4IO Williafjt Rtissel Dudley 



That these rudiments have a normal origin is seen in G, 

 (left hand fig.), a drawing from sections of the young pistils 

 and anthers where the latter appear to be as well-formed at 

 this stage as the ovaries. The number of anthers on a well- 

 developed staminate spadix is usually twenty or twenty-two, 

 (forty to forty-four lobes), the number on the rudimentary 

 ones on a similar pistillate spadix is about twenty, the pistils 

 eighteen or twenty, showing that on the staminate spadix an- 

 thers have not developed in place of pistils ; on the contrary, I 

 have not been able to find the least trace of a rudimentary 

 pistil either on the young or the mature staminate spadix. 



From this evidence, the derivation of one genus from the 

 other seems undoubted ; also it would appear that the stami- 

 nate spadix, and consequently the dioecious condition of the 

 genus had taken on a very decided character, and had 

 probably been brought about in recent geologic times, if we 

 are to judge by the persistence and character of the rudi- 

 mentary organs. 



In structure, the anthers have been described by Bentham 

 and Hooker and others as " like Zostera." The rudimentary 

 organs with their connective enable us to prove this assertion, 

 the pair, as in Zostera,* constituting but a single anther. 

 Sections of Phyllospadix anthers (Fig. J) also show that 

 each lobe has two pollen-sacs similar in appearance to those 

 seen in sections of the anther-lobes of our Pacific coast Zos- 

 tera. In one respect they difi"er, however. In Zostera the 



* A curious confusion in the description of Zostera anthers appears in 

 all the standard systematic works in England and America, viz., in 

 Bentham and Hooker, Gray's Manual, and Watson's Bot. of Cal. The 

 anther is described as '' one" and " one-celled " — a manifest contradic- 

 tion. If the anthers are single, between pistils, then they are not one- 

 celled. If they are one-celled, then what we have termed anther-lobes 

 must be considered as two anthers. Eickler in " Bliithendiagramme, " 

 and Ascherson in " Die Pflanzenfamilien," give a lucid and perfectly 

 correct diagnosis of the anther. Bentham and Hooker say of Phyllo- 

 spadix also, " anthera .... i-locularis," which may have led Mr. 

 Morong, even so late as the current year (1893), into the statement, 

 (Naiadacese of the U. S.), that Phyllospadix has " numerous sessile 

 stamens in two rows . . . i-celled." 



