Macbride — Notes on the Hydrophyllaceae 20 
‘PHACELIA INTEGRIFOLIA Torr., var. robusta, var. nov., planta 
robusta 4-12 dm. alta; caulibus ad medium 1 em. latis; foliis 
caulinis superioribus ovato-rotundatis s saepius 10 cm. longis 6 cm. 
eee Texas: Chinati Mts., 1882, Havard, no. 250 (Typr, Gray 
erb.). 
Only the upper portion of the plant is shown but the label bears 
the note ‘2-3 ft. high.” It was evidently a much larger plant 
than is usual for P. integrifolia in which species Dr. Gray included 
it. The leaves of the latter, however, are usually of an oblong 
rather than oval type. 
PHACELIA CRENULATA Torr., var. acco (Jones), comb. n 
P. ambigua Jones, Contrib. W. Bot. xii. 52 (1908). P. renulats 
Torr., var. vulgaris Brand, haha anise 6 iv. 251. 78 (1913) as to 
some specimens. 
The setose-hispid character of the pubescence on the stem dis- 
tinguishes the variety ambigua. The variety merges with the 
typical form but it is often well-marked and moreover almost if 
not entirely replaces in Arizona true P. crenulata. 
- PHACELIA CONGESTA Hook., var. rupestris (Greene), comb. 
nov. P. rupestris Greene, Leaflets i. 152 (1905); Brand, Pflan- 
zenreich iv. 251. 85 (1913); Wooten & Standley, Contrib. U. §. 
Nat. Herb. xix. 533 (1915). 
Hooker, 1. ¢., drew his description from plants collected by 
Drummond in the vicinity of Galveston, Texas, and specimens 
closely agreeing with the original diagnosis have been secured 
since only from the southern and eastern portions of the state. 
These are characterized by ample merely downy leaves and corol- 
las about 5 mm. long. Collections from farther north and west, 
however, exhibit great variation in the size of the leaves and the 
character of the pubescence. The leaves are generally smaller and 
the pubescence is often more pronounced on both leaves and stems. 
In New Mexico and far-western Texas there occurs a variation in 
which these tendencies culminate in extreme form to produce a 
plant with velvety leaves and more or less hispid stems. This state 
has received the name P. rupestris and Brand included the plant it 
represents in P. congesta. Wooten & Standley, 1. c., however, 
retained Greene’s name and thus excluded P. congesta from the 
flora of New Mexico. In this they were surely justified in so far as 
true P. congesta is concerned but examination of material from 
