.] ARREST OF AXIAL ROOT 97 



hereditary feature, it is now characteristic of all 

 monocotyledons, whether they be terrestrial or 

 aquatic in habit. 



My experiments with germinating seeds of the 

 water crowfoot (fig. 21) showed how the axial 

 root characteristic of terrestrial dicotyledons is 

 rarely (*) produced, but adventitious endogenous 

 roots are given off instead. An experiment with 

 mustard grown in water showed the same thing 

 (fig. 22). The apex became arrested while 

 secondary roots appeared at its base. Rather 

 long axial roots are produced in palms and Indian 

 corn when germinating, pointing to an ancestral 

 dicotyledonous habit, but they soon perish, 

 being followed by adventitious roots in ascend- 

 ing series. 



With regard to many minute anatomical 

 structures, I must refer the reader to the 

 details given in my paper, only mentioning that 

 a large number of additional facts can now 

 be added not recorded at the time ; such as 

 the origin of the root -cap, that it is early 

 differentiated and grows separately from the 

 initial cells of the root in monocotyledons and 

 the Nymphceacece, but it is not so in gymno- 

 sperms and dicotyledons generally. 



I have already alluded to aeriferous canals and 

 aerenchyma, as aeriferous tissue is called, as 

 characteristic of all aquatic plants, and as origin- 

 ally acquired by growing in water; but they 



N 



