REPORT OF THE SECRETARY. 73 



"2. The attachment of a dust filter of asbestus to the absorption apparatus pro- 

 duces results that are constantly lower than those obtained without the dust filter. 



"The most reliable method for the estimation of organic matter in air is that 

 known as Remsen's method, and is called Method Ilirt, in this research. ' The pum- 

 ice stone seems to be the best form of absorbent material, because it can be thor- 

 oghly cleansed by heat without changing its condition or usefulness. 



"4. Those methods which determine the organic matter from its reducing action 

 on permanganate do not seem to afford as satisfactory results as those in which 

 the organic matter is estimated as ammonia." 



No. 1038. Smithsonian Physical Tables, prepared by Thomas Gray. (Part of Vol. 

 XXXV of Smithsonian Miscellaneous Collections.) Octavo volume of xxxiv, 301 

 pages. 



"In connection with the system of meteorological observations established by the 

 Smithsonian Institution about 1850, a series of meteorological tables was compiled 

 by Dr. Arnold Guyot, at the request of Secretary Henry, aud was published in 1852. 

 A second edition was issued in 1857, and a third edition, with further amendments, 

 in 1859. Though primarily designed for meteorological observers reporting to the 

 Smithsonian Institution, the tables were so widely used by physicists thgity-after 

 twenty-five years of valuable service, the work was again revised and a fourth edi- 

 tion was published in 1884. In a few years the demand for the tables exhausted the 

 edition, and it appeared to Secretary Langley desirable to recast the work entirely, 

 rather than to undertake its revision again. It was decided to publish a new work 

 in three parts — Meteorological Tables, Geographical Tables, and Physical Tables — 

 each representative of the latest knowledge in its field, and independent of the 

 others, but the three forming a homogeneous series. Although thus historically 

 related to Dr. Guyot's tables, the present work is so entirely changed with respect 

 to material, arran<j,ement, and presentation that it is not a fifth edition of the older 

 tables, biit essentially a new publication." 



The first volume of the new series, the Meteorological Tables, appeared in 1893, and 

 a second edition was published in 1896. The second volume, the Geographical Tables, 

 prepared by Prof. R. S. Woodward, was published in 1894. The volume of Physical 

 Tables, forming the third of the series, was published during this year. It was pre- 

 pared by Prof. Thomas Gray, of the Rose Polytechnic Institute, Terre Haute, Ind. 



In the preface to the tables Professor Gray says : 



"In the space assigned to this book it was impossible to include, even approxi- 

 mately, all the physical data available. The object has been to make the tables easy 

 of reference and to contain the data most frequently required. In the subjects 

 included it has been necessary in many cases to make brief selections from a large 

 number of more or less discordant results obtained by ditterent experimenters. 

 I have endeavored, as far as possible, to compile the tables from papers which are 

 vouched for by well-known authorities, or which, from the method of experiment 

 au d the apparent caie taken in the investigation, seem likely to give reliable results. 



" Such matter as is commonly found in books of mathematical tables has not been 

 included, as it seemed better to utilize the space for physical data. Some tables of 

 a mathematical character which are useful to the physicist, and which are less easily 

 found, have been given. Many of these have been calculated for this book, and 

 where they have not been so calculated their source is given. 



"The authorities from which the physical data have been derived are quoted on 

 the same page with the table, and this is the case also with regard to explanations 

 of the meaning or use of the tabular numbers. In many cases the actual numbers 

 given in the tables are not to be found in the memoirs quoted. In such cases the 

 tabular numbers have been obtained by interpolation or calculation from the pub- 

 lished results. The reason for this is the desirability of uniform change of argument 

 in the tables, in order to save space and to facilitate comparison of results. Where 

 it seemed desirable the tables contain values both in metric and in British units, but 

 as a rule the centimeter, gram, and second have been used as fundamental units. 

 In the comparison of British aud metric units, and quantities expressed in them, the 

 meter has been taken as equal to 39.37 inches, which is the legal ratio in the United^ 

 States. It is hardly possible that a series of tables, such as those here given, involv-' 



'Method III. — Absorbent material, freshly ignited, fineljr granular pumice stone. 

 Absorption ajjparatus, a small glass tube, 20 cm. in length, consisting of a narrow 

 portion 4 cm. long and 3 mm. in its internal diameter, and an expanded portion 16 

 cm. long and 12 mm. in its internal diameter, similar in size and form to the absorp- 

 tion tube used by Remsen and by Abbott in their experiments. 



