iyaii,, 1921 NOTES ON DOVES IN THE RIO GRANDE VALLEY 21 
but gather in trees and just sit around. On such days all birds show a de- 
creased weight, doubtless directly due to empty crops. When scattered over 
the open mesas feeding on doveweed, a rainy day interferes with their feeding. 
They simply sit around under the lee side of bushes, keeping dry. Nevertheless 
they make their regular evening flight to water although it would seem easy to 
gather the necessary water off the stems and leaves of plants. On such days 
the birds killed on the flyways to water show decreased weight due to empty 
erops. 
The daily flight to water generally starts between 3:00 p. Mm. and 4:00 P. M. 
and reaches its height just before dark. Sometimes, where long distances must 
be traversed, the birds do not arrive at water until after dark. When the wind 
parallels the water flight, the flight going and coming will be at different lev- 
els; for instance with the wind blowing toward water, the birds going to water 
will fly high and with tremendous speed, while the birds beating their way bac« 
against the wind will fly so low as to be nearly hidden by ordinary sage-brush. 
Some hunters take advantage of this and shoot only the easy birds beating into 
the wind. 
The dove’s choice of watering places is peculiar. In the hills they lke 
springs, stock tanks, and open sandy ereeks. On the Rio Grande bottoms they 
use waste irrigation water spilled along open roads, grassy vegas provided the 
cover is short, and ponds with bare shores. It is generally supposed that they 
seek freedom from cover which might harbor natural enemies, but if this is the 
main factor governing their choice of watering places, then the shallow rills on 
the big open sandbars of the Rio Grande ought to suit them exactly. But I have 
never seen a dove light on a river sandbar. Here is an interesting problem to 
solve. | 
Many doves are killed yearly in New Mexico but no decrease in numbers is 
noticeable. The area adapted to doves is so enormous that the total population 
is probably very little affected by localized shooting. 
Albuquerque, New Mexico, October 18, 1920. 
CONCERNING THE STATUS OF THE SUPPOSED TWO RACES 
OF THE LONG-BILLED CURLEW 
By JOSEPH GRINNELL 
(Contribution from the Museum of Vertebrate Zoology of the University of California) 
N THE FALL of 1918 vertebrate collecting was carried on by a party from 
the California Museum of Vertebrate Zoology at Morro, San Luis Obispo 
County, California. Among the birds of interest obtained there was the 
series of Long-billed Curlew listed in the table presented farther along in this 
paper. The acquisition of so many specimens (fifteen) of this Curlew. taken 
from one locality and within a period of less than one month in a single year, 
seemed to afford ground for looking into the standing of the recently proposed 
two races of the species, namely Numenius americanus americanus Bechstein and 
Numenwus americanus occidentalis Woodhouse. This separation was first pro- 
