1921] Weatherhy,— Old Time Connecticut Botanists 173 



account of them may not be out of place. Among the willows, he 

 recognizes 29 species (fourteen or fifteen more than current manuals 

 admit in lis region", four of them indicated as new though only three 

 are properly published, and fourteen varieties, two unnamed, eleven 

 with nomina inula and one with a brief description. This one, Salix 

 tristis, var. monadelphia, is an apparently teratologic^ form in which 

 the filaments are uiited, as they normally are in S. purpurea. Bar- 

 ratt distributed his 29 species among eight sections, using as diagnostic 

 characters the time of appearance of the aments, whether before or 

 after the leaves, the point on the anient, apex, middle, or base, at 

 which the staminate flowers first appear, shape of leaves and type 

 of pubescence on them, color of scales, etc. A vast deal of patient 

 labor and observation went into this work and though the sections 

 are in some cases separated by what are now regarded as no more 

 than specific differences and some of the species are either hybrids 

 or divided by finer lines than later students have been able to draw, 

 yet the grouping is essentially that now in use in our current manuals. 



Barratt issued similar exsiccatae sets with printed title-page and 

 label-sheets, of the local Carices and Eupatoriums of the E. purpureum 

 group. The former contains (>7 species, mostly recognized now, and 

 some dozen varieties, only two of which, C. vulpinoidea, vars. glomer- 

 aia and ambigua? are described. Some of the varietal nomina inula 

 are applie 1 to segregates now recognized — the plants, for instance, 

 now known as Carex virescens, var. Swanii and C. vesicaria, var. 

 jejuna. Others seem to be based on wholly trivial and superficial 

 characters. A case in point is that of C. squarrosa and C. typhina, 

 related species which grow together in the Hood-plain of the Connect- 

 icut River. Barratt entirely overlooked, or thought of no account, 

 the characters of leaves, scales, form of spike and achene by which 

 they are separated, lumped both together under one species and 

 divided the aggregate into five unnamed varieties based entirely on 

 the number of fruiting spikes. 



In his work on the Kupatoria, he anticipates our latest treatment 

 by dividing the group of E. purpureum into four species, one new — to 

 which he adds three varieties. The new species, E. fi.siulosutn, is 



1 The former is a state of C. vulpinoidea with short, thick inflorescence; the latter 

 is, of course, C setacea, var. ambigua of Gray's Manual, C. xanthocarpa and C an- 

 nectens of Bicknell. 



