ON THE STRUCTURE AVD CLASSIFICATION OF SEE^ J^Q 



them for a future trial ; nor have I dissected the seeds of 

 the sea-weeds. 



The third class of seeds is a numerous one, and I have 3^ class 

 called it The canaliculated. Set fig. 10. It is distinguished Canillculata, 

 by a larger heart, with a curious sweep, which figure the 

 teats follow. The teats are numerous, and have the nou- 

 rishing vessels above them. This ch'ss takes in almost all 

 the papihonaceous, ringent, and many of the cruciform 

 flowers. The formation of the corculum (much as they may 

 differ in each seed) will still be found to have the mark of 

 this class ; which is principally a deep furrow beginning 

 •with the recess, runing on to the end of the primordial leaves, 

 and lengthening as the embryo increases. I have two or three 

 times found cotyledons in this passage, and I am rather in- 

 clined to believe, that farther search will show more, espe- 

 cially in the papilionaceous, which is also distinguished by 

 a curious sheath, that holds that jelly found constantly in 

 the pocket of the seed, and against which the primordial 

 leaves shoot. But 1 do not conceive, that more than four 

 cotyledons will ever be found. I have never seen more. 



The fourth class is the nonmammiferous, and is the one 4th class 



that differs most from the rest ; for it has neither recess, nor f*'^"'"*'"*"'- 



rerous. 

 teats. See PI. VI, fig. 1 and 2. The palms, and grasses, 



are included in this ; beside many odd plants, which it 

 would be useless in such a sketch to mention. The distin- 

 guishing marks of this class are the cotyledons proceeding 

 from the upper end of the corculum, instead of the usual 

 place ; this was the reason, that in the grasses botanists over- 

 looked them ; took the primordial leaf for the cotyledons, 

 and named them monocotyieilonoos. But had they dissected 

 the interior, they would have found, that they are placed 

 (with respect to the primordial leaf) exactly an in every other 

 plant ; both rising and branching iVom the same apparent 

 source. This is sufficient to prove, that these little leaves 

 (always given in those excellent drawings of Sowerby) are 

 really the cotyledons of the grasses ; and that they have al- 

 ways either two or four, as weli as the palms. The class is 

 also easily known by having the stalks running through the 

 corculum without impediment ; and the nourishing vessels 

 protruding on one side of the heart only, which has been the 

 N 2 cause 



4 



