XIII] 



THE PERISPORE 



261 



in it as a safe criterion is shaken by the fact that while it is present in 

 Blechmmi and Woodzvardia it is absent in the closely related Sadleria, Brainea, 

 and Doodya. It is clearly a feature 

 adopted late in Descent, and 

 apparently restricted to certain 

 circles of affinity. 



In Polypodiiini iiiibricatnin 

 the perispore takes the form of 

 fine hygroscopic fibres which lie 

 among the spores in the spor- 

 angia. They may be compared 

 to the elaters of Equisetiun, both 

 being specialised products of the 

 tapetum. It is suggested that 

 their biological use in this epi- 

 phytic Fern lies in fixing the 

 large spores upon the surface 

 of the tree-stems (G. Karsten, 

 Flora, Vol. 79, 1894, p. 86). The 

 perispore finds its most striking 

 development in the Salviniaceae. 

 In Azolla the limitation of the 

 "massulae" in the microsporan- 

 gium, and the formation of the " glochidia," and also of the swimming organ 

 in the megasporangium, are highly specialised types of perispore (Hannig, 

 Flora, Vol. 102, 1911, p. 243 ; also Vol. 103, p. 321). 



Spore-Output 

 By the definition of the sporangium given above the spore-mother-cells 

 form its essential feature. The structural characteristics of the sporangial 

 stalk and wall are developed in accordance with their number : and since 

 each divides to form four spores, all of which grow to maturity in normal 

 sporangia, the numerical output of spores per sporangium will give an ap- 

 proximate basis for their comparison. This was first suggested by Russow 

 ( Vergl. Unters. p. 86). An estimate of the spores of a single sporangium may 

 be made either by actual counting, which may be easy for small sporangia 

 where the numbers are relatively low : but this is a tedious method where 

 the numbers are large. In the latter case an approximate estimate may be 

 formed by the examination of transverse and longitudinal sections in the 

 stage of the spore-mother-cell. Subject to certain exceptions and limitations 

 the numbers thus obtained are approximately constant for the individual or 

 species, and accordingly they give a consistent basis for comparison of Ferns 



V\g. 257. Spores oi Aspidium trifoUatum, after Hannig. 

 a, /' = ripe spores with prickly exospore [ex.) and 

 transparent perispore {psp.), appearing like a loose 

 sac. ( X 500.) ^=part of the exospore and perispore 

 more highly magnified. 



