =o ea “14 t= ss sme melom me 2S lets .f afer » = Ss} fy “34 eee et ee ee oe ee oS oar aun = Sh @h ola @eo|24 5 = Sle6 6 6 =) Yh | aa ‘14 : a Soe & Slab oS oat 6.8 Slate So 1q (z) (1) (s) (1) (1) {ol i oS SOO ie Peed SEALE eet Sls =|= or 1m = oj {2} “14 g z ae St? = =m e = tlee we mes 2 = slo) @ =o] Ahh foo A i 8 t = S@ @ 7 leo Mee l= 6 6 2 S|> > 8 = se on “Aq = = eer ee = S le ee “Aq a 4 & Bh 16 gi Of = i 9 8 F 7 1 he q “A paGuise = Sgt = olo6 Moo l=6 ¢ 4 S68] 6 24) al “Aq A I= © = 56 Se leg we seels Ss we s/55 p92] ia “Aq (1) (@) (1) ( -£q ae Eee es SAF ene eae AS Gea en 5 SM | ES 9 oC) aE Homey “Aq sg ig ‘Aq ‘Ig “YO | 19 ‘Jd “Sad lad “GO | ‘19 Fa ‘sq 1g 49 | 39 sag Sa 1a 4 | 49 Ja “Aa ‘Id AO re s ; “Soitg korn UMOIG eq Wr qnuyseyy -uoytsodmog jo atjeuey | sanojog arqeqorg | pero, “SHV 10 SUNOTOD ATAALSIOAY -“s1d0qy (g) “TT pey oarysupooury : (L) ‘Seq foxy - (jg) ‘sretng Song * (h) erp s qT (1) ‘aourg “3g (Z) ‘sepey (Z) ‘puowseg (9) ‘Agtey puepoo + (0) ‘preroagy stg : (9) “eyuepoaory (0) ‘seTe Ay Jo court * (9) ‘Aqny woreg * (q) ‘umbsnay “4g * + (q) ‘onygsity : (9) ‘10Sangoe yy * + (9) ‘Aaquaeg f + (g) frermerg * (g) ‘og, ysouoxy > + (q) Spro8ueg Witson—The Inheritance of Coat Colour in Horses. 343 Perhaps the most interesting colour of all is the one set aside earlier in the paper for separate consideration, viz., roan. There are only a few entries of roans in the Thoroughbred and Clydesdale stud-books; but inquiry brought out the fact that there are more horses of this colour in both breeds than the stud-books would make out. It is an unpopular colour among Thoroughbreds and Clydesdales, and when a breeder sees, say, a bay foal “with a grey hair through” its coat, it is easy for him to believe the foal to be a bay and not a roan. On the other hand, roans are not altogether unpopular among Shires; and a good many are to be found in the stud-book, especially in the earlier volumes. We were therefore driven back upon the data collected from the first ten volumes of the Shire stud-book for examples of the be- haviour of roan. It has been stated already that, in collecting these data, several theories suggested themselves; and the one that roans are hybrids between grey and some one of the other colours seemed a priori the most likely. But greys never, or at any rate very seldom, and then doubtfully, produced roans with those other colours; and the crossing of roan with roan was too infrequent to give any guidance. LHventually it became evident that roan was separate from grey, but that its behaviour towards the other colours was similar, while at the same time it had a peculiarity of its own. Grey, as we have already found, is dominant in the other colours—brown, bay, black, and chestnut. This means that, in a population of the above colours, there can be no grey foal without, at least, one grey parent, and that, if the pedigree of a grey foal be followed backwards as far as it is known, a grey ancestor will be found in every generation. The following examples might be given :— A THoROUGHBRED Exampte.! A CLYDESDALE EXAMPLE. The Drone, grey (foaled 1823), < Kiss (foaled 1827). Blyth, grey (foaled 1836 or 1837). Trish Birdeatcher, X Whim, grey. Clydesdale Jock, grey. chestnut. Chanticleer, grey, X Birthday, bay or brown. Donald Blue, grey, X Mare, brown. Newminster, day, X Souvenir, grey. The Duke, X Mare, grey. chestnut. Strathconan, grey, X Sweet Violet, day. Sultan, XX Mare, grey. black. Linnaeus, grey, X Dulcie Agnes, brown. Glengarry, X Belle of Burnside, volow not stated. bay. Linny, grey, X Lady Blue, brown. Mains of \ x Rose of Meikle Folla, grey. Airies, § brown Sidestrand, grey (foaled 1903). Rozelle, X Blue Bell of Meikle Folla, grey. = bay. Rozabelle, grey (foaled 1907). 1 Sires are placed to the left, dams to the right. 344 Scientific Proceedings, Royal Dublin Society. The following pedigrees, similarly followed out, of the first three fillies and the first three colts met with in the last published volume (xxx) of the Shire stud-book will show that the roan has at least one roan ancestor in every generation,’ and that therefore roan behaves as a dominant to the other colours :— Tory, ved roan, X Bounce. Flower, roan. Roan Jumbo, roan. King ct Keele, X Darling, voan. ay. Muckton Nonsuch, day, X Muckton Violet, soan. Anottingley Royal, X Bankone Darling, roan. ay. Blythwood Conqueror, éay,. X North Cotes Primrose, ved roan. Filly, voan (foaled 1907). Ashwell Conquering Prince, bay, X Garston’s Lady, blwe roan. Filly, bay or ved roan (foaled 1908). Horse, roan. Samson, brown, X Mare (colowr not recorded). Emperor, 70an. England’s Wonder, roan, X Mare, roan. Carbon, brown, X Bonny, oan. Monaco, bay, < Trim, roan. Baroness, roan. Harold, drown, X Blue Bell, blue roan. Horbling Harold, blue roan, Primula, brown. Hitchin Magnet, day, X Silver Tail, blue roan. Anglian Harold, roan. 1 The Shire stud-book is comparatively young, and so the pedigrees are not long, Witson—The Inheritance of Coat Colour in Horses. 345 Vulcan, Slack, X Kit (Capes’), red roan. Prince Harold, 6lack, % Dunsmore Fashion II, roan. | Dunsmore Iron Duke, ean, % Lemington Royal Heroine, day. Forage Iron Duke, roan. Hereward, day, X Willow Lady, 7voan. | | Pratt’s Wallflower, black, X Willow Bounce, blue roan. | Blue Duke, blue roan. The comparison between roan and grey might be carried further. Grey is dominant to the other four ordinary colours. This means that when grey is crossed with these colours the number of grey foals ought to equal all the other colours together ; excepting that, grey being somewhat unpopular among Shires, a higher proportion of grey foals than of the other colours may not be entered at all, and thus the numbers for grey may be depressed. ‘I'he numbers from selected sires in the first ten volumes of the Shire stud-book are as follows :— Grey X chestnut gives Grey X black 99 Grey X bay 99 Grey X brown _,, Total of Oli, 1H A ae these 4 Gr colours. 8 13 14 3 = 35 18 2 6 g) 13 = 30 24 4 6 60 11 ~ 81 56 - 5 31 21 = 57 36 The figures for roans--iron-grey being counted among the roans—the progeny of roan mares in vols. x1. to xv. are as follows :— Roan X chestnut gives Roan X black 4 Roan X bay “9 Roan X brown Roan behaves therefore to Total of Qin, WL By ie these 4 Roan. colours. 7 2 7 2 = 18 8 - 5 - 2 = 7 12 4 2 30 ll = 47 41 1 2 14 7 = 25 20 these other colours very much like grey, their 346 Scientific Proceedings, Royal Dublin Society. dominant. But there is a very peculiar difference. The grey colour domi- nates the other colours below it entirely ; roan does not. The grey foal of a bay dam and a grey sire is grey; but the roan foal of, say, a bay dam and a roan sire will have the body a mixture of white or grey hairs in a back-ground of bay or any of the other usual colours, while, so far as can be seen as yet, the leg-markings will be those usually associated with the back-ground. A bay or brown roan will have black “points,” the legs of a chestnut-roan will be chestnut, and so on. And still further: the back-ground colours seem to behave to each other, so far as dominance is concerned, as they do when no grey hairs are present to make them roans. ‘That is to say: two chestnut-roans seem to produce only chestnuts and chestnut-roans, while two brown-roans seem to produce brown-roans, bay-roans, black-roans (perhaps also blue-roans), chestnut-roans, and another colour, strawberry-roan, which seems to depend rather upon the quantity of grey hairs present than upon the colour of the back-ground. It would almost seem as if the grey hairs in roan parents allowed the back- ground colours to work as they do ordinarily, but that they introduced them- selves to the coat of about every second foal. ‘These ideas have been suggested while tabulating the results of roan matings, but the data are still too few to say they are proved. The source of roan, that is of the white hairs that make other coats roan, has not been found. Nor are the relative positions of roan and grey clear. Roan behaves like grey with the other colours below grey; but the crossings between grey and roan are too few for any inference. Meantime let us set down a table showing the results of crossings with roans (vol. x1. to xv.), in which the roans are subdivided into roans, blue-roans, and iron-greys (which are probably grey-roans) :— Colours of TSR. Roan. Blue-Roan. Iron-Grey. Ch. Bl By. Br. Gr. ee En. Rn|Ch. Bl. By. Br. Gr &. Re Rn.|Ch. Bl. By. Br. Gr. G. Be Rn AN Ote faz) roe nl ie eae Sele wenn ue neta, Sy ny | 1 ie eee) UMA ee OTe ad Rew ING) Pacha eager ea ee a ne fl So = 4728480; 1 ah Ai eB Ones 1 aes ee Pa GB Se Lr To SQ a eS 2 T'9) | 2 9 Sb Se Pea er | = Beal CMe Te G essary Opa yal peace es ee ee = Tron-Grey, toy Wen eae eee ee Se eee Ib es SS, Blue=sRoans sili 2) a een a a 2 = i ny oe - - puis ee a ee SS ee Witson—The Inheritance of Coat Colour in Horses. B47 It will be seen from the above table that a few greys are produced when roans are mated with ordinary colours. If we could be assured that these are really greys, and that the parents are correctly described, the problem of the relative positions of roan and grey would be solved; but such assurance is impossible with colours so liable to confusion as roan and grey. Ifa similar table of grey matings with ordinary colours were made out, it would show the contrary phenomenon of roans being produced by the greys, and so would suggest that grey is dominant to roan. For the present, therefore, the relative positions of grey and roan must remain undecided. One other colour, dun, remains; but although some dun entries were seen in each of the three stud-books drawn upon, they were far too few for our purpose, and were not noted. But from statements published by Weldon and Cossar Ewart, and from information given by Mr. P. Macginnis as to foals produced by dun sires in county Londonderry, the following small table can be formed :— 1 dun X 1 chestnut gives - - 1 - - - 1 1 dun X 1 black 9 - 1 - - - = ® 1 dun X 2 bays 59 1 1 1 - = 1 1 1 dun X 1 dun 3 1 - 1 = = a 91 2 dun X 6 grey mares ,, - - - 2 2 - 4 It is apparent from the first two and the last of these matings that dun contains, i.e., is dominant to, chestnut, black, and bay. The matings of dun and grey give no indication. But from information received on Clare Island last July it seems certain that dun is dominant to brown also. Some time ago a dun Norse sire which was brought to the island left a number of dun foals from a population consisting chiefly of bay and brown mares; and these dun foals when themselves put to a brown sire threw some dun foals again. ‘This places dun dominant to brown, bay, black, and chestnut; but leaves its position with regard to grey and roan unknown.? It may be well to reiterate the statement made at the beginning of this paper, that stud-book data are far from being absolutely accurate, and that the validity of our conclusions must depend upon the success with which the inaccuracies of the stud-book have been smoothed out. It is hoped 1 One of these is called yellow. 2 Since this paper was read a case has been found, in Professor Cossar Ewart’s paper on ‘‘ The Multiple Origin of Horses and Ponies,”’ of two greys producing a dun. -This would show that dun is recessive to grey. SCIENT. PROC., R.D.S., VOL. XII., NO. XXVIII. 3K 348 Scientific Proceedings, Royal Dublin Society. such data may become more accurate in the future. Meantime this paper may lead to further observation and elucidation of the subject. In addition to those mentioned in the text, the following gentlemen have very kindly furnished information on the subject in question :— Mr. H. R. Rose, Chantilly Stud Farm ; Mr. Andrew Robertson, Dublin; Mr. James Barrie, Balmedie, Aberdeen ; Captain Greer, Curragh Grange ; the Rt. Hon. Frederick Wrench; Lord Rossmore; Lord Bradford; and Lord Derby. THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 29. APRIL, 1910. CHROME STEEL PERMANENT MAGNETS. BY W. BROWN, B.S8c., PROFESSOR OF APPLIED PHYSICS, ROYAL COLLEGE OF SCIENCE FOR IRELAND. [Authors alone are responsible for all opinions expressed in their Communications. } DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON. W.C. 1910. qqnsomian instit,, ty [ve QO, “ Price Sixpence. SUN lo Toa No on “tional Muse Roval Dublin Society. ee FOUNDED, A.D. 17381. INCORPORATED, 1749. EVENING SCIENTIFIC MEETINGS. Tur Scientific Meetings of the Society are held alternately at 4.30 p-m. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Editor. 849) 1] XXIX. CHROME STEEL PERMANENT MAGNETS. By W. BROWN, B.Sc., Professor of Applied Physics, Royal College of Science for Ireland. [Read Frpruary 22. Ordered for publication Marcu 8. Published Apri 21, 1910.] THE magnetic properties of chrome steel were—as far as I know—first investigated by Dr. Hopkinson! in 1885, by means of his yoke-and-bar method ; the two specimens tested contained (1) Cr = 0°621, C = 0'532, and (2) Cr = 1:195, © = 0°687. If we take the density’ of the samples when in the hard state to be 7-76, and divide the residual induction obtained by Hopkinson by 47 times the density, we find the magnetic moment per gramme of the two specimens to be about 88 and 81 respectively. In 1898 Mme. Curie? in her paper on the magnetic properties of tempered steels, gives the residual intensities for three chrome steels obtained with bars of square section 20 cm. long and one centimetre in the side. Taking the bar which has the chemical composition Cr = 2°48, C = 0°5, and assuming the density to be 7°76, we get magnetic moment per gramme of about 59. If, however, we take the value of the residual intensity which was obtained when the sample was in the form of a ring, the magnetic moment per gramme is then 82, which result is more comparable with that of Hopkinson, whose method of testing was also a closed circuit one. In 1900 Barrett, Brown, and Hadfield‘ published amongst other physical properties the results of magnetic tests of two chrome steels containing 3°25 and 9°59 per cent. of chromium. ‘The tests were made on rods of dimension ratio 200; and from the residual induction then obtained, and taking the densities’ of the materials as 7-765 and 7:703, the magnetic moments per gramme come out 87 and 109 respectively. Ina recent test® of the same 1 Phil. Trans, Roy. Society 1885, vol. clxxvi., part 2, p. 463. * Scient. Trans. Roy. Dublin Soc., vol. ix., p. 67. 5 Bull. de la Société d’ Encouragement, pp. 36-76. 4 Scient. Trans. Roy. Dublin Soc., vol. vii., p. 119. 5 Tbid., vol. xi., p. 67. 6 Thid., vol. xii., p. 318. SCIENT. PROC. R.D.S., VOL. XII., NO, XXIX. on 350 Scientific Proceedings, Royal Dublin Society. two steels when in the form of magnets with dimension ratio 33, the magnetic moments per gramme were 36 and 31. In order to test the effects of gradually increasing the amounts of chromium in steel magnets when the magnets were all of the same dimensions, thirteen specimens of steel with different amounts of chromium were obtained from Sir Robert A. Hadfield, F.R.S., which were each about 5 inches long and } inch diameter. They were all turned down to 0°3 cm. diameter and 10 em. long, giving a dimension-ratio of 33, and in order to get them all in the same physical state, they were individually raised to a bright red heat in a gas muffle furnace at a temperature of about 900° C.,! and dropped end on into 3 feet of cold water. The magnets were then cleaned, weighed, and magnetized in a field of 2000 c.g.s. units by putting them on the poles of a powerful electro-magnet. About twenty hours after being magnetized they were tested for the magnetic moment per gramme by the usual magnetometric method, the east and west position of Gauss; and immediately after the first observations each magnet was allowed to fall once through a height of one metre on to a block of glass and again tested, and then allowed to fall three times through the same vertical distance and tested once more, the percentage loss due to percussion being thus obtained. The magnets were let fall in every case with the true north or south-pointing pole downwards. The magnets were then put into a steam chamber and annealed for twenty-five hours in steam at 100° C. They were then magnetized and again steamed for five hours; and about twenty hours after the final heating, they were all tested for magnetic moment and loss by percussion in the same way as was done when they were glass-hard. The results for seven of the more important of the thirteen specimens are shown in Table I., which gives the chemical composition’—the amount of iron in each being estimated by difference—the magnetic moment per gramme when glass-hard and when annealed, and the last two columns gives the percentage loss in the magnetic moment per gramme due to the magnets falling end on four times through a height of one metre on to a thick block of glass. 1 Determined by means of a calorimeter. 2 Determined in the chemical laboratory of the Hecla Steel Works, Sheffield. Brown— Chrome Steel Permanent Magnets. 351 TaBLe I. Macenetic CHEMICAL ComrosITIoN. Moment PERCENTAGE Loss. per gramme. No. | Mark.} C Si Mn Cr WY Cu ae Annealed See | Annenea 1 993 0:88 | 0°19 | 0:28 | 1:75 38:2 41-2 31 1:9 2 517 0°86 | 1:96 | 0°40 | 1°96 50°4 50°8 27 0°5 3 518 0-76 | 1:02 | 0-29 | 2:11 52:5 52°2 33 2:2 4 1185F | 0:54 | 2:20 | 0:22 | 3°50 41-7 41-4 9°3 U2 b) 1255A | 0°85 | 0:31 | 0°50 | 5-79 1:83 38:7 39°3 6°5 7:0 6 1233A | 1:36 | 0°75 | 2-60 | 9-22 | 42:2 43°3 lets) 1:3 7 1249A | 0°48 | 0-29 | 0-82 | 1°74 | 1:90 | 1-91 | 44-0 41°3 6:4 3:2 The annealing has had very little effect on the magnetic moment, though in most cases the loss due to percussion has been greatly diminished. In the paper already referred to, Mme. Curie states that small quantities of silicon have little or no effect on the magnetic properties of a magnet. Assuming this to be case, if we plot from the above Table the percentage of chromium as abscissa, and the corresponding magnetic moments per gramme as ordinates, the resulting curve rises smoothly to a maximum at a point which would represent 2°5 per cent. of chromium, then drops to the point at 3°5 per cent. of chromium, and goes along in a practically horizontal direction. The point on the curve for magnet No. 5 is a little low, and that for No. 6 is rather high, which results are no doubt due to the presence of copper in the one case, and of manganese and high carbon in the latter case. From this temporary curve it seems probable that the best effect would be obtained with a magnet containing 23 per cent. of chromium giving a magnetic moment per gramme of about 54, and that a further addition of chromium would reduce the magnetic moment. Comparing Nos. 1 and 2, which have approximately the same amount of carbon, the extra manganese and chromium in No. 2 have raised its magnetic moment by about 19 per cent. (in the annealed condition), and diminish»d its percentage loss about four times; and in No. 3 the slightly less carbon and manganese, and the slight increase of chromium, have raised its moment about 23 per cent. Comparing Nos. 1 and 7, which have practically the same amount of chromium and the same magnetic moment in the annealed condition, it 352 Scientific Proceedings, Royal Dublin Society. would seem as if the effects of the diminished carbon and increased manganese in No. 7 were just balanced by the presence of the 1:9 per cent. of tungsten and the same amount of copper. No. 7, however, is less retentive, as will be seen from the percentage loss being nearly doubled ; and when this material was tested in the form of a thin magnet 10 cm. long, and having a dimension ratio of 60, its magnetic moment per gramme, when glass-hard, was found to be 55-7, and the percentage loss due to falling four times through a height of one metre was 8°1 as compared with 6°4, when the dimension ratio was 33. On the whole, therefore, the specimen No. 2 would be the best material to use in making a permanent magnet. | | | | 60 40 Magnetic Moment per gramme. Dimension-ratio. A good length of wire, one millimetre in diameter, was obtained of some- what the same composition as No. 5, but without the copper, viz. (C = 0°77, Si = 0:5, Mn = 0°61, and Cr =5:19) ; and from this wire fourteen magnets were made, varying in length from 20 cm. to 1 em. These magnets were hardened, one at a time, cleaned, weighed, and tested for magnetic moment per gramme, exactly as in the previous cases; and the results when plotted with the dimension-ratios as abscissee, and the corresponding magnetic moments as ordinates, gave a very smooth curve, as shown in figure. From the curve we see that a magnet 10 cm. long, or of dimension-ratio 100, is the most effective magnet to make of this material, and has a magnetic moment per gramme of 65, an increase of 68 per cent. on the value for No.5 in Table I., which has a dimension-ratio of 33. Brown— Chrome Steel Permanent Magnets. 3993 To check the point on the curve corresponding to dimension-ratio 100, five magnets were made of the same wire, each 10 em. long; and the mean of the magnetic moments gave the value 65:1. There was no perceptible loss in the moment when these magnets were let fall end on four times through a height of one metre on to the glass block. To show that the magnetic moment per gramme varies with the form of the sample tested, the values obtained by different observers for chrome steels (other than those in Table I.) have been brought together in able II., in which the percentages of carbon and chromium only are given. Tasxe II. | Magnetic Experimenter. Cr C eee Form of specimen. gramme. Hopkinson, . : : 0°621 0°532 88 Closed circuit. 1 es 1°195 0-687 | 81 S: 3 Mme. Curie, . . 5 2°486 0-501 82 | - 9 09 2°486 | 0:501 | 59 | i : z Square bar 20 cm. long, and Ue CASEM OAoNe Ge lcm. inthe side. — 3°445 1°069 68 2 “43 Barrett, Brown, and { oa Oe 87 Round bar of dimension-ratio Hadfield, 9-50 1-09 109 200. | | : { U2 | ie ae | Round bar of dimension-ratio Brown, - 6 : 33 tl 950 1-09 31) ; Comparing the values of the magnetic moments for the last two specimens in this Table with the values for Nos. 4 and 6 in Table I., which have approximately the same amounts of chromium, we find that the magnetic moments are less by about 13 and 28 per cent. respectively. ‘This is, no doubt, due to the presence of manganese in Nos. 4 and 6, ‘The chemical compositions of the last two specimens in Table II. are not known beyond the amounts of carbon, chromium, and iron. SOIENT. PROC. R.D.S., VOLe XIIey NO. XXIX. 3M THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 30. MAY, 1910. ON THE DISTRIBUTION OF MEAN ANNUAL RAINFALL AND AVERAGE NUMBER OF RAIN DAYS PER YEAR OVER AN AREA INCLUDING THE COUNTIES OF DUBLIN, WICKLOW, KILDARE, AND MEATH: A SPUD IN LOCAL VARIATION OF RATING AI BY WILLIAM J. LYONS, B.A., A.R.C.Sc. (Lonp.), ROYAL COLLEGE OF SOIENCE FOR IRELAND. A _attnsonian Instity,: fe % NOV 26 1910 Mai [COMMUNICATED BY PROFESSOR W. BROWN, B.SC (PLATE XXI.: MAP.) [Authors alone are responsib/e for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATEH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1910. Price One Shilling. Roval Mublin Society. AKAIR ARR R AAA FOUNDED, A.D. 1731. INCORPORATED, 1749. EVENING SCIENTIFIC MEETINGS. Tur Scientific Meetings of the Society are held alternately at 4.30 p.m. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and speed Ip the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Hditor. Brown— Chrome Steel Permanent Magnets. 353 To check the point on the curve corresponding to dimension-vatio 100, five magnets were made of the same wire, each 10 cm. long; and the mean of the magnetic moments gave the value 65:1. There was no perceptible loss in the moment when these magnets were let fall end on four times through a height of one metre on to the glass block. To show that the magnetic moment per gramme varies with the form of the sample tested, the values obtained by different observers for chrome steels (other than those in Table I.) have been brought together in Table IT., in which the percentages of carbon and chromium only are given. Vase II. | Magnetic Experimenter. Cr C pane Form of specimen. gramme. Hopkinson, . : 5 0-621 0°532 88 Closed circuit. A ec eit ete LSS 0-687 81 ie Mme. Curie, . 5 : 2-486 0°501 82 a ae / 2-486 0:501 59 \ 2-831 0-819 a Square bar 20 cm. long, and DY as! lcm. in the side. 37445 1-069 68 9.95 . Lyf Barrett, Brown, and if vice Whe) 87 | Round bar of dimension-ratio Hadfield, \ 9°50 1:09 109 \ 200. 3-25 0°43 36 Mh ee RATE Brown, . : ; { | Round bar of dimension-ratio \ 9°50 1:09 31 Boe Comparing the values of the magnetic moments for the last two specimens in this Table with the values for Nos. 4 and 6 in Table I., which have approximately the same amounts of chromium, we find that the magnetic moments are less by about 13 and 28 per cent. respectively. This is, no doubt, due to the presence of manganese in Nos. 4 and 6. The chemical compositions of the last two specimens in Table II. are not known beyond the amounts of carbon, chromium, and iron. SCIENT. PROC. R.D.S., VOL. XII., NO. XXIX. 3M XXX. ON THE DISTRIBUTION OF MEAN ANNUAL RAINFALL AND AVERAGE NUMBER OF RAIN-DAYS PER YEAR OVER AN AREA INCLUDING THE COUNTIES OF DUBLIN, WICKLOW, KILDARE, AND MEATH: A STUDY IN LOCAL VARIATION OF RAINFALL. By WILLIAM J. LYONS, B.A., A.R.C.Sc. (Lonp.), Royal College of Science for Ireland. [COMMUNICATED BY PROFESSOR W. BROWN, B.SC.] (Prats XXI.: Map.) [Read Fesruary 22. Ordered for Publication Marcu 8. Published May 19, 1910.] INTRODUCTION. Tue distribution of rainfall in space presents very marked variation, not only when considered for short periods of time, but even in the case of the annual and mean annual precipitation. The distribution, particularly that of the mean annual rainfall, depends on the relief or orographical features of the country; and this relation to the orography is found even in the local variation within very limited areas. In preparing a map to show the distribution of rainfall over a large area such as the British Isles, only the general features of the distribution can be represented; and in such cases, in assigning a value to the mean annual rainfall of a small part of the area, the local differences peculiar to that part are necessarily levelled down so as to represent the average for the district. ‘The study of these local differences is, however, often of more interest than that of the general distribution over large areas, and presents a problem in physical, rather than in mere statistical meteorology. The immediate causes of the formation of rain are very obscure, and existing theories rest on physical speculations rather than on meteorological data. No satisfactory theory is likely to be developed as the result of an increase in our knowledge of the physics of condensation and allied phenomena, unless there is a corresponding extension in our meteoro- logical study of the subject. The local variation of rainfall seems an Lyons—The Distribution of Mean Annual Rainfall. BOO important problem in this connexion. It is reasonable to suppose that over a small area the conditions existing in the upper and middle layers of the atmosphere are practically uniform at any time, and the rainfall distribution must be almost solely decided by the orographical features; hence the relation of the one to the other may be worked out all the more definitely for such limited areas. In only a few cases has this detailed study of local distribution of rainfall been effected. Special reference must be made to the work of Professor Hellmann,! in 1900, on the rainfall of Hast Prussia; to that of Professor Schreiber® on the rainfall of Saxony; and to the study of Silesia by Professor J. Partsch. In England the rainfall of the Lake District® was investigated by G. J. Symons, F.r.s., in 1897. In recent years the Geo- logical Survey of England and Wales has issued a series of memoirs‘ on the Water-Supply of some of the English Counties. Hach memoir contains a valuable report on the rainfall of the county and a detailed rainfall-map by Dr. H. R. Mill, Director of the British Rainfall Organization. Particutars oF AREA STUDIED. The present investigation into the rainfall of the counties near Dublin was undertaken by reason of the extremely marked yet simple relief features of the area, and the comparatively large number of records available. It proved, as anticipated, a very interesting study in local distribution, presenting a considerable range, and very abrupt variation in rainfall. Apart from its meteorological significance, the precipitation over this district is of importance, as the area includes the catchments of some important rivers, and the collecting-grounds of the Roundwood and Glenasmole Waterworks, which supply the city and suburbs of Dublin. The limits of the area for investigation were decided, not by reference to the irregular county boundaries, but by considerations of the configuration of the country, and the records available. The boundaries are very approxi- mately the meridians of longitude corresponding to 6° and 7° west, and the parallels of latitudes 52° 42’ and 63° 54’ north. The length is about 82 miles, the average breadth 34 miles, and the area approximately 2800 square miles. ‘he region lies due east of the Great Central Plain of Ireland, and is bounded on the east by the Irish Sea. ‘To the south of Dublin the Dublin and Wicklow Mountains form a continuous and clearly defined 1 Rainfall Chart of East Prussia. Hellmann: Lerlin, 1900. * Zeitschr. fur Gewasserkunde, Bd. i., 1900, p. 48. 3 « British Rainfall,’? 1897. See also years 1867, 1895, 1896, 1898. + Memoirs on Water-Supply. Geological Survey of England and Wales. 3u2 36 Scientific Proceedings, Royal Dublin Society. system, rising abruptly from the sea and culminating in Lugnaquilla at a height of 3039 feet. In the extreme south-west corner of the area due west of Carlow the land rises rather abruptly again to nearly 1000 feet. On the north of Dublin the country rises very gradually in a north-westerly direction, passing from 100 to 250 to 500 feet and higher, as we approach the north-west corner of the area. Wicklow Head, Bray Head, and the Hill of Howth form conspicuous elevations along the coast. The area includes all county Dublin, and nearly all of counties Wicklow, Kildare, and Meath, with parts of counties Wexford, Carlow, and Louth. Mean Annuat NRalnFatt. In considering the space-distribution of rainfall, the time-unit selected is usually the year; and this was adopted in the present paper for the sake of uniformity with other work on local variation, and in consideration of the data available. The year has the advantage as a time-unit in such work, in that it forms a complete seasonal cycle as regards insolation or solar influence, which is the most fundamental factor in meteorology. It must be recognized, however, that for many purposes the annual rainfall is not as important as that of shorter periods, and that in comparing the rainfall of two stations, it is possible to get the same annual precipitation for the two; yet the frequency, intensity, and average monthly distribution may be quite different in the two cases. Both the agriculturalist and civil engineer are more concerned with the rainfall over short critical periods than with the total for the year. There scems little probability of correlation between annual rainfall and such agricultural statistics as crop-yield; yet it 1s con- ceivable that such a correlation might be established with rainfall over certain critical months of the year. The engineer, in questions of water- power, town-supply, and drainage is concerned chiefly with fluctuation in rainfall, and especially with critical maxima and minima within lmited periods. It must thus be admitted that any complete discussion of distri- bution should take account, not only of the annual, but also of the monthly or seasonal rainfall. The term ‘‘ Mean Annual Rainfall” is open to different interpretations and needs some consideration. Apart from the looseness of its use in popular language, the word “mean” is used in different senses in scientific work. In dealing with a number of different values of the same quantity such as occur in experimental results or statistics, the mean value may be determined in different ways. When the quantity by its nature has neces- sarily an exact value, and the differences observed are due to “errors of observation,” the mean of the observed values is intended to give the “ true Lyons—The Distribution of Mean Annual Rainfall. 50¢ 2 value” or “most probable value” of the quantity. ‘I'o determine this value the observed numbers may be arranged in order of magnitude, and the middle term or ‘‘ median ” selected as the mean ; or the observed values may be divided into groups, all in each group having the same value. ‘Ihe value corresponding to the most numerous group may be taken as the mean (in this case sometimes called the “‘mode”). Lastly, the arithmetical or geo- metrical mean of all the observations may be taken, or a still more elaborate process of reduction adopted. In such cases as the determination of an exact physical magnitude the above methods will give results almost identical for the true value of the quantity, provided the observations are sufficiently numerous and trustworthy. This agreement is to be expected, and follows from the general “ Theory of Hrrors.” In dealing with such quantities as annual rainfall, however, the problem is somewhat different. ‘lhe quantity by its nature has no exact value, but varies within limits more or less wide. The expressions “true value” and “most probable value ” are meaningless, and the results obtained by different methods of reduction differ widely. It becomes a question of doubt and difficulty in all such cases to decide on the most satisfactory method of treating the numbers, and the mean becomes largely a matter of definition and convention. The ‘‘mean annual rainfall” is generally regarded as the arithmetical average of a sufficiently large number of consecutive years. Suggestions have been frequently made that the geometrical mean would be more satisfactory. Any theoretical advantage of one “mean” over another is, however, rendered very doubtful when we consider the great and irregular variation in annual rainfall and the very considerable sources of errors in all rainfall measurement. The arithmetical average of the annual rainfall for a period of years will differ for different periods, the differences being great when the length of the period is short. As a general rule, as the length of the period increases, the mean assumes a more constant value; and the limiting value to which it approximates for a very long record is regarded as the true mean. It has been found by Sir Alexander Binnie’ in a very exhaustive analysis of several records, extending from 50 to 97 years, that the mean of about 35 years was practically constant and probably differed from the true mean by not more than 2 per cent. Means computed for periods of 40 years have shown greater differences than those for 35 years; and it appears that a very much longer period than 35 years should be taken to give a more satisfactory result. ‘This interesting conclusion was also reached by Professor ' Minutes, Proc. Inst. Civil Engineers, vol. cix., p. 92. 398 Scientific Proceedings, Royal Dublin Society. J. Hann! in his analysis of the very long records of Padua, Klagenfurt, and Milan, and was further confirmed by Dr. H. R. Mill’ in a study of several long English records. It has been suggested that the particular value of the 35-year period is due to a periodicity in rainfall connected with the 35-year Brickner cycle. THe Recorps anpD tHE Meruops or Repucrion. The records used in the present investigation were, without exception, those published in the volumes of “ British Rainfall,” the annual publica- tion of the “ British Rainfall Organization” founded about fifty years ago by the late G. F. Symons, F.R.S. The Organization at present includes nearly 5000 observers for the British Isles. The gauges are open to occasional inspection, and all the returns are critically checked. Any doubtful return is referred back to the observer ; and if no correction is made, and the doubtful - character remains, the return is queried if published or omitted altogether from publication. There were available for the present paper seventy-three records of varying duration, as shown by the following figures :— Over 35 years’ duration 5 Records. From 30 to 385 BS ‘5 1 .; 25 ,, 30 3 s 4 . 20 ,, 25 as i, 4 5 15 ,, 20 A % 10 33 O » le 53 35 12 ee a) x I 5 * 15 5 Under 5) ee ie 22 op The fundamental period adopted in the case of the long records was the 35 years from 1873 to 1907. It was considered desirable to end with the year 1907, so as to omit the very exceptional wet year 1872. The method of deal-- ing with the records of less than 35 years’ duration so as to determine a mean on the 35-year basis was that well established in rainfall work. The rainfall over the short period was considered for the station in question, and also for one of the nearest long or fundamental records. The short-period average for the latter was estimated as a percentage of its known 35-year mean; and the short period average of the first station was increased or diminished in this part ii. a. * Minutes, Proc. Inst. Civil Engineers, vol. civ., p. 299. Lyons—The Distribution of Mean Annual Rainfall. 399 give very consistent and satisfactory results for records of even a few years’ duration, especially when the fundamental stations with which they are compared are not far distant. The assumption underlying the above method is that, for stations within a limited area, the rainfall varies similarly from year to year. ‘his assumption is seen to be justified by plotting on the same sheet the annual rainfall for such stations for a number of consecutive years. Such sheets were prepared respectively for the stations in each of the following five divisions—North Dublin, South Dublin, Central Dublin, Wicklow, Meath, and Kildare. The graphs showed a marked paraltelism for stations close together; and even for stations far apart and very differently conditioned, there is found a similarity in the run of the annual rainfall graphs. This is illustrated in the diagram, page 360, where the annual rainfall graphs are given for a few selected stations for the period 1885 to 1905. In reducing to a 35-year mean the records of a short-period station, the computation was independently made by reference to at least three long records. These long records included one, and, if possible, more than one, of the 35-year stations, or otherwise the longest records among neighbouring stations. ‘The greatest care and circumspection were used in selecting such stations of reference, and special attention was paid to similarity of conditions of the stations compared, and to the correspondence observable in their annual rainfall graphs. In one or two cases where a value for an individual year was missing in an otherwise complete record the value was estimated by interpolation, or by careful consideration of correspondence with the graphs of other stations. In only a few cases did any great discrepancy appear in the means computed by reference to different fundamental stations; and the results in such cases were most carefully considered. Sources oF Hrror anp ACCURACY ATTAINABLE IN RAINFALL MEASUREMENT. The values of mean annual rainfall submitted in the accompanying tables are the results of computation based on actual returns, and no correction of any kind has been attempted. ‘The “catch” of a rain-gauge is subject to several disturbing factors, nearly all connected with wind effect, and dependent on conditions of exposure. ‘There exist in consequence many and serious sources of error in the measurement of rainfall; but the effects are so irregular that attempts at correction are necessarily purely speculative, and in most cases injudicious. The wind by forming eddies and currents at the mouth of the gauge tends to reduce the amount of rain caught. As the height of the gauge above the ground is increased, the wind and the above effects are increased, and the gauge will usually register less than if placed at 360 Scientific Proceedings, Royal Dublin Society. > S06 oe Sane | ! > L hoe / A a at sy eal es MA L | Ta Be ir = — e06l ‘ aS 2061 “SS ‘ a Se eect ae 106! Bs 2 Moe eae : =a zak 006! aly is a Bal ; etal as = agai aca n 968 ii 7a . Ae ic (ac =a 68! ara are ake efainine CLP ta loci eaaal ; ~ é <— ; G6 S; BY i. a eg Of Bel | eee. 681 aes 2 peas 2681 ; ; Al # Rae 5 laas ie 1681 Ss wy =! SI 1 fa “4 iN -E 068! a —+ t ‘ina } ? eee I} v SH ina So 988i NGCRGn SSUe. N E - eee /881 SISSISIS_ | fe ote 9981 RBS Sema ccs ec *SONOUL UT [[RJUTL YT Variation in Annual Rainfall, 1885-1905. Lyons—The Distribution of Mean Annual Rainfall. 361 a lower level. A standard height of one foot above ground-level has been recommended ; and it is estimated, on the results of many tests, that for every extra foot up to about 10 feet the gauge loses 1 per cent. in catch. A correction on this estimate is made by Dr. H. R. Mill in his rainfall reports on the English counties. This correction seems, however, legitimate only if the other conditions of exposure are quite uniform; and, for many reasons, it was considered undesirable to adopt it in the present investi- gation. The two most important records, namely, Fitzwilliam Square’ and Fassaroe, Bray, would have needed modification. The gauge at Fitzwilliam Square is placed in a confined area where there is great pro- tection from wind, though little likelihood of the gauge being protected from the rain. This fact probably explains why the record at this station is almost uniformly greater than that kept at Trinity College, where the gauge is placed (in the Fellows’ Garden)? at a lower level, but in amore exposed position. With reference to the Fassaroe record, where the gauge is kept at 4 feet 6 inches above ground-level, I am informed by the observer, Mr. R. M. Barrington, m.a., that the “catch” at this level was generally greater than that found for comparison gauges placed lower. The position of a gauge with reference to rising ground is as important as the actual height above the ground. A gauge placed on a summit or slope is particularly exposed to the wind effect, and the catch will be too low. An example of this was found in the record at Tithewere, Roundwood, County Wicklow. The mean annual value for this station at 1000 feet appears too small in comparison with the numerous stations near it; and the deficiency is very probably due to the gauge being placed on the windward slope of a hill. The measurement of rainfall when all circumstances are considered is found to be necessarily only approximate. According to Dr. H. R. Mill,’ no one could hope to measure the annual rainfall of a place to a closer approxi- mation than 0°5 inch; and it is usually recognized that the estimate of mean annual rainfall is only correct to within 2 or 3 per cent. ‘Ihe rainfall values in the tables are in consequence quoted only to the first decimal place. The methods adopted for estimating the average number of rain days per year were in all respects similar to those employed for determining the mean annual rainfall. The limits of error in the record of rain days appear to be greater than in the case of rainfall. A rain day is defined as one during which ‘01 inch or more is recorded. When more than half 1 The gauge at Fitzwilliam Square is quoted in Table I. as at 1 ft. 6 inches aboye ground. It stands on a little mound which raises it about 4 feet above the general ground-level. 2 I am indebted to Sir John Moore, M.D. (who is responsible for the gauge at Fitswilliam Square), for his valuable opinion on this point. 3 Minutes, Proc. Inst. Civil Engineers, vol. cly., p. 373. SCIENT. PROC., R.D.S., VOL. XII., NO. XXX. 3N 362 Scientific Proceedings, Royal Dublin Society. this amouut is collected, the day must be reckoned as rainy; and when less than half is observed, the water is thrown away, and no rain registered. The personal equation enters largely in deciding in cases of such small quantities. In stating the average number of rain days per annum, the estimate was given only to the nearest 5 in the units’ place. The following tables give particulars of position of gauge, duration of record, values of mean annual rainfall and average number of rain days per year, for the more important stations in the area studied, and also for a few stations outside the area, the records of which were used as references. Taste I.—JJean Annual Rainfall of Dublin. i aie ie ; Heicur aBove _2 5 3 | 2 Zs os ae STATION. eg = re E - 28 zen ae 5 Bis | 3 2 ; [ee | #85 | eee a | & ae ey Sra ae feet | ft.) in’ Balbriggan (Laragh), . B 2 |). OY 0 |1868-1901| 384 | 30:1 | 30-1 195 a (Ardgillan), . .| 211 | 1 © |1893-1907| 15 | 28-9 | 29-3 190 Malahide, 3 3 0 . 26 1 0 | 1904-1908 5 24-6 28:0 185 Kilsallaghan (Corrstown), . . | 280 3 4 | 1875-1879 5 | 35°0 | 30°65 200 Glasnevin, : : : 5 ||) oS) 1 0 : 1878-1907 | 35 | 28:4 | 28-4 180 Phoenix Park, . ° ° . | 155 1 0 /|1873-1907| 35 28°2 28:2 215 Lucan, : . 6 . . | 105 | i 0 | 1898-1903) 11 29°5 28°2 170 Fitzwilliam Square, . D . | 54 1 6 |18738-1907) 35 | 27-7 | 27-7 195 Rathmines, . 0 3 . | 170 0 11 | 1892-1907} 16 | 27°65 27-9 195 Terenure, é 4 9 . | 125 0 6 {1892-1907} 16 | 27-2 | 27-5 190 Dundrum (Lynton), . : . | 200 1 0 (1903-1908 6 | 29-4 | 32:0 206 Stillorgan (White Cross), . . | 290 1 0 | 1905-1907 3 | 27-4 | 30° 190 Blackrock (Rockville), 6 .| 95 | 28 0 |1850-1873) 24 | 25-4 | 29:0 120 Monkstown (Easton Lodge), .| 90 | 1 0 | 1860-1884] 25 | 31-1 | 29:2 | 170 Kingstown (People’s Park), - | 48 1 @ | 1901-1907 7) 2024) 27 170 Dalkey (Belle Vue Pak), . .|148 | 1 3 |1877-1891| 15 | 27-0 | 27-1 | 170 Killiney (Cloneevin), ; . | 250 1 0 {1886-1907} 22 | 27-7 | 28-0 190 Ballybrack (Streamyille), . a | BO 1 6 /2899=1908)) 110) |) '29%5) |) 29-2 190 Carrickmines (Claremont), . . | 850 0 11 |1900-1908] 9 | 29-3 | 29%5 — Foxrock (Hillside), 0 . | 245 1 0 | 1904-1908 5) 29-1 3270 214 Glenasmole (Waterworks), . . | 514 1 © |1885-1908| 24 | 45:2 | 45°5 225 ee (Lodge), . . | 800 | 1 0 |1883-1892] 10 | 49-4] 53:0 | 240 90 (Friarstown), . 5 |) GO? 1 6 | 1883-1887 5 | 38:9 42°3 225 Lyons—The Distribution of Mean Annual Rainfall. 563 Supplementary Notes (Dublin). Mention must be made of some records in addition to those given above for County Dublin. Leeson Park (1905-8) and Upper Leeson Street (1883-90) show a slightly higher rainfall than that of Fitzwilliam Square. On the other hand, Donnybrook (1899-1902) is much lower; and Trinity College (1904-8) gives a record uniformly less than Fitzwilliam Square. Kecles Street (1873-7) agrees closely with Glasnevin. It is of interest to note that the annual graph of Glasnevin, though generally lower than that of the City and Phoenix Park up to 1890, lies subsequently higher. The cause of this anomaly is probably the changes made in the position and exposure of the gauge at Glasnevin about 1890.1 The records for Rush (1876), Skerries (1866-70), and Milverton Hall, Skerries (1907-8), agree very closely with the records quoted for the coast-stations given in the table. ‘The record for Lambay Island (1907-8) is remarkably low, and suggests a mean value of less than 27 inches. The record (1908) for Howth indicates a mean probably greater than 29 inches. 1Jn connexion with this matter Mr. F. W. Moore, Keeper of the Royal Botanic Gardens, Glasnevin, kindly writes in reply to enquiries as follows :— “ Royat Boranic GARDENS, GLAsNEVIN, DuBLIn, “ January 21st, 1910. ‘¢ Dear Sir, ‘«T have been hunting up the records of our rain-gauge. I find that a change was made in September, 1889, which would be likely to haye largely affected the record of the rainfall, as a collecting roof, which was formerly in use, was then changed to a standard gauge, which is still in use. Again, in July, 1901, a change was made which would be likely to affect the quantity registered, as the late Mr. Scott, of the Meteorological Office, considered that the railing which enclosed the rain-gauge at that time was likely to affect the amount registered by intercepting some of the fall. It was therefore moyed to another position, and I do not think there is anything in the present position more favourable to a heayier record than that to which the rain-gauge was moved in 1891. It is of somewhat lower level, 55 ft. On 30th November, 1899, the observer was changed, and the present observer took up the functions. This might perhaps influence the readings, but certainly not to a larger extent than 2 per cent. I think this answers all your questions; but if you wish for further information, please let me know. ‘“‘T am, yours truly, “FB. W. Moors.’’ [‘Tasies I]. anv III. an 2 364, Scientific Proceedings, Royal Dublin Society. Taste 11.—Mean Annual Rainfall of Meath and Kildare. HEIGHT ABOVE 3 | & ae q we | § |8_|e#e|sba STATION. : ro SP | Se | Soe | aos el Jal een ee PES) | ess | cae ¢2| 8 ae | 3 | eel kas | Fee D iS) ea a | So: be feet Comins | | | + Drumconrath (Aclare), 170 2 6 | 1892-1908 | 17 | 35:4 | 3675 215 Moynalty, 265 | 1 4 |1882-1908| 27 | 36:8 | 37-4 | 215 Oldcastle, 382 | 1 0 |1901-1905| 5 | 34-4 | 36:0 | 200 Kells, « 227 | 5 0 |1890-1908] 19 | 35-0 | 36-0 | 215 Navan (Balrath), 160 | 1 0 | 1876-1888] 13 | 30-7 | 31:2 | 160 Athboy, 227 | 1 8 |1889-1908] 20 | 33:5 | 342 | 180 Trim, 180. | 4 2 |1875-18s9| 15 | 38-4 | 33:0 | 175 Summerhill, 372 | 1 0 |1894-1908| 15 | 33:0 | 33:0 | 210 Williamstown (Clonee), 200 1 0 | 1873-1888, 16 | 30-1 | 30:3 | 200 Straffan, 240 | 2 0 |1888-1908| 26 | 30:9 | 32:0 | 1865 Monasterevin (Mooreabbey), 236 1 0 | 1898-1908) 11 | 31-4 32:8 | 200 Castledermot, — | 2 8 |1900-1908] 9 | 29:5 | 29:0 | 185 Ballymore- Eustace, _ 1 4 | 1875-1878 4 | 38°0 34°6 200 i Taste IIT.— Mean Annual Rainfall of Wicklow. STATION. Bray (Fassaroe), Greystones (Knockdolian), . Newcastle (Killadreenan), . Delgany (Inismore), Newtown Mount Kennedy, Annacarter, Tithewere, Roundwood (Knockatemple), ” (Vartry Lodge), Laragh (Glendalough), Rathnew (Clonmannon), Tinahely (Coolattin Park), . Carlow (Browne’s Hill), HEIGHT ABOVE E . a é as = 25 ; = ions a By 4 sae nS q ,2 1 E Pc Ee Peek ees & feet ft. in. 250 5 0 | 1873-1907} 35 40°7 40°7 195 68 i 1891-1907 17 33°8 34:2 180 255 i. 2 1897-1908 12 8474 35°0 180 220 1 0 | 1877-1880 4 39°0 37:0 210 330 — 1906-1908 3 33°3 38°0 — 750 HG 1899-1908} 10 42°3 43°4 = 1000 1 6 |1899-1908] 10 38°4 39°5 — 760 1 6 | 1899-1908; 10 40°8 41°8 = 800 1 6 |1899-1908} 10 42-1 43°2 = 720 1 0 | 1899-1908 10 44:1 45:2 — 460 2 9 | 1892-1907 16 57 0 58:0 215 27 1 0 | 1906-1908 3 30°9 34°8 — 427 1 0 | 1866-1873 8 37:5 42-0 — 291 arr 0 1873-1907 35 34°6 "34-6 200 ee Lyons—The Distribution of Mean Annual Rainfall. 365 Supplementary Notes (Wicklow). An interrupted record was kept at Parknasilloge (Hnniskerry), for the years 1877-80, 1884, 1886, 1889-91, 1898. The values are uniformly higher than at Fassaroe; and on reduction by reference to the latter, the mean for Enniskerry comes to about 45 inches. The isolated records for Bray, Pem- broke Lodge (1907-8), Florence Terrace (1881), and San Remo (19038), indicate a mean close to 35 inches. The annual values for Ashford (1893), Avondale (1908), and Shillelagh (1908), agree with the records for neigh- bouring stations. Tue IsonyveraL Map. The most effective manner of presenting the distribution of rainfall is by means of isohyetal or equal-rainfall lines drawn on a map. The stations with the numbers representing their mean annual rainfall values are marked on the map; and by reference to these stations, lines are drawn through points having probably the same mean rainfall. The principles of constructing such rainfall-maps are very fully discussed by Dr. Mill in the Quarterly Journal of the Royal Meteorological Society, vol. xxxiv., No. 146, April, 1908; and a very important critical discussion of the subject will be found in the Monthly Weather Review, United States Department of Agri- culture, Weather Bureau, April, 1902. The results obtained in the present investigation permitted the drawing of isohyetal lines corresponding to 28, 30, 32, 35, 40, 45, 50, and 60 inches of rainfall. Following a suggestion made some years ago by Symons, the increase in rainfall is indicated by increased thickness of the equal-rainfall lines. The 50- and 60-inch isohyetals are based on but few records, and, being to some extent doubtful, are represented by broken lines. Where the number of records was insufficient, the run of the isohyetal line was decided by reference to the contour and by careful consideration of the similarity of the conditions with those of other neighbouring places of known rainfall. This procedure unfortunately allows a great amount to individual judgment, and is open to criticism. It is, however, in the absence of records, the best approved method, and is found to be justified by experience. The degree of probability to be associated with the isohyetal lines is moreover likely to be greater over a small area than in the case of extended regions. The distribution as shown on the accompanying map (see Plate XXI.) is remarkable for the great range in precipitation, and for the very rapid gradient or sudden change in the rainfall immediately south of Dublin City. There is seen to be an area of mean annual rainfall less than 28 366 Scientifie Proceedings, Royal Dublin Society. inches, including Dublin as a centre, extending north to Malahide, west to Lucan, and taking in a narrow part of the coast about Kingstown, Dalkey, and Killiney. It seems very remarkable that the elevated ground at Daliey and neighbourhood should have such a low rainfall; but a careful investigation showed no evidence to doubt the values computed for Belle Vue Park and Cloneevin. The 28-inch line was pushed outside the coast immediately south of the city in consideration of the values found for Blackrock, Dundrum, Rathmines, and Terenure. The 30- and 32-inch lines were easily determined by the numerous records available. The 35-inch line presented two branches, one enclosing the high-lying land on tie north-west corner of the area; the other branch encircling the mountain system in the south from Carlow to Wicklow. This 35-inch line was brought outside Bray Head in consideration of the height of the latter and the elevations immediately inland. The 40-inch and 45-inch lines were decided by the records of Roundwood, Bray (Fassaroe), Enniskerry, and Glenas- ; mole. The 50- and 60-inch lines were based on the records for Glenasmole and Laragh (Glendalough). The following represents the areas included in the several zones of rainfall :— square miles. Under 28 inches, aa a 30 62:9 Between 28 and 30_,, er : as 151:0 fh OU aNC ete o ig ae 530°4 yO py BID). ing ae ats wi 876:3 ie Bey Le 2 ks in Ke 431-5 5 AO aes fone ae He a 342°6 NAS cont OMe che ae ae 143-0 a Omen OO rae wy Sh ie 97°8 Over 60 ,, te ae fis 144:0 Total, 2779°5 The rainfall of the area now treated in detail has been broadly repre- sented in previous rainfall-maps of Ireland, to which reference must be made. In 1868 Symons submitted a rainfall-map of the British Isles in his evidence before the Commission on river-pollution.t The map was based on the short period, 1860-1865. In 1884,’ and later in 1899,° Dr. A. Buchan prepared a rainfall-map of Ireland, the former based on the period 1860- 1888, the latter on the period 1866-1890. The latest and most reliable 1 Sixth Report of Commission on River Pollution, 1868. * Journ. Scottish Meteor. Soc., 3rd series, vol. vii., 1886, p. 131. 5 Bartholomew’s Physical Atlas, vol. iii., Meteorology, 1899. Lyons—The Distribution of Mean Annual Rainfull. 367 rainfall-maps of Ireland are those included by Dr. H. R. Mill in his valuable paper, ‘‘On the Distribution of Mean and Hxtreme Annual Rainfall over the British Isles”? in 1903.1 Mill considered the mean annual rainfall for the period 1870-1899, and also the rainfall for the extremely wet year 1872, and the exceptionally dry year 1887. The above maps, which were made on a small scale, gave only the more general features of the distribution of rain, and did not attempt to show the details of local variation. ‘he present map was originally prepared on a seale of | inch to the mile, or approximately 1 to 250,000. Special attention was given to the preparation of the plate, so as to deal with the well recognized difficulty of presenting at the same time the rainfall-distribution and the physical features. ‘This difficulty, which has been frequently discussed,” was met in the present case by representing the rainfall-distribution by the isohyetal lines and graduated tints; the relief features being brought out by showing the details of the river-systems and the heights of the more important elevations, as well as by carefully selected hill-shading in black. All roads were omitted to avoid complication of detail. The canals and railways were marked in for reference. GENERAL ResuLts AND ConcLUSIONS. The meteorological and physical importance of the local variation of rain makes it desirable to consider the results of the present investigation with special reference to the relation of rainfall to orography, and to the agreement of our results with facts elsewhere established on the subject. It is also necessary to review the current theoretical views on the problem of rainfall and relief in seeking an explanation of the observed facts. The distribution observed in the present area shows a clearly defined dependence on the marked configuration of the country. The highest ‘rainfall is found in the midst of the Wicklow Mountains; the lowest in the low-lying area enclosing Dublin and the valley of the Liffey near its entrance to the sea. On the east and west and north, the isohyetals from 35 to 60 follow closely the limits of the mountain system. ‘To the north- west of Dublin the rainfall increases very gradually and isin close agreement with the gradual increase in elevation. This connexion of rainfall with relief is seen in all parts of the world where the mean annual rainfall is charted. It is found not only in the general distribution over large areas, but even in the detailed charts of small ' Minutes, Proc. Inst. Civil Engineers, vol. cly., p. 293. 2 See Monthly Weather Review, U.S. Weather Bureau, vol xxx., No. 4, April, 1902, pp. 205-243. 368 Scientific Proceedings, Royal Dublin Society. areas as in the present case. It is moreover observable in districts the relief features of which are not very prominent, as was shown by Hellmann? in the case of Kast Prussia, and by Mill? in the maps of Sussex and Kent. Apart from the general fact that over elevated country the rainfall is greater than over the plains and low-lying districts, no quantitative or definite relation is discoverable between the annual rainfall and the absolute altitude, and many exceptions appear to the rule—the greater the height the greater the rainfall. The aspect of the hills with reference to the prevailing winds, and the detailed configurations of the mountain systems, are factors as important as actual height. Moreover, experience shows, and theory suggests, that the rainfall will depend on the slope, as distinct from the height of the mountain, and on the extent of the sea or evaporating area over which the wind has previously passed.? ‘To the lee-side the rainfall is usually found to be small, as if the mountains had removed the rain from the air-currents, and. thus sheltered the district leewards. These points are prominently established in the present distribution, especially when we consider the winds prevailing over the area. The ten years’ record, 1898-1907, at the Ordnance Survey, Phonix Park, Dublin, gives the following as the average number of days per year during which the wind blows in the particular direction stated :— N. NE. i. SE. 8. SW. W. NW. 18. 19. 30. 20. Phe 72. 104. lie It is easily seen that the westerly and south-westerly are the most prevalent winds; and we may regard them both as rain-bearing, more especially the south-west. As shown in the map, the increase in the rainfall to the south-west or windward side of the mountains is rather gradual, whereas on the north-east or lee side there is a very rapid decrease towards the city. ‘This effect is elsewhere recognized in the influence of mountains on precipitation. On the east coast of Wicklow the rainfall gradient is more gradual than that shown to the north-east. The easterly and south-easterly winds are responsible for a portion of the annual rainfall near the coast; and this effect very probably masks the sheltering action of the hills on the winds of westerly type. The particularly low rainfall of Kingstown and the elevated district near Dalkey and Killiney is probably due to the protecting influence of the Dublin Hills. It would be particularly interesting to study how the rainfall of these districts fares in the course of easterly and south- easterly winds. The rapid variation in the rainfall immediately south of Dublin, combined 1 loc. cit., page 355. * loc. cit., page 365. 3’ Dr. A. Buchan, Minutes, Proc. Inst. Civil Engineers, vol. cix., p. 138. 2 Lyons—The Distribution of Mean Annual Rainfall. 369 with the very narrow area of the zone under 28 inches, is an important feature of the present distribution. From Terenure, with a mean of 27:5 inches, to Glenasmole Lodge, with a mean of 53:0 inches, the distance is only a little over six miles. A very remarkable difference is also found between the rainfall at Glenasmole Lodge and Glenasmole Waterworks, within a distance of two miles. The record at the former is uniformly and considerably higher than at the latter. The annual rainfall graphs run practically parallel, but the mean of the Lodge record is 58, and that for the Waterworks is 45 inches. The gauge at the Lodge was unfortunately abandoned in 1889, owing to the site being affected by trees. The caretaker, Mr. O’Brien, who gives the closest personal attention to the rainfall and general weather of the district, finds both the frequency of showers and the intensity of rain greater at the Lodge than at the Waterworks. He tells of having frequently experienced very heavy rain at the former place during a day when the gauge at the Waterworks registered but a small amount; and the sudden rise of the river flowing from the Lodge has often indicated heavy and continuous rain at the upper valley when that registered at the Waterworks was insignificant. The increased precipitation at Glasnevin compared with the city is very interesting and difficult to explain, especially when we consider the low rainfall of the coast districts to the north. The exceptionally low value for Lambay Island is perhaps to be connected with the small rainfall usually observed near the sea; but the data for Lambay are too restricted to merit discussion at present. Passing from the lower limit to the higher limit of the rainfall, we find a considerable amount of interest in the record for Laragh (Glendalough). ‘The station is situated at a comparatively low level in the midst of the mountains, and to the north-east of Lugnaquilla (3039 feet). Its mean annual rainfall works out at 59 inches. In considering this record of Glendalough attention must be paid to the fact that on mountains a definite maximum zone of rainfall occurs ; and this maximum is found sometimes on the summit, sometimes on the windward, and occasionally on the leeward side. In the Lake District? the maximum occurs at an altitude of about 1500 feet ; and the striking fact is there noticed that the rainfall at 3000 feet is about 90 inches, whereas in the valley on the lee side the rainfall is about 140 inches. The Lake District being exposed to the practically direct approach of the rain-bearing winds from the sea is probably subject to more exceptional conditions than Wicklow. It is never- theless possible that the rainfall of Glendalough is not much lower than the 1 “ British Rainfall,”’ 1897. SOIENT. PROG, R.D.S., VOL. XII., NO. XXX. 30 370 Scientific Proceedings, Royal Dublin Society. rainfall at the higher levels around it. It is difficult, however, to extend the results obtained in other places to help to decide in such a matter as this. The position of the maximum zone depends on the absolute heights and details of the mountain system, on the prevailing winds, and on the average conditions of temperature and humidity." The maximum zone lies lower in winter than in summer by reason of the lower temperature and larger relative humidity, and in Wicklow the average rainfall for the summer is probably greater than for the winter. The fact that a rainfall of 53 inches was found for Glenasmole Lodge is, with the other considerations, a justification for considering a rainfall over 60 inches probable in the centre of Wicklow. With reference to the values found for the frequency of rainfall as repre- sented by the average number of rain days per year, it is remarkable that the difference generally is very small, and in cases quite negligible even for stations the rainfalls of which show the greatest contrast. Thus the fre- - quency of rain at Phoenix Park is 215, the same as at Glendalough, although the rainfall for the latter (59) is more than double the rainfall of the former (28°2). Again, there is observed a considerable difference in the rain-fre- quency for the neighbouring stations, Phoenix Park (215) and Straffan (180), whereas the difference in rainfall is small and in the opposite direction. It is perhaps possible to draw only questionable deductions from the results just mentioned as to the rain-frequency. It would seem, however, that the great differences observed in annual rainfall over a limited area are generally not due to rain days being more frequent in the mountains than in the low-lying districts, but rather to the rain being more intense, or possibly more prolonged in the course of each rainday. Further facts on these points are urgently needed. The question of rain-frequency, so far as the author knows, has not been previously considered in relation to rainfall-distribution ; and no confirmation of the present results can be given. THuORETICAL CONSIDERATIONS. When we proceed to seek the causes of the influence of relief on rainfall, we find the subject has not received very satisfactory treatment, and that existing views are vague and scarcely adequate to explain the several facts established. Older writers rested content with the suggestion that moist air was condensed by its contact with the cold mountain side. Itis now proved, however, that any condensation produced by such a process or by the mixing of warm with cold air-masses is insignificant. It is recognized that the most 1 <¢ |ehrbuch der Meteorologie,”’ J. Hann, 1901, pp. 350—352. Lyons—The Distribution of Mean Annual Rainfall. ovl frequent and most effective process in condensation is the adiabatic cooling produced by the expansion of ascending masses of air. The condensation will result at first in the formation of necessarily minute cloud-particles; and a second phase or process is involved in the development of rain-drops from these minute cloud-drops. The factors and conditions necessary for this second development are very obscure. The continued growth of a drop by further condensation is difficult to understand in most cases, owing to the latent heat involved in the process.! Such continued condensation could probably occur only if the expansion was rapid’ or sustained.* The develop- ment of rain-drops by the coalescence of the cloud-particles is a process also presenting physical difficulties ;* and it is well recognized that an important factor in this process of coalescence is the electrical conditions in the air.° Confining attention to the condensation phenomena, it is seen that the ascensional movement of air may be brought about in two ways, giving two distinct classes of rain. The first will be produced in the upper currents, associated with such barometric systems as cyclones and secondaries, and with such disturbances as thunder-squalls. In the second case, wind blow- ing towards mountains will be forced upwards ; and under suitable conditions condensation will take place. The latter effect will be local; and rains thus produced are called orographical rains, to distinguish them from the meteoro- logical rains associated with such atmospheric disturbances as cyclones. In considering the influence of relief on the rainfall, it is most important to distinguish between these two kinds of rain. It would seem reasonable to suppose that the relief-features would exercise very little, if any, influence on the condensation occurring in the case of the meteorological rains, the causes of which must be sought in the middle or higher layers of the atmosphere. This supposition should hold especially in the case of small areas, over which the conditions in such layers would probably be uniform. In this connexion it is important to note that Mill has recently stated that, in a study of the distribution of such rains as occur in cyclones and thunder-squalls, he finds no relation to the configuration of the country.® It is thus natural to look to the orographical rains as an explanation of the increased rainfall on elevated ground. ‘The theory of the formation of con- densation on mountain slopes was mathematically worked out in some detail, 1 Osborne Reynolds, Scientific Papers, vol. i., pp. 214-230. 2 Angot, ‘‘ Annales du Bureau Central Météorologique,”’ 1895. 3 Aitken, Proc. Roy. Soc., li., 1892, p. 408. 4 Aitken, ibid.; Lord Rayleigh, Phil. Mag., xlviii., 1899, pp. 321-337. 5 Lord Rayleigh, Proc. Roy. Soc., vol. xxviii., 1879, p. 406. 6 Quart. Journ., Roy. Meteor. Soc., vol. xxxiy., No. 146, April, 1908, pp. 72, 73. He regards it, however, as too soon to speak with certainty on this point. 372 Scientific Proceedings, Royal Dublin Society. by Prof. F. Pockels, Dresden, in 1901.1. Making certain assumptions, and considering only simple relief forms at right angles to the wind, Pockels showed that, in the case of high mountains at least, a maximum rain-zone would be found on the windward slope, and that the precipitation would depend on the slope rather than on the actual height of the mountain.? The conclusion was also given that mountains less than 500 or 600 metres high would not occasion rain under ordinary atmospheric conditions, either in summer or in winter. It is apparent that considerable difficulties arise in attempting to explain by reference to orographical or local rains only the influence of relief on rainfall. The influence as already noted is observed even in the case of small elevations; and in the case of moderate altitudes, such as were studied in the present paper, the effect is very marked. Over the present area and the British Isles generally a very considerable pro- portion of the total annual precipitation must be associated with meteoro-_ logical rains; and the difficulties of explaining the local variation by the orographical rains is increased. It would be necessary to suppose that such orographical rains were very frequent, or very intense, or long continued. The results for the number of rain days per year, given in the present analysis, do not appear to justify the assumption that rain is much more frequent in the mountains than in the plains. Apart from this, however, it is difficult to understand, on Pockels’ theory, or from general considerations, how the intensity and duration of the local rains could be such as to account sufli- ciently for the great differences observed in the rainfall. It must also be pointed out that local rains are more frequent in summer than in winter ; but in summer, when the relative humidity is low, and the temperature high, the influence of mountains in causing precipitation is less than in winter.* It is not desirable to go further into the discussion of this important matter in the present paper. A considerable amount of existing evidence has to be carefully balanced, and further information on several points obtained, before any finality of opinion can be reached. It may be suggested, however, with all the reserve necessary under the conditions of our present knowledge, that, apart from inducing condensation under favourable conditions, mountains increase the rainfall by their action on already condensed cloud-masses. It is not difficult to indicate how their action might probably help in the second process of 1 Ann. d. Physik, 1901 (4), vol. iv., p.459, and Monthly Weather Review, U. 8. Weather Bureau, April, 1901. 2 Hann’s “‘ Klimatologie,”’ vol. i., p. 298. 3 Prof. J. Hann, ‘‘ Lehrbuch der Meteorologie,”’ p. 350; Supan, Memoir ‘‘ On the Distribution of Rainfall over Islands and Continents of the Globe,’’ pp. 40-43. Lyons—The Distribution of Mean Annual Rainfall. 373 rain-making, viz. the development of rain-drops from the first minute particles of condensation. The vertical air-currents in the vicinity of mountains would impel upwards any approaching cloud-masses, the products of previous and independent condensation. This upward impulse would cause cooling and produce further condensation on the already formed cloud-particles; further- more, the eddying currents likely to occur in mountain systems would similarly help in the growth of rain-drops. If, as is probable, coalescence plays a part in rain-making, it is easy to see that mountains would tend to increase the frequency and intensity of collisions by the aerial disturbances set up, and the unequal velocities such disturbances would impart to different-sized drops. Moreover, if the coalescence process is affected by electrical con- ditions, and if, as suggested by Lord Rayleigh’s' experiments on colliding water-jets, coalescence is induced by differences in the electrical field, it is possible that the well-known inequalities’ in the atmospheric electrical field near mountains play a part in the formation of rain in such places. It is proposed to develop these suggestions in a subsequent communication. In conclusion the author would express his obligations to the various rainfall observers whose conscientious daily returns formed the groundwork of this investigation, and to the British Rainfall Organization for rendering the data so readily available for study. Indebtedness must also be expressed to Dr. H. R. Mill and Sir John Moore, M.D., D.Sc., for help in the matter of references. Acknowledgment should also be made of the special facilities for visiting the gauges and districts afforded by the Department of Agriculture and ‘Pechnical Instruction for Ireland. 1 loc. cit., p. 371. * Chwolson, ‘“‘ Traité de Physique,” tome iy., ler fascicule, 1910. SCIENT. PROC. R.D.S., VOL. XII., NO. XXX. 3e r, AVE NITEAG 62 -=s5 asnoyqybiy Ay1eg% ! (e ite yyMoH Py aAq spuejau) Fy EN) 1G aN Wis MAP OF MEAN ANNUAL RAINFALL OVER AREA INCLUDING COUNTIES DUBLIN, WICKLOW, MEATH, AND KILDARE. REFERENCE, Rainfall below 28 inches - - =a) in between 28 and 30 inches =a) » » 80and32 , Srey 2 » 32and 35 ,, PF ee. ” ” 35 and4o ,, 5s 6 2 : = ” 7p 40 and 45 eo = 2 6 5 ai aaa) ” above 45 inches aaa ——— SCALE OF MILEs. CALE OF MILES eee eee) fa # SCIENT. PROC. R. DUBL SOG., N'S.. Vou. XII. "PLATE XxI. ) Dunany Point IRISH SEA Rockabill * Lighthouse ngstown Seoatkey Island WICK LOW © Wiick/owHead Bewroan. Opney THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 31. JUNE, 1910. THE VAPOUR-PRESSURES, SPECIFIC VOLUMES, HEATS OF VAPORISATION, AND CRITICAL CONSTANTS OF THIRTY PURE SUBSTANCES. BY SYDNEY YOUNG, D.Sc., F.B.S., TRINITY COLLEGE, DUBLIN. —apsonian Instit, Gy” ? | NOV 26 1910 National Museu [Authors aloneare responsible for all opinions expressed in their Communications. } DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1910. Price Three Shillings. Roval Bubltw Society. RON Oa oN ON FOUNDED, A.D. 1731. INCORPORATED, 1749. EVENING SCIENTIFIC MEETINGS. Tue Scientific Meetings of the Society are held alternately at 4.30 p.m. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested +o forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay.. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Editor. Lyons—The Distribution of Mean Annual Rainfall. 373 rain-making, viz. the development of rain-drops from the first minute particles of condensation. The vertical air-currents in the vicinity of mountains would impel upwards any approaching cloud-masses, the products of previous and independent condensation. This upward impulse would cause cooling and produce further condensation on the already formed cloud-particles; further- more, the eddying currents likely to occur in mountain systems would similarly help in the growth of rain-drops. Ii, as is probable, coalescence plays a part in rain-making, it is easy to see that mountains would tend to increase the frequency and intensity of collisions by the aerial disturbances set up, and the unequal velocities such disturbances would impart to different-sized drops. Moreover, if the coalescence process is affected by electrical con- ditions, and if, as suggested by Lord Rayleigh’s! experiments on colliding water-jets, coalescence is induced by differences in the electrical field, it is possible that the well-known inequalities’ in the atmospheric electrical field near mountains play a part in the formation of rain in such places. It is proposed to develop these suggestions in a subsequent communication. In conclusion the author would express his obligations to the various rainfall observers whose conscientious daily returns formed the groundwork of this investigation, and to the British Rainfall Organization for rendering the data so readily available for study. Indebtedness must also be expressed to Dr. H. R. Mill and Sir John Moore, M.D., D.Sc., for help in the matter of references. Acknowledgment should also be made of the special facilities for visiting the gauges and districts afforded by the Department of Agriculture and Technical Instruction for Ireland. loc. cit., p. 371. 2 Chwolson, ‘* Traité de Physique,’’ tome iv., ler fascicule, 1910. SCIENT, PROC. R.D.S., VOL. XII,, NO. XXX. 3P r ae] XXXI. THE YVAPOUR-PRESSURES, SPECIFIC VOLUMES, HEA'TS OF VAPORISATION, AND CRITICAL CONSTANTS OF THIRTY PURE SUBSTANCES. By SYDNEY YOUNG, D.Sc., F.R.S., Trinity College, Dublin. (Read January 25. Ordered for Publication Frsruary 8. Published Junz 10, 1910.] THe kinetic theory in its original form is applicable only to theoretically perfect gases, the molecules, so far as their dimensions are concerned, being regarded as mathematical points. It is assumed that they are endowed with perfect elasticity, and that they do not attract one another. The extension of the theory to cover the case of real gases under high pressure and of liquids is due, in the first place, to Van der Waals,’ who proposed the fundamental equation of state :— (» + see = (0) = Jill The researches of Andrews, Amagat, and Ramsay and Young, however, led to the conclusion that the equation of Van der Waals, although reproducing the general form of the isothermals, did not give results in sufficiently close agreement with the experimentally determined data; and numerous modifications of the formula have been proposed by physicists, but no completely satisfactory equation has yet been devised. On the other hand, Van der Waals pointed out that by expressing the pressures as fractions of the critical pressure, the absolute temperatures as fractions of the absolute critical temperature, and the volumes of liquid and of vapour as fractions of the critical volume, a “reduced” equation could be derived from the original formula which should be applicable to all substances. Applying the reduced equation to the special case of liquids at their boiling-points and saturated vapours, it would follow that if the pressures of any two substances are proportional to their critical pressures, their boiling- points (expressed as absolute temperatures) should be proportional to their critical temperatures, and their volumes, both as liquid and as saturated vapour, should be proportional to their critical volumes. 1 Continuitat der gasformigen und fliissigen Zustandes, Youna— Vapour-Pressures, §¢., of Thirty Pure Substances. 375 Moreover, these generalisations regarding “corresponding ” pressures, temperatures, and volumes, might be true, even though the original equation required considerable modification. It was with the object of testing the correctness of these generalisations that I undertook a series of determinations of vapour-pressures, specific volumes, and critical constants of pure liquids. The data have been published in a series of papers in the ‘l'ransactions of the Chemical Society, the Proceedings of the Physical Society, and the Philosophical Magazine’; but, for the following reasons, complete data for any one substance are not to be found in a single paper :— 1. New methods have from time to time been devised, chiefly for the determination of the specific volumes of liquid and saturated vapour. 2. When the earliest papers were published, no satisfactory method of determining the critical volume of a substance was known; and it was shown by Gouy that, owing to the extreme compressibility of a substance in the immediate neighbourhood of the critical point, the direct method at first employed could not be regarded as accurate. It was, however, discovered by Cailletet and Mathias that the mean densities of liquid and saturated vapour, when plotted against the temperatures, appear to fall on a straight line (fig. 1, p. 376)?; and they showed that the point of intersection, C, of this rectilinear diameter, AC, with the closed curve of orthobaric volumes, VCZ, gives the critical density. Slight deviations from the law of Cailletet and Mathias were observed as a rule; but they appeared to be insignificant, and were attributed to errors of experiment; and the critical densities of the great majority of the substances were determined by this method. Subsequently, however, on close investigation of the whole of the results, it appeared that the deviations were too regular to be attributable to experimental error, and that the diameter must in most cases be regarded as slightly curved. A redetermination of the critical densities, taking the slight curvature of the diameter into account, was therefore necessary. 3. The methods employed for the determination of the volumes of a gram of the saturated vapours of most of the substances are of such a nature that the results increase in accuracy as the temperature rises. When the papers were first published, no special importance was attached to the data at the lowest temperatures (generally not far from the boiling-points under normal pressure); and it was considered sufficient to plot the logarithms of the 1 The experiments on ethyl ether, methyl alcohol, ethyl alcohol, propyl alcohol, and acetic acid (up to 290°) were carried out by Sir William Ramsay and the author at an earlier date. + In this figure the actual data for normal pentane are reproduced. 38F2 376 | Scientific Proceedings, Royal Dublin Society. specific volumes against the temperatures and to draw curves to pass as evenly as possible through the points. Now it is a matter of common experience that, in drawing a curve through a series of points, even when the errors of experiment are fairly regularly distributed, the deviation of the drawn curve from its true position is likely to be smaller in the middle portions than at the extremities. And when the trend of the curve—giving the values of iis considered, it is found that the errors are likely to be much greater at the extremities than near the middle. Density Fig. 1. In the case of the volumes of saturated vapour, however, the errors of experiment are usually greatest at the lowest temperatures; and therefore the deviation of the drawn curve from its true position, and, to a still greater extent, the error in the trend of the curve, is liable to be very much more pronounced at its lower extremity than in any other part. Having been asked by M. Chappuis, in 1907, to furnish a table of densities of saturated vapour under normal pressure, and of the rate of change of density per millimetre rise of pressure, for the “ secweil des constantes physiques,” it became necessary to consider the data at the lowest temperatures more carefully than had hitherto been done. An examination of the whole of the data leads to the conclusion that the Youne— Vapour-Pressures, &¢., of Thirty Pure Substances. 377 saturated vapours of all the substances investigated, with the single exception of acetic acid, behave more and more like perfect gases as the temperature falls, althongh the real vapour densities under normal pressure are in every case higher than those caleulated on the assumption that Boyle’s and Gay- Lussac’s laws are strictly true. In order to discover whether there was a serious error in the trend of any of the curves, the following device was adopted :—The volumes of a gram of vapour were calculated from the observed pressures for a range of 50 or 60 degrees from the boiling-point downwards on the assumption— (1) that the vapour behaved as a perfect gas, and (2) that the ratio of the actual to the theoretical density was constant and equal to that actually observed under a pressure of about two atmospheres (generally about 1-07 to 1:10). The logarithms of these theoretical volumes were plotted against the temperatures, as were also the logarithms of the observed volumes for a range of 100 or 120 degrees, that is to say, from about the boiling-point under normal pressure to a temperature about 100 degrees higher. It is evident that the curve for the real vapour should, if produced, lie between the two theoretical curves, and that, with falling temperature, it should approach that for a perfect gas. In this way it was found that, in drawing the original curves, too much weight had, in several cases, been attached to the least accurate observations at the lowest temperatures, but that, on the other hand, the results with iodobenzene at the lowest temperatures were more accurate than had been supposed. In most cases the reconstructed curves differed only slightly from those originally drawn. It was then found that the particular data required could best be obtained by plotting the logarithms of the densities of saturated vapours (water at 4°=1) against the logarithms of the pressures, guiding curves being drawn as before. The curves for the real vapours were found to approximate very closely indeed to straight lines at and near the normal pressure, so that the following formule could be employed : log s) = A +a log p, (1) ds a.s cae @) (where s is the density and p the pressure). The values of A, a, s, and di : e under normal pressure have been published in a separate paper.’ 1 Journ. de Physique, (4), vili., p. 5, 1909. 378 Scientific Proceedings, Royal Dublin Society. As the orthobarie volumes of vapour published at various dates in different journals are now found, in many vases, to require correction at the lower temperatures, they have been read from the reconstructed curves, and are given in the tables at the end of this paper. 4. In many of the papers, not only the specific but also the molecular volumes are given, and in some cases the molecular volumes only. But since the publication in 1889 of the first paper in which molecular volumes were recorded, the atomic weights of many of the elements have been redetermined with greatly increased accuracy. In order to avoid confusion, the atomic weights first adopted, those of F. W. Clarke (Constants of Nature, 1882), were adhered to in the whole series of papers, although it became more and more evident as time went on that they required considerable modification. In the present paper it has not been thought necessary to give the molecular volumes at all, as they can easily be calculated, if required, from the specific volumes. 3. After the vapour-pressures of each substance had been determined, the logarithms of the pressures were plotted against the temperatures, and curves were drawn through the points. Constants for Biot’s formula log p = a + ba’ + cf! were then in most cases calculated ; and the vapour-pressures were re- calculated from the formula, and compared with those read from the curves or actually observed. The constants and the recalculated pressures are given in the original papers for all the substances except the ten esters. The values for methyl formate and ethyl formate were actually calculated, but were not published at the time; those for the other eight esters were subsequently calculated by Dr. J. E. Mills, and the whole of them were published by him. For the great majority of substances the observations extend from 0° to the critical temperature; but with bromobenzene and iodobenzene it was not found possible to extend the temperature range beyond 270°. Also in the case of chlorobenzene the constants of Biot’s formula were calculated after determinations of pressure had been made up to 270°, but before those at higher temperatures had been carried out. It was noticed some little time ago both by Dr. Mills and by myself that although there is good agreement in all three cases between the observed and recalculated pressures, yet the pressures calculated for the critical temperatures are manifestly impossible. vidently, therefore, the trend ' Journ. Phys. Cliem., x., p. 1, 1906. Youne— Vapour-Pressures, §c., of Thirty Pure Substances. 379 of the curves at their upper extremities must have been incorrect; and it became advisable to recalculate the constants for Biot’s formula. In the original paper! it was pointed out that for any two of the four halogen derivatives of benzene, the ratio of the boiling-points (on the 7 absolute scale) was evidently the same at all equal pressures, or 7 = constant. The values of the ratios of the absolute temperatures of fluorobenzene to those of the three other compounds were first ascertained ; and the boiling- points of fluorobenzene, at a series of even pressures, were then recalculated from those of each of the other halogen derivatives. The mean of the three recalculated temperatures and the observed temperature of fluorobenzene at each pressure was taken to be the correct boiling-point at that pressure. Finally, from the corrected boiling-points of fluorobenzene, those of the other three substances were recalculated by means of the constants. Comparing these recalculated temperatures with the observed boiling- points, it appeared probable that there were small errors, chiefly at the highest temperatures. Taking these probable errors into account, new constants for Biot’s formula were calculated, and although the agreement between the re- calculated and observed pressures up to 270° was not quite so good as before, yet the calculated critical pressures were now satisfactory, notwithstanding the great extrapolation—in the case of iodobenzene from 270° to 448°C. The calculations of Dr. Mills had indicated the probable existence of errors—either of experiment or calculation—in several other cases, and in some of them it seemed not unlikely that the constants for Biot’s formula might be at fault. The constants were therefore calculated again for the following substances in addition to chlorobenzene, bromobenzene, and iodobenzene :—fluorobenzene, ethyl ether, ethyl propionate, methyl butyrate, and methyl isobutyrate. The whole of the data, pressures, specific volumes, &c., have been carefully revised, corrections being made where necessary ; and the collected data are given in the tables at the end of this paper. It has not, however, been thought necessary to republish the results of determinations of the compressibilities of the liquids or unsaturated vapours, as they are very voluminous and are not affected by the alterations in the orthobaric volumes. ‘These data are to be found in the following papers :— 1. Ethyl Ether, Ramsay and Young, Trans. Roy.Soe., elxxviii., p. 57, 1887. 2. Methyl Alcohol, R. and Y., ibid., clxxviii., p. 313, 1887. + Trans, Chem, Soc., ly., p, 486, 1889. 380 Scientific Proceedings, Royal Dublin Society. . Ethyl Alcohol, R. and Y., cbid., clxxvii., Pt. 1, p. 128, 1886. - Propyl Alcohol, R. and Y., idid., clxxx., p. 137, 1889. . Acetic Acid, RK. and Y., Trans. Chem. Soc., xlix., p. 790, 1886. . Isopentane, Young, Proc. Phys. Soc., xiii., pp. 602 and 658, 1895, and Zeitschr. physik. Chem., xxix., p. 193, 1899. 7. Normal Pentane, Rose-Innes and Young, Phil. Mag. (5), xlvii., p. 353, 1899, also (5), xlviii., p. 218, 1899, and (6), ii, p. 208, 1901. 8. Normal Hexane, Thomas and Young, Trans. Chem. Soc., Ixvii., p. 1071, 1895. 9. The vapour-pressures, specific volumes, and compressibilities of steam from 120° to 270° have been determined by Ramsay and Young, Trans. Roy. Soe., elxxxiii.a, p. 107, 1892. OS Oo PB Preparation of Material. The physical properties of a substance, more especially at or near its critical point, may be seriously affected by the presence of even a very small quantity of impurity; it is therefore of the utmost importance that the purification of the substances investigated should be carried out with the greatest possible care. Full details of the methods of preparation and purification of the thirty compounds are given in the original papers; but it may be useful here to state briefly the methods employed in each case. 1. Normal Pentane—This paraffin was separated from American petroleum by fractional distillation through a combined dephlegmator and “‘regulated temperature” still-head. The material from which the paraffin was obtained was Merck’s “pentane,” a light distillate from petroleum consisting chiefly of pentanes with relatively small quantities of butanes, hexanes, pentamethylene, &c. After three preliminary fractionations, the distillates coming over between about 28° and 37° (1030 grams) were shaken with (a) concentrated sulphuric acid, (4) mixtures of concentrated sulphuric and nitric acids, (c) water, (@) caustic potash, and (e) water. They were then distilled over phosphoric anhydride, and the fractional distillation was continued. The weight of pure normal pentane obtained after twenty-one fractionations was 175 grams, 101 grams of pure isopentane being also separated. Hach of the paraffins was finally collected in two fractions; and the specific gravities were determined by means of a Sprengel tube of the form recommended by Perkin. The weighings were in all cases corrected for the buoyancy of the ' Young and Thomas, Trans, Chem. Soc., Ixxi., pp. 440 and 446, 1897. Youne— Vapour-Pressures, §¢., of Thirty Pure Substances. 381 Specific gravity of normal pentane at 0°/4° . A = 0:64536 Bs 4 . B=0:64541 Boiling-point under normal pressure . : 36°3° Another specimen of normal pentane, recently separated from “petroleum ” by fractional distillation through an “evaporator” still-head of fifteen sections, after treatment with acids, &c., as above, had the same specific gravity ; but the boiling-point appeared to be alittle lower. The difference is doubtless due to an error in one of the thermometers; and it may be mentioned that the zero points of the German soda-glass thermometers originally used were not quite constant. The boiling-point calculated by means of Biot’s formula is 36:18°. 2. Normal Hexane.\—The specimen employed for the chief determinations was obtained from Kahlbaum, and had been prepared by the action of sodium on propyl iodide. It was treated with acids and caustic potash in the same way as normal pentane, and was fractionally distilled. ether Specific gravity at 0°/4° : : ; A = 0°67696 ep : B = 0:67697 Later determination with new specific- gravity tube x : : : 0:67703 Boiling-point under normal pressure . 68°95 A second specimen was separated from American petroleum (petroleum ether) by means of the combined dephlegmator and regulated temperature still-head ; but methyl pentamethylene, boiling at about 71:5°, is also present, and cannot be removed by fractional distillation. Both this hydrocarbon and isohexane are attacked much more readily than normal hexane by fuming nitric acid; and the eight fractions collected between about 66° and 69° were separately subjected to repeated and long-continued heating with the acid. After this treatment seven of the fractions boiled constantly at 68:95°, and the specific gravities of the five lowest varied only between 0:67693 and 0°67702. 3. Normal Heptane.—Dr. Thorpe very kindly placed his well-known specimen of normal heptane from Pinus sabiniana at my disposal. Specific gravity at 0°/4°, . : d : 0°70048° Boiling-point under normal pressure, ‘ B 98:48° 5 4. Normal octane..—The specimen investigated was obtained from Kahl- baum, and had been prepared from octyl iodide. The purification was in all 1 Thomas and Young, Trans. Chem. Soc., Ixyii., p. 1071, 1895; Young, idid., Ixxiii., p. 905,1898. 2 Trans. Chem. Soc., lxxiii., p. 675, 1898. 3 Thorpe, Trans. Chem. Soc., xxxvii., p. 213, 1880. * Trans. Chem. Soc., lxxvii., p. 1145, 1900. SCIENT. PROC. R.D.S., VOL. XII., NO, XXXI. 3G 382 Scientific Proceedings, Royal Dublin Society. respects similar to that of normal hexane, except that a more efficient still- head was employed. ‘The distillate, which came over at a constant tempera- ture, was collected in three fractions. Specifie gravity at 0°/4°, . i ! > 4S Olea) és 4 4 Z 5 . B=0°71847 . 1 = 0°71848 Later determination ANE new rate ; d 0:71854 Boiling-point under normal pressure 125°8°. 5. Isopentane..—Three specimens of this paraffin were obtained from different sources :— (2) From Kahlbaum’s “pentane.” This substance is obtained as a by- product in the preparation of amylene from amyl alcohol, the admixed amyl- ene being for the most part separated by means of bromine. The method of purification was similar to that employed for normal pentane; but after treatment with the mixed acids, and before that with caustic potash, the liquid was repeatedly shaken with concentrated sulphuric acid until no further orange coloration was noticeable. After complete removal of amylene by the acids, pure isopentane is easily obtained by fractional distillation. (b) From amyl iodide. The strongest possible hydrochloric acid was added drop by drop to amy] iodide, dissolved in five times its volume of absolute alcohol, in presence of zinc on which a little copper had been deposited. The temperature was maintained at 0°, and the action was complete in about twenty- four hours. The liquid was poured off and distilled from the water-bath until no more isoptane came over; the distillate was shaken with water several times to remove alcohol; bromine was added, to remove amylene, until the isopentane was permanently coloured; excess of bromine was removed by adding caustic potash; and the isopentane was then washed with water, dried with calcium chloride, and fractionally distilled. (c) From American petroleum. The separation of the two pentanes from petroleum has been described under normal pentane. The three specimens boiled constantly at 27-95° under normal pressure. Their specific gravities at 0°/4° were found to be :— (a) (2) (¢) 1. 0:68925 063935 1. 0°63931 2. 0-63922 2. 0:63929 6. Di-isopropy/.—Great difficulty was experienced in the preparation of ! Proc. Phys. Soc., xili., pp. 602, 658, and 666, 1895, and Zeitschr. physik Chem., xxix., p. 193, 1899; Trans. Chem. Soc., lxxi., p. 440, 1897. 2 Fortey and Young, Trans. Chem. Soc,, Ixxyii., p. 1126, 1900, Youne— Vapour-Pressures, &c., of Thirty Pure Substances. 383 this substance. ‘Two methods were employed, but neither of them gave very satisfactory results. (a) An ethereal solution of isopropyl iodide was treated with sodium, a little water being added’ to promote the action. A considerable amount of propylene was formed and carried away some of the di-isopropyl; but a suffi- cient quantity of the paraffin wasobtained. After removal of the greater part of the ether and unaltered iodide by distillation, the di-isopropyl was treated with strong sulphuric acid; and the pure paraffin was finally purified by fractional distillation through a five-column “evaporator” still-head. It was collected in two fractions. Specific gravity at 0°/4°, . seta se 5 A= OGM » x : ‘ : . B= 0°67945 Boiling-point under normal pressure, 58°1°. (6) A strong solution of potassium isobutyrate was electrolyzed in a cell of special construction. very precaution was used to avoid loss of di-iso- propyl; but under the best conditions the yield was only 1:5 grams from 100 grams of isobutyric acid, the theoretical amount being 49 grams. The chief product of the reaction was isopropyl isobutyrate, large quantities of pro- pylene being also formed. ‘The quantity of di-isopropyl obtained by this method was too small to admit of its complete purification. The specific gravity of the small impure specimen obtained by the second method was 0:6806, and on distillation the greater part came over at 58:0°, the temperature finally rising to 588°. The first specimen was employed for all the determinations. 7. Di-isobutyl2—This paraffin can be satisfactorily prepared either by the action of sodium on isobutyl bromide or by the electrolysis of a strong solution of potassium isovalerate. The action of sodium on isobutyl bromide is fairly rapid, and there was no difficulty in preparing a sufficient amount of the hydrocarbon. As the boiling- point of di-isobutyl (109:2°) is considerably higher than that of isobutyl bromide (92'3°), the paraffin could be separated fairly completely from the unaltered bromide by fractional distillation. The final purification was effected by treatment with a mixture of nitric acid and sulphuric acid, and subsequent fractional distillation through an efficient still-head. It is to be noted that pure isobutyl bromide cannot be obtained by distillation, because partial dissociation into isobutylene and hydrobromic acid takes place; and the products of dissociation recombine to some extent on cooling with formation of the tertiary bromide. A small quantity of ' Silva, Berichte, y., p. 984, 1872. ~ Fortey and Young, Joe. cit. 3G 2 384 Scientific Proceedings, Royal Dublin Society. hexamethyl ethane is therefore formed by the action of sodium on the bromide; but it is completely removed during the fractional distillation. A small quantity was isolated in the form of volatile colourless crystals. Specific gravity at 0°/4°, : : 071021 Boiling-point under normal pressure, : 109°2° The electrolytic method was found to give fairly satisfactory results ; but as a sufficient quantity of di-isobutyl had been prepared already by the first method, it was not considered necessary to purify the isovaleric acid by fractional distillation. About 20 grams of di-isobutyl boiling from 109°15° to 109:25° was obtained; but this specimen was not employed for the determination of the physical constants. The first specimen of di-isobutyl appeared to be quite pure; the boiling-point was constant; and the specific gravities of the two fractions agreed very well together. The vapour-pressure also at each temperature remained quite constant when the relative volumes of liquid and saturated vapour were altered. It has, however, been pointed out by Dr. Mills that there are certain anomalies in some of the physical constants for this substance ; and itis a remarkable fact that after being kept for two or three years in well-corked bottles, both specimens had evidently undergone some change, for a few colourless crystals were present in each, and the quantity of the crystalline deposit has continued very slowly to increase, but is still hardly large enough to allow of a satisfactory investiga- tion of the chemical nature of the substance. It seems just possible that the di-isobutyl may have already contained a minute amount of this impurity when its physical properties were determined. 8. Hexamethylene.i—A. small quantity of this substance was obtained by Miss E. C. Fortey by the long-continued fractional distillation of Galician petroleum. It boiled quite constantly at 808°, and its specific gravity at 0°/4°, 0°7903, agreed perfectly with that of a specimen prepared synthe- tically by Markownikoff.? It was found that it solidified partially in a freezing mixture, but that the freezing-point was very far from constant (about — 12° to — 7°) ; and it was evident that it contained a certain amount of another hydrocarbon, probably a heptane such as tri-methyl propyl methane, boiling at a temperature near to 81°, and inseparable from thie hexamethylene by fractional distillation. Fractional crystallization was therefore resorted to; and the less pure fractions (76° to 78°5°), leit over after the chemical investigation, were utilized in addition to the small quantity of material of constant boiling-point. Special methods of separation of the crystals from the mother liquor were ' Fortey and Young, Trans. Chem. Soc., Ixxy., p. 873, 1899. * Annalen, cccii., p. 1, 1898. Youne— Vapour- Pressures, §¢., of Thirty Pure Substances. 385 devised ; and after a series of systematic recrystallizations had been carried out, a product melting almost constantly at 4°7° was obtained. There was still, however, a slight fall of temperature before solidification was complete ; but the quantity of material was too small to allow of further purification, and it was thought that the amount of impurity must be very small indeed. I have been recently informed by Dr. Timmermans that a specimen of hexamethylene has been prepared by direct combination of benzene with hydrogen in presence of heated nickel, obtained by reduction of nickel oxide (method of Sabatier and Sanderan), and that the melting-point of this specimen is 6:5°. If this higher melting-point is correct, it is evident that the hexamethylene separated from Galician petroleum was not quite so pure as was hoped; but since the boiling-point of the admixed heptane is so close to that of hexamethylene, it is probable that the errors in the observed physical constants are only slight. Specific gravity at 0°/4°, . : : . 0°79675 Boiling-point under normal pressure, . _80°85° 9. Benzene—A large quantity of commercial benzene was distilled, and was then twice frozen. It was repeatedly shaken with strong sulphuric acid to remove thiophene, and was finally purified by fractional distillation. ‘he melting-point was 5°58°. Specific gravity at 0°/4°, . : 5 . 0:90006 Boiling-point under normal pressure, . 80°2° 10. Fluorobenzene..—This compound was prepared as described by Wallach? and Wallach and Heusler*® by the action of concentrated hydro- fluorie acid on benzene diazopiperidide, C,H, - N: N - NC;Hw. By fractional distillation a product of perfectly constant boiling-point was obtained. Specific gravity at 0°/4°, . ‘ : . 1:04658 Boiling-point under normal pressure, ; 85°2° 11-18. Chlorobenzene, Bromobenzene. Lodobensene.~—These substances were obtained from Kahlbaum, and were purified by fractional distillation. Specific gravity Boiling-point under at 0°/4°. normal pressure. Chlorobenzene, : . 1:12786 132:0° New specimen and new Tb, . 1:12805 Bromobenzene, ; 0 aziltsyy 156-0° New specimen and new i . 1:52178 Todobenzene, . : ; . 1:86059 188:45° 1 Trans. Chem. Soc., ly., p. 486, 1889. ’ Annalen, ccxliii., p. 219, 1888. * Annalen, cexxxy., p. 255, 1886. 4 Trans. Chem. Soc., loc. cit. 386 Scientific Proceedings, Royal Dublin Society. 14. Carbon tetrachlorides—A specimen which was obtained from Kahlbaum was fractionated, and a product of perfectly constant boiling-point was obtained. Specific gravity at 0°/4°, . ; . 163255 New specimen and new tube, . : . 163257 Boiling-point under normal pressure, g 165752 15. Stannie Chloride.-—A very pure specimen was obtained from Kahlbaum; the first portions of the distillate contained a little hydrate, after the removal of which the liquid boiled quite constantly. The receivers in which this very hygroscopic substance was collected were placed in a chamber containing phosphoric anhydride in order to prevent absorption of moisture by the distillate. This dry chamber was also used for many of the other substances investigated. ‘The chamber resembled an ordinary balance- case; a hole about an inch and a half in diameter, rather above the centre of . one side of the case, was closed by a piece of sheet indiarubber; and the delivery-tube from the still-head or distillation-bulb passed through a per- foration in the indiarubber. ‘The receiver could be inserted or removed by lifting the front side of the case. : In all the experiments with stannic chloride the greatest care was taken to prevent access of moisture. Specific gravity at 0°/4°, : . 2:27875 (‘Thorpe®) Boiling-point under normal pressure, 5 YM IULePIS 16. Ethyl Ether..—The ether was prepared by heating absolute alcohol with sulphuric acid in the usual manner. The distillate was shaken with caustic soda to remove any sulphurous acid, and was re-distilled. It was then left in contact with calcium chloride for some time to remove most of the alcohol. After re-distillation it was shaken repeatedly with small quantities of water ; it was then dried with calcium chloride and distilled again. Metallic sodium was then added from time to time until all evolution of gas ceased, and the ether was distilled from the sodium. It was afterwards left in contact with sodium until required, and was always re-distilled over sodium before being used. ‘The specific gravity was not determined; but the mean of the values given by Kopp, Pierre, Mendeléeff, and Perkin, :0'7362, was adopted. Boiling-point under normal pressure, 34°6°. 17-26, Ten Esters—Methy/ Formate to Methyl Isobutyrate.— Determinations were made in each case with two different specimens. ‘The one was obtained 3 [bid., xxxyii., p. 89, 1880. + Ramsay and Young, ‘l'rans. Roy. Soc., clxxviii., p. 57, 1887. ® Thomas and Young, Trans. Chem. Soc., Ixiii., p. 1191, 1898. Youne— Vapour-Pressures, &§c., of Thirty Pure Substances. 387 from Kalhbaum, the other was prepared from the alcohol and either the acid or the anhydride by the usual methods. After removal of any free acid by treatment with potassium carbonate, the ester was treated repeatedly with phosphoric anhydride until, after standing all night, the liquid had the appearance of a thin jelly. This appearance was never noticed until the whole of the water and free alcohol had been removed by the phosphoric anhydride. If necessary, the ester was distilled two or three times during the treatment with the anhydride. It was then fractionated until a product of constant boiling-point was obtained. The esters were always re-distilled over phosphoric anhydride immediately before use. Specific gravity Boiling-point under at 0°/4°. normal pressure. +O Methyl formate, A, . : . 1:00820 31:9 ¥ Bein : . 1:003818 319 Hithyl formate, A, . ; . 0:94807 54°3 5 Bic 6 ’ . 0°94796 54:35 Methyl Acetate, A, . X . 0:°95984 57°3 (P) 3 ease i 2 10;95929 57°15 Propyl formate, A, . : . 0:92866 80-9 3 Bis : . 0°92870 80°9 Ethyl acetate, A, : ; . 0:°92488 7715 % : : : . 0°92484 77:18 Methyl propionate, A, i . 0:93874 19-7 3b B, F . 093868 79°65 Propyl acetate, A, . : . 0°91015 101°55 3 Ae : 0 OOION 101-55 Hthyl propionate, A, : . 0°91288 99-0 53 BL 4 . 0-9) 242 99-0 Methyl butyrate, A, . : . 091994 102°8 3 1B}, 6 : . 0:92010 102-75 43 Oh 6 ; . 0:92016 102°75 ‘3 De : . 0:920038 102°75 Methyl isobutyrate, A, ; . 0:91186 92-3 59 Be be . 0:91126 92°3 © The following determinations of specific gravity were made at later dates with fresh specimens, and in several cases with a different specific-gravity tube. Methyl acetate, 0°95937. (Boiling-point, 57-1°.) Ethyl acetate, 0:92438, 0:92437, 0:92446. Propyl acetate, 0:91008. Ethyl propionate, 0:91251, 388 Scientific Proceedings, Royal Dublin Society. 27. Methyl Alcohol..—The specimen was prepared by the action of ammonia on re-crystallized methyl oxalate. The distillate was rectified, and was re-distilled, first over quicklime, and then over barium oxide. It was finally distilled six times over small quantities of sodium. As the boiling-point was not quite constant, the alcohol was fractionated. The boiling-point was 64:7° under normal pressure; the specific gravity was only determined at 22:94°, and was found to be 0°78909. It has since been observed? that the last traces of water can best be removed from methyl alcohol by distillation through a very efficient still- head. Purified in this way a specimen of the alcohol boiled at 64:7° under normal pressure; its specific gravity at 0°/4° was 0°81000. Ethyl Alcohol.A—Absolute alcohol was frequently distilled over lime, and finally over a little sodium. ‘he boiling-point was quite constant. The specific gravity was not determined, but the value 0°80633 was adopted. Ethyl and propyl alcohol can best be dehydrated by distillation with benzene through a very efficient still-head.t A specimen of ethyl alcohol, purified in this way, had the specific gravity 0:80634 at 0°/4°. By subse- quent distillation with normal hexane, the specific gravity fell. to 0°80627, owing probably to the removal of a trace of benzene. This value agrees very well with that of Mendeléeff, 080625. Boiling-point under normal pressure, 78°3°. Propyl Alcohol.—The specimen was procured from Kahlbaum. It was purified by fractional distillation, potassium carbonate being added to the lowest fractions each time to remove part of the water from the mixture of minimum boiling-point. Specific gravity at 0°/4°, : c 081929 Boiling-point under normal pressure, . 974° At a later date® a fresh quantity of propyl alcohol was obtained from Kahlbaum, and, after being purified by fractional distillation, was distilled with benzene through a very efficient still-head to remove the last traces of water. The specific gravity of this specimen at 0°/4° was 0°81923, and the boiling-point under normal pressure, 97:20°. Acetic Acid.'—The specimen employed for all the determinations had 1 Ramsay and Young, Trans. Roy. Soc., clxxyiii., p 313, 1887. 2 Fortey and Young, rans. Chem. Soc., Ixxxi., p. 717, 1902. 3 Ramsay and Young, Trans. Roy. Soc., clxxvii., pt. i., p. 128, 1886. 4 Young, Trans. Chem. Soc.,Ixxxi., p. 707, 1902; Fortey and Young, ie= ei) ~ « FL80GF66-1 | FISTS8G6-I | EFIZ9II-0 | 918908-I- | SeL6L89-0 | 986FL8-F- | 11800L-9 “Dot : “(OTT — <0) ploe o1e0W $Z0L¢966-1 | GFIF9TOO-0 | Lo9G0¢¢-0 || L8GGGS-E— | 6GOCT6E-I | FGE9FG-0 OLEGLE-F 0Z — “Dot "+ + foyoore TAdorg FEHESIEG-I | FELLEE00-0 | FEs0G09-0 | Z9GLZO-F— | TET90F9.g | STLEFO-0 080ZL0-¢ 10) se Toqooqe TATE 96166966-1 | 91F88666-1 | OLLE¢88-0 9S86GF-6— L8¢6F9G-1 896G0F-8I- | 960L08-66 Oe : “foyoore [ATQET 16888966-1 | $190F000-0 | 6SFLFOF-0 | 9SF6ES-2-— | cOL90FG.0 | OGLZLF-E OZIEST-0 “Dot 7 + foyer hqnqosr [Ate 99F61966-I | 8L161000-0 | S6zFOGF-0 | SLIIGS8-2- | 0F00LZ6-0 | 998ZSF-8 £0F690-¢- | 01+ "Oo} ore ‘ayerAqug [AqI9T $8086666-1 | 66199000-0 | 86LZ#8F-0 8986F0-E— 18990FZ-0 9LFOFL-1 GE9GE6-T Oat Oo? ; ‘ ‘oyeuordoad [Aqy of OF108G66-1 | ¥86F1000-0 | S600TSF-0 | FFGFZS-Z- | LLo99G0-1 | 00GZ6E-TL | 0G0666-2- | OL + “Od 7 7s “ovejeoul Adora 68ILF66-1 | 10669666-1 | Gs890T6Z-0 | L61896-I- | Og0ZT0G-0 | L0FG96-L2- | SFOFLZ-IT Oot * + ‘9yvurdoad [AyQOT 9ZEEES66-1 | 9FFF96G6-1 | 8IE0ZLZ-0 618018-1- FOFO9GS-0 OT98LT-2—- CSOSEF-01 “Oot y 8 ‘ ‘aqnyaor [AYE 1810966-1 | 6IFLL000-0 | FFIZELE-0 | FFIT9S-- | GIGSsEI-o || OGSLEs-T 09818¢-z OL = os} "+ + yeunt0 [Adour 69CTEC66-T | 68819666-1 | S8I01LG-0 | 69F998-I- | OLF0988-0 | LETG69-2— | 96GFOT-IT OL + “Dot oo nae ‘9yvzaov [ATO GIPS8FG6-1 | 99ZG1666-I | L166191-0 G80GEF-1- GEF0G19-0 L10€60-4- 66CE91-2 OL + 00} : : * foyrutog AY YT 8L86¢C66-1 | 0669F866-1 | GFEC099-T T¢9LEF-0- 199047451) OSFLFS-E— 1¥8166-9 OL — “Oct E e + ‘oqvuntog [MG I0 TL 81F86Z66-I | Scosegs6-1 | 6L9Egog-1 | GEcEF9-0- | FFFOSSe-0 | SOFFI9-E- | 069L08-9 OL + “Oct et ean “ore Aqag SILIE966-1 | Z1Z9Z000-0 | 1Tzg9FeF-0 | ZE68F8-Z- | S6GFSS9-0 | GFIEIS-F OFSPOT-T “Dot ; ; * “opmoryo oruueyg 99416966-1 | ¢¢89z000-0 | gteexe¢-0 | S99ELT-s- | 980Z9TL-0 | 8SFG0Z-¢ 0&ZL0¢-1- “Oot "+ fapHoryaeajzey woq.ted S6LG0966-1 | LEset666-I | seG1sss-0 | OGSFFF-G- | 619GGL9-0 | GSOGOL-y- | O8EIEE-L 08 = Ox - + + —_ ‘suazueqopoy GP66LE66-1 | LZ90E666-1 | 1LzGzL0-0 | FE6GL8-I- | G6IGZF9-0 | Zeeo6E-F— | 99T9ZO-L 08 — “Dot + + + uazueqoworg TGS9GGG6-T | ZLLES666-1 | 96E96eE-0 | LF6GST-c- | GL969L9-0 | L66ZSL.F— | PHIGEE-L “Dot ieee UC 2U2K O10) $166966-1 | 00F62000-0 | ExeLsoF-0 | LG0S9S-- | LOFSESS-0 | PE8Z6L-T &C1680-6 “Dot 2 88 SSNS ERNETFONOANLTT 68468466-1 | 0Z0EI000-0 | Oggsces-0 | 999FSI-G-— | OSFLESO-T | SOL9ZI-ZI | GPISTS-8— “Oct ae Si OS evar 8#180966-I | G116F000-0 | 96¢089-0 | SLLeeE-c- | ODISTLE-0 | T¢0FSE-c LOGHCF-1 “Dot "+ + Quay Ayjeuexe yy GLL61966-1 | 9ZFFOTO0-0 | 9FLL61F-0 | EOGST9-Z- | E9LSLZ6-I | 9869F8-0 | 160ST6-% OL = “Dot eee TENE $9088966-1 | 00Z6L666-1 | Tg0g0gs-0 | 6ZFGG69-I- | FOSTLZO-1 | 6FSGF9-OT- | 8LZZ00-FL | OT + “Oot ae ee Adordosr=r(y L60F8F66-1 | 19489666-1 | 000G261-0 | 96L¢19-I- 181L706-0 6E10E0-8- | GO9TLE-11 08 + “Dot : i : ‘ouvjuedosy 1¥866966-1 | 16GF€000-0 | GFSg6IF-0 | S¢esz9-c-— | 6E8G0E9-0 | 0989-4 0¢1068-0 OI — “Oot "+ + fgue}90 [RULLO NT 1L€96G66-1 | 80€S000-0 | STOOTOF-0 | S89L1G-4- | GE6LESE-0 | OFOTEF-Z SPIGFL-T “Dot eaaeac > guejdoy [eULON, GFL88966-1 | ESEZF000-0 | 99TSI9E-0 | I88s6z-c-— | O18G90E-0 | ¢L6L0Z-€ 90ZF0G-0 O01 + ‘Oot - + + Suexey [BUION — S098FF66-1 | 1899Z666-T | S6F0S60-0 | G8zEgez-I- | 919899-0 | OL6LEg-F— | O18ZZ9-2 0% + ‘Oct 7 +s ‘ouejuad [RULON g Soy » Soy 2 Soy 9 q Soy q v =} ‘gounjsqng Jo omEyy 92 + 09 + 9 = d SOT VIONYOA SaLOId WOA SLNVLSNOO Youne— Vapour-Pressures, &§c., of Thirty Pure Substances. 418 MODIFIED FORMULA OF CAILLETET AND MATHIAS. 8, = § tat + BL (+ ye). Name of Substance. So. a x 107, B x 1010 y x 1018 Normal pentane, . 5 4 0 073232 — 4610 — = Normal hexane, . a 0 : 63388 — 4445 — a Normal heptane, . ° S 4 0°3504 - 4192 — 621 — Normal octane, : j 3 3 0°3592 — 3986 — 960 — Tsopentane, . : ; : : 0°3202 — 4658 + 463 — Di-isopropyl, . : . : a 0°3401 — 4445 + 413 —_ Di-isoburtyl, . 5 F 0 : 0°3550 — 4116 — 592 — Hexamethylene, . ° . 0 0°3985 — 4685 + 791 — Benzene, : 9 : ; 5 0:4601 — 5248 + 693 — Fluorobenzene, : 3 5 5 075236 — 6000 + 293 — Chlorobenzene, : 2 a 0°5640 — 5337 — 509 — Bromobenzene, i 5 i 0 0-7609 — 6655 — 725 — Iodobenzene, . a ; 3 6 0:9303 — 7556 — 519 _— Carbon tetrachloride, 6 ‘ : 0°8165 — 9564 + 1480 — Stannic chloride, . : ; : 1:1387 — 12760 +977 — Ether, . a ° . : 0 0:3685 — 5377 — 475 — Methyl formate, . : ° : 0°5020 — 7013 — 665 — Ethyl formate, . ° F 0 04741 — 6251 — 694 — Methy] acetate, : 4 ° 5 0-4799 — 6280 — 1467 — Propyl formate, : : : ° 0°4647 — 5748 — 459 = Ethyl acetate, : 6 6 é 074624 — 5992 — 764 — Methyl propionate, . Bie se j 0:4696 — 5921 — 729 — Propyl acetate, c 9 a : 0°4553 — 5469 — 1124 — Ethyl propionate, . 9 : : 0:4564 — 5644 — 784 — Methyl butyrate, . é 5 : 04601 — 5430 — 906 — Methy] isobutyrate, : 0 0 0:4558 — 5593 — 689 = Methyl alcohol, 2 : s 3 04050 — 4479 + 1330 — 23760 Ethyl alcohol, 5 5 3 5 04028 — 3827 — 5940 + 661 Propyl alcohol 5 é : : 0-4095 — 3790 — 3750 — 5533 Acetic acid, . 3 . 5 ; 0°5355 — 5366 — 1191 -- SCIENT. PROC. R.D.S., VOL. XII., NO. XXXI. 3L Scientific Proceedings, Royal Dublin Society. 414 (Teor119) 991-8 = 8286-0 = (e6gz-0) | (eogz-0) 0 (c08-#) (gog-) fig = £9092 001GS | %-L6T 026-8 g+ 8ZES-0 0282-0 8912-0 | GLFZ-0 IL-8 G19-F CHO-F GP — CFOS 060% | S1-L6T 168-8 a= €686-0 9282-0 0602-0 | 0982-0 6L-F C8). 906-8 Ge= |. Manse 0906 | T-L6T 696-8 I+ FZEG-0 &ZES-0 6002-0 0F9Z-0 gG.9 066-7 88)-€ 66 — 16643 0z0e% | 0-261 61-8 i= $Z86-0 3286-0 0961-0 1692-0 = 001-4 911-8 OF — CS6FG G66FG | 6-961 OFT-S 6 = GZ8C-0 LGE6-0 61-0 0826-1) = G61-4 99-8 TF - 6167G 096%G | §-96T 910-8 I+ 9286-0 6283-0 I¥8I-0 6082-0 ZI-01 OSF-¢ 09¢-¢ OF — OIS8FZ OSS | G-961 918-3 6= 8686-0 0883-0 9FLT-0 166-0 8.21 081-6 OLSF-8 9g — 6Z9FZ 699FG 961 €89-% 7 eE86-0 LE8Z-0 609T-0 6908-0 99-81 1Z-9 6296-8 91 + TLZ¥G CCGG C61 99F-% pe GPES-0 9F8T-0 OFFT-0 868-0 86-61 6-9 C41 0-¢ 93 + 999% OFSES S61 966-6 i= 9480-0 1G83-0 6961-0 CFFE-0 89-46 88.) GZ06-G oF + OFS 00% 061 610-6 6= FLEG-0 9183-0 60IT-0 eF98-0 L¥-63 Z0-6 OGFL-3 cg + GCZIG O611Z 981 68-1 i) 4 Z0FG-0 10FZ-0 6860-0 | 1988-0 10-68 69-01 098¢-% Lt Z9E61 GFS6I 08st 801-1 0 SFF-0 SPFG-0 cgl0-0 | Z9TF-0 90-GF 19-81 0G0F-G gos egcgt OFS9T OLT 81C.T I+ F6FG-0 8653-0 1660-0 P6EF-0 GP-LP 6-91 0916 8% — GE0FT O90FT 091 PLP-T 0 0F9G-0 OFGZ-0 9140-0 F09F-0 68-ZE 00-12 OGLI-Z 13 + 92811 GOSIT OST 668-1 0 1896-0 1892-0 980-0 L8L¥-0 8G-9¢ 06-6 0680-4 0-3 — | 0-0686 686 OFT 0Z8-T i= 696-0 896-0 0180-0 LG6¥-0 8-09 06-28 GLI0-@ 0-91 + | 0-028 L818 Ost 896-1 0 6197-0 619-0 0920-0 1919-0 8F-F9 66-68 0866-1 GI -— | ¢-0F19 GPLO at $G6-1 0 6716-0 218-0 | ¥G0Z0-0 8FGS-0 18-49 ¥-6F 6606-1 GQ) = |) Grae S8PG OIL 061-1 i ILLG-0 0112-0 | 96910-0 1186-0 ¥6-59 ¢-19 C6C8-T 6M = 1-606F OFF 001 961-1 1 1182-0 9182-0 | 68Z10-0 8066-0 81-6. 9-LL GLI8-1 Oi sr || Googe S6FE 06 8ZI-T 0 8986-0 g98c-0 | ZLOT0-0 ¥Z96-0 ge-GL, 8-86 OSLL-1 TL + T-3FLZ C81z 08 FOL-T 0 6062-0 6062-0 | Z98L00-0 6819-0 LYLL G-LEL FOPL-T OG = 0-FIIG 61IZ OL 80-T 0 6663-0 G¢6Z-0 | 00900-0 0g8¢-0 10-08 1-991 6601-1 Ge = | 8-109 G09T 09 190-1 0 1008-0 1068-0 | ¢+¢F00-0 1466-0 81-28 0-02 9819-1 8% — | G-06IT S61T 0g 690-1 0 SF0S-0 8708-0 | 1900-0 6909-0 18-48 G-16 G649-1 fel = 8-498 S18 0F IFO-T 1 = 7608-0 460-0 | 19'Z00-0 | 919-0 91-48 0-80F 0229-1 O48 sp || (xi) 6-019 08 — 0 0F18-0 OFTe-0 (1100-0) | €9%9-0 = = 196G-T | Gt-§ + | ¢9-¢¢F G-0GF 0z — 0 9818-0 9818-0 eae 0) | 09¢€9-0 =_ = FOLG-1 | G8-1 + | G9-88¢ 8-182 OL — 1+ GEGE-0 188-0 (g000-0) | 6&¢F9-0 — = 9FGFS-T | GT-0 + | OF-881 GZ-E8T 0 = — = — — = = = = 36-0 — | GO-FIT 08-FIT Or - = — —_— - _ — — = = 0 68-19 68-19 0% - = — — — — — — == aS Gc-0 + | 0¢-8¢ 6-18 A093 = “sorsoTe/), “O70 O"0 *cUL pacegerg “TU S40 PL" |-nayeqnozeg | “poases Loge ~pmbr POE yorxv |e% TOTO | “Pp 10 samode 6 poyeangrg | PONT Sonal poyeanqeg payenyeg pmbry “UOT} 7 PERKIN HO) VEE) | apeasrueg 0 Ayisue(y “mode A pue pmbry jo wee fl ~BzLIOde A *soAIno WOT prey aany RONG ae Jo -viod way, 0} [eujay yoy fo o1ey + Azisue (y “mR % JO eTIN]O A ommssarg-aimode 4 “ANVINGd IVNYON 415 Youne— Vapour-Pressures, &c., of Thirty Pure Substances. (Teorgttg) Teg-2 — FPET-0 = FPSS-0 FPES-0 0 992-F 996-F A= EEEZS OSG | 8-F8% G66: ¢ + 8FEs-0 CPSZ-0 LO8T-0 288Z-0 ¥-01 6G-¢ 069F-€ 68 = 181% 0ZZGG 8G 081-2 y+ $¢83-0 6486-0 8GyT-0 0708-0 99-81 80-9 O68%-E (6 = OLYTZ 0681Z Seo GPF-G i= 9986-0 1983-0 GOFL-0 6288-0 1-61 (éliol} 0F00-8 @ 14603 64606 08% 861-6 e+ $886-0 0882-0 8061-0 1G98-0 61-43 18-8 CLI8-% 8G + S8L6L C9L6T 963 266-1 t= OLFZ-0 113-0 IL0T-0 018s-0 18-08 68-6 0969-3 EB 4p SeI81 OOTST 083 OLL-T 6 = CCFS-0 LG¥S-0 66810-0 POIF-0 6-18 99-21 OGZF-% or + 619GT LLGGT O1Z 029-1 0 6643-0 66FG-0 62890-0 cgat-0 GL-GF 08-ST 0166-3 9¢ + egsel 6FESI 00% 91G-T 0 8FoS-0 ePes-0 ee1¢0-0 OLGF-0 | O1-LF OF-61 O881-Z LG + LOFIT O8SIT 061 LEF-1 it 4F 8862-0 1863-0 8240-0 IGL¥-0 $6-0¢ @9-8% 0GOL-% #8 + | ¥-6996 1996 O8T GLE-T @ ar 893-0 0893-0 GLFS0-0 S16F-0 F8-FS 8-83 6480-3 L-0 — | 6-3a18 S318 OLT I1e-1 7+ 2196-0 $88z0-0 $90¢-0 g9-LG g-G8 OGL6-1 Bi = |) LePMNO S619 091 LGG-T a+ 61Lz-0 66660-0 10@¢-0 80-19 G-8F 6026-1 ¢.¢ — | g-909¢ 019g 0ST C1G-T La C91Z-0 99810-0 $FS¢-0 8-89 9-8¢ CTLS-1 0-e1 — | 0-€6¢F 909F OFT 641-1 i + 6082-0 F0G10-0 1946-0 §9-99 ¢-99 0668-1 Toi? a |) Ween 6ILE a LPI-T ats EG8Z-0 FG86-0 G0Z10-0 88g¢-0 66-69 6-88 F68L-1 7-0 + | F-60866 386% 0Z1 GZI-T 0 6682-0 6686-0 9¢600-0 S0L¢-0 LP-TL 9-401 FEGL-1 0-0 0-898 8G8s OIL SOT-T 0 FF6G-0 FF66-0 ¥9100-0 F1SG-0 SPS 1-861 LOGL-1 9-1 + | 9-281 9g8I 001 169-1 0 8866-0 8866-0 ¢8200-0 8164-0 16-1, 0-TIT 1689-1 ¥-% + | ¥-G0FL LOFT 06 10-1 i= ZE08-0 608-0 9F400-0 2209-0 eG.) 0-466 6099-1 0-0 0-Z90T 901 08 990-1 T= 1108-0 8208-0 | £9800-0 GZ19-0 61-64 |° 0-166 7209-1 | 00-6 — | 08-F8L 0-182 oL LG0-T 6= IGTS-0 ez1e-0 | 8800-0 1229-0 8-08 0-G0F C109-1 | ZF + | 39-196 6-99G 09 = (6 +) 99te-0 | (g9rg-0) (8100-0) | 8189-0 = = 8z89-T | 00-0 06-00% 6-00 0g = (G +) 0168-0 | (Z1%e-0) (g100-0) | GLP9-0 = = FOGG-1 | 98:0 — | $8-G2% 1-91 Ok = (@ +) GoS-0 (1¢c8-0) (6000-0) | ¢099-0 = > FLeS-1 | 00-1 — | 0F-FST PGS 08 = (I +) 66z8-0 | (00¢¢-0) (9000-0) | §699-0 = = e9I¢-T | 8¢-0 — | GF-6IT 0-061 0G = (0) Ppes-0 | (FEees-0) (¥000-0) | €899-0 = = e96r-L | 88-0 — | L9-FL 00-92 Or = (z +) gges-0 | (98ee-0) (000-0) | 6949-0 — = ZLL¥E-T | $9-0 — | 66-7 ChCP 0 = = = = = = = = = 00-0 06-6 06-26 0r— = = = = == = = = = CI-0 + | @-F1 OL-FT 0% — = = = = = = = = = G0) + | Grd 66-9 a0 = “SOLIO[B/) wy) aia) “UU “WU “mul "SqQ-"0T8Q |. bites “mode A an ‘modea | 01 *V Cee ol laenods NN) omni pega | ae: *sqQ—9[%) |. 29 | ‘peares Pees "| payeanyeg Weel guoty Vv [PERTOHO) | TE SERO) *apRistyud,) SOE -mmod b PAROS *saAdnd WOT PRA any [eorjexoa yyy, inode, puv prubry Fo uroyy yo J IF PVA . 0} yenjoy |— qyeoyy ~erodwa J, JO o1jByy “AJISUO “UIRID) B FO BUNTOA -ounssotg-inode A. “ANVXHA IVNUON 31 2 Scientific Proceedings, Royal Dublin Society. 416 ee ; | ({eo1st19) FG8-¢ = 1F8Z-0 = (1982-0) | (1#8Z-0) 0 (G13-F) (G12-F) ig = 66802 0840 8-99 LE1-8 = ShESZ-0 1982-0 G6S1-0 6182-0 09-8 086-9 O8FG-¢ = 6620 = €.99% 696-G e+ CFES-0 GFET-0 SLL1-0 1062-0 9-01 G69-G OOFF-S = 1e10z as 992 GELB c+ 0G€-0 GFEZ-0 1891-0 6208-0 = 81-9 0693-8 — FLS61 = 69 1G9-% e+ G¢8z-0 ZGET-0 SEST-0 9918-0 8-41 09-9 06S1-¢ (7) C6C61 O196T 79% PPPS 0 F98S-0 $98C-0 1681-0 ZEEe-0 = 9T-L 0100-8 = 17061 = 09% 608-2 0 GLET-0 GLEZ-0 1821-0 LGFE-0 06-12 LLL 0968-4 IF + IIS8T OLFST 09% LOI-Z i= 06&-0 1683-0 LII1-0 $998-0 = 66-8 0661-3 = 6LPLI — 96% 676-1 ¢+ LI¥Z-0 ZIFZ-0 19#60-0 LL88-0 GG-18 19-01 0819-4 or + 0G09T 0861 0Sz SZL-1 (tt Z9FG-0 1942-0 9410-0 LLI¥-0 GF-Le SF-81 OF6E-G 13+ 11881 O6LE1 04% 186-1 0 1063-0 1063-0 Z0090-0 FIPF-0 1€-6P 99-9T GG96-G Ly + LOST OI8II 0&G F8F-1 w= ZEEZ-0 $063-0 Z6S8F0-0 919F-0 9F-9F FF-06 G99T-% Got OSI0T SOOT 0% 00F-1 0 162-0 169-0 G00F0-0 S6LF-0 11-0 16-43 6980-4 9-71 + 9.8098 $668 O1Z 68-1 0 1792-0 1¥93-0 $0€80-0 ZG6F-0 LL-89 1Z-0€ G610-6 8-01 + 8-1L3L 1962 006 283-1 ot 9892-0 $89Z-0 11LZ0-0 9606-0 60-9¢ 8-98 0696-1 $.9 + 8-1019 ¢609 061 9FG-T Tes 6212-0 8Z13-0 BFZS0-0 GECS-0 BS-8G 9-FF G116-T 1-01 — 6-080¢ 1606 Ost 91-1 T+ ELL<-0 GLLZ-0 SF810-0 6984-0 8F-09 I-#¢ 0998-1 0:91 — 0-961F G1GF OLT L8T-1 T+ LI8Z-0 9186-0 ITST0-0 I8FS-0 69-29 6-99 GFZ8-1 L-91 — €-S8FE OctE 091 66.1 T+ 1982-0 098-0 ZGZ10-0 869-0 18-49 8-18 Z98L-1 OO = F-08L3 F813 Ost a iP 606-0 ¥06Z-0 | €82600-0 ITL9-0 I-19 G-Z01 OISL-T Ai = 1:92 6606 OFT 801-1 r= 8462-0 6762-0 | $9LL00-0 1289-0 18-69 8-831 O8TL-T 6-9 + 6-611 EGL1 Ost 80-1 i= 7662-0 £66Z-0 | ¢20900-0 9269-0 69-TL 9-591 GL89.1 T-¢ + L-ZL81 L981 0ZT 990-1 B= G£08-0 1e08-0 | 9000-0 1209-0 8-1 6.316 1669-1 €:9 + €-SS01 L¥01 OIL 90-1 eS 6106-0 0808-0 | L69¢00-0 FZ19-0 08-2 0-823 0889-1 1Z-0 — | &6-#6L 3-96) O0L 880-1 Ca IZ18-0 $Z18-0 | €0L200-0 8129-0 LL-LL 0-018 Z809-T 90-0 — | #2-88¢ 8-88¢ 06 080-1 0 ¢9Ts-0 cgTs-0 | 000Z00-0 1189-0 FP-6) 0:00¢ 9486-1 97-0 + | 90-20% 9.98% 08 180-1 0 8028-0 80ZE-0 | 09F100-0 G0F9-0 FF-08 0-989 129-1 1¥-0 + | LL-c08 €-B08 OL — c= 0628-0 1SZ8-0 (1100-0) | 169-0 = = 904-1 ZF-0 + | 38-606 6-80 09 — t= G668-0 €628-0 (1000-0) | 62¢9-0 = = 006¢-1 91-0 — | #L-0FT 6-0FT 0g — 0 Gees-0 ceee-0 (¢000-0) | ¢999-0 — a €00G-L | L£¢-0 — | 82-16 60-36 OF = 0 1188-0 LLe8-0 (g000-0) | 1¢29-0 = = E18F-1 87-0 — | 18-29 CE-89 08 _— I+ 0ZFE-0 61FE-0 (2000-0) | 9889-0 — = 6G9F-T #80 — | 91-98 0g.¢g 0z = UF Z9FE-0 198-0 (1000-0) | 0269-0 = = OSFF-1 10:0 + | 18.06 06-02 Or — 1+ 708-0 g0ce-0 | (L0000-0) | 8F002-0 = = 9LBF-1 0 GF IL CFI 0 SOTLO[BO "9°0 ‘0°O “mul “tut aeraeed o) *840-"91%0 |.nanem ETT mode A enn ‘smodva | 01% V PO 38 smodva | . payeingeg | PNET “8qO-"9[8D paqenyeg payeanqug pmnbry “u0N = “payepnoyeg | ‘pearesqg, apertniag| jo Aqrsueqy ‘node, pus pmbry jo uve ARON “SOAIND WOLF pvayy any [EERE AT eae, J -e1ad wa J, 0} [envoy yeoH = JO O1jey of 3 : ApIsua(y Wel @ JO OUUN]OA amnssotgq-inode A “UNVIdHH TIVNYON 417 Younc— Vapour-Pressures, &c., of Thirty Pure Substances. ({eoatt)) 698-8 = 1Z8Z-0 = (1382.0) | (2682-0) 0 (163-4) (166-4) Fate FELSI OgL8T | %-96Z S0F-G i= 6283.0 988T.0 9FET-0 | 9988-0 01-61 SPL 0GL6-% (har S6ILT 9BTLT 06 686-1 0 L0FG-0 10%Z.0 | 8860-0 | 8188-0 96-83 LT-O1 0619-% Go) = GFGFL ETOST 08% OLL-1 0 SFFS-0 8FFG-0 | 9TLL0-0 | S@IF-0 82-78 96-21 OSGF-% SF — LE6ZI 08621 OLG 69-1 G= 16FZ-0 £653.0 | £6z90-0 | F98F-0 1-68 10-91 0262-4 66 — 9STIT GSIIL 09% LGG-1 ot 6896.0 689.0 | SII¢0-0 | #9S%-0 08-GF #9-61 0961-4 Gee —| 3-81¢6 196 0% BSF-1 Ce 0892.0 8192.0 | LEZrO-0 | GELP-0 16-9F 9-83 GSI1-% ¢-F8—| ¢.F818 6128 OF LLe-1 0 FZ9G-0 $296.0 | F8FS0-0 | 106F-0 LG-6F 1-86 COFO0-% 0-1 + 0-9¢69 8469 08% S1e-1 I= 699-0 0196-0 | 2820-0 | £90¢-0 0-29 8-FE 0616-1 9-1 + 9-FL8¢ 8189 0% 993-1 0 8113-0 €1L3-0 | 9820-0 | 6819-0 88-46 8-6 OLZ6-T 6-1 + 6-806F LOGF 01% L8G-1 i= 9616.0 LGLZ.0 | ¢9610-0 | L189-0 19-96 6-08 8088-1 0-9 + 0-C0LF 6607 00 F0G-1 i= 0083-0 108%-0 | 1910-0 | TPFS-0 88-86 0-29 088-1 1.8 + 1.0688 B8Ee 061 FLL-L i= 8786-0 FF8Z-0 | 9IST0-0 | 9999-0 16-09 0-9 0008-1 9.0 P-CLLZ 9LLG OST SSL-1 0 1883-0 1883-0 14010-0 | 2999-0 9-29 ¥-86 LFOL-T 0-9 0-663 GGG OLT 81-1 T+ 0862-0 6263-0 | 16¢800-0 | ZLL¢-0 GL-B9 F911 968L-T 9% — F-C081 L081 091 ZIL-T 0 162-0 GL6G-0 | £98900-0 | 189-0 GF-99 L-GP1 TGOL-1 6-1 + 6-9GF I rca 091 60-1 ar 6108-0 FL0S-0 | $0FG00-0 | €269-0 83-89 0-481 IPL9+1 ¢.0 + G-FLIT PLIL OFT I80-1 I+ 8408-0 1¢08-0 | 61ZF00-0 | 1409-0 0-04 0-18 GLE9-T F-0 + F-668 668 ost FLO-1 0 0018-0 0018-0 | 4#Z800-0 | 8919-0 SLL 0-808 FIG9-1 6% + 6-679 ¥-9F9 0z1 = 0 GPIE-0 GFIE-0 2200-0) | 0929-0 = = GL6G-1 6-0 + 8-68F 6-18F a! = 0 F81g-0 F818-0 1100-0) | 19-0 = = GELG-T G9.1— | 96-198 9.668 001 = n= &ZEE-0 9668-0 (g100-0) | 8 -0 =a = GEGG-1 G6 — 0-193 G-896 06 == 0 196-0 198-0 6000-0) | 999-0 oe = 9286-1 0 8-FLT 8-FLI 08 = 0 808-0 8088-0 (9000-0) | 1199-0 = = LEIG-1 g9.0 + | ¢e.gIT 6-LIT OL = 0 6FE8-0 6Fe8-0 (4000-0) | ¥699-0 = a 686F-1 gc-0 + | OT-8L CG. LL 09 = i= 0668-0 1688-0 (g000-0) | 829-0 = = FSLE-1 8F-0+ | $8-6F G&-6F 09 = 0 1¢h8-0 Tee-0 2000-0) | 0989-0 = = LLG¥-T FI-0— | 12-08 8.08 OF = I+ GLPE-0 IL¥8-0 (1000-0) | %69-0 = = COPF-T 61-0 — | 12-81 0F-81 0g = T+ Z1G8-0 I1G8-0 1000-0) | %Z02-0 = = OFGF-T 01-0- | ¢8-01 CF-01 0% = I+ GS68.0 1648-0 = ZOTL-0 = = 080-1 0 29-9 9-9 OL = 0 66-0 G6G8-0 = SF81L-0 = = 8168-1 40-0- | 06-6 76:6 00 “SOLLO[VD) "0'0 i ba) maeedeeg “mur pategoeg ‘sqQ—0T%9 |. = Parcs “mode somali “mode A cADIL PSY Cee eA igor mode A. . Peas Bie “Sq Q-"918O, poyeinyeg paqnangeg pinbry “u0r} Vv “poye[nayey | “pedresq() -opeisnuag Jo Ausued | smodea pue pinbry Jo wroyy ~MAHIOUwA | soaano WoIy PLOY aan} [eorje1oey f, ee jo -wredway, 0} [enjoy yoo Jo O1yery *Aqisue (qT “URI B JO QUINOA *aanssa.tg -Ino0de A “ANVLOO 'IVNUON Scientific Proceedings, Royal Dublin Society. 418 (-yeo19119) GEL.€ = SFET-0 = (482-0) | (eh¢s-0) 0 (893-F) (892-#) Bik = ¢00¢z 810G% | $-L8T LEIS OS 9FS6-0 9682-0 TS61-0 1916-0 10-8 GGI-G 0369-§ KG = 62846 O88hG F-L81 96-6 if SF LEET-0 9FES-0 SES8T-0 L¢86-0 €F-01 CoF-G 0006-8 Ai = SILEG OS LEG L81 GSL.3 ieee 6SEG-0 &986-0 9L9T-0 8608-0 10-F1 ¢96-¢ 0G08-€ 06 — OSEFS OLEFS 98T 666-6 = 96-0 8¢E<-0 FLET-0 GPIE-0 L¥-91 GGs-9 O€ST-§ Sui? C6686 000%G ¢8T £0F-G ieee ¢9&6-0 9986-0 SLFI-0 I1é€-0 68-06 GO-L 000-€ 6G — 88686 OLEES esi C1G-G ib oF 6186-0 8186-0 861-0 86FE-0 G9-FG C6-L O8¢8.3 Se 69GGS OLGGG 081 €F0-G Crt 9686-0 8686-0 TOLT-0 F698-0 L1-66 80-6 0101-4 0G — 02606 01606 OLT GL8-1 0 £GFG-0 SGPG-0 L§€60-0 FL6E-0 86-FS IL-01 0G6¢.% WG = F606T CII6L OL LL9-T it 4p 693-0 85#6-0 686L0-0 90GF-0 LG-1F GL-T OSLE.% ces CSG9L O0ZE9T Oot L¥G-T 4= 166-0 F1GZ-0 FE8C0-0 CPEF-O 19.9% FI-LT 0096-3 ‘Cats FOSEL GO8EL 0ST PGP-T (6 op 6996-0 LEGs-.0 8240-0 SPOF-0 L0-1¢ C116 OSS1-Z ial = OZ9IT FE9lt OFT LLE-T Tiss $09Z-0 | ¢€09Z-0 1€880-0 97SF-0 L0-¢¢ 1-96 0ZL0-4 1-66 + 1-9016 0-F896 Os1 FIE-1 ee 0692-0 1296-0 90160-0 166F-0 G9-8G G-GE 1600.3 6-15 + 6-6608 0-SL08 061 9G: eu oor 9696-0 696-0 1960-0 OFTS-0 18-19 8-68 CCF6-1 §-61 + T9699 8-8¢9 Orr SIZ: Te ok TPLZ-0 OFLZ-0 0G0Z0-0 8169-0 $8-9 ¢-6F OF68-T 0-6 + C-£GE¢ ¢-GhEG Oot G8I-1 0 1816-0 L8L16-0 TG910-0 S1PS-0 T¥-L9 1-19 CLES8-1 SFI + 9-C6GF €-186P 06 9¢T-T (= GE8C-0 FE8S-0 L8Z10-0 OFoS-0 08-69 LLL ¢c08-1 8-21 + 8-00F§ 0-€88& 08 ¥81-1 bas 8182-0 6186-0 S1010-0 9¢9¢.0 c6-1L 1-86 6191-1 fey = 0-8996 L-9696 OL 601-1 Tack: 6666-0 FG66-0 | L€8L00-0 6919-0 CS-FL 9-261 6ZEL-1 6:0 — 9-CE06 ¢-9806 09 980-1 0 0166-0 0162-0 | 996°00-0 T88¢-0 98-91 6-L9T CO0L-T Gog” = G-SESL L-GEST 0g 190-1 Tea L10€-0 9108-0 | 9¢FF00-0 886¢-0 G1-6L F-F66 0019-1 AG = T-Télt C-OFIT OF 020-1 0 6908-0 6906-0 | 996€00-0 609-0 SP-18 G-90€ SLF9-1 OO = FE-C18 0-618 0g 8£0-1 To 6018-0 OLIS-0 | $FEZ00-0 9619-0 1F-€8 9-96P IF19-1 68-0 + 6¢-GL¢ G-GLG 0G ge0-1 Diese. ecls-0 9¢1¢-0 | 0¢9100-0 669-0 EL-F8 0-909 C88¢-1 61-9 + 6G-068 F-068 OL = 0 GOGE-0 GO0GE-0 (1100-0) | 0g6¢9-0 —e = FP9G-T 68-0 + $L-193 GE-L96 0 = = = = = =e od =, lic) = C6-€91 C0-F91 OWS ore = aa = id = = =F = 90-0 + 90-001 0-00T 06 — = = a =n = = = = = G30 — | 08-8¢ e¢.8¢ o08 — *sallo[2Q “00 +09 “UU “TU “TUL “sqO- 980 |. noyen'| :poaaasa “anode A ~prnbt sanodv A. rOL X V Peete 2 anode ih eee eee “*8Q07 ID |. z peyeinyeg payemnyeg prmory “moT} Vv porepno[ep | “pearoesq¢ -operSrjueg yo Ayrsneq -anodeA puv pmbry jo uvayy sezuiode A | oaano WOIZ pray any [eotjo100q J, eee jo -vroduoy, 0} enjoy qra yy Bd OROTS BT “Aqisuaq “WRI B FO OTUNTOA ‘ornssarg -ino0de A “ANVLINHdOST 419 Youne— Vapour-Pressures, &c., of Thirty Pure Substances. ; (ont) IFL-8 = I1¥Z-0 = (11#Z-0) | (11¥-0) 0 (SF1-F) (SF1-F) 0g - 08g 0988% | Ge-1z% 106-2 )h ate C1FZ-0 80FZ-0 6FST-0 | 1962-0 = O1F-g 00LE-8 = $408 — 6.9% FE9. i= BBFG-0 SZFG-0 6F9T-0 | 8618-0 1G-F1 | ¢90-9 C9BT-8 g = G85 88903 a4 CEF-S T+ IS¥Z-0 O8FG-0 SS8FI-0 | ZL8e-0 = 61-9 0996.2 = 99616 = 82 Wizara 0 SFFS-0 SPFS-0 IZgt-0 | ¢9¢e-0 FL-GG 1-1 0608. eg + Z901G 6001 0% 010-2 0 09FZ-0 0943-0 €9I1T-0 | 8¢Lg-0 02-9 09-8 0199-4 Lg + C0661 SFS6I 912 068-1 Ti 98FG-0 L8FG-0 Z9860-0 | 8868-0 60-18 1-01 910-4 0g + SSZ8T 80ZST 01Z 969-1 0 6663-0 6BG6-0 1810-0 | FLZF-0 CI-Le LL-@ C688. 6g + CSLSOT $6961 00% 81¢-1 t= 1193-0 6196-0 19g90-0 | 80S¥-0 L8-1F GL -G1 C8IZ-% opt 61ST GLPET 061 F9P-1 T+ F19Z-0 819%-0 9160-0 | G0L¥-0 98-cF 11-61 QSZ1-% il 4 98ST GgcTL OST 80F-T 0 1992-0 199-0 G6ZF0-0 | S88F-0 98-67 08-83 OLF0-% GB — 1816 9086 OLT FPES-1 0 1026-0 1012-0 1ZGg0-0 | 6F0¢-0 01-2¢ GS-86 C086-1 (= FEB 9928 09T 166-1 e+ SFLZ-0 0FLZ-0 6Z8z0-0 | 1619-0 OT-1g G-F8 OFZ6-T g = 6189 $889 0ST GPG-1 at 1812-0 e8Z-0 $9860-0 | 889-0 CF-8G 8 GF SFL8-1 0% — 1696 LILS OFT F0G-T 6 4 0882-0 8282-0 LZ610-0 | #9FS-0 0-19 6-19 0088-1 p= SLOF LLOF 081 LOT-1 ot F183-0 GL8Z-0 ¢ceto-0 | 68¢¢-0 19-89 -49 1682-1 GBP G-86LE 0618 ya Sel-T I+ 1162-0 916-0 8210-0 | 8024-0 88-69 1-08 0ZEL-1 OT = 0-EF08 FFE OLT SIT-1 0 0962-0 0962-0 | 1Z6600-0 | 1z8¢-0 FI-89 8-001 OSTL-T Sot = 1-603 I1¥Z 00T 880-1 0 008-0 F008-0 | G8LL00-0 | 1TS6¢-0 OF-0), G-831 0989-1 6-7 + 6-088T 918 06 990-1 6= SF08-0 0608-0 | 0900-0 | 6809-0 F9-GL 0-991 669-1 8-0 + 8-FFPL PPPL 08 890-1 a= 6608-0 6608-0 | 619F00-0 | +F19-0 $P-PL .91 9109-1 NGG = €-0601 $601 OL 0F0-1 e= 9818-0 6g1e-0 | 6LF800-0 | &FZ9-0 03-92, $18 6109-1 (= 8-908 108 09 080-1 6= 0818-0 ZS1g-0 | 1L¢z00-0 | g8ge9-0 06-L1 0-688 LLLG-1 10 - 1-F8¢ 8-786 0g = Ls FOGE-0 CBE-0 (8100-0) | %gt9-0 = = LPSS-T 0-1 + 9-Z1F 9-11F OF = i= 898-0 6976-0 (100-0) | ¢z¢9-0 — = 9289-1 Boll = 8-886 1-98 0g a = G1ee-0 E1ee-0 (6000-0) | 4199-0 = = S11S-T | 90-1 — ¥-681 ¢¢-061 02 = 0 1gg¢-0 1gee-0 (9000-0) | 2019-0 = = 606F-1 | 02-1 — | §%-er “GF-8E1 OT = a+ 1068-0 6688-0 (4000-0) | 8619-0 — = LILF-1 | 01-0 + | G0-92 6-9) 0 — = = = = = = = = 02-0 + | OF-¢F 02-4 01 — “SoLo[e “0°0 "0°0 “Ulu paeederg bacoueeg 8407 °I%)). n "padres anode A. *pmnbt ‘mode, | +01 * V EE ee BD vanod@A | by peyemaeg | 7 ae EIOs OO eae paving payemyeg pmbry : sun Vv poyelnoVy | “peadrasqg -opexyqu0g arses srnode\ puv pinbrT jo uray = ¥ i A *SOAIND WLOIF PVs 9iny ' -e1edmay, 0} jenqoy yea, jo o1jery ~Ayisue(y “WBIN) B JO OULNTCA ammssarg-imode A, "TACLOUCOSTIC Scientific Proceedings, Royal Dublin Society. 420 ((eo1411p) 018-8 — 9982-0 = (998¢-0) | (9982-0) 0 (133-¥) (186-4) 03 - OF9ST 09981 8-913 I19-@ Cie 8186-0 0882.0 GLET-0 | 1818-0 ¥G-FT 98-9 OLET-¢ = 90081 = FL 918% qs 9682-0 103.0 TGEI-0 | Z8FS.0 9-61 1G.L OIL8-G b+ 6Z69T G069T 016 826: T iL OFFZ-0 1#FZ.0 | 66960-0 3168-0 £6-16 1g-01° 0966-2 6 + 099F1 IG9FI 09% GeL-T 0 P8F6-0 #8F3.0 | 08910-0 661F-0 $E-88 0-81 CO8E-Z i FS9GT 8991 092 $09-T 0 8266-0 8282.0 | $3690-0 PEPF-0 18-18 10-91 GSGS-G Ci — $8801 G060I OFZ 809-1 6 oP £163-0 1493.0 | 6040-0 S89F-0 88-0F €9-61 G8CI-Z fe a 6186 C186 08% 9GF-T I+ 9196-0 G197.0 | ZO0GEO-0 OI8F-0 08-8F 8-82 G610-Z p t+ IF6L Le6L 023 L9g-T it 0992-0 6992.0 | F8FE0-0 0L6F-0 88-9F 1-8 0210-3 | Ge — | ¢.6zL9 SeL9 OIG I1g-1 it ap 8013-0 GOLZ-0 | ¥1820-0 LI1G.0 €8-SF 8-F8 GFS6-T ge = G.999¢ 019¢ 006 696-1 T+ LPLZ-0 9FLZ-0 | 9830-0 6969.0 ST-1¢ L-GF 0£06-T 0% — | O0-SEL% OFLE 061 Feo it 4 0622-0 6817-0 | 69610-0 $88S-0 c0-8¢ c0-1¢ LLGS-T 8-2 — | 6.0868 8868 O8T 606-1 1 oe 8883-0 GE8Z-0 | &19T10-0 $06¢-0 OL-Fg 0-29 GLIS-T Acie = || oie SPE OLI 981-1 I+ LL8Z-0 9182-0 | IGE10-0 0299-0 §6-9¢ L-SL G6LL+T FFL — | 9-089 Gr9G 09T LOT-1 T+ 0%62-0 6163-0 | OL010-0 BEL9-0 06-L¢ G.86 LPPL-T G3 + | G-81IZ 91IZ 0ST GFL-1 w= 6966-0 2962-0 | 0F¢800-0 1#8¢-0 81-66 T-L11 GEIL-T 1-0 — | 6-891 6891 OFT 611-1 i= 6008-0 9008-0 | 062900-0 G¥69-0 08-29 8-SFT 1G89-T G:C eECeGGell 6181 Ost 960-1 6= 1¥08-0 670-0 | 61GG00-0 9705-0 10-#9 9-161 0FS9-T 9-4 + | 9-201 OZOT 0Z1 FL0-1 i= 0608-0 1608-0 | 066800-0 €F19-0 8-99 9-09 08Z9-T BO sp || Gol2i T-LLL O11 8g0-T Te GETE-0 egis.0 | 410800-0 9829-0 1-89 G-188 9609-T | 98-4 + | GT-18¢ 8-816 ne FPO-1 0 ¢L1g-0 CLIE.0 | 686Z00-0 8689-0 $0-0L L-9FF $08S-T Oot = || Cxixg? 9-967 06 a 0 11Z8-0 L1B8-0 (9100-0) | LT#9-0 = = $8oo-T Aa) = 6-F08 G.c0g 08 = 1+ 6928-0 8EZE.0 (Z100-0) | +0¢9-0 = = 9189-1 0 8-81 €-81G OL = I+ 1088-0 0088-0 (g000-0) | %6¢9-0 = = OLTE-T | GT-0 + | G8.GPT aiat 09 = I+ ShEs-0 GPSE-0 (¢000-0) | 1299-0 — _ LLGF-T | #Z-0 + | 61-96 6-66 0¢ = 0 #888-0 P8SE-0 (F000°0) | 929-0 _ _— F8LF-1 | ZI-0 — | 89-19 8-19 OF = il 4 9ZFE-0 CZPE-0 (Z000-0) | 8Fg9-0 _— _ ZO9F-T | 1g-0 — | 61-88 ¢-8¢ 0g = 0 89Fe-0 89F8-0 (Z000-0) | 269-0 = _ GGFF-T | 80-0 — | LL-8% 60-86 0% == t= 6068-0 Orgg-0 (1000-0) | 6102-0 _— — ShGF-T 0 00-81 0-81 Or = i= 0gg8-0 1¢¢8.0 ae TeOTL:0 - _— O80F-T = IT-L = 00 “SOLLo[ey) “00 “070 “Tur berger mategueg $407 IR |. nopR9 | “paaaas smode A Rane ‘mode, | 01x V Po}E[NoTRH | “paartasqg wiOdeK se payeanyeg prnbry aera : i sm 3 "SOAJ If pve aan jeonax00q J, odvA puev pinbry yo uray Jo SOAIND WOLF PROT gaainer 0} [enjoy eo z Jo oney ~Aytsuacy “meLD B FO eTINTO A somnssarg-inode A. TTALNAOST IA 421 Youne— Vapour-Pressures, &¢., of Thirty Pure Substanees. “UU 82Z0€ = 08% 03 pezeL100 + 0f6-61G IVa (wap) 901-8 — CELZ-0 = (ce2zz-0) | (¢82z-0) 0 (999-8) (999-8) 9% — GSZ08 09208 08 988-4 01- 6813-0 6416-0 COIZ-0 £688-0 8L-11 CL L¥6-% = $1866 = 613 866-3 0 SF13-0 ShL3-0 CG81-0 GF98-0 8-91 68- 9FL-% 99 — OF16G 9066 LLG £98-% I - 1912-0 Z9LZ-0 F891-0 1688-0 =< 1-9 019-3 = $9086 = PLZ 806-Z I- 8113-0 6LLZ-0 E8F1-0 GZIF-0 GL+9B 86-9 FOES er + 0899 19996 OLZ 166-1 tH 6612-0 1082-0 8P71-0 FGEF-0 = 10:8 166-2 = 8206 = 69% €18-1 (i 0282-0 ZZSZ-0 ITIT-0 88SF-0 91-68 00-6 090-4 t+ 19F8G 6EFSS 092 | GO2-1 0 £982-0 €98Z-0 | $¢060-0 0Z8F-0 60-1F 0-11 CFL0-G 9¢ + 69606 660% 096 | §8¢-1 1+ 1062-0 9062-0 | 9610-0 £906-0 91.9% FE-81 0GL6-1 le 196LT SL6LT OFS | 86%-1 I+ 6763-0 8h6Z-0 | 09%90-0 1L2¢-0 86-6 00-91 0168-1 1g + GE9ST 1091 08% | 007-1 I + 6666-0 1666-0 | 6FZ¢0-0 9949-0 99-8 60-61 LZE8-1 ct + 9ZESI ILGg1 023 | 98-1 I + 9808-0 ceoe-0 | LerF0-0 9696-0 09-98 9-36 GLLL-T 0 IG9TI IS9TI 01Z | PSO e+ 0808-0 4108-0 | g&Le0-0 08L¢-0 1-69 GL-9 GOEL-1 Ge + 0866 L166 002 | OL%-1 t+ FZIS-0 OZIS-0 | OFTE0-0 926-0 81-39 ¢8-1E ¢189-1 G.oL + G.46F8 Z8FS 061 | G6G-1 g + S9Ts-0 c91g-0 | ¢29Z0-0 1909-0 G1-99 1-88 G8F9-1 6-¢ + 6-661 POLL 081 L61-T at G1E-0 O1ZE-0 | €61Z0-0, | 009-0 96-19 9-9F 0819-1 1-61 — 6-109 1#09 OLT 991-1 ie + GCZE8-0 FGE-0 | STIST0-0 6689-0 11-69 0-¢¢ OT89-T 1-08 — €-100¢ 3Z0¢ 091 | QFI-T I+ 0088-0 6668-0 | e0¢t0-0 SFF9-0 FL TL GF-99 6086-1 Gt + aga SEF 0ST SIL-T 0 PFSE-0 FPSS-0 | LG10-0 ¢9¢9.0 7G-FL, G-18 $86G-1 0 0-988 G98 OFT 860-1 I - 6888-0 0688-0 | 0¢6600-0 0899-0 T¥-92 G-001 OLGF-T 1-9 + 1-601 S0LZ O8T LLO-T i= FEFE-0 G&FS-0 | 9600-0 1619-0 82-8) 9-CZ1 9CLF-T #8 + F-GG1Z ZEIG 031 690-1 I O8Fs-0 18Fs-0 | 108900-0 8689-0 IL-18 L-8S1 96FF-T OF + 0-169 L891 O11 9F0-T t= ezee.0 9Z¢8-0 | 886F00-0 €00L-0 10-88 C-Z0G 6LGF-T $< + 8-908T FOST 001 F20-1 g- 69¢8-0 GLGE-0 | 1T8E00-0 90TL-0 66-8 F:Z9G SL0F-T #8-0 + | 8-366 766 06 920-1 @= G198-0 1198-0 | 868Z00-0 901-0 GL-98 1:Gt 8188-1 68-1 — | 16-68 8-1PL 08 = gs 1998-0 #996-0 (6z00-0) | 9082-0 = = 8898-1 6-0 — | 1¢-0F¢ 8-0FS 02 —- re 1018-0 8018-0 (9100-0) | 00FL-0 = ae FIGE-1 L0-T + | 10-98 0-488 09 = ES 8918-0 #o1¢-0 (1100-0) | 96FL-0 = = 1¥ge-I | G1-0 — | ¢0-69¢ G-696 0¢ = 0 6618-0 6618-0 (1000-0) | T6¢Z-0 S = PLIST 78-0 + | 6F-G8T 9-181 OF = 0 GESE-0 CF8e-0 (000-0) | ¢s92-0 = = Z108-1 6I-L — } 91-081 Gg. 161 08 = 0 1688-0 1688-0 (g000-0) | 6222-0 a = GG8c-1 | 78-0 — | 84-92 6-92 0Z == 0 se6g-0 8868-0 (6000-0) | #282-0 = = 0018-1 70-0 + | 60-LF G0-LF OL = Tt ar G6se-0 #868-0 | (¥1000-0) | ¢2962-0 = = Tese-t 08-0) | esc GG-LG 00 | mS) 8.) 179) ‘070 baa} moegees paeogees "mu | “sqQ-"oTe9 |. 3 “node A. “ smode, | 701 x V EARS MED) | EEE SEERNO) «mode, payeinyeg aa -8qQ—"2189) 3 pro Fe) ee, 7 [PRIMO "PAHO | spesvan is 1 s "SOA. ; E ean [eo14aI00U J, mode, pue pmbry jo uvayy 70 OAINI WOIF PVIIT Beda 0} [enoy - JeoTy cereal ~Aqisua TURIN) @ JO OUINTOA samnssotgq-inode A “UNHTABLOANV XU 3M SCIENT. PROC. R.D.S., VOL. XII., NO. XXXT. Y- Society an i yal Dubl. O01 ngs, R 7) fic Proceed i Seient 42 *eUaZTOG PI[OS » : ({e019119) G).8 = cF0e-0 = (¢#08-0) | (¢¥08-0) 0 (83-8) (#8z-8) FI + |. Gorge c6ege | ¢.88z a a Goa Daa ss 986-0 = = 869.3 ip = 6L19€ 9ZC9E | 0-886 = = ie se = SLOF-0 = = 1G¢F-G 1 = CPECE 9EFCS 1-986 = =e a aaa ea S1GF-0 a = GELe.s cG — 69SFE FOGEE | €.F86 LEG: i= 9808-0 1806-0 O099T-0 FICF-0 EF-LG €0-9 FG1C-% Ot = GLLEE G8LEE 083 FE6-1 0 CSTé-0 Céitg-0 L8Z1-0 F86F-0 61-98 fifiol £900-Z (8 4p C8886 GE88G OLZ GPL-L it ar PSIS-0 81-0 8E0T-0 8689-0 G8-8F F9-6 OLL8-1 Linch 9LESE 66896 096 019-1 0 GEGE-0 GE6E-0 FE¢80-0 609¢-0 LV-6h 69-11 LGSL-T GE + FIGGG C81GG 0G6 Og-T is 18G&-0 868-0 881L0-0 T¢8¢-0 IL-F¢ GO-FL 1602-1 PTL ar 69661 GfE61 OFS 66P-T 0 I&6é-0 1ééé-0 11690-0 ¢909-0 1g-8¢ €L-91 LS¥9-1 OS CT89L C89L 08% 198-1 G+ O8E8-0 81E¢-0 S10¢0-0 GeZ9-0 FG-69 6-61 986¢-1 (op 9G9FL IGGFL 026 106-1 i? 30 OSFE-0 9GFE-0 LOGF0-0 GEPI-0 L8-¢9 LL-&@ 9FGG-T 66 + SFL ESPCL O1Z LGG-T t= 6LFS-0 OStE-0 9F2€0-0 ¢099-0 18-89 06-86 6E1¢-1 (SIl SP €990T OS90T 006 LIZ ib oP 6698-0 8698-0 1L6Z0-0 8¢19-0 9L-TL 9-86 86LF-1T UP + F-65706 F06 061 ISt-1 G+ 61¢¢-0 LLGs-0 180-0 9069-0 69-41 G-OF OSFF-T 68 + G-GC9L LI9L OST O9T-T (8 4p 6698-0 969€-0 1800-0 €F0L-0 89-92 6-LT S6LF-1 GSTS I-FLE9 9889 OLT 96T-T 0 6196-0 6198-0 FEL10-0 C8Il-0 F6-SL L-LE SI6E-T [Eg3]b = 6-18Z¢ 00g¢e O9T IGT-T Cia 6GLE-0 LGLE-O LEPLO-0 OTEL-0 GL-08 9-69 0898-1 (60) = 8-PSEP CeeTr OST SOL-T it. SP O8LE-0 6148-0 9LT10-0 OFFL-O G8-G8 0-¢8 OFFE-T 0 0-068 O0GSE OFT 680-1 Lim 1€8€-0 GE8E-0 | 199600-0 89¢2-0 FL-F8 L-F01 FIGE-T 6g + 6-F68Z 1Z8G O€T LLO-T b> GS88E-0 F88E-0 G191L00-0 6691-0 8g.98 €-0&T 0008-1 1.8 + 1-880 0866 ya 80-1 Ses GE6E-0 Gé66-0 | Gb0900-0 6082-0 I1-68 ¢.¢91 9084-1 6-6 + GSFLT 6EL1 OIT 0-1 ee €866-0 1868-0 F0LF00-0 LZ6L-0 T¥-16 9-Z1G C19G-1 B43} Ir 6: PEL cest 001 10-1 Go PEOF-0 6E0F-0 | O19800-0 TF08-0 19-66 0-LLE 98FS-1 he ar T-9101 8001 06 €60-T 0 980F-0 980F-0 | G&L600-0 GFI8-0 CF-C6 0-99€ 812Z 88-L = G9-SEL 0-G¢) 08 GG0:1 (9 or LETF-0 PELP-0 0F0Z00-0 8hZ8-0 01-96 €-06F PGIG-1 91-0 — OF-LES 9L-Sh¢ OL — 9 5p 681F-0 98TF-0 (2100-0) LGEs-0 = aoe 9961-1 10-0 + 8¢-88¢ 1¢-886 09 = i OFGF-0 6EGF-0 pene 99F8-0 = = ZISI-I | 49-0 + | 16-89% 08-89% 0¢ — 0 Z6GP-0 Z6GE-0 (s000-0) 9168-0 = = T99T-T 88:0 + 80-18T 0Z-08T OF = I- FFEF-0 CEEF-0 (9000-0) 6898-0 — — PIGT-1 61-0 + FG-SIL GF-LIT 0g = i= 966F-0 168-0 (F000-0) 0618-0 = = LLET-1 €¢-0 + 99-FL SI-L 0G == 0 6EPE-0 6PFF-0 (Z000-0) 688-0 = = GPGI-1 43-0 + S¥-CF 61-¢F OL = 0 10¢F-0 10¢F-0 G1000-0 90006-0 bere rae FOLII-1 0 $9.96 #PS-9G 0 — = => = = = = ra om ilo = €8-FL ¥16-F1 o0L — “S918 OTRO) "9°0 o9)°a) “UL UL “THU “TUL $40 IF) |-poyernozeg | -poaaos eodeN “prnbr crisity || dou 7 [EO NO “mode j pajamas | "80-9180 |. : poqwanqug Reeetiad “pmbry “u0ry = po}znoreg | ‘peartesqg, erences jo Aqisue(q ‘mode A pur pmbry yo uve ; ~ezt10de A TIN “AQoomrp any yeorer00q as te ue Jo UHORESIQEEREL || PAE -vrodwoy, 0} [enjoy peo JO O1jery -Aqsuoq “meIN & JO OTUNIOA -oanssarg-inode A ‘HNAZNAD 423 ‘ly Pure Substances. dc., of Thi (OMG SSULES, 2 Youne— Vapour-Pr¢ (Teng) | 961-8 = 1¥S8-0 = (1¥¢8-0) | (1Fg8-0) 0 FE8-G PZ8-G 0F- GLESE G16gE | G¢-98o | = = = = = = = = =o 86 - aga sogee | 2-68 = = = = = = = = = 19 — socae 69¢zE | 8-88 SFE. v= 6168.0 8868.0 F£0G-0 | $&1¢.0 8-0 26-F G8¥6-1 1p OLITE LIGIE Ox || 816-1 0 1898.0 1898.0 GeS1-0 | 681¢.0 6.66 1G-9 POPL-T 8z + 16813 898s OLE TLL-T T+ 9698.0 6698.0 9221-0 | $919.0 G9.cg GI-8 GGC9-1 ret F668 09686 096 689-1 C= FSL8.0 9618.9 8001-0 | ¥#069.0 66.0F 6-6 GLEG.T se+ FF60G 9060 09 Se¢.T o= €188.0 C18s.0 €0FS80-0 | 6819.0 LO.FF 06-11 6GLF-T spt 60Z8T I9IST OFZ 6GF-1 I+ 1188.0 0188.0 18010-0 | 9801.0 Se-LF 1G-F1 81GF-1 9¢ + O9LST FOLCT 08% 068-1 ot 0868.0 8Z68.0 106¢0-0 | 9ZL.0 18.06 16-91 COLE-T 8% + eLGel OSGst 0% 18-1 T+ 6868-0 8868.0 896F0-0 | O8FL-0 $0.8¢ 81-06 6986-1 8 + GZ9OTT LIOIT 01% 662-1 e+ SF0F-0 GhOF.0 F81¥0-0 | TL9L.0 GE.GG 6-86 Ge08-1 I+ —| 6-686 6686 003 LGB. y+ LOTF-0 S01F.0 96FE0-0 | Le8L.0 8-16 9-86 LGLE-1 ¢-01— | ¢.-¢9sg 9188 061 166-1 Cai 991-0 FOLP-0 1160-0 | 4808-0 11-09 GE-FE GPEGT 6-ZI +] 6-810 9002 OST 061-1 e+ CZEF-0 GOGF.0 €1FZ0-0 | 08-0 18-39 iaa 0613-1 6-0 —| T-0F8¢ IPS OLT FOT-T e+ F8GF-0 18GF-0 Z6610-0 | $988.0 1g-#9 0G LO6L-T 1-71 — | 6-FI8F 6G8F 091 6ST-T (3 +p SPSF-0 IPSF.0 6Z910-0 | 61¢8.0 8-99 ¥19 SeLI-T 9-0 +] 9.6268 6268 0ST 811-1 p+ GOFF-0 S6SF-0 91810-0 | ¢998.0 ¢)-89 0-92 OFST-T 9:0 —| #F-ILIe GLIE OFT 060-1 e+ I9FF-0 SGPP-0 €9010-0 | T18s.0 0-12 0:66 6FSL-L list = || feeyaseE 08S OST 610-1 il sr 0Z&F-0 6ISF-0 | LFE800-0 | 4469.0 80-82 8-611 99LT-T #0 —| 9.8861 6861 0Z1 GC0-1 i= O8SF-0 I8¢F-0 | ZEG900-0 | 9606.0 G0-¢), 1-891 $660-T 6-0 —| T-IPST GPST OIL 860-1 g= 6897-0 GPIF-0 | G80G00-0 | ¢8Ez6.0 OT-L2 9-861 180-1 Goi =|) TeQ2iti LLIT 00T 10-1 a= 669F-0 ZOLF-0 | 9F8800-0 | 9986.0 69-8), 0:09 9190-1 98-4 — | #1-088 0-888 06 820-1 v= 8G1F-0 Z9LF-0 | $88Z00-0 | 96F6-0 10-08 L.9F8 TE20-1 69-1 +] 6F-9F9 6-FF9 08 = a= SISF-0 IZS¥-0 (1300-0) | %%96-0 al = 8680-1 08-1 + | 08-49% G.89F OL = g= LL8¥-0 OSSF-0 (C100-0) | +16-0 = = £920-1 FE-1 + | FF-968 1-66E 09 = g= LE6t-0« | OF6F-0 (0100-0) | 0186-0 = = GEL0-1 #F-0 — | 9F-8%G 6-866 0g = IL ar 166F-0 966-0 (1000-0) | 9866-0 == = F100-1 18:0 — | SL-8F1 9-6F1 OF aaa IE 4 9¢0¢-0 ¢e0c.0 (c000-0) | F010-T = = 1686-0 09:0 — | 10-96 19-96 0g = ot 9ITS-0 FI1G-0 (€000-0) | ¢8z0-T = = 0816-0 29-0 — | 66-6¢ #4-09 0G = a+ 9116-0 FL1C.0 (Z000-0) | L¥g0-T = = 6996-0 Ge-0 —| ¢0-9¢ 0F-98 Or = e+ 9869-0 €&6¢-0 | (Z1000-0) | ge9ro.T — = FOGCG.0 80:0 — | $8.06 66-06 0 = = = = = = 60-0 — | $G-11 G.I (He = = = | tc = = = 10-0 — | 80-9 C1-9 6 = | ic) (49) “O°0 “oO paoogaog “mu “TUL = BIOS PleO) | ; Enanaer sanodvy | , cancttigy, || oly jf EEO ENO “modu, poveaniyg | PNT S(O) (40) peeing paywanjyg | “PMOYT poy r o “PEFTMTO | “paAxoeqO | -onmcrque a Jo Ajisueq 4 : | -ezitodeA | “saAimo “Ay}00aTp = [eor}ax0ay{, mode, puv pmnbry jo uray Jo moO prox pray fine 0 — —— : — yeoyy 4 Fo o1jey | “Aqisua(y | “UBL eB yo 9tun[o A ‘ounssarg-aimode |. “ANAZNYAOWON TA 3 mu 2 Scientific Proceedings, Royal Dublin Society. 424 (qe01}1t9) 9LL-8 a Po96-0 = P96-0 F698-0 0 LEL-G LEL-G 11 9p LEGES 90688 | 6-698 = = = = =a OOFF-0 = == SLE: Ce = S6LEE FOSS | 8-898 = roe = — a 0&99-0 wer a 808-1 681 + GPGOE S010 8-8h§ = = = = = I1F9-0 = =e 9g¢-1 OLT + COLFS GE6EG | G-6ZE 688-1 = es =F = = 66-41 ae = SSL = 08Z 668-1 esr GOTF-0 €91F-0 1660-0 PESL-0 99-66 ZE-06 FILET GB = 6FLIT S8IIT OLG 086-1 ar SIGP-0 LIG¥-0 GL1F0-0 9108-0 G1-Gq 16-86 PLEG-1 Go)E = G-9896 L¥96 096 886-1 {bere FLGP-0 GLOF-0 OF¢E80-0 9618-0 0¢-L¢ 66-86 1066-1 8-16 + 8-168 OLG8 066 LOG-T G+ OSéh-0 8Z8F-0 00060-0 9588-0 6F-6G &-86 8961-1 €-%§ + | €:0802 8F0L OFZ OST-T Lee oger-0 9SEF-0 PESCO-0 8198-0 06-19 6-68 OPLI-1 FFG + | F-8109 686¢ 0&6 LGL-T 6 = LPPF-0 SPPF-0 GPIG0-0 GL98-0 68-69 9-9F ISGT-1 G-9T +] G-GL0G 9¢0¢ 066 OFL-T = LOFF-0 S6PP-0 80810-0 G188-0 61-#9 §-G¢ PPEL-L G-8l + G-LEGF 666P 016 OGI-T 0 &SSP-0 €¢GF-0 9010-0 G¢68-0 §8-99 $99 LOLL-T Wok, ar T8696 16gé 00G GOI-T 0 S09F-0 809F-0 LPZ10-0 1606-0 LE-L9 6-08 6660-1 irs Cuesta ¥-G06G G06G 061 ¢80-T Le G99F-0 €99F-0 &G010-0 ¥G66-0 96-89 GL-16 1780-1 0 0-0L83 OLED OST GLO-T = SILT-0 6ILF-0 | GLE800-0 $CE6-0 80-04 F-611 0690-T Cio = G- E161 O0Z61 OLT 990-1 Dee SLLP-0 PLLY-0 | $81900-0 08F6-0 96-1L b-LPL SFCO-T 8-9 — || L-L0ST Pest Oot 8¢0-T bar 8G8F-0 LESF+0 | GEFSO0-0 6696-0 LE-GL T-P81 81F0-1T tl = 8:-FOGL 9061 OST GG0-T 0 &88F-0 €S8F-0 | 9TEh00-0 6616-0 98-EL 1-186 6860-1 63) ¢-166 F686 OFT 6G0-1 Gt LE6P-0 Gg6P-0 | 8800-0 9686-0 FG-FL ¢-G66 9910-1 ¢0-0 + | 00-612 96-81 081 ie 0 &66F-0 &66F-0 (9200-0) 0966-0 aa a OF00-T CT-0 — | $9-6F¢ 08-GFS a fia 6 = 1¥0¢-0 6F0G-0 (6100-0) 6100-1 = =e 6666-0 SIl-0 — | OF-Z0F GS-GOP OLT = g= 101-0 01-0 | (F100-0) | 8610-1 = = 1186-0 | £0-0 —| GL-666 | $1.66 00r a Gg 9¢1¢-0 8ST¢-0 Eee 0) G060-T 7 == $016-0 1-0 + | 0¢-806 C§-80G 06 = ois 01GG-0 6166-0 (s000-0) 61F0-1 ae ara 8666-0 96-0 +] I1-G#I GL-PFL 08 aa 6 = 6966-0 9965-0 (9000-0) 9660-1 ror, aoe 0066-0 9¢-0 + | 9F-86 06-16 OL re 6 = SIég-0 0G8¢-0 (4000-0) 9690-1 = <= GOP6-0 89-0 — | 96-49 FS-G9 09 = 0 GLEG-0 GLEG-0 (g000-0) GFLO-T aa ae 6066-0 G¥0) = | OS 50F 86-1F 0¢ = G+ 9GFG-0 FGPE-0 (3000-0) 9F80-1 2 sas 0666-0 6¢-0 — | TL-96 00-96 OF S 6+ OSFS-0 SLES-0 (1000-0) $960-1 = =, 6Z16-0 1-0 — | G&-GT GP. GT 08 ce if ar ee 0 GEG¢-0 (1000-0) G90I-T se == 0F06-0 0 91-8 91-8 06 =r il SF 18¢¢-0 98¢¢-0 a LLIL-1 i a 6G68-0 10-0 — | 6L-% 98°F Or = Ear 0F9-0 6£9¢-0 (Z0000-0) | 9821-1 68°68 ie $9988-0 60-0 —| 6F-G GS-6 00) *SOLI0[V) *0°0 *9'0 “ULUL “UU WUE 810~ 9180). payamnayeg | :peatesqa anode *pinbry “mode A 101 X V “mode, pazenyzes ‘sqQ-"2[tQ |. ; payBinyeg payeinqeg por UO Vv EID | EO ‘opRisuad jo Aqisuod -mode pur pmbry Jo uva -eziodeA | “sadmno *A]goertp earn [woneroay | mori g WN yo Wolf pray} pray omiedne 0} jenjoy yeoyy L Jo onva -ginssaig-inode A “Aj}sua(] “WUBI B JO OWIN[OA “ANAZNAAOUOTHO 425 Youne—Vapour-Pressures, &c., of Thirty Pure Substances. “UU Z16E¢ sv ainsseld [voto Suryey, + ‘poututajap ApQo011p jou yng ‘pazemnyss y. (TvonHO) 4608-8 SoSF-0 = (egst-0) | (s¢sF-0) 0 (190-2) (190-2) (12 -) cesses | x(t6ge) | (268) 696-1 = == = = = = 1-13 = = 8-808 = 082 PGG-1 or + 09L¢-0 0¢2¢-0 | 910F0-0 6601-1 PS-FF 6-46 0106-0 8.93 — G-8L0L COLL OLZ 61-1 b+ 0g8¢-0 928¢-0 | LzFg0-0 OTST-T GL-GP 6-66 ZFSS8-0 RL = €-8909 6109 09 GLT-T ios 006¢-0 1066-0 | 16¢0-0 OIST-T 08-9F GFE 8898-0 4-01 + L-0L1¢ o9T¢ 096 9¢T-1 de 0169-0 6966-0 | GSFZ0-0 689-1 96-LF $-0F eggs.0 1-91 + L-9L87 098F OFS OSt-T 6 = 0%09-0 GF09-0 | 64060-0 9L8T-T PP-6F 1-8 0ZF8-0 8.61 + €-819¢ 6o98 08% eIT-1 p+ 0119-0 9019-0 | SFL10-0 1806-1 9F-0¢ €-1¢ g0gs-0 Gr + ¢.190¢ scog 083 e0T-1 I + 6119-0 SL19-0 | SGF10-0 O1GG-T 18-19 9-89 0618-0 0-01 + 0-189 LEGS 01Z €60-1 a= 6739-0 €929-0 | 60Z10-0 C8es-1 G6-GG 1-28 $108.0 yr + ¥-6L06 £106 003 ¢go-1 6 + 6189-0 119-0 | 0¢6600-0 FEGS-1 $0-8¢ ¢-001 $161.0 G8 4b 8.1891 F891 061 80-1 - 1889-0 6889-0 | 411800-0 1696-1 08-8¢ G-SE1 9181.0 hig 4p ¢.GG8T 1¢81 OST @10-1 it oe 1GF9-0 9¢79-0 | 9900-0 LP8G-1 SF-FG P-ZCL F8LL.0 30 = G-9L01 LLOL OLT 690-1 6 - 4h 6Zg9-0 8G¢9-0 | 1F%400-0 ¥666-1 1G-9¢ 8-061 9691-0 | 8[-6 — | &8-eF8 0-958 091 690-1 i ar 6669-0 ¥699-0 | ¢Z1F00-0 OPIS-1 60-9¢ FGFS LO9L-0 | G9-€ + | 02-0¢9 0-679 0ST = t -4 ¢999-0 7999-0 (1800-0) | ¢66¢-T = = €0GL-0 | 19-1 + | LF-L6F 8-C6F OFT = ii st GE19-0 1819-0 (FZ00-0) | 6&Fe-T = = 1FFL-0 | GS-0 + | OT-e2e 69-618 a == i -= 0089-0 1089-0 (6100-0) | ¢s¢g-T = = 6981-0 | 40-0 + | 96-412 6-FLS all == gs g989-0 1189-0 (F100-0) | 6318-1 = = #8ZL-0 | 60-0 + | 61-861 1-861 O11 = 0 1869-0 1669-0 (0100-0) | 988-1 = = 81GL0 | 98-0 — | F2-0F1 L-TPL 00r = g2= F002-0 1001-0 (1000-0) | 900%-1 r= = OFIL-0 | 8-0 — | OF-16 61-16 06 = t= GL0L-0 €101-0 (€000-0) | GPLF-T = = TLOL-0 | 27-0 — | ¢)-¢9 66-99 08 = I + 6811-0 SE1L-0 (000-0) | GLaF-I == = LOOL-0 | 28-0 = | 81-8% GG-eF OL = I + 10GL-0 | °9082-0 (2000-0) | TIFF-T = = 6269-0 | 80-0 — | €¢-1% 19-1 09 = 0 FLEL-0 PLGL-0 (1000-0) | 9FGF-1 = = ¢189-0 | 60-0 + | 66-91 96-91 0¢ = T -4 reh-0 182-0 (1000-0) | Z89F-T = = 1189-0 | 31-0 + | TT-01 66-6 OF = 0 804-0 80FL-0 | 0000-0 CL8F-T = = 0919-0 | 1-0 + | 62-¢ 19.¢ 0g = a + 9151-0 FLFL-0 = SP6F-1 = = 0699-0 a = a 0G _ 0 GPEL-0 GFSL-0 = $80¢-1 = = 0899-0 _ = = OL = 0 6092-0 6094-0 = B81ZG-T = = 11L¢9-0 = = = 00 “S9TLLOTBO 0°) 0° “tau “mu “wu *S¢0- 8D "paqeypoyv. "paArasg “anode A “ptnbt “node A. yOL XV Eee aca |e 1D “mode A, or pazeinyeg yal eal eee : poyengeg poyenqeg una On EOS NGO) PPA) ee SEO “apVIstzUaL yo Ajyisuaqy oan@ikan\ joie wrnnA To TET -vz0deA | “soaano “AToaatp W ain i [Royeroeyy| PX YEA (a) W jo WOIf pvary pear ~eiaduiay, oF Tenpoy qeoy poner sl *AqIsua(y “Tv v JO BUINIOA *aanssartg-inodv A “ANUZNHAONOUT vety. Soe in , Royal Dubl mgs tifie Proceed clen 426 yrems Area aq pos 02% Aojeq pur ye semsserd payypnoyvo oy} UT UoKyeIEy[R ey} esn¥oeq “LTNUMOF s JOU SBA JT ‘aSLe] YYAOTOS SI QPP JV odnsserd poyeMTyse PUL poyE[NI[vo 94} UaeAjeq eoTELey Ip OY Y “TMM ZLEGE Sv eansserd [woryTIO Suryey, + QOL OF S]ULJSUOD oY} oJLTNITeIer 0} Arwssodeu JYySNOTy *paaput Snompy ‘peurmejep Ayoorrp jou ynq “payeuarysy » (eon) L08L-8 = = aE (F18¢-0) | (F18¢-0) 0 (0@L-1) (002-1) 0-118 — gegee| x(c16ee)| »(8FF) S9T-T i = i = = a , 9-68 i == 8-68P7 a 0086 SPL-T 6+ GGGL-0 9162-0 F09G0-0 GLIP-T 66-07 ¥:8& 9901-0 8-68 — G-G8LE C18é OLE OGI-1 ilar €0E1-0 Z0EL-0 00GG0-0 PSEP-L OF-1P CPSP 6969-0 feof = 8-008 8168 096 101-1 Taz GSEL-0 €8¢L-0 EE810-0 I8¢F-1 SP-OP 0-4¢ 8989-0 F-01 — 9-6896 OOLG 0&6 880-1 0 09FL-0 09FL-0 SgST0-0 FOLFE-1T LG-8F 6:79 €1L9-0 eels = 8-GPZS 9G OFG FLO-T (gan 8891-0 ¢g¢gL.0 6610-0 Th6F-1 CL-bP G-LL €699-0 6-0 + 6-FS81 FSst 0&6 190-1 g+ 9191-0 IT9L.0 9L010-0 CTIG-T SL-PF 6-66 9199-0 c.0 + G-06GT OGST 066 $90-1 @Q= €692-0 GOLL-0 | 688800-0 9TS¢-T GE -SP G-3Il 6699-0 Fy Gia a ¢-FEGL aera OIG 190-1 0 TLLL-O TLLL-0 | 846200-0 OLFE-1 08-F PLE F9P9-0 GO-L + G0-66 0-166 006 090-1 gt Sh8L-0 782-0 | §06900-0 L69¢-1 €6-9F ¥-691 6689-0 19-5 — 66-882 0-662 061 690-1 ite 9Z6L-0 6-0 | $¢2F00-0 €08¢-1 69-9F §-11G 8669-0 86-0 + 66-819 L-819 O8T = ie €008-0 $008.0 (1800-0 GL6E-1 = =S 1969-0 60:0 — G9-6LF L-6LP OLT ier 0 1808-0 1808-0 6200-0) FE19-1 = — 8619-0 08-0 — 0¢-998 §-L9E O9T = 0 8¢18-0 8418-0 ZG00-0 $669-1 ae oe LE19-0 06:0 — 08-916 L-91G O¢T =a 0 6868-0 ¢&z8-0 (1100-0 79-1 eas — 8409-0 91-0 — FL-F06 6-406 OFL —- ‘Cine G1E8-0 G1E8-0 (2100-0 1199-1 = es 8109-0 GI-0 + GF-S8FL 6-8PL Ost aia 0 8868-0 8868-0 6000-0 L9L9-1 os = $969-0 eF-0 + €8-COL F-COr 0ZI = Ciena 99%8-0 6978-0 1000-0. GE69-1 a ag 9069-0 10-0 + 68-81 88-62 OIL = 0 GPE8-0 GPSS-0 (g000-0) 6L0L-1 ar a cgsc.0 02-0 + SF-0¢ &@-0S oor = g) = 6198-0 6698-0 (000-0 LYGL*1 ae ay 8619-0 10-0 + LG-€8 08-66 06 ao Feo 9698-0 1698-0 €000-0 T6EL-1 RAS = 0¢L¢-0 0-0 — 9L-1G 8L-1Z 08 = ($ ar TLL8-0 8918-0 (Z000-0) Cech. = = 6015-0 40-0 + 69-E1 G9-ET OL a B= 888-0 1¢8s-0 (1000-0 GOLL-T = =e 6F9¢-0 G0-0 + c6-8 08-8 09 i 0 668-0 FG68-0 1000-0) SF8L-1 a ox €099-0 10-0 + 66-7 G8-h 0S fae 9+ 0006-0 4668-0 a 6861-1 ae aoe 6e¢¢-0 0.0 + 61-6 FL-Z OF re i or 9106-0 ¢106-0 (z0000-0) 6F18-1 a au O1gG-0 0-0 + 6G-1 8P-1 0€ = g= 6S16-0 FS16-0 ay 8068-1 — = G9FS-0 has a a 0G aa i= 1666-0 F&66-0 = LOFS-1 a = CIPS -0 =a am = OT a 0 6086-0 6086-0 ei 6¢2098-T Kae aE 9FLES-0 a =a = 0 “saLtopen ‘00 “0:0 cireaued “Ur “UU | FOO) eee ae || ; “mode A é crusty 0X V payepnaep | “paasesq¢ poyengug pmbry pees payeangng eae “pmbry 10) : {O- 1%) |-pozefnopeg | “paaresqQ pega jo Aytsueqy 2 [eRe | -ezmodvA | ‘saamo | *A]o01rp v RA) anode, pur pmbry fo uvoy, | z E amy [eotjetoy gy, ; jo wor proy| pray | -vieduiay, 0} jen}woy a qvoyy Jo o1ey 2 “Aqisuaqy “wey B JO BWINIOA saanssorg-inode A ‘UNHZNHAOGOI 427 y Pure Substances. rt z w-Pressures, Se , of Th Youna— Vapor (qeoryt49) 089-€ 916¢.0 os (919¢-0) | (9L9¢-0) 0 (FE6L-T) | (862-1) 19 - FIFE I8IFE | G1-€8% eh x sk Se fants oe — — — 09 — SLOFE sete | G0-88z SoF-G 6 = 8094.0 6198-0 16G8-0 FE9L-0 Sv-01 81:3 O1g-T 0 CZSZE GZ8ZE 08% 620-6 iat 1696-0 8899-0 OILZ-0 9998-0 96:4 69-¢ PST-L I + C6686 1668 016 g18-1 0 8119-0 SLLG-0 9F1Z-0 60F6-0 10-02% 99-F 8290-1 7 BEEGS 8ZSEs 09% 199:T Tes 998¢-0 198¢-0 FGLT-0 0866-0 61-86 OL-¢ 1200-1 go 6076S FLFGS 09 EGG-T Teese 6466.0 #969-0 FOFT-0 FEFO-T 96.95 88-9 6L&6-0 Ne 9661 88961 OF GLF-T 0 FF09-0 +F09-0 ZEE1-0 1980-1 F9-L1G ZI-8 1126-0 or + 990L1 9G02L1 083 COF-T Tig ge19-0 PS19-0 0F01-0 LGGT-T CF-66 79-6 1068-0 6 = S6LF1 GOLFT 023 6FS-T 0 GZz9-0 ZG69-0 | 18280-0 9961-1 60-18 8-11 9498-0 0 6GLZ1 6SLZI O1G 108-1 ? = 1169-0 GTg9-0 | STFL0-0 8881-1 19-28 8-81 ZIF8-0 fh te SP601 98601 00% 696-1 iss 10F9-0 8079-0 | 0G%y0-0 G61G-T 0-7 00-9T G0Z8-0 LST + | 1-9%86 1186 061 GOG-T 9 = 1649-0 1659-0 | 6FZE0-0 OLFZ-T 0F-gg 60-61 6108-0 GP — | 8-F68L 668), O8T S8T-1 g = 6869-0 6869-0 198%0-0 FELS-T 88-98 6-2 Se81-0 1-6 — | 6-1899 1#99 OLT O9L-T 0 8199-0 $199-0 | 0¢980-0 G86-1 66-18 ¥-16 8011-0 ¥-71 — | 9-FoGG 6gGg 09T FPL ey + $919.0 0949-0 | 0F080-0 C1GE-1 68-88 6-68 19¢1-0 9-7 + | 9-6¢¢F cacy 0ST FELL @ 4p ¢¢s9-0 0¢89-0 00¢60-0 O¢FE-T 66-6§ 0-0F GetL-0 1-0 + T-GZL¢ GoLe OFT 901-1 @ 4b 1¥69-0 G¥69-0 | 180&0-0 0898-1 86-0F 1-6 OTSL-0 TL + | 1-6008 G00E OsT 680-T 9 + 6801-0 Se0L-0 | 689T0-0 G06E-1 80-3 0-19 611-0 8-6 + | 8-00%@ 1686 0GI 610-1 y+ IST. @ LZ1L-0 | LO&T0-0 FOIF-T- 60-8F C.9) 0801-0 %-6 + | ¥-688T 08ST OIL 190-T 0 SCGL- SZ6L-0 | LZ010-0 SPer-1 CL-PF ¥-16 6169-0 g-L + | 8-F9FT LOFT 00T 6F0-T to. Gh LISL-0 | $¢6200-0 FOGF-1 91-Gh 19 1189-0 0-6 + | O-LITT IIT 06 170-1 g= OTFL-0 SIFL-0 | €50900-0 COLP-1 00-9F ¥-F91 gL19-0 = G8-9E8 08 980-1 i = £0G2-0 ¥0GL-0 | 0L¢400-0 696F-T LL-9F 8-81 899-0 0 8-819 8-819 OL a ‘Ci L6GL-0 66¢2-0 (¢00-0) GOTG-T = ae F6¢9-0 19-T + | L9-0F% 0-68F 09 = 6 = 1691-0 8691-0 (4200-0) | T9891 == a 01g9-0 | 16-0 — | 60-808 0-608 0¢ = G2 egLL-0 LSLL-0 (1100-0) | L¢¢¢-1 = = 8ZF9-0 | 61-0 — | 14-016 6-016 OF = i = 6181-0 0881-0 (G100-0) | SPLG-1 = = 0989-0 | 40-0 + | $9-681 9-681 08 = T 3 FL6L-0 8164-0 (g000-0) | 686¢-T = = #129-0 | 90-0 + | 19-68 6G.68 0Z = 0 6908-0 6908-0 (¢000-0) | G&T9-1 = = 6619-0 | 1-0 — | 1¢-¢¢ G9-GG 01 = T + G918-0 F9I8-0 (2000. 0) | 3¢ GE9-1 = = 2219-0 0 80-8 80-88 0 — a = = = = 80-0 + | 68-ST 18-81 () es = — — — = = = = = 98-0 + | 60.01 66-6 Ae = “SOLLO[R{) ba) 0'0 “wut “mu “TUL FAIO= OU) *paye[na[eg | “padresqg soe “pmbry smodeA | 701% V “mode, ean PeFBUnyES _ |-sqo="0179 payemyeg payeanjug | PMONT ae Vv TERRA WES OAD ‘opvasiquey ee “mode, pur pinbry yo uvayy me BA *SOAINO TOIT pvary aan} : -vied may, 0} [enjoy - eo Jo o1jeyy Ajisueq. “MUL @ JO oUINTOA ganssorgq-inode A “ACTUOTHOVULEL NOPUVO ciety. 0) S nm yal Dubli ) Scientific Proceedings, Rc eX - Aqisua (qT “WIRIN) B FO DUINIOA ‘aanssargq-anode A (Teortt9) OFL-€ = 6IFL-0 = (61%L-0) | (61¥L-0) 0 (S¥8-1) (8F§-T) Wor 0808% 6L08Z | 1-818 So 7 z= == am a) cars — aa GIot+ F6GGG 61096 Org a = Sr saa i= = ar — = = GSEGG =a 00€ =a cme sul — ram Ss = re a me GPL6I = 066 | Lb9-T (Sc 1681-0 F68L-0 0916-0 8696-1 09-¢T 69-F 8EéL-0 ‘AO = IGE41 FGPLI 086 6-1 C= €§L08-0 €108-0 G18I-0 61GF-1 OT-L1' GGG SE0L-0 0G = O6TET O86ST OLG SOF-L ar CE1s-6 FEls-0 0ZE1-0 LYLF-1 0¢-8T 8¢-9 1819-0 Che = GPGET 066ET 096 [68-1 8+ 8968-0 0668-0 O8ZT-0 1GGE-1 LL-61 18-1 0L¢9-0 Se = SSFIT FISIL 096 T¥é-1 9+ I8é8-0 ¢1E8-0 801-0 199G-1T 98-06 66:6 6889-0 0 C166 C166 OFG F66-1 G+ FOE8-0) GO0¢8-0 67 160-0 0609-T 16-16 66-0T 169-0 (AB a0 60¢8 GLYES 0&6 996-1 Cia 1698-0 0698-0 8610-0 88F9-T 68-66 F6-61 ¢909-0 61 + LGGL SEcL 063 [CGT j= 0218-0 6218-0 1990-0 9989-1 19-83% Gé-S1 6664-0 iB ar L¥I9 9119 O1Z C6L-1 Hie FL88-0 €888-0 6¢F90-0 FCGL-T LG-4G GE-81 908¢.0 (XG ar S9Tg CPIG 00% 99T-T = 8668-0 006-0 cS¢r0-0 9¢¢)-T 8E-cS 60-66 9696-0 Go + 60EF ESP 061 SéI-1 lies GG16-0 €616-0 8180-0 €L84-1 66-96 GL-9G C6G¢.0 @- 4p 19¢¢ eccs OST 6IL-T ar 9PS6-0 CFG6-0 11080-0 GS8I8:1 00-23 C-6E 00¢¢-0 Qe ea F16G 0G6G OLT SOL-T Hr 0LE6-0 9966-6 1260-0 I8f8-1 69-16 8-6¢ IL¥¢-0 § = 6286 8986 O9T 680-1 ar C6F6-0 88F6-0 §€0G0-0 GLL8-1 CE-8G G-6F LGEG-0 Op, = 0-S88T C68T O¢T GLO-T (or 0696-0 1196-0 8T910-0 €L06-1 G1-6G 8-19 SFGS-0 (8) = 0-16FL L6FI OFT 1¢0-1 4p ChL6-0 GFL6-0 FLEIO-0 1966-1 68-62 G-8) 9919-0 HG = 0-G91T 21 O&T 9F0-T {bar 0186-0 6986-0 166600-0 6696-1 F¢-0¢ 8-001 6604-0 Oa7 > F-168 F-468 OGI 160-1 IS 6666-0 9666-0 0T9100-0 9166-T LI-1¢ P-TéL 1606-0 ig = L-GL9 €-GL19 OIL 660-1 lies TG10-T GG10-1 1¥1¢00-0 9810-6 9L-1¢ O-FLT Pl6F-0 GG + ¢.86P 0-96F O00T == iar L¥GO-T 9FG0-T (Z¥00-0) OSF0-G <= = O6SF-0 9T + 6-96 9-098 06 = 6 > GLEO-T FLEO-1 (100-0) LTL0-6 a = LGS¥-0 8-0 + G-LGE 1.966 08 = ats 66F0-T 86F0-1 (6Z00-0) 6160-6 = a S9LF-0 i) = 8-SLT 6-841 OL = G+ GG90-T §690-T (G100-0) LEGIL-G = = OLLF-0 Uo <= T-1G1 G-G61 09 = e+ 1GL0-1 810-1 (0100-0) LSPI-Z = = PS9F-0 |31-0 — 89-62 08-62 0¢ = it 6180-1 8280-1 (2000-0) 6ELI-S = ae 86¢F-0 |90-0 + 88-0¢ 8-08 OF = Gt GOOT-1 sOOT-T (F000-0) G00G-% = = Chch-O |}Z1-0 + GP LE 06-18 08 = 0 GSIT-1 GSIT-L (Z000-0) 69GG-G = = G6FF-0 | P10 + GL-ST 8¢-8T 0G = Chae 6261-1 G9GI-T (1000-0) 8606-6 = = OFFF-0 |68-0 + GL-O1 §s-0L Or — k= LS81-T POET-T (60000-0) | ¢1822-2 =z = 88EF-0 |¢8-0 + 88-¢ 6G-¢ 0 = = = es a om Fae i; a COSCO 80-6 8L-G o0L = *SaLlO[VO SOR CAPA) Teraree “UU “TUL : : | Q 3 | §40— 91% |-payzmozeg | -paaras anode A\ ‘pmbt | -mode, | 301 XV pozENo[ey | “Ps 410 EN paqeanqeg | PMOFT eee | poqwingeg rece | *prnbry “u01} ME T 9 -poqemmopeg “paatasq Q) cepramag JOUNEEREC. smodv, pur pmbry jo uvoyy CARON I cesycaaey WoOIF PvE any eenerce uu eS FO | -viadwag, 0} yenjoy wor, JO O1FBY “ACTYOTHO OINNVILS o> | = ae R | + | at) S s | (jeo14119) 80-1 8 + GLFS-0 LOFs-0 | 6L0¢00-0 | ¥#9L9-0 64-08 6-961 ESL¥-1 G0) +] B.ORei 9LEI 0¢ % 390-1 g + 69FS-0 9978-0 | T&L¢00-0 | 689-0 88-28 0-89 GOGF-T 0 0-126 0-126 0F & 660-1 I + FGGE-0 €698-0 | 119600-0 | 610)-0 81-68 9-818 LEGP-T — 99-179 | — 0g QQ FPO-1 0 LLGE-0 11¢8-0 | OL8100-0 | ¢erZ-0 #G-18 L-F8G 910F-T | 0-6 + | 00-cFF 8-68F 0% & 20-1 I + 1298-0 0g98-0 | F9Z100-0| gFa2-0 G1-06 T6L 1618-1 | 08-0 + | 09-162 8-066 OL SS 820-1 0 6898-0 ¢g9e-0 | 148000-0 | 981-0 69-36 6061 | 8848-1 | 11-0 + | 10-48T 6-F8T 0 = = = = = as = = = = 01-0 —| 06-IT §-31T OL - Sy “SOLO[R) "0 hp) “UU “wu aries NS 3 Z 8907 120 |-payemoreg | -poaaos “anode “pmb i ‘mode, | 501 XV pazenorey | “Pp 10 AER payeangeg | PONT edger Zz payeangeg — = payninqeg *prnbry “u0T} Vv “pojze|noyeD | “pedsesq GC, ‘opvaSiqzuer 5 eae “mode, pur pinbry fo uray es “SOAINO WOIZ pay any : > 0} [enjoy vey puxedure jo o1ey ; “AjIsua(y “WUBIN) e jo auny[o\ ‘aanssatg-inodr A ‘GHHLA TARDY SCIENT. PROC. R.D.S., VOL. XII., NO. XXXI. ty. . in Socwe yal Dubl Ko. edings, fie I’ rocee Seienti 430 (Teont19) GE6-2 — 68FE-0 — (68-0) | (688-0) 0 (998-2) (998.2) 86 — GE6FP OS 0SF 0-F1G LPG-S SI €6F8-0 118-0 C986-0 LOTF-0 ¥S-0O1 6F-E CCOF-S gi? = GCOFP OLOFF G-€1G 940-§ SH 96F8-0 POCE-0 1893-0 SGEF-0 LE-E1 EL. COLES G6 — OTEht OLFFF §1G LE8-6 teak FOGE-0 0068-0 ISZ-0 6PSF-0 €9-L1 80-7 861-3 WG = FOLEF SOLer GIG 86-6 p= 8148.0 BEGE-0 S81Z-0 | L8F-0 86-26 LG-P 0660-4 = B0GGF O1SGF 01Z 808-6 r= L¥SS.0 16¢8.0 G98T-0 1hZS-0 81-0 LEG 0806-1 es + S9LOF $8006 906 F£0- [ieee 0698.0 1696-0 FEC1-0 8¢9¢.0 08-88 9¢-9 GLOL-T pert 6&89¢E G399¢ 006 98L-T Lasts F99E-0 6996-0 SLIT-0 8P19-0 98-8F 67:8 6969-1 981+ 9SL1E OgcTs 061 Geo-1 i? oP 9ELE.0 GELE-0 FEF60-0 1669-0 8P-9¢ 09-01 96e¢-1 SoIt+ S9ITLZ OF0LZ Ost GGS-T OaNor 6085-0 FOSE-0 ¥E910-0 F¥89-0 96-69 O11 TL9F-1 oL + 6808Z ST0EG OLT 6SF-1 i 4p I88&-0 OS8s-0 1€690-0 9ETL-0 96-89 G0-9T SLOF-1 ie = 6LF6I 0061 O9T ¢98-T Crem €S68-0 CC68-0 £90¢0-0 60FL-0 8G-EL GL-61 80¢¢-1 LG SOS9T OSE9T OST 808-1 0 CZOF-0, CZOF-0 FGLFO-0 8€91-0 L6-LL GC-FG 6608-1 Sh > LEGET OLCET OFT 096-1 0 L60F-0, L60F-0 PFEEO-0 0981-0 80-38 6-66 €CL6-1 Fl + 6IIII COLIT Ost FIST Lig se: S9LF-0 691F-0 $8960-0 0208-0 66-98 GLE 6686-1 6-08 + 6-9F06 9106 OGL I8t-T iets LPGF-0 OFGF-0 09160-0 F9G8-0 1¥-68 €-9F 0013-1 0 0-8L62 SLGL OIT GST 0 GLEP-0 GLEF-0 €L10-0 GOFS-0 08-66 ¢0-8¢ T&81-1 8-6 + 8-E81¢ PLLG 00T PCLT 0 PSEP-0 FSEP-0 0¢E10-0 FE98-0 LL-96 C0-FL G8EL-T 6-01 + G-PECH PGF 06 660:T oor CCOFT-0 ESFP-0 6F010-0 £088-0 18-86 §-96 O9ET-T Uk) = §-10¢¢ 80¢¢ 08 LLO-T Gita 96CF-0 FGCP-0 G&0800-0 8968-0 18-TOL ¢-FEL IST1-T P-FL — 9-8096 EL9G OL CC0-T 0 LOGF-0 L6SF-0 6€0900-0 §€16-0 IL-€0L 9-C9T 660-1 0-01 — 0-TS861 1661 09 G60-1 (Gn L99F-0 699F-0 | 9CFF00-0 ¥626-0 CF-SOL F426 0910-1 6-9 = 1-SFF1 GFL 0S £60-T 6 = SELP-O OFLF-0 9€6600-0 L¥¥G-0 c@- 111 0-608 ¢8¢0-T (3) Sr §-6601 6601 OF TLO-1 1 608F-0 OLSF-0 162200-0 866-0 L1G-FIL G-98P 61F0-1T 8o-¢ + 8F-ETL 6-L0L 08 = 0 OSSF-0 OSSF-0 (9100-0) CPL6-0 ae = 690-1 L0-€ + LY-6LF F-9LF 0G = 0 0S6F-0 OS6F-0 (1100-0) | 6886-0 = ao GILO-T | 09-1 + | 00-118 4-608 ol ae Beate 0206-0 610¢-.0 (1000-0) | 6Tg00-T = = 8966-0 66-1 — IL-861 0-C6T 0 = = ie: Ss cen a a ae eae, GS-6 — SI-S1T G9-LT1 OL == coe! = = ms == = aS = 18-6 — 68-9 L-19 06 — “SOLlO[VD) *“0°0 “oO “uuL “wuUr “WU *sqQ-"9[29 |. are ‘anode A. eat Ode SOLXV paye[noyeg | “paarasqg Lae payeinqeg PILOT OID payeimyeg eats a *pmbrT “uot Vv DREAD) | OG) | oa borey yo Ayisuagy . LARS -ezmodeA |, pviol}ues) : g amodeA puv pmubry yo uray 3 SOA.IND WOIF pvary arny [eatje1090q J, jo -riadwe y, 01 jenjpy — —- wo = —- FO O1VY - “ApIsuaqy “MvIQ @ JO OUIN[O A -alssatg-inode A “HLVNGOX TABLA 431 y Pure Substances. rth 3 of Th AB rsp - Pressures, & Youne— Vapour (18919119) 68-8 = GEZE-0 = (zgze-0) | (eae-0) 0 (760-8) (¥60-8) 66 — 9gFEE cegce | 3-¢8% 188-3 ¢+ 0FZ8-0 CEZE-0 €982-0 | LITF-0 16-81 6-4 066F-% = SISte = #83 8ZL-% 9 + 9F68-0 0FZE-0 8612-0 ISGF-0 89-91 GG-F 0988-3 = OFETE = tg ISt-Z r+ 998-0 Z9ZE-0 0681-0 | E9t-0 61-66 66:9 CLOTS pos 1968 39668 08% 991-2 6 = 6668-0 1088-0 L8S1-0 | 10-0 1S-62 08-9 CF66-1 = CFLOE = a 100-2 o> GES8-0 FEEE-0 6181-0 | 0626-0 L¥-FE GGL 6068-1 gir + 04986 LESS 026 F9L-T 62 1688-0 6688-0 €L01-0 | FLG-0 | 88-3 GS-6 OLFL-+1 ect + FOL¥EG 109%% OI 19-1 = S9FE-0 FOFE-0 16980-0 | 9909-0 8-67 09-11 F8F9-1 SZI + ILGIG SFIIG 002s L0G-1 0 828-0 8ZG8-0 ST0L0-0 | 889-0 GL-FS GZ: FL 9ELG-1 GII + F9IST GG08T 061 PGP-1 le £6¢8-0 G6S8-0 1¥1¢0-0 | 0199-0 FF-66 F-L1 6C1-1 Ly + GOFST goss 081 98-1 0 8998-0 8698.0 68L40-0 8F89-0 Z1-89 1-14 SL9OF-1 9 + 69601 $9661 OLI 908-1 0 SZLE-0 SOLE-0 91880-0 | 8402-0 16-29 8-96 691F-1 Ai = Ogsol L¥SO1 091 896-1 0 L8LE-0 1818-0 $9180-0 LOZL-0 £9-0L 9-18 6LLE-T G01 +{| &-F968 $968 0g1 PICT 0 BG88:0 E88-0 $996 SPPL-0 Sl-FL 0-68 LZPE-1 9-11 —]| ¥-SP8L 09e2 OFI IST-1 = 9168-0 8168-0 €10Z0-0 | 8292-0 OG-LL GC-8F OLIg-1 6-¢ +] 6-696¢ ¥G6¢ 0g SPI-L 0 1868-0 1868-0 LG910-0 | 96LL-0 9-08 8-09 LGSG-T 8:0 —| L-LLLF SLLF OZI 031-1 a+ CEOF-0 EF0F-0 ZIET0-0 | ¢¢62-0 91-68 G-9L TL93-1 0-T —| O-I8Z¢ ZSLE OIL 660-1 it ap 601F-0 SO1F-0 ZE010-0 | ZIT8-0 #L:G8 6-96 LEST 1:0 —| 6-646 096 001 610-1 I+ BLIF-0 ILIF-0 | ¥66100-0 | 9&8-0 GE-88 1-61 FOIS-T 10 +) 4.6966 6966 06 090-1 lets 98ZF-0 CeZF-0 | $60900-0 | 60%8-0 16-06 0-F91 ZOST-1 Ib +] T-O1LT 90LT 08 $F0-1 Tee 008F-0 6624-0 | OLEF00-0 | ZEgs-0 GS-86 8-816 e691-1 = G-99E1 = 02 1g0-1 6 tb e98F-0 19gF-0 | 0L8800-0 6898-0 28-56 1-966 60G1-1 6-€ —| 9-L16 ¢-126 09 GZ0-1 0 9GFF-0 9ZFF-0 | 88FG00-0 | 1288-0 6-16 G-01F 6281-1 0 +) #-6F9 F-679 0¢ a 0 06FF-0 06FF-0 (1100-0) | 968-0 = = LOIT-T | 68-0 +] GG-LtF L-9F% OF = t= COCF-0 SSCF-0 (Z100-0) | 4606-0 = == 9660-1 | 00-4 + | 0-666 08-162 0g — I- 919F-0 L19¥-0 (8000-0) | 9%6-0 = = 6880-1 | OFT +) &6-86T cG-Z61 0G = 6.4r SL9F-0 9L9F-0 (c000-0)| 986-0 = a 00L0-T | GL-0 + | OI-TZT GE-0G1 01 = r= IFLE-0 CHLE-0 (000. 0) | ZOSt6-0 = = SFc0-1 | 1-0 +] L¢-2L PECL 0 = = = = 2 = = = = ¥0-0 +) #9-1F 09-17 01 - ait = Se = — = = — — 11-0 +] 09-26 6F-GS 0% — On “SOTLO[B/) bah) 0°O “wu “mu paregetg “sqQ-9189 |. ? “mn0ode A 4 carat, | ag 27 payzpnayeg | *paasesq¢, vere payenqeg pmbry Lane payengeg ‘A | spmbrT ‘u0n “payyfnoreg | “pearosqg | yo Ajrsuacy @ Db eS -vztaode j M nH me) it = "1m0ae me PIndT 0 uvd o “soArINO TOL Nets) amy [eono109q ‘A pay prnord # W gO ee -v10d wd J, oy yenqoy eo Jo o1yeyy “Aqisua( “UB. e Jo emny[oO A ‘omssarg-node A “ALVANYOR TAHLY 3.N Scientific Proceedings, Royal Dublin Society. 4532 “mode A poyeinyeg jo Ayisuaqy [eonje100qy, 0} [enjoy a OAWEL 5) || ++t+t+ 141 BAH OOCOHMONSCAHANNSCHAMNMNMONANR as sts Ost + + } ee "80-9180 70OL X V (reonup) CG7E-0 = (Zeze-0) | (ceee-0) 0 (¢10-8) (410-8) BR = OsTgs G1sGe | 21-88% 9628-0 09Z8-0 629z-0 | 9668-0 OL-IL 96-8 080¢.z = 6284 — 4 $968.0 1968-0 8823-0 | 9&GF-0 ¢8-G1 18-4 0998. 0 FESHE FREES 28S LLZS-0 SLZE-0 8206-0 | L&&F-0 66-06 86-7 0606-2 a+ ogee seses 08% 8628-0 1628-0 9LLI-0 | SI8F-0 68-96 £9-¢ GGL0-2 = LE6ig — ha LbEE-0 SF88-0 9IF1-0 | T8Z¢-0 18-48 90-2 9868-1 66 + F088z G0L8z 0% G1F8-0 91FS-0 1601-0 | TPL¢-0 99-8F 11-6 SIPL-1 gs + LELEZ 6F9FG 01z F8PS-0 €8Fé-0 | 8¢980-0 | 0079-0 96-09 6¢-T1 8689-1 19 + 9F1IZ esos 00% 96-0 gcgs-0 | €6690-0 | O1F9-0 F1-9¢ 08-FI 109¢-T 6T + O96LT IFGLT 061 1298-0 0696-0 | %89¢0-0 | TL99-0 00-19 09-L1 166F-1 Fe + ZOTST SIIST OST 6898-0 es9g-0 | 86¢t0-0 | 2069-0 61-99 CL-1Z LLP¥-1 t= 16921 B69ZT OLT 9¢18-0 e¢16-0 | Ig2¢0-0 | e8TZ-0 96-69 08-92 OGOF-T 1% = GFOOT 99¢0T 09T FZ8E-0 1Z8¢-0 | 920¢0-0 | 6882-0 89-8), 0-88 929-1 Z-LI— | 8-898 ZOL8 0ST T68¢-0 688E-0 | F2720-0 | Zg¢2-0 88-91 CL-0F 9168-1 T-L1—| 6-280 O0TL OFT 1468-0 9266-0 | OL610-0 | GILL-0 6-61 GL-06 1966-1 G-0l—| &-FIL¢ Cole 0g GZOF-0 GZ0F-0 | OLGTO-0 | 682-0 18-28 1-89 OL9G-1 G0 —| 9-g¢cr 9G¢F 021 0604-0 Z60r-0 | 6EZ10-0 | 0908-0 1-48 1-08 90FG-1 ¥-0 +] F-8¢e PSCE OIL 9eTF-0 6G1F-0 | TL9600-0 | 12z8-0 68-88 F801 €91Z-1 TT +] 1T-6L22 SLLZ 001 GBCF-0 FCCF-0 | OFFLOO-0 | FL1E8-0 ZI-16 ¥-FE1 FGI £9 +] ¢.0z1z 11S 06 18G¥-0 88zF-0 | 819¢00-0 | 6198-0 10-46 0-821 SSLT-1 6-8 +] 6-88cT e8¢1 08 GGSF-0 ZceF-0 | S61F00-0 | %998-0 1F-96 C883 PRST-1 = G-L9I1 — OL LIFF-0 ctFF-0 | 9L0¢00-0 | 0088-0 66-86 1-98. S9ST-1 0g-T +] 08-se8 09-188 09 I8FF-0 ISFF-0 | Z1Zz00-0 | 6g68-0 F£-001 0-29F LSIT-1 £0-0 +] g81-ss¢ G1-8g¢ 0¢ 9FGF-0 CEGF-0 (6100-0) | ¢206-0 _ — 610T-T 00-T + | OF-1OF 0F-00F 0F 609F-0 609-0 (0100-0) | $0%6-0 — _ 0980-1 ¢0-0 +] 08-¢9z 61.99% 0g ELor-0 €LOF-0 (1000-0) | $¢g6-0 _— _ 6010-1 GF-0 +] ZZ-0LT 08-691 0% 9814-0 C8 LF-0 (g000-0) | ¢9#6-0 — — 6960-1 11-0 + | Z0-SOt 8-F0T Or 6614-0 8614-0 (g000-0) | Zg6¢6-0 — — FGFO-T 80-0 +] gt-29 01-29 0 -- — — — — _— _ 0 G1-¢e GT-ge or - — — -- — — — — 91-0 — | 68-81 0-61 AN = “SOLLO[VD) °0°0 “0°0 masegerg “Ulu “TUL “Tur *paqe[noyey | ‘pearesqa een *pinbryT ; : smodeA | -ymbr : EOD enramarc “paaresqg peyeingeg anetat ot v peyemnoyeD ‘opraSiquag ‘inodvA pue pinbry jo uray | “RZTOUBA | sg nano WOIF pvory aan} | an -eiodway, “Aqtsuaq “TURE B JO OTUNOA sornssorg-mnode A “ALVLEOV TAHLAN 433 ‘ULVNYOd TAdOUd ((eargt19) es 698-8 = (608-0) = (2608-0) | (8608-0) 0 (g83-g (€83-¢) G9 — 88F0E 09F0E | 98-49% g LEP-G aoe GZIE-0 9ZTE-0 SFS8I-0 | F0FF-0 0-1 IF-¢ (0122. t= 6183 0Z69z 092 S 800-2 (qe 1818-0 €818-0 OFEI-0 | &0¢-0 66-18 9F-L 0066-1 66 + GOLFG S694 082 Ss 891-1 0 128-0 TFZ8.0 CHOL-0 | 8&FS.0 CF-68 19-6 0688-1 oe + LL¥IG CCFIG OFZ 2 919-1 at 1088-0 668-0 €0F80-0 | LGL¢.0 S1-9F 6-11 OLEL-1 #9 + 6ZE8T COFST 08% S FIG-T e+ 0988-0 1988.0 16890-0 | 209-0 79-6 G71 0099.1 or + GI6ST OL8ST 02% R FSF 9+ 0ZFE-0 PIFE-0 $69¢0-0 | 6929-0 EF-8¢ oG-LT 1169-1 61 + F6GeT CLGSI 01Z SS 698-1 0 6LF8-0 6LFE-0 LIL¥0-0 | . L8¥9-0 16-9¢ G-1G 91FG.T Gl = CPSTI O9STT 00% Ss O1g-T 6= 8868-0 0FSE.0 1680-0 | 1699-0 8-09 1G CFGF-1 6-9 + 6-016 F816 061 aq 816-1 = 1668-0 8668-0 9280-0 | €289-0 08-29 6-08 OS¢F-1 1-81 — €-8C18 LLIS OSs G8G-1 T+ 1698-0 9698.0 1990-0 | S0L-0 8F-99 G-18 C61F-T 6-61 — T-L219 L6L9 OLT => 861-1 4 GTLE-0 €1Ls-0 6L1Z0-0 | 6062-0 63-89 6-GF IL8E-1 9-93 — P8146 Gogg 091 BS LOT-T 6 4 GLL8-0 SLL8-0 OLL10-0 | 6982-0 6-01 ¢.9¢ OLG8-1 9-31 — P-CbGP Sccr 0ST ws LET-T 0 8E88-0 €88e-0 GZFIO-0 | €2GL-0 FL-E) g-0L G6GE-1 6-1 — 1-699 9198 OFT N LIVI 0 G68E-0 G68-0 OFTTO-0 | 0292-0 96-1 L-L8 1808-1 0-1 = 0-F16z G16 og << O0T-1 i= 0668-0 1g68-0 | €£0600-0 | TIT82-0 FI-8L 1-011 GO8Z-T lt - 6-983 8866 at S 680-1 0 600F-0 600F-0 | L40L00-0 | 1462-0 09-08 6-1FT F8GC-1 ee = L-LOLT OLLT OIL SS 890-1 0 190F-0 190%-0 | @&FS00-0 | 0808-0 99-68 I-F81 9LET-1 jolt = 9-8FS1 SPST 001 = $90-T G= 9ZIF-0 SZIF-0 | LOTFOO-0 | F1Z8-0 16-¥8 G-8h CLIG-T Gil = ¢-Z00T FOOT 06 a LE0-T = FSIF-0 981F-0 | OF0g00-0 | TFE8-0 6F-L8 0-668 6861-1 91 — 6-G8L, G.FS, 08 = $£0-1 = GPEF-0 FPCF-0 | LGG00-0 | . 9948-0 60-68 0-6FF GIST-L 0 6-86¢ 6-86 OL S = 6o= 00EF-0 GOEF-0 (9100-0) | $8¢8-0 = = FPOT-T | LPO + | Lg-e9¢ 6-496 09 Ss = 0 6GEF-0 6GEF-0 (T100-0) | s048-0 = = P8FI-T | Shel — | 16-24% 63 0g % = t= 91FF-0 LIFF-0 (g000-0) | 288-0 = = 6ZEI-1 | 94-0 — | ¥e-2e9T 9-891 OF Se = 0 GLEF-0 CLEF-0 (9000-0) | €#6S-0 = = GSII-T | €0-0 + | &T-FOT T-F01 0g a = har GESF-0 1ESF-0 (g000-0) | 806-0 = = OFOT-T | €1-0 + | ¢0.79 6-89 03 : — e+ 069F-0 L8GF-0 (1000-0) | 216-0 = = G060-1 0 G8-18 G8-LE Or S = e+ LEOF-0 PPOF-0 (1000-0) | 9826-0 = = 89L0-T | 20-0 + | aF-1z OF-1Z 0 s — — — — — — — = = 91-0 + | 9¢.11 OF-TT Ol — Sy *SOLLO[B) Ore “To rederg irate S “sqQ-"0T8Q |, 5 “mode A amine | -modea | sor xv | Pormelsd) ‘pedtesaQ arte poyrmngeg | PMONT een 2 peyeanyeg payeinjeg “pmbry ion v “paye[nayey | ‘poasesq¢ eraLaTaaA =) EO A SLHET -modyA pur pmnbry fo uvayy TRZNOGBA | -sgamno WOLF pray mn} iS) EBERT ae Jo -ereduay, KR Ub 03 [eno yeoH P| “Ayisueq “WeIN B JO oUINTOA ‘oInssotg mmodvA ety. ) Soe Ui yal Dubl t Ro . ngs, fie Proceed Scvent 434 ({e01}119) 66-8 = LLOE-0 = (1108-0) | (4208-0) 0 (0gz-¢) (0¢Z-8) | ¢-991 — I1L8@ 1-0@ = = = = = oS = == = 61 — 1L98% 0S@ ¥66-G ps 7808-0 1608-0 882z-0 | 6888-0 80-21 L8-F G89G-% FOL — 9968 676 LL9-% 6.59 1608-0 6608-0 9661-0 | §61F-0 GI-L1 10-¢ OFS8E-% Foes ILPLZ L¥G LLP gt O11E-0 GOTS-0 ZOST-0 | LOFF-0 61-06 GG.g 0ZLE-Z Ng = £6996 zg C61-G e+ GFIE-0 6E18-0 66FI-0 | S8LLF-0 LI-LZ 19-9 0860-2 Gel ap GZ8FZ OFG F88-1 0 908-0 9038-0 ISTI-0 | 18%¢-0 G0-98 8-8 C868-1 SZI + S681 08% 969-T 0 6968-0 698-0 G0680-0 | S9G-0 89-GF 8-11 GOLL-+1 sg + EFSSl 026 096-1 t+ GEEe-0 8288-0 8Z1L0-0 | FF6G-0 86-85 £0-F1 8289-1 2, oF 689ST O1Z 99F-T 0 c6ge-0 G68E-0 16L¢0-0 | 0129-0 TL-6G CB-LT FOL9-T G9 SE SSZEl 00% 168-1 0 8GF8-0 SoFE-0 1¢L40-0 | IFF9-0 ‘OF-9¢ 0-12 GGEG-1 6G + FSIIT 061 LES-T i= 0z¢e-0 1298-0 €88¢0-0 | 999-0 18-6¢ GL-G 1809-1 IT +] 1-616 OST 980-1 I + g8¢e-0 Z8EE-0 ¢9Ig0-0 | 8F89-0 1-89 9-18 ZO9F-T ¥-61 —]| 9-G6LL OLT LPG-T 0 Gh9E-0 GF9E-0 116Z0-0 | €€02-0 16-69 8-8 61GF-1 9:93 —| ¥F-aP89 091 806-1 = S0L8-0 60LE-0 04020-0 | O1Z2-0 80-69 £-8F 6988-1 G01 —| 4-L¢1¢ 0ST FOL-T I- OLLE-0 TLLS-0 0910-0 | SZ&L-0 FG-GL 9-09 goce-1 PLT —| 9-8F1F OFT Set-1 0 BESE-0 ZESE-0 FIELO-0 | 892-0 69-FL 1-92 GLEE-1 0-1 —| 0-L6¢8 Ost 601-1 I+ F688-0 8686-0 0010-0 | §892-0 €9-LL T-L6 9108-1 €0 —| 2-98¢¢ OZT 180-1 0 9968-0 9968-0 | 900800-0 | 1882-0 66-62 6-FEI OLLG-T #8 = —| T1-8661 OIL IL0-T 0 LIOF-0 L10F-0 | $S1900-0 | ZL6L-0 C1-Z8 ¥-291 FPSE-T GG +! G-6I1ST 001 090-1 6= 6L0F-0 180F-0 | LL9F00-0 | ZII8-0 10-8 8-81 838-1 $b +! #F-PSIT 06 890-1 0 OFIF-0 OFIF-0 | G6FE00-0 | SFZ8-0 81-68 1-983 6C1G:1 | $86 —| 18-088 08 LFO-T 0 10ZF-0 10ZF-0 | 19z00-0 | 9288-8 GPL 6-068 6E61-1 | LLL —| $¢.463 OL =r (Gis 19GF-0 89GF-0 (8100-0) | s0¢8-0 = = SCLI-I | €0-:0 —| 18-917 09 = 6 = BSEF-0 PSEF-0 (Z100-0) | 998-0 = = 6LST-T | 22-0 + | ZP-Z8d 0g me C= S8EF-0 G8er-0 (000-0) | Z9L8-0 = — €lFI-I | 1-0 +] 98-98 i i ae C= EbPP-0 GFFP-0 (¢000-0) | ¢888-0 = = GGCI-T | 81-0 + | §8-8IT 08 om 0 FOSF-0 FOSE-0 (¢000-0) | 006-0 = oa GOIT-IT | 62-0 +] 60-82 03 = 0 F9SF-0 F9CP-0 (Z000-0) | 2216-0 = = 6960-1 | GF-0 +] ZI-SF 01 = (64P FZ9F-0 BC9F-0 (1000-0) | 98%%6-0 = = 8180-1 0 08-43 0 = — — — = — = = = 10:0 + | G0-8T ol - = — = _— = — — = = 40:0 +) 68-9 sis = *SOTLO[BD 0°90 “O° “wut “Tu *SqO-9[%) |. ln sanodz A. Sara smodvA | 701 XV Pee SicbeseatO srnodv A poyeinjeg | PHOT *sqQO- 9180 0 nevennied sent *pmnbry “10 S “poye[noley | “pedtesqQ -apersqueg uOAs yeu yat ‘rnodvA puv pinbry jo uveyy ; pezode *SOAIND WOIT PVIIT any Teoner0e mL ae #0 -eroduay, 0} [enjoVW yea Jo onvy “Aysuaq “WRI a4 jo auny[o A sanssorg-inode A “ALVIGOV TABLE 435 ‘ty Pure Substances. re., of Thi § Youne— Vapour-Pressures, (jeon119) 606-8 = FZ18-0 = (Fa1¢-0) | (F3T¢-0) 0 (102-8) (103-8) 0ST — Z886G 3e00e | #16 886-6 ga ge18-0 SE18-0 7660-0 | 6868-0 01-81 98-7 GIIG-% 681 — 9086 CPP6G 99% GPL pt 6818-0 GeIs-9 SI1Z-0 | IG1F-0 19-S1 GL-P 060F-% egl — 0068 66066 GGG 109-4 9 + IG18-0 GFls-0 0681-0 | 10FF-0 6-61 6-8 0ZLZ-G F0) GOSS G0G8Z 89 F08-G e+ 0118-0 cg Ig. GLOT-0 | Ge9r-0 08-43 16-9 OSF1-Z 0) = 18693 1669 09 910-6 Gap G0Z8-0 1618-0 SIFI-0 | 916F-0 00-08 GO-L 1600-2 Ge = 08092 GI1GZ GG 966-1 ¢ + €868-0 8268-0 981-0 | 0669-0 1F-F8 60-8 8916-1 0 GZESS GEES 0F% S1L-1 9 - 6628-0 1088-0 79960-0 | ¢89¢-0 LL-TP 8-01 GPLL-1 101 + 1010 00008 08% F8G-1 = 8g88.0 0988-0 Z1SL0-0 | 8869-0 FL-LP 08-21 0F89-1 War TSSLT O9TLI 02% 68-1 o= 1Z¥8-0 EZPE-0 0690-0 | 1029-0 19-19 ¢9-GT IL19-1 (i) L89F1 CZ9F1 OIG OLF- 1 I= E8FS-0 F8PS-0 9860-0 | $PF9-0 19-9¢ 01-61 G1G¢.1 19 + axa 08821 002 PGE-1 0 GFGS.0 CHos-0 0z8F0-0 | 1999-0 6-89 G1-8G GZ0G-1 ay + COFOL OZFOL 061 c0g-1 0 9098-0 9098-0 agee0-0 | 9689-0 0-29 1-83 GSGF-1 0 0-1E8L8 LEL8 Ost 193-1 0 8998-0 8998-0 106Z0-0 | ¥0L-0 Z0-99 OF-FE C61F-T pi, —= || Peketeral CPL, OLI 81Z-1 a+ 0818-0 8ZLE-0 9980-0 | 1202-0 GT-89 CPF 6F88-1 6.26 — | &-786S 1G6¢ 091 F8L-1 G+ BGLE-0 0618-0 c0610-0 | 0682-0 8-0 G.39 1898-1 #-01 —| 9-618 O&SF 0g 9ST-1 0 £988.0 gg88-0 6ZS10-0 | 8962-0 OF-SL, ¥-99 OFZE-1 Gol =|) ToiWlee 8888 al SEI-T I+ F168-0 8168-0 FIZI0-0 | GOLL-0 0-94, ¥-28 8166-1 GON cha rea 08 TL08 ost 801-1 0 F168-0 $16E-0 | 699600-0 | Z&8d-0 1G-8L ¢-F01 GElZ-1 0-3 —| 0-70%Z 904% 0Z1 060-1 T+ 9607-0 Ge0F-0 | 9FFL00-0 | 9662-0 9-08 g-PS1 9096-1 7-0 — | 8-981 FOS OIL ¥LO-1 @ 160F-0 1607-0 .| PILG00-0 | LET8-0 61-68 0-GLT 6866-1 3-0 +] %-90FT 9071 001 960-1 i= LGIF-0 ScIF-0 | LOSF00-0 | 228-0 02-8 G-G8G 180-1 $6 +] F-14501 CFOL 06 9F0-1 eo LIGF-0 0ZzF-0 | 661800-0 | 80F8-0 10-18 9-218 PEST-T | Ge-9 — | 99-492 0-TLL 08 8g0-1 (oo SLGF-0 O8cr-0 .| Tgec00-0 | 1898-0 6-88 0-66F SI4I-1 | 90-6 — | F6-9%9 0-87 OL = 6= SeeF-0 OPSF-0 (9100-0) | $998-0 = = IFGT-T 0 08-088 $-088 09 = Ga S6EF-0 LOPP-0 (1100-0) | 0628-0 = ss LLST-T I-T +] 08-1g¢ 1-993 0g = b= SOFP-0 O9FF-0 (1000-0) | @168-0 = = IGGI-T | 82-0 + | 89-691 §-691 OF = T= LIGE-0 SIGF-0 (¢000-0) | Zg06-0 = = GLOT-1 | 60-0 + | 68-L01 8-101 0g = I+ SLSF-0 LLGF-0 (2000-0) | 116-0 = = 8260-1 | 40-0 — | S1-99 3-99. 0% = G+ Le9P-0 Ge9F-0 (2000-0) | 8926-0 = = 0610-1 | 80:0 + | 86-88 G8-88 Or = ot 969F-0 F69F-0 (1000-0) | 1L886-0 = aS $¢90-1 0 06-12 6-16 0 es = = = = = = = = 11-0 + | GL-T1 ¢¢-11 or - = = _ = = = == — = 8z-0 + | 86-5 ¢9.G 006 = “SOLLO[BI) 209 Mi) “TUL a togoeg “UU "sqQ-'2189 |. wet “mode A eae sinodv A ct x va EPH) WOO anode \ payeangeg, | BeuD ea -3q0—9[89 payernyng payeinyeg *pmnbry “m01y a *paze[no[eg | pearesqQ count homey jo Aqisueq inode, puv pmbry fo uray “BZNOdBA | osoAino WOI, PLary aany [eaerooyy eee #0 -vioduray, OF jenpy — yea fo o1yey “Aqisuaqy “MBI B JO BTUNTOA -ammsserg-node A. “ALVNOIdOUd TARLAN Scientific Proceedings, Royal Dublin Society. 436 | (1e14119) 86-8 = 196-0 = (466-0) | (2962-0) 0 (288-8) (288-8) 66- 86676 L@CSS | B-9LE 646-2 QQ = 7966-0 6963-0 691-0 | 6918-0 SL-11 19-F 0869-2 991- F8CFG OSLEG GLB 699-6 Il — 9163-0 1863-0 ZI6I-0 | $90F-0 LT-91 86-S GL9OF-S LGT— $0683 090% €L3 90F-Z o> $662-0 1666-0 1991-0 | S&8¥-0 19-03 0-9 0808-% lity $168 086% OLG 661-G o1- 8108-0 8208-0 CHFT-0 | T19F-0 66-96 26-9 G891-% 07 — CF9IZ GS9IZ 996 116-1 6 = eG0g-0 1608-0 COZI-0 | 806F-0 01-08 08:8 6180-4 = 6F861 6C861 09 891-1 Ca 9118-0 FLIS-0 0660-0 | 68Z¢-0 PP-LE 69-01 9068-1 eg + SZIL1 OG0LT 083 C191 oP GL18-0 GLIE-0 91¢10-9 | 989¢-0 OF-P G81 $06L-T 6G + FOLPL GLOFT OFZ 904-1 I + 9868-0 G&Z8-0 $4190-0 | ¢¢8¢-0 18-9F 7-91 SLOL-T sg + S91 GOSZI 4 61F-T 1 + 9668-0 6668-0 2060-0 | 1809-0 8L-0¢ 6-61 6ZP9-T gg + 81901 0Z90T 0% SPST tS 6g8e-0 9988-0 | SIIF0-0 | 1089-0 66-4 8.46 6989-1 se + 1106 $168 01Z 16-1 0 FLPE-0 FLPS-0 0660-0 | 88F9-0 8T-LG 6.63 FLPS-1 SI + 969), 89), 00 096-1 Tt + ELF8-0 BLES-0 81120-0 | 1999-0 10-09 0-98 666F-T ar + 1829 CLZ9 061 O1Z-T z+ gees-0 1g98-0 | 8920-0 | ¢889-0 08-29 L-FP OL9F-T 1 + 061g 68I¢ O81 OST-T it = 0668-0 1698-0 SFSI0-0 | 4669-0 80-69 L-FG G6GF-1 63 — OFGF 696F OLT SPIT 0 6F98-0 6F98-0 68FI0-0 | 6FIL-0 99-19 1-19 1868-1 B= EPS ihazs 091 921-1 0 8018-0 8028-0 | G6I10-0 | 1662-0 61-69 1-88 | GOLS-T b- SPLS LPLG OST LO-T 0 G91-0 G9L8-0 | L6FG600-0 | GeFrL-o. $8-1L §-G0I OSF8-T ¢ - 991% IL1Z al 980-1 0 8788-0 8z8e-0 | OFFL00-0 | I2¢L-0 O1-FL F-F81 6028-1 4p L891 Egor Ost 690-1 = 1888-0 0sse-0 | 092¢00-0 | 3024-0 e8-9L 9-811 P866-1 T + F661 £681 0Z1 140-1 0 1868-0 1868-0 | G0FF00-0 | O88L-0 80-8), 0-186 TLL-1 G0 — 6-616 0-916 OIL 190-1 0 6668-0 6668-0 | 82gg00-0 | 1964-0 08:6), G.008 1966-1 Let = 6-121 9-831, 00T GFO-1 0 ZS0F-0 BG0F-0 | 29%Z00-0 | 6208-0 69-18 0-0 LLEG-1 9-1 - G-86 8-F6¢ 06 _ t= SOTF-0 60IF-0 (1100-0) | 10%8-0 == == 8616-1 LZ = L-O18 8-ZL8 08 = @ = POTP-0 991F-0 (Z100-0) | 0Zg8-0 a = 0203-1 PLOT 6-466 8-16 OL a I - 1GGF-0 BCGF-0 (6000-0) | ¢gFs-0 — = CSST-T $0 + T-ZL1 8-TLT 09 = 6 = LLGP-0 6LZF-0 (9000-0) | 1¢¢8.0 = os PO9T-T 0 GZI1 G-Z1T 0g = i = BESP-0 secr-0 (7000-0) | $993-0 = = SPST-T | 1-0 — | 49-02 8-0 OF = 0 S8sPr-0 SSEF-0 (000-0) | $228-0 — == S6EI-T | 10-0 + IL-GF L:GF 0g = tT + FEPP-O STFP-0 (Z000-0) | #888-0 = = 9921-1 | 01-0 — | 00-¢% 1-6 0% = I + S6FF-0 LOFF-0 (1000-0) | ¢668-0 = = OZIT-T | 20-0 + | %6-ET 6-81 Or = (6 oP e9¢P-0 1¢cF-0 | (40000-0)| 9016-0 = = 1860-1 | 0-0 — 8e-) FL 0 = = = = ~- = — = = 01-0 + | OL-¢ 09-8 AO = *SOTIOTVO, hi) 0"0 “WU “WU “TUL S10~PIR) *paqvnayRg | *pearasq¢ anodeAt *pmbryT | sanode A. p0OL XV «mode, 7 poyenjusg "sqo—o[RQ |. i ile ream) OE | cae 7 RAMON] PAPO | oonssnney ‘ G 2 <6 "SOAIND WOIT pva' ean jeoniex0a ty, Imodva pee pinbry jo uray 10 Bp SEL | een cont enjoy {wo jo Oryeyy ~Aqisua(y “TUB. B FO ATIN[OA “omssa17-in0de A “ALVLHOV TAdOUd 437 y Pure Substances. e wrt Younc— Vapour- Pressures, §¢., of Th (Tear) 806-8 = 6962-0 = (¢96¢-0) | (9962-0) 0 (eL8-8) (eL8-8) LZI — 06092 LIZGS | 6-BLZ a i zs = = — _ = = 19 — E6946 099% | ¢-1Lz 869-2 ioe 862-0 1860-0 LG61-0 | 810F-0 69-91 I1-¢ O68F-Z ce — 0L0%% CO1FZ OLz 9L¥-Z 9+ 6666-0 6863-0 TL1-0 | L6GF-0 IF-61 TL-¢ 0998. 8 — L8g¢% C1FSG 89% 466-6 + 108-0 0108-0 Z9ST-0 | 6947-0 G1-8% 07-9 SOFC. = F6EGG 00%Zz 69% 260-6 gt ¥F08-0 1408-0 LESI-0 | FFLF-0 8-1 SF, LLOL-Z 9 = 6180 8680 09% E81 t= FOTS-0 coTE-0 0€0I-0 | I8TS-0 66-58 1L-6 £086-T Ie = SF6LT OL6LT 097 G19-1 e+ #91g-0 | Z9L-0 | 0€z80-0 | To0¢e-0 £6-0F CL-31 SL18-T a = S0FST SCFST 0% 166.1 0 Goce-0 | ¢zze-0 | 219990-0 | #8Lc-0 99-FF 0-81 062-1 sl + OTST CPIST 08% 97-1 r= 828-0 | ¢8Z8-0 | GeFeo-0 | 2209-0 49-8 ¥-81 1669-1 Qos O6IIT 61IT 0% 68-1 T= rree-0 | GhEee-o0 | F9FFO-0 | €F29-0 68-16 F-6G 1109-1 8-1 + | g-L¢r6 9876 OZ S&E-1 I- FOFE-0 | GOFE-0 | 91980-0 | &FF9-0 99-#¢ B-L2 1GE¢-1 0-8 + | 0-376, FS6L 002 L8G-1 T+ F9FE-0 89F8-0 | Z10¢0-0 | ¢299-0 09-L¢ G-€8 ¥60¢-1 €-1 +] ¢.0z99 6199 061 6FG-1 a+ 858-0 1Z¢8-0 | 69%20-0 | 9619-0 46-66 ¢-0F OILF-T GSI — | g-ELF¢ L8¥G O8T O1G-T e+ 88-0 6168-0 | 400z0-0 | 8&69-0 19-29 6-6F IL8F-1 1-06 — | 6-48*F cost OLT GLT-T e+ 1498-0 8898-0 | 1910-0 | SITL-0 91-99 6-19 GCOF-T ¥-61 — | 9-Le98 1g9¢ 09T FLL T+ 6698-0 8698-0 | Z6Z10-0 | 219%4-0 69-19 FLL O9LE-1 8-3 — | &-L166 0262, OST SIT-1 I+ 618.0 8G18-0 | FZ010-0 | sTFL-0 61-02 1-16 6848-1 ¥¢ — | 9.018% 918% OFT 260-1 e+ LI8€-0 FIS€-0 | 00800-0 | SF¢L-0 ¥8-6), 0-31 8PE-T 0-* + | 0-GogT 1081 O8T ZL0-1 I- 9188.0 1188-0 | 0900-0 | %692-0 LI-SL 8-191 1008-1 93 + | 9.8881 9881 a 890-1 I- FE6E-0 868-0 | 8FL¥00-0 | Z82-0 FO-LL 9-016 ZS8LZ-T 8-3 + | ¢-0g0T FOL Olt L¥0-1 I- 668-0 668-0 | 08¢g00-0 | 1962-0 80-61 8-612 LLSG-1 | 12-6 — | 6L-6LL 0-982 001 940-1 o- 0¢0F-0 GS0F-0 | L9Z00-0 | 2208-0 64-08 0-FLE 188G-l | 68: — | II-L9¢ G-69¢ 06 = S= LOTF-0 OITF-0 (6100-0) | 10z8-0 = = F6IZ-T | oF-0 — | GT-goF 9-80 08 = TS C9TF-0 991F-0 (€100-0) | 0z88-0 = = 6106-1 | 6-0 — | 19-622 6-616 OL = g= GCBF-0 CCF-0 (6000-0) | OFF8-0 = = 6FST-T | 0-0 + | 0¢-8sT 0-881 09 = t= OSGF-0 T8ZF-0 (9000-0) | 1¢¢8-0 = = L891-1 | 18-0 + | 18-8eT 0-831 0g = L= LEeP-0 SEeF-0 (4000-0) | ZL98-0 = = GSSI-T | 11-0 + | 10-82 06-21 OF = gs F68F-0 L6EF-0 (000-0) | 1628-0 = = GLEI-T | 61-0 — | 99-L% GLLE 0g = 0 ISFF-0 IGFF-0 (000-0) | 1068-0 = = C8oI-T | OT-0 + | 98-22 GL-LG 0Z = 6 80SF-0 90¢F-0 (1000-0) | 7106-0 = = 8601-1 | #0-0 + | 6g-¢T GG-GT OT = ot F9SF-0 | %99F-0 | (¢0000-0)| 4216-0 69-16 = 0961-T | 10-0 + | 18.8 08-8 0 = = = = = = = = = S10 + | 0G-% 60-7 oT — *SoTLO[VD) “O°0 ‘00 “mur “TU palegeeg *8qQ-"9[RQ |, : -rnodv A tpt cassia | AO Vy [PEI | EN smodeq | ap pojertang | PE *sqQ-"9T89 |. ; poyeiniag payengeg | “PMbrT “u0T} + LPEIDIEO) | EON ETIIO) rey 4 eee “mode A pux pmbry jo uray reoden *SOAINO TOIT pvoy ein} oy ie 4 oo < BURCIOGUCH § Jo onyey “AjIsue(y “WRI @ JO oWNIOA “ornssorg-anode A “ALVNOIAOUd TAHLY 30 SCIENT, PROC, R,D.S., VOL. XII., NO. XXXI, n Society. U yal Dubl ings, Ro, fic Proceed y Scient 438 206-2 — 008-0 = (Z00€-0) | (008-0) 0 (1¢8-8) (188-8) G2 = 9169 6909 816-6 Z+ 6008-0 1008-0 10ZZ-0 | ZI8¢-0 91-11 9-F 6869-6 6F 11963 0996 979-2 e- 1208-0 FZ08-0 S¥6I-0 | OOTF-0 SF-9L S1-¢ 068F-Z im = 6084 OZ8FG B8E-G te 6808-0 8806-0 1691-0 | 988F-0 96-1 16-9 L6LG-Z f= 88183 G6LE IZ1-G 0 6908-0 6908-0 QIFI-0 | 1ZL¥-0 96-92 90-2 OSIT-Z 6 GIGS 091% 698-1 0 8218-0 8218-0 1601-0 | 991¢-0 FF-F8 11-6 8986-1 g@ = ZSI61 S8I6I 189-1 0 L81&-0 1818-0 96980-0 | gGoge-0 T1-0F GIT F9I8-1 Bl = Seco OSc9T SLo-T G+ 9468-0 FPEE-0 SFIL0-0 | SLLG-0 FL-FP 0-F1 SOSLeT 9g — F6IFL O8F1 ESP-T e+ F08S-0 10€8-0 | 8F8¢0-0 | 8109-0 80:8F T-L1 9199-1 fit sr GEIZI GOTSI PIPL y- 988-0 1988-0 Tg8F0-0 | 1629-0 18-1¢ 1-06 1669-1 iil 4p 96Z01 G8Z01 ShS-1 i= TGPE-0 BCFE-0 8960-0 | 8FP9-0 89-Fg G9 019-1 GL +1] &-1698 #898 966-1 i= 6LF8-0 O8F8-0 8980-0 | ¢$g99-0 IF-Lg 9:08 110-1 0 0-L83L L8GL 098-T T+ 9868-0 cece.g | 6690-0 | 0089-0 1L:6¢ 60-18 GOLF-T 6-0 +] 6-8909 €909 GS-1 e+ G6g8-0 £68-0 | $1%0-0 | 969-0 00-29 GL-oF 098F-T 0-91 —]| 0-F00¢ 0Z0¢ 961-1 t+ B498-0 1¢98-0 | LOSTO-0 | ZzIL-0 16-F9 cg-cG¢ IFOF-T G-61 —| %-160% I11F 991-1 if ae 6028-0 8048-0 | 6&FI0-0 | 0222-0 £9-99 69-89 9GL8-T 8-91 —| &11ge szeg OFL-T I+ LOLS-0 9918-0 S9110-0 | SIPL-0 8-89 9.98 98FE-1 £2 —| £-6792 199% OZI-T tt ar 8Z88-0 ZZSE-0 | ¥6B600-0 | Teez-0 78-01 9-LOT SFES-1 1:9 —| 6-860¢ 001z OII-T T+ 088¢-0 6188-0 | $¢8200-0 | 4892-0 18-1, 0-981 S108-1 0-¢ +] 0-689T LZ9I 60-1 i= 9868-0 1868-0 | 80L¢00-0 | 9182-0 18-FL G-SLT FOLG-T | OL-6 +) L-GS3T Siva 80-1 I- £668-0 #668-0 | PLE400-0 | GF6L-0 60-9 9-8Z% 18G%-1 | SF-> + | $F-cF6 0-1¥6 FL0-1 I- 6F0F-0 090F-0 | 00gg00-0 | 908-0 O8-LL 0-808 #68G-1 | #80 —| 98-002 1.00 = i= GOTF-0 901F-0 | (4200-0)| g8Ts-0 = a €1ZG-1 | $-T + | #6-80¢ 0-106 = B= I9TF-0 S91P-0 (2100-0) | gogs-0 = = L80Z-1 | 48-0 —| ¢9-098 P-198 _- 0 LIG¥-0 LIG¥-0 (G100-0) | <2Fs-0 = = SL8T-T | 28:0 —| &-6FS $.082 — T+ GLEF-0 TLZF-0 (g000-0) | ¢geg-0 = = QTLT-T | 98:0 +] 8-291 CLOT — T+ SZSF-0 LZEF-0 (c000-0) | 698-0 — = Z9ST-1 | 01-0 —| $¢-60T 69.601 — i sp E88F-0 G8EF-0 (g000-0) | 0928-0 = = QIFI-T | 90-0 —| 1-69 03-69 — I- LEFF-0 SEFF-0 (2000-0) | €228-0 = a OLZI-T | 60-0 +] 40-GF o6-1F — 0 B6FF-0 B6FF-0 (1000-0) | 868-0 =— = ISII-T | 10-0 —| 9-42 8.4% — 0 LEST-0 LECH. (1000-0) | 606-0 == = 8660-I | 11-0 —| 69-8T 08-81 — 0 109F-0 109%-0 | (0000-0) | 900z6-0 = _ 6980-T | 0:0 —| La-2 08-2 — = = — = — — _— = 01-0 +] G9-¢ G¢.g T=) A(0) (279) ‘o°0 0"0 “wu “Uru “TUL Eq Os 8) “poye[noyey) *“peatesq (¢ node ~pmnbry smodv,a | 01x V | sanodu, es j *SdO7 9180 | payeanyzeg pareanyeg ‘pinbryT “m0ry Vv “paze[noTVy | *peatesq¢, eae -mmodeA pue pmbry jo uvayy aa *SOAIND WLOIT pray 0} [enjoy yoy fo oney “AjIsue(y “WUBI @ JO 9WMN[OA. -ainssargq-inode A. ([eo13119) €+18Z 08Z 81% GLE OLG 094 096 OFS 08% 0GZ OZ 00Z 061 OST OLT “epeisiyuay ain} -viodmay, “ALVYALN TAHLAN 39 4 y Pure Substances. 6 t WY Youne — Vapour-Pressures, &c., of Th i} ) ) ((eoHD) 998-8 = G108-0 = (Z10€-0) | (G108-0) 0 (0z8-€) (08-8) 1ZI- 61996 OFLGS | GG-L9% 696-G Ol = 6108-0 6608-0 8923-0 0618-0 92-01 I} {869-6 66 — IFGSS CHEST ¢-996 FOL-G Ih = 8Z08-0 GE0E-0 &£06-0 90F-0 &o-F1 66-4 OSLF-G 0) = OGLEG OSLEG G9 906-6 Sinton OF0E-0 8F08-0 SE8T-0 8°GF-0 68-21 PPG GSPE-G Lv = EOFS 080F% 696 966-6 [ie 8908-0 6906-0 &G91-0 C6PP-0 16-14 91-9 CECE G + GE08G 080E% 096 9ET-G i oF 1808-0 9806-0 LEFT-0 CELP-0 FG-GG 96-9 OZII-Z oI + GELIG OPLIZ 996 696-1 6 = LITE-0 6ITE-0 SIZT-0 1609-0 L1-06 16-8 G166-1T el + OFG6L SG661 096 &GL-1 Cues 9L18-0 FLIE-0 £1960-0 9866-0 90-98 F-01 G9¢8-1 i? or PILL O6TLT OFS 609-1 @ ar 9868-0 1865-0 GGLLO-0 069¢-0 &0-1F [6-61 GLEL-T g¢ + SoLFL OOLFT 0&6 $08.1 0 C668 -0 6668-0 6890-0 1966-0 91-SF 6-ST GLLO-T ee + £096T OLSZT 06 SIFT (f= FOEE-0 LGSs-0 1F1LS0-0 0069-0 68-8F SP-61 6619-1 GL + GOLOT 06901 O1G FGE-1 ORS: GIFS-0 L1¥&-0 8GGF0-0 IL¥9-0 £0-6¢ C9-6 86¢9-T G61 + ¢-0806 TL06 006 COE-T 0 IL¥S-0 ILPE-0 06F£0-0 £699-0 99-PG 99-86 891-1 G6 + G-99¢L Lge) 061 696-1 € + 0&S8-0 LEGE-0 6980-0 L9L9-0 61-L¢ G8-FE SLLP-1 0-L =] 0-0629 1669 Ost G3G-1 @ ar 1896-0 F8CE-0 SFE60-0 £869-0 69-6¢ 99-GF POPP Gol —| ¢-§si¢ 961¢ OLT 061-1 (3 ar 998-0 €F98-0 £0610-0 9604-0 G1-69 Co-oG G60F-T PLT — | 9-086F 8hOP 09T O9T-1 Cats €0L8-0 TOLE-0 €&S10-0 8PZL-0 1G-¥9 G6-G9 9618-1 Goll = |] The AE SIbs Oct S&1-T a + 1948-0 618-0 ¥GC10-0 96&2-0 68-99 L-18 0696-1 GL — | 8-61 S616 OFT SILT (6 9p 0G8E-0 SI8&-0 812600-0 6E¢L-0 08-89 6-G0T FIGE-T OF + G-9F1G G-GFIG 08st OOT-T ad LL88-0 8188-0 | 8¢9200-0 0892-0 G9-02 1-161 1608-1 0-F +] 0-999T1 0-G99T 0G G80-1 6 = ¢&68-0 1866-0 | 8800-0 GI8L-0 LL-GL 0-0LT 9626-1 GG +] ¢-GL21 0-0L81 OIL 890-T B= 6668-0 {666-0 | GLFFOO-0 CP6L-0 LL-VL 9-866 9896-1 G8-G + | Z8-Fo6 0-666 001 190-T OS 6F0F-0 T¢0F-0 19&&00-0 6908-0 GE-9L G-L66 6686-1 tPF — | 9¢-G0L 0-202 06 a toed LOTF-0 SOIF-0 (¢z00-0) | Z6T8-0 a = L0GG-1 88-0 +) 88-¢0¢ 0-90¢ 08 =e ae FITP-0 COTF-0 {S100-0) | Zrgs-0 a == 006-1 0¢-0 + | 0L-G¢8 G-9¢8 OL = 6 = OZCF-0 GGGP-0 (€100-0) T&F8-0 = = T98T-T 06-0 — | 09-8%3% 8-EF% 09 a 0 LLGP-0 LLGF-0 (6000-0) | 9¥¢8-0 = an LOLT-T 10-0 + | L0-G9T 0-291 0¢ zi 0 FEEF-0 FEEF-0 (9000-0) | z998-0 = =F GPGT-T 92-0 — | FF-FOT L-FO1 OF x 0 06EF-0 O6EF-0 (000-0) 9LL8-0 = ios PET-T 8h-0 — | 16-49 c7-¢9 0g =a Tisst: LYFT-O OFPF-0 (000-0) 0688-0 a i GPCI-T 60-0 — | 88-8¢ 06-88 0G = 0 GOSF-0 GOGF-0 (1000-0) £006-0 ae ar LOIL-1 01:0 —| 08-22 OF-GG OT ae Task 8ooF-0 LGGP-0 (40000-0) | 1g116-0 = 2 €260-T ¢0-0 + | 06-31 CL-GL 0 = rx is oe 5 Po See = ee G10 + | $§-9 66-9 o0L — “SoTlo[ey) *0°0 O°9 medgerg “TUL “mur *sqQ—"9TBQ, ‘ “node, : Pero once paye[noreg | ‘paaresqy Re ‘ payeanyg pmbry HOR : Leen payeanqeg pmbry ae i a peyepnoyeg | ‘pearosqg “quencinnag oer oan ‘anode, pue pmnbry Jo uveyy se VA | -saaano Wor pray A Sine 0} [enjoy qeayy pearl 70 O18, *Aqisuaqq “MUvIN) B JO eWN]OA ‘aanssarg-inode A ‘HLVYUALNAOST TAHA 302 ety. ‘ 7 n Soe ‘the Royal Dubli gs of © Proceedin 440 (teoHN9) 6G0.F _ = = (8ZLZ-0)| (6628-0) 0 (F19-8) (F29-8) 66 + 69169 0996 OFS = = — — = S ay 3 as: LGI+ 8966¢ GPI6E C686 este = = ma SLST-0 ze re SEG se Gris 06L8¢ TPL8¢ 6&6 140-8 BG + FE16-0 G1ILZ-0 6811-0 GS98.0 &6-bP 6G-G OTSL-G Gis OTES8¢ 6658S G-8E6 68S crp+ 8812-0 £696-0 IS91-0 cOLE.0 66-0G G6-G 0669-3 $96r 6E8L¢ 9QLGLS 886 = = = me = 0288-0 rae aa 0169-2 gg - 0069¢ 6694 LEG 899-6 E+ PGLZ-0 0€L6-0 COST-0 GG6E.0 99-19 $9-9 0869-4 LPEt TL69¢ bO9GG 9&6 61¢-G fh SP OLLG-0 6916-0 ISST-0 GPIF-0 ST-0L FOL OSLF-Z GIGt IG1hG 68689 PEG 868-6 0 9812-0 9823-0 LLZI-0 C6GP-0 SLLL 8-2 O8ZE-G OLI+ SLEZS GO0ZGE GES 160-3 Shek 1086-0 8616.0 LS8I1-0 OLPP-0 L¥-F8 GP 0896-6 LEor T990G FLFOG 08% 160-Z 0 6882-0 6686.0 001-0 GLOV-0 0S-66 16-6 0661-6 eer O€¢OF L6G9F GGG Sh6-1 g) > 9186-0 688.0 €980-0 006F-0 84-211 89-11 OLF0-6 sort ANG ELGGP 033 STL-T Spe 8F66-0 $966.0 1G¢90-0 GG69.0 8L-FET €&-GT 0606-1 8h — GGLGE OLLEE OIG 669-1 6 = LIOE-0 610€.0 1090-0 0&9¢.0 ¥8-1G1 OL-61 0808-1 Ga = 8896% 18166 006 LL¥-1 Te 808-0 G806.0 O10F0-0 OLLG-0 $9-G9L F6-F6 0seL-1 AB Shh G19FG 061 FOF-1 leas SPIS-0 6h 16.0 98T60-0 0869-0 QT-LLT OF-TE 0ZL9-1 G4 = 91002 68006 Osl PPE-L oP O1ZE-0 9068-0 9930-0 0919-0 GG-S8L 69-66 069-1 i, = S191 C6Z9T OLT 666-1 0 OLGE-0 OLGE-0 F6610-0 OFE9-0 FE-861 91-0¢ OLLG-T Gi6 = 66661 LOOST 09T 6FS-1T Cheat 8686-0 9GEE-0 910-0 C6F9-0 §1-906 00-9 OOFS-T (i 98801 OT GIG-1 (AP F8EE-0 188-0 91Z10-0 0799-0 GL-916 GG-E8 090S-1 el + 1108 OFT OST-1 @) ar SEPS-0 GEPE-0 64€600-0 0L19-0 LO-F66 9-90T OLLF-1 8-6 — (G74!) 0st IST-T @ op 16FE-0 98PS-0 GPIL00-0 0069-0 00-28% 0-0FT O6FF-1 60 + GLP 0Z1 961-1 7 oP TFEE-0 LEGE.0 9LE¢00-0 0601-0 16-686 0-981 CRSP 1-0 + 19g¢ OIL FOL-T i or 166-0 06¢8-0 ¥86600-0 OFTL-0 10-966 0-193 OLOF-T Goya 1693 oot 80-1 ie oF T¥96-0 OF9S-0 106600-0 0861-0 91-696 0-FPE O6LE-T ing > LOST 06 690-1 eat 6898-0 8898-0 #80Z00-0 CGSL-0 96-846 8-6LF 0098-1 il) = T¥EL 08 990-1 (g = GgLs-0 LELE-0 ¢9P100-0 O9FL-0 T¢-£96 L-389 COFE-T — = OL PFO-T Cats 182é-0 §8L6-0 900100-0 GccL.0 T¥-696 8-666 OFZE-1 O10 — 0-969 09 C801 Ox oe 9286-0 8286-0 6€19000-0 0¢91-0 PL-PL3 PSPL €L0-1 Lb-e + 0-90% 0g 860-T hase IL88-0 GL8E-0 F6EF000-0 OFLL-O 81-116 9L3G 0666-1 fA) = G-096 OP 1G0:1 ee 9168-0 6166-0 | G142000-0 CZ8L-0 10-686 8096 0816-1 83-0 + 0-09T 0€ LIO-T oP 1968-0 8968-0 £691000-0 SI6L-0 $G-F86 6689 PEIG-T 06:0 — 0-96 0G 10-1 if Sr COOF-0 FOOF-0 9660000-0 8008-0 9§-L86 OFOOT 88h6-1 8F-0 — L-¥G Or 600-1 0 0¢0F-0 OGOF-0 960000-0 0018-0 L1-686 OOSLT 9PEG-1 0 9-66 0 = ro =F = cae — = = — id) = ¢-GT oO — “SoTLOTep) O°0 OO “mu “TUL z 840-18) “popeynoyee "poaAlos “mode A *pmbr sanode, | 301 x ¥ | 22veIM2l8O | ‘Peatesqo een poraingeg | PDE ae peywinyeg newer oTEInIZS “pinbry “W0T} Vv “paze[nayey | “peatesqg -apeisjua9 jo Ayisuaq smodv, puv pmbry Jo urvoyy SCAROHON, || ceansaaay Wolf pvory any Teoneren ek ue -e1oduia J 0} [enjoy yea 1) Jo oney “Apisua(y ‘uv & JO OUINTO\ ‘oinsserg-anode A TIOHOOTV "AHL C—! i) = ; (jvo1129) j 920-F = GEL6-0 = (¢¢22-0) | (¢¢22-0) 0 (0g9-8) (0g9.8) 06 + OGLE | OG8LF L-8h8 S GST-€ fs = 0916-0 T6L6-0 F916-0 61FE-0 1-36 69°F 6.6 Loot SOLE 90GLF €-6FS S 96-6 me €916-0 8916-0 0661-0 9FSE-0 F-86 Z0-G 028.6 oGit 690LF LI69F BEG S T9L-G 0 OLLE-0 OLL6-0 GEST-0 COLE-0 0-9§ GP-G 669-6 Sats 88Z9F OLG9F 1¥G = 619-6 ar LLLO-0 OLLG-0 GTILI-0 68-0 6-0F §8-¢ F19.6 el + 61SCP POGGE OFS = C10-Z Le sx GF86-0 86-0 CSIl-0 OSCF-0 9-02 18-8 861-6 S1Zt 68E8E 9LI8é 0&G DR SLL-1 0 9066-0 9066-0 F80-0 826F-0 6-88 LIT L10-6 66 + 96168 L608 06% 86¢-T CP 8962-0 €162-0 ¢¢90-0 1666-0 6-SOL LG-GT 068.1 7 or GZ89G 16893 01% © LOF-T tt 0&08-0 Sé0s-0 80¢0-0 89¢¢-0 9-911 69-61 964.1 Gil ar G81GS FIGS 00Z = OLE-T (6. 4° 2608-0 0608-0 1660-0 Z8LG-0 F-861 81-96 C6GL.1 eect CsIsl SLI8T 061 Q 66-1 Sao I¢1s-0 8FIE-0 SI1g0-0 869-0 6-681 I-28 TL19.1 4p POLFL SOLFL OST > 886-1 (Bor 8068-0 068-0 9FPSO-0 919-0 F-SFL 88-0F 669-1 ( 4P 8S8IT 9S8IL OLT = 61-1 fi, 4p L9CE-0 0968-0 9T610-0 6Z59-0 6-961 61-34 08¢.T 6-8h + 6-60%6 9966 O9T Ss LGT-1 3 ar GGEE-0 | GIEE-0 S8FI0-0 6879-0 L-F91 G-L9 IPG. 1 L-6F + 1-898 9GEL 0ST i= FEI-1 QP S4EE-0 | ELSS-0 6S110-0 1899-0 T-TL1 8-98 80S-T 9:06 + 9-989¢ 999¢ OFT < 601-1 {ace [§1€-0 8ErE-0 11800-0 6819-0 9-LLT 0-F1I SLP-T fSo]E => L-SLEF OGEF 0&1 = 880-1 a= 98FS-0 C6FE-0 8900-0 £669-0 G-P8l 6-1ST PPP-L = 0-868 = yas os 890-1 hit = 9896-0 696-0 98400-0 LG01-0 €-061 6-406 LIV-T 8-§ + 8-698% 998% OIL Ry L¥0-1 GES L8G&-0 9698-0 1¢€00-0 LGTL-0 T-L61 6-986 GL6E-1 =a §-G691 car 00T sp 860-1 CAS 9698-0 898-0 06600-0 TEGL-0 9-106 ¢-666 T6LE-1 (ote oS ¢-981L F611 06 a GG0-T if ap $896-0 §898-0 1100-0 8PrEL-0 $-906 0-¢1¢ 6096-1 = 18-118 a 08 & 060-T G+ TELé-0 66LE-0 61100-0 9FFL-0 6-606 6-648 OSFE-1 ee 16-0F¢ = OL S T10-1 § 4p LLLE-O FLLE-0 064000-0 T#SL-0 F-16 FLCI 19GE-1 = 1G-0¢¢ nae 09 = 100-1 Soak 2686-0 6188-0 90¢000-0 91-0 0-916 LL61 TOIS-1 = 8-616 rs" 0¢ a €00-1 (oP 9986-0 986-0 ST€000-0 BGLL-O L-81G OLTE 0966-1 = GP-SEL ae OF RS €00-T 6G Ww 8068-0 9068-0 161000-0 0182-0 1-066 1hGg F086-1 G00 0+ IT-82 90-82 0€ Q €00-T if oP 6F68-0 8h6E-0 TI1000-0 681-0 9-066 $006 8996-1 60-0 — L6-6F 00-FF 0G de €00-1 Ls 6868-0 0668-0 690000-0 6161-0 G-1GG PILOT 896-1 40:0 — LEG LLG OL ~ €00-1T 8 = 8Z0F-0 IS0F-0 €60000-0 6908-0 6-066 OFLOS 60FG-T 0 FE-Z1 |: HG-G of) Sy *soL10[2Q) “970 yn) “THO “TOU “ULUL iS} XS ae omit G§O=AELO | weserenaaay lo mode |. | smodea | s01 XV eel ee sanode A US ae *sqO-"2180 : parnmans events 4 *pmbry wy & “payenopey | “peatesq¢, -apeasqueg 2) TaHnGeee. ‘mode, pur pinbrry jo uvayy es PA | -soaano WOLF pray geen ba 0} [enjoy yoo My fo onjey : “AjIsua(y “UIBIN) B JO AUINTOA samnsserg-inode \ "TIOHOOTY AHL Scientific Proceedings, Royal Dublin Society. 442 (Teoryt19) 866-8 = FELZ-0 = (4g12-0) | (FE2z-0) 0 (sg9-) (g¢9-8) Opi = LEGLE OZI8E | 4-896 = = a = = 0ses-0 = = 696-6 = = = G.$93 = — = = = 0GF8-0 = = 668-6 = = = G1-89% L8¥-% 4p 0913-0 8912-0 OT9T-0 4068-0 G.gg 03-9 199-2 e61 — so6ge S019 09% 10-3 Nie= 9Z8Z-0 8982-0 OSIT-0 CCCF-0 9-06 06-8 O1Z- te + 10808 GgLog 09 Ges.T 0% — B68Z-0 166-0 060-0 0Z6F-0 $89 90-11 GE0-Z 69 + $9696 F619 OFZ $99.5 6 -= 8962-0 1963-0 F0L0-0 0829-0 0-42 13-41 G16-1 9, + 0822S FIGS 08% FOG. 1 0 108-0 1Z08-0 9460-0 GSFS-0 $68 00-81 828-1 el = 19981 6L98T 0% GGP-T @ + €808-0 8106-0 GFF0-0 CTLG-0 G-F6 69-86 OSL-T ie = 6ES¢1 GLSST 01% Ggg-1 9 + eF1S-0 1818-0 €¢¢0-0 0269-0 ara 8:83 689-1 ar 60861 10821 00% F08-T @) ap G0ZE-0 9618-0 6860-0 0119-0 9-601 ¥-98 189-1 ig = CFFOL 99401 061 926-1 Gar 0968-0 €66E-0 GZG0-0 | 869-0 $-91T GPF 166-1 $-gg +] 8-818 S8es Ost S1Z-T Q 4p 918-0 gee-0 | OL210-0 0SF9-0 | 8-80T 9G 09¢-T 8-2 +} 98-1699 G699 OLT FSET i + OLE8-0 6986-0 | O8gto-0 0099-0 0-621 8-3) C1G-1 9:01 —] #-86¢ $96 091 SG-T I + FZFS-0 SaFs-0 | 09010-0 0FL9-0 8.G81 6-86 PSFeL T¢ +] T-LG0F BS0F 0ST SZI-1 6 = 918-0 8148-0 | ¢0800-0 6189-0 F-GP1 0-31 GG¥-1 g-9 +] €-0808 F108 OFT GOT-T I = 1GG8-0 gz¢e-0 | ¢0900-0 669-0 0-LF1 0-991 OgF-T 6% +] 69686 £606 (a 680-1 I - 9168-0 Lhg8-0 | &FF00-0 OTTL-0 0-8ST 0-933 90F-1 0:9 —| O-LL9T S891 ya 490-1 0 9298-0 9298-0 | 0&g00-0 0ZZL-0 0-681 0-218 egg-T Cp, = || tacteraitis 9081 oll 90-1 6 = 6198-0 F198-0 | 9200-0 GZEL-0 0-F9T 0-8hF ggg-1 19-9 —| 68-988 @-ZF8 00T 80-1 0 OZLE-0 OZLE-0 | 9G100-0 GZFL-0 | 0-691 0-869 LEST 68-¢ —| TI-s9¢ 0-FL¢ 06 020-1 0 G918-0 e918-0 | #0100-0 0Z¢L-0 | 0-841 0-886 oge-1 69-0 —| Te-ans 0-918 08 = (ap OT8g-0 8088-0 (9000-0) | 0194-0 = = PIS-1 PPL + | PF-0F6 0-68 ol = USP eeg6-0 3988-0 (5000-0) | OO0LL-0 | = a 666-1 16-1 +) 16-841 0-LF1 09 = o + 9688-0 F688-0 (g000-0) | $82L-0 | = == G8G-1 08-I +] 00-68 BLS 0S = 6 = 1868-0 6868-0 (000-0) | GLS2-0 | == = 013-1 G60 +] GI-1¢ 0 OF = 6 = 8168-0 0868-0 (1000-0) | 0964-0 | = aa 996-1 69:0 + | €1-8¢ 9-16 0g = i= LIOF-0 S10F-0 (1000-0) | ¢g08-0 | = = CPST 82-0 + | 8L-FI C.F 0G = (6 4 LG0F-0 | ¢O0F-0 =. | one@ = = E&E-1 81-0 +] 68-2 96+) Or == B= G60F-0 | L60F-0 _ 086Ts-0 | — = 10-1 c0-0 + | 6F€ PPE 00 “SaLlo[e/) "0°0 ‘0°0 “mul baetgeeg “TUL *SqQ-"9[O |. ; ye “mnodv A ait saoton || aOR peyepnoyeg | poatesqy ay poyeinqeg pmbry SAUDE peyeanqeg payeinqeg | “PMOYT “1074 ean jberuisoleo) apeateed Ol aeamuen Ia) sue ‘mode, puv prubry jo uray TARRY "saATO WOLF pray | any [eotjeto0eyy, ed : Jo -e1odmay, 0} [Rup = —-—-—-——- — - yea Jo o1yey : | -Aqtsue(qy “TUBAL e jo euuNn[OA ‘eanssaiq-imode A. "TIOHOOTV IAdOUd 443 y Pure Substances. arty Youne— Vapour-Pressures, §c., of Th 666-7 961-6 ISF-G 9F6-G 601-6 810-6 8F6-1 168-1 6F8-T LI8-T 884-1 TGL-1 862-1 60L-T ¥69-T 089-1 699-1 699-1 $99-1 699-1 699-1 G69-T 602-1 STL-T 682-1 F9L-1 86L-T OF8-T 688-1 GE6-1 G86-1 sanodv A. payezmyeg jo Ayisuecy [eote109q], 0} [enjoy Jo o1gery (Teon19) =% aa: fase (90¢¢-0)] (9098-0) 0 (6g8-2) (28-2) 19 + OFS O0FSF | 9-128 ao 8Igé-0 x = CT9F-0 = = LOT-G Gi ae: OSSGF EOoGF 0ZE L+ 8196-0 ILE8-0 8IL1-0| &6FS-0 LL-LE £8-¢ PPS-1 GGG + 89TLE £7698 org (ion 8896-0 0F9E-0 11-0} 0966.0 L6-8P GG-L 9089-1 696 + GIESE £7068 008 Vitrwe 6696-0 £0L8-0 €L01-0| F&&9-0 TG-L¢ GE-6 L8L9-1 Oi [S616 TC9LE 066 (3 Sr 6918-0 9GL6-0 €880-0 | 6299-0 8F-89 &&-11 980¢-T 89 = SSOFS EGLPS 086 i sw 6188-0 818é-0 G9E10-0} 0069-0 &€-89 89-1 GOFF iQ = 06606 19906 016 mals 6186-0 9188-0 69190-0 9814-0 96-GL C6-91 PLOF-T gy = LOGLT 6LELT 096 0 OF6E-0 OFG6E-0 £91¢0-0| 982-0 GG.cL 88-61 0698-1 is = GESFI 998FT 086 ee 8668-0 GO0P-0 LGEF0-0| TLSL-0 81-82 GL+86 LOGE-T X= GLYCI LOGE OFZ oe. 890P-0 €90F-0 969€0-0| F9LL-0 91-08 09-26 0682-1 t+ 9CFOT GOFOT 0&6 = 9TTF-0 ZGLF-0 1200-0 | 1462-0 60-68 IL-€€ 696-1 fl ap GG98 8£98 066 (3 = 9LI¥-0 6LIF-0 886Z0-0| 6018-0 LI-¥8 61-0F GEES: @ oP PelL SGIL O1Z Tes PEGF-0 GEGF-0 6500-0 | 928-0 GG-G8 PL-8P 0016-1 ar 9E8¢ &E8¢ 002 1 oar G6GP-0 1667-0 T8910-0 &1F8-0 FL-98 8h-6G 9881-1 Cua GELP SELF 061 Vast OCET-0 9OFEP-0 OLETO-0| $998.0 TL-L8 16-GL 6891-1 16 = 6086 &E86 Ost @} oP 60FF-0 LOFF-0 F8010-0 ¥698-0 ¢9-06 €-06 GOST-L Ko = FE0E 8208 OLT Soar 99FF-0 6SPP-0 18800-0| 6688-0 £9-68 L-GIT 9GET-T (nee G6ES 60FZ O9T () ar S6SF-0 LISF-0 €0400-0| 968-0 FL-06 G:GPL SSIL-L Ol = £981 GIST OST 8 + I8GF-0 &LSF-0 GTS¢00-0| 1606-0 €8-16 €-18T 000T-T og + T&L 181 OFT C= LE9F-0 689F-0 GLEF00-0 | GE66-0 TL-66 6-886 8280-1 (Hi? ar €80T OFOT OST 5 = F690-0 169F-0 14G€00-0 6986-0 8&-F6 L-G0E 6890-1 = F08 Ss OGL iA OGLF-0 FCLY-0 8900-0 | €8F6-0 62-66 1-SOF GFCO-T Gol = 8-08¢ 9-28 Ort Pom GO8P-0 608F-0 S&8100-0! 666-0 68-66 9-Gh¢ 81F0-1 Qo) aF T-L1F GOLF 00T Pe GI8P-0 9984-0 88E100-0 8116-0 88-16 GLPL 0660-1 6:0 + 1-866 8-266 06 i Ae ST6F-0 GG6F-0 696000-0 | %&86-0 69-16 SPOr 8910-T GO + £-606 T-G0G 08 Pes €L6P-0 LLGY-0 €49000-0 8766-0 &P-06 98F1 6900-1 156 Keser 0-981 T-LE1 OL = 6609-0 GE0G-0 | 129F000-0| 0900-T 69-68 PITS 0766-0 ¥9-0 + | ¥6-88 £-88 09 Ci ¥806-0 6806-0 | 001€000-0 CLI0-1 F1-88 96GE 8286-0 96-0 +] 99-9¢ 6-99 0¢ es 8E19-0 &F1S-0 | G10G000-0} +F820-T 60-48 OL6P PGL6-0 LL-0 + | LL-vS 0-%E OF Go 6619-0 961¢-0 | ¥9G1000-0 6660-1 88-68 OT6L £696-0 IL-0 +] 19-06 6-61 0€ 6+ 8P6S-0 9F6¢-0 | 79L0000-0 1670-1 ¢0-48 08081 3E¢6-0 £00 —| &2-TT 08-11 06 7 409 1089-0 1669-0 = 6660-1 = rae OF6-0 40:0 +] 88-9 F§-9 OL @) ar GcEs-0 6PE¢-0 = 01690-T = = 8h&6-0 06-0 —| 08-8 02-8 of Bi) 6 (0) (719) oh) o"0 paurdueg Barcus “uLUr AO) *payena[eg | padrasq¢, MOON *prnbry BULAN “inode j b PevEIES j *8qQ-"0[89, =| swcrycerencey || RN “Hon ~ payenoyeg | *peatesqg OpuLaNTe) sinode A puv prnbry fo uvayy ees ®A | -soaamo UWLOIZ pvory ean} t -vroduay, yeoyy -Aqisuaqy “TUBIN) B JO OMINTOA ‘dIOV OILHOV einsserg-inode A cL ip cae 2 THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XID. (N.8.), No. 32. MAY, 1910. A SIMPLE FORM OF OPEN-SCALE ISOTHERMAL AIR BAROMETER. BY W. F. BARRETT, F-.RB.S. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATE, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1910. Price Sixpence. Roval Dublin Society. DADA Rm FOUNDED, A.D. 1781. INCORPORATED, 1749. EVENING SCIENTIFIC MEETINGS. Tur Scientific Meetings of the Society are held alternately at 4.30 p.m. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to’ read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Editor. 143 DS S S = Ss S SS w SS = Ss y Pure rey 0 Con Of Mp ) SUreS, S: Youne— Vapour-Pre | (Teo) 366-1 = = = (90¢8-0)| (909¢-0) 0 (2¢8-2) (z¢8-2) 19 + 19¥8h O0FEr | 9-128 == ort 81G&-0 Em = C19F-0 = net LOT-S ee OSSGP SO9Ch 068 | 961-6 [ap 8LE8-0 IL-0 STL1-0 SGPS-0 LL-LE €8-¢ PPS-T GGG + S9TLE EP69E ole IGF-G Gz 8&9E-0 OF9E-0 1€€1-0 0969-0 L6-8F GS-2 9089-1 696 + GLSZE SPOGE 008 9FG-G Ve: 6696-0 6018-0 ELOT-0 $§69-0 16-L¢ GE-6 L8L¢-1 Ola TS616 TG9LG 064 601-2 (aa 6GL8-0 9¢18-0 €880-0 6299-0 8F-69 6é-11 G80¢-1 th) = CCOFG SGLFG 086 810-2 iL SP 6188-0 8188-0 69¢10-0| 0069-0 66-89 89-1 G6FF-1 iQ = 06606 T9906 OLG 8F6-1 & + 6L8¢-0 9188-0 ¢9190-0 9811-0 96-61 GG-9T FLOP T og LOGLL 6LGL1 096 168-1 0 OF6S-0 OF6E-0 8910-0 F9EL-0 Gg.GL 86-61 06¢8-T ie = GE8t1 998F1 096 GPS8-1 v= 8668-0 GOOF-0 LGEF0-0 TL9L-0 81-82 61-86 LOGE Ch = GLPGI LOGZI 0G LI8-T Cec SGOF-0 £90F-0 9Z920-0 POLL-O 9T-08 09-16 0682-1 t+ 9GFOT COFOT 0&3 882-1 = OTIF-0 GGLF-0 1G0€0-0 I¥6L-0 60-28 I1-&& £69¢°1 JN ar g¢98 8698 066 IGL-T 8 = 9LIF-0 6LIF-0 88420-0 6018-0 LI-¥8 61-0F GEES-1 @ ap PSIL SOIL O01 83-1 = FEGP-0 GSGP-0 6S0Z0-0 [968-0 GG.¢s PL-8P O00TZ-T (8 Ap 9€8¢ E8¢ 006 60L-T I + Z6GF-0 1GaF-0 I8910-0| &IF8-0 FL-98 8F-6¢ 9881-1 (6 ar GELP SELF 061 469-1 $+ 0G2F-0 9FEF-0 OLE10-0| &¢¢g-0 TL-L8 16-61 6891-1 — 6088 S888 Ost 089-1 8+ 60FF-0 LOFP-0 F8010-0 $698-0 G9-06 §-06 GOST-T io = FE08 890€ OLT 699-1 fh a 99FF:0 6SPF-0 18800-0 6288-0 £9-68 L-G1T 9GETT JE => C686 60% oot 6¢9-T @) 9F SGSF-0 LIGP-0 60L00-0 6968-0 FL-06 G:GFL SSIL-T Gil = S98T GL8T OST $g9-T 8 + I8¢h-0 SLSR-0 STS900-0 1606-0 68-16 6-181 O000T+T og + ISh1 8&1 OFT 699-1 6 = LE9F-0 6E9F-0 GLG¥00-0 866-0 I1-€6 6-886 8280-1 ey + S801 OFOT O&T 699-1 (iS F69F-0 L69F-0 116€00-0 6966-0 8E-F6 L-G0€ 6890-1 = 08 = 061 669-1 bE OGLF-0 PGLP-0 89FZ00-0 €8h6-0 61-66 1-G0F CFG0-1 Ebi = 8-084 9-G8¢ OIL €OL-T Bis COSF-0 6O8F-0 €&8100-0 6696-0 68-66 9-GFG 8LF0-T 90 + I-LI¥ G-91P 0oT STL-1 iS G98F-0 998F-0 SES100-0 8116-0 88-16 G-LUL 0660-T 6-0 + 1-66 8-666 06 681-1 v= S16F-0 GG6F-0 696000-0 G§86-0 66-16 Sol 8910-1 G0 + §-G0G 1-606 08 FOL-T VS €L6F-0 LL6F-0 €19000-0 8P66-0 6-06 981 6900-1 1a ee oe? 0-9€T T-L81 OL 862-1 Sines 6G0¢-0 GE0°-0 1Z9F000-0 0900-1 69-68 F9IG 0P66-0 49-0 3 o+ $6-88 €-88 09 OF8-1 G = F80¢-0 6806-0 | 001€000-0 GLIO-1 F1-88 96GE 8686-0 98-0 +] 9¢.9¢ 6:9¢ 0g 688-1 One 8€1¢-0 SFI¢-0 | 610Z000-0 $860-1 60-18 OL6h ¥GL6-0 LL-0 + | LL-P8 0-F§ OF ZE6-1 e- ©61¢-0 961G-0 | #9Z1000-0| Z620-1 88-8 O16, £296-0 IL-0 +] 19-06 6-61 0€ 86-1 G+ 8465-0 9FZ¢-0 | $9L0000-0 160-1 ¢0-F8 O80ET 3846-0 0:0) Sal Sher 08-11 06 — i 10€¢-0 1669-0 = 660-1 oF er OFF6-0 40-0 +1] 88-9 bE-9 or = §) ar GGEés-0 6FSG-0 i 01690-T ane mn 8hE6-0 06-0 — | O08-€ 0¢-€ 00 *SoLIO[eD) BRD °9°0 *ULUL “UU “uur *8q0-9]8) |. 4 node A : anedane (sony pozepnayeg | “poatosqg anaes poyeinqeg pmnbry LO PERE) || = =| povanuvect | eiera “u0T} o 9 /-poyemnozug | “poarosqo |, g o Ayisue(y F Lata) -eztiodeA | | ey sehe I) aEaieiGoue anode pue pmbry jo uvoy, jo SOAIND WOIE proxy Chum) on jEnoy ae ywory meses SRE Jo o1yuyy -Aqsua(y “UII B JO OUINOA ‘aansserg-anode A. ‘dIOV OLLGOV 3P SCIENT. PROC. R.D.S., VOL. XII., NO. XXXI. [ 444 3 XXXII. A SIMPLE FORM OF OPEN-SCALE ISOTHERMAL AIR BAROMETER. By W. F. BARRETT, F.R.S. [Read Frsruanry 22. Ordered for Publication May 10. Published May 19, 1910.] A LONG-NECKED glass flask filled with air and inverted in a wide-mouthed bottle of water is to be seen in some cottages, where it is used as a weather- glass. If kept at a uniform temperature, the changes in atmospheric pressure will, of course, be indicated by the rise and fall of the water in the neck of the flask. But as this cottagers’ weather-glass is usually kept on the mantel- piece, the changes of temperature to which it is exposed vitiate its use as a simple form of barometer. If, however, the flask could be rendered impervious to temperature change, it would form an effective and very sensitive barometer. It occurred to me some years ago that this isothermal condition might to some extent be attained by using one of Sir James Dewav’s liquid air flasks. As is well known, in order to preserve air in the liquid state, Sir J. Dewar devised a glass flask with double walls, the space between being highly exhausted and the walls coated with a reflecting surface of deposited mercury or polished silver. Such a flask was fitted with a thick rubber cork, through which passed a glass quill tube open at both ends, having a liquid index to record changes of volume. Any degree of sensitiveness can easily be attained by altering the ratio between the volume of air in the flask and in the index-tube. In order to protect the latter from sudden temperature changes, the quill tube is encircled with a larger glass tube, closed at its upper end, and having a plug of cotton wool encircling the annular space between the two tubes at the lower end. A scale is now attached, the graduation of which can be easily made by comparing the rise and fall of the index with that of an ordinary barometer when both are enclosed in a receiver and submitted to slight exhaustion by an air-pump. A range of from 6 to 10 inches to 1 inch of barometric change was found to be con- venient. Bending the index-tube at right angles after it leaves the flask, so that the scale is now horizontal, makes the arrangement more convenient; and to increase the portability of the instrument, the long index-tube can be coiled spirally. Barrerr—A Form of Open-Scale Isothermal Aw Barometer. 445 Instead of a short liquid index, another arrangement, which was found convenient in practice, was to pour a little mercury—or other liquid having a low vapour-pressure, such as creosote—into the Dewar flask, and allow the lower end of the quill index-tube to dip into the liquid. The displacement of the air in the flask, caused by pushing in the thick rubber cork, was found sufficient to raise the level of the liquid to a convenient height in the index- tube. My former lecture assistant at the Royal College of Science for Ireland, Mr. George Hughes, made up one of these isothermal air barometers for me more than a year ago, and has had it under cbservation ever since. He has found its variations agree extremely well with the ordinary mercurial barometer. The cost of such an instrument is trifling; and as a sensitive weather-glass it may be found useful, although it cannot pretend to be an instrument of precision, or to take the place of the ordinary barometer, until temperature changes are wholly excluded. P.S.—Since the foregoing paper was read before the Society, my atten- tion has been called to the fact that the publication of my new weather- glass has been forestalled by a paper recently published in the Comptes Rendus. In this paper, read before the Paris Academy of Sciences on December 6, 1909, M. G. Carpentier describes an isothermal air barometer, made by the Marquis de Montrichard, which is practically the same as the one I have here described and have had in use for over twelve months. A model is depicted, resembling an aneroid barometer, in which the graduated index-tube is coiled into a spiral form and separated by an opaque glass division from the small Dewar air reservoir below. It is stated that the inventor has lately made an ingenious addition to the apparatus whereby an absolutely uniform temperature may be maintained in the air reservoir. This consists in placing some ice inside the Dewar flask, or, more conveniently, the ice may be dropped from outside into a little well or receptacle made in the side of the air reservoir. No doubt, if the neck of the aperture to the well be made small and plugged with cotton-wool, the ice will not need renewal for some days; and if this be so, the apparatus may, for certain purposes, become a useful laboratory instrument. Stl a ny win te ; Phe 1 Mariel Sts i ‘ THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.S.), No. 33. JULY, 1910. AGRICULTURAL SEEDS AND THEIR WEED IMPURITIES: A SOURCE OF IRELAND’S ALIEN FLORA. BY T. JOHNSON, D.Sc., F.L.S., PROFESSOR OF BOTANY IN THE ROYAL COLLEGE OF SCIENCE, DUBLIN ; AND iu sansoniaa inca Se ‘hege. MISS BR. HENSMAN. (poy... “\ hy. aio, Oia! iusew™ (PLATES XxXIl., XXIII.) [Authors alone are responsible for all opinions expressed in their Communications DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY LEINSTER HOUSE, DUBLIN, WILLIAMS AND NORGATEH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1910. Price One Shilling. Roval Dublin Society. FOUNDED, A.D. 1731. INCORPORATED, 1749. EVENING SCIENTIFIC MEETINGS. Tur Scientific Meetings of the Society are held alternately at 4.30 p.m. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Hditor. Barrert—A Form of Open-Scale Isothermal Air Barometer. 4465 Instead of a short liquid index, another arrangement, which was found convenient in practice, was to pour a little mercury—or other liquid having a low vapour-pressure, such as creosote—into the Dewar flask, and allow the lower end of the quill index-tube to dip into the liquid. The displacement of the air in the flask, caused by pushing in the thick rubber cork, was found sufficient to raise the level of the liquid to a convenient height in the index- tube. My former lecture assistant at the Royal College of Science for Ireland, Mr. George Hughes, made up one of these isothermal air barometers for me more than a year ago, and has had it under observation ever since. He has found its variations agree extremely well with the ordinary mercurial barometer. The cost of such an instrument is trifling; and as a sensitive weather-glass it may be found useful, although it cannot pretend to be an instrument of precision, or to take the place of the ordinary barometer, until temperature changes are wholly excluded. P.S.—Since the foregoing paper was read before the Society, my atten- tion has been called to the fact that the publication of my new weather- glass has been forestalled by a paper recently published in the Comptes Rendus. In this paper, read before the Paris Academy of Sciences on December 6, 1909, M. G. Carpentier describes an isothermal air barometer, made by the Marquis de Montrichard, which is practically the same as the one I have here described and have had in use for over twelve months. A model is depicted, resembling an aneroid barometer, in which the graduated index-tube is coiled into a spiral form and separated by an opaque glass division from the small Dewar air reservoir below. It i8 stated that the inventor has lately made an ingenious addition to the apparatus whereby an absolutely uniform temperature may be maintained in the air reservoir. This consists in placing some ice inside the Dewar flask, or, more conveniently, the ice may be dropped from outside into a little well or receptacle made in the side of the air reservoir. No doubt, if the neck of the aperture to the well be made small and plugged with cotton-wool, the ice will not need renewal for some days; and if this be so, the apparatus may, for certain purposes, become a useful laboratory instrument. SOIENT. PROC. R.D.S., VOL. XII., NO. XXXII. BQ (ie 1446" XX XIII. AGRICULTURAL SEEDS AND THEIR WEED IMPURITIES : A SOURCE OF IRELAND’S ALIEN FLORA. By T. JOHNSON, D.Sc, F.LS., Professor of Botany in the Royal College of Science, Dublin ; AND MISS R. HENSMAN. (Pratrs XXII., XXIII.) [Read Marcu 22, 1910. Received for Publication Aprit 12. Published Juny 22, 1910.] Mempers of the Royal Dublin Society have been for years in a position to have seeds tested by the Society's Consulting Botanist ; and, in the year 1899, one of us (T. J.) undertook to report on the germinating power of certain clover and grass seeds for the Congested Districts Board. But it was not until the year 1900, when the Department of Agriculture and Technical Instruction for Iveland started the first and still the only official Seed-testing Station in the United Kingdom (charging farmers 3d. and seedsmen 3s.—now 2s.—for a full report on each sample), that seed-testing on an extensive scale was carried out in Ireland. We have been in charge of the work (1) of the Station from the first, beginning with an investigation of the flax-seed supply. During the period 1900-1909 inclusive, Reports have been made on the genuineness, purity, aud germinating power of 11,000 samples, the yearly average being now more than 2000. The following ‘l'ables (p. 447) show the average germinating percentage of these samples. The results obtained by testing seeds of the same species in the well-known Ziirich Station (2) and by the twenty-one different German Stations (3) are incorporated for comparison in T'able II. The Tables show apparently that the seeds bought and sown by the farmers in Ireland have in many cases a lower germinating power than those sown by the farmers abroad. The Department’s Station, as stated in the first annual Report, differs from all others in that, while giving separate reports on the purity and germinating percentage of the seed, it submits to the germination tests, seeds which, though complete in outward appearance, may be without a kernel, i.e, “chaff.” Such seeds are treated as impurities in other Stations. Jounson AND Hensman—Agricultural Seeds, Se. 447 Taste I., showing the average percentage of germination of the chief agricul- tural seeds tested at the Station during the years 1900-1909 inclusive :— 1900- 1902 | 1903 1904 1905 1906 1907 1908 1909 Perennial Rye, . 68 78 ag) 82 82 76 74 72 Italian Rye, . o || 77 78 82 77 74 73 68 Timothy, 4 5 78 92 91 93 89 81 90 93 Cocksfoot, ' . 61 53 54 59 53 56 5a 61 Meadow Fescue, ; 82 75 76 83 75 69 63 61 Sheep’s Fescue, . | 59 26 61 36 47 51 31 21 Hard Fescue, . : 68 1 65 58 49 52 42 37 Meadow Foxtail, : 55 84 39 49 35 38 37 38 Smooth stalked Meadow Grass, . 30 37 57 65 — — 14 33 Rough stalked Meadow Grass, : 77 QD | OB 94 = 57 54 55 Crested Dog’s Tail, s || 8 60 | 64 67 80 75 63 60 Wheat, . | = cB || oi 87 85 = 85 61 Oats, C D 5 88 86 | 69 88 90 93 87 83 Barleysairs soe ee 90 73 88 93 86 84 79 Rye, 5 0 — 46 58 87 47 — 24 68 Red Clover, 3 ; 82 82, 4h* | 88, 4h | 87, 3h | 87, 3h | 87, 8h 75, 5 82, 5h Alsike, . 0 : 74 78, 6h 82, 8h | 76, 6h 76, 6h | 81, 8h 73, 7 76, 9h White Clover, . C 58 59, 10h | 71, 11h | 72, 10h} 73, 12h | 78, 8h 75, 10 | 69, 10h Trefoil, . 2 || 8 | 7, aa | 91,2h | 90, 1h | 72, 1h | 81, 2h | 62,2 | 42, 3h Swedes Turnip, .| 84 87 88 87 86 84 87 85 Mangold, - | 189 141 116 123 136 152 124 131 Flax, ; a 95 Gy | Oe on 94 92 89 89 Taste II. shows the average annual percentage of purity and germination of the chief agricultural seeds tested at the Station, the total number of samples of each kind tested, and, for comparison, the average percentage of germination of the same-named seeds, as tested in Ziirich and also in the twenty-one German stations :— Germination percentage | Total number of samples tested in Irish Purity} Ireland | Ziirich oe Sree : a Perennial Rye (Loliwm perenne, L.), : - 94 76 82 82 1427 Italian Rye (Lolium italicum), ° : 6 90 74 50 17 1087 Timothy (Phlewm pratense, I,.), 6 ; 98 88 92 89 498 Cocksfoot (Dactylis glomerata, Tidy . 91 56 83 78 554 Meadow Fescue (Festuca pratensis, Huds. ), : 97 73 85 74 319 Sheep’s Fescue (Festuca ovina, L.), 5 3 98 41 74 — 39 Hard Fescue (Festuca dwriuscula), A 5 98 46 65 —_ 95 Meadow Foxtail (Alopecurus pratensis, L., 93 39 68 61 183 Smooth-stalked Meadow Grass (Poa pratensis, L. ih 98 39 67 61 10 Rough-stalked Meadow Grass (Poa trivialis, ty ), | 97 66 75 = 24 Crested Dog’s Tail Cree, us cristatus, L. i) 0 96 69 76 _ val Wheat (Lriticwn), . 5 99 83 87 83 61 Oats (Avena sativa), : : 2 . : 99 85 88 87 571 Barley (Hordeum), 9 c : 5 - | 100 84 84 91 315 Rye (Secale cereale), ° o 97 56 87 — 38 Red Clover (Zrifolium pr atense, L. 4 : : 95 84, 4h} 91 85 1444 Alsike (Trifolium hybridum, aly : - 5 93 We OX) 88 83 586 White ee (Trifolium repens, L.), é : 93 69,10h) 80 78 519 Trefoil (Medicago lupulina, 1.), - : i 97 70, 2h} 78 81 101 Swede, Turnip (Brassica), : ‘ ‘ ; 99 86 87 90 759 Mangold (Beta vulgaris), : : ‘ : 99 132 127 = 513 Flax (Linum usitatissimum, L.), . 2 . 98 91 83 83 1625 *h = ‘hard’ seed, i.e. a seed which has a hard coat and does not germinate within the ordinary limit of time, but may need several years. Some Stations allow two_out of three such seeds as viable, and count them accordingly. 38Q2 448 Scientific Proceedings, Royal Dublin Society. This method naturally lowers the germination percentage in, e.g. such chaffy seeds as meadow foxtail. The Department considers that its method gives a truer idea of the quality of the seed. As one of us (T. J.) has ceased to be responsible for the work of the Seed-testing Station, owing to the increased administrative work necessitated by the operations of the Weeds and Seeds Act (Ireland), passed in December, 1909, the time appeared opportune to summarize certain results of the Station’s work during the ten years it has been in operation. Itisnot much more than a generation ago that seed-buyers in any part of the world were made to realize the necessity of testing the seeds offered for sale. To Nobbe, of Tharandt, in Saxony, the credit is mainly due of being the pioneer of seed-testing. In his book, “Die Samenkunde,” amongst other revelations, is that of a letter from a Hamburg to a Dublin firm, offering, at a cheap rate, for mixing with clover seed, stones so like true seed as to deceive the buyer. Fraud of this type is now, let us hope,-impossible ; but still it is often of practical importance to be in a position to say what is the actual origin of a seed, i.e. the region from which it comes. Stebler states, e.g., that Chili red clover, Utah lucerne, and New Zealand cocksfoot are of little value; that cocksfoot of forest growth grows smaller than cocksfoot from the meadow, and that American meadow fescue falls a victim torust. Hence arises the necessity of reliable means of detecting the source of supply of a seed. Wittmack was the first scientific man to interest himself in this question, and in 18738 he recognized a red clover as American owing to the presence in it of seeds of Ambrosia. In 1875 the Ziirich Station began the same kind of investigation; and in 1906, Dr. Stebler, the head of the Station, gave us at Dresden an illustrated account (4) of the results of its work. Sometimes the appearance of the seed itselfi—the metallic lustre of its coat—proclaims its origin. (‘The lustre of Hnglish “ cow-grass”’ (perennial or red clover) is said to be usually artificial.) In some cases particles of soil (e.g. black soil from Russia) or other inert matter suffice. Weed-seeds in the sample are, however, the best guide, though caution is necessary in drawing conclusions from them. Stebler calls those weed-seeds, which indicate the origin of the seed, sowrce- indicators. Other seeds not so reliable, but still helpful, he calls companion- seeds. He divides the seed-supplying districts of the world into the following regions :— 1, Sourn European. (South France, Italy, Spain.) Coronilla scorpioides, Koch. (Arthiolobium scorpioides, Desv.), and Ammi majus, L., in a sample are sure signs of the S. European origin of the seed. JOHNSON AND HensmMan—Agricultural Seeds, &c. 449 2. West Evropran. (Great Britain, N. France, Netherlands.) Alopecurus agrestis, Li. (A. myosuroides, Huds.), is a typical weed, usually found without the glumes, as the naked caryopsis in red clover and in grasses. Cavum Petroselinum, Benth. et Hook. (Petroselinum segetum Koch.), occurs in French red clover. Certain of the West Kuropean weeds are winter-green or winter-annuals, but incapable of standing the severer winters of Hast Hurope. Such are, Stebler states, Alopecurus agrestis, Li, Valerianella dentata, Poll. (Valerianella Morisonii, DC.), Geranium molle, L., G. pusillum, L., G'. dissectum, L. 3. Nort American. (United States and Canada.) Since the American weeds flower late, and are rarely ripe in Mid-EKurope when the European crops in which they may be growing are cut, the presence of American weed-seeds in a sample is a reliable sign of the American origin of the seed. Such are Panicum capillare, L., P. dichotomum, P. virgatum, P. clandestinum, Li., Paspalum ciliati- folium; Cuscuta arvensis, Beyr. (in red clover occasionally, and in lucerne); Plantago aristata, Gray, in red clover and meadow fescue; Ambrosia artemistifolia, L., and A. trifida, L.; Rudbeckia hirta, L., in Timothy, and many others. 4. Austrantan. (Australia and New Zealand.) Agrostis Forster, R. & SS. in meadow foxtail, > ” ” » 1:015 | 12°21 ”» ” ” ” ” 0-90 9) 10°98 Myers ab: i is ik 1-012 | 1217 Feb. 1910. K 3 . 1171 | 14:06] 145 abe Move 3 1-002 | 12-05 EMEA sees » xind, 0:958 | 11°53] 160 ” ” ” ” fruit, 1-029 | 12°37 159 Aug. 1909. | Citrus Aurantium, fruit, 1-400 | 16°84 ” ” 9 ” ” 1-525 | 18°34 ” oe ” ” ” 1-462 | 17°60 deat & 3 1:598 | 19-29 Feb. 1910. <3 EY way 1:020 | 12-27 ” ” ” »” ” 5 : 3 1:031 | 12:40 164 a a Citrus Aurantium, fruit, Tangerine var., 1-289 | 15°51] 208 Mar. s aA » . 1:267 | 15:24] 231 Aug. 1909. | Lycopersicum esculentum, fruit, 0°658 | 7:91 Feb 1910. * a at a 0-656 | 7-89 Coens ch x AR 4. 0-575 | 6:92] 110 ” » ” ” ” » 0-506 6°09 122 »” ” ” ” 29 ” 0-494 5:94 162 a ne Pyrus malus, fruit, . 1-571 | 18-90 ” ” ” ” 1-653 | 19-88} 168 ” » ”? ” : 1-473 | 17°72 165 i i Vitis vinifera, fruit, 2:131 | 25:64) 127 ” ” ” ” ” ° . 2-306 27-73 125 sh cs Rheum officinale, red stems, 0-542 | 6:52 83 ” ” ” ” » . 5 0-542 6°52 76 Mar. 1910. | Apium graveolens, white bases of leaves, 0-944 | 11°36] 125 ah os Allium Cepa, bulb, x , 0-663} 7:97] 146 Aa oe Beta vulgaris, root, 1-082 | 13-01] 214 ” ” ” ” ” 1-090 | 18-11 215 ” ” ” 2» ” : . 1-041 | 12°52) 202 ing a Brassica Rapa, white, root, 0-608 | 7:31} 110 iene » » » | kept,* 0-925 | 11:92] 101 > ” ” ” ” 0:856 10-30 118 a Pe ( an », Swede, root, 0-890 | 10:70] 1381 ” oe ” ” kept,* 1-028 | 12-36 153 ” ” ”? ” 1-009 12°14 146 ss 5 Helianthus tuberosus, tuber, 1:153 | 13°87] 371 ” ” ” ” 9 1-062 12°76 394 ” ” ” »” ” . 1-653 18°67 370 Be 9 Solanum tuberosum, tuber, 0-544 | 6:54) 148 2) » ” oe) ” . 0-552 6°64 149 9 Dp ” ” », kept,* 0-612 | 7:36] 148 > ” ” ” ” ; 0-586 7:05 140 April, 1910. op ” ” 0°538 | 6:47) 159 ” ” ” ” ” 0-561 6°75 142 * For details, see Discussion of Results. Arxins—Osmotie Pressures of some Plant Organs. 469 The above results seem to show that while large variations are met with in the osmotic pressures of various fruits and of the same fruit at various stages, tubers, on the contrary, vary but little in the same species, though there is a considerable range of variation between various species. This result was not unexpected ; for tubers exist under very uniform conditions in nature, and usually possess a large store of starch or other reserve material capable of being mobilized and of thus restoring equilibrium. In this they differ from the leaves, which are organs for elaborating compounds with little or no store-material. The latter are consequently very liable to fluctuation in pressure, as shown in a former paper. (Dixon and Atkins, loc. cit.) The determination of the mean molecular weights frequently throws light on the metabolism, a fall pointing to formation of starch or some other insoluble material or possibly to exhaustion due to respiration. A rise, on the other hand, indicates the conversion of colloids into soluble starches, mono- and di-saccharides, or synthesis of some sugars by assimilation. Conclusions. 1. As the result of over fifty measurements, the deduction may be made that similar plant organs of the same species have approximately equal osmotic pressures. (‘This does not apply to leaves, for which see Dixon and Atkins, loc. cit.) 2. The osmotic pressure of the solutes in the fruits examined varied from about 6 to 30 atm., and the mean mol. wt. from 110 to 231. 3. The pressure in the underground organs studied ranged from 6:5 to 18-7 atm., while the molecular weight of the solutes ranged from 101 to 394. 4. The red stem of Rheum officinale gave the lowest recorded mean molecular weight, viz. 76, with an osmotic pressure of 6°52 atm. 5. The tuber of Helianthus tuberosus gave the highest mean molecular weight found, viz. 394. This paper is properly a continuation of a joint paper with Prof. H. H. Dixon; and I have much pleasure in thanking him for his advice throughout. ad tte eh THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.8.), No. 39. JULY, 1910. THE SEPARATE INHERITANCE OF QUANTITY AND QUALITY IN COWS’ MILK. BY JAMES WILSON, M.A., B.8Sc., PROFESSOR OF AGRICULTURE IN THE ROYAL 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 NORGATEH, ame 14, HENRIETTA STREET, COVENT GARDEN, LONDON, wh. ion Insti 1910. Price Sixpence. Roval Dublin Society. ADAIR RAR ARR AR FOUNDED, A.D. 1731. INCORPORATED, 1749 EOS EVENING SCIENTIFIC MEETINGS. Tur Scientific Meetings of the Society are held alternately at 4.30 pm. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Hditor. Arkins— Osmotic Pressures of some Plant Organs. 469 The above results seem to show that while large variations are met with in the osmotic pressures of various fruits and of the same fruit at various stages, tubers, on the contrary, vary but little in the same species, though there is a considerable range of variation between various species. ‘lhis result was not unexpected ; for tubers exist under very uniform conditions in nature, and usually possess a large store of starch or other reserve material capable of being mobilized and of thus restoring equilibrium. In this they differ from the leaves, which are organs for elaborating compounds with little or no store-material. The latter are consequently very liable to fluctuation in pressure, as shown in a former paper. (Dixon and Atkins, loc. cit.) The determination of the mean molecular weights frequently throws light on the metabolism, a fall pointing to formation of starch or some other insoluble material or possibly to exhaustion due to respiration. A rise, on the other hand, indicates the conversion of colloids into soluble starches, mono- and di-saccharides, or synthesis of some sugars by assimilation. Conclusions. 1. As the result of over fifty measurements, the deduction may be made that similar plant organs of the same species have approximately equal osmotic pressures. (This does not apply to leaves, for which see Dixon and Atkins, loc. cit.) 2. The osmotic pressure of the solutes in the fruits examined varied from about 6 to 30 atm., and the mean mol. wt. from 110 to 281. 3. The pressure in the underground organs studied ranged from 6:5 to 18-7 atm., while the molecular weight of the solutes ranged from 101 to 394. 4. The red stem of Rheum officinale gave the lowest recorded mean molecular weight, viz. 76, with an osmotic pressure of 6°52 atm. 5. The tuber of Helianthus tuberosus gave the highest mean molecular weight found, viz. 394. This paper is properly a continuation of a joint paper with Prof. H. H. Dixon; and I have much pleasure in thanking him for his advice throughout. SCIENT. PROC, R.D.S., VOL. XII., NO. XXXV. 38uU fae XXXY. THE SEPARATE INHERITANCE OF QUANTITY AND QUALITY IN COWS’ MILK. By JAMES WILSON, M.A., B.Sc. Professor of Agriculture in the Royal College of Science, Dublin. [Read May 24. Ordered for Publication Junz 14. Published Juny 28, 1910.] It is a very general opinion that the milk of high-yielding cows is usually poorer and that of low-yielding cows richer in quality. ‘The origin of this opinion lies, perhaps, in the fact that some breeds, like the Dutch, that are at the top of the scale with regard to quantity, are nearer the bottom with regard to ‘quality; while others, like the Jersey, that are at the top with regard to quality, are lower down with regard to quantity. No one denies that there are many exceptions to the rule, but these exceptions are generally regarded as departures in some degree from the normal. At the same time it would be difficult to find any cautious writer vehemently dissenting from, or assenting to, the general opinion, for the reason that, till recently, no large body of data bearing upon the question had been brought together. Now, however, in the “ Report of Milk Records for Season 1908,” recently published by the Ayrshire Cattle Milk Records Committee, we are furnished with data from which it is possible to obtain a ruling. The systematic milk-testing scheme for Ayrshire cattle, the inception and success of which are almost entirely due to the late Mr. John Speir, Kt. St. O., of Newton, near Glasgow, was inaugurated in 1903; and 1342 cows were tested in that year. The records of 81382 cows are contained in the report for 1908. ‘lhe scheme is worked as follows :—The Ayrshire cattle country is divided into districts each containing either about ten or eleven or about seventeen or eighteen farms coming under the scheme. ‘I’o each district is appointed a trained official whose duty it is to visit all his farms in rotation. If there are eleven farms in his district, he can visit each of them once a fortnight ; if there are eighteen, he can visit each once every three weeks. The official reaches a farm in the afternoon, stays over night, and leaves next Witso n—Inheritance of Quantity and Quality in Cows’ Milk. 471 forenoon. Thus, he is present while the cows are milked in the evening and in the morning. He weighs the milk of each cow and takes a sample evening and morning. The weights are recorded and the samples sent on to an analyst, who tests them for the amount of fat they contain. Thus, means for deter- mining every cow’s yield and the quality of her milk are ascertained either every fourteen or every twenty-one days; and when the means are all collected, the total can be worked out for the year or for such time as the work is continued. It cannot be maintained, of course, that this method is absolutely accurate. More accurate results would have been obtained if the observations had been taken more frequently—most of all if they had been taken daily—but they are sufficiently accurate for the purpose of the Ayrshire cattle-breeders ; and when it is considered that, given a sufficient number of cases, the errors below the line are cancelled by those above, the results are sufficient for our purpose also. By grouping the cows according to the quality of their milk and eliminating all that have been observed for less than thirty weeks and which, in consequence, might not show the true approximate average, we find that the bulk of Ayrshire cows’ milk contains from 3°3 to 4 per cent. of fat, but that there are many cows with milk both above and below these limits: some running over 5 and others below 3 per cent. It is from this fact, which is expressed graphically in diagram 1 (p. 472), together with the fact that the quantity of Ayrshire cows’ milk varies in a similar manner, that we can infer that the yield and quality of cows’ milk are separately inherited. If we group together all the low-yielding cows, and find their milk invariably high in quality, we may infer that low yield and high quality are of the nature of concomitant variations. If we group the high-yielding cows together, and find their milk invariably of low quality, we may infer that high yield and low quality run together. But if we take these groups and any other groups we can form, and find that the quality varies the same way in them all—that is that there are low qualities, high qualities, and medium qualities in every one of them—then we are justified in inferring that the quantity and quality of the milk are independent of each other. And this is what we do find. But, to get a fair view of the case, we must eliminate some other cows in addition to those that were eliminated to form our first diagram. In the Ayrshire cattle country there are two main systems of dairying : one in which milk is produced throughout the year, and the other in which it is produced from early spring till late autumn chiefly for cheese-making. 3u 2 472 Scientifie Proceedings, Royal Dublin Society. Dracram 1. Diagram showing the numbers of Ayrshire cows giving milk of ditferent qualities. NUMBER OF COWS 600 NK 600 2) 580 580 560 560 C) S40 oy 540 520 520 500 S00 480 480 460 460 ® Lo) 440) 440 420 420 400 400 ° 380 ; 380 ¢ 360 360 340 340 320 320 300 300 + 280 280 260 260 240 240 220 220 200 200 Nn 180 Es 180 160 160 140 {40 120 o 120 t 100 100 80 BO 60 y 60 t © ot 40 ¢ 40 20 20 ° Lele PERCENTAGEOF)O ROMO -AMNPT HON DHAOTVNTNHONROAROTN FAT IN THE MILKS i] + a Witson—Inheritance of Quantity and Quality in Cows’ Milk. 473 In the former case, cows are calving at all times, while, in the latter the majority calve early in spring and are dry during the dead of winter ; and the records are kept in each case to meet the system of dairying: that is throughout the year in the one and during the producing season in the other. Under the whole-year system, there are cows whose full lactation periods do not fall in any one year, while, under the cheese-making system, the records of late-calving cows are kept for only a portion of their lactation period. At the same time, under both systems, there are many cows too young to have attained what might be called their normal yield; while, under the whole- year system especially, records are kept of cows that have milked far beyond the normal nine or ten months. To eliminate all such cases as far as possible, we make use only of the records of cows aged four years and more that have been tested for not less than thirty and not more than forty weeks. The cows whose records we can use may be separated into four divisions of not very unequal numbers—viz., those yielding less than 500 gallons, those from 500 to 600, those from 600 to 700, and those yielding over 700 gallons. They could have been separated into more divisions than four; but, on consideration, this was found unnecessary. Among the lowest-yielding cows, many give well below the 500, and, among the higher-yielding cows, a con- siderable number give over 1000 gallons. But, in working through the figures, it was seen that, if the very lowest- and the very highest-yielding cows were given separate divisions to themselves, it would have made no material difference to the general result. The results of this separation are expressed in diagrams 2, 3, 4, and 5. These show that the quality of the milk given by any one of the four divisions is almost the same as that given by the other three. The curious drop in the number of cows under the 500-gallon standard giving milk containing 3:6 per cent. of fat may be taken as of no vital importance. It is a phenomenon not strange in statistics. [| Diagrams. 474 Scientific Proceedings, Royal Dublin Society. Diacram 2. Diagram showing the qualities of the milk from cows giving less than 500 gallons. NUMBER OF COWS oO bi) 70 70 60 50 40 50 PERCENTAGE OF FAT IN THE MILK DiaGRAmM 3. Diagram showing the qualities of the milk from cows giving from 500 to 600 gallons. NUMBER OF cows 120 120 lo 110 100 100 90 90 80° 80 70 70 60 60 50 | 50 40 40 30 30 20 20 10 16 PERCENTAGE : L = OF FAT IN OF DMOSTAMYEHOHR DDOT VHF HOHRDDO = THE MILK ov ” A ee eae : xe) 476 Scientific Proceedings, Royal Dublin Society. Diacram 4. Diagram showing the qualities of the milk from cows giving from 600 to 700 gallons. NUMBER OF COWS 110 100 90 80 70- 60 50 40 30 20 PERCENTAGE OF FAT IN THE MILK. OND Va 3:6 EKZ/ 110 100 90 80 70 60 50 40 50 20 Witson—Inheritance of Quantity and Quality in Cows’ Milk. 477 DiaGRam 5. Diagram showing the qualities of the milk from cows giving over 700 gallons. NUMBER OF SOE —— = = Ln?) 70 a) 70 m 0 i) “i) 60 : 60 ¢ ") 50 40 30 ™ ™» wy 2) 20 2) 10 ; —-ava”m PERCENTAGE OF) © © © FAT INTHE MILK | Os N) The similarity between the quality of the milk given by the four divisions of cows can be made still clearer if we reduce the figures represented by the last four diagrams to percentages, and then combine the results obtained in this way in a diagram in which the four divisions are represented on the same basis. [Tasie, SOIENT. PROC. R.D.S., VOL. XI., NO. XXXV. 3x 478 Scientific Proceedings, Royal Dublin Society. TABLE SHOWING THE NUMBERS OF COWS GIVING DIFFERENT YIELDS AND THE COMPOSITION OF THEIR MILK. 2, apis |e pg Sapte "stn aas|afere | teee of fat in junder 500 over 700] Seems | yielding |, ste ons yen the milk,| gallons. | £20) | anone, | alons. || “eitone. [600 gatlone-(700gallons} gallons, gallons. | gallo: gallons gallons./700gallons,) g 2:5 2-6 1 1 17 12 2-7 2:8 4 4 3 “68 “42 37 2:9 10 7 4 1 71 “TA +49 “19 3-0 7 6 13 8 1:20 “63 1:60 152 $1 16 21 17 14 2°73 2-22 2°10 2-67 3-2 17 41 35 31 2°91 4:33 4:32 5:90 3-3 26 63 63 38 4-44 6°65 7:79 7:24 3-4 56 99 64 52 9°58 10°44 791 9:90 3-5 66 103 103 59 11-28 10°86 12°73 11-24 3-6 52 107 106 66 8°89 11:29 13-10 12°57 3-7 67 125 99 70 11:45 13-19 1224 1333 3-8 69 105 86 61 ||, 11-79 11:08 10:63 11-62 3-9 61 89 62 43 | 10°43 9°38 7:66 8-19 40 | 36 67 54 35 6:15 7:07 6:67 6°66 4-1 30 61 45 19 5:13 5:38 5:56 3°62 4-2 23 25 27 iil eee 2°64 3:34 2:86 4:3 16 18 10 9 || a7 1:90 1-23 L-71 4-4 9 4 5 2 1:54 “42 “62 °38 4-5 6 3 2 | 1:02 39 24 | 4-6 7 7 7 1 aen@ 74 86 +19 4% 3 1 2 “51 +10 24 “19 48 1 1 17 “10 4-9 1 1 10 12 5-0 1 “17 5-1 1 VG 585 94s | 809 | 625 Witson—Inheritance of Quantity and Quality in Cows’ Milk. 479 Dracram 6, Diagram showing the percentage numbers of cows giving different yields and the qualities of their milk. 7 under 500 500 - 600 3 Va 600-700 26 27° 28 29 SO#ST 32 33° 34 SS°36 37 38 539 GO 4? 42 43 44 AS 46 47 48 49 $0 $I 52 Percentage of fat in milk’ It will then be seen how substantial is the agreement among all four divisions. It might be objected that the line representing the cows giving 700 gallons bends a little more to the left than the others, both in rising and descending; but this does not substantially modify the general result. Indeed it may lend emphasis to it, since there is a tendency among farmers to retain good-yielding cows, even if their milk be poor, and thus to increase the proportion of poorer milk among them. ze ¥ Te Seer ee a ar aires y | K uh i { , ie un - t . \ NPC NO Syn arse ae hing a ele havea c OTS ieasildl eatrepermy oie Oh EU Oa " 7 F Esta hs 8 ’ yee we 2a) Suh i 4 o WE CR IMT TS: a LY Thttober 3 ; URE eS Ba SS AO HES THE SCIENTIFIC PROCEEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.8.), No. 36. SEPTEMBER, 1910. MECHANICAL STRESS AND MAGNETISATION OF IRON. Part III. BY WILLIAM BROWN, B.Sc., PROFESSOR OF APPLIED PHYSICS, ROYAL COLLEGE OF SCIENCE FOR IRELAND. [Authors alone are responsible for all opinions expressed in their Communications. | DUBLIN: PUBLISHED BY THE ROYAL DUBLIN SOCIETY, LEINSTER HOUSE, DUBLIN. WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1910. sahsonian dnszis ~ RS ig nay (3s NOV 26 1910 “anal! Neuse Price One Shilling. Roval Dublin Society. een ee FOUNDED, A.D. 1731. INCORPORATED, 1749 EVENING SCIENTIFIC MEETINGS. Tur Scientific Meetings of the Society are held alternately at 4.30 p.m. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested +o forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be get down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Editor. Witson—Inheritance of Quantity and Quality in Cows? Milk. 479 Diacram 6. Diagram showing the percentage numbers of cows giving different yields and the qualities of their milk. under 500 500 - 600 600-700 26 27 28 29 30 351 32 33 34 55 56 37 35 59 9O 41 42 43 44 45 46 47 48 49 50 ST $2 Percentage of fat in milk It will then be seen how substantial is the agreement among all four divisions. It might be objected that the line representing the cows giving 700 gallons bends a little more to the left than the others, both in rising and descending; but this does not substantially modify the general result. Indeed it may lend emphasis to it, since there is a tendency among farmers to retain good-yielding cows, even if their milk be poor, and thus to increase the proportion of poorer milk among them. SCIENT. PROC. R.D.S,, VOL. XII., NO. XXXV. 3Y [ 40 J XXXVI. MECHANICAL STRESS AND MAGNETISATION OF IRON: Part III. BY WILLIAM BROWN, B.8c., Professor of Applied Physics, Royal College of Science for Ireland. [Read May 24, Ordered for Publication June 14. Published Sepremper 2, 1910.] A CONSIDERABLE amount of work has been done on various effects of mechanical stress on the magnetisation of metals, and the report on the subject by H. Nagaoka! tabulates all the data known up to 1899. Amongst the subsequent work the following bear on the subject of the present paper :—Nagaoka and Honda’ have shown that when a wire of nickel- iron alloy containing nickel up to 45 per cent. is subjected to simultaneous longitudinal and circular magnetic fields, the direction of twist of the free end is the same as that of an iron wire. Shimizu and Tanakadate*® have studied the Wiedemann effect at high temperatures, and have shown that for iron, nickel, and tungsten steels, it vanishes at the critical temperatures of the metals. In the first and second parts of this paper the writer has published‘ the results of some experiments obtained with soft iron wires, in which the longitudinal magnetism, the circular magnetism, the longitudinal load, and the cross-sectional area of the wires were varied. The present results were obtained with iron wires of different tempers, or different degrees of magnetic softness, different lengths, and different diameters. Five different degrees of hardness were adopted, which were obtained by heating the wires to a cherry-red heat by means of a broad Bunsen burner, when they were hung vertically, and subjected to different longitudinal loads. In the experiments for testing the effect of temper on the torsion and magnetism, No. 16 size wires of the best Swedish charcoal-iron were employed, this size of wire being easier to manipulate than larger or smaller wires. In order, in the first place, to get the wire as soft as possible, it was suspended from the ceiling of a darkened room by means of a three-jaw self- 1 Rapports du Congrés International de Physique, Paris, 1900, vol. i1., pp. 536-556. 2 Compt. Rend., 1902, tom. cxxxiy.; and Phil. Mag., 1902, 6th Ser., vol. iv. 3 Proc. Phys. Math. Society, Tokyo, Oct., 1906. *Scient. Proc. Roy. Dub. Soc., 1909, vol. xii., pp. 101 and 175. Brown—WMechanical Stress and Magnetisation of Iron. 481 centring clutch, and allowed to hang loosely under its own weight only. It was then raised to a cherry-red heat by heating from the top downwards, so that the hot air ascending tended to anneal the part just heated, the degree of hardness thus obtained being here distinguished by the symbol Hy. The next degree of hardness was obtained by the same process, but with a weight on the lower end of the wire amounting to 10° grammes per sq. cm.; this hardness we call H,, and the other degrees were obtained by hanging on weights of 2 x 10°, 2:5 x 10°, and 3 x 10° grammes per sq. cm. respectively, and their hardnesses are distinguished by the symbols H,, H:.; and Hs. The first heating of the wire did not change its diameter, which was in every case carefully measured when the oxide had been all cleared off by means of emery paper; new wires of full No. 16 size were taken for each of the other tempers, so as to get all the wires of different degrees of magnetic softness, and very approximately of the same diameter. To get a.measure of the hardness of the wire, two tests were applied, first the simple rigidity of each was measured; and, secondly, the electrical conductivity of each was determined by comparison with a standard of electrolytic copper wire whose conductivity was 101:01, Matthiessen’s standard being 100. The simple rigidity was measured by a statical method in which the horizontal form of Searle’s torsion apparatus was employed. This apparatus consists essentially of a spindle, mounted on ball- bearings, with a self-centring three-jaw clutch at one end of the spindle, and a wheel at the other end, by means of which the required couple can be applied. One end of the wire under test is caught in the self-centring clutch, and the other end fixed in a firm support at a suitable distance (in this case 403 cm.) from the clutch. In the apparatus used for these experiments, the end of the spindle at the wheel was prolonged about 5 cm., and a plane mirror fixed on it with its reflecting surface in the same plane as the axis of the spindle; and the twist of the wire under test was observed in the usual way by means of a vertical seale and telescope with cross-hairs, the scale being 64°5 cm. from the mirror. In observing the deflection on the scale, or the twist of the end of the wire, the maximum of which was about 6°, four or more different weights were used, and a double reading taken in each case by hanging the weight from either side of the rim of the wheel. The values of the scale-readings were then plotted against the corresponding values of the weights, and a straight line drawn from the origin through the points; the mean values of the weight and twist were picked off for calculating the rigidity. The scale used was divided into millimetres, and could be easily read to 0-2 mm., so that an error in the scale-reading of 0°2 mm. made an error in the values By 2 482 Scientific Proceedings, Royal Dublin Society. obtained for the rigidity of about 0-2 per cent.; and in measuring the diameter of the wire a standard micrometer screw-gauge was used, in which an error in reading of 545 mm. made an error of 0°3 per cent. in the value of the rigidity, so that the values of the rigidity obtained may be reckoned correct to within the half of one per cent.; the same error in estimating the diameter of the wire would cause an error of about the half of one per cent. in the value of the electrical conductivity ; both methods were therefore used as a check on one another. The results are collected in Table I., and shown as curves in fig. 1 The rigidity is expressed in grammes per square centimetre, and the electrical conductivity as a percentage of pure copper. Tasie I. Cross-sectional Rigidity Electrical area of the f x 10° conductivity wire Hest grammes per | pure copper, x 10-8 sq. em. sq. cm. 100. 20°80 Ho 774 14:3 20°73 Hy 788 14:0 20°64 He 803 13°6 20:46 Ha.5 812 12°5 20°35 Hs 823 gpl Rigidity x 106 Conductivity Fic. 1.—Load on wire (grammes per sq. cm. x 10°) when it was heated. Brown—WMechanical Stress and Magnetisation of Iron. 488 In plotting the results of these experiments on rigidity and electrical conductivity, an arbitrary scale of hardness has been taken on the axis of abscissee, that is, the load in grammes per sq. em. which was on the end of the wire when it was heated has been taken as a measure of the hardness. Against this, on the axis of ordinates the simple rigidity has been plotted in the one case, and in the other case the percentage electrical conductivity compared with pure copper. The conductivity curve here given is similar to those obtained by Strouhal and Barus! in their work on the specific resistance and temper of steel wire. In order to observe the amount of twist on the end of the wire when it is subjected to different longitudinal loads, and when placed in different longitudinal magnetic fields—the maximum current sent through the wire in all cases was at the rate of 100 amperes per sq. em.—the same long solenoid with accessories was used as was employed in previous work.? ‘The wire under test was fixed so as to hang vertically, and as near as possible in the middle of the solenoid, whose total length was 237 em.; the distance from the mirror on the vibrator at the lower end of the wire to the millimetre scale was 116°5 cm., and for every millimetre of deflection on the scale, the end of the wire was twisted through an angle of 88 seconds. With a certain load on the end of the wire and a given longitudinal magnetic field round it, an electric current was sent through the wire, and its value slowly increased up to the maximum of 100 amps. per sq. em.—so as to give a constant circular magnetism—and the steady deflection read off the millimetre scale; the current was then gradually diminished to zero, reversed, and again slowly increased to the maximum, and the scale-reading again observed on the other side of the zero; the mean of these two readings was then taken as the true value of the twist for that magnetic field. The longitudinal load on the wire being kept the same, a similar process was gone through for twelve different magnetic fields up to a maximum of 14 ¢c.g.s. units. The longitudinal load on the wire was now increased, and a series of observations were made, similar to what was done for the first load, and so on for five different values of the load. Before doing these, however, the wire, which had been strongly magnetised during the previous experiments, had to be carefully de-magnetised each time; this was a somewhat troublesome operation, but was overcome by means of a reversing key in the circuit of the solenoid; by continually reversing the diminishing current through the solenoid, and decreasing the load on the end of the wire and the 1 Wied. Ann., 1883, vol. xx., p. 621. 2 Scient. Proc. Roy. Dub. Soc., 1909, vol. xii., pp. 115 and 183. 484 Scientific Proceedings, Royal Dublin Society. current through it, the wire could be de-magnetised as far as the vertical component of the Harth’s magnetism would permit. This was tested by putting a certain current round the solenoid in such a way and of such a value as to annul the vertical component of the Harth’s magnetism inside it; and then a current sent through the wire gave little or no twist if the wire was quite de-magnetised. The results for five different longitudinal loads on a No. 16 soft-iron wire of cross-sectional area 20°80 x 107° sq. cm., and simple rigidity 774 x 10° grammes per sq. cm., and electrical conductivity 14°31, are shown in 3 Table IL., and as curves in fig. 2. Tasin II. oneitadinal Twist or deflection on the scale in mm. when the wire is loaded with 8 = grammes per sq. cm. x 10° :— magnetic field Eh 05 1-0 1-5 2 3 0-45 18:5 17 16 15 12°5 0°75 30 22 18-5 16°5 14°5 1:0 35 29 24:5 20 17°5 1°5 39 36 33°5 29°5 24 2-0 40:2 387°5 35:5 33-5 29 2°5 40°65 38 36 34 29-5 3 40 37-2 35:2 33°5 29 4 38:2 35°5 33:5 82 27°65 6 34 31:8 30 28°5 24-5 9 28 26:4 25 23°5 20°5 12 23 21:5 20 19 17 14 19°5 18 17 16 14:5 Area of cyclic curve in 22°5 20°8 18°8 17:2 13:8 "sq. em. Twist in scale divs. (mm.) oO oF Ua) 19 Fre. 2.—Magnetic Field. Brown—WMechanical Stress and Magnetisation of Iron. 485 From the curves we see that the maximum twist or deflection on the seale occurs for all loads when the wire is in a longitudinal magnetic field of 2-5 units. It for this field we plot the values of the load as abscissaa and the values of the corresponding maximum deflections as ordinates, the points will be found to lie very approximately in a straight line, and show that when the load on the wire is éncreased six times the twist or deflection on the scale is decreased 27 per cent. By comparing the twist obtained with this wire when the load was 10° grammes per sq. cm. and when in a longitudinal magnetic field of 2.5 o.g.s. units with the twist obtained with another No. 16 iron wire with the same load and in the same magnetic field,’ we get from Table I., above, a deflection of 38 mm., and in the former case the deflection when reduced to the present conditions becomes 36 mm. This difference of 2 mm. is no doubt due to (1), the former wire being annealed by means of a Bunsen flame when it was suspended in a horizontal direction, and the present wire being annealed by the flame when it was hanging vertically and heated from the top downwards, (2) the cross-sectional area of the former wire being 20°6 x 10° sq. cm., and that of the latter being 20°8 x 10% sq. cm., therefore the maximum current through the two wires was 2°06 and 2°08 amperes respectively. This straight-line relation between the load and twist holds very approximately for other longitudinal magnetic fields higher than 2°5 c.g.s. units, and would no doubt also hold for lower fields, if sufficient points had been determined on the rising parts of the curves. ‘The magnetic field of 2°5 units was chosen because the highest deflections occur with this field. In order to test how the circular magnetism changed with the increased longitudinal load on the wire, the longitudinal magnetic field round the wire was kept constant at 2°5 c.g.s. units, and a complete cyclic curve taken when the wire was stretched by each of the five different loads. An electric current was sent through the wire which was increased by steps up to a maximum of 2:08 amperes, then decreased to zero, reversed and increased to a negative maximum, then back once more to the positive maximum, the twist on the end of the wire or deflection on the scale being read off at each step of the cycle. A complete cycle was taken in this way when each load was on the wire, and the results plotted on millimetre paper where on the axis of abscissee two centimetres represented one ampere and on the axis of ordinates one centi- metre represented ten divisions on the scale; the total area of each cyclic curve was then measured in sq. cm., and the results are shown in the last line of 1 Scient. Proc. Roy. Dub. Soc., 1909, vol. xii., p. 186. 486 Scientific Proceedings, Royal Dublin Society. Table II. If we plot as abscissee the values of the load on the wire, and as ordinates the corresponding values of the areas of the cyclic curves, the points will be found to lie in a straight line and show that when the load is increased six times the circular magnetism or area of the cyclic curve is decreased about 38 per cent. These results confirm those obtained previously with a wire which was tested under the same conditions of load and magnetic field. In the former test the area of the cyclic curve when the wire was loaded with 10° grammes per sq. cm. was 30 sq. cm., which value when reduced from the scale and distance used then to the scale and distance used now becomes 20°1 sq. cm. as against 20°8 sq. cm. for the new wire, which was both a little thicker and softer. The above results were obtained when the wire was the full length of 226 com., and in a uniform longitudinal magnetic field throughout its entire length ; tests were now made when the wire was 0°8 times and 0:5 times this length respectively, so as to find out how the twist and circular magnetism varied when the length of the wire was altered to 181 em. and 113 cm. The top suspension for the wire was made longer to suit the lengths required, the wire still remaining in a uniform longitudinal magnetic field throughout its whole length, and each length was put through the same series of tests as was done with the full-length wire. That is, the maximum twist was observed for the given current density in the wire of 100 amperes per sq. cm.; when the wire was placed in longitudinal magnetic fields of twelve different strengths, and also when subjected to three different values of longitudinal load, instead of the five loads used with the long wire. As in the previous case when the wire was of the full length, the greatest twist occurred when it was in a magnetic field of 2°5 units, and at this point a complete cycle was taken in each case. The results obtained are shown in Table III., and one set of the results for each of the three lengths when the load on the wire was 0°5 x 10° grammes per sq. cm. are exhibited as curves in fig. 3; the results for the other two loads when plotted show the same general contour in the shape of the curves with the maximum gradually decreasing. Brown—WMechanical Stress and Magnetisation of Iron. 487 Tasie III. x sos ; c . ; 2 ao is eS Twist or deflection on the scale in mm. when subjected to the following oa a En x longitudinal magnetic fields :— Ba OE aS 2 2 2 ot | is Be leis ae 2 S$ B¢e| 045 | 0-75 | 10 | 15 | 20 | 25 3 4 6 9 12 14 |228 S| jee 4 0-5 | 185 | 30 35 39 40:2 | 40°65 | 40 38°2 | 34 28 23 19°56 | £2°5 226) 1:5 | 16 18°5 | 24°5 | 33-5 | 35°5 | 36 35°2 | 33°5 | 30 25 20 17 18-8 3°0 | 12°56 | 14°5 | 17-5 | 24 29 29°5 | 29 27-5 | 24°5 | 20°5 | 17 15 13°8 05 | 14 20 25 30 32 32°5 | 31°5 | 29°5 | 26 215 | 17-5 | 16 17-9 | 181 | 1°5 | 12 14 17 24°5 | 28°5 | 29 28°5 | 27 24 19°5 | 15:5 | 13 15°4 3:0 9 11 13 19 23 23°5 | 23 22 19°5 | 16 13 115 | 114 0-5 8 12 16 19 20 20°5 | 20 19 17 14 115 975) 11-6) )) 113 15 7 8°5 | 10 14 17 18 17-5 | 16°5 | 14°65 | 11:5 9 75 | 10:0 | S | 3:0 | 6 C2 13 14-8 | 16 14:8 | 14°56 | 12°5 | 105 | 8 7 70 | Twist in scale divs. (mm.) Fre. 3.—Magnetic field. If for the three different loads we plot the values of the lengths of the wires as absecissee, and as ordinates the corresponding values of the maximum twist when the wire was in a longitudinal magnetic field of 2°5 units, the points will be found to lie very approximately in a straight line in each of the three cases, and each of these straight lines when produced will pass through the origin. These results show that the twist of the free end of the wire is directly proportional to the Jength of the SCIENT. PROC. R.D.S., VOL. XII., NO. XXXVI. 3 Zz 4887) Scientific Proceedings, Royal Dublin Society. wire; and when the length of the wire is halved, the mean maximum twist is decreased about 50 per cent. in each case when the three loads are used. This straight-line relation between the length of the wire and the twist also holds very approximately for fields higher than 2:5 units, showing that the curves are practically parallel. In the same way by plotting the values of the lengths of the wires as abscissee against the corresponding values of the areas of the cyclic curves as ordinates, when the wires were subjected to the three different longitudinal loads, the points will be found to lie in a straight line in the three cases, and these lines also when produced will pass through the origin and show that when the length of the wire is halved the circular magnetism when the wire is in a magnetic field of 2°5 units is decreased about 48 per cent. Again, if for the three different lengths of wire we plot the values of the load as abscissee, and as ordinates the corresponding values of the circular magnetism or areas of the cyclic curves in a magnetic field of 2°5 units, the — points will be found to lie in a straight line in each of the three cases, and show that when the load is inereased six times the circular magnetism is decreased about 35 per cent. Also for the three different lengths of wire by plotting the values of the load on the wire as abscisse, and as ordinates the corresponding values of the maximum twist when the wire is in a magnetic field of 2°5 units, the points will be found to lie in a straight line in each of the three cases, and show that when the load on the wire is increased six times the mean maximum twist is decreased about 27 per cent. These three latter straight lines, when produced, cut the axis of abscissee at a mean point representing aload of 8:2 x 10° grammes per sq. cm., which is very approximately the elastic limit of the wire, and shows that with the given current density in the wire and in the given magnetic field of 2°5 units if the wire be stretched near to its elastic limit, there will be little or no twist on the free end. ‘This confirms an observation made in previous work,’ where it was shown that no transient current would be produced by twisting a wire placed in a longitudinal magnetic field if the load on the end of the wire was such as to stretch it to, or nearly to, the elastic limit. With the arrangement of apparatus used in these experiments it was not possible to put such a large weight on the end of the wire; it was reckoned, however, that a steel wire of the same diameter, with a very small weight on the end to keep it straight, would be at least as hard as the iron wire when loaded near to its elastic limit. A steel wire of rigidity 830 x 10° grammes per sq. cm. and electrical conductivity 9:6, ' Scient. Proc. Roy. Dub. Soc., 1909, vol. xi, p. 110. Brown—WMechanical Stress and Magnetisation of Iron. 489 was thereupon taken and placed in various longitudinal magnetic fields inside the solenoid, and there was no twist whatever of the free end of the wire when a large or small current was sent through it. Another wire was now taken and prepared to give the degree of hardness H,, as already described above; the rigidity was measured and found to be 788 x 10° grammes per sq. cm., and the electrical conductivity 14:0 as compared with pure copper. The cross-sectional area of the wire was 20:73 10% sq. cm., and it was tested in precisely the same manner as was done with the soft iron wire of hardness H,—that is, for twelve different strengths of longitudinal magnetic fields, for three different lengths, and for three different longitudinal loads. The results are here shown in Table IV. TABLE IV. x ee c - soo S Pes Twist or deflection on the scale in mm. when subjected to the following 2 5g p 2 Xx longitudinal magnetic fields :— Sg Sj|oag Ra Be le 28 | a8 & 8 B5| 045 | 0-75 | 1:0 | 15 | 20 | 25 8 4 6 9 12 14 & 3 cm. 3°0 7-5 | 10 12 16°5 | 19°5 | 20 19°8 | 19 17:5 | 15 12°5 | 11:5 0°5 6°5 9 12°5 | 16 17:5 | 18:5 | 18:0 | 16°5 | 15 12°5 | 10°5 9 1138 15 6 (a) 9 12°5 | 15-5 | 16°56 | 16 15 13 1075 8°5 7 3:0 5 6 7 9 12 12°56 | 12:2 | 12 10°5 9 75 6°5 As before, if for the three different loads we plot as abscisse the lengths of the wires, and as ordinates (1) the maximum twist, and (2) the areas of the cyclic-curves when the dongitudinal magnetic field round the wire is 2:5 units, the three straight lines in both cases (1) and (2) when produced all meet at the origin. This shows that when the length of the wire is halved, both the maximum twist and circular magnetism or area of the cycle- curves are decreased about 50 per cent. Also, if for the three different lengths of wire we plot the values of the loads on the wires as abscissa, and the corresponding values of the maximum twist as ordinates, it will be found 322 0-5 | 18 28°5 | 32°5 | 35°5 | 37 87°5 | 86°5 | 34°5 | 30°65 | 25 20°56 | 17°5 | 18°5 490 Scientifie Proceedings, Royal Dublin Society. that when the load is increased six times, the mean maximum twist is decreased about 82 per cent. These three straight lines when produced cut the axis of abscissee at a point representing a mean load of 8:25 10° grammes per sq. em., which is about the same value as that found for the softer wire. Another wire was now taken and treated in the same way as in the previous cases, but with a load on the end of 2x 10° grammes per sq. cm., and was made into the state of hardness called H,. The wire was prepared in the usual way, the rigidity measured and found to be 803 x 10° grammes per sq. cm., and the electrical conductivity 13°6 compared with pure copper. The cross-sectional area of the wire was 20:64 107* sq. cm., and it was tested in precisely the same manner as the previous wires, and the results are shown in Table Y. TABLE V. a SSS 5 2 5 2 . [aot iS Boo Twist or deflection on the scale in mm. when subjected to the following fo 2 ag ie 2 Xx longitudinal magnetic fields : — Sa a> |S8¢ Se tA jo BO i ais 5 g Sh 5/045 | 075 | 1:0 | 15 | 20 | 2-5 3 4 6 9 12 14/236 4H I | aos 0°56 | 16 25 30 33°5 | 34 34°2 | 34 31°8 | 28 23 19 16°5 | 18 226 |- 1:5 | 13 18°5 | 22°5 | 26 27°5 | 28°5 | 27 26 23°5 | 20°5 | 17 1575 | 14 3°0 | 10 12 14 19 20 21 20 19 17 14 11 9 7 181 15 8-5 | 11 13 18 22°5 | 23 22°5 | 21°56 { 19°5 | 16 13°5 | 11°56 | 11 These results, when plotted, show the same aspect of curves as was found for the soft iron wires, and show that when the length of the wire is halved, both the circular magnetism or areas of the cycle-curve when the wire is in a longitudinal magnetic field of 2°5 units, and the maximum twist, are decreased about 50 per cent. Also, for three different lengths by plotting the load as abscissa, and the corresponding values of the maximum twist as ordinates, it will be found that when the load is ‘ncreased six times, the mean maximum twist is decreased about 38 per cent.; and when these three straight lines are produced, they 05 | 11 16 20 25 27 28 27°8 | 26 23 18:5 | 14°7 | 12°5 | 14:3 3:0 7 8:5 | 10°56 | 14°5 | 17 POG}, |) Teel |} ahe/ WGP) |) UI || abilor? | Ly 55 7 | 06 | 62) 7 [92 | 142 | 165 faz5 | 168 |16 [145 [a2 |100| 65 | 8 1138 15 8. |} 66 75 | 11 14 14:5 | 14 13:5 | 12:2 | 10 8 6°5 7 3:0 4 4:8 5:8 8 10 10-2 | 10 9°8 9°5 8:5 a 6°5 378 Brown Mechanical Stress and Magneiisation of Iron. 491 meet the axis of abscissee at a mean point corresponding to a load on the wire of about 8x 10° grammes per sq. em. It is interesting to notice, for the above three states of hardness H,, H,, and H,, that as the length of the wire is diminished and the load on the wire increased, the twists gradually approach to the same value as the longitudinal magnetic field round the wire is increased. Two new wires were now taken and prepared to give the degrees of hardness H,.; and H;; these were tested when at the full length of 226 cm. only for three loads and for twelve different magnetic fields. The rigidity of H:.; was 812 x 10° grammes per sq. em., its electrical conductivity 12°5, and its cross-sectional area 20°46 x 10~* sq. cm. The wire H; had a rigidity of 823x 10° grammes per sq. em., electrical conductivity 9:1, and cross-sectional area 20°35 10% sq. em. ‘The results of the tests are shown together in Table VI. TABLE VI. od - HD” , log HS 5 & a) Twist or deflection on the scale in mm. when subjected to the following ea a. | BX longitudinal magnetic fields :— Sa fo} q 3. dl g : ae aSzOg Ss ne t 5 desi | eH B.Ao|S & >! 0-45 | 0-75 | 1:0 | 15 | 2:0 | 25 3 4 6 9 12 14 |S828 eS | < 0-5 | 15°5 | 19°5 | 28 26 28 29 28°5 | 27°5 | 24°5 | 20-5 | 16-5 | 14 13°5 226 | 15 | 13 Nay |) ake 20 22°5 | 24 23°T | 23 20°5 | 17 13°5 | 11 10°8 2.5 | 3:0 | 11 12°5 | 13°5 | 15:5 | 17 17°56 | 17:2 | 17 16 12 9 75 6°5 05/18 |15 | 16:5 )18 | 188 |)19 | 18-7 |18 |16 |13 | 10-7 | 9-2 | 10-4 226 | | 15 | 11-5 | 18-2 | 14:5 | 16 TU7/ jay 16°5 | 15:5 | 14 11-2 9°5 8-2 8°6 Hs | | 370 | 8:5 | 10 1l 12 18 18°5 | 18-2 | 12°5 | 11:5 9°5 8:5 75 6:2 As will be seen from fig. 1, page 482, these two wires have departed from the straight-line relation between the rigidity and the load on the end of the wire when heated, exhibited by- the first three, that is, they have become more rigid or harder in proportion; the electric conductivities have also diminished at a more rapid rate. When, however, the loads are plotted against either the maximum twist in a field of 2°5 units, or the areas of the cyclic curves, they still show the straight-line relation found with the previous three tempered wires. In all the experiments on these wires in different states of magnetic softness, the maximum twist takes place in a longitudinal magnetic field of about 2°5 ¢.g.s. units. In the paper by Nagaoka and Honda,’ already 1 Phil. Mag., 1902, 6th Ser., vol. iv., p. 61. 492 Scientific Proceedings, Royal Dublin Society. referred to, the results are given of a test on an iron wire when the current through the wire was kept constant, and the longitudinal magnetic field round it varied, and it was found that the maximum twist occurred in a magnetic field of about 20 units. The wire used by them was 21 em. long and 0:98 mm. in diameter; the current sent through it was 6 amperes, or at the rate of nearly 800 amperes per sq. cm. ; that is, about eight times the current density used in most of the experiments described in this paper, and the maximum twist found by Nagaoka and Honda occurred in a magnetic field about eight times that found by the present writer with a wire about 16 mm. in diameter. In order to test experimentally how the magnitude of the magnetic field— in which the maximum twist occurs—varies when currents of different value are sent through the wire, a No. 17 iron wire was taken (in the physical state in which it came from the manufacturer) and prepared for testing in the usual manner. ‘The wire was placed in the solenoid, with a load on the end equivalent to 10° grammes per sq. cm., and the twist measured when the wire was subjected to twelve different longitudinal magnetic fields. Three sets of observations were taken when the current through the wire was 1°5, 3, and 6 amperes respectively, or when the current densities were 95, 190, and 380 amperes per sq. cm. he results obtained are shown in ‘able VII. Taste VII. dnservine | Ns can Geena ae magnetic field TEL, 95 190 380 we. Vee | oR 39:5 0°75 12:7 7 40°5 1:0 145 28:5 41 1:5 17 31 42°5 2-0 18-5 33°5 44 2-5 19 35°5 45°7 3 18-7 37 47 4 18:5 39 49°5 6 17 40:5 53-6 9 15 37-5 575 12 13 34 57:5 14 115 32 65-5 Brown—WMechanical Stress and Magnetisation of Iron. 498 If from the above table we plot as abscisse the values of the longitudinal magnetic fields round the wire, and as ordinates the corresponding values of the twist in each of the three cases, we obtain three smooth curves which rise gradually to a maximum, and then fall away more slowly. The maximum twist occursin different magnetic fields in the three cases—viz., about 2°5, 5:5, and 11 units respectively ; and if we further plot the values of the current through the wire as abscisse, and as ordinates the corresponding values of the magnetic field in which the maximum twist occurs, the three points will be found to lie very approximately in a straight line, which line, when produced, passes through the origin. This shows that the longitudinal magnetic field in which the maximum twist occurs is directly proportional to the current density in the wire. The results—collected from the tables above—for wires of very approximately the same diameter, but of different degrees of magnetic softness, are shown as curves in fig. 4. a SScs S| Twist in scale divs. (mm.) oO po . JO (py Fie. 4.—Magnetic field. All these wires were of the full length, 226 cm., and all loaded at the rate of 0:5 x 10° grammes per sq. cm. ‘I'he top curve is that obtained with the softest wire H,, and the lowest curve that obtained with the hardest wire H, ; the intermediate curves are those obtained with the wires of intermediate hardness H,, H2, and H;.; respectively. The wires at the different tempers were very approximately of the same diameter (Table I., p. 482), and the maximum twist occurred in all cases when the wire wasin a longitudinal magnetic field of 2°5 c.g.s. units. 494 Scientific Proceedings, Royal Dublin Society. The curves in fig. 5 show how the maximum twist—when the wire is in a longitudinal magnetic field of 2°5 units—varies with the hardness of the wire, and when the longitudinal load on the wire is varied. The top curve represents the results obtained with a load of 0°5 x 10° grammes per sq. cm., the middle curve with a load of 1:5 x 10° grammes per sq. cm., and the lowest curve when a load of 3:0 x 10° grammes per sq. em. was on the wire. a a ee Cae ee fe Tse Sait | ikea PSS PASS Twist in scale divs. (mm.) ~ sy 770 790 370 850 Vig. 5.—Rigidity (grammes per sq. cm. x 10°). These curves show that as the hardness is increased from H) to H, the twist is decreased 16 per cent. for the small load, about 29 per cent. for the high load, and about 22 per cent. for the intermediate load, The same general forms of curves are obtained if the results are plotted for any other longitudinal magnetic field. In fact, as the load on the wire and longitudinal magnetic field round it are increased, the relation between the rigidity and twist tends toward a straight line, and in the lower curve of fig. 5 the twist is decreased about 53 per cent., when the hardness is increased from H, to H, on the arbitrary scale. Experiments were now made to find out how the twist of the free end of the wire varied when the diameter or cross-sectional area of the wire was changed. The wires, as received from the manufacturer,’ were all in the same physical state, and had all received the same heat-treatment; that is, they were put in an annealing furnace and raised to a bright, cherry-red heat, then allowed to cool naturally until chevy could be handled. 1 For this set of experiments many attempts were 5 realy ml various methods tried to get the different-sized wires all in the same soft state Ho, but with no success owing to the great length of wire employed. It was found, however, on inquiry, that the wires as delivered by the manufacturer were all in the same physical state as regards heat-treatment, and they were therefore tested in that state. Brown— Mechanical Stress and Magnetisation of Iron. 495 Five of these wires of different sizes were taken and put through the tests, with a load on the end of each wire equivalent to 10° grammes per sq. cm. ; and the twist of the free end measured when each wire was placed successively in twelve different longitudinal magnetic fields. In every case the current density in the wire was 100 amps. per sq. cm. The three thicker wires were tested for simple rigidity, and the mean value obtained was 815 x 10° grammes per sq. em., which falls about the position H,.; in our arbitrary scale of temper. i The results obtained are shown here in Table VIII., and as curves in fig. 6. Taste VIII. Longitudinal Twist or deflectionien tis scale i a syhentte cross-sectional area magnetic wire x sq. cm. is as follows :— field = - - a Ei 6-06 11:39 20:6 34-7 53-9 0°45 6 (a) 85 10 12 0°75 7 8°5 10 12 14°5 1:0 ee 9:5 11°5 13°5 17 1:5 10 Tay 14°5 16°5 21 2:0 12°5 14°5 Wy 19 23°5 2°5 14 16 18 21 25°5 3 13°5 15°5 er 22 27 4 13 14°5 17 22 28 6 11°45 13 15°5 20 26 | 9 9°5 11 13°5 17 22°5 12 75 9 11 14°5 19 14 6 Yi) 9°5 12°5 17 = ZB 3E a e = Fie. 6.—Magnetie field. SCIENT. PROC. R.D.S , VOL. XII., NO. XXXVI. 4a 496 Scientific Proceedings, Royal Dublin Society. When the wires are of different diameters, the maximum twist occurs in magnetic fields of different strengths; in the top curve in fig. 6, which gives the results obtained with the thickest wire, the maximum twist takes place in a field of 4 units, and the lowest curve which was obtained with the smallest wire has its maximum twist in a field of about 2:3 units. By plotting the cross-sectional areas of the wires as abscissee, and the corre- sponding values of the twist, for any field, as ordinates, the points all lie very approximately in a straight line in each case; and when the cross- sectional area of the wire is increased about nine times, the increase in the twist is larger as the field is increased. Thus in a magnetic field of 3 units the twist is doubled, and in a field of 6 units the twist is increased about 2 times, and about 22 times in a field of 14 c.g.s. units. In an application of Kirchofi’s theory to the Wiedemann effect, Nagaoka and Honda’ have shown theoretically that for a given longitudinal current through the wire, the twist of the free end is inversely proportional to the square of the radius. In order to test this experimentally, three wires were taken of different diameters (viz. Nos. 15, 16, and 17), all in the same physical condition, as delivered by the manufacturer, and tested in twelve different longitudinal magnetic fields. ‘The wires were each 226 cm. long, the load on each was 10° grammes per square em., and the current through each was 15 amperes or of current densities 95, 72:8, and 57:5 amperes per sq. em. respectively. The tests were made in the usual way, that is, with a given longitudinal magnetic field round the wire: the current through it was raised to 1:5 amperes, and the twist or deflection on the scale read off; the current was then diminished to zero, reversed, again raised to 1-5 amperes, and the twist again observed; the mean of the two readings was then taken as the true twist for that magnetic field. The same process was gone through for fields up to 14 ¢.g.s. units. + Phil. Mag., 1902, 6th Ser., vol. iv., p. 68- Brown—WMechanical Stress and Magnetisation of Iron. 497 TABLE EX. | Twist in scale-divisions (mm.) when the cross- | eee | sectional area of wire x 10-8 sq. cm. is :— | | field | Ht. 15:8 20°6 | 26-0 | = = | | 0-45 10:2 85 7-5 | 0-75 12-7 10 | 8-2 | 1-0 1455 11 9 | 15 17 13 10-2 | 2-0 85 14-2 11-2 | | 2-5 | 9 15 115 3 8:7 14:7 11-2 4 8°5 | 14:5 | 11 | 6 17 13-5 10°5 | 9 15 11°5 9 | 12 13 10 | 75 14 115 9 | 7 | | The results are shown in able IX., and if these be plotted with the values of the longitudinal magnetic fields as abscissze, and the corresponding values of the twist as ordinates, the theoretical considerations are seen to be practically confirmed, that.is, that the twist is inversely proportional to the cross- sectional area of the wire, or directly proportional to the current density in the wire, for any given longitudinal magnetic field. This is seen when we plot the values of the current density in the wire, against the corresponding values of the twist ; the points will then be found to lie in a straight line which when produced passes through the origin. Also, by plotting the values of the cross-sectional area of the wire as abscissee, and the corresponding values of the twist as ordinates, the points will lie in a curve which gradually approaches the axis of abscisse. Since the twist varies inversely as the radius of the wire squared, and the cross-sectional area of the wire varies as the radius squared, the slope of the curve varies inversely as the radius of the wire to the fourth power; and therefore, as the radius increases, dy/dz decreases rapidly, and becomes zero at infinity. 498 Scientific Proceedings, Royal Dublin Society. SuMMARY. In parts I and II of this paper,’ where the results of experiments on soft iron wires only are given, the main conclusions arrived at were :—(1) The amount of transient current produced by twisting a soft iron wire placed in a longitudinal magnetic field, is inversely proportional to the load on the end of the wire, for a range from about 10° to 3 x 10° grammes per square cem., and when the load was increased about three times the transient current was about halved. (2) With a given longitudinal magnetic field round the wire, and a given longitudinal stress per sq. em. on it, the transient current produced by twisting the wire was found to be directly proportional to its cross-sectional area. (3) With a given longitudinal magnetic field round the wire, and a given current density in it, the twist of the free end of the wire was found to be inversely proportional to the load ; and when the load was inereased seven times, the twist was decreased about 14 per cent. (4) The current density in the wires being the same, the maximum twist of the free end was found to be directly proportional to the cross-sectional area of the wire, and when the cross-sectional area of the wire was increased 4-7 times, the maximum twist was increased about three times. For the experiments which form the subject of the present Part III of the paper, a new batch of the best Swedish charcoal iron was obtained, which is of practically the same quality as was used in the previous experiments. : With the soft iron wire of hardness H, when placed in a longitudinal magnetic field of 2°5 units, the twist of the free end is inversely proportional to the load on the wire, and when the load is increased six times, the twist is decreased 27 per cent. This diminution is about 7 per cent. more than was found with a previous soft iron wire tested under the same conditions, but this can be easily accounted for by the fact that the latter wire was a little softer and a little larger than the former one. In all the three states of hardness H,, H,, and H,, the twist of the free end is proportional to the length of the wire, and when the length of the wire is halved, both the area of the cyclic curve in a field of 2°5 units and the twist are decreased 50 per cent.; also in the three cases the twist is inversely proportional to the load on the end of the wire, and the harder the wire, the more the twist is diminished as the load is increased. When the load is increased six times, the twist is decreased 27 per cent. in the wire of 1 Scient. Proc. Roy. Dublin Soc., 1909, vol. xii., pp. 101 and 174. Brown—Mechanical Stress and Magnetisution of Iron. 499 temper H,, 32 per cent. in the wire H,, and 38 per cent. in the wire H., i.e., increased load has more effect on a hard wire than on a soft one. When the curves relating to load and twist in the three tempered wires H, to H. were produced, they all cut the axis of abscisse at points approximately equivalent to the elastic limit of the wire. The twist diminishes as the simple rigidity increases, and also as the load increases. When the hardness is increased from H, to H.,, the twist is decreased 16, 22, and 29 per cent. respectively, when the loads on the wires were 05, 15, and 3:0 x 10° grammes per sq. cm. When the wire becomes harder—between certain limits H, to H.— as the length of the wire is decreased, and the load on it increased, the twists gradually approach the same value as the surrounding longitudinal magnetic field is increased. In wires of different degrees of magnetic softness, the longitudinal magnetic field in which the maximum twist occurs is independent of the load on the end of the wire between certain limits. With iron wires in the same physical state but of different diameters, the maximum twist occurs in longitudinal magnetic fields of different strengths— the thicker the wire the higher is the magnetic field in which the maximum twist occurs. In a magnetic field of three units the twist is dowb/ed when the cross-sectional area of the wire is increased nine times. The magnitude of the longitudinal magnetic field in which the maximum twist occurs is directly proportional to the current density in the wire. With a given load on the wire, and a given current through it, the twist in a given longitudinal magnetic field is inversely proportional tothe cross- sectional area of the wire, or directly proportional to the current density in ib. SCIENT. PROC. R.D.S., VOLs XII., NOs XXXVIs 45 THE SCIENTIFIC PROCKEDINGS OF THE ROYAL DUBLIN SOCIETY. Vol. XII. (N.8.), No. 37. DECEMBER, 1910. MECHANICAL STRESS AND MAGNETISATION OF NICKEL. Pawn I BY WILLIAM BROWN, B.8Sc., PROFESSOR OF APPLIED PHYSICS, ROYAL COLLEGE OF SCIENCE FOR IRELAND. [Authors alone are responsible for all opinions expressed in their Communications. | x Greonian insti, DUBLIN: (~ AUG PUBLISHED BY THE ROYAL DUBLIN SOTHTK, LEINSTER HOUSE, DUBLIN. Ona\ BN WILLIAMS AND NORGATH, 14, HENRIETTA STREET, COVENT GARDEN, LONDON, W.C. 1910. Price One Shilling. Roval Bublin Society. FOUNDED, A.D. 1731. INCORPORATED, 1749 EVENING SCIENTIFIC MEETINGS. Tun Scientific Meetings of the Society are held alternately at 4.30 p.m. and 8 p.m. on the third Tuesday of every month of the Session (November to June). Authors desiring to read Papers before the Society are requested to forward their Communications to the Registrar of the Royal Dublin Society at least ten days prior to each Meeting, as no Paper can be set down for reading until examined and approved by the Science Committee. The copyright of Papers read becomes the property of the Society, and such as are considered suitable for the purpose will be printed with the least possible delay. Authors are requested to hand in their MS. and necessary Illustrations in a complete form, and ready for transmission to the Editor. Brown—WMechanical Stress and Magnetisation of Iron. 499 temper H,, 32 per cent. in the wire H,, and 88 per cent. in the wire alo, 1.e., increased load has more effect on a hard wire than on a soft one. When the curves relating to load and twist in the three tempered wires H, to H, were produced, they all cut the axis of abscisse at points approximately equivalent to the elastic limit of the wire. The twist diminishes as the simple rigidity increases, and also as the load increases. When the hardness is increased from H, to H., the twist is decreased 16, 22, and 29 per cent. respectively, when the loads on the wires were 0:5, 1°5, and 3:0 x 10° grammes per sq. cm. When the wire becomes harder—between certain limits H, to H.— as the length of the wire is decreased, and the load on it increased, the twists gradually approach the same value as the surrounding longitudinal magnetic field is increased. In wires of different degrees of magnetic softness, the longitudinal magnetic field in which the maximum twist occurs is independent of the load on the end of the wire between certain limits. With iron wires in the same physical state but of different diameters, the maximum twist occurs in longitudinal magnetic fields of different strengths— the thicker the wire the higher is the magnetic field in which the maximum twist occurs. In a magnetic field of three units the twist is dowbled when the cross-sectional area of the wire is increased nine times. The magnitude of the longitudinal magnetic field in which the maximum twist occurs is directly proportional to the current density in the wire. With a given load on the wire, and a given current through it, the twist in a given longitudinal magnetic field is inversely proportional to the cross- sectional area of the wire, or directly proportional to the current density in it. SCIENT. PROC. R.D.S., VOL. XII., NO. XXXVI. 4B [| oO 7 XXXVII. MECHANICAL STRESS AND MAGNETISATION OF NICKEL. Part I. By WILLIAM BROWN, B.Sc., Professor of Applied Physics, Royal College of Science for Ireland. (Ordered for Publication Ocrovrr 21. Published Drcemprn 4, 1910.) In 1883 it was observed by Knott! that the torsion produced on a nickel wire under the combined influence of longitudinal and circular magnetism was in a reverse direction to that produced on an iron wire; and a summary of thie subsequent work done in this part of the subject up to 1900 is given in Nagaoka’s report on Magnetostriction for the International Congress on Physics.2 As far as the present writer knows, no further work—bearing mainly on the twist of nickel wires when magnetised longitudinally and circularly—has been done except that of Shimizu and ‘lanakadate,* who found that the Wiedemann effect in nickel practically vanishes at the critical temperature of the metal. The present paper gives the results of some experiments on the magnetisation and torsion of nickel wire, the apparatus and experimental arrangements being the same as were employed by the author in the work on mechanical stress and magnetisation of iron.’ In this case, also, five different degrees of hardness were adopted, which were obtained by heating the wires to a cherry-red heat (in a darkened room) by means of a broad Bunsen burner when they were hung vertically and subjected to different longitudinal loads. In the experiments for testing the effect of temper on the torsion and magnetism, No. 16 size wires of pure nickel’ were employed, this size of wire being easier to manipulate than wires of smaller or larger diameter. In order, in the first place, to get the wire as soft as possible, it was suspended from the ceiling of a room by means of a three-jaw self-centering 1 ‘Trans. Roy. Soc. Kdin., vol. xxxii. * Paris, 1900, vol. ii., pp. 586-556, 3 Phys. Math. Soc. Tokyo, Proc., 1906. 4 Scient. Proc. Roy. Dublin Soe., vol. xii., No. 86, pp. 480-499. 5 For the chemical analysis I am indebted to Mr. A. (’ Farrelly, M.A., Lecturer in Organic Chemistry in the Royal College of Science, who found the following percentage composition :— Ni = 98:9; Fe =0:79; Si02=0'18; Cu=0:11; C=0°02; Zn=a trace. Brown— Mechanical Stress and Magnetisation of Nickel. 501 clutch, and allowed to hang loosely under its own weight only. It was then raised to a cherry-red heat—by means of the Bunsen burner—-by heating it from the top downwards; this was done three times to make sure that the wire was as soft as it could be made; and the hardness thus obtained we here call Ho. The highest degree of hardness was obtained by the same process, the wire in this case being heated when a weight or load was on the lower end of it equivalent to 10° grammes per sq. cem.; this degree of hardness we call H,; and the intermediate degrees were obtained in the same way when the loads on the end of the wires were } x 10°, 3 x 10°, and $ x 10° grammes per sq. cm. respectively, and these degrees of hardness are indicated by Hy, H:, and H:. Fresh No. 16 nickel wires were taken in each case to obtain the various degrees of hardness ; and the largest load of 10° grammes per sq. cm. did not appreciably change the diameter of the wire, which was taken as 0:163 cm. in all the tempers. It was found that the nickel wire was much easier to harden by this means than an iron wire of the same size; thus, an iron wire when heated with a load on its lower end of 10° grammes per sq. cm. had its rigidity increased about 1-8 per cent., whereas a nickel wire treated in the same way had its rigidity increased by about 6 per cent. The simple rigidity and electrical conductivity of each wire were measured, the former by means of a slightly modified form of Searle’s horizontal torsion apparatus, and the latter by comparison with a wire of pure copper. The results are collected in Table I, where the rigidity is expressed in grammes per square centimetre, and the electrical conductivity as a percentage of pure copper, the cross-sectional area of each wire being very approximately 20°8 x 10% sq. ems. Tanne I. | Load on the end Ka Bese Electrical con- of wire when Temper or | Rigidity ductivity pure heated. Grammes hardness. x 105. eee TaN pper, 190. per sq. em. : 3 B ze a 0 Hy 707 18:2 0-25 % 10 | TEL 718 ies 4 [cs50m | H, 728 | 17-5 0-75 ,, H, 739 | 17-0 : | he » H, 750 | 16°6 | Tf these results be plotted by taking on the axis of abscissee an arbitrary 432 502 Scientific Proceedings, Royal Dublin Society. scale of harduess—that is, the number in the first column, or the load in grammes per square centimetre which was on the lower end of the wire when it was being heated—and as ordinates the corresponding values of (1) the rigidity, and (2) the electrical conductivity, the points will be found in each case to lie very approximately in a straight line. In order to observe the amount of twist on the end of the wire when it was subjected to different longitudinal loads, and when placed in different longitudinal magnetic fields, and with constant circular magnetism (that is, with a maximum current through the wire in each case at the rate of 100 amperes per sq. em.), the same long solenoid with accessories was used as was employed in previous work.! The wire under test was fixed so as to hang vertically and as near as possible in the middle of the solenoid, the total length of which was 237 cms. ; the distance from the mirror on the vibrator or weight on the lower end of the wire to the millimetre scale on which the twists or deflexions were read was 116-5 cms., and for every millimetre of deflexion on this scale the end of the wire was twisted through an angle of 88 seconds. With a certain load on the end of the wire, and a given longitudinal magnetic field round it, an electric current was sent through the wire, and its value slowly increased up to the maximum of 100 amperes per sq. cm.; and the steady deflection read off the millimetre scale. he current was then gradually diminished to zero, reversed, and again slowly increased to the maximum, and the scale-reading once more observed on the other side of zero; the mean of these two readings was then taken as the true value of the twist for that longitudinal magnetic field. The longitudinal load on the wire being kept the same, a similar process was gone through for other fifteen different longitudinal magnetic fields up to a maximum of 50 ¢.g.s. units. The load on the wire was now increased, and a series of observations made similar to what was done for the first load ; again the load on the wire was increased, and another series of observations taken, giving results for three different loads in all. The above series of observations or tests were made on each of the five tempered wires, which were each 226 cms. long and 0:163 cm. in diameter. The results obtained are given in Table IT. 1 Scient. Proc. Roy. Dub. Soc., vol. xil., pp. 115-183, and pp. 480-499. 05 ~ & Brown— Mechanical Stress and Magnetisation of Nickel. IP 8-1F GG G-GP Ih @.L§ €& 66 96 GG SI GEL GL OL Ge) ¢.9 0-§ GS €¢ 8-88 C.F ¢.cg 98 98 cE ES GE 6G LG &G IG G.LT GT G1 Ty LG C16 G86 ¢.66 G08 G-1¢ ¢-GE GS ve G-S¢ 18 ¢.8G ¢.GZ FG GG 1G ¢.0 °.9¢ GLE G-8¢ 6¢ 8-L1¢ GG OF OF GFE 8G eo GeLT SIL art fe 8 fh 0-€ (Gi GS CF S-9F ¢-SF 1¢ €¢ G-GE 0¢ G-CF 6§ ¢.8s LG 8G 61 G.cT G.T ty | C.F§ 9g ORS ¢.6€ 152 SP CF G.LF 8P L¥. Si €.6¢ ¢.F§ ¢-1§ 8G G-9G ¢.0) | GL FL GOL G-8h 9L G-OL ¢-19 “EG G.9F 6-68 ¢.0€ GG ent FL G-O1 Pe a 0-6 c.ge G.9G 6-64 6:69 6-9 G89 Th G-0L 99 §-S¢ 6F IF G:GE LG C.0G 9L G1 ty cf L¥ 6F G-1G ¢.g¢ 9¢ G.8¢ 8-09 G-69 ¢.19 §-9¢ G-1¢ G-SF 8s GS 8G c.() ¥8 L8 06 v6 £6 98 9L G-¢9 cn OF | vee a0G tan | GI é-T1 €.6 0-€ 69 ¢.G9 G-69 gL LL 18 v8 GS 9L L9 ce Cr ¢.Fe 8G G1G @.JT G1 ty 6P ¢.Z¢ ¢.c¢ 6¢ ¢.69 @-99 8-02 FL GL C.8L ¢.99 09 8-67 G-1f GE €.8G ¢.0 pees @-101 BLOT. SIL OTT FIL 901 G6 G18 8-01 8¢ 8-GF G-GE 8-16 GOL GL 8 0-€ G.SL ¢-Sh last ¢.68 8-46 GOL Gol ¢- 101 C6 ¥8 89 GG .8¢ ¢-0§ GG SL ¢.T OF €.L¢ 19 g-F9 69 ¥L ¢.6) 98 ¢-16 G-56 8-€6 98 GL G69 6F 6.9 9G ¢.() 0g cD (7 cs 0€ Gs 06 9T €1 OL Z G ‘ca G T Gt-0 5 ee = — Spey ousvUL [eUIpUpSUOT SutMoTpoF oy} 0} pazoalqns st oILA\ OY} WA SaTJATUI[[IM Ut a[vOS LO WOTXapep IO STAT, s5 E 7 S JOE Senn y, 504 Scientific Proceedings, Royal Dublin Society. These results exhibit several peculiarities and differences from those obtained with iron wires in the same physical state and when tested under the same conditions. It is well known that the twist produced in the free end of a nickel wire is in the opposite direction to that produced in an iron wire when tested in the same way’; also that the maximum twist in iron takes place in the same longitudinal magnetic field whatever the load on the wire may be, within certain limits, and that the curve relating the longitudinal magnetic field round the iron wire and the twist of its free end is lower the greater the load on the end of the wire.” i100} Twist in scale divs. (mm.) (0) 20 40 60 Magnetic Field H. Fic. 1.—Wire of constant temper Ho, and under three different loads. One set of results for nickel wire is shown as curves in fig. 1—that is, for the temper or hardness Hy from Table II, where the abscissee are the values of the longitudinal magnetic field round the wire, and the ordinates the corresponding values of the twist of the free end of the wire expressed in millimetres on the scale. ‘Taking the curves in fig. 1 in order before they eross one another, the top or left-hand curve is that obtained with the small load on the wire, the lower or right-hand curve that obtained when the high load was on, and the middle curve that obtained with the intermediate load of 1°5 x 10° grammes per sq. cm. on the wire. 1 Knott, Trans. Roy. Soc. Kdin,, yol. xxxii. * Scient. Proc. Roy. Dub. Soe., vol. xii., No. 36, p. 484, Brown— Mechanical Stress and Maguetisution of Nickel. 500 It will be seen that each of the three curves rises quickly to a maximum and diminishes more gradually ; and the longitudinal magnetic field in which the maximum twist occurs changes with the value of the longitudinal load ou the wire.’ The longitudinal magnetic field in which the maximum twist occurs, for a given load, is the same for all the wires, though they are of different degrees of hardness, that is, in a field of 13 units for the small load, in a field of 385 units for the large load, and in a field of about 20 units for the middle load of 1:5 x 10° grammes per sq. cm. In Table II the maximum twists are indicated by means of heavy-faced type. Again, for all the wires of different degrees of hardness, the curves cross each other in the same longi- tudinal magnetic field; thus the curves obtained with loads (0°5 x 10°) and (1:5 x 10°) cross in a magnetic field of 13 units, under loads (0°5 x 10°) and (8 x 10°) in a field of 18:5 units, and under loads (1°5 x 10°) and (8 x 10°) they cross in a magnetic field of 23 units; these relations hold for wires of different lengths as well as for wires of different tempers. 100 Twist in scale diys. (mm.) ou is) fo) 20 7 20 _ 60 Magnetic Field H. : Fic. 2.—Wivres of different tempers and the load constant. Taking from Table II the results for wires of the same length (226 ems.) and with the same longitudinal load on the end (1°5 x 10° grammes per sq. em.), but of different degrees of temper or magnetic softness, and plotting as 1 Nagaoka, Phil. Mag., 1908, yol. xxix., p. 123 (using a nickel wire 30 cms. long. and 1 mm. diameter), found the maximum transient exwrrent produced in the wire (when the free end was twisted through 60°) to take place in a longitudinal magnetic field of 25 units with no load on the wire, and in fields of 45 and 70 when the loads on the wire were 3°82 x 10° and 7-64 x 10° grammes per sq. cm. respectively. 506 Scientific Proceedings, Royal Dublin Society. abscissee the values of the longitudinal magnetic field, and as ordinates the corresponding values of the twist, we obtain the curves shown in fig. 2 (p. 505), from which we see that the longitudinal magnetic field in which the maximum twist occurs is independent of the temper of the wire. The top curve is that given by the softest wire of temper H,, and the lowest curve that given by the wire of temper H,; and the intermediate curves represent the intermediate tempers or degrees of magnetic softness. Again, from the same table, if we plot on the axis of abscisse the values of the rigidity of the wires at different tempers, and as ordinates the corre- sponding values of the maximum twist for the five wires when they are subjected to the three different longitudinal loads, it will be found that the points in each of the three cases lie very approximately in a straight line. When these three straight lines are produced, they cut the axis of abscisse at practically the same point which represents a rigidity of 774 x 10° grammes per sq. em. ; and therefore a wire of this rigidity should give little or no twist when subjected to simultaneous longitudinal and circular magnetism. An attempt was made to test this experimentally ; but the nickel could not be hardened beyond a certain limit by means of hanging a weight on the end of it and raising it to a bright cherry-red heat. Weights equivalent to 1:5 x 10° and 2 x 10° grammes per sq. cm. were hung on the wire when it was being heated, and the highest rigidity obtained was 757 x 10° grammes per sq. cm.,—that is, a little above the wire of temper H,—and the twists for the various longitudinal magnetic fields (with a load on the wire of 1:5 x 10° grammes per sq. cm.) when plotted gave a flat curve situated below the lowest curve in fig. 2, and having a maximum twist of approximately 25 millimetres on the scale, which value of maximum twist falls practically on the straight line corresponding to the given load. This, however, shows that the graphs relating the maximum twist and hardness or temper of the wires, for the different longitudinal loads, are not really straight lines, but are flat curves, which for the higher tempers very gradually approach the axis of abscisse. In order now to test the effect of varying the length of the wire, a fresh No. 16 nickel wire was taken and prepared for experiment in the usual way, without getting it to any definite temper—which was, however, between the tempers Hz; and H;, judging from the values of the twists obtained. Three different lengths were tested, each under three different loads, and in sixteen different longitudinal magnetic fields. All the results are given in Table III; and one set of them is shown as curves in fig. 3—that is, the results obtained when the longitudinal load on the wire was 1:5 x 10° grammes per sq. em. 507 Brown—Mechanical Stress und Magnetisation of Nickel. | | G96 8-98 | LE § 9¢ FS ¢-06 GLE FG 0G | | | | CLG | €:9G £66 8-06 8-1€ C-6E €-6E tg 6G C.9G | GEG FG CFG GG G-9G CLG €-66 0€ 0€ ¢-86 6¢ 09 | 8-09 OL) |) Ge || Gag || BANE || Goer OF FE OF Clie Ci ¢.0¢ GE $¢ G-PE COC C.6F C.F 66 OF e-1F SF CFF G-OF CSF oe €-6F LY ¢.8¢ ce C.98 8& C.68 ¢.1F toi CFF C.GF OF @.19 C.0L €-G) €.GL OL ¢9 8-S8¢ e.[e CE GLE ce Le 6¢ 19 89 09 ¢-F9 (as) G8¢ Gg | GS 1) ole | Pade e¢ 9¢ | ¢-g¢ | 3.6¢ | ¢.6¢ 9¢ og cP OF GE 0€ GS 06 9T 81 Or --:Sp[ay oyouseur [vurpnjsaoy Surojpoy oy} 0} paysolqns st att Oy} UTA SOAJOWUIT[IUL UI B[VS TO WOTXOHep 10 IST, S:cn icon || (cg ¢.9 CF Gre 0-8 GG | ¢-8T al |) ow |] Gag L Got || BL Overs || Gok SI GCI | GOT c.0) | 9% | ¢-0G FL | ¢.0L L ¢ 0-8 1g | 8-08 | 8.2% SI rail Or G1 ISL Gi g og | s-F SI cl ¢.0 C7 | Gop || Gre re || G6 1% | 910-0 6.8% Ke) GG a 9 0-€ Ger | G-G8 | G96 | Go | ¢-FT Il GT 9GG G.67 | ¢-eF 9¢ | ¢.0g 6. || Gopi G0) HEL pee! 40 L g 8 td T 90 |. 28 BS BSB >| oo a8 S|) sie x B 2 g S), err ce Bes | oF TI] “1av NO, XXXVII. $., VOL. X10. SCIENT, PROC, R.D 50S Scientific Proceedings, Royal Dublin Society. The curves show that the longitudinal Joad on the wire being constant, the longitudinal magnetic field in which the maximum twist occurs is independent of the length of the wire. : Siig ee Li! ak | ae) aaa | i) — A SS Se | Ss 69 A olen ees c ‘Twist in scale divs. (mm.) (e) 20 40 60 Magnetic Field H. Fic. 3.—Wire with the temper and load constant, and varying length. If we plot the values of the length of the wires as abscissee, and as ordinates the corresponding values of the twist, for any longitudinal magnetic field, the points will be found to lie very approximately in a straight line in each case; and these straight lines when produced pass through the origin, which shows that the twist is directly proportional to the length of the wire. If for the three different lengths we plot the values of the load as abscissee, and as ordinates the corresponding values of maximum twist, the points im each of the three cases will be found to lie in a straight line. These three straight lines when produced cut the axis of ordinates at different points, which points indicate the values of the maximum twist. which would be obtained if the wire were tested with no load on it, thus :— | Length of wire | Maximum twist | Ratio of max. ; | (cms.) | REGO OF Tonge: | forno load. | twists. | 226 | i 58 | 1 181 | 0-8 47 0-81 | 113 0-5 | 29 | 0-5 Brown— Mechanical Stress und Magnetisution of Nickel. 509 Again, by plotting the values of the load on the wire as abscisse, and as ordinates the corresponding values of the longitudinal magnetic fields in which the maximum twist occurs, the points will be found to lie in a straight line which, when produced, cuts the axis of ordinates at the point 10 on the scale. This indicates that, if the wire were tested with no longitudinal load on it, the maximum twist would occur when a longitudinal magnetic field of 10 units was round the wire. In order to test this experimentally, a thin aluminium tube holding a mirror, and a piece of iron wire for making contact with the mercury, was hung on the lower end of the wire, the whole weighing only 33 grammes, giving a load per sq. em. of 0:016 x 10° grammes. The results of the test are shown in Table ITI, in the first line under the length 181 cms.; and when the curve is drawn, the maximum twist occurs when the wire is surrounded by a longitudinal magnetic field of 10-3 units, which practically confirms the observation made above. [- ] ‘Twist in scale divs. (mm.) Magnetic Field H. Fic. 4.—Wire of constant temper and under three different loads. ‘The curves relating magnetic field and twist in nickel wires, with respect to the longitudinal load, are different from those given by an iron wire, inasmuch as, after a certain longitudinal magnetic field (about 24 units), the twists produced on a highly loaded wire are greater than the twists given by 402 510 Scientific Proceedings, Royal Dublin Society. one more lightly loaded. This was seen in fig. 1 (p. 504), and is also seen in fig. 4 (p. 509), which is the graph of the results obtained with a wire of hardness between H: and Hg, and of length 181 ems. To the right of the figure (over the point representing a magnetic field of 50 units) the top curve is obtained witha load on the wire of 3 x 10° grammes per sq. em., the lowest curve being that obtained with a load of 0°5 x 10° grammes per sq. cm., and the middle curve that obtained with the medium load of 15 x 10° grammes per sq. cm. It was thought possible that these curves might re-cross if higher longitudinal magnetic fields were applied to the wire under the three separate loads; a test was therefore made in which longitudinal magnetic fields up to a maximum of 200 units were applied ; and the results are given in Table IV, which shows that the curves maintain their relative positions even in that high field, and the slope of the curves, when the results were plotted, show that they are not likely to re-cross for even higher fields.’ Tasix LV. Twist or deflection on scale (in mm.) when the wire Longitudinal is loaded with grammes per sq. em. x 105:— Magnetic Field | _ Hi. | 05 1:5 3:0 | 50 39 | 46- 56 75 | 31 B75 47°5 100 | 25°2 30°8 39°8 125 21 25°5 32°5 150 18 21°5 | 27 | 176 16 18°2 22 200 | 14°75 15°65 182 | I | { In some previous work’ with iron wires it was found that the longitudinal magnetic field in which the maximum twist occurred was directly propor- tional to the current density in the wire, and tests were made to find out if nickel wire obeyed the same law. :, 1'The long solenoid at present in use for these experiments could not conveniently carry a current to give magnetic fields higher than 200 units. A new solenoid, howeyer, is being wound, by means of which wires 1 metre long can be tested in much higher fields. *Scient. Proc. Roy. Dub. Soc., vol. xii., p. 493. Brown—Mechanical Stress and Magnetisation of Nickel. 511 A fresh No. 16 nickel wire was taken and prepared for experiment in the usual way, without attempting to get any definite degree of hardness ; the heating, however, left the wire in a state of magnetic softness lying between H, and Hy, of our arbitrary scale, judging from the values of the twists subsequently obtained. The length of the wire was 226 cms.; and the longitudinal load on it during the tests was 0°5 x 10° grammes per sq. cm.; and four sets of observations were made when the current densities in the wire were 50, 100, 150, and 200 amperes per sq. cm. respectively; the results are shown in Table V, and as curves in fig. 5. ‘Twist in scale divs. (mm.) 40 Magnetic Field H. Fic. 5.—Wire with load and length constant and yarying current density. [Tasin V, 512 Scientific Proceedings, Royal Dublin Society. TABLE V. Twist or deflection on the scale (in mm.) when the current densities in Longitudinal the wire were :— Magnetic Field x fl. | 5 50 100 150 200 0-45 19 ol 44 57 1 21 35 50 65 2 24 45 63-8 8d 3 27 52 79 | 109 b) 32 66 100 153 7 36 74 110 46 10 40 80 121 160 18 41 82 123 164 16 39°5 80 120 159 20 388 77 115 152 25 36°5 73 109 144 30 34°5 69 103 136 35 32°5 65 97 128 40 30°65 61 92 122 45 29 58 86 115 50 27 54 81 10875 From the curves in fig. 5 it will be seen that, unlike the iron wire, the maximum twists with the nickel take place in the same longitudinal magnetic field of about 13 c.g.s. units. If for any field we plot the values of the current density in the wire as abscissee, and as ordinates the corresponding values of the twist, the points will be found to lie, in every case, very approximately in a straight line, which, when produced, passes through the origin, showing that the twist of the free end of the wire is directly proportional to the current density in the wire. Experiments were now made in order to find out how the twist of the free end of the wire varied when the cross-sectional area of the wire was changed. Five wires were taken—Nos. 15 to 20, inclusive—and prepared to have the ‘degree ‘of hardness H,. Each wire was raised to a bright cherry-red heat Brown— Mechanical Stress and Magnetisation of Nickel. 518 three times (when hanging loosely in a vertical position with no load on it), so as to make sure that it was as soft as the material and method of softening would permit. The wires (226 cms. in length) were then put through the tests, with a longitudinal load on the end of each equivalent to 10° grammes per sq. em.— that is, the twists were observed in the usual way when the wires were successively subjected to sixteen different longitudinal magnetic fields, the current through the wire in all cases being at the rate of 100 amperes per square centimetre. The results obtained are given in Table VI. Taste VI. ‘Twist or deflection on the scale (in mm.) when the cross-sectional area of the | | Longitudinal wire x 10 sq. em. is :— | Magnetic Field eee H. 61 11-7 15:9 20:8 26:3 a (sso = 0°45 9 14 25 30 41 1 14 20 30°5 36 47 2 23°5 31 40°5 46 56 3 33 40 48°5 55 63°5 Hy) 49-5 56 62 67 75 7 63 68°5 73°5 17 83:5 10 78 82 85 89 93 13 87 90 92 95 98 16 85:5 89 91 94°5 97-5 20 81:5 | 85 87-5 90°5 94 25 77 | SO | 82°45 85 89 30 722 75 78 SO | 83°53 15) 67°75 70 73 7A | 79 40 63 66 68 70°5 | 73°2 i) 58°5 61:5 | 63°5 66 | 69 50 54-5 575 59-5 62 | 65 | | It we plot these results in able VI in curves with the values of the longitudinal magnetic fields as abscisse, and as ordinates the corresponding values of the twist in each case, the curves will be found to be very symmetrical and practically parallel to one another, the maximum twist in N 514 Scientific Proceedings, Royal Dublin Society. each case occurring in the same longitudinal magnetic field of about 13 ¢.g.s. units. This also is different from what takes place with an iron wire when it is tested in the same way and under the same conditions, for with iron the longitudinal magnetic field in which the maximum twist occurs is different for different-sized wires.! In the case of nickel wires, as will be seen from Table VI, when the cross-sectional area of the wire is increased about 4:3 times, the increase in the twist is larger as the longitudinal magnetic field is increased; thus in a field of 13 units (the field in which the maximum twist occurs in each case) the twist is increased 12°6 per cent., in a field of 30 units 16-7 per cent., and 19°3 per cent. in a field of 50 e.g:s. units. It has been shown—with iron wires—that the twist was directly pro- portional to the current-density in the wire for any given longitudinal magnetic field round it. Table V above gives the results of tests made with a No. 16 nickel wire, with a longitudinal load on it of 0°65 x 10° grammes per ~ sq. em., when different currents were sent through the wire. Experiments were now made with three wires of different diameters, and the same value of current through each—that is, the current densities were 100, 724, and 55-2 amperes per sq. em. respectively. The three wires were tested in the state of magnetic softness, represented by H, on our arbitrary scale, and were Nos. 16, 17, and 18, s.w.g. The length of each wire was 226 cms., and the longitudinal load on each 10° grammes per sq. cm. ‘The results obtained are given in Table VII. 'Scient. Prov. Roy. Dub. Soc., vol. xii., p. 495. Brown—WMechanical Stress and Magnetisation of Nickel. 515 Taste VII. Twist or deflexion on the scale (in mm.) when Tovernaciaal ee ee area of the wire x 10-° Magnetic Field an REA H. } 11:7 15:9 20°8 0-45 14 | 10 75 1 20 16 18 2 31 23°5 18°5 | 3 40 30 23 | 5 56 41°5 31 7 68°5 50 875 10 82 59 46 13 90 64:5 49 16 89 | 64 48 20 ge | @ 46:5 25 80 | 585 44 30 78 | 55 41:5 35 70 | 515 39 | 40 66 | 48 36°5 45 61°5 | 45 34:5 50 57-5 | 41°5 32 If the results given in Table VII be plotted with the values of the longitudinal magnetic fields as abscissee, and as ordinates the corresponding values of the twists, the three curves show that the maximum twist occurs in the same longitudinal magnetic field of about 13 units in each case. By plotting as abscissze the values of the current densities in the wire, and as ordinates the corresponding values of the twists for any longitudinal magnetic field, the points in every case will lie very approximately in a straight line, which, when produced, passes through the origin, which again shows that the twist is directly proportional to the current density in the wire. Some differences in the behaviour of iron and nickel when subjected to simultaneous longitudinal and circular magnetism are shown in fig. 6 (p. 516), where one of the curves should be drawn below the axis of the abscissa to indicate that iron and nickel twist in opposite directions, but both are here shown above the axis for convenience of reference. SCIENT. PROC. R.D.S., VOL. XII., NO. XXXVII, 4p 516 Scientific Proceedings, Royal Dublin Society. The wires were tested under exactly the same conditions, and were of the same length (226 cms.) and the same diameter (0°163 cm.) ; had the same longitudinal load (0:5 x 10° grammes per sq. cm.), and the current density was the same in each—viz. 100 amperes per sq. cm. Both were made as soft as the nature of the material and method of annealing would permit. The simple rigidity of the iron wire was 774 = 10°, and of the nickel wire 707 x 10° grammes per sq. cm. The maximum twist for the iron (40°5 mm.) took place in a longitudinal magnetic field of 2°5 units, and the maximum twist for nickel (95 mm.) in a field of 15 units. Twist in scale divs. (mm.) Magnetic Field H. Fic. 6.—Similar wires of nickel and iron contrasted. SUMMARY. [V.B.—Some results previously obtained with iron wire are also given here for convenience of contrast and reference. | 1. Soft nickel wire heated to a bright cherry-red heat when loaded at the rate of 10° grammes per sq. cm. has its rigidity cncreased about 6 per cent., and its electrical conductivity decreased 8°8 per cent. Soft iron wire when treated in the same way has its rigidity increased 1:8 per cent., and its electrical conductivity decreased 2°1 per cent. Brown— Mechanical Stress and Mugnetisation of Nickel. 517 2. With nickel, beyond a certain value of longitudinal magnetic field, the twist increases with the load on the wire. With iron, in all fields, the twist decreases with the load. 3. With nickel, the longitudinal magnetic field in which the maximum twist occurs is independent of the hardness of the wire, but is different for different loads. ‘With iron the magnetic field in which the maximum twist occurs is independent of the load. 4, With nickel, the curves obtained with differently loaded wires, cross one another in definite longitudinal magnetic fields, whatever the hardness or length of the wire. With iron the curves never cross, whatever the load or length of the wire. 5. With nickel, the longitudinal magnetic field in which the maximum twist occurs is independent of the length of the wire, the load being constant. The same holds with iron wire. 6. With nickel, the twist is directly proportional to the length of the wire. The same holds with iron wire. 7. With nickel, the longitudinal magnetic field in which the maximum twist occurs is independent of the current density in the wire. With iron it changes with the current density. 8. With nickel, the twist is directly proportional to the current density in the wire, whether with different currents through the same wire or the same current through wires of different diameters. The same holds with iron wire. 9. With nickel wires of different cross-sectional areas, the maximum twist occurs in the same longitudinal magnetic field. With iron the maximum twist takes place in different longitudinal magnetic fields. 10. With nickel, when the cross-sectional area of the wire is increased 4:3 times, the increase in the twist is larger the greater the longitudinal magnetic field: thus, in a field of 13 units, the twist is increased 12-6 per cent.; in a field of 30 units, 16-7 per cent.; and in a field of 50 cgs. units, 19°3 per cent. With iron for the same increase in the cross-sectional area of the wire, the twist is increased about 75 per cent. in a field of 14 c.g.s. units. 11. When the soft wires of nickel and iron (of the same length, diameter, and under the same longitudinal load) are tested under the same conditions, the maximum twist for nickel (95 mm.) occurs in a longitudinal magnetic field of 13 units, and the maximum twist for iron (40:5 mm.) takes place in a field of 2°5 c.g.s. units. 518 Scientific Proceedings, Royal Dublin Society. Note added November, 1910.—In continuation of this work a fresh nickel wire was prepared in the usual manner, to give the degree of magnetic softness between H: and Hs; of the arbitrary scale used above, and was tested when subjected to siz different longitudinal loads, in longitudinal magnetic fields up to 200 ¢.g.s. units. The current sent through the wire was at the same rate as before, viz. 100 amperes per sq. cm.; and it was found that the relation between the longitudinal magnetic field in which the maximum twist occurs, and the longitudinal load on the wire when it was being tested, still follows very approximately the straight-line law; but the maximum twist attains its highest value when the load on the wire is at the rate of 4 x 10° grammes per sq. cm., as shown in the following table :— Tontthe wire in | Maximum twist | genetic feld in grammes per sq. cm. aa i eeoule: which the maximum x 10°, twist occurs. 0°5 66 13 1°5 75 20 3 81 35 4 86-5" 44 5 86-5 54 6 86'5 64 ( S10 ) INDEX TO VOLUME XII. Absorption of Water by Seeds (Arxtns), 30. Agricultural Seeds and their Weed Impu- rities: A Source of Ireland’s Alien Flora (JoHnson and Hensman), 446. Ammonia produced from Atmospheric Ni- trogen (WOLTERECK), 54. Analytical Machine (Lupeats), 77. Atkins (W. R.G.). Absorption of Water by Seeds, 38. Cryoscopic Determination of the Osmotic Pressures of some Plant Organs, 463. Osmotic Pressures of the Blood and Eggs of Birds, 123. See Dixon and ATErns, 275. Barometer: A Simple Form of Open-scale Isothermal Air Barometer (BARRETT), Aaa Barrett (W. F.). Methods of Determining the Amount of Light scattered from rough surfaces, 190. A New Form of Polarimeter for the Measurement of the Refractive Index of Opaque Bodies, 198. A Simple Form of Open-scale Iso- thermal Air Barometer, 444. Barrington (R. M.). A New British and two New Irish Birds, 18. Beeswax, Analysis of (Ryay), 210. ~ Benzidine, Value of, for the Detection of Minute Traces of Blood (McWErEnNery), 216. Birds: A New British and two New Irish Birds (Barrineron), 18. Brown ‘W.). Mechanical Stress and Mag- netisation of Iron, 101, 175, 480. Mechanical Stress and Magnetisa- tion of Nickel: Part L., 500, Permanent Steel Magnets, 312. Chrome Steel Permanent Magnets, 349, SCIENT. PROG. R.D.S. VOL. XII., INDEX. Cattle, Origin of the Dexter-Kerry Breed of (Witson), 1. Cattle, Scandinavian Origin uf the Hornless Cattle of the British Isles (Watson), 146. Chrome Steel Permanent Magnets (Brown), 349, Chrysophlyctis endobiotica, Schilb. (Potato- Wart or Black Scab), and other Chytri- diacece (JoHNson), 131. Colours of Highland Cattle (Wttson), 66. Cow’s Milk, Separate Inheritance of Quan- tity and Quality in (W1tson), 470. Cryoscopie Determination of the Osmotic Pressures of some Plant Organs, 463. D’Albe (E. E. Fournier), On Photography by Reflection under Contact, 97. Dendrobium, Pollination of certain Species of (Kerr), 47. Dew at Kimberley, some Observations of (Surron), 266. Ditton (Thomas), see Ryan and Ditton. Dixon (H. H.). Vitality and the Trans- mission of Water through the Cells of Plants, 21. Note on the Tensile Strength of Water, 60. Dixon (H. H.) and W. R.G. Arxrys. On Osmotic Pressure in Plants; and on a Thermo-Electric Method of determining Freezing- Points, 275. Eassie (J/ajor l'.). Some Variations in the Skeleton of the Domestic Horse and their Significance, 321, Fossil Hare of the Ossiferous Fissures of Ightham, Kent, and on the Recent Hares of the Lepus vuriabilis Group, 225. Hensman (Diss R.), see Jonnson and HENSMAN. 4h 520 Index. Hinton (M. A.C.). On the Fossil Hare of the Ossiferous Fissures of Ightham, Kent, and on the Recent Hares of the Lepus variabilis Group, 225, Horse: Variations in the Skeleton of the Domestic (Easstn), 321. Horses, Inheritance of Coat Colour in (Witson), 331, Inheritance of Coat Colour in Horses (Witsown), 331. Tron ; Mechanical Stress and Magnetisation of (Brown), 101, 175, 480. Johnson (T.). Chrysophlyctis endobiotica, Schilb. (Potato-Wart or Black Scab), and other Chytridiaces, 131. Further Observations on Powdery Potato-Seab, Spongospora subterranea (Wallr.), 165, Johnson (T.) and Miss R. Hensman: Agricultural Seeds and their Weed Im- purities: A Source of Ireland’s Alien Flora, 446. Kerr (A. F. G.). Notes on the Pollination of certain Species of Dendrobium, 47. Light scattered from rough Surfaces, Methods Determining the Amount of (BARREtr), 190. Ludgate (P. E.). Ona proposed Analytical Machine, 77. Lyons (W. J.). On the Distribution of Mean Annual Rainfall and average Number of Rain Days per year over an Avea including the Counties of Dublin, Wicklow, Kildare, and Meath: A Study in Local Variation of Rainfall, 354. McWeeney (E. J.). On the Value of Benzidine for the Detection of Minute Traces of Blood, 216. Mechanical Stress and Magnetisation of Iron (Brown), 101, 175, 480. Mechanical Stress and Magnetisation of Nickel (Brown), 500. Montanin and Montana (Montan) Waxes (Kyaw and Drnton), 202. Moss (R. J.). Taxine in Irish Yew, 92. New British (A), and two New Irish Birds (Barrineron), 18. Nickel, Mechanical Stress and Magnetisa- tion of (Brown), 500. Observations of Dew at Kimberley (Surron), 266. Origin of the Dexter-Kerry Breed of Cattle (Wutson), 1. Osmotic Pressure in Plants; and on a Thermo-Electric Method of determining Freezing-Points (Dixon and AvrxKqys), 275, Osmotic Pressures of the Blood of Eggs and Birds, 123. Osmotic Pressures of some Plant Organs, Cryoscopic Determination of (ATKINS), 463. Permanent Steel Magnets (Brown), 312. Photography by Reflection under Contact (D’AnBe), 97. Polarimeter for the Measurement of the Refractive Index of Opaque Bodies (Barrett), 198. Pollination of certain Species of Dendro- brum (Kern), 47. Production of Ammonia from Atmospheric Nitrogen (WoLTERUCE), 54, Rainfall and average Number of Rain Days per year over an Area including the Counties of Dubliv, Wicklow, Kildare, and Meath (Lyons), 354. Ryan (Hugh). The Analysis of Beeswax, 210. Ryan (Hugh) and Thomas Ditton. On Montanin and Montana (Montan) Waxes, 202. Scandinavian Origin of the Hornless Cattle of the British Isles (W1tson), 145. Separate Inheritance of Quantity and Qua- lity in-Cow’s Milk (Witson), 470. Spongospora subterranea (Wallr.), Pow- dery Potato-Scab (Jonson), 165. Sutton (J. R.). Some Observations of Dew at Kimberley, 266. Taxine in Irish Yew (Moss), 92. Tensile Strength of Water (Drxon), 60. Thermo-Electric Method of determining Freezing-Points (Drxon and ATKINS), 275. Index. Vapour-Pressures, Specific Volumes, Heats of Vaporisation, and Critical Constants of Thirty Pure Substances (Youne), 374. Variations in the Skeleton of the Domestic Horse and their Significance (NAsstE), 321, Vitality and the Transmission of Water through the Cells of Plants (Drxon), 21. Wilson (J.). Colours of Highland Cattle, 66. Inheritance of Coat Colour in Horses, 331. 521 Wilson (J.). Origin of the Dexter-Kerry Breed of Cattle, 1. Scandinavian Origin of the Horn- less Cattle of the British Isles, 145. Separate Inheritance of Quantity and Quality in Cow’s Milk, 470. Woltereck (H. C.). Production of Am- monia from Atmospheric Nitrogen, 54. Young (Sydney). The Vapour-Pressures, Specific Volumes, Heats of Vaporisation, and Critical Constants of Thirty Pure Substances, 374. END OF VOLUME XII. | 16, 17. SCIENTIFIC PROCEEDINGS. VOLUME XII, . The Origin of the Dexter-Kerry Breed of Cattle. By Jamms Witson, M.a., B.sc. (Plates I-IV.) (January 20,1909.) Is. . A New British and two New Irish Birds. By Ricuarp M. Barrieton, ™.a. (February 2, 1909.) 6d. . Vitality and the Transmission of Water through the Stems of Plants. By Henry H. Dixon, sc.p., r.n.s. (March 15,1909.) 6d. . The Absorption of Water by Seeds. By W. R. Getsron Arxins, B.A. (March 16, 1909.) 6d. . Notes on the Pollination of Certain Species of Dendrobium. By A. F. G. Kerr, u.p. (Plates V., VI.) (April 3, 1909.) 6d. . Production of Ammonia from Atmospheric Nitrogen. By Herman C. Wourerscx, px.p. (April 3, 1909.) 6d) . Note on the Tensile Strength of Water. By Henry H. Drxon, sc.p., res. (April 5, 1909.) 6d. . The Colours of Highland Cattle. By James Witson, m.a., B.sc. (Plate VII.) (May 15, 1909.) 6d. . On a Proposed Analytical Machine. By Percy H. Lupears. (April 28, 1909.) 6d. . The Taxine in Irish Yew. By Ricwarp J. Moss, r.t.c., F.c.s. (April 30, 1909.) 6d. . On Photography by Reflection under Contact. By EH. EH. Fournmr p’Ausg, B.SC., A.R.C.SC., M.R.I.A. (Plate VIII.) (May 10,1909.) 6d. . Mechanical Stress and Magnetisation of Iron. Part 1. By Wit.1am Brown, B.sc. (May 28,1909.) 1s. . The Osmotic Pressures of the Blood and Hees of Birds. By W. R. Guxsron Arxins, B.A. (May 28, 1909.) 6d. - Chrysophlyctis endobiotica, Schilb. (Potato-Wart or Black Scab), and other Chytridiacee. By T. Jonson, D.sc., F.u.s. (Plates IX.-XI.) (June 16, 1909.) 1s. . The Scandinavian Origin of the Hornless Cattle of the British Isles. By James WILSON, M.A, B.Sc. (June 19, 1909.) 1s. Further Observations on Powdery Potato-Scab, Spongospora subterranea (Wallr.). By T. Jounson, p.sc., r.u.s. (Plates XII.-XIV.) (July, 1909). 1s. Mechanical Stress and Maenetisation of Iron. Part 2. By Wiuiam Brown, B.sc. (June 21,1909.) 1s. 22. 23. bl. SCIENTIFIC PROCEEDINGS—continued. Methods of Determining the Amount of Light scattered from Rough Surfaces. By W. F. Barrerr, r.n.s. (July 27,1909.) 6d. . A New Form of Polarimeter for the Measurement of the Refractive Index of Opaque Bodies. By W. F. Barrerr, r.zs. (July 27, 1909.) 6d. ). On Montanin and Montana (Montan) Waxes. By Huen Ryan, m.a., D.se., F.R.U.1., and Tuomas Ditton, ma. (July 17, 1909.) 6d. . The Analysis of Beeswax. By Hueu Ryan, m.a., p.s¢c., r.R.u1. (July 17, 1909.) 6d. On the Value of Benzidine for the Detection of Minute Traces of Blood. By Hi. J. McWereney, u.a.,u.pv. (August 14,1909.) 6d. On the Fossil Hare of the Ossiferous Fissures of Ightham, Kent, and on the Recent Hares of the Lepus variabilis Group. By Marrm A. C. Hinton. (Plate XV.) (September 8, 1909.) 1s. 6d. 4. Some Observations of Dew at Kimberley. By J. R. Surron, m.a., se.p. (January 15, 1910) 64d. . On Osmotic Pressure in Plants; and on a Thermo-Hlectric Method of Determining Freezing-Points. By Henry H. Dixon, sc.p., F.R.s., and W. R. Gexston Arxins, ma. (February 28,1910.) 1s. 6d. . Permanent Steel Magnets. By W. Brown, B.sc. (February 21,1910.) 6d. . Some Variations in the Skeleton of the Domestic Horse and their Significance. By Masor F. Hassm, p.s.o. (Plates XVI.-XX.) (March 8, 1910.) 1s. . The Inheritance of Coat Colour in Horses. By James Wiison, M.A., B.SC. (April, 1910.) 1s. . Chrome Steel Permanent Magnets. By W. Brown, s.sc. (April, 1910.) 6d. On the Distribution of Mean Annual Rainfall and average number of Rain Days per year over an area including the counties of Dublin, Wicklow, Kildare, and Meath: A Study in Local Variation of Rainfall. By Witu1aM J. Lyons, B.a., a.n.c.se. (Lonp.). (Plate XXI.: Map.) (May, 1910.) 1s. The Vapour-Pressures, Specific Volumes, Heats of Vaporisation, and Critical Constants of Thirty Pure Substances. By Sypney Youne, D.sc., F.R.S. (June, 1910.) 3s. A Simple Form of Open-Scale Isothermal Air Barometer. By W. F. Barret, v.r.s. (May 19, 1910.) 6d. . Agricultural Seeds and their Weed Impurities: A Source of Ireland’s Alien Flora. By T. Jonson, p.sc., F.u.s., and Miss R. Hensman. (Plates XXII., XXIII.) (July, 1910.) 1s. . Cryoscopic Determination of the Osmotic Pressures of some Plant Organs. By W. R. Geusron Arxins, m.a, (June 380, 1910.) 6d. . The Separate Inheritance of Quantity and Quality in Cows’ Milk. By James Wison, m.A., B.sc. (July 29, 1910.) 6d. Mechanical Stress and Magnetisation of Iron: Part 3. By W. Brown, 8.sc. (September, 1910.) 1s. - Mechanical Stress and Magnetisation of Nickel: Part I. By W. Brown, B.sc. (November, 1910.) 1s. DUBLIN: PRINTED AT THE UNIVERSITY PRESS BY PONSONBY AND GIBBS. sake Mr Prats na Hay RAT a Peabe Kens ae < SMITHSONIAN INSTITUTION LIBRARIES WA 3 9088 01303 4194