XII] POSITION OF THE SORUS 217 



in near relation to it, as in Pteridiicm or the Hymenophyllaceae. The two 

 positions, the superficial and the marginal, graduate into one another by- 

 many intermediate steps. But the position held by the sori of a species, 

 genus, or even of still larger groups, is as a rule definite, showing that it is 

 not a readily fluctuating character, but is so far constant in the individual 

 species or genus that it may be depended on for comparative purposes. 



So marked is the constancy of position of the sorus in genera and species 

 that the few exceptions which occur become notable. The most familiar is 

 that of a variety of PolysticJmm aciileatum from the uplands of Ceylon, 

 described as Aspidmrn {P.) aiiomalum, Hk. and Arn. (Hooker's Sp. Fil. 

 Vol. iv, p. 27). Here normally constructed sori appear on the upper instead 

 of the lower surface of the leaf. Since no intermediate stages appear, either 

 in the individual or in related species or varieties, the natural conclusion is 

 that there has been a transfer bodily of the factors of normal initiation of sori 

 into an abnormal position. Such an instance as this, being anomalous, should 

 not be held to affect the general argument on the position of the sori. 



The question will naturally arise whether the marginal or the superficial 

 position of the sorus was the prior state. A general probability that the 

 superficial is relatively late and derivative arises from the fact that it is 

 prevalent in modern Ferns. It is physiologically intelligible that parts so 

 important shall not be unduly exposed : moreover the downward direction 

 would favour the dissemination of the spores. But more direct evidence is 

 afforded by reference to those Ferns which by comparison and by fossil 

 history are held as primitive, viz. the Simplices. This will give a general 

 basis of fact upon which a reasonable opinion may rest. Further comparison 

 may be made of the position of the sorus in genera, species, and varieties 

 related to one another. And in particular of its position in the successive 

 ontogenetic stages of the individual. From such sources it may be shown 

 that a transition from the marginal to the superficial position has occurred 

 repeatedly in the evolution of Ferns, and that it may be traced even in the 

 ontogeny of some of them. 



Among the Simplices, the Botryopterideae, Zygopterideae, Ophioglossa- 

 ceae, Schizaeaceae, and Osnmnda show a distal or marginal position of the 

 sporangia. The Marattiaceae, Gleicheniaceae, Matonineae, and the genus 

 Todea bear their sporangia superficially. Osniunda itself bears the sporangia 

 as a rule in tassels at a margin of the attenuated fertile region of the leaf. 

 But von Goebel points to those sporophylls that are half modified into tro- 

 phophylls : here the insertion of the sporangia is superficial, following the 

 veins, in a manner similar to those of normal specimens of Todea{Y'\^.2io,io-,, 

 p. II 39). This supports his general statement that the more the sporophyll is 

 developed as a dorsiventral foliage leaf the more are the sporangia restricted 

 to the lower surface. The facts for the Simplices are in general accord with 



