MISCELLANEOUS PAPERS. 397 



of prairie fire extending north and south along the horizon. He soon 

 appeared fully above the horizon and looked like an enormous elon- 

 gated red football lying on its side with the ends north and soutli. 



Three minutes of time had passed. Eight minutes of time is 

 equivalent to 2^ of longitude, amounting in the latitude of Great 

 Bend to 109 miles. Hence the sun was seen at a time when it must 

 have been fully 2° below the horizon and ought to be just rising in 

 eastern Marion county, or even Chase or Butler. Presently the rift 

 of sky and the upper horizon disappeared altogether, sun and all, and 

 the true sun began to rise a second time, still about one and one-half 

 minutes before his stated time ; but this time only slightly flattened. 



Looking to the north over Cheyenne Bottom the entire plain was 

 alive with a wavy, tremulous motion, like hot air rising from a furnace. 

 Atlanta and Larned had now disappeared. The temperature was 34" 

 and the show was over. 



Passing down to the surface of the upper stratum of cold air, the 

 mirrored appearance was still faintly to be seen ; but it was being 

 broken up by a slight breeze flowing from the northwest. At the 

 station at 7:20 the temperature was 22^, a rise of 7" in about half an 

 hour. Passing on down into the valley the temperature was found to 

 be 15°, a rise of 9 '^. Continuing into the creek bottom the mercury 

 dropped to 10^, a rise of 8^ from the first observation. This showed 

 that the breezes stirring on the hill had not yet reached the bottom. 



Mirage of this character is usually most noticeable, not when the 

 temperature is so low as in the case noted, but when the temperature 

 is close to the freezing-point. Then the cold frosty air flows into the 

 valley, leaving the warm air on the high lands. That is why frost ap- 

 pears first in valleys and low grounds, and why there is less frost in 

 spring and fall on a hilltop. Perhaps that was true in the case noted, 

 as shown by the temperature taken from the stone heap on the sum- 

 mit. 



One of the most notable cases of refracting mirage that has come 

 under my observation was one that occurred on the night of November 

 2, 1899, at Topeka. This was not known as a mirage. It would have 

 passed entirely unnoticed had not a preconcerted arrangement been 

 made over a wide extent of territory in order to determine the visi- 

 bility of certain phenomena. 



There were to be some fireworks burned in Topeka in honor of the 

 return of the Twentieth Kansas from the Philippines. At my solici- 

 tation, the Topeka Capital offered a series of prizes to the ladies of 

 the country surrounding Topeka who should be fortunate enough to 

 see the fireworks at the greatest distance. Therefore, thousands of 



