May, 1901] 



Kellerman — Syndesmon. 



109 



form of flowers and leaves almost without, exception, the second 

 (and third when present) do the same; if one varies in any respect 

 rarely does the remainder fail to follow suit. This can be seen in 

 the tabulation where two or more stems are indicated — both or all 

 are given (except in Nos. 29, 30 and 49) as observed, in the successive 

 serial numbers. Another instance of the persistency of an idiosyn- 

 cracy, as we may call it, was observed in some Syndesmons taken 

 from the woods by a gardener at Springfield, Ohio, over forty years 

 ago. The flowers were bountifully double, and the plants have each 

 year since faithfully presented the same striking peculiarity. 



The tabulation that follows is based on specimens from Toledo 

 (Lucas Co.), Nos. 1-30; from St, Marys (Auglaize Co.), Nos. 31-48; 

 from Steuben ville (Jefferson Co.), Nos. 49-65; from West Mansfield 

 (Logan Co.), Nos. 66-76; from Rendville (Perry Co.), Nos. 77-88; and 

 from Columbus, Nos. 89-100. The number of stems to each plant is 

 given in the second column; then follow in order the number of 

 flowers to each stem, the number of simple leaves with length of 

 their petioles in milimeters, the number of compound leaves with 

 length (also in milimeters) of their petioles and finally of their 

 petiolules. 



TABULATION. 



