peratures used would be of great economic value. Indeed, without such knowledge improvements of metallurgic processes must be slow, uncertain, and expensive, and must result oftentimes from chance. In many metallurgic processes a nice adjustment of temperature is essential to success. But until quite recently pyrometers have found small favor with practical men, chiefly because efficient instruments were not easily available.

In short, pyrometry is in touch with the industrial arts at many points, and an investigation more detailed than needful for purely geologic work seemed warranted. When the work was begun,' about twelve years ago, almost no practical method of pyrometry was in use or satisfactorily vouched for. True, the great researches of Pouillet, Becquerel, and Deville and Troost were already published; but, unfortunately, the results were so frequently discordant that the methods were looked upon with distrust. This was especially true with thermoelectric pyrometry, and it is a pleasure to note that our laboratory was the first to revive this invaluable method.

A full account of the work done in pyrometry is given in the publications of the Survey. These books begin with a detailed historical review of the subject so arranged as to best meet the needs of the practical investigator. The experiments begun by Drs. Barus and Hallock were planned on a broad scale, and, following Deville and Troost, the general scheme was to thermally experiment with large masses of substance. But as it became necessary to change the quarters of the laboratory, the original plan had to be abandoned, and the subsequent researches of Dr. Barus took a new point of departure. The chief aim of the work was to standardize the thermocouple for high temperatures. The problem required that this couple should consist of platinum combined with some equally refractory and thermoelectrically invariable alloy of platinum. Platinum-iridium was the metal principally used. Methods of calibration based on known boiling points and convenient forms

'Annual reports of the director of the Geological Survey, 1881-'82, p. 9; ibid, 1883'84, p. 62; ibid, 1886-'87, p. 192. Bull. No. 54, U. S. Geological Survey, 1889, preface, pp. 17-21.

Bull. No. 54 and 103, U. S. Geol. Survey.

not been made for the purpose of obtaining Museum material, they are very extensive and include a full exhibit of the faunas contained in the upper Paleozoic formation, of the Rocky mountain region. Large collections have also been made from typical localities in the Southern and Middle states in order to obtain the material for a reference collection for use when studying the collections brought in by Geological Survey parties engaged in work in areas that have not hitherto been surveyed.

COLLECTIONS FROM MESOZOIC ROCKS.

The total number of specimens of Mesozoic fossils collected by this division is estimated at 32,000. Of this number 13,636, actually counted, have been transferred to the National Museum and entered in the catalogue, the remainder constituting material mostly unstudied. This estimate does not include several small collections of fossil plants and Paleozoic and Cenozoic invertebrates that have been transferred to other divisions of the Survey.

The collections made by the geologists of the Survey in this division have been indispensable to the prosecution of its work and they are indispensable to its completion if that should ever be attempted. They will always remain of the greatest importance for reference by those who shall prosecute similar work in future.

COLLECTIONS FROM CENOZOIC ROCKS.

There are at present about 200,000 fossils from the Tertiary in charge of Mr. Dall, of which number about 50,000 were collected prior to the existence of the present Survey. Of this collection about 60,000 specimens have been studied, put in order, and turned over to the National Museum, a part of which are on exhibition in the Museum building.

In bulk the collection is the largest in the world relating to the American Tertiary, and, by reason of the data accompanying the specimens, is in scientific interest the most important. Nearly all of the material collected by the Survey has its geologic and geographic position properly recorded from the col

lector's note books.

COLLECTIONS OF FOSSIL VERTEBRATES.

Prof. Marsh reports that the number of large boxes containing vertebrate fossils shipped from the West during the past ten years was over one thousand; that the average weight of these boxes was several hundred pounds; that during the past few the boxes have increased in size, averagyears ing from 600 to 800 pounds, while a number of them weighed from one to three tons each. Several hundred small boxes containing vertebrate fossils, many of the specimens being of great value, were also sent in as a part of the collections made in the West by this division. He regards the scientific value of the entire collection as far greater than that of any other collection of fossils made by any other geological survey in the world. This estimate is indorsed by some of the best scientific authorities in this country and in Europe.

A collection contained in seventy-two boxes was shipped to Washington in 1886, and in 1891 all the available case room provided by the National Museum was filled up by a collection of great value and scientific interest.

COLLECTIONS OF FOSSIL PLANTS.

Since its organization, in 1881, the Division of Paleobotany had received from the geologists of the Survey and had collected by its own members, 33,000 specimens. Of these, 14,600 have been transferred to the collections of the National Museum. Studies of the remaining portion of the collection have not yet been finished, and these specimens, although contained in the laboratories of the National Museum, have not yet been formally transferred to that institution. Their value to the cause of science is very great, as many of them could not be duplicated and they are the types upon which species and genera have been founded, and, in a number of cases, form the basis for geologic determination Prof. William M. Fontaine has added about 6,000 specimens to the collections of the Survey, of which 2,500 have been sent to the National Museum and are included in the 33,000 speci mens mentioned above.

COLLECTIONS OF FOSSIL INSECTS.

The number of specimens belonging to the Survey, collected by this division, is 2,190. The specimens are not as valuable for the purposes of correlation as those from other divisions, but they are of great scientific interest. They afford an almost virgin field for the investigator, and when the localities where they occur have been properly explored they will furnish an immense number of different forms of life relating so closely to those now living that conclusions can be drawn concerning the climate and interrelation of life in past ages (from our knowledge of the distribution and dependence of existing life) such as no other field will be able to offer. They are, as far as known, coufined entirely to fresh-water deposits, and they promise to be of much value in the correlation of the sedimentary series of the great interior lakes of Cretaceous and Tertiary time.

PREPARATION FOR COLUMBIAN EXPOSITION.

In connection with the geologic exhibit made at the World's Columbian Exposition at Chicago, the Paleontologic Division of the Survey prepared an exhibit to illustrate the stratigraphic succession of the faunas and floras in the sedimentary rocks of the United States. This is supplemented by maps, sections, and a typical series of rocks illustrating the sediments in which the fossils occur, and forms one of the principal features of the Survey exhibit.

PUBLICATIONS.

The paleontologic branch has from time to time announced the results of its researches in numerous treatises, which have been published by the survey in 8 monographs, 26 bulletins, and 10 accompanying papers in the annual reports, aggregating more than 8,000 printed pages, with more than 1,000 illustration plates and several thousand text figures. For detailed information concerning these publications, the reader is referred to the advertisement printed in the preliminary pages, herewith, and especially to Bulletin 100, giving a list and index of the books of the survey.

In addition to the official reports numerous papers have been published in magazines and journals: By W. H. Dall and assistants, 10; by C. A. White and assistants, 36; by C. D. Walcott and assistants, 48; by O. C. Marsh and assistants, 42; by S. H. Scudder and assistants, 33; by L. F. Ward and assistants, 40; by Alpheus Hyatt, 11.

GENERAL SUMMARY OF WORK IN TERRESTRIAL PHYSICS, 1880-'92.

RIGIDITY.

Among the first of the geologically important properties of bodies investigated in the physical laboratory was rigidity. It was thought inexpedient to begin with the study of elasticity, for geologic agencies usually act gradually throughout enormous periods, and hence the time variation of rigidity (or viscosity) seemed to be immediately important. It was also thought that, after the nature of viscosity was thoroughly apprehended, the geologic relations of rigidity, regarded as an elastic resistance, would more appropriately come up for discussion.

Much consideration was given to the choice of material suitable for the work. It seemed impracticable to begin with rocks, because of the difficulty in shaping or adapting them to the experiments; nor did it seem wise to seek the elementary facts of complex phenomena in substances which, as a rule, possess neither definiteness of chemical character nor physical homogeneity. Moreover, it was believed that if all the data to be gathered related to some substance of great industrial importance, the work, aside from its theoretic aspects, might ulteriorly be productive of results of economic value. Accordingly the group of iron carbides (wrought iron, steel, cast iron) was selected for experiment.

There was still another argument in favor of this choice. One member of the group, steel, has the unique property of passing continuously from an extreme of brittle hardness to an equally pronounced extreme of soft plasticity. By the aid of an operation so simple as tempering, therefore, this single metal is capable of representing an almost infinite variety of bodies.