III.] REVERSIONS IN LEAF-FORMS. 97 



isms founded upon so fragmentary and scant material 

 as those applied to fossil plants; and yet I cannot help 

 feeling that some of these contemporaneous variations 

 are reversions to very old types. I was first led to this 

 opinion by a study of the sports in ginkgo leaves, and in 

 finding them suggestive of Mesozoic types. "This va- 

 riation in leaf characters," I wrote at the time,* "recalls 

 the geologic history of the ginkgo, for it appears to be 

 true that leaves upon the young and vigorous shoots of 

 trees are more like their ancestoi-s than are the leaves 

 upon old plants or less vigorous shoots, as if there is 

 some such genealogical record in leaves as there is in 

 the development of embryos in animals." Subsequent 

 observation has strengthened my belief in the atavistic 

 origin of many of these abnormal forms, and this expla- 

 nation of them is exactly in line with the characters of 

 reversions in animals and in cultivated plants. It would, 

 of course, be futile to attempt any discussion of the 

 merits of the specific types proposed by palaeobotanists, 

 but in those cases, like the ginkgo, where the geologic 

 types are fairly well marked, constant and frequent, and 

 where the similar contemporaneous variations are rare, 

 there is apparently good reason for regarding contem- 

 poraneous forms as fitful recollections of an ancient state ; 

 and this supposition finds additional support in the 

 ginkgo, because the species is becoming extinct, a fact 

 which also applies to the tulip -tree, which is now much 

 restricted in its distribution. I am further reinforced in 

 this view by Ward's excellent study of the evolution of 

 the plane-tree, for, in this instance, it seems to be well 

 determined that the geologic type has fairly well marked 

 specific characters, and the auricular or peltate base upon 



* American Garden, xii. 262 (1891). 

 7 SUE. 



