schaeffer: donaciini of the new world. 99 
form and sculpture of prothorax in which they vary so much that two 
or three specimens could be selected, which look distinct enough from 
each other to be considered good species. However, intermediate 
specimens make it very difficult to separate even certain specimens as a 
variety. Moreover, there are specimens so close to some porosicollis 
which are troublesome to place correctly. In certain specimens the 
prothorax in form and sculpture is very much hke that of some specimens 
of porosicollis, in a few as in tuberculijrons and others more or less like 
distincta. I have given a name to the more extreme variation which I 
have from several locahties and which I hope will make it a little easier 
to determine this species. 
Dr. Leconte in his revision identifies Melsheimer's biimpressa (blue 
form) and aurichalcea (cupreous form) as emarginata and specimens of 
the latter are so labelled in his collection. However, Melsheimer de- 
scribes the hind femora of both as reddish at base — entirely metallic in 
emarginata — which first called my attention to the possible misidentifi- 
cation of Melsheimer's species. Melsheimer's types are supposed to be 
in the Leconte collection but as stated above the specimens so labelled 
are emarginata. At the suggestion of Mr. Banks I looked for Melsheimer 
Donacias in the general collection and found there the two blue specimens 
(biimpressa) and a cupreous specimen (aurichalcea) which agree in every 
respect with the descriptions and are without doubt his types. They 
are not emarginata but the same as torosa Lee. The distincta of Mr. 
Leng's revision is not Leconte's species but the cupreous form of this 
species. The elevated sutural margin of elytra near apex is said by 
Mr. Leng to be sinuate in his distincta but this is a mistake, the sutural 
margin is as in subtilis and allies. 
The blue, bluish-green, purple and black specimens occur more 
commonly in the males than in the females, as I have seen only two blue 
females among the numerous specimens from a number of different 
localities. 
Typical biimpressa have the ventral segments, especially the first, 
rather sparsely punctate and very sparsely pubescent with very short, 
white or cinereous hairs, which are often almost invisible at middle of 
first ventral segment; prothorax generally as wide at apex as long, or 
very nearly so, the sides behind the lateral tubercles more or less con- 
verging towards base and scarcely sinuate, the punctuation very variable 
from very fine and sparse with finely impressed lines between the 
punctures to relatively moderately coarsely and closely punctate; the 
intervals of the elytra moderately finely rugose, near the scutellum 
