FLORIN: SYSTEMATICS OF THE GYMNOSPERMS 357 



lateral type. In Whittleseya they are campanulate structures, built up of a 

 single whorl of numerous united sporangia, immersed in sterile tissue and en- 

 closing a central cavity. Aulacotheca, another extreme, consists of narrow, 

 hollow, seedlike capsules made up of whorled sporangia. Halle suggested that 

 the synangium of this group may be derived from a terminal tuft of cyclically 

 arranged sporangia. Dolorotheca, also a genus of campanulate pteridosperm 

 male fructifications, differs from the Whittleseyineae by the tubular microspo- 

 rangias not forming a single whorl. This genus was later studied by Schopf 

 (1948; cf. Baxter, 1949), who assembled its species in the subgroup Dolerothe- 

 cineae. Here, the sporangia are arranged in biseriate rows radiating from the 

 center of the fructification. Other microsporangiate fructifications were de- 

 scribed by Walton (1931, 1949a) and Read (1946). In Diplopteridium the api- 

 cal, dichotomized portion of the main rachis of the frond is the synangia-bearing 

 part. Alcicornopteris has tufts of free sporangia borne terminally on dicho- 

 tomous branchlets. Lacoea (Read) consists of cupular, spore-bearing organs 

 attached to slender rachises and believed to enclose elongate tubular sporangia 

 on a convex receptacle. 



Information on seed-bearing pteridosperm fructifications was given in some 

 cases. The Calathospermum cupule was the first many-ovular type of paleozoic 

 age in which the order and arrangements of the ovules could be studied. The 

 presence of a crescentic bundle in its stalk suggested to Walton (1949b) that the 

 whole structure was morphologically equivalent to an inrolled or folded frond 

 or part of a frond. Salpingo stoma (Gordon, 1941) is a similar many-ovular 

 cupule. The summit of the lagenostome of their ovules is prolonged into a tubu- 

 lar structure called the salpinx, and there is no micropyle. In later pterido- 

 sperms, a micropyle was formed and the salpinx became stronglj^ reduced or dis- 

 appeared. Our knowledge of the anatomy of pteridosperm seeds was furthered 

 by several writers, who dealt particularly with those of the Trigonocarpales 

 (Hoskins and Cross, 1946, and others). Seward (1917) had classified the paleo- 

 zoic seeds into the Lagenostomales, Trigonocarpales, and Cardiocarpales. The 

 first two groups corresponded to the radiosperms of Oliver (1904), and the third 

 to his platysperms. Loubiere (1938) distinguished instead between the Nerto- 

 caryales, the Mesocaryales, and the Acrocaryales. In the first group the nucellus 

 and integument are fused, in the second they are free except at the base; the 

 third group is based on Leptotesta. But Emberger (1944; cf. Arnold 1938, 1948) 

 disputed Loubiere's interpretation of this genus, and was inclined to refer it 

 to the Nertocaryales. 



Caytoniales and Related Groups 



The Caytoniaceae became much better known than they were at the end of 

 the 1920's (Harris, 1932-1937, 1933, 1940a, 1940b, 1941b, 1951a). Caytonia 

 is a pinnate megasporophyll, with the pinnae attached to a rachis and bearing 

 ovules in rows on their incurved adaxial surfaces. The "fruit" is no carpel, and 

 its "stigma" is not stigmatic, but merely a lip. The pollination was gymno- 

 spermous. The inner part of the "fruit" wall was thick and fleshy, embedding 

 the seeds, and narrow canals led from the micropyles towards the lip. The micro- 

 sporophyll (Caytonanthus) consists of a pinnately branched rachis, the lateral 



