^>, v.^. g; i^ I r-i 1' .£ I .^'"> IMAGE EVALUATION TEST TARGET (MT-3) 1.0 I.I US ^^ itt iiii 12.2 Fholographic Sdmoes CarpojcBton 1-25 1 1.4 pi.6 CIHM/ICMH Microfiche Series. CIHM/ICMH Collection de m Canadian Inttituta for Historical IMicroraproductions / institut Canadian da microraproductions liiatorlquas Technical and Bibliographic Notas/Notas tachniquas at bibliouraphiquaa Tha Institute has attamptad to obtain tha bast original copy availabia for filming. Faaturas of this copy which may ba bibllographlcally unique, which may alter any of tha images in the reproduction, or which may significantly change the usual method of filming, are checked below. P D D D Coloured covers/ Couverture de couleur I I Covers damaged/ Couverture endommagAe Covers restored and/or laminated/ Couverture restaurte at/ou pallicuMe Cover title missing/ Le titre de couverture manque Coloured maps/ Cartes gAographiques en couleur Coloured ink (i.e. other than blue or black)/ Encre de couleur (i.e. autre que bleue ou noire) I I Coloured plates and/or illustrations/ Planches et/ou illustrations an couleur Bound with other material/ Rail* avec d'autres documents Tight binding may causa shadows or distortion along interior margin/ La raliure serrAa paut causer de I'ombre ou de la distortion la long de la marge IntMeure Blank leaves added during restoration may appear within the text. Whenever possible, these have been omitted from filming/ II se peut que certalnes pages blanches ajout^as lors d'une restauration apparaissent dans le texte, male, lorsque cela Atait possible, ces pages n'ont pas 4t4 filmAas. L'Institut a microfilm* le meilieur exemplaire qu'il lui a At* possible de se procurer. Les details da cat exemplaire qui sont peut-Atre uniques du point de vue bibllographique, qui peuvent modifier une image reproduite, ou qui peuvent exiger une modification dans la mAthode normala de filmaga sont Indiqute ci-dessous. D J? D D D D D D D D Coloured pages/ Pages de couleur Pages damaged/ Pages endommagtes Pages restored and/or laminated/ Pages reataurtes et/ou pellicultes Pages discoloured, stained or foxed/ Pages dAcolortes, tachetAes ou piquAes Pages detached/ Pages d6tachAes Showthrough/ Transparence Quality of print varies/ Quality inAgale de I'lmpression includes supplementary material/ Comprand du matArldl suppMmentaire Only edition available/ Seule Mition diaponible Pages wholly or partially obscured by errata slips, tissues, etc., have been refilmed to ensure the best possible image/ Lea pages totaiement ou partiellement obscurcies par un feuillet d'errata, une pelure, etc., ont 4t* filmtes A nouveau de fa^on A obtenir la meilleure Image possible. Additional comments:/ Commentairea supplemental res: This item is filmed at the reduction ratio checked below/ ■ Ce document est filmA au taux de rMuction indiquA ci-deaaous. ^ 10X 14X 18X 22X »X aOX 1 C X 1 12X 1«X aox MX 2BX ax tails du odifiar una maga Tha copy filmad hara has baan raproducad thanks to tha ganarosity of: Library of Congrass Photodupiication Sarvica Tha imagas appaaring hara ara tha bast quality posslbia considaring tha condition and lagiblllty of tha original copy and in kaaping with tha filming contract spacif icatlons. L'axamplaira film* fut raproduit grAca A la gAnArositA da: Library of Congrass Photodupiication Service Laa imagas suivantas ont At4 raproduitas avac la plus grand soin, compta tanu da la condition at da la nattatA da I'axampiaira filmA, at an conformity avac las conditions du contrat da filmaga. Original copies in printed paper covers ara filmad beginning with the front cover and ending on the last page with a printed or illustrated impres- sion, or the back cover when appropriate. All other original copies are filmed beginning on the first page wKh a printed or illuetratad impres- sion, and ending on the last page with a printed or illustrated impreasion. IS Lee exemplaires originaux dont la couverture en papier est ImprimAe sont filmte en commen9ant par la premier plat at en termlnant soit par la darnlAre page qui comporte une empreinte d'Impresslon ou d'illustration, soit par la second plat, selon le cas. Tous las autrea exemplaires originaux sont filmte en commen^ant par la pramlAre page qui comporte une empreinte d'impression ou d'illustration at an termlnant par la darnlAre page qui comporte une telle empreinte. The laat recorded frame on each microfiche shall contain the symbol -^ (meaning "CON- TINUED"), or the symbol ▼ (meaning "END"), whichever applies. Un des symboles suivants apparattra sur la darniire image de cheque microfiche, selon le cae: le symbols — ► signifie "A SUIVRE", le symbole V signifie "FIN". Maps, plataa, charts, etc., may be filmed at different reduction ratioa. Thoae too large to be entirely included in one expoaure ara filmed beginning in the upper left hand corner, left to right and top to bottom, as many framaa as required. This following diagrama illustrate the method: Lea cartas, planchae, tableaux, etc., peuvent Atre filmte A des teux de reduction difftlrents. Lorsque le document est trop grand pour Atre reproduit en un saul cliche, 11 aet f llmA A partir da Tangle supArleur gauche, de gauche A droite, et de haut en bas, en prenant la nombre d'lmagaa nteassaira. Lea diagrammes aulvants illustrant le mAthode. errata I to t I pelure, on A ■mmmbbI 1 2 3 ax 1 2 3 4 5 6 :¥'■.- ■J.. 6^ i'/. .MS REPORT TO TUK ^•!*»'^1'^'V SECRETARY OF THf 1 AVY ui( RECENT IMPROVEMENTS IN ASTRjIpim INSTRMNTS: ■-:.? ','1 ■1' >■'■.( :)| ■I"' By SIMON NF.WOOMB, FBOVBBROB, W. «• WAVV. , -- ' HHMTOfi^MMM ' i884. - ^ ' , ) L,i£^ f -M ■-t..V ■•■.v.*-»^*Me^?f?^*^rrv-r^rV'SP->-v }-<■'' ■ \i*-, i''.v,,vt>'r ■ ."r^."--_:i(A,-..'^.!-r, .-av-.':^.- V - ■I . u. 1 ft _jf.v "i %f^ 'W i'l^S I t REPORT TO THa SECRETARY OF THE NAVY oir RECENT DIPROYBMENTS IN ASTRONOMICAL INSTRUMENTS. Bv SIMON NEWOOMB. noraasoB. v. & vxrr. "WASfllNQTOK: OOTBBIfMSRS PUIKTIKO 0»»101. 1884. *^ ' .aMK'i waa i&ffin w jw a aMMw^wiwwi*^^ •»•■■■. ■: I 46th Congress, 1st Session. SENATE. Ex. Doo. No. 96. MESSAGE FROM THE PRESIDENT OF THE UNITED STATES, TRANSMITTING i " A communication from the Secretary of the Navy, including a report on recent improvements in astronomical instruments. Fbbruary 12, 1884.— Read and referred to the Committfle on Naval Affairs and ordered to be printed. To the Senate: I transmit a communication, under date of the 8th instant, addressed to me by the Secretary of the Navy, covering a repoi;t of Prof. Simon Newcomb, United States Navy, on the subject of recent improvements in astronomical observatories, instruments and methods of observation, as noted during his visit to the principal observatories of Europe, in the year 1883, made iu pursuance of orders of the Navy Department. The request of the Secretary is commended to the consideration of Gongress. ExBCUTivB Mansion, February 11, 1884. CHESTER A. ABTHUB. II JTavy Department, y\/,i3hington, February 8, 1884. Sib : I have the honor to transmit the report of Prof. Simon Newcomb, United States Navy, on recent improvements in astronomical observa- tories, instruments and methods of observation, as noted during a visit to the principal observatories of Europe in the spring and summer of 1883, made in pursuance of orders of this Department. The report is considered of sufficient importance to be laid before Congress and printed . Very respectfully, your obedi^t servant, WM. E. CHANDLER, Secretary of the Navy. The Pbb8idbnt« ■Immm iiiniiniiiiiiiiiiii IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. mA Nautical Almanac Office, Navy Department, 'Wcaihxngton^ I), C, January 20, 1884. Sir: In pursuance of onlers from the Department I visited certain of the leading observatories on the continent of Europe during the past year for the purpose of collecting information respecting the most recent improvements in astronomical instruments and methods of observation. I now have the honor to submit the following report upon the knowl- edge thus gained. The heads of this report are not arranged with respect to the different establishments visited, but with respect to the different kinds of instruments, all that relates to each instrument being collected together, even when the material was gathered at various places. The establishments visited fh>m which valuable information was gath- ered, were the observatories of Paris, Neuchatel, Geneva, Vienna, Ber- lin, Potsdam, Leyden, and Strassburg, and the workshop of the Messrs. Bepsold at Haml>urg. At the latter place I enjoyed the opportunity of meeting Director Struve, of the Pulkowa Observatory, and of discuss- ing with him and the Bepsolds the plans of the great 30-inch refhiotor, the objective of which had just been completed by the Messrs. Olark. It is both a duty and a pleasure to acknowledge the very cordial re- ception I met from the directors and astronomers of the various obser- vatories, and the facilities which were everywhere afforded me for the execution of the mission with which I was charged. In every case the fullest liberty was accorded me to make as critical an examination of every iwint as circumstances permitted. In this report it is not practicable to present that exhaustive discus- sion of the subject of recent instruments which might have been ex- pected, for the reason that within the limited time at my disposal it was not possible to prepare the detailed drawings and make the tests which would have been necessary for that purpose. I shall therefore confine my report to such special points as appear most important to persons who may intend to found new observatories, or to design or purchase new astronomical instmments. THE great yiENNA TBLBSOOPB. Among the InstmmentB which I have examined, that to which most interest now attatches is the great telescope recently completed for the Imperial Observatory at Vienna, by Howard Ombb, esq., of Dublin. It is the largest refracting telescope in actual use at the present time, RUMENTS. iSPABTMBNT, January 29, 1884. 1 1 visited certain of 3pe dnrlog the past itinjc the most recent hods of observation, art upon the kuowl- I not arranged with with respect to the ich instrument being gathered at various aformation was gath- Qeneva, Vienna, Ber- rkshop of the Messrs. ed the opportunity of fctory, and of disouss- reat 30-inch reflractor, by the Messrs. Olark. » the very cordial re- of the various obser- B afforded me for the d. In every case the cal an examination of at exhaustive discus- might have been ex- at my disposal it was make the tests which hall therefore confine important to persons to design or purchase Bd, that to which most ntly completed for the ^mbb, esq., of Dublin. se at the present time, IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. 8 being of one inch greater aperture than that of the Naval ObHcrvutory at Washington. The contract waM made with Mr. Grubb in 1875, but, owing to diflBculties in procuring glass disks of the necessary size and purity, it was not completed until 1881. Further delay occurred in mounting, so that it was barely ready for active work at the time of my visit in April last. I made as careftil and critical an examination of its working as was possible during the unfavorable weather which pre- vailed at that time at Yien.ia. My examination was principally in the nature of a comparison of its working with that of the Washington telescope. General etyle of mounHng. — In its main features the telescope is mounted on the same fundamental plan as that at Washington, each being on the (German plan, with the Gtorman system of counterpoises and with a steel tube. In both, the rapid motion in declination is by means of a rope attached to the two ends of the tube; that in right as- cension by a system of wheel-work. In both, the clock-work is in the pier below the instrument. The leading points of difference are, that the mounting of the Vienna telescope is much larger, stronger, and heavier in all its parts, that the appllMices for using it are more elab> orate and numerous, and that an elaborate system of friction rollers in declination is provided, while the Washiflgton telescope has none. A more convenient system of illuminating the field and the divisions on the several circles has also been introduced. As a piece of mechanical engineering it reflects great credit upon its designer and constructor. Eaee of motion. — In moving the Vienna telescope one is at first struck with the tact that mere weight is a serious drawback in the management of such an instrument, but, when the motion is once commenced, the movement in right ascension is almost as easy as in the Washington telescope. It is, however, very different in declination. For reasons which neither Dr. Weiss nor myself was able to perceive, the Motion rollers seemed to be of little benefit in easing the motion in declination, which was much more difficult than in the Washington telescope, and, in &ot, quite a tax upon the strength of the observer at the eye-piece. The quick motion for setting in right' ascension is made below the end of the polar axis by turning a steel steering-wheel. This appliance is in every way inferior to the system at Washington, where the same motion ia effected by an endless rope hung over a grooved wheel which the observer palls hand over hand. By this axnuigement the observer at the Washington teleooope can make the required motion, with his IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. '- :. 'if • iH'il ■-*'! eyes fixed upon the tcIcHcope or upon the vernier, as he maj' desire, and without giving any thought to the motion of the hands. But the handles of the steel wheel are much less convenient to take hold of than a rope ; and if the motion is at all rapid the operator cannot turn his eyes to the moving tele8coi»e without danger of his knuckles being struck by the steel handles as he attempts to take hold of them without look- ing. The necessity for oare in this respect makes the motion hesitating and laborious, at least to one unaccustomed to it. The clock motion. — On the system of the Messrs. Clark, applied in the Washiugton telescope, the screw which turns the sector does not take hold of the circumference of the latter directly, but gears into a complete worm-wheel, around the axis of which is wrapped a pair of brass or steel bauds which also enwrap the arc of the sector. By this arrangement the toothed wheel makes a nearly complete revolution while the sector is moving through its arc, and the effect of the small unavoidable irreg- ularities in the working of the screw is diminished in the ratio of the arc of the sector to the circumference of the wheel. Whatever advan- tages this arrangement may have in small Instruments, I think that in large ones they are more than counterbalanced by the irregularities arising from the elasticity of the band, combined with the variations of firiotion, and the action of wind and other forces operating to vary tbo* uniform motion of the telescope. Owing to this elasticity, the eftioct of the wind or of any slight pressure by the observer on the eye-piece is many times greater in the Washington than in the Vienna instrnmoat. But it did not appear to me that the firmness of the connection in the latter instrument between the support of the turning screw and the tube of the telescope was as great as was expected by those who lay stress on large and stable mountings. I found that by a simple press- ure of the thumb-nail upon the tube of the Vienna telescope the point- ing in right ascension could be changed by several seconds so as to throw an object entirely away from the wire. The main question is, however, the steadiness of motion when no pressure whatever is applied by the observer, and so far I have found no large telescope which is entirely satisfactory. The Vienna telescope was not supplied with a micrometer at the time of my examination, BO that I could not test its motion as thoroughly as I wished to; but by bringing the planet Uranus in the edge of the field I found that when the clock was going there was a constant irregular movement in right ascension, the amount of which I estimated as between one and rRUMENTS. • 8 ho mny desire, and u hands. But the to takehoUl of than cannot tarn his eyes uvkles heing struck them without look- the motion hesitating Clark, applied in the sector does not take gears into a complete A pair of brass or steel By this arrangement ition while the sector all nnavoidable irreg- ed in the ratio of the b1. Whatever advan- ments, I think that in by the irregularities with the variations of operating to vary tho» elasticity, the effect of ver on the eye-piece is ;he "Vienna instrumont. if the connection in the turning screw and the Bcted by those who lay that by a simple press- ana telescope the point- everal seconds so as to less of motion when no and so far I have found . The Vienna telescope ime of my examination, ;hly as I wished to; but if the field I found that t irregular movement in bted as between one and 1MPR0VBMENT8 IN ASTRONOMICAL INSTRUMENTS. 5 two HcuundH of arc. TIiIh niovcnivnt had no regiihir p«>ri(Ml, and there- fore did not Heem to be connected witli any defect in the flgnrc or mo- tion of the Hcrew. Its irre(;nhir period, if I may use the term, viiried from the smalleHt nppreciublo amount to two or tlir<>e 8euondH of time. Itti nioHt probable cauHe seemed to be the variable friction of tlie mo- tion in riglit ancenHion and CHpecially of the friction rollerH by which the polar axiH is Hiipporte'S34I8V'. : t!»«isa»»«tn(^' '.*" ■ 6 IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. pointing in declination, except when the observer was at the eye-piece. This want, combined with the great force necessary to move the tele- scope in declination, makes its pointing a diflBcult and troublesome operation. The observer must first set the telescope by pure guess-work. He has then to mount to the eye-piece, wherever it may be, look into the niicroscope, and note the reading of the circle. He has then to with- draw his eye, and by considerable muscular exertion to make another guess, which he again tests by reading the circle. Thus the pointing is to be made by a series of trials which are so troublesome that I found the observers were in the habit of mounting to the top of the cylinder of the dome and finding the pointing in declination by moving the tele- scope around the horizon. I remark, in this connection, that the Washington telescope has a coarse setting which the observer can read from any point below the telescope with the aid of an opera-glass. This setting is sufficiently accurate to bring any object whose position is known into the field of view of the finder, and near its center. Objective. — ^The proper fi^ring of an objective so as to give the best possible image, is justly considered th<) most difficult task in the con- struction of a large telescope. Especial interest, therefore, attaches to Mr. Orubb's success with the objective. The atmospheric conditions during my visit were unfavorable to the finest tests, but I succeeded in making such examination as the cirouihstances admitted of en three evenings. On the first trial, the image was found to be defective, owing to a want of a4iastment of the glass itself. This was corrected next day by Director Weiss. On the second trial I found a well-marked spherical aberration, which seemed, however, to be very regular from center to circumference. But there had been a fall of temperature and the dome had been opened only a short time; circumstances under which the Washington telescope always exhibited the same phenomena. On the third evening, the dome had been opened long enough to nearly equalize the internal and external temperatures. So far as I could judge the character of the image was perfect, there being no appearance of those rings of different focal length, which are so often found in large objectives. As I hacl not used a large telescope for some eight years, I could not feel that my judgment was of the most critical kind, but I am persuaded that if any defects exist, they are so minute as not to interfere in any important degree with the finest performance of the instrument. The color correction is less than in the Washington telescope. The result is that the blue areole around brilliant objects is much less striking. rRUMENTS. iras at the eye-piece, py to move the tele- It and troublesome by pure guess-work, t may be, look into He has then to with- )n to make another Thus the pointing blesome that I found top of the cylinder I by moving the tele- rton telescope has a any point below the etting is sufficiently »wn into the field of 10 as to give the best cult task in the con- iherefore, attaches to nospheric conditions », bat I succeeded in idmitted of en three l» be defective, owing 1 was corrected next found a well-marked be very regular from 11 of temperature and astances under which bme phenomena. On ng enough to nearly 3o far as I could judge Dg no appearance of ) often found in large >r some eight years, I critical kind, but I am ate as not to interfere oe of the instrument, igton telescope. The 3 is much less strikiug. IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. 7 THE GREAT DOMES AT PARIS AND VIENNA. The proper perforoiauce of a large telescope is so much affected by the character of the dome in which it is placed, that the latter may be regarded as of equal importance with the mounting of the telescope. All my experience, however, leads me to the conclusion that there is no decided superiority in any special form of dome, but that the princi- pal difference in the working arises from the quality of the workman- ship. In choosing among a number of proposed forms we can only say that that which is best constructed is the best. The Vienna dome was constructed by Mr. Grubb. It is built of iron, is 45 feet in external diameter, and weighs fifteen tons. Its working leaves nothing to be desired, except that its great weight renders its motion somewhat cumbrous. By moving it a short distance it appeared that one man could turn it in eight minutes. Mr. Grubb says that when first mounted a weight of seven poands on the rope was sufficient to start it. I did not test this by actual trial, but cannot resist the con- clusion that much more than seven pounds is now required. The drum of the dome is of thick massive brick- work. I cannot but regard this sort of base as objectionable, even when pierced with nu- merous openings, as in the present case, owing to the difficulty of se- curing equable temperatures inside and outside. Heretofore three methods of supporting a turning dome have been proposed: I. Cn wheels, fixed either to the dome or to the base on which the dome rolls. II. On a system of rollers connected by a live ring. III. On cannon balls. I conceive that the choice should lie between the last two. The can- non-ball system works the best of all so far as ease of motion is con- cerned ; its drawback is the difficulty of keeping the balls at anything like equal distances apart. The system of rollers in a live ring was in- vented by Mr. Grubb, and is employed at Washington as well as at Vienna. It has the advantage of always working well, but is more troublesome in construction and requires more force than cannon balls. To these three systems the French astronomers propose, in their new great dome, to add a fourth, by floating the dome in an annular trough forming the top of the drum. The base of the dome will then be a float- ing annular caisson. It woald be hasardoas to predict in advance how ■' ; i-..- <.. M l:;i 8 IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. this ingenious plan will work. It will certainly have the advantage that a slow motion can be given with less expenditure of power than on any other system ; but if the motion is at all rapid I am inclined to suspect that the friction of the fluid will be equal to that of the rollers on the other system. The difficulty which I should principally fear is the leaking of the caisson. The freezing of the water will be avoided by impregnating it with chloride of magnesium. It is intended to con- struct a dome on this principle 20 meters in diameter for the great re- fracting telescope now being constructed. It is feared, however, that the practical completion of the work will be long postponed, owing to the necessity of finding a better foundation for the structure than is now afforded by the grounds of the observatory. THE GREAT BUSSUN TELESCOPE. In 1879 Privy Counselor Otto Von Struve, director of the Pulkowa Observatory, visited this country and contracted with the Messrs. Clark for the construction of an objective 30 inches in aperture. It was com- pieced and delivered during the year 1882. The mounting is now being completed by the Messrs. Bepsold, of Hamburg. Although still nnfln- ished, I was desirous of gaining all the information possible respecting its construction, and therefore visited Hamburg for the puri>08e of ex- amining its parts. The following are some essential points in the struc- ture: The most striking feature of the instrument will be the absence of fHction rollers from the declination axis. With so large an instniment the firiction on the declination axis will be too great to admit of the tele- scope being conveniently turned either by hand or by a rope attached to the two ends, as at Washington and Vienna. The quick motion in declination will be given by a system of cog-wheels turned by an axis passing through the polar axis of the instrument and coincident with it. This axis will be turned by a crank at the lower end, or by the ob- server taking hold of the circumference of a wheel, at choice. Although the turning of the crank is a more convenient motion for the purpose than that of taking hold of the handles of a steering-wheel, I do not consider it so convenient as pulling a rope. This system of wheel- work will also be connected with the axis of a crank at the eye-piece which the observer can take hold of and turn without leaving the eye-end of the telescope. A second crank will be furnished for the motion in right ascension. 5TRUMENTS. have the advantage diture of power than apid I am inclined to to that of the rollers aid principally fear is trater will be avoided It is intended to con- Bter for the great re- feared, however, that postponed, owing to structure than is now DPB. •ector of the Pulkowa with the Messrs. Clark bpertnre. It was com- mounting is now being Although still nnfln- 9n possible respecting • for the puriK)se of ex- tial points in the strnc- will be the absence of 90 large an instniment Mit to admit of the tele- or by a rope attached . The quick motion in eels turned by an axis it and coincident with )wer end, or by the ob- )l, at choice. Although notion for the purpose leering- wheel, I do not is system of wheel- work at the eye-piece which leaving the eye-end of for the motion in right IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. 9 Instead of using a sector for the clock motion the screw will gear into a complete wheel about two meters in diameter. The trouble of having to turn the sector back will thus be avoided. The illumination of the finding circles and the arrangements for reading them will, in their results, be similar to those used on other large telescopes: that is, the arrangement will be such that the observer can read either circle from the eye-piece. The system of illuminating the field wires, micrometer, position circle, &c., though extensively employed in Europe, is so little known in this country that attention should be called to it. The side of the telescope at a convenient distance above the eye-end is pierced by an opening on the opposite side from the declination axis. Through t&is opening passes a conical tube parallel to the declination axis. At the outer end of this tube is a reflector inclined at an angle of 45 degrees to the axis of the cone, but turning- on an axis coincident with that of the cone. The illuminating lamp shines upon this reflector and turns upon the same axis with it. It is also hung upon gimbals so as to turn upon a secondary axis coincident with the axis of its own line of light. The resnlt of this arrangement is that the lamp always hangs vertically, whatever the position of the telescope, and that the horizontal beam of rays thrown ftom it always strikes the mirror at an angle of 45 de^^rees in such a way as to throw the light directly through the conical tube and into the telescope. The slightly divergent beam which fills the cone is divided into two or three concentric portions. One of these is reflected upward to the object-glass, and by reflection firmn the glass itself illuminates the field of view. Another portion shines upon four whitened surfaces around the sides of the micrometer, by which both sets of wires are illuminated. The portion of light which is not needed for this purpose is so arranged as to illuminate the two verniers of the position circle and the heads of the micrometer. So far as I could judge, the working of this plan leaves nothiL„' to be desired in the way of convenience to the observer. Worthy of special attention are the> eye-piece micrometers now made by the Messrs. Bepsold. They include every contrivance necessary for rapid and convenient use. Support of the polar axis. — Another important feature, which has been applied by the Bepsolds in their other large instruments, is the method of supporting the iralar axis. This axis has to bear a large part of the instrnment, counterpoises included. Asjordinarily made, it is necessa- rily snbleot to an end thrust equal, in our latitude, to two-thirds the ,.^;;M*r«W«' ^"aWirflVi-ilr' • V 10 IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. weight of the instrument. How to support this thrust without inter- fering with the ease and freedom of motion has been one of the difficult problems in mounting a telescope. In the Bepsold instrument the thrust is nearly avoided by supporting the polar axis upon a vertical friction -wheel under the center of gravity of the entire instrument. Oounterpoises can be placed at the lower end of the axis so as to bal- ance the instrument upon this wheel. So far as I can judge, this plan leaves nothing to be desired. PBACTIOAL OONOLUSION8. I have been led by the examination above described, combined with some experience in the use of the Washington telescope, to some con- clusions resiiecting the most appropriate features in the mounting of an instrument of the larger size. They may be here enumerated for the consideration of those engaged in cbnstnictions of this kind. I. I think that in order to secure the necessary stiffness with the least weight the axes should be hollow. The material can then be made comparatively thin. It is true that the larger the axis the greater the friction. But the mass of metal in the interior of the axis contributes so little to its stiffness that the external diameter will have to be in- creased very little to secure the same stiffness with the hollow axis as with the solid one. II. It is not worth while to supply the declination axis with friction rollers unless experiment and research shall show that they can be made more effective than they appear to be in the Vienna instrument III. The best quick motion in right ascension is that adopted in the Washington telescope, where the observer i)ulls an endless rope hand over hand, and can lock and unlock the gearing which connects the turning-wheel with the telescope at pleasure. lY. If, as is possible, the quick motion in declination, by means of a loose rope attached to the two ends of the telescope, requires too strong a pull, the best method of giving this motion is through a gearing turned by an axis passing centrally through the polar axis on the Bepsold plan. But it is preferable to have this motion made by turning a crank or pulling a rope rather than by taking hold of a wheel. V. Coarse divided wheels should be supplied, so that the observer while turning the instrument can constantly see its approximate point- ing. It is better if this coarse reading can be made with the naked eye, as is the case In the right ascension movement of the Washington tele- TRUMENT8. bhrast without inter- in one of the difficult sold instrument the axis upon a vertical e entire instrument, the axis so as to bat - can judge, this plan ibed, combined with Bscope, to some con- 1 the mounting of an enumerated for the ;his kind. y stiffness with the ial can then be made axis the greater the the axis contributes ' will have to be in- h the hollow axis as on axis with fHotion lat they can be made a instrument that adopted in the a endless rope hand which connects the ition, by means of a ), requires too strong igh a gearing turned bxis on the Bepsold B by turning a crank rheel. 10 that the observer B approximate point- ) with the naked eye, he Washington tele- IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. 11 scope. The declination circle being farther from the observer, it has to be read with an opera-glass if more than a coarse fraction of a degree is required. By such an arrangement the telescope can always be set by the quick motion so nearly that any object sought shall be in the field of view of the finder. In nine cases out of ten this will be all that is required in practical use. It should never be forgotten that in all quick motions it is very desirable that the observer shall be able to keep his eye upon the movements of the telescope itself in order to save him firom any apprehension, even a groundless one, that something may be going wrong. VI. The slow motion should if possible be endless. There is no dif- ficulty in making it so in right ascension ; though there may be in dec- lination. VII. When the instrument is so large that there is an interval of three feet or more between the center of the polar axis and the side of the tube, the screw which communicates the clock movement should be geared into a complete circle rather than into a sector. The use of the metal band to multiply the effective radius of the wheel offers no advan- tage in the case of large instruments to compensate for the disadvan- tage of want of stability arising from elasticity of the band and its fast- enings. VIII. In this connection should be considered the question of apply- ign the system of Aing, which consists in giving a dock-motion to the verniers of the right-ascension circle so that their position shall repre- sent sidereal time. Every practical astronomer is familiar with the trouble in setting an ordinary equatorial, arising firom the necessity of having to calculate the constantly varying hour angle of the object on which he points. With the Greenwich arrangement there is no such trouble. The damping-wheel being once set to sidereal time, the ob- server has only to set the other one to the constant right ascension of the object. It is true that practical difficulty arises in the usual con- struction, owing to the fact that the vernier on the gear-wheel will from time to time be on every point of tfa circle. But this difficulty can, I think, be obviated by appropriate arrangements. IX. A dock motion which can be kept up by water or other power is greatly preferable to any system which requires an assistant to wind up a weight. X. The entire practicability of illuminating the divisions of the circles by a lamp and of reading these divisions firom the eye-end of the telescope m0iitiiimimmm»mtitm 4^ I 1 ^ - ii /*."' '^^J^i ' '^^^'^f^l Is^^t^^ B 12 IMPBOVEMENTS IN ASTRONOMICAL INSTRUMENTS. has been so completely demonstrated that all large instruments should be supplied with this arrangement. It can hardly be doubted that elec- tric lights will hereafter take the place of lamps for this purpose. XI. The systeVn of illuminating wires, field, micrometer-head, &c., by a single lamp, which shall be vertical in all positions, has been so per- fected by the Eepsolds that it leaves nothing to be desired. XII. The Washington plan of having the whole micrometer plate, in- cluding both fixed and movable wires, moved by a fine screw which has not necessarily^ a divided head, offers such a convenience in setting that it should always be adopted. XIII. The old system of having a single finder on that side of the telescope which is opposite the declination axis becomes very inconven- ient in a large i^nstrument, owing to the necessity of setting the slit in the dome not only to the telescope but to the finder. The plan adopted in the Vienna telescope of having two finders, of which one shall be above and the other below the telescope when the latter is in the meridian, obviates this difficulty and should always be adopted. BEFLECTINO TELESCOPES IN FRANCE. It is well known to all who have given attention to this subject that the optical pertbrmance of great reflecting telescopes has never been proportional to their size, and that the mechanical difficulties of keeping a large reflector in proper figure in different positions have been appar- ently insurmountable. A plan of supporting a large mirror, devised by the Messrs. Henry, has been adopted in Paris, which it is hoped may ob- viate this difficulty. It consists, in principle, in supporting the mirror upon a mass of metal of a form similar to that of the mirror, the surfiEkoe of which is ground to fit the lower surface of the mirror with accuracy when the latter is in proper shape. If the mirror rested directly in contact with this second si^rface no advantage would be gained, since the backing itself would bend as readily as the mirror. Therefore be- tween the two is inserted a thin stratum of some elastic substanqe. M. Henry has found a sheet of fine flannel to give the best results. The effect of the sheet is to diminish the flexure of the mirror by a fraction depending upon its stiffness and upon the elasticity of the flannel. Theoretically it may be considered imperfect, because, in order to act, some stiffhess is required in the mirror itself. A perfectly flexible mir- ror would bend just as much with the flannel as without it But flex- ure of the mirror can, it appears to me, be reduced to quite a small ■MIlMMMMMibB ^STRUMENTS. ge instrumeuts should y be doubted that elec- for this purpose, croiueter-head, &c., by tious, has been so per- be desired. e micrometer plate, in- a fine screw which has enieuce in setting that ler on that side of the tecomes very inconven- Dy of setting the slit la ler. The plan adopted m the bending of the mirror, but no such effect shows itself, at least in any striking degree, in the Paris instrument, which is of about ten English inches aperture. THE BTRABSBUBO MKBIDIAM OIBOLB. This instrument is commonly considered to embody the latest concep- tions in astronomical mechanics. Its general design is founded on that adopted in the great meridian circle of the Harvard College Observ- atory which was constructed by Troughton and Simms, of London. The original design of the latter instrument is, it is understood, largely due to the late Professor Winlock. The most essential modification of the older plan is that the T's and the reading microscopes, instead of being supported upon piers of stone, are borne by a massive metal foundation, the tops of the piers being below the level of the bottoms of the circles. The drawbacks arising fh>m the unequal contraction and expansion of the stone piers, under the influence of variations of temperature, are thns almost entirely avoided, because the metallic supports rapidly assume the temperature of the surrounding air and of the instrument. Every part of the instniment bears the impress of the thought and care devoted to its construction both by the makers (the Messrs. Bepsold) and by Professor Winnecke, the director of the observatory. Even the form of the piers of masonry which support it and its collimators is highly original. The base of the principal pier is smaller than usna!, an^ the amount of material in it is still farther diminished by building it in the form of a Greek cross. The collimators are supported on cy- lindrical piers of the usual oonstmotion. Each of these three piers is protected from changes of temperature by having a hollow cylinder of brick built up around it from the ground. The thickness of the wall of this cylinder is that of one brick. To insure stability the difBarent cyl- inders are connected together by brick arches, but these arches do not exert any pressure upon the interior piers supporting the instrument, which rests only on their foundation. A degree of stability is thns secured which I believe has never before been reached. But this may be in great part due to the excellence of tlie foundation. <8TRUMENT8. work of a certain oUms f increased. The form ich the highest optical the double reflection. 9 the images from the itself, at least in any I of about ten English BOLB. K>dy the latest concep- lign is founded on that 'vard College Observ- nms, of London. The nderstood, largely due al modifloation of the sopes, instead of being sive metal foundation, bottoms of the circles. Jon and expansion of temperature, are thus }port8 rapidly assume e instrument. BS of the thought and } (the Messrs. Repsold) >servatory. Even the and its collimators is is smaller than usual, iminished by building are supported on cy- of these three piers is g a hollow cylinder of lickness of the wall of ility the different cyl- it these arches do not >rting the instrument, le of stability is thus Mshed. nee of tlie foundation. IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. 15 The observatory building rests u^n a stratum of gravel so clean and pure that in case of a flood in the Rhine the water permeates through the gravel to the base of the piers. It might be supposed that water thus penetrating the foundation would produce an injurious effect upon the stability, but such is not found to be the case. The fact that gravel forms the best foundation for an astronomical instimment has long been understood by those who have given attention to the subject. But I do not know of any other case in which the saturation of the gravel with water has been experienced. I may mention in this connection that a solid bed-rock might be even better than gravel were it covered with so deep a layer of soft earth that ii would not be affected by daily or annual changes of temperature. Experience has, however, shown that for want of these conditions being llilfllled a solid rock forms a very unsafe foundation. An interesting example of this is afforded by the observatory at Neuchatel, which is erected at the base of the Jura Mountains. The annual change in the ' pointing of the meridian circle is so great that Dr. Hirsch has recently published an investigation of the subject, showing that the mountain undergoes an annual change to an extent which has never before been remarked. In order to obtain an accurate estimate of the stability of the Strass* burg instrument, I requested the acting director. Dr. Schur, to allow me to transcribe the instrumental oorrectfons during as long a period as practicable. The following are the values of the three instrumental con- stants which depend on the deviation of the axis of rotation from a true east and west line. Column • gives the level correction ; n, the distance of the line of collimation east of the pole; m the deviation of the same line at its point of intersection with the equator. Of these constants n is more accurately determined than either of the others, and its stability affords a test of the stability of the instrument both in level and admnth. i. n. M. 1888. Jane 17 Jnly 3 10 14 15 17 80 87 1. .00 -.08 -.03 : +.03 -.02 +.06 +.01 +.11 +.03 +.06 +.11 «. -.03 -.08 -.11 +.01 +.06 .00 -.01 +.06 -.18 .?. .'' 19 IMPBOV£MENTS IN A8TH0N0MICAL INSTRUMENTS. Sept, Oct. 1889. Aug. 34 i<5 89 1 It 6 V Itf 35 85 4 7 13 17 19 81 8» 86 88 30 31 1 1 3 6 (. +.13 +.89 +.09 +.06 +.01 03 08 01 01 01 83 I Not. loo +.08 +.08 .00 +.01 +.06 —.01 +.01 -■.04 --.03 --.05 -}-.01 -..00 --.03 f *• m. *. f. +.18 -.01 +.06 +.04 +.UJ —.11 .00 +.01 +.09 -.06 +.80 -.19 +.08 —.08 +.02 +.01 -.01 .00 +.07 +.96 +.05 -.05 -.05 --.18 -.04 --.17 -..06 —.06 --.04 -.03 —.02 +.13 +.08 -.10 —.03 +.04 -.18 -■.05 -.08 -.01 ...06 +.03 --.01 .00 -..06 -.06 +.08 —.08 It will be seen that if we take the mean value of n, 0*.06, as a constant for the four months the average deviation of the observed values from this mean will be less than 0*.06. A portion of these deviations are due to errors of determination and the accidental deviations of a temporary character due to changes in th» temperature of the different parts of the instrument, produced by the impacts of air currents. The stability of the nadir point appeared much less satisfactory. This correction was not determined with sufficient regularity to admit of its changes being the subject of a positive calculation, and Dr. Sohur expressed himself not entirely satisfied of the accuracy of the determi- lyitions. It seems scarcely possible that the polar pointing of the in- stmment could be so steady if the nadir point were subject to consider- able changes. The stone piers are terminated on top by horizontal faces, in which are set the two iron supports of the T's and of the microscopes. The base of each is an iron frame 18 inches from east to west, by 24 inches flrom north to south, which is set on the top of the stone and held in place by iMing bedded in cement. Upon this sets the base of the microscope bolder, which rests on three feet, and may be adjusted horisontally by screws. I should myself suppose that greater security against aooi. ■"•MaHUMMMi rSTRUMENTS. / -.01 f.64 -.11 -.06 -.19 -.08 ■f-.OJ .00 ■f.SW -.05 +.18 +.17 —.06 —.03 +.13 -.10 +.04 -.08 -.01 +.03 .00 -.06 —.08 of n, 0».06, as a constant le observed values from bhese deviations are due iviations of a temporary he different parts of the ■ents. much less satisfactory, ient regularity to admit loulation, and Dr. Sohur iccuraoy of the determi- )lar pointing of the in- rere subject to consider- orizontal faces, in which r the microscopes. The ast to west, by 24 inches le stone and held in place le base of the microscope acyusted horisontally by Br security against aoci- IMPROVEMENTS IN ASTRONOMICAL INSTRUMENTS. 17 dental change would have been secured if the top face of the frame had each been planed to lit the base of the holder instead of having the latter on feet, provided some side support were added to guard against any possible minute rocking motion. The mioroitcopes are carried on the external surface of a cylinder, which is oast in the same piece with the base last described. The fhlorum on whioh rests each lever supporting the weights of the instrument at one end and the counterpoise at the other is itself sup- imrted upon the cylinder carrying the microeoopes. This feature of the iu8t.niment has been objected to on the ground that the mioroscoiie caniers should not be sab|}eeted to so great a pressure, which may. pos- sibly b« subject to vibratory changes as the instrument tnrns upon the friction rollers. Although it is quite possible that no actual evil results from this cause, it must yet be regarded as a not improbable source ot danger to the steadiness of the microscope holders ; I therefore con- sider that it would be better to adopt some other system of support- ing the instramait) perhaps upon i^btfs passing centrally through the microscope holders and set in the pier below. The arrangauieiits far okmpiiig the miforosoopes to the cylinder leave nothtef : to b« deeteed, provided tJie fwrner ate once got into their pwper poaiMon. Butth«ta«kofsetlUig«miflra«eppe«ftwitisonoediaarranged is •atratawlir laboiioas, aud aoBie addittonal meehanism fi»r eflboting Hiis isxlMterillew w4M0htlM'divisione ea ,tiieeiiole«recttt,aiid'the adaplaiisu ofiMn AUoroeospe to Ihfsir reading, are nadonbtodlar Ibe gwtort ipaiaite afidilioallit in «ks aoMteaolion^a iMfktA. oind* Af > Die first order. I am oonvinoed that muoh must yet be done to •eoBn^tke beat KMuMs in tUft teapeM. Ls tM' Btg— li iwg oiral% aail In th*otlier lee— jil u s tM irtsof th » Bi ie n l ii S ) itha;diaB|Btog(iif theaiiol^ia i s d u t w d to twoAal. .iiiftro- diMiiitoaUlM iMfethafc, a»tlH»-«iaiiietoi! Ht^liufaiiahrt^— y msm,ai,a, «lwp Miouit i« iiMp«Miti«fc«f .^.division tka>diaaeCcr af <eeii determined for every 5 degrees. The best method of making a numeri- cal estimate of the accidental errors of division seems to be to compare the error of each division with the mean of the errors of the two aton- ing onett differing by 6 degrees. I examined the table of errors through - a portion of the circle with the result that the mean error as thus de- termined is 0''.32, while the maximum is 0".63. It will be interesting to compare this with the Washington circle. On page 37 of the Washington Observations for 1860 is found a table of the errors of the two circles of the Washington instrument. Treat- ing them in the same way, we find: Waabington, circle A: mean error, (K'.28; maximum, W.79. Washington, circle B: mean error, 0".21; maximum, 0".31. Straasbnrg circle: mean error, niiecutive divisions ou the original circle. iSnpposiug the copies U* be absolutely perfect and the original circle to have remained unchanged, all the Repsold instruments should exhibit the «iaiiie errors of division, but I am uot aware whether the uumliering on thedifferentcircles is made to correspond to the samedivisiousof the orig- iaal circles, a pipceeding which would be necessary to test the uniformity. Ill this connection I may remaric that during my visit to the Bepsold establishment they allowed me to examine the divisions on the original circle with the aid of a powerful micrometer microscope attached to it, and used in makingacopy. So fiir as I conldjudge from acursory exami- nation there were no sensible accidental errors of division in the minute a]>ace over which my observations extended. If so, the errors of division in the oopy must be mainly accidental, and perhaps due to elasticity in the mounting of the cuttiiig tool combined with irreguhkrity in the resist- ance it meets with in cutting the metal. Should the errors of division be wholly doe to this cause, we could expect no correspondence betwee^ those on different instruments. Themioraaoopes are about two feet long, and their absolute power is, I think, somewhat less than in the Washington instrument It did uot appear to me, however, firom examining the divisions, that they, would bear any higher power with adyantage. The edges appear deficient in straightness and sharpness, and this appearance is exaggerated by the numerous discolorations upon the silvered surface. The probable error of a single setting of a microscope appeared to be about double of that in the Washington drole, or 0".2 to 0".3 against (KMO to O^Mfi. From these facts I am led to the oonolnsion that an improved system in the con- struction of eiroles is a desideratum. It is true that the necessary probable ervm of astronomical observa- tions arising tnm unavoidable disturbing causes is such that no great additional aoonraoy in single observations would be obtainable by a more aeenrate reading of the circle. The ottfect of increased accuracy of reading is to ftMilitate the determination of errors of division. The latter must be determined with a precidon corresponding, not merely to that of a single obeervati<», but to the mean of a great number of ob- servations. To do this without an enormous expenditure of labor, the j w rt ^w ij m i tt twi w^ i fv'; • ^^ ■ '. ; : ^ ■ , ■_ "■■ V. - -^■^,;::^ , , '■-<'■ "-> ■< \ ' '^•T }■•■'. ■ ■_v,^',- - -^; •--■-- . -.L--V .■■/ ■■*..' '' . J fc '^ '"■ 1 ,.■:- A ■ ) (#■■'■ .■ ■?V'.'-,i 20 IMPROVEMENTS IN A8TRC50MICAL INSTRUMENTS. microscopes mast read with sucb precision that a single (l^terdlination of each division will snffice. The greatest improvement in this direiBtion wonld be made by the introduction of glass circles which have lately been proposed by several American physicists. The practicability of this innovation can, however, only be determined by experiment. Withont pretending to decide, at present, whether glass or metal will prove to be the best material, I do feel that astronomers onght not to rest satisfied with a degree of aecnracy so far behind that reached by the working physicists, who cat one or two thousand divisions to the millimeter, and space them so evenly that their inequalities defy direct n^easorement. Ih tb6 Strassburg' circle an innovation has been made, designed to render nnneceesatty the determination of more titan a limited number of divisibn errors. Oiie of the circles is ditid«d only to every degtvp, and fbut of these degrees, distftnt 90 degrees from each other, are diilded to every two minutes. Thus there are in all 480 divisions on tAie eircle^ and the errors of these can b6 determined with great precision without an inordina(t6 expenditure of labor. With the eirole thus divMed an arc of any required lengtii can be measured, one of whose tenninf shall lie in the degree which is finely divided, and the other on ob* of the entire degrees. To do this it is necessary to adjust the eIrMe ou'th^ axis with each obeamition in such way thait ^at observation shall be made upon the finely divided part, while the nadir or hforiMNotal point shall IMl upon an entire degree. The latter point must then twssparaiely dMArminied for each astronotfdcal observation. I cannot tUnk bat tbsfe the labdr of doing this exceeds the advantage gaiLed by it MXBOUBT BABin FOB KADIB POOfT. I fouWdiit Stbas^bbi'lri I«!j^dMn atid ofthibr oontin«iMaf cbhii)lii!«ltiJM ot$iitiH*h^^ flliig in im fbt avbiflih^ thM dllllotfty, hone of which ate eiitiM^ iAti*- ftaictbfy. Th6 ntb tif a cb^p^r btein to hbld thib mercery #iAiiii^»^«ft In Germany elurly in the ptM^t cibttti^, bat the oondiftons li«iciehind that reached by onsand divisions to the inequalities defy direct /■♦v. been madej designed to uan a limited nnmber of ily to every degmr, and leh other, ore divided to I divisions on fibe circle, great predsion without B drole thns divided an e of whose terminf shall ;he other on ob6 of the accost the dMLe oittM lat obeervatton shall be mdiror iMnriMHrtal: point t must then be ■epamlely I camot think bat that i;aii. only and the copper of the eontaining veaseL I found that in making flD'Obeewation of the nadir point the heavy walking of men avannd the reogoi produced no dietorbance whatever, and evMi stamping on the floor ia the neighborhood of the inatrament only caused a monentary di w i p pearenee «t ti» refleotM image of the wirea la the aftplioatioB of titia form of biMun it is desimble that the plate ahoold be considerably larger tlian the olgective in order that the layer of mercury may be equal to the latter and yet have plenty of margin aMmnd it oh the plate. MHrraoM OP SBTSBMiMnro plbxub% bio. In the use , )■'■'. 1 . f. ; ; ' ■^ -•^.i ,j .'• ■■■ ■ 1 ■ " '■ '^^-i ^>'".. '?.^ * • 1 »i '%' \ '■ , i,"" .^'} T'l y. /i' ^ '■- n > :■:'"-■ 1 ■ '■ fe V '- *:'., -/■.'-', -w»r- :,^-^, 1 ».-,- "■■ 1'^ :% ^';;i these disoordanous to the effect of refraotioo caused by the diftiere nt temiteratares of different strata of the air in the observing room and in the instrument. It is therefore absolutely necessary to success that the determination should be made when there is a perfect uniformity of temperature inside, around, and above the instrument. This was found practicable only in periods of long-continued rain, which cooled off the roof of the building and thus prevented an accumulation of warm air in the upper part of the room. This method has generally been employed only when the instrument was horizontal, and its application in other positions is so troublesome that it has seldom been undertaken. At Paris, however, I found in use on the new.Bischoffisheim circle a vertical collimator which, although apparently not intended for that purpose, could readily be used to deter- mine the flexure in a vertical position. The collimator itself is supported in a horizontal position upon a standard on the east pier, around which it turns upon a vertical axis. When placed in position for use its ob- jective is over the center of the telescope. To receive the rays from the latter a reflecting prism is placed in front of its objective. Thoa the result is optically the same as if the collimator looked vertically down into the telescope of the main instrument. There would be no diffloolty in setting the collimator either upon its reflected image in the baaiM of quicksilver below, or on another vertical telescope below the floor. The apparatus would then be available for the determination of flexure. Mlk lokwy'8 method of mbasubino flbxubb. An ingenious plan has recently been proposed by Mr. Loewy, vioe-, director of the Paris Observatory, for determining the flexure of the telescope in all positions. It has been sa fully described in the OompiM Rwdfu and other publications that I need only here give its general principle. A small glass instrument which combines the function of a lens and a reflector is placed in the central cube of the telescope. The flexures of the two ends of the telescope, relatively to the reflecting surface of this glass, are separately and independently determined in all posttionif of the instrument It is assumed that the glass, b^ing in the neutral axte of the telescope, the astronomical effect of the flexure will be given b;^ the diffierence in the flexures of the two ends relatively to the glass. Ingenious and well considered as this method is, I cannot consider it' reliable for determining the effects of flexure, because it leaves out of mmm I8TBUMEMTS. kosed by the difterent observing room and in ary to snccesa that the perfect uniformity of uent. This was found i, which cooled off the imulation of warm air ir when the instrument dous is so troublesome lowever, I found in use nator which, altliough eadily be used to deter- ator itself is supported last pier, around which position for use its ob- loeive the rays from the ) objective. Thus th« Looked vertically down I would be no diffloulj^ d image in the basin »8cope below the floor, itermination of flexuce. a FLEXUSB. 3d by Mr. Loewy, vice-, ing the flexure of the 9Soribed in the Oompte* here give its general (hnction of a lens and slesQope. The flexures he refleeting surfiioe of mined in all posttioncf of iing in the neutral axiik lexnre will be giv«n bf relatively to the glass, is, I oannot consider it Ijecanse it leaves out of IMPBOVEMENTS IN A8TB0N0MICAL INSTRUMENTS. 23 mmmm*' account the bending of the parts of the instrument between the central cube and the divisions on the circle. In the Washington circle there is a well-marked flexnre of the circle itself, which may be expressed by saying that if the central cube revolves uniformly the circumference of the circle does not revolve uniformly but is affected with a periodic inequality. Hence, to make the determination tree firom all sources of error, the flexure must be determined by a direct comparison of the op- tical axis of the telescope with the reading of the circle divisions under the microecopeti. This can be done only by pairs of opposing collima- tors on the usual or Besselian plan, or on some other plan by which rays can be sent in the same straight line in two opposite directions. OOLLDIATOBS AND MSBISIAN MASKS. l*be old-fashioned system of placing meridian marks at such a dis- tance that they could be observed tiirongh the telescope of the transit circle without changing the astronomical focus may be regarded as now entirely abandoned, owing to the bad effect produced on the images by the passage of the light through several miles of air near the ground. The present system is to place the meridian miurk at a distance of 100 or 200 yards and to render the rays emulating flrom it parallel by a lena of long focus. The plan of putting this lens as a cap over the ob- jective has be«i abandoned, owing to its displacement of the optical center of the combined system of the lens and the ol|jeo^gla88. It is therefore usaally fixed on top of a pier. Bnt at Straasbnrg a differ- ent system is adopted. No lens of long Ibons is osed at all, bat fhe tel- escope is pointed directiy upon the waridian mark and the rays are brought to a focus in the plane of the spider Unes by means of a lens of short focus which can be slid into tiie «ye-pieoe of the tdesoope. On this plan the position of the image will depend upon that of the small lens— a dependence which I think ought to be avoided. The best sys- tem seems to me to be that of the fixed objective of long focus. At Strasabnrg the fixity of the meridian mark is aMoied by support- ing it on a very firm stone foundation and protecting it by a frame build- ing from the rays of the ran. The neoeasitiy of these inecaationB is too obvious to require any comment upon them. At Paris a very ingenious system of refiecting collimators is applied to the great transit circle. As, however, this system was probably adopted only because there was no room for collimators of the usual construction, I did not deem it necessary to prepare a description of u IMPROVEMEINTS IN A80SONOMICAL IN8TSUM£MTS. them. A system of fiducial lines which may be adopted with advantage in ordinary collimators is, however, worthy of note. Instead of setting one spider line upon the image of another, the spider line in one colli- uiator is replaced by a fine transparent line through a narrow band of some oiiaque substance on a plate of glass. Thus when the observer looks at the image of this band in the other collimator he sees in the center of the field a fine horizontal or vertical bright line ou which he can set the dark line of bis collimator with great precision. This plan does not, however, so far as I can see, readily permit th« setting to be made by means of the dark band collimator. Whether this limitation is a serious defect is a question on which opinions may differ. OB8BRVAT0BT BUILDINGS. In the coarse of my journey I had the opportunity of visiting two new observatories of the first class erected within the past law years: one, the Imperial Observatory at Yienna, the other the Astro«Physi«al Observatory at Potsdam. It is generally necessary to design aa •b- servatory with especial reference to the eharactor of the observatians to be made and the objects to be parsued. To this may be add«d the frequent neoeasity Ibr gratifying some public taate witb leapeet ti>>arohi- tectnre. For these reasons one observatory cannot weU serve as a oodel for another; bat there are certain special features which would work equally well under neariy all conditions, and whioh ace therefore worthy of oonsiderAtion in building any observatovy. In?the Yienna Obaervatoij tiie arehitactoral Meatent predeminafles. I did not observe any new featuie of eapcoial impovtaiMe to the deaigii- era «f ftitare observatories exioept those already noted. ■The Potsdam Obserroftofy, as its name implies, was desifaedfwith especial vefennce to physical tribserrakions upon the kewrenly bodies. This braaeh of astrmomy in its pnsent developsMiit te |M new>tiiat every estebUshment ibr proeecating it has ts be ^aoned with refersooe to the speotelw«riE to be doae. Hence, nstfrtlhstonding thufeldM^ob- serratoty in qoestion is, in itsow t i t asd dssign» ons sf ths most pegftst yet bnilt, those featnies of it wUoh it woald 'be advisabls toineofpo- rate in another establishment, baiIt,periM^ for another purpose, would generally ooour to the designers of sasbaa establishment The following are, however, well worthy of ooasideration in all plans of new observa- tories. The sffiBct of the sun's rays npoa the metal roof of the baiUing is to hsat I I mrmmm mmm advantage 1 of setting k onecoUi- »w band of 6 observer aees iu the I which he This plan ttingtobe limitation Biting two bw years : NPhysieal gn aaob- mvatioiis idd«dthe tto>ardii- tsamodel mid work re worthy Nninales. ttd«aigB< raed<^witti lyboditt. new>ttet refirMioe *tk» ob- it poefiwt ineorpo- le^wodid following obnerva- istohoat IMPBOVEMEMTS IN ASTRONOMICAL INSTRUMENTS. 25 I I the air in immediate contact with the roof and thns to injure the definition of the heavenly body seen firom within the observing rooms in the day- time. To avoid this difficnlty the roof is covered with soil and sodded with green turf and thus kept as cool as the surrounding air, how hot soever the sun's rays may be. It is of course necessary to keep the turf watered. Possibly any other absorbent substance might answer the purpose of the turf, but the latter has at least the advantage of cheapness. It is a common feature of all, or nearly all the continental observa- tories, that quarters are provided for the astronomers, generally in the building itself. This offers the great advantage that the astron- omers are nearly always near their instruments, and may be regarded as absolutely essential to the efficiency of any large observatory, espe- cially if it is not in the midst of a city. At Potsdam the houses of some of the astronomers are separated from the building, but the general rule is that the building accommodates the director and such of the assist- ants as are engaged in actual observations, together with their fomilies. This collection of several fiunilies in the same building is more accord- ant with European habits, than with our own, and the question of its introduction among us can be settled only by careful consideration. I can only say that I noticed no serious inconvenience arising firom it. GLOOXS. In most astronomical work of the first dass^ especially in meridian observations, the perfection of the dock is as necessary as that of any other installment But it seems to be an observed fact that no certain way has yet been found of securing an approach to perfection in the rate of the dock. All we can say is, that dooks of marvelous excellence are now md then made, sometimes by one maker and sometimes by an- other, and that of these docks some are permanentiy good while others, in the ooune of time, deteriorate. I found a few examples of docks pre- serving their rate with remarkable uniformity through oonsiderable pe- riods. One of these is the Normal dock of the Berlin Observatory, made by Tiede. It is indosed in an air-tight case in wder to prevent changes of rate arising from variations in the barometric pressure. The tem- perature compensation is unfortunately imperfect, so that the rate is sub- ject to an annual change. This fact has prevented the exact discussion to which I desired to subject it It would seem, however, ifcom a cur- sory examination, the materials for which were courteously afforded me S. Ex. 96 3 WS IMPBOVSMENTS IN ASTBONOMIOAL INSTRUMBNTS. by Professor Fdr»ter, that the aanaal change from temperaturo does not exceed 10 or 15 seoonds per year, and that when this is allowed for the differences between the actual and the computed errors will be a very few seconds per year. In recent times the docks Aimished by Bowhtt, of Amsterdam, have secured a reputation for uniform excellence which has never been surpassed ; that is, instead of being able to occasionally turn out a dodk of remarkable excellence, all the clocks of this artist, so far as they have been discussed, are of the first class. The following exhibit of the observed and computed errors of one of his clocks through a period of nearly two yean has been selected, not from- a belief that this particular clock was better than others, but be- cause the data toe the examinatiMrl, toiaer, OoM«-95. [FopMUUi roB courirtED dailt mu*- +fl^-'IS» - .MM9 of one of loted, not I, bat be- »b$enatory P 1 e onsatis* kg the ex- ck shoald be foonded upon its errors, determined from time to time through a pe- riod of not less than a year. These errors should be exhibited in coD' nection with the mean temperature of the clock-room, and if the dock is not in an air-tight case the height of the barometer should also be given. A calculated error should then be carried through the whole period, in which the corrections for temperature and height of the barom- eter should be introduced. A clock which stands this test well may be presumed beyond doubt to keep its rate during short intervals, which is generally the important point. It is very common to present as sufficient data forjudging of a clock an exhibit of its daily rates firom time to time. If these rates were really determined with the last degree of accuracy they might be sufficient for the purpose. But as found in practice they will be the result, not merely of the actual rates of the clock, but of various personal differ- ences among the observers and changes in the pointing of the instru- ment as well as the accidental errors of observation. From these causes , although the clock were perfect, we might expect an apparent difference of several hundredths of a second between its apparent rate on succes- sive days. The barometric change in the rates of all clocks of the usual construc- tion is so important a drawback that it shoald no longer be tolerated in work of the first class. Two methods have been proposed : the one, that already mentioned, of inclosing the dock in an air-tight case; the other, to supply it with a baromelric OQmpensation. The latter method is undoubtedly the easiest, but where the necessary perfection of ar- rangements can be secured the former must be considered greatly preferable. The grounds of preference are that the air can be exhausted from the case to any extent, thus diminishing its resistance to the mo- tion of the pendulum and permitting a diminution in the driving power. Again, if, instead of air, the case be filled with some gas which does not act on the oU, the slow oxidation of the latter may be prevented^ It may therefore be expected that under this system a clock oould be al- lowed to remain undisturbed for a longer period than under any other. Yery respectfliUy, your obedient servant, SIMON NBWOOMB, ProfMsor, United 8tate$ Kwvy. Hon. W. E. Ghandlbb, Seeretary of the Nary. mm p«*