206 



SCIENCE. 



[Vol. VL, No. 136 



on the fillet without transferrence to books. In re- 

 sponse to inquiry, he stated that the diflaculty in get- 

 ting a circuit through a clock-pendulum and globule 

 of mercury, which would be absolutely sure to close 

 every second, might be entirely overcome by having 

 the mercury pure and making sure of good connec- 

 tions; that the difference between commercial and 

 pure mercury was a very marked one in this case. 

 Mr. J. A. Brashear of Pittsburgh called attention to 

 the growing importance of chronographic records in 

 all employment of men and machinery, and described 

 a very perfect system in use in some manufactories. 

 Professor Newton referred to the very convenient 

 system of the Repsolds, for printing rapidly the set- 

 tings of micrometer-screws, as well illustrated upon 

 the new Yale college heliometer in charge of Dr. 

 Elkin. Prof. C. S. Peirce of the coast survey called 

 attention to the great gain this would be in record- 

 ing the readings of micrometer screws in the com- 

 parison of standards of length where rapidity was 

 highly desirable, and especially the avoidance of the 

 necessity of removing the eye continually from the 

 eye-piece to read off the head. Professor Paul 

 alluded to his hope of soon applying Repsold's appa- 

 ratus where rapidity was of the first importance; viz., 

 in recording the settings of the position-circle of 

 the Nicol-prism in Professor Pickering's method of 

 observing the eclipses of Jupiter's satellites, where 

 as many settings as possible are wanted while the 

 satellite is entering or leaving the shadow; and he 

 said he hoped, with a chronograph-key in one hand, 

 and managing the Nicol and its printer with the 

 other, to be able to secure the record of the times and 

 settings of the Nicol-circle every two or three seconds, 

 woiking entirely in the dark, and keeping it up as 

 long as desired. 



The next paper, by Prof. J. Burkitt Webb, de- 

 scribed a method of using polar coordinates, by 

 transferring the origin from the centre to the end of 

 the unit-radius, — thus substituting (r— 1) for r, — and 

 then using the length of the arc and the distance out 

 from its end upon the radius vector, as x and y are 

 used in rectangular coordinates. He found this a 

 very convenient transformation in the application 

 of polar coordinates to the discussion of Amsler's 

 planimeter; and pointing out that by substituting 

 infinity for unit-radius in the equations thus trans- 

 formed, they were reduced to those of rectangular 

 coordinates, he thought this transformation of polar 

 coordinates might be found generally useful. 



The only paper on Monday was a presentation by 

 Mr. C. H. Rockwell of Tarrytown, N.Y., of some re- 

 sults of his observations for time and latitude with the 

 almucantar, an instrument devised by Mr. Chandler 

 of the Harvard-college observatory a year or two ago, 

 which promises at least to furnish an entirely new 

 and radically different method of attacking the ques- 

 tion of absolute positions of the stars, and very prob- 

 ably far to surpass all others in accuracy, on account 

 of its freedom from systematic errors. The results 

 thus far published by Mr. Chandler seem fully to con- 

 firm all that was expected of the instrument; and it 

 is probably not too much to say, that it is the most 



important addition of the present century to the in- 

 struments and methods used in the determination of 

 absolute star-positions. The sources of systematic 

 error would seem to be almost wholly reduced to 

 those of varying personal equation in the observation 

 of transits at all speeds, and at all inclinations and 

 directions over horizontal wires, and to possible sys- 

 tematic difference in atmospheric refraction in differ- 

 ent azimuths. Mr. Eockwell exhibited some results, 

 simply copied from his observing-books, illustrating 

 the methods of reduction for time and latitude ob- 

 servations, and showing the degree of accuracy that 

 can be attained by the instrument in both these di- 

 rections. They served to show that the instrument 

 when duplicated will give equally good results with 

 the one first constructed; and their consideration 

 gave rise to a very interesting discussion, partici- 

 pated in by many members, as to the character of 

 work the instrument might be expected to do, in the 

 course of which Mr. Rockwell answered, in a very 

 entertaining way, many questions, put by various 

 members, as to the details of observing and reducing, 

 which were not before clearly understood on account 

 of the novelty of the work. One of the most impor- 

 tant problems which the instrument is specially 

 adai)ted to investigate, and one which we hope Mr. 

 Qhandler will soon find time to undertake, is the de- 

 termination of the declination of fundamental stars 

 south of the equator, tying them to northern stars 

 at corresponding zenith distances below the pole. 

 This would seem to be by far the best, perhaps the 

 only, method of connecting these together in a way 

 that shall be free from systematic error. 



PROCEEDINGS OF THE SECTION OF 

 PHYSICS. 



The first paper read before the section of physics 

 was by Prof. S. P. Langley, on the spectra of some 

 sources of invisible radiations, and on the recognition 

 of hitherto unmeasured wave-lengths. The meas- 

 urement of infra red wave-lengths has heretofore 

 been confined to those found within the range of the 

 solar-heat spectrum. It is of interest to know 

 whether there are other wave-lengths than those 

 found in the sun's heat, so that we may perhaps ex- 

 plain how it is that the surface heat of one planet is 

 maintained in spite of the ready radiations of extreme 

 solar heat through the atmosphere. Our knowledge 

 of wave-lengths is comparatively recent, as Fraun- 

 hofer gave the first accurate measures in 1823. His 

 range of values was from .00036 to .00075 of a milli- 

 metre. The use of the florescent eye-piece and pho- 

 tography has extended the range. The extreme range 

 of the normal eye is from about .00036 to .00081 of a 

 millimetre, or a little over one octave. It has been 

 known since the time of the first Herschel, that heat 

 radiations existed below the range of vision; but it 

 was supposed that glass absorbed this dark heat. In 

 1881 Professor Langley found that common glass was 

 diathermanous to all the dark rays which come to us 



