104 CAEPOLITHES. 



or, possibly the better plan, by adding the word cycadean or 

 Cycadacece as a descriptive epithet to Garpolithes. If some such 

 course as this were generally followed, there would be less cause 

 for the not altogether unwarranted criticisms, which students of 

 recent plants are in the habit of passing on the misplaced 

 dogmatism of palasobotanists. Our records of fossil plants ought 

 surely to be sufficiently trustworthy, to be made use of by 

 botanists in compiling statistics of the geological history of 

 any class or family of plants. It must be admitted that to 

 attempt a history of plant development or distribution in the 

 various epochs of the earth's history, by simply accepting as 

 reliable data the examples of fossil plants, or fragments of 

 plants, described under the names of existing genera, or desig- 

 nated by terms plainly suggestive of botanical affinity, would 

 lead the too trustful student into hopeless error. Occasionally 

 a fossil seed may exhibit some definite and characteristic form, 

 for which some special specific term might be added, but in 

 the majority of cases where the individual differences are merely 

 those of size or slight variation in shape the use of specific 

 terms is to be deprecated. In Fontaine's Potomac Flora several 

 seeds are recorded as species of Carpolithes, the genus being 

 used in this instance for the "nut-like seeds of conifers." 1 

 Under the genus Cycadinocarpm the same author places "various 

 horny seeds which resemble those of cycadean plants more than 

 those of conifers." 3 It is admitted by Fontaine that the correct 

 placing of these seeds is impossible ; his species are founded in 

 some cases on very slight differences in size and shape, and can 

 have but little taxonomic value. Saporta 3 has instituted various 

 species for the French Jurassic seeds referred to gymnosperms ; 

 some of these show fairly well-marked characteristic features, 

 but in others it would be difficult to justify the adoption of 

 specific designations. In a recent paper by Dawson, several 

 gymnospermous seeds are wisely grouped together as examples 

 of Carpolithes.^ 



1 p. 264. 



2 p. 270. 



3 Loc. cit. pp. 238-245. 

 * Dawson (2), p. 90. 



