122 VARIATIONS IX NUMBER OF SPORAXGIA 



the products have been rounded off: they will be best distinguished in 

 the early stages of development. A good, clear example is found in the 

 genus Gleichenia : the sorus of most species of the genus is of the radiate, 

 uniseriate type, in which the sporangia form a series round the receptacle, 

 while the centre of it is unoccupied by sporangia : there is good comparative 

 reason to believe that this was the primitive type of sorus for the genus. 

 But in G. dichotoma additional sporangia, one or more, are often seated 

 in the central unoccupied space (Fig. 63). Examination of early stages 

 of development shows that these sporangia are not a result of fission of 

 sporangia typically present, but develop from new primordia not represented 

 in other species (Fig. 64). The interpolation, of which this is a simple 

 case, becomes a more prominent feature in the Ferns which palaeontology 



FIG. 63. 



u-h, sori of Gleichenia dichotoma., Willd. Showing one or more sporangia in the 

 centre of the sorus, usually in this genus vacant, i-k, sporangia of Gl. circinata, Sed. 

 /-, sporangia of G7. dichotoma. a-h x about 14. 1-11X50. 



teaches us to regard as the more recent. The succession of the sporangia 

 characteristic of those Ferns which will be styled on that account the 

 ( iradatae, may be recognised as a repeated basipetal interpolation of 

 sporangia on the receptacle of the sorus (Pig. 65 A). In the Mixtae it is 

 still more prominent in the sorus, for new primordia, formed in continued 

 succession, appear scattered between those which are more advanced 

 ( l-'ig. 66). But interpolation may take place not only of sporangia, but 

 also of new spore-bearing organs between those typically present. This 

 is exemplified in the interpolated stamens in certain flowers, the number 

 of which is closely related to their bulk, and to the space available for 

 them on the floral receptacle. Such interpolation of sporangium-bearing 

 parts will clearly result in an increase in number of sporangia. The 

 prevalence of interpolation thus seen in the Leptosporangiate Ferns, has 

 given an entirely fictitious importance to this phenomenon in the general 



