ON THE DEVELOPMENT AND STRUCTURE OF MOSSES. 11 



jectionable, I should prefer that of ' Drpphyton/ which simply 

 indicates that it is the forerunner of the true plant. 



b. After a time, whatever privileged portion of the threads 

 may give rise to a bud, fibrous rootlets strike downwards from 

 the base, and the bud itself is gradually elongated upwards, 

 according to the character of the plant, into the true axis, 

 clothed with its proper foliage, and in time giving rise to the 

 true fruit, whether male or female, on the same or different 

 plants. These threads are sometimes persistent, as in Ephe- 

 merum serratum and Ephemerella recurvifolia, but in general 

 they vanish almost entirely long before the plant has arrived 

 at maturity. 



In Sphagnum the course is not precisely the same. The 

 first result of germination, instead of a thread as in the more 

 typical Mosses, is a scale-like expansion (Plate 1, fig. 1) re- 

 sembling closely the young state of a Jungermannia, and pro- 

 ducing buds from the notches of the margin. 



The prothalloid stage of Mosses must riot be confounded 

 with a growth of a very similar appearance, which takes place 

 from the rootlets of some Mosses, as for example, in Pogona- 

 tum aloides (Plate 19, fig. 2), which is produced after the 

 death of the old plant, and forms a green velvety mass, which 

 at first sight cannot be distinguished from a true prothallus. 

 Conferva velutina of authors owes its origin to such a growth, 

 and another supposed Conferva is due to a similar development 

 in Schistostega osmundacea (Plate 14, fig. 4), the necklace- 

 like ultimate joints of which refract light so strongly that it 

 has been supposed to be phosphorescent. 



Rootlets, it has just been observed, are produced at the base 

 of the fertile buds, and this at a very early stage of their 

 development. They are for the most part more slender than 

 the primary threads, more or less distinctly but obliquely ar- 



