291 



resistance will be frequently encountered from above, 

 because of the uneven deposition of the Eern tissue. 



Bulbils, it will be noticed, are sometimes produced mi the 

 upper surface as in \V . Orientals and sometimes on the 

 under surface as in C. bulbifera, and on the radiis as in 

 Polystichum. We may reasonably infer that it is entirely 

 a matter el' convenience as to where the sori are produced, 

 and that a reasonable amount of latitude is allowed to 

 ferns in general and that they are at liberty to make use 

 oi either the front or the back door according to circum- 

 stances. 



Most of the marginate, muricate and fleshy Scols are 

 bulbiferous under close moist culture, and these bulbils 

 almost invariably occur on the upper surface, possibly 

 because of the greater amount of light on the upper as 

 compared with the under surface. There is also some subtle 

 connection between sori and bulbils, as is seen from the 

 number of instances in which immature sori can by abnormal 

 atmospheric conditions be transformed into bulbils, e.g., A.f i . 

 plu mo sum elegans, p. Druery, p. clivaricatum, Cyst, sem- 

 pervirens and several others. Scolopenclrium bulbils in 

 their earliest stages occur in small brownish white patches, 

 and look uncommonly like immature sori, as are best seen 

 on abortive fronds of A.f i. plumosum elegans, when what 

 would eventually be developed into hundreds of spores seem 

 by some inexplicable process of fusion to develop into one 

 bulbil. It is not improbable that when in this mobile half- 

 and-half condition, they might if taken in hand just at the 

 psychological moment, be persuaded to alter the programme 

 and develop into sori instead of bulbils. Just as in the 

 instances above cited the reverse operation is known to 

 take place. 



If then, Scolopendriums find it most convenient to emit 



