August 21, 1890] 



NATURE 



393 



required, and no wall presses are used. Although the 

 ultimate potential storage capacity of the building when 

 complete may be extended to 600,000 volumes, at present 

 only about 100,000 have to be provided for. 



The arrangement of the books is according to a modi- 

 fication of the decimal system of Mr. Dewey, of the State 

 Library of New York. It is claimed for this system that 

 it brings together on the shelves all works of cognate 

 character, be they general or specific. 



Within the space available here it is not possible to 

 fully illustrate this system, but a few lines may be devoted 

 to explaining the general principles of the method. The 

 whole Library must be regarded as being divided into nine 

 libraries, numbered as follows: (i) Philosophy, (2) Religion, 

 &c., (5) Natural Science, and up to (9) History. Each of 

 these is again divisible, if necessary, into nine parts. 



Thus the number 54 represents the 4th division 

 (Chemistry) of the 5th class (Natural Science) ; 541 

 represents Theoretical Chemistry, and 5412 represents 

 the 2nd division (Atomic Theory) of Theoretical Che- 

 mistry. 



Every book as it is received in the Library will receive 

 a number, which will at the same time indicate its place 

 on the shelves, and be a summary of its contents. 



When he receives the title from a reader, the attendant 

 will, after a little practice, be able to go directly and without 

 reference to the place marks, to the exact shelf or quarter 

 of the stores by simply translating the title into its 

 corresponding number. 



One advantage of the system is that the special works 

 contained in the Library on a given subject can always be 

 seen together at a glance. It is needless to point out 

 that the complex character of many books will furnish 

 complex exceptions to the more simple nomenclature. 



The Library, as an all-round, modern, working student's 

 library in science and literature, is a very valuable one, 

 though it is not at present so in the sense that it contains 

 any particular literary treasures. 



The administration of these several institutions, to- 

 gether with the Botanic Gardens at Glasnevin, and the 

 Metropolitan School of Art, is carried on by the Science 

 and Art Department, which is represented locally by the 

 Director of the Science and Art Museum, under whom 

 there are heads of the several departments and institu- 

 tions. Two local bodies were created in 1877 to aid the 

 Department in the supervision of these institutions, one 

 the Board of Visitors of the Museum and Botanic Gardens, 

 and the other the Council of Trustees of the National 

 Library, the functions of the latter including the selection 

 of books for the Library. 



The total cost of the several institutions is provided 

 in the annual estimates of the Science and Art Depart- 

 ment which are voted by Parliament. 



COMPARISON OF THE SPECTRA OF NEBULA 

 AND STARS OF GROUPS I. AND II. WITH 

 THOSE OF COMETS AND A UROR^} 



II. 



General Comparisons. 



T N the preceding article I showed that the spectra of 

 ■*• nebulce, aurorae, bright-line stars, and stars of Group 

 II. are closely related to the spectra of comets. In the 

 table which follows, all the spectra are brought together 

 and compared. It is not sufficient to show that each group 

 resembles comets in some respects, as each one might 

 have some feature which was absent in the other. I 

 therefore give the following table to show how far they 

 resemble each other. In the last column the dark bands 



' Continued from p. 345. 

 NO. 1086, VOL. 42] 



which are simply due to absence of radiation, and are 

 not really absorption-bands, are omitted. 



It will be seen that there are three flutings which run 

 through the five columns, namely, 468-474, 517, and 558— 

 these are due to carbon and manganese, and are the 

 familiar cometary bands ; four more— hydrogen 486, mag- 

 nesium 500, magnesium 521, and lead 546 — occur in four 

 out of the five columns. Out of the thirty-four lines or 

 flutings given, there are nineteen which occur in less than 

 three columns, but this number is greatly reduced when 

 slight differences of temperature, masking effects, and the 

 exceptional cohditions of comets are taken into account. 



It is now universally agreed that comets are swarms of 

 meteorites, and the tables which I have given show that 

 nebulae, bright-line stars, stars with mixed flutings, and 

 the aurora, have spectra closely resembling those of 

 comets, the special features of which are the carbon 

 bands, to which I have recently added the absorption 

 bands of manganese and lead ; all are therefore probably 

 meteoritic phenomena. 



The following is a list of the bodies which contain 

 either one or both of the carbon flutings near 517 and 

 468-474, the latter being a group of flutings, which, as I 

 have before shown (Roy. Soc Proc, vol. 35, p. 167), 

 sometimes has its point of maximum brightness shifted 

 from 474 to 468. The fluting near 564 has been omitted 

 from the table, as it is generally masked, either by con- 

 tinuous spectrum or by the superposition of the fluting of 

 manganese near 558. The wave-lengths given are as 

 measured by the various observers stated. 



The spectrum of the aurora is added for the sake of 

 completeness. 



It will be seen from the table that the record of the 

 presence of carbon is unbroken from a planetary nebula 

 through stars with bright lines to those resembling o Her- 

 culis, i.e. entirely through Groups I. and II, of my 

 classification. 



