BOTANICAL SUBJECTS. 221 



attachment of the indusium is on the side. In Asplenium 

 lanceolatum^ Huds., and others, the attachment is also at the side, 

 though elongated, as if it were a confluence of smaller side bracts. 

 All the bract-like indusia emerge from the veins, as ordinary leaves 

 do from stems, and in Lygodium, each separate sporangium, or 

 spore-case, has a separate subtending bract. Some indusia are 

 more like wings of stems than bracts. All indusia appear pro- 

 ducible by a confluence of hairs. The radiating hairs of Eleagnus 

 might represent the origin of the peltate indusium or Aspidium. 



Again, in Scolopendrium vulgare, Sm., we have two adjacent 

 veins giving off two elongated indusia, like the wings of stems, 

 the free edges of which face each other, and give cover to two 

 rows of sporangia, the whole being indistinguishable from a one- 

 celled elongated anther. 



The fertile pinnee of Cryptogramme crispa, Br., and Blechnum 

 boreale, Sw., as already stated, also lead the mind in the same 

 direction, viz., that the morphological origin of the anther, as we 

 find it in phaenogams, is probably to be found in the sori of ferns, 

 with their single or double indusia (adjacent sori becoming 

 confluent or fused). 



In other words, the sorus appears to be a transition form 

 between the reproductive leaf of Porphyra laciniata, and the 

 anthers of phaenogams. I consider the leaf of this alga as the 

 rudiment of the specialized stamen of the higher plants. It is a 

 simple frond, the vegetative cells of which produce pollen or 

 ovules, or both. 



Curiously enough, the pollen of phcenogams follows the 

 quaternary division of the pollen of Alg^e. Thuret says of 

 Cutleria mtdtifida, p. 22 : " II est curieux de retrouver ici une 

 exemple de cette division quaternaire, si frequente dans les organes 

 reproducteurs des cryptogames." He shows the same quaternary 

 division in Porphyra laciniata. Is this resemblance between 

 tetraspores and pollen grains accidental or has it a deeper 

 ineaning ? 



In the anther of Liriodendron and Asarum Canadense we 

 appear to have a repetition, in a much modified form, of a double 

 indusiate sorus, and in this case the anthers dehisce dorsally, like 

 the two indusia of Pteris aqiiilina and others. 



