Fossil Ice Crystals 5 



other; and on other rocks they would be found crossing in 

 close networks. In the summer of 1904 I again found my 

 ice markings on a layer of arenaceous limestone in the same 

 formation in the Big Bend country in Texas. This time I 

 collected some specimens which were subsequently photo- 

 graphed. One of these photographs is shown in Plate II. 

 Again in 1906 I noticed the same markings on some thin 

 sandy flags which occur in the Del Rio clay near the city of 

 Del Rio. In this case the needle-like crystals were somewhat 

 more slender than those previously seen, and some were 

 slightly curved and somewhat more elongated. These of 

 course interested me as showing the occurrence of freezing 

 temperatures no doubt at a somewhat earlier time than that 

 pertaining to my previous observations. 



During all these years my residence was in Illinois, and 

 I was naturally watching for similar markings in recent 

 mud, resulting from late and early frosts. I found them in 

 the fall of 1909. At this time some excavations were being 

 made in the loess in Rock Island, when some rains fell in 

 the late fall. These rains evidently happened to give the 

 mud the amount of moisture necessary for such crystals to 

 develop, as the ground froze. The rains had washed the 

 loess extensively, and I found a number of places where it 

 lay redistributed, with a fairly smooth surface. It was 

 evident that the moisture content of the ground, together 

 with the temperature conditions, determined the size 

 and the closeness of the frozen patterns. In places 

 the crystals were long and slender, in others they 

 were short and stout. At some points they were straight 

 and in others slightly curved. Here and there the patterns 

 were close enough to resemble the fine lines which we some- 

 times notice in the hoarfrost on windowpanes. In other 

 places the crystals occurred in radiating groups, and else- 

 where they would form scattered separate units. For 

 preserving a record of what I saw, I poured plaster over 

 several patterns and had these casts photographed, as ap- 

 pears in Plates VIII. IX, X. Placing these side by side with 



