PROCEEDINGS OF THE CHEMICAL SOCIETY APRIL 1959 THE PLACE OF CHEMISTRY. X.* IN THE SCOTTISH UNIVERSITIES By ANDREW KENT (GLASGOW UNIVERSITY) WHEN Galileo was promoting mechanical philo- new natural philosophy. In the fluid situation of re- sophy Bacon was laying the foundation of Chem- covery and overhaul then prevalent in Scottish estab- ical. ...The improvement of experimental knowledge lishments novelties had more instant appeal than in was left to Boyle who took up Chemistry where the stabler south as the northern universities in the Bacon left it. ...Boyle is almost the first chemist that eighteenth century replaced their antique “regent” is worth reading; but after him came many famous system by a new and specialised professoriate there ones particularly the great Newton....Lastly came was further opportunity for modernisation. Boerhaave . . . the chief reformer of the Method and The impulse for the first provision of chemical Language of the Chemists (who) has made great dis- curricula in Scotland came from the desire to pro- coveries in Chemistry particularly in fire . . .” vide in continental fashion university schools of -William Cullen Lectures. instruction in medicine which would supplement existing facilities for the examination of intending Since St. Andrews dating from 1413 is the oldest physicians. Thanks to Paracelsus such plans in the Scots university and Edinburgh founded in 1582 the eighteenth century inevitably included chemistry; youngest they have all at some time taught variants and they were bound also to reflect the influence of of the standard medieval curricula with a degree of Leyden’s Boerhaave whose Elernenta Chemiae ex-emphasis on Aristotelian ideas of the structure pounded his successful redaction of alchemy and of matter.Before a more modern approach to chem- iatro-chemistry to an empirical factual recital. istry was anywhere attempted two later influences Some sixty years after the appointment of had impinged effectively on the national outlook. D’Avisonne (ne Davidson) at Paris the town The Calvinist religion and politics of John Knox council of Edinburgh appointed in 1713 Dr. James led to a general replacement of French cultural in- Crawford (Crauford) as their first professor of fluence by Dutch. The Glorious Revolution of 1688 Physic and Chemistry and this project gained its paved the way for the advent of a notably advanced second wind when in 1726 Dr.Andrew Plummer pro- and forward scientific outlook stimulated largely by fessed the chemical sector of a joint-chair at the Isaac Newton’s triumphant demonstration of the College in Chemistry and Medicine. In 1747 Dr. * For earlier articles in this series see Proc. Chem. SOC.,1957,185 190,273,276,313; 1958,93,96,135,307; 1959,52. 109 William Cullen filled the new lectureship in Chem- istry at Glasgow the regius chair was provided in 1818. Similar developments at St. Andrews were somewhat delayed by the lack of an adequate hospital and a concurrent shortage of cadavers;l but a course in Chemistry and Chemical Pharmacy under Dr.Robert Briggs was authorised in 181 1 and a first appointment to the (Purdie) Chair of Chemistry duly transpired in 1840. Crawford and Plummer were both pupils of Boerhaave. “Plummer’s still reminds us of their iatro-chemical interests although this doctor’s status in science was well recognised. His own pupil Cullen was to become in 1757 his successor as teacher of Chemistry and Medicine; but not before he had sounded at Glasgow a clear and compelling call to recognise the status of Chemistry as a separate branch of science. Thomas Thomson held him the first to do so and the first to lecture in Scots instead of Latin.3 In curricula exposition and breadth of appeal Cullen and his successor Joseph Black set new standards rapidly extended by their students to other institutions at home and in America.With their professional background and their academic purpose in medicine they still contrived emphasis on Chemistry not as a simple nurse of nostrums but as a new and powerful physical science and they opened their classes to every type of enthusiast. ORIGINAL APPROACHES If such developments portrayed the response of medical science events in Aberdeen underline the extent to which Galileo and Newton had energised all departments of Scots philosophy. In both univer- sities of this city various experiments in the teaching of chemistry were undertaken in the Faculties of Arts before in 1793 Marischal College4 announced Dr. George French-who had studied under Cullen at Edinburgh-as its first professor of Chemistry; and before Patrick Forbes provided a course on Chemistry and Mineralogy at King’s College in 1817 which was also an Arts class and indeed a compul- sory subject superseding Willian Ogilvie’s venture.If French was grandson to a litster (dyer) Forbes like Ogilvie was a professor of Humanity (Latin) and in early emphasis on the contribution of science to general culture these northernmost universities- whose union came only in 1860-together set a valuable precedent still echoed in the Arts curricula at Aberdeen. Both courses were of more immediate effect when a joint medical school was established in 1819. A third development was based neither on PROCEEDINGS medicine nor on general culture but emphatically if prematurely on the values of a science-centred and inexpensive education.This was John Anderson’s University established without an official charter by those eighty-one trustees to whom this erstwhile Glasgow College professor had entrusted El000 and his dream of a cultivated democracy. In 1796 within a few months of Anderson’s death they appointed Thomas Garrett later of the Royal Institution he was followed by George Birkbeck whose admiration of Glasgow’s Mechanics Institute would fructify later as a constituent college of London University. So successful was the chemical section of their lec- tures on natural philosophy that already in 1804 the trustees appointed Andrew Ure as their first professor of Chemistry.Thus early in the nineteenth century Scotland had six established courses in chemistry of university rank apart from those in Mechanics Institutes the powerful extra-mural schools in Edinburgh or other teaching enterprises. It is clear that the ferment of science had proliferated headily in the century before. If David Gregory Francis Hutcheson Adam Smith David Hume and “Common-sense” Reid typified the more erudite devotees there were few specialists in our sense of the term. Crawford Edinburgh’s first chemist later accepted a chair of Hebrew :John Anderson had once professed Oriental Languages Ogilvie and Forbes were Aberdeen divines:St. Andrew’s first professor Arthur Connell had been trained as a lawyer. Yet were they all Newtonian.The public response to their efforts was and remains astonishing. Nine hundred and seventy-two students enrolled in Garnett’s third course of Chem-istry. Andrew Ure took three classes in 1806 two of which had enrolments of some 400.Cullen’s class at Edinburgh was nearly 150 strong Black’s was over 200 and Hope in 1827 issued tickets to 575. So late as 1878 the first Graham lecture promoted by the chemical section of GIasgow’s Philosophical Society attracted an attendance of one thousand. Three factors may be advanced in partial explana- tion of such statistics. First it must be borne in mind that until about 1840 chemistry included heat light electricity and magnetism :as mineralogy it covered much of modern geology.Chemistry was almost the natural history of the inanimate affording to its adepts wide and colourful fields of exposition. Secondly the startling facts of pneumatic chem- istry originating in Black’s “Magnesia Alba” and still a world’s wonder shortly provoked the exciting- ly novel views of Mons. Lavoisier-translated and on Robert Walmsley “Veterum Laudes,” p. 121. * A calomel (Hg2C12) confection retained in the Pharmacopeias for more than two centuries. a Thomas Thomson “A History of Chemistry,” Vol. I London 1830 p. 315. Marischal College and University like the Town’s College of Edinburgh was a post-reformation foundation. APRIL1959 sale in Scotland within a few weeks-and then the astonishing atomism of Mr. Dalton. These were great days in which to be alive and to be an amateur of chemistry.Finally and of special note to Scotland partly ruined by the American War of Independance there was wide interest in the industrial potential of this new knowledge. Black invested in an unlucky soda venture but Watt turned his views on heat to mas- sive account :James Hutton was enterprising in am- monium chloride provision (non-cameline) William Irvine was an expert on glass-manufacture. The MacIntoshes Josias Gamble Charles Tennant and the rest became agents of a Chemical Revolution in industry. John Galt has written of Scotland in this era that “The minds of men were excited to new enterprises a new genius as it were had descended upon the earth.” Much of this drive originated in the new schools of chemistry whose vitality was ex-pressed through some of the earliest chemical societies on record.The Chemical Society of Edinburgh University dates back to 1785. There was a shorter-lived Chemical Club in Glasgow College at the same time this and a second venture presided over by William Ramsay grandpgre had strong in- dustrial associations. Such bodies catalysed the change of natural philosophers into chemical specialists. PRACTICAL INSTRUCTION Chemistry’s intrusion in the older academic fabric did not go quite unresisted by argument on ex-pediency legitimacy or status. The universities were as poor as they were ambitious and this new science came late to the table of endowments. Against the poverty of the new departments and the meagre equipment of the “preparation rooms” it is difficult to overestimate their research achievements or their sustained vitality.Edinburgh’s first appointee was warned “not to expect any salary” Glasgow was launched on an unexpected windfall of E30 and forty years later the entire equipment was inventoried on a single sheet of notepaper. In Marischal College the early lectures were circumscribed because “we have not apparatus to fill up a much longer course” at the neighbouring King’s the first graduating course was floated on the relics of earlier ventures “the three-foot table with 220 and some glass mostly broken.” Sir James Irvine speaks of St. Andrew’s first lecturer as “provided with no class-room no laboratory no apparatus and no place on the time- table” however the United Colleges sent their Quaestor to a sale of Thomas Thomson’s surplus apparatus where he procured a job-lot costing E40.5 In chemistry our academic institutions were en-countering with distaste the sharp and continuous cost of scientific education; yet worse was to come for it later became evident that laboratory instruction of students was a pedagogic necessity at university levels.Thomson had experimented with such instruction in his extra-mural work in Edinburgh appointed professor at Glasgow he demanded an undergradu- ate laboratory and by 1820 had obtained it. Hope in Edinburgh soon followed suit. (Liebig’s similar equipment was set up at Giessen in 1824.) Thomas Graham at Anderson’s began a practical course in 1831.In the same year Thomson achieved the con- struction of a new department the “Shuttle Street Laboratories” of Glasgow University specially de- signed for such teaching at a greatly grudged ex- penditure of S5,000.6 His pupil Thomas Clark in- augurated practical chemistry at Marischal College in 1833 William Gregory followed up at King’s in 1840 this member of his great academic clan having studied under Hope and Liebig. Largely for financial reasons such equipment at St. Andrews had to await the appointment of Thomas Purdie in 1884; and he avoided the inevitable senate storms by financing the construction himself. These changes implied some extra emphasis on the scientific st-udents a rapid rise in their quality and the production of a stream of young chemists most notably from Glasgow and Edinburgh who were of a new “professional” type filling many posts of importance in the newer English institutions and industries.They played a notable part in the founda- tion of “The Chemical Society,” first headed by a Scots professor Graham of London University. Their intensive training underlined the need to reduce and homogeneate the syllabus. By 1840 heat had been transferred with similar intangibles to physics interest in mineralogy died out slowly perhaps because of private and profitable exercises in quanti- tative analysis; and a traditional attention to agri- cultural chemistry did nothing to improve the underdevelopment in organic studies.Prominent students like Gregory Playfair Ramsay Crum Brown Purdie and Scott Couper had to look abroad for inspiration in this subject (and also for suitable scientific degrees). In the period post 1850 this deficiency continued. Moreover organised research on fundamental topics was sporadic and of little significance research schools in our modern sense were long unknown. A reasonable explanation if a poor excuse lay in the chronic financial shortages affecting staffing building and equipment. TEACHERS & TEXTS There did result a powerful sequence of text-books Sir James Irvine “Chemistry Centenary Lecture,” a pamphlet Edinburgh 1941 p. 11. * Andrew Kent “Glasgow University Gazette,” 1956 no. 25 p. 10. based on the new modes of instruction.The later works of Thomas Thomson those of Graham Gregory R. D. Thomson and Greville Williams6 were widely influential in mid-century and later while the Scottish centres still maintained the degree of distinctiveness they had established in the previous century. Yet the influence of The Chemical Society the broadened basis of the Royal Society the activa- tion of new departments in London and the pro- vinces all tended to fuse the northern schools into the general British field. Frederick Penny came to Anderson’s frsm London in 1839; Brazier appointed to the Chair at Aberdeen in 1862 was an early pro- duct of the Royal College of Chemistry Even more suggestive was the appointment of Wm. Dittmar at Anderson’s in 1874 only three years after his countryman Gustav Bischof had accepted the new Young appointment in Technical Chemistry at the same college.’ This second half-century was a period of great teachers.Gregory Crum Brown John Ferguson Fyfe Brazier and Heddle could all claim pedagogic skill. Their academic status was effective in intra- mural policy while the public status of the profes- sion was enhanced by the activities of Sir Lyon Playfair M.P. and the Master of the Mint. New opportunities occurred not only at Anderson’s but also when Carnelly of Owen’s College became the first (Baxter) professor of Chemistry at University (Queen’s) College Dundee; and when the Heriot- Watt College in Edinburgh inducted W. H. Perkin Jr.to a similar post in 1887. Victorian prosperity had other useful repercus- sions. Public funds became available on an en-lightened scale though reaching a total of only f40,OOO for all purposes in 1893. Professorial salaries were gradually made independent of fees erstwhile “private” assistants were officially paid. The department at Edinburgh acquired new “palatial” accommodation in 1884. Endowed lectureships in metallurgy and organic chemistry appeared before the century’s end; yet Glasgow made the first such ap- pointment in physical chemistry only in 1904 an opportunity furiously developed by Frederick Soddy.s (Scotland’s academic appointments still lack a chair of physical chemistry.) Probably the chief scientific event of this Victorian period was the appeal of Thomas Purdie at St.Andrews in 1884 for student-and-other support of his proposed school of research. This disciple of PROCEEDINGS Wislicenus and Edward Frankland founded the St. Andrews school of stereochemistry and being Purdie proceeded then to finance it and to achieve a properly equipped centre by 1905 which (Sir) James Irvine would inherit in 1909. Further twentieth-century developments are most- ly too recent for historical review. The Carnegie Trust from 1901 gave effective assistance at last with the financial burden of modern research :everywhere its largesse was welcome the groups around Purdie Soddy Japp and (Sir) James Walker had stimulus and s~pport.~ The establishment of the first Universities Grants Commission in 1919 opened even wider horizons.In this same year of post-war reconstruction T. S. Patterson became the first (Gardiner) professor of Organic Chemistry at Glasgow ;loand Edinburgh’s appointment of George Barger to a new chair of Chemistry in relation to Medicine neatly returned the courtesy of Dr. James Crawford‘s promotion some two centuries before-a full circle which must delimit this survey. A RECAPITULATION In summary then chemistry in Scotland found its roots in iatro-chemistry and its inspiration in Galileo Newton and Boyle. There were some important scientific societies but the nurture of modern science was a university achievement assisted by a degree of disorganisation which provided a near tabula rasa for reformers of the early eighteenth century.In Aberdeen the cultural status in Anderson’s the social significance of the science was boldly set forth; the earlier and more general development was pur- sued in faculties of medicine with a syllabus incor- porating a substantial amount of physics and geology. This supervisory association persisted despite the foundation of science faculties until in the times of such as Japp and G. G. Henderson it had outlived its useful function; yet the sturdy atti- tude of Cullen and Black asserted from the beginning the full status of this new branch of physical science whose classes attracted every kind of enthusiast including dissenters from New England Ulster and elsewhere. The Newtonian fervour of Scottish philosophy in the eighteenth century was a major factor in the astonishing response to chemical praelection in the nineteenth.The departments whose cost had been already lamented now proceeded to the provision of Anderson’s University or Institution was later incorporated in the Royal Technical College Glasgow now the Royal College of Science and Technology. * Sir Alexander Fleck “BioFaphical Memoirs of Fellows of the Royal Society,” 1957 Vol. 111 p. 206. * “Record of Fellows &c. The Carnegie Trust Edinburgh 1935; A. Logan Turner “History of University of Edinburgh 1883-1933,” Edinburgh 1933 p. 47. lo In this same year (Sir) I. M. Heilbron occupied a similar chair at the Royal Technical College Glasgow. Neither appointment has been continued in the original form.APRIL1959 laboratory instruction against the heavy prolonged resistance of economists and neo-scholastics. It was largely through chemistry that the northern univer- sities were eventually reconciled to the cost of scientific teaching and the new equipment as in Shuttle Street produced the first professional chem- ists. On the other hand the following half-century was a period of consolidation in which European progress in organic and physical chemistry or in the promotion of research schools was rather slowly followed up. The first years of this century saw pro- vision of Carnegie finances to support research schools which now cover the whole spectrum of interests in large and well-equipped departments.There are few populations which produce a similar proportion of university graduates in the subject and General references this has relation to an old still vigorous tradition of chemistry as a very important unit in the scientific sector of the national culture which now enjoys both the prestige of departmental size and the dignity of long standing. It is little wonder that the Scots chemists have a bent to historical studies. They share a contact thin yet clear with the Physica of the classical quadrivium a primary position in the practical manifestation of modern science involving a long association with Medicine and Arts which antedated the Science faculties a pioneering insistence upon expensive laboratories and quick proof of their profit in teach-ing and research; and a list of scientists who have added lustre for centuries to their subject and their colleges.Alexander Findlay “The Teaching of Chemistry in the Universities of Aberdeen,” Aberdeen 1935. Andrew Kent (Ed.) “An Eighteenth Century Lectureship in Chemistry,” Glasgow 1950. John E. Mackenzie “The Chair of Chemistry in the University of Edinburgh,” J. Chem. Educ. 1935 12 503. John Read “Men of Science The Chemists,” in “Veterum Laudes,” (J. B. Salmond Ed.) Edinburgh 1950. (For details of more recent developments see the articles “Schools of Chemistry in Great Britain,” in J. Roy. Znsi. Chem. 1953-4.) NEWS AND ANNOUNCEMENTS New Members of Council.-The following new appointments to the Council were announced at the Annual General Meeting Vice-presidents who have filled the Ofice of President Sir Robert Robinson Honorary Secretary Professor A.W. Johnson Vice-Presidents who have not filled the Ofice of President Professor F. Bergel Elected Ordinary Members of Council Constituency I Dr. C. B. Amphlett Dr. H. T. Openshaw Dr. W. A. Waters Constituency LI Professor C. H. Hassall Constituency IV Dr. I. J. Faulkner Professor G. E. Coates Dr. A. H. Lamberton Corday-Morgan Medal and Prize.-The Corday-Morgan Medal and Prize for 1957 has been awarded to Professor G. W. Kenner (Heath Harrison Profes- sor of Organic Chemistry University of Liverpool) in consideration of his contributions to synthetic organic chemistry particularly in the field of peptide and nucleotide chemistry.This Award consisting of a Silver Medal and a monetary Prize of 200 Guineas is made annually to the chemist of either sex and of British Nationality who in the judgment of the Council of The Chemical Society has published during the year in question the most meritorious contribution to experimental chemistry and who has not at the date of publica- tion attained the age of thirty-six years. Copies of the rules governing the Award may be obtained from the General Secretary of the Society. Applications or recommendations in respect of the Award for the year 1958 must be received not later than December 31st 1959 and applications for the Award for 1959 are due before the end of 1960.Election of New Fellows.-1 18 Candidates whose names were published in the Proceedings for February have been elected to the Fellowship. Chemical Society Library.-The Library will be closed for the Whitsun Holiday on May 18th and 19th 1959. Elections to The Royal Society.-The following Fellows of The Chemical Society have been elected to the Fellowship of The Royal Society F. Bergel Professor of Chemistry in the University of London at the Chester Beatty Research Institute of the Institute of Cancer Research Royal Cancer Hospital. Distinguished for his work in synthetical organic chemistry in the field of vitamins and drugs. R.J. W.Le FZvre Professor of Chemistry in the University of Sydney and Head of the School of Chemistry.Distinguished for his studies of the physical properties of organic compounds. G. V. Rajwor Feeney Professor of Physical Metal- lurgy in the University of Birmingham and Head of the Department of Physical Metallurgy. Dis-tinguished for his research on the structure of metals and alloys. R. E. Richards Lecturer in Chemistry at the University of Oxford Physical Chemistry Labora- tory. Distinguished for his work on nuclear magnetic resonance and its application to chemical problems. R. Spence C.B. Chief Chemist and Head of Chemistry Division Atomic Energy Research Establishment Harwell. Distinguished for his work on the chemistry of radioactive elements and for his leadership of the Chemistry Division of Harwell.The Chemical Council.-The Chemical Council has elected Honorary Officers for 1959 as follows Chairman Dr. J. W. Cook F.R.S. Vice-Chancellor of the University of Exeter ;Vice-chairman Profes- sor Harold Burton ;Honorary Treasurer Dr. Frank Hartley; Honorary Secretary Dr. J. Chatt. Conferences and Symposia.-The 1959 Gordon Research Conference on Radiation Chemistry will be held at New Hampton School New Hampton New Hampshire U.S.A. Enquiries should be ad- dressed to Mr. W. George Parks Director Depart- ment of Chemistry University of Rhode Island Kingston Rhode Island U.S.A. An International Symposium on Macromolecules sponsored by the International Union of Pure and Applied Chemistry will be held in Wiesbaden Germany on October 12-16th 1959.Enquiries should be addressed to Dr. W. Mauss Wiesbaden- Biebrich Rheingaustrasse 25 (Kalle & Co. A.G.) Germany. Transfer of D.S.I.R. Stations.-Three D.S.I.R. Stations-all concerned with the preservation storage handling or protection of foodstuffs-will be transferred to the Agricultural Research Council on July 1st. They are the Ditton Laboratory at Larkfield Kent the Low Temperature Research Station at Cambridge and the Pest Infestation Laboratory at Slough. The agreement between the Council for Scientific and Industrial Research and the Agricultural Re- search Council-which has received the approval of the Lord President of the Council (Lord Hailsham)- was made because the two Councils believe that it is in the national interest to bring all Government research on food other than fish under one body.The aim is to make the research more effective. Translation of Papers on Iron and Steel.-The British Iron and Steel Industry Translation Service run on a co-operative basis by leading British iron PROCEEDINGS and steel companies with the British Iron and Steel Research Association and The Iron and Steel Insti- tute will arrange a translation of any article con- nected with the manufacture of iron and steel. Translations required urgently are put in hand straight away; before the invoice is sent details are circulated to 150 steel companies and others to try to find others to share the cost. When cost is a factor the enquiry is circulated first and translation is not begun unless two or more other requests are received the price including illustrations is less than half the net translating cost.Personal.-Pro fessor S. J. AngyaZ has been awarded the H. G. Smith Memorial Medal for 1958 for his work on sulphonamides the Sommelet re action and the cyclitols. The Medal is awarded every year by the Royal Australian Chemical Institute to the member who in the opinion of the Council of the Institute has contributed most to the development of some branch of chemical science. Dr. H. J. Barber has been awarded the degree of D.Sc. of London University for work in the field of chemotherapy. Professor D. H. R. Barton has been selected as first recipient of the Roger Adams Award in Organic Chemistry.This Award is sponsored by the American Chemical Society by Organic Reactions Inc. and by Organic Synthesis Inc. and consists of a medal and an honorarium of $5000. It is to be given biennially to a single individual without re-gard to nationality for outstanding contributions to research in organic chemistry. Dr. A. H. Beckett has been awarded the D.Sc. degree of the University of London for his work in the field of medicinal and pharmaceutical chemistry. The Society Medal of the Society of Chemical Industry which is awarded not more than once every two years for conspicuous services to applied chem- istry or to the Society has been awarded for 1959 to Dr. Francis H. Carr C.B.E. Sir Alexander Fleck who has been named the first Visiting Fellow of the American Section of the Society of Chemical Industry has spent the first two weeks in April in the United States visiting the main chemical centres of the country.Sir Alexander Fleck has accepted the Presidency of the Industrial Co-partnership Association. The Honorary Degree of Sc.D. has been conferred by the University of Dublin upon Professor E. L. Hirst Professor of Organic Chemistry in Edinburgh University. The Senate of the University of London has con- ferred the title of Fellow of University College London upon Professor R. S. Nyholm. Dr. F. L. M. Pattison has been appointed Professor and Head of the Department of Chemistry at the APRIL1959 University of Western Ontario London Ontario Canada effective from July lst 1959.Dr. D. Traill Research Director of the Nobel Division of Imperial Chemical Industries Limited has been elected a Fellow of the Royal Society of Edinburgh. Dr. G. 0.Aspinall will be spending the months of September and October 1959 visiting a number of centres in the United States and Canada to lecture. Professor D. M. Newitt Head of the Department and Courtauld Professor of Chemical Engineering at the Imperial College of Science and Technology visited Spain in February to discuss the training of chemical engineers at the invitation of the Director of the Instituto Quimico de Sarria Barcelona. Dr. F. G. T. 0.Torto the Society’s Local Repre- sentative for Ghana will visit this country during the period June-September 1959.Mr. K. L. Butcher Brotherton Lecturer in Chemical Engineering at the University of Leeds has been promoted to the title and status of Senior Lecturer from October 1st. Mr. R. B. Fisher Demonstrator in Biochemistry at Oxford and Lecturer of Wadham College Oxford has been appointed to the Chair of Chemistry in relation to Medicine at Edinburgh University in succession to Professor G. F. Marrian who resigns on September 30th. Dr. T. W. Goodwin has been appointed to the Chair of Agricultural Chemistry at Aberystwyth University in succession to Professor R. 0.Davies who retires at the end of the present session. Mr. N. H. Haddock of Imperial Chemical In-dustries Limited has been transferred from photo- graphic chemical research to undertake special researches in the field of dyestuffs chemistry.The Eirrl of Halsbury has been invited to join the Board of Lancashire Dynamo Holdings in April as a Vice-chairman. Dr. A. K. Kiang has been appointed Senior Lecturer in Chemistry at the University of Malaya. Dr. A. C.Monkhouse who was recently Acting Director of Fuel Research of the Department of Scientific and Industrial Research has been ap-pointed Adviser to the Athlone Fellowships Scheme. Dr. E. N. Morgan of Parke Davis & Co.Ltd. has been appointed to the newly created post of Research Manager at their Hounslow laboratories. Dr. K. T. Potts has joined the staff of the Depart- ment of Organic Chemistry at the University of Adelaide. Dr. W. 1.Pumphrey Manager of the Research Department of Murex Welding Processes Ltd. Waltham Cross Herts. has been elected to the Court of Governors of the University of Birmingham as a representative of the Guild of Graduates. Dr. W. Rigby has been appointed Lecturer in the Department of Organic Chemistry at the University of Leeds from October 1st. Mr. 0.G. Weller who has held the position of Technical Manager of the Uganda Development Corporation Ltd. for some years has succeeded the late Mr. Trefor Davies as technical executive officer of A.B.C.M. responsible for the work of the chem- ical engineering research and advisory service committee the instrumentation advisory service committee and the information exchange com-mittee. Mr. J. Wood has accepted a post at the Woodstock Agricultural Research Centre of Shell Research Ltd.near Sittingbourne. FORTHCOMING SCIENTIFIC MEETINGS London Thursday May 7th at 7.30 p.m. Meeting for the Reading of Original Papers. “Recent Developments in the Chemistry of the Ipecacuanha Alkaloids,” by A. R. Battersby R. Binks G. C. Davidson B. J. T. Harper and S. Garratt. “Aldol Pinacol and Benzoin-type Re- actions of dl-Pyrroline 1-Oxides,” by R. F. C. Brown V. M. Clark M. Lamchen B. Sklarz and Sir Alexander Todd. “Perpendicular Conjugation in Some Octahedral Metallophthalocyanine Deriva- tives,” by J. A. Elvidge and A. B. P. Lever. To be held in the Rooms of the Society Burlington House W.l. Coffeewill be served in the Library from 7p.m.(Abstracts of the Papers are available from the General Secretary.) Thursday June 4th at 7.30 p.m. Meeting for the Reading of Original Papers “Kinetics and Orientation of Some Epoxide Ring- opening Reactions” by N. B. Chapman N. s.Isaacs and R. E. Parker. “Aromatic reactivity. Part Ill. Cleavage of Substituted Phenyltrimethylsilanes by Sulphuric Acid in Acetic Acid-Water,” by F. B. Deans and C. Eaborn. “The Heats and Entropies of Ionisation of Some Aromatic and N-Heteroaromatic Amines,” by J. J. Elliott and S. F. Mason. To be held in the Rooms of the Society Burlington House W. 1. Coffee will be served in the Library from 7 p.m. (Abstracts of the Papers are available from the General Secretary.) Birmingham Friday May lst at 4.30p.m.Lecture “The Structure and Reactivity of Reduced Metallic Surfaces,” by Professor K. W. Sykes M.A. D.Phi1. Joint Meeting with Birmingham University Chemical Society to be held in the Chemistry Department The University. Friday May 29th at 7 p.m. Lecture “Polyethers,” by Professor G. Gee Ph.D. F.R.S. To be given at Courtaulds Ltd. Coventry. Durham Monday May 4th at 5 p.m. Lecture “Carbon-14 Compounds,” by Dr. J. R. Catch. Joint Meeting with the Durham Colleges Chemical Society to be held in the Science Labora- tories The University. Exeter Friday May 22nd at 5 p.m. Lecture “Optical Rotatory Dispersion in Structural Organic Chemistry,” by Dr. W. Klyne. To be given in the Washington Singer Laboratories Prince of Wales Road.Irish Republic April 29th May lst and May 4th Lecture “A Generation of Chemotherapeutic Re- search,” by Dr. F. L. Rose O.B.E. F.R.I.C. F.R.S. Joint Meeting with the Institute of Chemistry of Ireland the Royal Institute of Chemistry and the Society of Chemical Industry. To be held as follows April 29th at 7.45 p.m. in the University Chemical Laboratory Trinity College Dublin. May lst at 7.45 p.m. at University College Cork. May 4th in the Chemistry Department University College Galway. Friday May 8th at 7.45 p.m. Lecture “The Nature of the Active Centre of Chymotrypsin and Other Esterases,” by Professor H. N. Rydon DSc. D.Phil. F.R.I.C. Joint Meeting PROCEEDINGS with the Werner Society to be held in the University Chemical Laboratory Trinity College Dublin.Manchester Friday May lst at 7.15 p.m. Lecture “Characterising the Useful Properties of Starch,” by Dr. T. J. Schock. Joint Meeting with the Royal Institute of Chemistry and the Society of Chemical Industry to be held in the Chemistry Lecture Theatre The University. Northern Ireland Tuesday May 5th at 7.45 p.m. Official Meeting and Tilden Lecture “Nucleotides and Bacterial Cell-wall Components,” by Professor J. Baddiley D.Sc. Ph.D. To be given in the Chem- istry Department Queen’s University Belfast. Oxford (Joint Meetings with Oxford University Alembic Club to be held in the Inorganic Chemistry Lecture Theatre.) Monday May 4th at 8.15 p.m. Lecture “Magnetic Properties of Molecules,” by Dr.J. A. Pople. Monday May 18th at 8.15 p.m. Lecture “Some Unnatural Products,” by Professor E. R. H. Jones D.Sc. F.R.S. Monday June 8th’ at 8.15 p.m. Lecture “Graphite and its Crystal Compounds,” by Professor A. R. Ubbelohde D.Sc. F.R.S. St. Andrews and Dundee Friday May lst at 5.15 p.m. Lecture “Chemical Effects due to Fission Frag- ments,” by Dr. R. Spence C.B. D.Sc. F.R.I.C. Joint Meeting with the University Chemical Society to be held in the Chemistry Department St. Salvator’s College St. Andrews. CENTENARY LECTURE* Spectra of Free Radicals By G. HERZBERG RESEARCH OTTAWA, (NATIONAL COUNCIL CANADA) INthis lecture I propose to look at certain chemical problems from the point of view of the physicist and more particularly of the spectroscopist.Chemists have studied free radicals for many decades. A large variety of these radicals has been postulated in in-numerable chemical reactions Sometimes the argu- ments for the temporary existence of a free radical in a chemical reaction are somewhat ambiguous or so they seem to a physicist. As a spectroscopist one would naturally want to observe the spectrum of a free radical before one would feel entirely convinced of its presence in a chemical reaction a flame an explosion an electric discharge etc. In addition once a spectrum has been obtained it is possible to * Delivered before the Chemical Society in London on October 16th 1958; also in Dundee on October 15th 1958; in Southampton on October 17th 1958; in Sheffield on October 20th 1958; and in Liverpool on October 21st 1958.APRIL1959 117 ~~ ~ ~~~ ~ ~~ derive from it useful information about the structure of the radical its energy levels and possibly its dis-sociation energies all of which may be of importance for an understanding of its behaviour. The spectra of many diatomic free radicals are well known. In emission for example the OH NH CH C, and CN radicals are observed in all sorts of flames in explosions and in electric discharges. Most of them also occur in emission in the spectra of comets and some in the spectrum of the night glow. In absorption the same free diatomic radicals can be observed in suitable gases at high temperature. They have also been observed in the spectra of low- temperature stars and a few even in interstellar absorption.The structure of these diatomic free radicals is well understood and I do not propose to discuss them in this lecture. Until about ten years ago very little definite spectroscopic information was available about poly- atomic free radicals. To be sure several emission spectra observed in flames or discharges were ascribed to polyatomic free radicals (for example the hydrocarbon flame bands to HCO the a-bands of ammonia to NH,) but there was no certainty of such an assignment since no detailed analysis of these spectra was available and they had not been observed in absorption under conditions under which the presence of these radicals could be con- sidered as certain.The first polyatomic free radical that was definitely identified by its spectrum was CF,. This spectrum was first observed in emission in a discharge through CF by Venkateswarlul and later in absorption in a similar discharge by Barrow and his collaborators.2 The structure of the spectrum and the conditions of observation led unambiguously to the result that it was due to CF,. Long before this identification of CF by Venkateswarlu and Barrow my own interest in the spectra of polyatomic free radicals had been stimu- lated by certain astrophysical problems. Indeed it is quite remarkable to what extent in the field of free-radical spectroscopy there is mutual stimulation between astrophysics and chemistry. In the present lecture I shall restrict myself mainly to a discussion of the work of my own laboratory which means at the same time a restriction to very simple polyatomic free radicals.THEC3AND Sic RADICALS Ever since the spectra of comets have been observed a group of bands near 4050 A has been known. The emission of these bands is restricted mainly to the nucleus of the comet but there it is very prominent. While most of the other features of cometary spectra were readily identified as due to the various diatomic free radicals already mentioned the interpretation of the 4050 group remained in doubt for many years. In 1942 I succeeded in reproducing the 4050 group in the laboratory under conditions which were suggested by the assumption that this group is due to the CH2 radical.3 However even though the production of these bands in a discharge tube seems to require the presence both of carbon and of hydrogen it was shown in 1949 by Monfils and Rosen4 that the substitution of deuterium for hydrogen produced a spectrum identical with the original one.The absence of any isotope shift in the spectrum (conked in our laboratory under much higher resolution by Douglas) of course ruled out any possibility that the spectrum was due to CH,. In order to ascertain whether carbon was present in the molecule responsible for the 4050 group Douglas5 investigated this spectrum using 50% and 100% 13C. In the experiments with 50% 13C the principal band was found to be replaced by six bands suggesting that the molecule responsible contains three carbon atoms.The simple fine structure of the principal band shows that the molecule is linear and the intensity alternation in this band in pure 13Cand the absence of alternate lines in 12C give further proof that the molecule responsible is the free C3 radical. The spacing in the rotational structure leads to a C-C distance of 1-28A. Unfortunately the remainder of the 4050group is so complex that in spite of con- siderable work in various laboratories it has not been analysed in detail. Therefore the vibrational frequencies of C are not known. The C3 molecule is a radical which up to the time of its identification by spectroscopic means had not even been postulated in any chemical reaction. How- ever since that time its occurrence in the explosion of acetylene-oxygen mixtures has been demonstrated by means of its absorption spectrum by Norrish Porter and Thrush6 and confirmed by Ramsay.It has also been found in the photolysis of diacetylene.' Furthermore it has been observed by Garton* in a carbon tube furnace at a temperature of 2800"c and has been shown by mass-spectrometric techniques to be one of the principal molecules evaporating from Venkateswarlu Phys. Rev.,1950. 77. 676. Laird Andrews and Barrow Trans.' Faraduy SOC.,1950 46 803. Herzberg. Astrophys. J. 1942 96 314. Monfils and Rosen Natrtre 1949 164 713. Douglas Astrophys. J. 1951 114 466; Clusius and Douglas Canad.J. Phys. 1954 32 319. Norrish Porter and Thrush Proc. Roy.SOC.,1953 A 216 165. Callomon and Ramsay Canad. J. Phys. 1957,35 129. * Garton Proc. Phys. SOC.,1953 66 A 848. graphite at high temperature^.^ By means of its emis- sion spectrum C3 has been observed to be present in various flames.lo Finally there is fairly strong evidence that C3 occurs in the atmospheres of low- temperature carbon stars.ll More recently a radical similar to C3 has been found in the spectra of stars as well as in the labora- tory namely the radical Sic,. For many years a series of very strong absorption bands was known in the spectra of a few carbon stars.12 These bands were reproduced a few years ago at Ottawa by Kleman13 in a carbon tube furnace charged with pure silicon at a temperature of 2200" c. The conditions of pro- duction of this spectrum make it certain that the molecule responsible contains both carbon and silicon.The vibrational structure of the spectrum together with the very narrow rotational structure make it extremely likely that the carrier of this spectrum is the free Sic molecule in its unsym- metrical form :Si-C-C. Again this molecule has later been observed by mass-spectrometric means14 in the evaporation of silicon carbide at high temperatures although the mass spectrum can of course not dis- tinguish between symmetrical and unsymmetrical Sic2. FLASH NH, PHOTOLYSIS It is natural to attempt to observe the absorption spectra of free radicals during the photodecomposi- tion of an appropriate parent compound. Thus early attempts were made in 194243 to observe the spectrum of CH (at that time believed to be the 4050 group) in absorption in photo-decomposed keten (CH,=CO).The failure of these early experiments could be accounted for by either of two reasons the stationary concentration of the CH radicals was in- sufficient to produce an absorption spectrum or the spectrum was not in the region investigated. As it turned out in this specific case both reasons applied. In order to overcome the first difficulty that is in order to obtain much higher concentrations of free radicals than are obtainable in ordinary continuous photolyses a flash-photolysis technique was deve- loped first by Norrish and Porter15 in Cambridge and independently in Ottawa16 and Pasadena.17 As shown by Fig.1 [facing p. 1231 in this technique an absorption tube T containing the parent substance PROCEEDINGS is irradiated by a discharge tubeP through which a condenser of high capacity (up to 600 PFat a voltage of up to 10,OOO v) is discharged. The extremely bright flash of short duration (20-1000 psec.) pro-duces momentarily a high concentration of photo-decomposition products that is free radicals in the absorption tube. During this time the light from a second flash tube is sent through the absorption tube to the slit of a spectrograph. The second flash tube is a fairly narrow tube which gives the con- tinuous spectrum necessary for obtaining an absorp- tion spectrum. A considerable increase in the sensitivity of detec- tion or in other words in the intensity of the resulting absorption spectra can be obtained by the use of the technique of multiple traversals through the absorp- tion tube first described by Whitels and further developed in our 1ab0ratory.l~ This technique makes it possible to have an absorbing path in the photo- decomposed gas of 40or even 100 metres when the absorption tube is only one metre long.Two other important assets in our experiments were the avail- ability of high spectroscopic resolution and the pos- sibility of extending the observable region into the vacuum-ultraviolet. High resolution is necessary if one wants to detect very narrow absorption lines. Such narrow absorption lines can often not be ob-served at all with inferior resolution.The use of high resolution proved decisive in the first example of a polyatomic free-radical spectrum that we found. In the past the so-called cc-bands of ammonia occurring in oxy-ammonia flames in electric discharges through ammonia in cometary spectra and in fluorescence produced by far-ultra- violet light in ammonia had been ascribed to various molecules NH NH, NH, N,H4 and others although NH seemed the most likely of these. Since photochemical investigations indicate that ammonia photodecomposes by the reaction NH + hv+NH2 + H it seemed worthwhile to attempt a flash-photolysis experiment of the type described using ammonia as parent compound. If the a-bands are in fact due to NH they should occur in absorption in such an experiment.This actually turned out to be the case as shown by Fig. 2 which gives the spectrum of an Chupka and Inghram J. Phys. Chem. 1955 59 100; Thorn and Winslow J. Chem. Phys. 1957,26 186. loDurie Proc. Roy. Soc. 1952 A 211 110; Kiess and Bass J. Chem. Phys. 1954 22 569. l1 McKellar Swings and Rao Monthly Notices Roy. Astron. Suc. 1953 113 571. l2 McKellar J. Roy. Astron. Soc. Canada 1947 61 147. l3 Kleman Astrophys. J. 1956 123 162. l4 Drowart De Marie and Inghram J. Chem Phys. 1958,29 1015. l5 Norrish and Porter Nature 1949 164 658. l8 Herzberg and Ramsay Discuss. Faraday SOC.,1950 9 80; Ramsay J. Chenz. Phys. 1952,20 1920; Callomon and Ramsay Canad. J. Phys. 1957,35 129; Ramsay Ann. New York Acad. Sci. 1957,67,485. l7 Davidson Marshall Larsh jun.and Carrington J. Chem. Phys. 1951 19 1311. lS White J. Opt. SOC. Amer. 1942 32 285. Bernstein and Herzberg J. Chem. Phys. 1948 16 30. APRIL1959 oxy-ammonia flame and of the flash photolysis of ammonia taken with the same 21-ft. grating spectro- graph. Nearly every line of the absorption spectrum of photodecomposed ammonia is matched by a line of the emission spectrum of the flame. In order to be absolutely certain about the carrier of the spectrum experiments were carried out both with ammonia containing 16N and with deutero- ammonia. The former gave lines that were shifted by small amounts from the lines obtained with ordinary ammonia while the latter gave a completely different spectrum as would be expected because of the large change of mass.There is therefore no question but that both nitrogen and hydrogen are present in the molecule responsible for this spectrum. The NH spectrum is a many-line spectrum that is it consists of a large number of single lines whose arrangement in the form of bands is quite indistinct. Without the use of high resolution this spectrum would not have been observed in absorption at all. The measurement and analysis of the NH2 spectrum was carried out by Ramsay20 and that of ND2 by Ramsay and Dressler.,l It took several years before the clue to the analysis of this extremely complicated spectrum was found but this has now been ac-complished and a fairly complete analysis is avail- able. This analysis confirms in every way the assumption that the molecule responsible is the free NH radical and in addition supplies information about the structure of NH both in the ground state and in the excited electronic state.In the ground state the molecule is non-linear with an angle of 103-5". In the upper state it is linear. Table 1 gives the geometrical data that have been obtained. In the TABLE 1. Structure of the NH radical LHNH Ground state 1 -024 103" 23' Excited state 0.976 180" lower state the molecule is somewhat similar in struc-ture to H,O that is it is an asymmetric top and this asymmetry accounts for the complexity of the spectrum. Because of the large change of angle in going from the ground state to the excited state there is a long series of absorption bands corresponding to increasing excitation of the bending vibration in the upper state.This accounts for the large extent of the spectrum. Unfortunately it has not yet been possible to determine the vibrational frequencies of the molecule in the ground state. A band system similar to that of NH was found by Ramsay, in the flash photolysis of PH,. Large isotope shifts occur when PH is replaced by PD and therefore there is little question that the spectrum is due to the PH (or PD,) radical even though a fine-structure analysis has not yet been accomplished. THE RADICALS HCO AND HNO As mentioned before the hydrocarbon flame bands which occur in almost all hydrocarbon flames in the region 2-100 A have been ascribed for many years to the HCO molecule (Vaidya).It was thought that by an investigation of the flash photo- lysis of suitable parent compounds it would be pos- sible to obtain these bands in absorption and thereby establish with more certainty whether or not they are due to HCO. However in the region of these hydro- carbon flame bands no absorption was found in the flash photolysis of H,CO CH,.CHO and (HCO),. On the other hand a series of extremely weak bands were observed in the red and the green part of the spectrum.= In this case the use of multiple traversals in the absorption tube proved invaluable in increas- ing the intensity of the bands and making it possible to obtain a complete analysis.24 In Fig. 3 one of the bands observed in the flash photolysis of CH,CHO and the corresponding band found in that of CD,CDO are shown.The bands have an extremely simple structure consisting of a single P Q and R branch. There is a long progres- sion of these bands for both isotopic molecules as shown schematically in Fig. 4. The existence of a large isotope shift in going from the normal to the deuterated parent compound shows that the mole- cule involved does contain hydrogen. The narrow- ness of the fine structure immediately shows that there must be two heavier atoms present in addition to hydrogen and the various parent compounds studied leave no doubt that these two other atom are C and 0,even though this has not been checked by isotopic substitutions. That the radical really is HCO is in agreement with a good deal of photo- chemical work on H,CO CH,CHO and (HCO),.The very simple fine structure of the HCO absorp-tion bands might lead one to believe that the mole cule is linear in both the upper and the lower state. However the observation of a long progression of bands corresponding to the bending vibration and of a very large combination defect between the Q branches on the one hand and the P and R branches on the other proves conclusively that the molecule is non-linear in the ground state and linear in the ex- cited state. The geometrical parameters obtained in both states are given in Table 2. 20Ramsay,J. Chem. Phys. 1956 25,188; Mem. SOC.Roy. Sci.Liige 1957,18,471. 21 Ramsay and Dressler J. Chem. Phys.1957 27 971 ;Phil. Trans. in the press. 22 Ramsay Nature 1956 178 374. 23 Ramsay J. Chem. Phys. 1953,21,96Q. 24 Herzberg and Ramsay Proc. Roy. SOC.,1955 233,34. PROCEEDINGS TABLE 2. Structure of the HCO radical r,(CH) r,(CO) LHCO Ground state (1.08 8,assumed) 1.19 A 119" 30' Excited state (1.07 8,assumed) 1.183 A 180" One may ask how the spectrum can be so simple if the lower state is non-linear since that implies that the molecule is an asymmetric top in the lower state. However even in the bent configuration the mole- cule is still approximately a symmetric top. Only one or two K values of the lower state can combine with a given I value (vibrational angular momentum) in the upper state. Even with this simplification one would still expect a number of overlapping sub- bands where only one is observed.One has therefore to assume that energy levels with I > 0 in the upper state are strongly predissociated and thus not ob- served under high resolution. This explanation was confirmed by the observation under low resolution of diffuse absorption bands halfway between the sharp bands (see Fig. 4). These diffuse bands ap- parently correspond to I = 1 in the upper state which alternates with I = 0 as the lowest sub-level for each value of the vibrational quantum number of the bending vibration. The diffuseness of some of the bands means of course that the dissociation energy D(H-CO) of HCO must be less than the energy corresponding to these bands that is less than 37-7 kcal./mole.The existence of diffuse bands at a comparatively long wavelength in a simple molecule like HCO may have an interesting astrophysical application. A number of diffuse interstellar lines are known which have so far defied all attempts at identification. While none of them agrees with the diffuse bands of HCO it is at least possible that another triatomic molecule of the type HXY might also have a low dissociation energy and thus might be responsible for the diffuse interstellar lines. If so and if this molecule could be found it would clearly throw considerable light on the chemistry of the interstellar medium. Until now no connection has been found between the HCO absorption bands and the hydrocarbon flame bands.The flash-photolysis investigations do not exclude the possibility that the hydrocarbon flame bands also belong to HCO and have the same lower state as the absorption bands but the upper state must certainly be different. The upper state of the flame bands must be much higher and in it the molecule is in all probability bent thus accounting for the great complexity of these bands. When the DCO spectrum was first investigated an additional band was found at 7550 A with a very different structure from the other DCO bands. The fact that it was impossible to find the analogue of this band for HCO led to the conclusion that it must have been due to an impurity introduced in the pre- paration of CD,.CDO. It was shown by Dalb~~~ at Ottawa that the 7550 A band mentioned is actually due to DNO.He did find the corresponding band of HNO by studying the photodecomposition of nitro- ethane and of nitromethane and also in the photo- reaction between NH and NO. It appears probable that the original band observed in the photolysis of CD,CDO was due to the presence of C2D,-N02as an impurity. In addition to the band at 7550 A two further much weaker bands were found at 7900 and 6850 A corresponding to the excitation of the NH bending and NO stretching vibrations in the upper state. The isotope shift again proves that there is only one hydrogen atom present in the molecule. We have not carried out experiments with 15N or l80 but there can be little doubt that nitrogen and oxygen are the other two atoms involved in the molecule under consideration.The fine structure of the two bands is fairly readily analysed. Fig. 5 shows the principal band of HNO. It consists of a number of sub-bands each with simple P,Q and R branch structure. This difference from the HCO spectrum shows that the molecule is non-linear in both the upper and the lower state. The geometrical data derived for this molecule from the spectrum are given in Table 3. TABLE 3. Structure of the HNO radical r,(NH) r,(NO) LHNO Ground state 1.063 A 1.212 8 108" 35' Excited state 1.036 8 1-2418 116" 15' The lifetime of HNO under the conditions of the photolysis experiments was found to be approxi- mately 0.1 second while HCO has a lifetime of only 50 microseconds.This difference is presumably con- nected with the fact that HCO has an odd electron while HNO has not. THERADICALS NCO AND NCS A few years ago Holland and Style26 observed a group of bands near 4400 A in the fluorescence of ethyl isocyanate excited by far-ultraviolet light. The same bands occurred also in a high-frequency dis- charge through the same compound. similarly in methyl thiocyanate a number of bands were ob-served in the region 3750-4850 8 under similar conditions. More recently Dixon and Ramsay26 at Ottawa have obtained these bands in absorption in the flash photolysis of various compounds containing the NCO and the NCS group respectively. The obser- vation of these bands in absorption in flash photolysis confirms the assumption by Holland and Style that f6 Dalby Canad.J.Phys. 1958 36 1336. 26 Holland Style Dixon and Ramsay Nature 1958 182 336. APRIL1959 121 ~~ ~~ ~ these bands are due to the free NCO and NCS radicals. For NCO in absorption an additional band system was found in the region 2600-3200 A. On the Ottawa plates taken under high resolution the fine structure of all the bands is very well resolved as shown for one of the bands of thelong-wavelength system of NCO in Fig. 6. These bands are shaded to the violet while all the other bands are shaded to the red. The structure of these bands shows conclusively that both molecules are linear in all observed states. The NCO radical is isoelectronic with the C02* molecule. It is therefore not surprising that the ground state of both NCO and NCS are found to be of the same type as that of CO,+ namely 2n.Also the excited states are similar to those m CO,+. While a preliminary value for the moment of inertia of NCO has been obtained it is not possible to derive from it the two internuclear distances; this would require the observation of the spectrum of an iso- topic molecule. It appears probable that the NCO and NCS free radicals observed in the flash photolysis are primary products in the photolyses that is arise from reactions such as CH,-NCO + hv -+ CH + NCO THERADICAL CH The discovery of a spectrum of the free CH radical was one of the principal aims in our pro- gramme of investigations of free-radical spectra.It took many years before a CH spectrum was in fact observed. Many unsuccessful attempts were made in the region from 9000 to 2300 A by studying the photolysis of numerous compounds which were thought to be suitable using different pressures long absorbing paths and high resolution. Since CH is isoelectronic with NH one might have expected a transition in the visible region. However if the nature of this predicted electronic transition is considered following Walsh,2' it is found that it would be a for- bidden transition if the molecule were planar. This may account for the failure to find any absorption in the visible and the near-ultraviolet region. Therefore experiments were started in the vacuum-ultraviolet where one finds strong Rydberg transitions at least near the ionisation limit for almost all molecules that have stable excited states near this limit.How- ever the difficulties of working with flash photolysis in the vacuum-ultraviolet are considerable. Because of the strong absorption of the parent molecules only extremely low pressures (of the order of 0.01 mm.) can be used and at these low pressures the absorp- tion of the radical that is wanted may be too weak. Another difficulty is for example scattered light from the visible and the near-ultraviolet region which may obscure the very much weaker continuum in the vacuum-ul traviolet. In spite of these difficulties a Rydberg series was eventually observed= in the flash photolysis of di- methylmercury.This Rydberg series starts at 1500 A and consists of five members some of which may be seen in Fig. 7. Later the same Rydberg series was found in the flash photolysis of a number of other parent compounds such as CH,CHO (CH,),CO (CH,),N, CH,I and CH,Br. Since there is strong photochemical evidence that CH radicals are pro- duced in the photodecomposition of all these mole- cules it appeared extremely likely that the observed Rydberg series is in fact due to free CH,. A corres-ponding slightly shifted series was found in deuter- ated dimethylmercury (see Fig. 7). Finally under the conditions under which this Rydberg series appeared most prominently an additional diffuse band was observed at 2160 which makes the observation of CH much simpler since it does not require the use of a vacuum-spectrograph.The corresponding band of the fully deuterated compound is at 2144 8 and shows some fine structure. Spectra were also taken with intermediate concentrations of deuterium and show clearly (see Fig. 8) that four isotopic species are present as would be expected if the molecule responsible is CH (that is CH, CH,D CHD2 and CD,). From the limit of the Rydberg series one obtains a precise value for the ionisation potential of the CH radical. One finds 9.840ev for CH and 9.832ev for CD, results that agree most satisfactorily with the (less accurate) mass-spectrometric value of 9-90 f 0.1 ev. The fine structure of the CD band (see Fig. 8) was further investigated. It shows clear evidence of a slight intensity alternation.According to theory such an intensity alternation in what appears to be a so-called parallel band can occur only if the molecule is planar in at least one of the two states involved. Since there is only a single prominent band accom- panied by two or three very much weaker bands it follows from an application of the Franck-Condon principle that if the molecule is non-planar in one of the two states it cannot deviate from planarity very much in the other. However the weaker bands are difficult to interpret on the assumption that both states are planar. We therefore come to the provi- sional conclusion that the CH molecule deviates slightly from a planar configuration in the ground state while it is planar in all the observed excited 27 Walsh J.1953 2260. 28 Herzberg and Shoosmith Canad.J. Phys. 1956 34,523. z 22 PROCEEDINGS states. This would be in agreement with the theoret- ical prediction that the CH3+molecule is planar. Further analysis of the rotational structure of the CD band at 2144 A yields an approximate value for the moment of inertia of CD in the ground state about an axis at right angles to the symmetry axis. This value is 5.769 x g. ern.,. It is of course not sufficient to determine completely the structure of the CD molecule and the lack of a fine structure in CH makes it impossible to obtain a second moment of inertia that would allow a complete structure determination. However the following are two values for the CD distance corresponding to two different assumptions for the angle fi of the CD bond with the symmetry axis /3 = 90" ro(CD) = 1.072 A fi = 75" ro(CD) = 1-0618 For smaller values of /s ro would be still smaller.In view of the known C-D distances in other molecules it seems extremely unlikely that r, for CH could be less than 1.060 and therefore one will have to con- clude that the angle is larger than 75" in agreement with the tentative conclusions already mentioned. Even for a planar configuration (18 = 90') the C-D distance comes out distinctly smaller than in methane where it is 1.093 A. The smallness of the C-D distance may be related to the high C-H bond dissociation energy in CH compared with that in CH,. Extensive experiments have been carried out in order to find at longer wavelengths the predicted for- bidden transition in CH, but without success.If this transition could be found it would enable one to obtain much more definite information about the geometrical structure of CH, in particular about the question of the exact height of the pyramid in the ground state. The only hope at present of obtaining this information lies either in the study of the infrared spectrum of CH or in the study of the analogue of the 2160 A band in CT,. Both investigations are being attempted at Ottawa. THERADICALSCH AND C,H As mentioned earlier when the 4050 group was still believed to be due to CH, various attempts were made to obtain it in absorption under conditions under which a high concentration of CH could be expected.When it became clear that the 4050 group is not due to CH, attempts were made to study carefully the whole accessible region of the spectrum down to 1200 A in order to see whether or not the true CH spectrum could not be found. For most of th experiments the photolysis of keten (CH,=CO) was used for the production of the CH radicals. Since the ionisation potential of CH according to mass spectrometric data is approximately 11 -9 ev,29 that is appreciably higher than that of CH, it was expected that any Rydberg series of CH would be at correspondingly shorter wavelengths. Unfortu- nately at these short wavelengths the parent absorp- tion is very strong and very low pressures (less than 10 p) had to be used.No absorption that may be due to CH was found in this region. This failure may be due to the fact that the partial pressure of the CH radicals was too low or to the fact that there is no simple Rydberg series in CH,. At somewhat longer wavelengths fairly strong absorption bands of CO were found in absorption in consequence of the photolysis showing that photo- decomposition actually does occur. Some of the CO bands arise from vibrationally excited CO molecules (up to V" = 3) indicating either that the CO mole-cules formed in the photolysis are vibrationally ex- cited or that the temperature resulting from the flash photolysis is so high that a high vibrational tempera- ture results. Fig. 9 shows a small section of the spectra obtained.One of the hot bands is visible in the photolysis spectrum but not in the spectrum of the residue. Also a high rotational temperature is clearly indicated by the much greater length of the branches during photolysis than afterwards. In order to reduce the temperature during photolysis the keten was mixed with helium argon or hydrogen in the ratio 100 1. The photolysis spectra under these conditions showed much shorter branches in the CO bands corresponding roughly to room temperature but the "hot" bands with V" = 1 and 2 were still present; that is CO like other mole- cules loses its vibrational energy only very slowly. Since the converse also appears to hold that is CO acquires vibrational energy very slowly it follows that the first alternative mentioned previously applies namely that the CO molecule is formed with vibrational energy in the photolysis of keten.In the visible and the near-ultraviolet region much longer paths could be used for the study of the CH absorption because of the much lower or absent absorption of the parent molecule. However in spite of considerable time and effort no discrete absorp- tion was found except the CH band at 3143 A which occurred both in pure CH,=CO and in the 100 1 helium-keten mixture as shown in Fig. 10. The presence of CH may be ascribed to the photodecom- position of CH in a weak continuous-absorption region. The failure to find a discrete absorption of CH says of course nothing about the existence and stability of the free CH radical.It is entirely possible that CH has only continuous absorptions in the Langer Hipple and Stevenson J. Chem. Phys. 1954,22 1836. FIG.2. Etiiixvion spectrrrtn oj an oxy-anmonia flame coniparcd with tire absorption spcctrcm i!f photo&-coniposed anitironiu obtained by flash photolysis. (Both spectra wwe taken with the sattie 2 I :ft. grating. Only two stiiull sections of the very extensive spectrum are shown. Soins detail has been lost in reproduction.) FIG.3. Absorption bands of HCO arid DCO obtained in the flash photolysis of acetaldehyde arid cleutero-acetaldehyde. FIG.5. Principal crbsorption band of HNO obtaiticrt in the flash photolysis FIG.6. Strongest absorption band of‘ nitrotnetliarre.( Tlic nrrttibers at the top of each strip indicate the K’-K” of NCO obtained in the flash valires of‘ tlic srrh-hanrls. Tlic spectrunr M?USobtained in the first order of a photolysis of HNCO. 3 5-ft. zruting spcct ro~~rapli .) FIG. 7. Rydberg series of bands of CH and CD obtained in the flash photolysis of dimethylmercury and deuterodimethylmercury. (For comparison the spectrum of the source in vacuum and the absorption spectrum of the residue are shown. The Rydberg series of CH is marked at the bottom.) CD FIG. 8. The 2150 A absorption band of free methyl obtained with various concentrations of deuterium in the parent compound (acetone). FIG.9. CO bands in the flash photolysis of keten. (CO is weakly present in the absorption spectrum of the parent compound because of photolysis produced by the light source.The 1-1 band of CO is present only during photolysis not in the spectrum of the residue.) Slit I I 600pF so0 v 4z C FIG.1. Flash-photolysis apparatus. FIG.10.CH band in the flash photolysis of a 100 :1 helium-keten mixture. (The white spectral lines are emission lines in the source.) APRIL1959 123 accessible region and these are far more difficult to establish and identify with certitude than discrete absorptions. It may be recalled that the CH4 mole- cule has continuous absorptions only in the whole region so far investigated (other than the infrared). Thus if CH were a free radical it would be extremely difficult to detect spectroscopically.Another molecule in which we have been interested for some time and whose spectrum we have not yet found is the C2H radical. A promising parent com- pound for obtaining free C2H appeared to be pro- piolaldehyde (HC2CHO) since the aldehyde group gives rise to a fairly strong absorption in the near ultraviolet which will in all probability lead to photo- decomposition and since it is easy to establish the occurrence of photodecomposition by an observa- tion of the known HCO bands. The HCO bands were indeed observed in the photolysis of propiolaldehyde but in spite of much effort no discrete absorption that might be ascribed to C2H was found. That C2H actually was present under the conditions of the experiment follows indirectly from the observation of acetylene (C2H& and particularly diacetylene (C,Ha absorption bands in the spectrum of the 30 Norrish Proc.Chem. SOC. 1958 247. Ramsay Adv. Spectroscopy,in the press. residue. It appears that the only way in which di- acetylene can be formed on photolysis of propiol- aldehyde is by the recombination of two C2H radicals formed in this photolysis. Experiments were also carried out with a 100 1 mixture of helium and propiolaldehyde. In this mix- ture the spectra of C2H2and C4H2 occur even during the photolysis and not only in the residue as in the case of the undiluted gas. Moreover a few “hot” bands of C,H2 have been found under these conditions. CONCLUSION A large amount of work on free polyatomic radicals has been carried out in other laboratories.It is not possible to give here an adequate summary of it but reference may be made to Norrish’s Liversidge lecture30 and to a forthcoming review article by Ram~ay.~l Further work in this field will be of importance both from the point of view of the chemist interested in establishing the presence of free radicals in various chemical reactions and from the point of view of the physicist interested in the structure of simple molecules. COMMUNICATIONS Perpendicular Conjugation in Some Octahedral Metallophthalocyanine Derivatives By J. A. ELVIDGE and A. B. P. LEVER CHEMISTRY LABORATORIES COLLEGE S.W.7) (ORGANIC RESEARCH IMPERIAL LONDON WEhave prepared true octahedral (6-co-ordinate) metal complexes in the phthalocyanine series for the first time.Only a few 6-co-ordinate solvates such as the diamagnetic dipyridineferrous phthalocyanine,l have been obtained before. Of our new compounds dihydroxymanganesew phthalocyanine and aquo-hydroxychromiumm phthalocyanine are unique in being dibasic acids. No examples are known in in- organic chemistry of the displacement of a proton from a hydroxy-group attached directly to a chelated metal atom. Evidently there is considerable stabilisa- tion of the anions of the above phthalocyanine derivatives and this may be achieved through sharing of the negative charge with the aromatic .rr-electron system. For this to occur there must be conjugation between the mutually perpendicular phthalocyanine and oxy-groups.Perpendicular conjugation has not previously been considered but appears possible through pxd orbital overlap. Barrett. Frve. and Linstead. J.. 1938. 1157. Manganous phthalocyanine2 (peff34.34B.M. 6+ 9) with methanolic potassium cyanide yields hydroxy- cyanomanganesew phthalocyanine PcMn(CN)OH (where Pc = phthalocyanine residue) in which the manganese is in the 6-co-ordinate quadrivalent state (perf3.88 B.M. 6 -5). Similarly methanolic sodium hydroxide yields dihydroxymanganesew phthalocyanine PcMn(OH), which on dissolution in the saturated reagent affords disodium dioxy- manganesew phthalocyanine [PcMnO2I2-2Na+ (perf 4.00 B.M. 6-49). Electrophoresis demonstrated the anionic character.Chromium phthalocyanines have now been authenticated for the first time.* Chromium phthalo- cyanine PcCr (unit cell 3-4 x 42.5 A by electron- diffraction) is readily oxidised but is stabilised by co-ordination with pyridine the magnetic moment (peff 3.16 B.M. 6 -35) confirming the 6-co-ordinate structure for the solvate PcCr,2py (py = pyridine). Barrett; Dent and Linstead J.‘ 1936 1719. See Figgis and Nyholm J. 1959 331 ;the present complexes obey the Curie-Weiss law over the range studied i.e. 110-295”~. ‘Cf. ref. 2 and Anderson Bradbrook Cook and Linstead J. 1938 1151. Chromiumm p hthalocy anine is obtained from chromic acetate and phthalonitrile as the4-co-ordinate hydrox- ide [PcCr]+OH- (peft. 4.03B.M. 8 -15) (migration of the coloured ion to the cathode).This with aqueous- ethanolic acid gives aquohydroxychromic phthalo- cyanine PcCr(H,O)OH which is not a loose hydrate but a 6-co-ordinate tervalent chromium complex (peff3.52 B.M. 8 -10) its reconversion into [PcCr]+OH- requires heating at 400°/10-6 mm. AquohydroxychromiumI" phthalocyanine with meth- anolic sodium hydroxide gives disodium oxyhydroxy- chromium111 phthalocyanine [PcCr(OH)0I2-2Na+ (peff3.91 B.M. 8 -155)" (migration of the coloured ion to the anode). Pyridine likewise dis- places two protons to yield the dipridinium salt [PcCr(OH)0I2-2pyH+ (pu,ff4.06 B.M. 6 -15) an alternative structure [PcCr(OH),]-pyH+,py is eliminated by the stability at 150"/20 mm. and the infrared absorption. The feasibility of n-bond formation by overlap of p-with d-orbitals (cf.A) has been discussed by Craig and Padd~ck.~ Because adjacent lobes of a dzz-or &-orbital have opposite signs full d px conjugation round a ring (i.e. degeneracy) is possible only with even numbers of the two contributing types of atom. A suitable metal atom M bound by the phthalo- cyanine ligand could therefore be conjugated only to one or other of a pair of nitrogen atoms at a time by d px overlap. A valence-bond partial representation is shown at (B). Only a small gain in resonance energy is likely from this conjugation because the four equivalent sets of canonical structures cannot be hybridised. Nevertheless this limited conjugation between metal and planar phthalocyanine ligand will allow of further conjugation with other groups at- tached pzrpendicularly to the metal.A p-orbital of PROCEEDINGS such a group could overlap with one of the d-orbitals already considered and give rise to a 7-bond to the metal along the z-axis. There would be perpendicular conjugation and attendant possibilities of resonance (cf. C). *fr\ These ideas make it possible to account for the stability (Le. existence) of the anions [PcMn0,I2- and [PcCr(OH)0I2- through a sharing of the negative charge on oxygen with the rest of the molecule as indicated by the partial formulae (D). This degeneracy through perpendicular conj uga- tion is possible only if dzz-and/or d,,-orbitals of the metal are vacant so it might be observed in octa- hedral complexes of titanium and vanadium besides chromium (spin-free) and manganese'" (spin-free) but not in those of the next transition metals iron cobalt nickel.The ultraviolet- and the visible-light absorption of the analogues of vitamin B, (B,,, etc.) are closely similar to one anotherY6 but those of the octahedral phthalocyanines show variations which are too large to be ascribed to inductive effects and are presumably a manifestation of the perpendicular conjugation. We thank Dr. I. S. Kerr for electron-diffraction measurements and the Department of Scientific and Industrial Research for a maintenance grant (to A.B.P.L.). (Received February 23rd 1959.) * The value of 8 renders uncertain the interpretation of peff but tervalency for the metal follows from other results.Craig and Paddock Nature 1958 181 1052. 6 Kaczka Wolf Kuehl and Folkers J. Arner. Chern. Soc. 1951 73 3569. Trimethyl-lead Hydride By R. DUFFY and A. K. HOLLIDAY (THEUNIVERSITY, LIVERPOCK) lead forms a very unstable hydride PbH methyl-lead borohydride Me,PbBH ; this decom- ALTHOUGH and stable tetra-alkyl derivatives PbR, no inter-posed to lead hydrogen and methyl diboranes. As mediate alkyl-lead hydrides [e.g. R,PbH] have an a1 ternative means of preparing trimethyl-lead hitherto been prepared. borohydride the reaction of trimethyl-lead chloride The reaction of tetramethyl-lead with aluminium and potassium borohydride in liquid ammonia has borohydride was shown1 to form the unstable tri- been investigated.The initial reaction expected here Holliday and Jeffers J. Inorg. Nuclear Chem. 1958 6 134. APRIL1959 is Me,PbCl + KBH4 -+ Me,PbBH + KCl and when equimolar amounts of reactants were used at -33" potassium chloride was precipitated. The sol- vent ammonia was then removed slowly at -78" without appreciable loss of volatile boron or lead compounds. When the residue was warmed to -5" some volatile material rapidly distilled off (over a very short path) and was all immediately condensed at liquid-nitrogen temperature. This material con- tained all the lead and no boron. No hydrogen or methane was produced during its formation and the remaining solid residue contained only potassium chloride and a substance of empirical formula BH,,NH,.The condensate melted at about -loo" without evolution of gas to a homogeneous colour- less liquid. The only known alkyl-lead compound of comparable volatility is tetramethyl-lead m.p. -27 ". All these facts point to the conclusion that the substance obtained is trimethyl-lead hydride formed by the reaction Me,PbBH -+Me,PbH + BH,,NH,. This was supported by the reaction (qualitatively and quantitatively controlled) with hydrogen chloride which yielded hydrogen and trimethyl-lead chloride in accordance with the equation Me,PbH + HCl +-Me,PbCl + H,. The trimethyl-lead hydride was thermally very unstable decomposing slowly above -100" to a red solid stable at -78" and methane (CH, Me,PbH = 1:2). This indicated that the red solid might be pentamethyl-dilead hydride 2Me,PbH + Me,Pb.PbMe,H + MeH.Further decomposition of the latter gave lead tetramethyl-lead hydrogen and methane in amountscorresponding to Me5Pb,H. The use of liquid ammonia as a solvent for metathetical borohydride reactions may be a useful general method of preparing hydrides since it is clear that the donor power of ammonia can effect fission of a BH group from a borohydride so leaving the hydride. (Received February loth 1959.) The Action of 6oCo-y-Radiation on Propene and Ally1 Alcohol in Aqueous Solution By P. G. CLAY,J. WEISS and J. WHISTON UNIVERSITY NEWCASTLE (KING'SCOLLEGE OF DURHAM ON TYNE,1) Propme.-When aqueous solutions of propene-oxygen (1:l; total press. 1 atm.) are irradiated with 6oCo-y-rays the main product is a hydroperoxide formed with the comparatively high yield of G 2.6 (G = molecules/lOO ev).The yield of the oxidation products associated with a break of C-C bonds was relatively small viz. G(forma1dehyde) 'L 0.4; G(formic acid) r,0-4; G(aceta1dehyde) r,0-4;and G(propiona1dehyde) r,0.2. Traces of acetone were also detected; in addition hydrogen peroxide is formed with G(H,O,) = 1-9. The total yields of the organic radiation products (hydroperoxides + aldehydes + formic acid) were somewhat higher at pH 1.2 (G = 4.0)than at pH 5.5 (G = 2.9). When the hydroperoxide in the irradiated solution was decomposed by ferrous sulphate propylene glycol was produced a good material balance was obtained between the hydroperoxide decom- posed and the glycol and ferric salt produced.This and some related observations suggest that the hydroperoxide is a hydroxyhydroperoxide CH,.CH(02H)*CH,*OH. When however aqueous solutions of propene- oxygen are irradiated in the presence of ferrous salt the major product was propionaldehyde which was formed in the relatively high yield of G r,20 which must be due to some chain process. AZZyZ Alcohol.-When allyl alcohol (IO-,M) was Clay Johnson and Weiss J. 1958 2175. irradiated in aqueous solution in the absence of oxygen acraldehyde and propylene glycol were formed both with a yield of G = 1.2. In the presence of oxygen (1 atm.) irradiation of allyl alcohol soh- tions at pH 1.2 gave acraldehyde in a yield of G = 1.2 glycollaldehyde G = 1.7 and formaldehyde G = 1.5 in addition to an organic hydroperoxide which was also formed with a yield of G = 2.Further in-vestigation has shown that here the glycollaldehyde and formaldehyde are secondary products from the decomposition of the organic hydroperoxide during the analytical procedure. From an investigation of the rate of liberation of iodine from iodide by the irradiated solution it appeared that two hydro- peroxides were formed. The most obvious route to acraldehyde would be by initial hydrogen abstrac- tion by the hydroxyl radicals (produced from the water) whereas formation of the hydroperoxide would presumably be initiated by addition of a hydroxyl radical to the double bond. The relatively high yields of the organic hydro- peroxides particularly from allyl alcohol suggest that in addition to hydroxyl radicals hydrogen atoms and/or HO radicals may b capable of adding to the double bond as in the reactions of ethylene previously studied.l One of us (P.G.C.) is indebted to the British Oxygen Company for the award of a Research Fellowship and another (J.Whiston) thanks the D.S.I.R. for a maintenance grant. I26 PROCEEDINGS The Structure of Diaboline By A. R. BATTERSBY and H. F. HODSON (THEUNIVERSITY, BRISTOL) THEtertiary base diaboline was first isolated as its crystalline hydrochloride from the bark of Strychnus diaboli Sandwith by King1 who assigned to it the molecular formula C21H2603N2. Bader Schlittler and Schwarz2 pointed out that the ultraviolet spectrum of diaboline is characteristic of an N-acylindoline; the alkaloid was proved to be an N-acetylindoline by acid-hydrolysis to deacetyl-diaboline which showed an indoline absorption spectrum.The Swiss workers tentatively suggested a crypt~phenoli&~ N-acetyl-7-hydroxyindolinestruc-ture for diaboline on the basis of their reported alkali-solubility of diaboline and deacetyldiaboline and the similarity of certain colour reactions of di-aboline to those of vomicine. It was however emphasised that the evidence for a phenolic hydroxyl group was by no means conclusive. Our re-examination of the tertiary bases from S. diaboli by partition chromatography5 of the hydro- chlorides on cellulose has so far given two crystalline alkaloids both N-acylindolines.The major alkaloid representing at least 0-1% of the dry bark has been shown to be diaboline by a rigorous comparison with the late Dr. H. King's materia1.l As noted by Witkop and Patri~k,~ the infrared carbonyl stretching frequencies of diaboline and its 0-acetyl derivatives are at variance with an N-acetyl- 7-hydroxyindoline structure for the base. It has now been shown that no phenolic hydroxyl group is present in diaboline and that deacetyldiaboline is in fact the Wieland-Gumlich aldehyde6 (I).The experi- mental evidence and reasoning which led us to make this direct comparison will be described in the full paper now in preparation. Considerable care was taken with the proof of identity because of the difficulties caused by dimorphism2 in this series; the comparison involved m.p.and mixed m.p. of the base m.p. 211-213" (Bader et aL2 record m.p. 197-198") the picrate m.p. 235-237" and the methopicrate,* m.p. 234-234". Also the infrared spectra of the base and its hydrochloride and metho- chloride are superimposable upon those of the Wieland-Gumlich aldehyde and the corresponding derivatives; the corresponding salts in the two series have identical R values in two solvent systems; the rotations of deacetyldiaboline and the Wieland- Gumlich aldehyde are identical. Diaboline is thus N,-acetyl Wieland-Gumlich aldehyde (11) C2,H2,03N2 and a reassessment of the published172 analytical data for diaboline and its salts shows that they are in moderate agreement with this formulation.However the figures given for deacetyldiaboline by Bader et aL2 lead directly to the correct molecular formula for this base. The Wieland-Gumlich aldehyde (I) has been isolated as the base' (caracurine-VIJ) and as the Nb-metho-derivative* (hemitoxiferine-I) from Strych-nus tuxifera. Diaboline (11) is thus intermediate in OH complexity between these alkaloids and strychnine (JII) and its occurrence in S. diaboli has considerable interest in connection with the later stages of the proposed biogenetic schemeg for strychnine. Grateful acknowledgement is made to the Medical Research Council for financial support and to Dr. J. Walker who very generously provided the diaboline and S.diaboli bark for our work. (Received February 25th 1959.) * The m.p. of this salt was incorrectly reported8 as 134-136'. King J. 1949 955. Bader Schlittler and Schwarz Helv.Chim. Acta 1953 36,1256. Witkop and Patrick J. Amer. Chem. SOC.,1954 76 5603. * Orazi Corral Holker and Djerassi J. Org. Chem. 1956 21,979; Snyder Strohmayer and Mooney J. Amer. Chem. SOC.,1958 80 3708. Schmid Kebrle and Karrer Helv. Chim. Acta 1952 35 1864. Wieland and Gumlich Annalen 1932 494 191; Wieland and Kaziro ibid. 1933 506 60;Anet and Robinson J. 1955 2253. Bernauer Pavanaran von Philipsborn Schmid and Karrer Helv. Chim.Acta 1958 41,1405. Battersby and Hodson Proc. Chem. SOC.,1958 287. Woodward Nature 1948 162,155. APRIL1959 127 The Structure and Biogenesis of Trichothecin By J.FISHMAN, E. R. H. JONES,G. LOWE,and M. C. WHITING (DYSONPERRINSLABORATORY UNIVERSITY) OXFORD FREEMAN,GILL and WARING~have recently described an extensive investigation of the structure of trichothecin an antifungal metabolic product of Trichothecium roseurn Link. Their array of experi- mental evidence was originally explained in terms of structures (I) and (Ia). Through a generous gift of material and access to their unpublished results we have been able to continue the study and finally clinch structure (I). Mild hydrolysis of trichothecin yielded trichothec- olone (II) oxidised with chromic acid to tricho- thecodione (111) which with 10% aqueous R *\ -O Me*CHrCH.CO.O u u/ partially reduced acid (IX) gave p-xyloquinol (VIII) along with the product (VII).It had been hoped at the outset of our investiga- tions that a further study of isotrichothecolone,' an alkali-isomerisation product of trichothecolone (II) would help to throw light on the trichothecin structure. However complex rearrangements occur in this reaction and also in the formation of another product allodihydrotrichothecolone,obtained when zinc dust is added to the alkaline reaction mixture. Nevertheless their respective structures (X)and (XI) had profound implications for trichothecin stereo- (I4 6 0 0 0- H-7HO- (IV) (V) sodium hydroxide at 100" was found1 to give a mixture of p-xyloquinone and the quinol. No frag-ment from the other carbocyclic ring could be isolated.An impure compound C8H1202was ob- tainedl analogously from the acid (IV)and it seemed that further investigation of this acid offered the best opportunity of obtaining and separating identifiable fission products. Under carefully controlled condi- tions treatment of the acid (IV) with alkali yielded the more stable stereoisomer2 of 2,5-dimethylcycio- hexane-1,4-dione (VI) and 2-methyl-3-oxocyclopent-1-enecarboxylic acid3 (VIZ). The racemisation of five asymmetric centres is noteworthy. The process can be regarded as a reverse-Michael reaction (cf. IV) facilitated by the release of a considerable amount of compression energy associated with the stereochemistry indicated and followed by cleavage (cf.V) at the /%position to the carbonyl group. Under the anaerobic conditions chosen the intermediate hydroxycyclohexenone could not become aromatic and hence yielded the diketone (VJ) but the only Freeman Gill and Waring J. 1959 1105. 0)7JO Ho2cQ chemistry (cf. I) which is determined with the exception of the one centre. The structure (I) for trichothecin does not obey the classical isoprene rule but it can be derived from a tri-isoprenoid precursor by a 1,3-or a double 1,2-methyl group migration. The accompanying but structurally unrelated metabolites rosololactone and rosenonolactone have been found4 to conform to the biogenetic isoprene rule.5 [2-14C]Mevalonic lactone (XII) was incorporated (ca.0-5%) into trichothecin (I).Degradations mainly employing the above fission reaction indicate a dis- tribution of radioactivity as in (XIV) and hence support a biogenesis of the carbon skeleton from three mevalonate units (XIJI) involving two succes- sive 1,2-methyl group shifts.[A 1,3-shift would demand a labelled atom either in the methylene group of the four-membered ring or in a methyl group of the diketone (VI); both possibilities were eliminated.] Stephenson untmblished work. * Sufter and Schfittler HeEv. Chim. Acta 1949,32,1860. We are indebted to Dr. Schlittler for a sample of the acid obtained by degradation of picrotoxin. Birch Rickards and Smith Proc. Chem. SOC.,1958 223; Britt and Arigoni ibid. 1958 224. Ruzicka Eschenmoser and Heusser Experientiu 1953,9,357 ;Cornforth Cornforth Popjak and Youhotsky-Gore Biochern.J. 1958 69 146. None of the mevalonate radioactivity was found in the cis-crotonate group whereas when the fungus was grown on a medium containing sodium [1-14C] acetate (incorporation ca. 0.2%) 95 % of the activity resided there. Acetate is clearly used up much faster CO,H (XI 0 (XI11) PROCEEDINGS for crotonate than for mevalonate synthesis and our mevalonate experiment provides further support for the suggestion6 that the acetate -+ mevalonate process is irreversible. (Received February 19th 1959.) 0m (XIV) OCO-CH=CHM~ Birch English Massy-Westropp and Smith J. 1958 369. The Use of Spherical Samples in High-resolution Nuclear Resonance Spectroscopy By P. HIGHAM and R.E. RICHARDS (PHYSICAL LABORATORY, CHEMISTRY OXFORD) ITis common practice to use cylindrical sample tubes in high-resolution nuclear resonance spectroscopy because these are easy both to fill and to use. There are however two disadvantages associated with this type of sample container first that a considerable volume of the sample is unused because the radio frequency coil is usually much shorter than the liquid column; secondly that if an external reference sample is used in a central or annular capillary the chemical shifts measured must be corrected for the difference of susceptibilities of the two liquids. The first disadvantage can be overcome to some extent by using a plug in the glass cylinder made e.g. of Ny1on.l The second disadvantage can be over- come by using an internal reference compound al- though this method can cause serious difficulties when aromatic solvents are being used.Both of these difficulties are overcome by the use of spherical sample holderse2 Glass ampoules with spherical cavities3 are difficult to make and therefore expensive; they are also very difficult to fill and empty especially if solutions of involatile solutes are used. Small glass bulbs can be blown and fitted inside a normal cylindrical sample tube,4 but it is almost impossible to obtain perfectly spherical containers in this way and the filling factor produced is inevitably low. Varian Associates. Tech. Bull. Vol. II. No. 3. 1959. A method of preparing spherical samples has been devised which both overcomes the difficulties described above and is relatively simple to carry out.A cylindrical sample tube is filled with a suitable gel. We have used a gel prepared from 10 g. of normal “laboratory gelatine” in 25 ml. of boiled-out water. A glass syringe is made by drawing a long spill on a piece of capillary tubing and a stiff rubber teat is fixed to the end of the tubing. The syringe is cali- brated by weight or with the aid of an “Agla” micro- meter syringe. The sample tube of gelatine is warmed to 35-40’ and held horizontally. A calculated volume (about 0.03 ml.) of the sample is injected into the gelatine at about the correct height from the bottom of the tube and the sample forms a bubble in the gelatine matrix.The tube is then rotated slowly while the gel cools and the tube tilted so as to obtain a perfectly spherical bubble at just the correct posi- tion. If the sample and gel are of the same density it is easy to adjust the position of the bubble in the tube by adding or removing gelatine from below the bubble by means of a second glass syringe. When the gel has cooled the sample tube can be used in the ordinary way except that only low spinning speeds can be permitted to avoid distortion of the bubble. This method has been used in the laboratory for studying samples when only small quantities are Andrew “Nuclear Magnetic Resonance,” Cambridge Univ. Press 1955 p. 78. Primas and Gunthard Helv. Phys. Acta 1957 30 3 15. Bothner-by and Glick J.Chem. Phys. 1957,26 1647. APRIL1959 available or when it was desirable to use an external standard without susceptibility correction. Its only limitations are that the sample must be immiscible and unreactive with water and that there is a strong resonance from the water in the gel which obscures a small region of the spectrum. The usual volume of sample used is 0.03 ml. and in general a minimum amount of solute of mole is required although with a low signal noise ratio smaller amounts can be studied. When this method is used the amount of material for a given signal :noise ratio is about one-third of that required in a cylindrical sample tube with a plug. The resolving power obtained is at least as good as roc..n 40 Mc./sec. which have been selected5 as the best available values measured in cylindrical samples. From the magnetic susceptibilities of the samples6 these shifts have been corrected for the effects of susceptibility by using the classical 27~~/3 bulk susceptibility effect for cylinders. The third column gives comparable values obtained by the gel- bubble method. These figures have been obtained by multiplying our measurements by 40/29.92 because our spectrometer operates at a frequency of 29.92 Mc./sec. but no correction for susceptibility was applied. The agreement with the corrected values is within the accuracy of the susceptibility measure- ments. IOcps. FIG.1. Nitrobenzene in (a) bubble and (b) cylindrical tube. Chemical shifts (c./sec.) at 40 Mc./sec.Compound Uncorr. shifts from Corr. values Bubble values Shifts with cyclohexane published data as internal reference Chloroform Benzene Toluene ring Toluene methyl Cyclohexane -33.9 -25.0 -23.4 -20.3 0 0 0 -16.7* 3-1 3.7 3.2 -11.6* 199.4 200.0 200.0 183*8* 212.2 212.5 212.2 212.5 * Note anomalous effect of aromatic solvents. that observed with cylindrical samples. The Figure shows recordings of the hydrogen resonance of nitro- benzene run in a cylindrical tube and in a spherical bubble. An additional advantage is concerned with the accurate location of peaks. The water in the gel pro- vides a very convenient external standard and because of the spherical shape of the sample correc- tions for susceptibility differences are unnecessary.The accuracy obtainable is indicated by the measure- ments in the Table of the positions of the resonances of a number of organic liquids. The first column gives the chemical shifts in cycles per second at Reilly Mellon Newsletter No. 2 1958 p. 5. Handbook of Chemistry and Physics 37th edn. 1956. 'Bother-by and Glick J. Chem. Phys. 1957 26 1651. The same compounds were studied by using cyclo- hexane (5%) as an internal standard in cylindrical tubes and the chemical shifts are recorded in column 5. The values confirm the view that in aro- matic solvents significant errors can OCCU~.~ A considerable number of materials has been tried for use with samples which are miscible with or re- active to water.We have not yet found a material which does not suffer from one of the disadvantages that (a) it does not wet glass (b) it sets too sharply (c) it has a complex spectrum of its own or (d) it melts at too high a temperature. (Received March 5th 1959.) PROCEEDINGS Electron Spin Resonance in Crystals of the Amminoperoxydicobalt Complexes By E. E. SCHNEIDER and J. WEISS [DEPARTMENT OF CHEMISTRY OF mfys~cs(E.E.S.) and DEPARTMENT (J. W.) KING'SCOLLEGE UNIVERSITY OF DURHAM ON TYNE] NEWCASTLE WITH reference to the recent Communication1 on the electron spin resonance spectrum of the [(NH3)5Co-0,Co(NH3)5]5+ ion in solution we report our investigations with single crystals of the ammino per oxydico bal t sulp ha t e .Our measurements were carried out at 9.5 and 32 kMc./sec. with as- semblies of small (aligned) single crystals. The resonance spectrum is strongly orientation-dependent and consists in general of two peaks corresponding to two non-equivalent molecules in the unit cell. The principal components of the g-factor derived from a preliminary analysis were Bernal Ebsworth and Weil Proc. Chem. Soc. 1959 57. g, = 2.12 and gL = 1.99. These results and the large orientation-dependent width of the individual lines (200-500 gauss) are consistent with the sharing of" the positive hole between the two cobalt ions. The large spin-orbit coupling of Co4+ and the hyperline interaction arising from the nuclear moment of 59C0 would account for the large g-shift (gll-go) = 0.12 and the large width of the resonance.The results of' Bernal et a2.l for solutions support these findings. Some of these results were presented at the Colloque Ampkre in Paris July 1958. More detailed studies with individual single crystals are in progress. (Received March 19th 1959.). Substituted Manganese Carbonyl Derivatives By R. S. NYHOLM and D. V. RAMANARAO (UNIVERSITY COLLEGE GOWER STREET LONDON W.C. 1) VERYfew compounds are known in which one or more of the CO groups of dimanganese decacarbonyl have been replaced by a neutral molecule e.g. R3P. An example of these is the triphenylphosphine com- pound [Mn(CO),Ph,PIo prepared by Hieber and Wagner.l This complex is monomeric in benzene so the manganese atom is five-covalent.It is paramag- netic with the magnetic moment expected for one un- paired electron. We have been studying the behaviour of Mn,(CO), with o-phenylenebisdimethylarsine (D) and find that two CO groups per manganese atom may be replaced readily. Two derivatives have been isolated both of which have the empirical formula MnD(CO), one being monomeric and the other dimeric in freezing benzene. The two forms are prepared as follows with properties as shown. Typical analytical data are Dimer C 36.9; H 3.9; As 34.0; Mn 12.6. Monomer C 37.1; H 3.9; As 35.1; Mn 12-75. Cl,H1603As,Mn requires C 36.7; H 3-8; As 35-3; Mn 12.9%. In a Nujol mull and in chloroform solution the infrared spectra of the two forms are identical bands being observed at 1957 and 1860 cm.-l in Nujol and 1958 (s) and 1905 cm.-l (s) in chloroform.Both the compounds dissolve in chloroform but only the monomer can be isolated subsequently. However the spectra of the two forms do differ in carbon di- sulphide; the dimer shows bands at 1944 (s) 1927 (s) and 1885 (s) cm.-l whereas the monomer absorbs at 1965 (s) and 1918 (s) cm.-l. Further the melting points of the dimer and the monomer are unchanged Hieber and Wagner 2.Naturforsch. 1957 12b,478. on recovery of material from carbon disulphide solutions. It seems reasonable to suggest that initially one diarsine molecule is attached to each Mn atom in the dimer and that on further heating the Mn-Mn bond is broken. One might have expected an equilibrium between the monomer and the dimer depending upon the temperature.However it is found that al- though the dimer is converted into the monomer on further heating or melting reconversion into the dimer on cooling to room temperature does not take place. We therefore regard the dimer as a metastable intermediate in the formation of the (stable) mono- mer. In these two compounds we have an unusual example of polymeric isomerism. If either form is heated at a higher temperature for a long time more CO is lost and a more strongly paramagnetic com-pound (p -3.6 B.M.) probably Mn(CO),D is formed. This unstable derivative is being investigated. When titrated spectrophotometrically with bro- mine in chloroform both forms react with the loss of one mole of CO per Mn atom a sharp end-point occurring after two equivalents of bromine have been added-a reaction which does not occur with Mn,(CO), under these conditions.A pale yellow bi- valent manganese compound MnD(CO),Br, can be isolated (Found C 25-3; H 3-5; Br 29.4; Mn 10.1. C1,H16As,02Br,Mn requires C 25.8; H 2.9; Br 28.7; Mn 9.9%). The compound is paramag- netic (peff 2.3 B.M.) indicating one unpaired electron as expected for an octahedral spin-paired APRIL1959 131 Action of the diarsine (D) on Mn,(CO), when heated together in vacuo for 6 hr. ~ 130" Mn,(CO)10 160" I I 4 Heat Or Dimer [Mn(CO),D] -dissol. in CHCI ---+ Monomer [Mn(CO),D]O M.p. 292" M.p. 170" Diamagnetic Paramagnetic (p1-76B.M.) Both the monomer and the dimer are pale yellow and are non-electrolytes in MeNO,.complex of Mn(x1). This complex decomposes at 202" and is virtually a non-electrolyte in nitromethane. When iodine is used as oxidant only one equivalent is absorbed and no carbon monoxide is evolved. The product is a non-electrolyte and is diamagnetic it is clearly the octahedral Mn(1) derivative MnD(CO),I of interest since few Mn(1) compounds are known. The authors are indebted to the Ethyl Corporation (U.S.A.) for a gift of manganese carbonyl and to the Indian Government for an award (to one D.V.R.R.) also to Mr. C. Barraclough who is investigating the infrared spectra of these compounds. (Received February 30th 1959.) The Nuclear Magnetic Resonance Spectrum of a New Plant Amino-acid; Evidence for a Pyrazole Ring By L.FOWDEN, F. F. NOE,J. H. RIDD,and R. F. M. WHITE COLLEGE STREET W.C. 1) (UNIVERSITY GOWER LONDON THEwork on heterocyclic compounds carried out by two of us (J.H.R. and R.F.M.W.) suggests that proton magnetic resonance can be of use in dis- tinguishing between glyoxaline and pyrazole rings and in identifying the position of substituents. The proton magnetic resonance spectrum of glyoxaline has already been discussed;l in deuterium oxide the spectrum consists of two lines of relative areas 1 :2 assigned to the protons at the 2-and the 4(5)-posi- tion respectively (see Fig. a). A third line not shown in the Figure is caused by the protons moving between the nitrogen atom and the residual water in the solvent.In 1-methylglyoxaline the 4-and the 5-position are no longer equivalent but the spectrum of a solution of 1-methylglyoxaline in deuterium oxide shows two peaks in the low-field region of areas 1 :2 (Fig. b),so that under these conditions the chemical shift between the 4-and the 5-proton must be neglig- ible. The nuclear magnetic resonance spectra of solu- tions of pyrazole and 1-methylpyrazole in deuterium oxide indicate that there is little chemical shift be- tween the 3-and the 5-proton and that these are spin-spin coupled with the 2-proton; the larger peak is split into a doublet and the smaller into a triplet with separations of the multiplet components of about 2.5 c./sec. These spectra are illustrated dia- gramatically in the Figs.c and d the length of the line indicating the relative peak areas; the fine structure is omitted. Gillespie Grimison Ridd and White J. 1958 3228. In the spectrum of glyoxaline and I-methyl-glyoxaline the peak arising from the unique hydrogen atom occurs at a lower field than that arising from the equivalent pair of hydrogen atoms but in the spectrum of pyrazole and 1-methyl-pyrazole the reverse is true and the separation of the peaks is much greater. These differences probably suffice to distinguish N-substituted glyoxalines from N-subs tituted pyrazoles. Concurrently with the above work two of us (L.F. and F.F.N.) have extracted a new amino-acid from water-melon seed (Citrullus vulguris var. Tom Watson). The compound C6H902N3 is isomeric with histidine.It is considerably less basic than histidine and with hydrogen iodide gives rise to alanine (detected chromatographically) amongst other products. If the molecule contains an alanine residue then from the empirical formula this is probably attached to a glyoxaline or a pyrazole ring. The fission of the alanine residue from the ring by hydrogen iodide suggests that the linkage is to a ring-ni t rogen at om. The nuclear magnetic resonance spectrum of this amino-acid in 3~-sodium deuteroxide in deuterium oxide is shown in the Fig. e. In the low-field region two peaks of relative areas 2 1 are obtained as shown in the Figure their relative positions and the separation between them are characteristic of a PROCEEDINGS probably a solvent effect; the spectrum of l-methyl- pyrazole is modified in a similar way in a similar a rI medium.The two peaks showed indications of spin- spin interaction but this was not resolved probably b rI I because the viscous medium caused viscosity broad- =I I dl I I =I 8 I b 0 D c/sec. 700 50 i s’b H--t Proton magnetic resonance spectra with a water capillary as the reference zero. a Glyoxaline. b 1-Methylglyoxaline. c Pyrazole. d 1-Methylpyrazole. e New amino-acid. a-d in D,O; e in ~N-N~OD-D,~. pyrazole ring. The shift of the two peaks to lower field relative to pyrazole and 1-methylpyrazole is ening of the peaks. The peaks in the high-field region are assigned to the alanine residue; the peak at 20 c./sec.is a doublet and that at 50 c./sec. is a triplet with components separated by 6 c./sec. This fine structure suggests that the pyrazole ring is at- tached to the /?-carbon atom of alanine. Thus the spectrum of the amino-acid suggests that it is p-1-pyrazolylalanine. This compound appears to be the first example of a naturally occurring pyrazole derivative. Two of US (L.F. and F.F.N.) thank the Ferny- Morse Seed Co. (California) for the gift of the water- melon seed and one of us (F.F.N.) is grateful for an American Cancer Society Fellowship- (Received March 6th 1959.) Effect of Benzene as a Solvent on the Photochlorination of Butyryl Valeryl and Hexanoyl Chloride By H. J. DEN HERTOG and P. SMIT (LABORATORY CHEMISTRY UNIVERSTTY, OF ORGANIC OF THE AGRICULTURAL WAGENINGEN, THENETHERLANDS) RECENTLY G.A. Russell described a remarkable effect of some aromatic solvents and certain non- aromatic liquids on the photochlorination of several carbon compounds.l They accentuated the difference in reactivity of carbon-hydrogen bonds in these com- pounds. In his third paper Russell concluded that the relative reactivities determined mainly by the stability of the incipient free radicals are very sensitive to changes in solvents but that there is little solvent effect on relative reactivities determined chiefly by polar effects. This conclusion seems to contradict results of an investigation on the photochlorination of fatty acid derivatives which is in progress in our laboratory.2 We treated a mixture of 1 mol.of fatty acid chloride and 4 mols. of benzene at 20” with 0.35 mol. of chlorine under irradiation with ultraviolet light. The composition of the mixture of monochloro- derivatives formed was established by fractional dis- tillation in vacuo and countercurrent distribution of the lower fractions. The results are summarised in the Table together with those previously obtained when chlorinating the acid chloride in absence of the solvent.2 Relative reactivity towards Substrate Solvent chlorine* at position stated B Y 6 E Butyryl C6H6 1.0 0.35 -chloride None 1.0 0.6 -Valery1 C6H6 1.0 4.3 0.75 -chloride i None 1.0 3.05 1.2 -Hexanoyl C6H6 1-0 4.5 15.8 1.05 chloride None 1.0 3.6 5.1 1-75 * The reactivities of the C-H bonds at a-positions are omitted as the amounts of a-isomers formed are very small.The divergent reactivities of the Cp-H bonds are assumed to be 1.0. Hexanoyl chloride gives themost information about a possible solvent effect upon the deactivation of C-H bonds by an electron-withdrawing group; its molecule contains a series of Csec-H bonds with strongly decreasing eIectron density from 6 to a. The ratio of the reactivities towards free chlorine of the C,H and C,-H bonds in hexanoyl chloride amounts to 5.1/1-75 = 2.9. This value is but slightly lower than that given by Russell for the relative re- activities of Csec-H and Cprim-H bonds in pentane Russell J. Amer. Chem. SOC. 1957 79 2977; 1958 80 4987 4997 5002.Cf. den Hertog de Vries and van Bragt Rec. Trav. chim. 1955 74 1561;Smit and den Hertog ibid. 1958,77 73. APRIL1959 133 ~ ~~ (3-7) which only depend on the resonance stabilisa- tion of the incipient free radicals. Thus it appears again that the inductive effect of the carbonyl chloride group is scarcely felt at the 6-and the E-position and that the enhancement of the relative reactivities of these bonds by addition of benzene as a solvent from 2.9 to 15.0 only must be ascribed to the resonancestabilisationof theincipient radicals in ques- tion. Whereas the ratio (react.C,-H) (react. C,-H) increases to five-fold the relative reactivities of Cs-H and CB-H bonds increase from 5.1 to 15-8 (i.e.,more than three-fold) by the solvent effect.The change in relative reactivities of C,-H and the C,-H bonds would probably yield a still higher ratio. Thus in hexanoyl chloride the polar effect is not less im- portant than the resonance effect in determining the change in reactivity of carbon-hydrogen bonds by addition of a complex-forming solvent. For comparison the relative reactivities of the C-€1 bonds in valeryl and butyryl chloride are given all depending on both the resonance and polar effect. In propionyl chloride the relative reactivities at 01 and 18 are about the same (2 5)whether benzene is used as a solvent or not; this fact again elucidates the importance of the polar effect. (Received February 6th 1959.) The Mass Spectra of Phosphorus(v) Halides By T.KENNEDY,D. S. PAYNE, R. I. REED,and W. SNEDDEN (THEUNIVERSITY W.2) GLASGOW MASS-SPECTROSCOPIC data for dichlorotrifluorophos- phorane up to a mass of 160 have been given in a recent paper.l The range of masses accessible with the instrument then employed has been extended enabling reliable results on masses up to at least 400 to be obtained. Samples of “phosphorus penta- chloride” [PCl,+] [PcI6-] and of a new compound recently prepared by two of us (T.K. and D.S.P.) of empirical composition P,CI,F have now been examined. The two substances were prepared for examination under strictly anhydrous conditions and were ad- mitted through taps greased with KEL-F grease. The vapour pressures of the solids were both sufficiently high for the experiment to be conducted without the sample being heated.An ionisation energy of 50 ev was employed. The ion species were identified by counting from known background masses of residual hydrocarbon gases and confirmed by observation of the mass spread due to the isotopes of chlorine. The cracking pattern of “phosphorus penta- chloride” showed peaks which were assigned to the ion species P,Cl, P,Cl, P2C15 P,Cl, P,Cl, and P,C1 as well as a collection of peaks corresponding to species containing one phosphorus atom. For P,Cl,F peaks were assigned to the ion species P2ClgF P,Cl,F P,Cl,F P,C16F P,Cl,F P,Cl,F P,Cl,F and P,ClF in addition to fragments contain- ing one phosphorus atom. Several recent investigations have indicated the existence of polymerised metallic halides in the vapour state.However whilst it has been known for many years that the halides of phosphorus(v) are based largely on ionic lattices the considerable amount of information on the vapour densities particularly of “phosphorus pentachloride,” has provided evidence only of dissociation never of association. Attempts by one of us (D.S.P.) to measure the molecular weight of “phosphorus penta- chloride” in the vapour phase at relatively low temperatures under conditions in which the dissocia- tion is known to be very small gave no indication of association within the limits of measurement. It is clear however from the mass spectra that associa- tion does occur. The existence of P2X, units (where X represents Cl or F) and associated fragments in the mass-spectrometer could be explained by a collision hypothesis e.g.PX + PX,+ -+ P2Xlo+ etc. but in view of the absence of fragments containing two fluorine atoms in the experiment with P,Cl,F and the relatively small amounts of PX species likely to be present this seems unlikely. It is also possible that a parent of even higher molecular weight occurs and that this on ionisation breaks up to give P,X,, etc.. units. However the most reasonable hypothesis is that there are P,X, units in equilibrium with the solid at room temperature. The solid being of the type [PX,+] [PX,-1 the units are formed essentially by evaporation of a cation-anion pair from the surface of the crystal.In the P,X, and other P units a halogen bridge involving a four-membered ring could be possible. This would be in keeping with many other polymeric halides and indeed might be likely in that the result- ing complex would have octahedrally disposed halogens around the phosphorus atoms rather than the trigonal bipyramid arrangement as found in the case of the “phosphorus pentachloride” molecule.2 The exact composition of the vapour with respect to monomeric and dimeric units in equilibrium with the solid and the detailed structure of these units remains to be investigated. (Received February 19th 1959.) Kennedy and Payne J. 1959 1228. a Brockway and Beach J. Amer. Chem. SOC.,1938,60 1836. 134 PROCEEDINGS Echitamine By T.R. GOVINDACHARI and S. RAJAPPA (DEPARTMENT INDIA) OF CHEMISTRY PRESIDENCY COLLEGE MADRAS GOODSON and HENRY^ isolated the alkaloid echit- amine as its “hydrochloride,” C,,H,,O,N,,HCl from various A lstonia species. Besides clearing up disputed points in earlier investigations2 of the alka- loids of Alstonia scholaris they obtained the follow- ing information The free base could not be obtained by basification of the hydrochloride. Mild alkaline hydrolysis leads to “demethylechitamine” and methanol indicating the presence of a methoxy-carbonyl group. Distillation of “echitamine hydro- chloride” with alkali yields methylamine and a substance giving indole colour reactions. We now find that the alkaloid contains one OMe one NMe and no CMe group.Potentiometric titra- tion against N/lOO-sodium hydroxide indicates that the salt is not a hydrochloride but a quaternary chloride. The alkaloid shows absorption maxima in ethanol at 235 and 295 mp (log E 3.93 3.59 un- affected by addition of acid. From this it is inferred that the alkaloid is a dihydroindole derivative with N and Nb separated by not more than two carbon atoms,3 Nbbeing the quaternary centre carrying the rNMe+ group. The chloride shows infrared bands at 3.04 (OH) 3.17 (NH) 5-79 (ester) and 13.2 p (0-disubstituted benzene) and no band near 4.15 (=NH+) or 7.2 p (CMe). The ultraviolet absorption of the acetyl derivative’ (Amax. 235 295 mp; log E 3.94 3.50) is very similar to that of echitamine chloride and the infrared spectrum exhibits bands at 2-92 (NH) 5.70 (ester) and 8.14 p (OAc).Mild alkaline hydrolysis of the alkaloid gave “demethylechitamine,” which must be a betaine since potentiometric titration still indicated the presence of a quaternary centre. Catalytic reduction of echitamine chloride at atmospheric pressure gave a tertiary base (a mixture of stereoisomers) showing absorption at 247 and 307 mp indicating that it is a dihydroindole. In acid solution there is a hypsochromic shift of both these bands suggesting that the two nitrogen atoms are separated by only one carbon atom.3 The base had infrared absorption bands at 3.57 (>NMe)4 and 7.20 p (CMe). The ready formation of this tertiary base with newly formed >NMe and CMe groups indicates the presence of an allylamine system:5 >C:dCH,.+NMe< -+ >CH-&HMe + MeN< Acetylation of the tertiary base gives an N,Ac derivative as indicated by infrared (amide band at 6.0 p) and ultraviolet spectra (Amax.252; hinfl. 282 mp) and a positive Otto reaction. Hence the alkaloid must have a free N,H group. Dehydrogenation of the tertiary base by selenium gives besides indoles a basic product isolated as the picrate m.p. 253” (decomp.). The base liberated from the picrate has Amax. in ethanol at 245 and 307 mp and Amax. in ethanolic hydrochloric acid at 245 327 and 355 mp. The spectrum of a neutral solution is strikingly similar to that of 1’-methylpyr-rolo(2’ 3’-3 :4)q~inoline.~ The infrared spectrum contains no NH band but a well-defined band at 7.22 p (CMe).We thank the Council of Scientific and Industrial Research for the award of a Senior Research Fellowship (to S.R.). (Received February 24th 1959.) Goodson and Henry J. 1925 127 1640; Goodson J. 1932 2626. a Gorup-Besanez Annalen 1875 176,88; Hesse ibid. p. 326; 1880 203 144; Ber. 1880 13 1841; Harnack Arch. exp. Path. Pharm. 1877,7 126; Ber. 1878,11,2004; 1880,13 1648. Hodson and Smith J. 1957 1877. Braunholtz Ebsworth Mann and Sheppard J. 1958 2780. Bickel Schmid and Karrer Helv. Chim. Ada 1955 38 649. Eiter Munafsh. 1948 79 17; Eiter and Nagy ibid. 1949 80 607. OBITUARY NOTICE CECIL HENRY DESCH 1874-1958 CECIL HENRY DESCH F.R.S. died in a London hospital on Thursday June 19th 1958.This dis- tinguished metallurgical scientist whose published work is known in all parts of the world was 84 years of age and had lived in retirement for some years. Born in 1874 in London son of Henry Thomas Desch he was educated at Birkbeck School Kings-land and Finsbury Technical College London. From there he went to the University of Wiirzburg where he obtaiqed the degree of Ph.D. and subsequently APRIL1959 he returned to University College London later receiving the degree of D.Sc. of the University of London. He was trained as an organic chemist but joined the metallurgical department of King’s College London in 1902 and became research assistant to the late Professor A. K. Huntington. In 1909 the year of his marriage to Elison Ann daughter of Professor W.Ivison Macadam of Edinburgh he was appointed lecturer in Metal- lurgical Chemistry at the University of Glasgow where he remained until the end of World War I. For the next two years he was Professor of Metallurgy at the Royal Technical College Glasgow until in 1920 on the death of Dr. J. 0. Arnold he became Professor of Metallurgy at Sheffield University occupying the Chair for 11 years. In 1931 Dr. Desch was appointed Superintendent of the Metallurgy Department of the National Physical Laboratory Teddington in succession to Dr. W. Rosenhain but did not take up his duties there until the following year because of his visit at the end of 1931 to the United States of America where he delivered the George Fisher Baker course of lectures on the chemistry of solids at Cornell University Ithaca New York.In 1934 this course of lectures was printed in book form and entitled “The Chemistry of Solids.” After retiring from his position at the National Physical Laboratory in 1939 Dr. Desch was soon directly connected with industry. In 1943 he was appointed to the Board of Directors of Richard Thomas & Company Ltd. and was responsible for the direction of the research and development acti- vities of the firm; upon the amalgamation with Baldwins Ltd. he resigned his directorship and was subsequently associated with the Whitehead Iron and Steel Company. In the academic field Dr. Desch has received wide recognition. He was elected a Fellow of the Royal Society in 1923 was President of the Faraday Society from 1926-1928 President of the Institute of Metals from 1938-1940 and President of the Iron and Steel Institute from 1946-1948 having been awarded the Bessemer Gold Medal in 1938 for his work in the advancement of metallurgy.He was also awarded in 1941 the Platinum Medal of the Insti- tute of Metals was a corresponding member of the AcadCmie des Sciences and held the honorary degrees of Glasgow and Leoben(South Austria) Universities. In addition to his well known writings in the metallurgical and chemical fields he was interested in sociology and as a member of the Institute of Sociology he presented a number of papers on this subject. His published work includes the popular “Metal- lography” (1910) which was translated into German in 1914 and has run to several editions and the well known “Intermetallic Compounds ” (1 914).During Dr. Desch’s long career a major change took place in the outlook on metallurgical teaching and research. In his early years studies in inorganic chemical analysis formed a large part of the metal- lurgist’s training; subsequently the applications of physics physical chemistry and thermodynamics in the study of metals were all greatly extended. Desch himself did a great deal to foster these developments both by his teaching and in the post-graduate re- search laboratory which he set up and equipped in Sheffield in the early ’20s. Here physicists chemists and engineers worked together with metallurgists on many problems in physical metallurgy.His many degree students will remember the out- standing clarity of his lectures. His post-graduate students still remark on the phenomenal detail and scope of his knowledge of metallurgical literature and recall his quiet kindness and diligence in ensuring that they received the full credit for any original work which they accomplished. C. SYKES. APPLICATIONS FOR FELLOWSHIP (Fellows wishing to lodge objections to the election of these candidates should communicate with the Honorary Secretaries within ten days of the publication of this issue of Proceedings. Such objections will be treated as confidential. The forms of application are available in the Rooms of the Society for inspection by Fellows.) Ahmad Sanjid Shamin M.Sc.54 Noel Street Hyson Green Nottingham. Arene Eugene 0. Tedder Hall University College Ibadan Nigeria. Bailey Jennifer Vivien. Highfield Hall Omdurman Road Highfield Southampton. Barnett Gerald Wilton. Flat 4 60 Wymers Wood Road Burnham Bucks. Bhatia Iqbal Singh M.Sc. Ph.D. Indian Tea Association Tocklai Experimental Station Cinnamara P.O., Assam. India. Bhatia Yog Raj M.Sc. 1 Petersfield Avenue Slough Bucks. Bishop Robert John. 29 Kingswood Avenue Bromley Kent. Bourn Patricia Margaret B.Sc. Imperial Chemical Industries Limited Akers Research Laboratories The Frythe Welwyn Herts. Bryan Robert Finlay Ph.D. Laboratorium fur Organ- ische Chemie Eidg. Technische Hochschule Zurich 6, Switzerland.Bullock Pete Bradley. Divisional Laboratory 67 Wellington Road Edgbaston Birmingham. Burrows Brian Frederick B.Sc. Department of Chem-istry The University Bristol 8. Burton Ronald Eliot B.A. 392 Harold’s Cross Road Terenure Dublin. Cazes Jack B.S. 614 Junius Street Brooklyn 12 N.Y., U.S.A. Corbin James Lee B.A. Kedzie Chemical Laboratory Michigan State University East Lansing Michigan U.S.A. Cox Alan. 32 Keswick Grove Seedley Salford 6 Lancs. Critchley John Philip Ph.D. Truro Strawberry Lane Wilmslow Cheshire. Dailey William Sebastian. 53 Courtenay Gardens, Upminster Essex. Dale Anthony Richard. 7 Waller Road Beaconsfield Bucks. Davidson James Alan. 19 Gilroy Way Orpington Kent. Dikko Muhammed.University College Ibadan Nigeria. Domsky Irving Isaac Ph.D. Sterling Chemistry Labora- tory Yale University New Haven Conn. U.S.A. Douraghizadeh Khosrow. 60 Idmiston Road West Norwood London S.E.27. Duffy John Alastair Ph.D. 50 Leahurst Crescent, Harborne Birmingham 17. Foot John Rodney. 34 Carnock Road Manadon, Plymouth Devon. German Reginald James B.Sc. Department of Chem- istry The University Birmingham 15. Goh Siew Kee.Central Medical Store General Hospital Kuching Sarawak. Golfier Michel Jean Francois. Laboratoire de Chimie Organique &ole Polytechnique 17 Rue Descartks Paris 5 France. Hardv. Frederick Edward. B.Sc. 47 Bolingbroke Street. -Heaton Newcastle on Tyne 6. Hayman Nigel Ward. 29 Bartlemus Road Oxford.Melgen Lon Endre B.A. 2703 Yale Station New Haven Conn. U.S.A. Hiscock Alan Keith. 44 Lime Grove New Malden Surrey. Hiskey Richard Grant A.B. M.S. Ph.D. Department of Chemistry University of North Carolina Chapel Hill North Carolina USA. Hopton Frederick James B.Sc. 49 Oak Road Hoxfield Bristol 7. Hutchinson John. 14 Hirst Grove Darlington Co. Durham. Israel Mervyn M.S. Ph.D. Department of Chemistry, University of Michigan Ann Arbor Michigan U.S.A. Jaffe Fred Ph.D. W. R. Grace & Co. Washington Research Center Clarksville Md. U.S.A. Jones Richard Arnold Yardley B.A. M.S. University Chemical Laboratory Lensfield Road Cambridge. Last Walter Allan B.A. 49 Queen Anne Street London w.l. Marinier Bernard Ph.D. University Chemical Labora- tory Lensfield Road Cambridge.Mohrig Jerry Robert B.S. Department of Chemistry, University of Colorado Boulder Colorado U.S.A. Moore Francis Hugh B.Sc. Department of Chemistry, University of Auckland Auckland New Zealand. Morcom Ronald Edgar B.Sc. A.R.I.C. 20 Oakwood Avenue Borehamwood Herts. Mundye Anthony Arnold John. 73 Verulam Road Greenford Middx. Nash Glynn Michael. 77 Tudor Drive Watford Herts. Ndupu Bismarck Livingstone Nwobi. University College Ibadan Nigeria. Neve Sidney George B.%. 27 Gableson Avenue, Brighton 5 Sussex. Newton Alan Branford. Needler Hall Cottingham Nr. Hull Yorks. Nutten Albert John Ph.D. F.R.I.C. Government Chemist’s Laboratory Hong Kong. Nwaji Martin Nwege. Hall Two University College Ibadan Nigeria.Oke Olusegun Ladimeji. Mellamby Hall University College Ibadan Nigeria. Pincock Richard E. B.S. A.M. Converse Chemical Laboratory Harvard University Oxford Street, Cambridge 38 Mass. U.S.A. Prout Ernest George M.Sc. Ph.D. Department of Chemistry Rhodes University Grahamstown South Africa. Poe Anthony John M.A. B.Sc. 87 Westgate Chichester Sussex. Rafiq Muhammad B.Sc. B.Pharm. Department of Chemistrv. School of Pharmacy. Brunswick Sauare. London -W.C. 1. Ranslev. Derek Leonard. B.Sc. Sterling Chemistry Laboiatory Yale University Prospect St.,New Haven Conn. U.S.A. Ratcliffe Neil Thomas. 33 White Hart Road Chalvey Slough Bucks. Redford David George B.Sc. 32 Welbeck Avenue Bromley Kent. Ringold Howard J. Ph.D.Research Laboratories, Syntex S.A. Apdo. Postal 2679 Mexico D.F. Mexico. Roberts John Angel. c/o Investigation & Research Div. Private Bag 28 Ministry of Animal Health Kaduna Northern Nigeria. Roberts Vivien Gabrielle. 74 Primrose Avenue Chadwell Heath Essex. Rodwell Kenneth B.A. 68 Browning Road Enfield Middx. Saracoglu Saim. Department of Chemistry King’s College London W.C.2. Sheldrick Peter A.B. 2607 Piedmont Avenue Berkeley. 4 California U.S.A. Sherman Warren Vincent B.Sc. 98 Herent Drive Ilford Essex. Shipley Dudley Keith B.Sc. A.R.I.C. 25 Oakfield Avenue Birstall Leicester. Speroni Giovanni D.Chem. Via Gino Capponi 9, Florence Italy. Steyermark Paul R. Ph.D. W. R. Grace & Co. Research & Development Division Clarksville Maryland, U.S.A.Taylor David Robin B.A. c/o Rectory 1 Range Road Whalley Range Manchester 16. Teibo Benjamin Olanrewaju. Kuti Hall University College Ibadan Nigeria. Wain John Gordon. 248A Quannby Road Huddersfield Yorks. Wallace Thomas Joseph M.S. Department of Chemistry University of Connecticut Storrs Conn. U.S.A. Wessely Friedrich D.Phil. Wahringerstrasse 38 Vienna IX Austria. Willcocks Doreen V. 39 Bay Street Mosman Sydney N .S.W. Australia.