70 REPORT OF THE SECRETARY. 



May, September, and October, otberwise frequently too warm. At x>resent the 

 change of temperature of the inner room between these warmer and the coldest win- 

 ter months is only a fraction of a degree centigrade, and during an hour's observa- 

 tion it is generally less than one-tenth degree, the control being automatic. 



2. Many holographs of the infra-red solar spectrum have been taken, which, in 

 consequence of these improvements, have yielded results threefold richer in "real" 

 detail corresponding to solar and telluric absorption lines than any hitherto obtained. 



3. About 40 of these holographs have been compared, as described in the Appen- 

 dix to the Secretary's Report for 1896, and 21 of the most perfect have been measured 

 upou the comparator to determine the positions of the deflections found to be "real," 

 or, in other words, corresponding to either solar or telluric absorption lines. These 

 comparator measurements included about 44,000 separate observations. 



There have thus been found over 700 absorption lines in the infra- red solar spec- 

 trum between wave lengths 0.76 /.i and 6.0 /a, an increase of about 500 over last years' 

 results. 



4. With the purpose of making a more accurate determination of the wave lengths 

 corresponding to the well-determined positions of the absorption lines discovered 

 in the rock-salt prismatic spectrum, a very exact comparison of the dispersion of 

 rock salt and fluorite has been made. This comparison will allow the indirect 

 employment of certain recent and apparently very accurate determinations of the 

 wave lengths in the fluorite prismatic spectrum. Apparatus has been made ready 

 and certain preliminary observations have been taken to directly measure the dis- 

 persion of rock salt. It is hoped that these steps will result in furnishing the wave 

 lengths of the infra-red absorption lines to a degree of accuracy corresponding to 

 the exactness of the determination of their prismatic deviations. 



5. Many interesting instances of local variations in the absorption have been 

 noticed. Among these by far the most striking is a great decrease in the absorption 

 at the longer wave-length side of the great band ip at about 1.4 //. This change 

 occurred about February 15, 18U8, and caused the holographs to take on quite a dif- 

 ferent form at the place in question. This new form continued through the months 

 of March and April, but in the month of May the usual form was gradually restored. 

 It is found, by reference to former holographs, that this marked decrease in absorp- 

 tion at this point takes place annually at about the same period, which coincides 

 (fortuitously or otherwise) very nearly with that at which there is the greatest 

 activity of growth in the vegetable kingdom. This raises the question whether the 

 growth of vegetation does not abstract from the air great quantities of some select- 

 ively absorbing vapor active in absorption at this wave length. 



Whether such be the case or not future investigation must determine, but enough 

 variations in the absorption have been observed to indicate that the Observatory 

 is now in condition to make advances along the line indicated in the Secretary's 

 report for 1892, in which is pointed out the important relations of astrophysics to 

 meteorology. 



FULLER DETAILS OF THE WORK OP THE YEAR. 



The best design for a sensitive galvanometer of the Thomson reflecting type was 

 investigated, with the following results: The most suitable electrical resistance of 

 a galvanometer for the holographic work proposed was determined. Expressions 

 were deduced from which the size and relative efficiency of coils of the best form 

 and of a given resistance, but differently wound, could be computed. Computations 

 were made which determined the best sizes of wire and the best apportionment of 

 resistance for a coil of the most suitable total resistance wound in three sections. 

 Four such coils were wound. Several experiments were made to determine the best 

 construction of galvanometer needle, and two needles were made which are found 

 to be very satisfactory. The coils were mounted in a galvanometer case constructed 

 at the Observatory shop, and the needle was suspended by a quartz fiber of exceeding 

 fineness. 



