.161 



With but few exceptions, the composites within our bounds do not come into 

 full flower until July and August. As a rule the tlowering season is long, many 

 genera bhissoming alnindantly from July until checked by the frosts. From the 

 middle of August they determine the [diysiognomy of the vegetation over the 

 entire area of the State. This is especially true in the prairie region and in open 

 fields. Indeed, the great majority <»f the composites of Indiana are found in their 

 greatest abundance and lu.xuriauce in dry soils and in regions exposed to the full 

 action of the sun. Thej' seem to be xerophytes of the xerophytes. 



Some species of Eupatorium, Liatris spicata and other forms, however, furnish 

 exceptions as regards dryness of soil, while Polymnw and a few others give excep- 

 tion as to light. Certainly in no other family in our area can xerophytic adap- 

 tations be so satisfactorily studied. 



While the flowering season is so extended, and the consequent number of 

 achenes formed enormous, it is probable that but a small proportion of them 

 germinate. Theseedlings, also, in all cases in which experiments were tried, were 

 remarkably sensitive to changes in temperature and moisture. Almost every 

 other form used was more hardy in the seedling stage than the com{)ositae. Ex- 

 ceptions to this were the Ambroaias and Laetiica Canadensis L. In the series of 

 experiments the percentage of seeds germinating was vei\v small in the compositie, 

 with the exception of Arctium, where the per cents, in three experiments were, 87. o, 

 80, and 87.5. In Bidens 20 per cent, was the highest, in Laduca 25 per cent., in ^»?i- 

 brosia 20 per cent., -while in Cnicng out of three plantings of 30 achenes each, 

 only two achenes germinated. Under the same conditions Abiitilon Avicennn' 

 Gaertn, in two experiments gave 100 per cent., and in a third, 96.7. The seedlings 

 of this plant were extremely hardy, withstanding wide ranges of temperature and 

 moisture. Solanum nigrum L , Datura stramonium L., and Serophularia nodosa L., 

 Marilandica Gray, invariably showed germination per cents, above eighty-five. 

 The plants, other than composites, are introduced simply for purposes of com- 

 parison. The data given above are derived from a large number of germination 

 experiments conducted in the Laboratories of Purdue university. In these ex- 

 periments I have endeavored to eliminate possible error, and to give, so far as 

 could be determined, natural condition-!. The experiments cover some 30 com- 

 posite species distributed among 15 genera, and 50 species representing families 

 other than the compositiv. The material was gathered in almost every instance 

 with extreme care in order that conclusions might be based upon known condi- 

 tions. So far as the experiments go concerning composita^, I am convinced that 

 the distribution of this family is largely limited, first, by the small germination 

 percentage of the achenes; second, by the extreme sensitiveness of the seedlings 



