peated stress. In the case of cold-stretched steel, for low stresses the fatigue strength is actually less than for the same steel before stretching.

These phenomena, and others that illustrate the complexity of this problem, afford abundant opportunity for further research. The membership of the committee includes representatives of educational institutions, the Bureau of Standards, and several large industrial establishments. The work was divided among the members, two dealing with its metallographic features, two with machines for testing, two with mechanics of the materials involved, and one with a survey of the subject from the standpoint of the steel manufacturer. The results already obtained promise much for the future success of this undertaking.

Scores of other illustrations of effective coöperation in research might be given, especially in astronomy, where each of the 32 committees of the International Astronomical Union (p. 412 ff.) is constituted for the purpose of organizing coöperative investigations. In spite of the length of this list of committees, it cannot be said that astronomy offers any unique possibilities of joint action. The division of the sky among widely separated observers is only a single means of coöperation, which may be paralleled in geology, paleontology, geography, botany, zoölogy, meteorology, geodesy, terrestrial magnetism and other branches of geophysics, and in many other departments of science. Most of the larger problems of physics and chemistry, though open to study in any laboratory, could be attacked to advantage by coöperating groups. In fact, it may be doubted whether research in any field of science or its applications would not benefit greatly by some form of coöperative attack.

As for the fear of central control, and of interference with personal liberty and individual initiative, which has been entertained by some men of science, it certainly is not warranted by the facts. Coöperative research should always be purely voluntary, and the development of improved methods of obser

vation and novel modes of procedure, not foreseen in preparing the original scheme, should invariably be encouraged. They may occasionally upset some adopted plan of action, but if the coöperating investigators are following the wrong path, or neglecting easily available means of improving their results, the sooner this is discovered the better for all concerned.

XXIII

THE INTERNATIONAL ORGANIZATION OF

TH

RESEARCH

GEORGE ELLERY HALE

HE progress of research, and the rapid advance of knowledge along particular lines, have naturally resulted in the highly specialized organization of science of the present day. Two centuries ago the Royal Society of London and the Paris Academy of Sciences could easily embrace the whole range of science, and include in their membership essentially all of the able investigators of England and France. The establishment of the Linnean Society in 1788 marked the beginning of a dispersive movement that has continued ever since. The Geological Society was instituted in 1807 and the Royal Astronomical Society in 1820, partly as the result of the accumulation of valuable observations too extensive for the Royal Society to publish. One by one the recognized branches of science took definite form, developed a large group of adherents, and a special society resulted. Engineering and medicine experienced the same progress, large general societies being followed by special organizations occupying particular fields. Civil, mechanical, mining, and electrical were the first great subdivisions of engineering, but recent years have brought increased specialization, and we now have societies of naval, illuminating, automotive, and refrigerating engineering, followed within the past year by a society devoted to electric welding. Medicine has also divided into many elements, and it is safe to predict that new special bodies will continue to arise as workers multiply.

This rapid progress of subdivision indicates, of course, a most healthy condition of affairs. Without intense concentration of interest and effort science and its applications could never have attained their present advanced position. Specialization in research, and the association of specialists in groups, must, therefore, be regarded as signs of progress. But the consequences of this movement are not wholly advantageous. The separation of investigators of somewhat different tastes has deprived them of many mutual benefits. In each branch of science instruments and methods are devised to meet particular needs. These may have many applications, direct or indirect, in other quarters, but for the most part they remain the undisputed possession of those acquainted with the special journals in which they are described. How often does the physicist consult the proceedings of a psychological society, or the engineer a journal of physiology? Yet problems often arise in which the experience gained in remote fields would be invaluable. More often, when some large problem is open to attack, its aspects are so various that investigators representing a dozen branches of science may be needed to deal with it effectively. It then appears most clearly how the artificial partition of knowledge, and the erection of barriers between those concerned with its increase and those most interested in its applications, must hamper effort and delay progress.

Without specifying other reasons, it is plain that the increase of specialization, instead of rendering unnecessary organizations dealing with science as a whole, has served to emphasize their possibilities. In fact, it may be doubted whether there was ever a time in the history of science when such bodies could render a greater service. The rise of astrophysics and physical chemistry is evidence enough of the advantage of bridging the gaps between diverging branches, and the great national academies of science are in a position to contribute in large measure toward this end.

In a previous chapter (1) we have seen how the National Academy of Sciences, acting on the approach of war, estab

lished the National Research Council. The immediate purpose in view was to effect a working federation of research agencies, without regard for the distinctions which have divided them into classes, and kept them from acting together. The organization then effected was a temporary one, designed for war service, and open to reconstruction to meet the needs of peace. In chapter 24 Dr. Angell describes the present organization of the Council and the nature of its work. We may therefore turn to the question of the international organization of science.

The international scientific associations that existed before the war were of several distinct types. Some devoted their efforts to the establishment of uniform standards of measure, others organized coöperative researches, while the majority held occasional congresses for the personal interchange of views. The most important body of the first type is the International Metric Commission, with its International Bureau of Weights and Measures at Sèvres, chiefly concerned with fundamental standards of length and mass. Other bodies of this class dealt with electrical units and standards, the standardization of the nomenclature and ratings of electrical apparatus and machinery, uniformity in the methods of testing materials, annual revision of the tables of atomic weights, annual publication of physical and chemical constants, the science and art of illumination, collaboration in the publication of astronomical ephemerides, uniformity in meteorological observations and their reduction, the determination of standards of wave-length, the classification of stellar spectra, and the unification of time standards.

Men of science interested primarily in coöperative research organized the International Chart of the Heavens, observations to determine the variation of latitude, seismological observations, explorations of the sea, solar observations, studies of the brain, and other investigations. Other international organizations dealt with agricultural information and statistics, physiological instruments, the telegraphic distribution of astronomical information, the preparation of a joint map of the