1913] NELSON & MACBRIDE~WESTERN PLANTS 379 
CASTILLEJA MINIATA Dougl.—Special effort was made to secure 
as full a series as possible in this genus in order to see something of 
the degree of variability within specific lines. No species shows how 
great this variability is better than C. miniata. Incidentally it may 
be remarked that the making of synonyms is not always due to the 
variability in the specimens representing a species, but quite as often 
to errors in the original descriptions which authors continue to 
copy and with which we constantly compare new material. Another 
source of error is found in the habit, more or less prevalent, of 
naming up material by comparison alone. A has a specimen 
slightly aberrant; B names his by comparison with A’s, and lets it 
pass though evidently somewhat different; C, having B’s plant, 
names his material accordingly, and so on. Look through any well 
filled species cover and see these facts illustrated. 
In the original description of C. miniata the galea is said to 
exceed the corolla tube. On the assumption that this was so for 
the species (it probably was on the type sheet) some segregates have 
been made by various authors that had better not have been made. 
C. miniata material measured by this yardstick contains few (if any) 
specimens that are typical. Dr. RyDBERG, for example, in working 
all the material for his Flora of Montana had but one number 
(RyDBERG and Bessey 4965, Wolf Creek, 1897). Unfortunately, 
too, other sheets of even this same number (4965) show only 
corollas in which the galea is shorter than the tube. The character 
is therefore more or less unreliable in this species, and hence 
probably in others. : 
Another character in Castilleja which is often misinterpreted is 
the root system. Collectors almost always pull the stems loose 
from the root or caudex. Being usually numerous in the clumps 
they are decumbent at base, and more or less rooted at the lower 
nodes. This often gives them the appearance of having been 
detached from a running rootstock when such was not the case. 
Of course it is generally known that the color of the bracts 
varies, and this gives the plants a very different look even when the 
color of calyx and corolla remains (as it should) fairly constant. 
These observations were induced by the necessity of distributing several of 
our numbers as C. miniata in spite of the fact that they refuse to accord fully 
