654 



SCIENCE. 



[N. S. Vol. XVII. No. 434. 



veloped entirely within the tissue of the 

 sporophyte, in the most intimate anatom- 

 ical and, physiological contact and depend- 

 ence upon the latter, and is quite incapable 

 of developing the sex cells without the 

 direct cooperation of the sporophyte. It 

 is plain that a part at least of the work 

 of nourishing and preparing the sex cells 

 for their functions, assumed by the pro- 

 thallus in the fern, has in the phanerogam 

 been transferred from the rudimentary 

 prothallium to the highly developed sporo- 

 phyte. The morphological line between 

 gametophyte and sporophyte can still be 

 traced (though only through recondite 

 comparative researches), but the physio- 

 logical, and to a great extent the struc- 

 tural, line between the two has vanished. 

 The gametophyte, therefore, does not con- 

 stitute a 'generation' in the sense in which 

 the word was originally used in the ferns, 

 for the physiological equivalent of the 

 sexual generation of the ferns is, in the 

 phanerogam, the gametophyte plus part of 

 the sporophyte.* Not only are the tissues 

 of the sporophyte in the immediate vicinity 

 of the gametophyte specialized to aid 'the 

 latter in its work of developing the sex 

 cells, but this is true (though to a lesser 

 extent) of the sporophyte tissue for long 

 distances away, even to the confines of the 

 parts we call stamens and pistils, so that 

 I can not doubt that some at least of the 

 attributes properly belonging to a 'sexual 

 generation' have been transferred that far 

 back from the gametophyte into the sporo- 

 phyte. It is no objection to this view that 



* The intimate physiological iuterlooking of 

 gametophyte and sporophyte is strikingly illus- 

 trated in the phenomena of polyembryony, where 

 the sporophyte (nucellus) has acquired the power 

 of producing embrj^os within the embryo sac, 

 which embryos, although purely asexual, have the 

 general form and course of development of em- 

 bryos produced by the gametophyte. The phys- 

 iological equivalence of perisperm and endosperm 

 points in the same direction. 



I can not tell where in the ascending series 

 the 'sexuality' begins to pass over to the 

 sporophyte; even if we knew precisely the 

 actual stages in the evolution from fern to 

 phanerogam (which we do not), and even 

 if we were agreed upon a definition of 

 sexuality (which we are not), it might 

 still be impossible to tell precisely, so 

 subtle are the gradations of natural pro- 

 cesses, and so regardless are they of de- 

 finable categories. 



The sum of my argument, then, is this 

 — that in the phanerogams the physiolog- 

 ical line between the two 'generations' has 

 vanished, and that a large part of the 

 original function and attributes of the 

 gametophyte has been transferred to the 

 sporophyte which has had its tissues spe- 

 cialized to that end; hence the gameto- 

 phyte of the phanerogams is no longer a 

 'sexual generation' in point of physiolog- 

 ical fact, and it is misleading to use the 

 name as an expression of morphological 

 relations; sex not being confined to the 

 gametophyte, the sex terminology can not 

 be. 



Only the morphological line remains to 

 mark off the two generations in the phan- 

 erogam, but it is precisely this fact which 

 has caused the whole difficulty. Morphol- 

 ogists have found so great a satisfaction in 

 tracing the intricate but beautiful homol- 

 ogies from fern to phanerogam, that their 

 attention has become centered exclusively 

 upon the morphological phases of the sub- 

 ject, to the exclusion of its physiological 

 aspects. They have forgotten that sexu- 

 ality is more a matter of physiology than 

 of morphology, and that function cuts 

 across morphological boundaries in the 

 most irrelevant manner. They have fallen 

 into that error, against which Goebel has 

 so forcibly warned us, of attempting to 

 interpret morphology without reference to 

 function, a method which can lead only to 



