722 NATURE December 15, 1945 Vo l. 156 THE INDIAN STATISTICAL work but subsequently revoked their decision and asked the Hon. Secretary at the end of March 1942 INSTITUTE to proceed with the Jute survey. By this time a good portion of Burma had been occupied by the HE continued activity and growth during the Japanese, and the supply of rice from that country T war period of the Indian Statistical Institute is had been cut off. Apprehending a serious deteriora• a striking example of the benefits which may flow tion in the food situation (not because of the physical from the stimulus of war conditions, when individual volume of the import which was small but because initiative is unhampered in putting itself at the of the possible effect on prices) the Hon. Secretary service of Government and to meet the needs and submitted a definite scheme to the Government of deficiencies of existing official organization. Perhaps India at the end of March 1942 for extending the nowhere than in India have the prevailing conditions sample survey to cover the paddy crop in Bengal. been more propitious, or the need greater, for the In the course of the next few months he also dis• active use of agencies run at private risk for public cussed the question with a large number of officials purposes. in Bengal but all his efforts failed completely. In the The Institute, as it has now developed, has many absence of reliable statistics both the provincial and facets : on the educational side equally as a training the Central Governments were left entirely in the ground for computors and routine statisticians, and dark regarding the supply position of rice at the end as a centre of postgraduate research in the most far• of 1942. This made it impossible for Government to reaching branches of the mathematical theory of pursue a consistent food policy on any objective statistics and experimental design; as a professional basis. The weakness of the statistical position was institute and learned society bringing together all thus an important factor in bringing about the schools of thought in Indian statistics; as an agency deplorable food crisis in Bengal in 1943." employed by departments of Government and It was not, therefore, until after the famine that advisory bodies, in the essential work of collecting, the Institute was enabled to show what its organiza• scrutinizing and digesting the facts upon which tion could do with the immense problem of sampling administrative decisions must depend. The achieve• the 70,000 square miles of agricultural Bengal. ment of co-operation among the many able men Adequate accuracy in areas under crops is much more needed to guide these various activities has been the difficult to attain than is yield for unit areas. Some work of an applied mathematician, Prof. P. C. 59,000 grids each of 2·25 acres, chosen on a system Mahalanobis, formerly professor of physics, acting as of stratified random sampling, were needed for honorary secretary to the Institute. He was this year Bengal, the whole being divided into zone cells of elected fellow of the Royal Society. sixty-four square miles and sub-cells of one square There can be no doubt that accurate knowledge mile, approximately the area occupied by a single by the Government of Bengal of the amount of rice village. A feature of great importance for Indian available in the Province would have obviated the conditions, and worthy of study elsewhere, is the food crisis of 1943, in which approximately one duplication of the system in interpenetrating net• million lives were lost, by forestalling panic and works, so that entirely independent pairs of estimates cutting the ground from under the food speculators. are available for each area. This not only facilitates The story is tqld in the report for 1943-44 : the administrative checking of gross negligence, or "The most notable progress in the year under misunderstood instructions, but also enables the pre• review was achieved in the sample survey of crops. cision of the final estimate to be assessed as it really It would be remembered that work was started on a is by including all causes, human as well as physical, small scale with a total expenditure of about Rs. 7,000 which contribute to inaccuracy in the result. on an exploratory survey of the jute crop in Bengal There is perhaps no other organization in which in 193 7. This was the beginning of a five-year scheme practical and theoretical work are more thoroughly for the improvement of jute statistics which was integrated. The combinatorial investigations of R. C. financed jointly by the Indian Central Jute Com• Bose and A. Bhattacharya and the studies of multi• mittee and the Government of Bengal. The"sampling variate distributions of Mahalanobis and S. N. Roy technique, developed in the course of a gradually supply not only the general plan but also very expanding series of surveys culminating in the full detailed guidance to the two hundred or so workers scale provincial work of 1941, which demonstrated of the field and computing staffs. beyond dispute the possibility of securing by this method a final estimate of jute acreage with a margin of error of only two or three per cent at a cost of one-fifteenth or one-twentieth of that of a complete EXCAVATIONS AT HYRAX HILL, enumeration. A general account of the five-year scheme with discussion of theoretical foundations was NAKURU, KENYA COLONY given in a memoir prepared by the Hon. Secretary in LTHOUGH uncertainties in detail may still 1942 which is being published in the Philosophical A exist, the general structure of the prehistoric Transactions o{ the Royal Society of London. story of Kenya as laid down by Dr. L. S. B. Leakey "The Hon. Secretary had been pressing from the is now accepted by most prehistorians. The various very beginning of the five-year scheme in 1937 for climatic changes that have occurred and their cor• the extension of the method to cover paddy and other relations with the more important cultures found are important crops in Bengal. Each year from 1938 to also clear. Following on a major pluvial phase called 1942 he repeated his efforts but without success. In Gamblian and correlated with the Kenya upper 1942 the Indian Central Jute Committee had ex• palreolithic, there were two merely wet phases pressed their complete satisfaction with the method separated by a dryer period. These are known of the sample survey and had recommended its con• respectively as the Makalian and the Nakuru wet tinuance by Government in future. The Government phases. They are post-palreolithic in age, and the of Bengal however decided at first to discontinue the latter is not of great antiquity.

© 1945 Nature Publishing Group No. 3972 December 15, 1945 NATURE 723 Mary D. Leakey, in a paper entitled "Report on overcome them. That the subject of the conference the Excavations at Hyrax Hill, Nakuru, Kenya is of vital interest to the chemical industries, as Colony, 1937-1938" (Trans. Roy. Soc. S. Africa, 30, indeed it is to all industries, was shown by the Pt. 4), describes excavations made at sites on an remarkable attendance, the lecture theatre at the eminence which was under water during the period Royal Institution being filled and almost over-filled. covered by the last major pluvial (Gamblian), and Apart from all other benefits which accrue from a along the sides of which occur beach-levels of the conference of this kind, the personal contacts which subsequent wet phases. Correlation of the archoo• are made between the makers and users of instru• ological and geological records is therefore simple. A ments provide ample justification for holding it, and number of sites, including some stone-walled en• it was a happy thought on the part of the organizers closures and a low mound, were investigated, as well to arrange that the instrument firms should hold as a group of pit dwellings. There were found : ( 1) a private exhibitions of their instruments on the neolithic occupation site and cemetery, (2) a Gumban following day. (neolithic) pit-dwelling village, and (3) a later, prob• In the opening address to the conference, given ably not very ancient, level connected with the by the president of the Institute of Physics, Sir stone-walled enclosures and associated burial pits Frank Smith, two main points were made. The first and with an industry which shows the influence of was the interesting one that although in industry Arab traders. the scientific instrument is usually regarded as the The neolithic occupation site lies on the 335-ft. handmaiden of industrial processes, experience shows beach (Makalian) and is therefore subsequent to its that the emergence of a new type of instrument has formation ; the Gumban pit-dwelling is somewhat often been responsible for the birth of a new industry. younger in age and contains an industry (Gumban B) There is, for example, a great industry now in which is well known elsewhere and has been assigned existence catering for the requirements of radar, and to the maximum period of the later (Nakuru) wet it is indisputable that radar, and the industry phase. From the earlier neolithic levels there were associated with it, could not and would not exist brought to light no less than eighteen skeletons-a unless an instrument were available for measuring number female-in contracted burials. These are time intervals smaller than one millionth of a second. described in detail by Dr. Leakey. Some decades ago, Sir J. J. Thomson devised the The associated industry clearly shows its upper first instrument capable of achieving such measure• palooolithic origin, for it must be remembered that in ment. This was the cathode ray tube, which later East Africa there were no devastating climatic became the cathode ray oscillograph. Without this changes of the magnitude of those which obtained in instrument, and without also the wireless valve which Europe ; cultures were not swept away but ripened was a development of an instrument (the one-way and went to seed ; the past continued to influence rectifier of Sir Ambrose Fleming), there could not the growth of the present much more than was the have been any radar nor, in consequence, the great case in Europe. Thus in the levels above mentioned, new industry which has developed around it. Equally there were fo1md a whole series of obsidian tools spectacular examples could be cited; but it is prob• clearly derived from normal Kenya upper palooolithic ably true that nearly every big advance in industry types. In the Gumban levels there were stone bowls has been the result of knowledge obtained in the first and pottery typical of the Gumban B neolithic. place by the use of a new instrument. There is, '.rhe report contains 409 pages and is well illustrated. therefore, every incentive for industry to be instru• All concerned in the post-palooolithic cultures of ment-conscious, apart from that arising from the Kenya will find it very interesting. benefits obtained by the use of instruments which is M. c. BURKITT. ancillary to industrial processes. Britain has been warned, and is acutely aware, that the maintenance of its standard of living is dependent upon an increased export trade. The achievement of AUTOMATIC CONTROL AND an increased export trade is dependent upon our ability to produce articles of exceptional and standard RECORDING IN CHEMICAL AND quality at competitive prices. The material articles OTHER PROCESSES which can be exported can be classed either as special products or as articles capable of being produced by HE Institution of Chemical Engineers, the mass-production methods. It is in the production of T Institute of Physics and the Chemical Engineering this latter category that a realization of the necessity Group of the Society of Chemical Industry made of industrial instrumentation is essential. Only by arrangements more than a year and a half ago to the complete control of processes can an exceptional hold a joint conference in the autumn of 1944 on quality of product be made standard, and only by recording and controlling in the chemical industries. the avoidance of waste and the lowering of costs The incidence of flying bombs and of rockets, how• resulting from control can competitive prices be ever, made it inadvisable to gather together a large offered. number of key technicians in London at that time, In some general observations on the use of re• and the conference was postponed indefinitely. This corders and controllers, the director of the British has proved to be a fortunate circumstance ; for the Scientific Instrument Research Association, Mr. A. J. conference was held on October 19 of this year, when Philpot, pointed out that the general industrial the industries interested were not completely occupied structure of Britain, with its preponderance of rela• with war-time requirements, but were turning their tively small concerns, has been favourable to the attention once again to the problems of ordinary creation and maintenance of a system based on the commercial production. In consequence, the con• maximum use of individual craftsmanship and a ference has been held at a time when industry is not large exploitation of personal knowledge and ex• merely aware of certain technical deficiencies but also perience. Such a system served us well during a is actually planning to do everything possible to long period when industrial processes were relatively

© 1945 Nature Publishing Group