44 Kansas Academy of science. 



ing their specific heat to be on the average that of CO^, for about 12° rise in tem- 

 perature 0.9 per cent, must be lost. 



4. The heat of the decomposition of the KCIO3 ^^^ KNO3. This can be only 

 approximately determined for the apparatus. 



(a.) Two grams of pure carbon under combustion in the calorimeter gave 11,643- 

 gram-degrees thermal effect. It is known that the perfect combustion of this amount 

 of pure carbon should have produced 1(),1()0 gram-degrees. The total loss is there- 

 fore 28 per cent. Of this, considering 3.3 per cent, as due to causes already men- 

 tioned in 2 and 3, 24.7 per cent, must be required for the decomposition of the 

 oxygen-producing mixture. 



[b.) In order to determine this amount in another way, if possible, the heat re- 

 quired for the decomposition of the 22 grams oxygen mixture was calculated in 

 accordance with the methods and tables given by Pattison Muir.* This amount i& 

 as follows: 



16.5 grams KCIO^ in decomposition gives out 1312 gram-degrees, 



5.5 " KNO3 " " absorbs 0144 " 



Amount absorbed for 22 grams, oxygen mixture 4832 ' ' 



" '' by apparatus, by radiation, etc., (33 per cent) 533 '' 



Total auiouni absorbed 5365 " 



2 grams pure carbon should give 16,160 ' ' 



The loss, therefore, (by calculation,) is 33 per cent., which does not'differ greatly 

 from that obtained by (a), or 28 per cent. 



(c.) No attempt to estimate the heating powers from the chemical analysis of the 

 coal is considered valuable. In this connection, M. L. Grunow, in the Engineering 

 and Mining Journal, vol. XVIII, states: "It is evident that the manner of combina- 

 tion of the elements in coals is too variable to make it possible to determine their 

 true heating power from simple elementary analysis." 



5. By error in the amount of coal taken. All the coal was weighed in balances 

 reading to 0.5 milligrams. A possible error of 1 milligram in the weighing would 

 affect the result by only 0.05 of 1 per cent., a negligible quantity. 



6. By error in amount of water taken. An error of 3 grams was a possible limit 

 with the balances employed, and this would introduce an error of 0.15 of 1 per cent., 

 a quantity also negligible. 



From the above results it is evident that the correction of 10 per cent., recom- 

 mended by the manufacturers of the apparatus, and hitherto adopted by experi- 

 menters, is far too small. The total correction is probably not far from 30 per cent. 



In the following tables, corrections of both 10 and 30 per cent, are given, the 

 former to render the results readily comparable with those of other experimenters 

 with the same apparatus. 



DE.-CKIPTION OF THE COAL BEDS, BEGINNING WITH THE LOWEST. t 



A. The Leavenworth vein, at Lansing and Leavenworth: 24 inch vein, about 720 



feet below surface. 



B. The Cherokee, in Crawford and Cherokee counties, at Pittsburg, Weir City and 



vicinity: 42-inch vein; from outcropping to 70 feet deep. 



C. Upper vein in Crawford county, now worked as strippings and slopes; 18-incb 



vein. 



D. The Fort Scott coals: chiefly slojjes and strippings; 28 inches — 14 inches. 



E. The Linn county bed, at Pleasanton, La Cygne, <fec.: 14-inch vein; 60 to 100 feet 



deep. 



F. The Osage county veins, at Osage City, Carbondale, Burlingame and vicinity: 



Vein 18 to 20 inches; 35 to 85 feet deep. 



G. Franklin county coal. 

 H. Cloud county coal. 



* The Elements of Thermal Chemistry: Muir. 1885. Appendix I. 



tSee also "Composition of Kansas Coals," by Prof. E. H. S. Bailey, in this volume, p. 46. 



