336 IOWA ACADEMY OF SCIENCE 



to the descriptions. Mr. Hillman also is the author of Farmers' 

 Bulletin 428, 1911, U. S. D. A., which gives valuable lists of 

 adulterants of commercial seeds. He also has published num- 

 erous other bulletins but the author considers the above men- 

 tioned the most comprehensive. ''The Seeds of the Blue 

 Grasses," Bulletin 84 of Bureau of Plant Industry, U. S. D. A., 

 1905, by Edgar Brown and P. H. Hillman, includes a key to 

 the seeds of six species of Boa as found in commercial seeds. 

 Aside from this last mentioned paper, no literature has been 

 found which concerns itself with a key which will aid in the 

 systematic determination of different seeds. 



METHODS OF STUDY. 



The specimens studied were collected for the most part dur- 

 ing the summer and fall of 1911. These were carefully labelled 

 with the common and scientific names. An attempt was then 

 made to describe each seed carefully, using the external char- 

 acters which were most evident with the use of an ordinary 

 hand-lens. From these descriptions and with the use of the 

 specimens constantly for reference, the key was constructed. 

 Drawings were used in preference to photographs because it 

 was believed that by them the distinguishing characters could 

 be brought out to a better advantage. The seeds of the grasses 

 were described as they appear with the glumes or scales re- 

 moved, at least where these are easily removed by rubbing 

 between the thumb and fore-finger. Seeds vary considerably 

 and where there was any doubt as to which one of two descrip- 

 tions would serve, the seed was entered under both headings. 

 In this way the opportunities of misinterpretation were mini- 

 mized. A given seed may be entered many times, due to its 

 variations. This might appear to be a bad condition of affairs 

 and lead one to believe that better distinguishing characters 

 might have been chosen. A trial of the key in actual testing, 

 however, should prove to the individual that the arrangement 

 into distinct groups according to size facilitates the quick 

 determination of a seed. It should be noted that even in the 

 most complex instance but twenty steps are required to make 

 tlic determination. 



