144 NINTH REPORT. 



the volume of a stream, without increasing its load, as by stream capture 

 may occasion the development of terraces by allowing the stream to deepen 

 its channel. Barriers may cause flood plains, their removal will cause the 

 stream to cut more or less deeply into the plain above, leaving terraces. 

 The recession of a falls through a floodplain convert such parts of it as re- 

 main, into terraces.' Prof. I. C. Russell has described terrace formation 

 lateral to the Menominee river hy cutting down through l-)road alluvial 

 plains, the highest terrace being the plain itself, the lower terraces were 

 eroded in subsequent flood plains. Again he describes terrace formations 

 adjacent to Green Bay which are due to subsequent lowering of the lake 

 and down cutting of the stream channel.- ' , 



TEMPERATURE. 



Finally I would like to be permitted to digress a minute and call atten- 

 tion to the very favorable opportunity of studying the effect of water tem- 

 peratures oh the insular climate of Lower Michigan. The prevailing wind 

 in Lower Michigan is from the southwest, so that members of the Academy 

 living on the Lake Michigan shore can have an opportunity of determining 

 the results and relationship of the exact effect of water temperature on that 

 of the land. An average of the isothermal lines shows the same temperature 

 averages extending 46 miles farther northward on the west side of the Lower 

 Peninsula as compared to the east side. Comparing the results on the east 

 and west sides of Lake Michigan we have an average of 36 miles farther 

 north than on the east coast of Wisconsin in crossing the lake. During the 

 latter part of August the average water temperature is about 4° greater 

 than that of the air. Also during this time the ratio of change of water 

 temperature relative to that of the air temperature is less than 34.4%. In 

 this factor taking into account the prevailing southwesterly winds, we find 

 a partial explanation of the insular climate of Lower Michigan. Moreover 

 the temperature of the water, as a rule, being greater than that of the air 

 from about 7 p. m. until about 9 a. m. the following day, the tendency would 

 be to increase the temperature of the adjacent windward shores. On the 

 other hand, the air temperature being greater during the remainder of the 

 day, the water would tend to establish an equilibrium by reducing the air 

 temperature the mean range probably approaching the mean of the aver- 

 age ranges of air and water temperatures. The more prolonged period of 

 average higher water temperature is doubtless the greater factor in this 

 ciuestion.^ 



Observations of this character to be of general value should be taken at 

 least once a week throughout the year and should be continuous for 24 hours, 

 readings being taken every hour beginning at say 6 p. m. Readings of the 

 air temperature on the land and of the surface water temperature at the end 

 of a pier where there is 15-20 feet of water should be taken as nearly together 

 as possiljle. As to the proper way in which this work can be carried on in- 

 formation from C. F. Schneider, head of the State Weather Bureau, at Grand 

 Rapids could be obtained, and perhaps the use of an accurate thermometer 

 suitable for the purpose. The results would be of very considerable scien- 

 tific value. 



Lansing, Michigan, March 26, 1907. 



1 Chamberlin & Salisbury, Geology, Vol. I, p. 193-198. 



2 I. C. Russell, Ann. Rep. Mich. Geol. Surv., 1906, pp. 77-78. 



3 Michigan Academy of Science, 7th annual report, p. 40-43, and Monthly Weather 

 Review, Washington, D. C, Dec. 1905. 



