Ift2 



to heat and moisture changes. A. series of pots containing seedlings was subjected 

 to an artificial drought of five days. Of the eleven species of composites all except 

 Ambrona died. Of ten species of other families, only Scrophularia nodosa Marilandica 

 died. Eepetitions of the experiment showed similar results. Another line of ex- 

 periments showed that the composite seedlings were unable to withstand any con- 

 siderable change in temperature, being much more affected by temperature in- 

 crease than by its decrease. An increase of 5° C, from 25° C. to 30° C , usually 

 proving sufficient to kill them or greatly retard their growth. When it is remem- 

 bered that the distribution of composites is for the most part in dry soil, in places 

 exposed to the full force of the sun, it is apparent that large numbers of seedlings 

 must perish. It is possible that the danger of a spread of these forms through 

 seed dissemination has been overestimated. 



Another fact indicated by the experiments was that the achenes of the earlier 

 and later flowers were rarely viable, this being especially true in Helicnithus. 



It is somewhat surprising that in a family so dominating in species and indi- 

 viduals there is not included a greater number of "worst weeds." Considering 

 the immense size of the family, the number is astonishingly small. 



Taraxacum invades the lawns; the Lactucas, Gnicus, Arctium, and Erigeron the 

 fields ; but none of them compare in noxious features with forms from other fam- 

 ilies. Ambrosia, which overruns waste fields, I find is considered by the farmers 

 as a positive benefit to the land. Erigeron, which a few years ago was a great 

 annoyance, seems to have yielded to cultivation, and to have practically lost its 

 place among bad weeds. Doubtless in some places it is still annoying, but the 

 evidence is that it disappears from carefully cultivated fields. Chrysanthemum 

 Leucanthemum L. is certainly bad, but is of restricted range. Bidens is annoying 

 to the sheep-raiser, but does not otherwise rise to the rank of a "bad" weed. For 

 the most part the composita? seem perfectly content to occupy waste places, and 

 readily yield to man the possession of the soil. 



So far as I have been able to discover, none of the species are poisonous, if I 

 except a few reported instances of poisoning by forms of Cnieus. Most of these 

 cases, I think, can be referred to personal idiosyncrasy. I have tested all the 

 forms of Cnicus upon myself and upon numbers of students without results other 

 than were referable to the mechanical action of the prickles. Xaiithium Cana- 

 dense Mill, is said to be poisonous to the touch. ^ If this be true, the forms found 

 in the State, X. spmosum L. and X. strumarium L., are to be regarded with sus- 

 picion by persons susceptible to plant poisoning. It is to be remembered, how- 

 ever, that even the known poisonous plants are only poisonous to a small percent- 



^Vfhite, Dermatitis vtnr.naia. Boston, 1887. 



