No. 494] 



PARASITIC PLANTS 



103 



those groups is of normal structure. Every member of the 

 four remaining groups, however, begins life in an embryo 

 which is simple and filiform and without cotyledons, rad- 

 icle and plumule, although the flower in which it is pro- 

 duced is of normal phenogamous structure. ^Ioivovcr, 

 although the simple abnormal embryo is physically iden- 

 tical for each of the four groups just mentioned, the 

 resulting forms of parasitism are too widely different for 

 each group to suggest for them even a remote community 

 of origin. 



A remarkable fact concerning group IV is that the two 

 genera which compose it, Cuscuta and Cassytha, belong 

 to widely different families, namely, Convolvulaeese and 

 Lauracese, respectively, and that the respective genera 

 prevail in distantly separated parts of the world. Both 

 genera are endowed with a single parasitic impress which 

 distinguishes and dominates them equally in both habi- 

 tude and somatic structure. That impress also separates 

 all the species and individual plants of the whole group 

 from normal plants and from all other parasites. The 

 habits of this group, as shown by our well-known dodders, 

 are widely different from those of all the other parasitic 

 groups. They are all annual plants and consequently 

 the whole life history of each species is crowded into a 

 single season, which is shortened by late spring germi- 

 nation and early frosts. Therefore all the characteristics 

 of the whole group lie dormant in the simple filiform 

 embryo of every dodder seed for more than half of each 

 year ; and yet every one of those characteristics is invari- 

 ably heritable and constant. Difficult as it is to under- 

 stand how every individual member of such a distinctly 

 defined double group of annual plants could have assumed 

 their abnormal characteristics either slowly or suddenly, 

 and attained a world-wide distribution, it is still more 

 difficult to understand how two such diverse genera could 

 have assumed identical parasitic characters. It is almost 

 superfluous to add that the habits and structure of no 

 known plant offers any suggestion of a gradual origina- 



