638 



SCIENCE 



[N. S. Vol. XLIV. No. 1140 



pages 28 to 33, the conclusion has been 

 reached that here, also, the gumbo is so related 

 to the drift that it is undoubtedly the thor- 

 oughly weathered product of the Illinoian drift. 

 As a result of the field investigations and 

 the chemical studies it is now proposed that 

 the somewhat indefinite term " gumbo " be no 

 longer used for these super-drift clays, but 

 that the name "gumbotil" be used. Gum- 

 botil is, therefore, a gray to dark-colored, 

 thoroughly leached, non-laminated, deoxidized 

 clay, very sticky and breaking with a starch- 

 like fracture when wet, very hard and tena- 

 cious when dry, and which is, chiefly, the result 

 of weathering of drift. The name is intended 

 to suggest the nature of the material and its 

 origin, and it is thought best to use a simple 

 rather than a compound word. Field work 

 has already established the fact that in Iowa 

 there are three gumbotils, the Nebraskan 

 gumbotil, the Kansan gumbotil and the Illi- 

 noian gumbotil. 



Geo. F. Kay 



University or Iowa 



THE EVOLUTION OF HERBS 



The article by Edmund W. Sinnott, pub- 

 lished last week in Science, U ■ 291, supports 

 conclusions on this subject arrived at from 

 quite a different standpoint. 



The idea that trees are primitive forms is 

 involved in the proposition advanced by Henry 

 L. Clarke, in the American Naturalist, 27: 

 769-81, September, 1893, that in their order 

 of blooming the generalized precede the spe- 

 cialized. 



My observations were based only on ento- 

 mophilous flowers, 493 native and 61 intro- 

 duced. 



If we assume that the earliest, least special- 

 ized, and primitive plants form the earliest 

 maxima and succeed in regular order, we shall 

 have for indigenous plants the following re- 

 sults according to the time of the maxima : 



Trees April 27-May 8 



Woody climbers June 13-15 



Shrubs June 21-23 



Perennial herbs August 2-6 



Annuals and biennials . . . August 30-September 6 



And this seems to be the probable order of 

 their development. The original plants having 

 the most freedom developed large size and 

 occupied the most favorable positions. The 

 less favored could become reduced to shrubs 

 and finally to herbaceous perennials, and 

 occupy many positions which were unfavorable 

 for trees or with which trees did not interfere. 

 The habits of perennial herbs are better under- 

 stood if we suppose that they had to compete 

 with trees, or rather avoid competition with 

 them, from the first. The annuals developing 

 later were able to find many temporary situa- 

 tions unfavorable for woody plants or peren- 

 nial herbs. The primitive Angiosperms were 

 probably trees, like Magnoliaceas, Anonacese 

 and Lauracese. 



Another general characteristic of blooming 

 seasons is that the earliest, most generalized, 

 most primitive plants have the shortest sea- 

 sons, while the most specialized, most recent, 

 and latest arrivals have the longest seasons. 

 Arranging the vegetative forms according to 

 their average blooming seasons, we have the 

 following order : 



Days 



Woody climbers 36.5 



Trees 39.4 



Shrubs 42.7 



Perennial herbs 57.1 



Annuals or biennials 75.1 



Cosmopolitan 80.4 



Introduced 117.3 



Except for trees and woody climbers, the 

 order is the same as for the maxima. 



Charles Eobertson 

 Carlinville, III., 

 September 6, 1916 



HORSE FLESH AND THE DIET OF EARLY MAN 



To the Editor of Science: In Science, for 

 September 22, is published a letter on the 

 "Animal Diet of Early Man," which dis- 

 cusses the subject with reference to possible 

 evidence drawn from tapeworms and their 

 hosts. In this connection, the writer of the 

 letter speaks of the horse as food, as follows: 



There is nothing to show that horses were not 

 eaten, unless the rather widespread abhorrence of 

 eating horse flesh at the present time can be con- 



