224 MORPHOLOGY OF THE ANGIOSPERMS 



building of the tubular follicle. In the follicles of the Ranunculaceae, 

 an elongated fertile strip is claimed to be built up by a cross zone more 

 active than that of Rammciilus. The carpel of Calycanthus is cited as 

 showing steps in the transition from the primitive achene to the follicle: 

 its short stipe and ovary represent a first stage in the evolution of a 

 longer stipe and a longer tubular ovary; the two ovules, a step from the 

 one of the achene to the many of the follicle. The follicle, with its 

 many ovules on this long, fertile strip built by the cross-zone meristem, 

 is interpreted as a higher type of carpel. (This interpretation of follicle- 

 achene relationship is the reverse of that based on much generally ac- 

 cepted evidence from comparative form, vestigial structure, and vascular 

 supply. ) 



Epeltate carpels "show no evidence of peltation" at any stage; they 

 have no cross zone. The carpel primordium arises crescent- or horse- 

 shoe-shaped in cross section and remains so; the closure of the carpel is 

 wholly ontogenetic, and there is, therefore, "no strictly tubular part" — 

 that is, a part formed by a cross zone. Few taxa with carpels of this 

 type are listed: Butomaceae, Hydrocharitaceae, Aponogetonaceae, 

 Nymphaeaceae, Alismataceae, Scheuchzeriaceae. (Both follicles and 

 achenes are present in this group of families. ) The absence of a cross 

 zone in these plants is considered remarkable and explained only as 

 the result of reduction. Whenever a stipe is present in epeltate forms, 

 this theory maintains that there must have been originally a cross zone. 



Lines are not clearly drawn between the types of peltate carpels, and 

 opinions as to the most primitive form seem to differ. The epeltate 

 forms are considered reduced and derived from the peltate by loss of 

 the cross zone. The latent peltate (achene) form is considered less ad- 

 vanced than the manifest peltate. The manifest peltate form is called 

 the "basic form from which the others are derived." 



Abundant evidence from anatomy, ontogeny, and comparative mor- 

 phology of the families listed as having epeltate carpels, supports the 

 view that the peltate theory reads the series in carpel form in the wrong 

 direction; simplicity, here, is advanced, not primitive. The theory that 

 the peltate form is basic for the carpel is not supported by comparisons 

 throughout the angiosperms. If leaf form were interpreted similarly, 

 the shield-shaped and pitcher-shaped leaves would be considered basic 

 types. The taxa considered epeltate are phylogenetically primitive, ac- 

 cording to the evidence of ontogeny, anatomy, and Hower structure, 

 and, in them, the peltate form should be seen in its simplest expression, 

 if it is basic. Carpels of these taxa are interpreted as reduced from 

 peltate types. The peltate theory assumes as primitive a carpel of the 

 achene type, with a single ovule borne on the cross zone, a secondarily 



