Terry A. Vaughan

Family DIDELPHIDAE

Didelphis marsupialis virginiana Kerr

Virginia Opossum

The opossum is common in and near small towns and cultivated areas at the Pacific base of the mountain range and does not thrive away from human habitation; extensive trapping in the coastal sage and chaparral belts produced no specimens except immediately adjacent to citrus groves. Pequegnat (1951:47) mentions that opossums in the Santa Ana Mountains of southern California are in the lower parts of the larger canyons, especially near human habitation.

Specimens examined.—Los Angeles County: Claremont, 1600 ft., 2 (PC).

Family TALPIDAE

Scapanus latimanus occultus Grinnell and Swarth

California Mole

Workings of moles were found on the Pacific slope of the mountains from 1600 feet at Claremont up to 7500 feet on Blue Ridge, and on the Pacific slope beneath basin sagebrush in Cajon Canyon one mile from desert slope Joshua-tree flats, but not on the desert slope, although moles probably occur on that slope in some of the places where there is suitable habitat.

Near Camp Baldy in the sandy soil beneath groves of alders moles seemed to be especially abundant. Although common on the coastal face of the range, moles shunned compact, dry, or rocky soils. In the greasewood chaparral one-half mile west of the mouth of Palmer Canyon, where the soil was hard and rocky, mole tunnels were in soft soil that had accumulated at the edge of a fire road beneath a steep road cut. The assumption is that this accumulation contained insects attractive, as food, to the moles.

Specimens examined, 2: Los Angeles County: Camp Baldy, 4200 ft., 1(PC); Claremont, 1600 ft., 1(PC).

Family SORICIDAE

Sorex obscurus parvidens Jackson

Dusky Shrew

Jackson (1928:124) recorded a specimen from Camp Baldy, 4200 feet, San Antonio Canyon.

Sorex ornatus ornatus Merriam

Ornate Shrew

Both of my specimens were taken amid riparian growth on the Pacific slope of the range.

Specimens examined, 2: Los Angeles County: San Antonio Canyon, 3500 ft., 1; Cobal Canyon, 5 mi. N Claremont, 1800 ft., 1 (PC).

Notiosorex crawfordi crawfordi (Coues)

Gray Shrew

One was taken in 1946 beneath a woodpile on the campus of Norton School, two miles northeast of Claremont, and examined by Dr. W. E. Pequegnat.

Family VESPERTILIONIDAE

Myotis yumanensis sociabilis H. W. Grinnell

Yuma Myotis

A female was taken in lower San Antonio Canyon, 2800 feet elevation, on September 27, 1951.

Myotis evotis evotis (J. A. Allen)

Long-eared Myotis

This species was observed and collected at several stations ranging from 2800 feet elevation in San Antonio Canyon, to Blue Ridge at 8200 feet, and down the desert slope to 6000 feet at Jackson Lake. This distribution encompasses most of the chaparral and yellow pine forest associations. Within these areas, however, this bat shows marked habitat preferences.

Woodland habitats seem to be preferred by evotis. At several ponds in lower San Antonio Canyon this bat was observed repeatedly as it foraged over the water and coursed low between rows of alders and Baccharis. At Blue Ridge in September, 1951, these bats foraged approximately six feet above the ground beneath the canopy of coniferous foliage and between the trunks of the trees.

Most of the bats were taken by stretching fine wires above the surface of a pond as outlined by Borell (1937:478). Collecting was generally carried on until at least 11:00 p. m., and the time at which each bat was taken at the pond was recorded, thereby making possible a rough estimate of the pre-midnight forage period of each bat commonly collected at the ponds. Usually bats taken at the start of their supposed forage period had empty or nearly empty stomachs, whereas those taken towards the end of their forage period had full or nearly full stomachs. M. evotis usually first appeared just at dark, well after the pipistrelles and California myotis had begun foraging. The forage period of evotis seemed to begin approximately 30 minutes after sunset and to end approximately two and one-quarter hours later.

Individuals of this species were taken from May 4, to October 14, 1951. A female taken on May 19, 1951, in San Antonio Canyon, carried one minute embryo, and one taken in the same locality on June 8, had one embryo four millimeters in length.

Specimens examined.—Total, 12, distributed as follows: Los Angeles County: San Antonio Canyon, 2800 ft., 11; Claremont, 1100 ft., 1 (P.C.).

Myotis volans interior Miller

Interior Long-legged Bat

Although seldom found to be plentiful, this bat was recorded from many points on both the coastal and desert slopes of the mountains. Specimens were taken in the chaparral association in San Antonio Canyon, near Jackson Lake among yellow pines, and in Mescal Canyon at the upper limit of the Joshua tree woodland. Bats, probably volans, were noted over sage flats at 8000 feet elevation on Blue Ridge. The only place where these bats appeared to be numerous was Jackson Lake on the interior slope; there, on September 19, 1951, volans appeared with the pipistrelles, and was the most common bat before dark.

An individual of this species taken on October 28, 1951, in a short mine-shaft in the pinyon belt at the head of Grandview Canyon was slow in its movements and felt as cold as the walls of the tunnel. It was late afternoon and the temperature outside the cave was below 40°F. The floor of the tunnel was covered with the hind wings of large moths of the genus Catocala; volans probably hung in the cave while eating them.

The series of volans from the San Gabriels shows that the two color phases of this bat both occur in the area. Two specimens from Jackson Lake contrast sharply with the rest of the series in their dark coloration. Benson (1949:50) states that color variation in a series of volans from a given locality may be striking.

This bat was collected in San Antonio Canyon from 50 minutes after sundown to two hours and 40 minutes after sundown. In this area these bats did not visit the ponds in large numbers as they seemed to do on the desert slope.

A female taken on May 29, 1951, contained one embryo nearly at term.

Specimens examined.—Total, 9, distributed as follows: Los Angeles County: Mescal Canyon, 8 mi. E and 5 mi. S Llano, 4900 ft., 1; 3 mi. W Big Pines, Swarthout Valley, 6000 ft., 3; San Antonio Canyon, 2800 ft., 5.

Myotis californicus californicus (Audubon and Bachman)

California Myotis

On the Pacific face of the mountain range this bat was recorded commonly below approximately 5000 feet elevation, where it seemed to be most common in the oak woodland of canyons. On the desert slope it was collected at Jackson Lake in yellow pine woodland, in Mescal Canyon in the juniper belt, and bats presumably of this species were observed at several points in the pinyon-juniper woodland.

Individuals of this species were often observed foraging from five to ten feet above the ground around the alders and Baccharis near San Antonio Creek, but they did not fly so low or so near the vegetation as did Myotis evotis. Here they were taken from 18 minutes to 55 minutes after sunset; this indicates an early and short forage period.

This bat may be active even in winter. On February 8, 1952, in lower San Antonio Canyon, a bat, probably of this species, was noted foraging; and collecting in early November, 1951, yielded specimens.

On May 22, 1951, a female obtained in San Antonio Canyon had one five-millimeter embryo, and subsequently all the females examined had embryos until June 12, when collecting was discontinued.

Specimens examined.—Total, 16, distributed as follows: Los Angeles County: Mescal Canyon, 4800 ft., 2; Jackson Lake, 6000 ft., 1 (PC); San Antonio Canyon, 3900 ft., 1; San Antonio Canyon, 2800 ft., 12.

Pipistrellus hesperus merriami (Dobson)

Western Pipistrelle

This is the most obvious if not the most common bat of the lower coastal slopes of the San Gabriels. In the spring and fall of 1951 individuals were noted from 1700 feet in the coastal sage scrub association to the white fir forests on Blue Ridge at 8200 feet elevation and were commonest in the rocky canyons of the lower Pacific slope below 4000 feet, and usually foraged near the steep canyon sides high above the canyon bottoms.

Pipistrelles were generally the first bats to appear in the evening, although the times of their appearance were irregular. In April and May, in lower San Antonio Canyon, they appeared from 28 minutes before sunset to 30 minutes after sunset, with the average time of appearance eight and one-half minutes after sunset. Like Myotis californicus this pipistrelle seemed to have a short and early foraging period. No pipistrelles were recorded at ponds later than one hour and five minutes after sunset, and usually they were not seen later than 40 minutes after sunset. Most of the specimens taken later than one half hour after sunset had full stomachs. More than 50 pipistrelles were captured at the ponds in San Antonio Canyon; six were kept for specimens. This species is probably present in the area throughout the winter. Pipistrelles were active in early April in Evey Canyon, were observed in early November in San Antonio Canyon, and on January 26, 1952, an individual was noted foraging near the mouth of Palmer Canyon. They are probably not active in winter on the colder desert slope of the mountains.

Pipistrelles often foraged in loose flocks of about half a dozen individuals. On many occasions these groups were first seen foraging high up above the canyon bottom, then, as it grew darker, they descended and foraged within 50 or 100 feet of the floor of the canyon. Immediately before dark these groups seemed to have forage beats; one minute several pipistrelles would be overhead, and the next minute none would be in sight.

A female taken in San Antonio Canyon on June 8, 1951, contained two five-millimeter embryos.

Specimens examined.—Total, 6, distributed as follows: Los Angeles County: San Antonio Canyon, 2800 ft., 5; Evey Canyon, 2400 ft., 1.

Pipistrellus hesperus hesperus (H. Allen)

Western Pipistrelle

This species was common in the spring and autumn of 1951 from the lower edge of the yellow pine forest down into the belt of Joshua trees. In early April on the desert slope at 4800 feet in Mescal Canyon, pipistrelles foraged on evenings when it was windy but not cold. On cold evenings (when the temperature was below roughly 45°F) none was seen. On windy nights the pipistrelles often forsook their usual high forage habits and foraged 15 feet or so above the ground where the vegetation and outcrops of rock broke the force of the wind. In 1951 no pipistrelles were noted on the desert slope later than October 15.

Specimens examined.—Los Angeles County: Mescal Canyon, 4800 ft., 4.

Eptesicus fuscus bernardinus Rhoads

Big Brown Bat

This bat was on the coastal slope from the sage scrub association at 1100 feet, up to 8000 feet on Blue Ridge, and on the desert slope down to the upper edge of the Joshua tree belt at 4800 feet in Mescal Canyon. It was the most common bat at the ponds in San Antonio Canyon in May and June of 1951, but in September and October of the same year none was obtained there.

On the Pacific slope of the San Gabriels the big brown bats segregate according to sex in the spring, the males occupying the foothills and mountains and the females the level valley floor at the coastal base of the range. Of 70 big brown bats captured in May and June of 1951, at the ponds in San Antonio Canyon, only one was a female. A large colony of more than 200 individuals in a barn near Covina, in the citrus belt, was composed of only females.

Times of capture of this bat at the ponds in San Antonio Canyon ranged from ten minutes after sunset to two hours and thirty minutes after sunset. Generally these bats came to the ponds in groups of several individuals, and often more than a dozen were captured in the course of an evening's collecting.

Specimens examined.—Total, 7, distributed as follows: Los Angeles County: Mescal Canyon, 4800 ft., 1; San Antonio Canyon, 2800 ft., 2; Covina, 1100 ft., 4 (2PC).

Lasiurus borealis teleotis (H. Allen)

Red Bat

One female was taken on September 30, 1951, in San Antonio Canyon, at 2800 feet elevation. The descriptions which the citrus growers of the Claremont and Glendora vicinity give of the bats they find occasionally hanging in their citrus trees accurately describe this species. Its seasonal occurrence there is unknown.

Lasiurus cinereus cinereus (Pasilot de Beauvois)

Hoary Bat

Specimens were collected in spring in 1951 at elevations of 2800 and 3200 feet in San Antonio Canyon, on the coastal slope, and in Mescal Canyon at 4900 feet, on the desert slope. Large, fast flying bats, probably of this species, were seen at Jackson Lake, 6000 feet elevation, on October 15, 1951.

Hoary bats are present in the San Gabriels in the fall, winter, and spring. In 1951 the last spring specimen was taken on June 11, in Mescal Canyon; then collecting was discontinued until late September when the first hoary bat was taken on the thirtieth of that month. From this date on into the winter hoary bats were recorded regularly. They seemed to be as common in early June as in most of April and May; possibly some remain in the San Gabriels throughout the summer.

In spring these bats seem to segregate by sex; of twelve kept as specimens and at least an equal number captured and released only one was a female. All were captured above 2800 feet.

Hoary bats seem to have a long pre-midnight forage period, having been captured at ponds from 21 minutes after sunset, to three hours and 26 minutes after sunset. Generally those taken early had empty stomachs and those taken later had full stomachs. On the night of May 24, 1951, a hoary bat captured two hours and five minutes after sunset had only a partially full stomach.

On May 25, 1951, an unusual concentration of hoary bats was observed at a pond at about 3200 feet elevation, in San Antonio Canyon (Vaughan, 1953). The day had been clear and warm, one of the first summerlike days of spring. Beginning at 30 minutes after sundown hoary bats were collected until two hours and 35 minutes after sundown; in this period 22 were caught and at least as many more observed. Many were released after being examined, whereupon they hung on the foliage of nearby alders to rest and dry themselves. This concentration of hoary bats may have been due to a sudden beginning of migration with a resultant concentration of bats at certain altitudinal belts. The warm weather might have set off the migration. On evenings that followed subsequent hot days no such concentration of hoary bats was seen. B. P. Bole (Hall 1946:156) observed a concentration of hoary bats on August 28, 1932, in Esmeralda County, Nevada.

Several captive Myotis californicus in a jar next to a pond in San Antonio Canyon set up a squeaking which seemed to attract a hoary bat. Repeatedly the large bat swooped over the jar.

Specimens examined.—Total, 12, distributed as follows: Los Angeles County: Mescal Canyon, 4900 ft., 2; San Antonio Canyon, 3200 ft., 2; San Antonio Canyon, 2800 ft., 8.

Antrozous pallidus pacificus Merriam

Pallid Bat

The pallid bat is probably the most common and characteristic bat of the citrus belt at the Pacific base of the mountains. Only once, on May 4, 1951, was this bat taken in the mountains. On that night two individuals were collected at 2800 feet in San Antonio Canyon. All of the other specimens and observations were from colonies in old barns and outbuildings in the citrus belt where these bats are found in spring, summer, and fall.

The impression gained by examining many mixed colonies of Antrozous and Tadarida was that the former greatly outnumbered the latter. For example, a small colony of bats in an old barn near San Dimas Wash consisted of about thirty pallid bats and five freetails.

Large numbers of wings of moths of the family Sphingidae, and legs and parts of the heads of Jerusalem crickets (Stenopelmatus fuscus) were beneath an Antrozous night-roosting place in a barn near Upland.

Pallid bats were collected in 1951, from April 16 to October 17 but probably were active in the area into November.

Each of two pregnant females taken two miles northeast of San Dimas on April 20, 1951, carried two embryos 4 millimeters long.

Specimens examined.—Total, 6, distributed as follows: Los Angeles County: 2 mi. NE San Dimas, 1200 ft., 2 (1PC); Ontario, 1100 ft., 4 (3PC).

Family MOLOSSIDAE

Tadarida mexicana (Saussure)

Mexican Free-tailed Bat

This bat, regularly met with in the citrus belt at the coastal base of the range, occurred in small numbers with colonies of Antrozous, and was once found with a colony of Eptesicus near Covina. None of the females taken in April 1951 was pregnant.

Specimens examined.—Los Angeles County: 2 mi. NE San Dimas, 1200 ft., 4.

Eumops perotis californicus (Merriam)

Mastiff Bat

H. W. Grinnell (1918:373) mentioned individuals collected at Sierra Madre (at the coastal base of the San Gabriels west of the study area), and Sanborn (1932:351) reported specimens from Covina and Azusa. Probably this bat occurs locally all along the coastal base of the range.

Family LEPORIDAE

Lepus californicus bennettii Gray

California Jack Rabbit

This species was found in the coastal sage belt from Cajon Wash west to San Gabriel Canyon and was most plentiful in thin stands of sagebrush, and in and around citrus groves. Because of their preference for semi-open country, jack rabbits are absent from much of the coastal belt of sagebrush where the brush is fairly continuous, and they never were observed in the chaparral association.

Coyotes catch many jack rabbits and regularly forage around the foothill borders of the citrus groves for cottontails and jack rabbits.

A female examined on February 19, 1951, was pregnant, and one taken on March 15, 1951, carried three small embryos.

Specimens examined.—San Bernardino County: 2 mi. NW Upland, 1600 ft., 3 (PC).

Lepus californicus deserticola Mearns

California Jack Rabbit

There was sign of jack rabbits along the desert slope of the San Gabriels up to about 6700 feet, one-half mile west of Big Pines. They were fairly common in the Joshua tree belt, occurred less commonly in the juniper belt, and were present locally in small numbers in the pinyon-juniper association.

The population seemed to be at a low ebb from 1948 to 1952, when field work was done on the desert slope. I often hiked for an hour or more on the desert or juniper-covered benches without seeing a jack rabbit. The species was commoner in washes where as many as eleven were noted in two hours' hiking.

In December, 1951, below Graham Canyon, the leaves on large areas of many nearly recumbent Joshua trees had been gnawed down to their bases, and jack rabbit feces covered the ground next to these gnawings. Probably the Joshua tree is an emergency food used by the rabbits only when other food is scarce.

In years when the population of jack rabbits is not low they serve as a major food for coyotes. In the Joshua tree belt below Mescal Canyon, jack rabbit remains were fairly common in coyote feces, and tracks repeatedly showed where some coyote had pursued a jack rabbit for a short distance. A large male bobcat trapped in the juniper belt in Graham Canyon had deer hair and jack rabbit remains in its stomach.

Specimens examined.—Total, 7, distributed as follows: Los Angeles County: 6 mi. E and 1 mi. S Llano, 3500 ft., 4; Mescal Canyon, 4800 ft., 3.

Sylvilagus audubonii sanctidiegi (Miller)

Audubon Cottontail

Cottontails are common in the coastal sage scrub association and in and around citrus groves, but generally penetrate the mountains no farther than the lower limit of the chaparral association. They are everywhere on coastal alluvial slopes, except in the barren washes, and prefer patches of prickly-pear and often are loathe to leave its protection. After completely destroying a large patch of prickly-pear in the course of examining a wood rat house in the center of the cactus, I found hiding, in the main nest chamber of the house, a cottontail that dashed from its hiding place only when poked forceably with the handle of a hoe.

Cottontails are seldom above the sage belt in the chaparral associations, although along firebreaks and roads they occasionally occur there. Habitually cottontails escape predators in partly open terrain offering retreats such as low, thick brush, rock piles, and cactus patches; but on open ground beneath dense chaparral, cottontails may be vulnerable to predation.

Examinations of feces and stomach contents of the coyote reveals that it preys more heavily on cottontails than on any other wild species. Remains of several cottontails eaten by raptors were found in the sage belt.

In April, 1951, many young cottontails were found dead on roads in the sage belt, and a newly born cottontail was in the stomach of a coyote trapped four miles north of Claremont, on February 7, 1952.

Specimens examined.—Total, 3, distributed as follows: Los Angeles County: mouth of San Antonio Canyon, 2000 ft., 1 (PC). San Bernardino County: 2 mi. NW Upland, 1600 ft., 2 (PC).

Sylvilagus audubonii arizonae (J. A. Allen)

Audubon Cottontail

This subspecies was recorded on the interior slope from 5200 feet elevation, as at the head of Grandview Canyon, down into the desert, and was common in the sagebrush flats of the upper pinyon-juniper association. Piles of feces under thick oak and mountain-mahogany chaparral indicated that the rabbits often sought shelter there. Adequate cover is a requirement for this rabbit on the desert slope of the San Gabriels; in the juniper and Joshua tree belts the species occurs in washes where there is fairly heavy brush, and only occasionally elsewhere. In the foothills, when frightened from cover in one small wash cottontails often run up over an adjacent low ridge and seek cover in the brush of the next wash. In the wash below Graham Canyon tracks and observations showed that cottontails were taking refuge in deserted burrows of kit foxes.

In the pinyon-juniper association cottontails and jack rabbits probably occur in roughly equal numbers, but in the Joshua tree belt cottontails seem far less numerous than jack rabbits. In the course of a two hour hike in lower Mescal Wash, at about 3500 feet, eleven jack rabbits and two cottontails were noted.

Specimens examined.—Total, 2, distributed as follows: Los Angeles County: 6 mi. E and 1 mi. S Llano, 3500 ft., 1; Mescal Canyon, 4800 ft., 1.

Sylvilagus bachmani cinerascens (J. A. Allen)

Brush Rabbit

Brush rabbits inhabit the Pacific slope of the mountains from about 1200 feet in the coastal sagebrush belt up to at least 4500 feet in the chaparral, and are the only lagomorphs found commonly above the lower edge of the chaparral association. Here they were often on steep slopes beneath extensive and nearly impenetrable tracts of chaparral.

The ecologic niche of the brush rabbit is in brush where the plants form continuous thickets with little open ground. In the coastal sagebrush flats, areas supporting only scattered bushes are uninhabited by brush rabbits, while areas grown to extensive tracts of brush harbor them. When the brush rabbit's mode of escape from its enemies is considered, the reason for their habitat preference becomes more clear. Almost invariably these rabbits seek escape by running through the densest portions of the brush, never appearing in the open; in this way they travel quickly away from the source of danger without being observed. Because they avoid being seen in the open, and do not seek safety largely through running ability, they need continuous stretches of brush for escape. While hunting in the coastal sagebrush belt I have repeatedly seen frightened brush rabbits turn and dart beneath the bushes a few feet from a human being rather than be driven into the open.

A great horned owl shot in March, 1951, in the sage belt, had in its stomach the remains of a freshly killed adult brush rabbit. Although coyotes and brush rabbits often occur in the same general sections of the sage flats, remains of these rabbits have been notably scarce in coyote feces from these areas. This is probably because the coyote hunts along clearings and in open brushland, precisely the type of habitat avoided by brush rabbits.

Family SCIURIDAE

Sciurus griseus anthonyi Mearns

Western Gray Squirrel

Gray squirrels were on both slopes of the San Gabriels in oak woodland. A gray squirrel was observed in April of 1948, as it climbed a telephone pole adjacent to an orange grove near Cucamonga. This, and one noted bounding up a slope of greasewood chaparral near Cattle Canyon, were the only gray squirrels seen in areas which were not grown to oaks or adjacent to oak woodland. In the lower foothills gray squirrels were invariably found in association with valley oak, this plant forming limited woodland areas in canyon bottoms. In the upper chaparral association the squirrels frequented the large scrub oaks growing on talus slopes and canyon sides. In the yellow pine woodland, gray squirrels are restricted to black oaks, often where they formed mixed stands with the conifers. On the interior slope these squirrels were found only at the lower edge of the yellow pine woodland where black oaks are common. There, in the vicinity of Big Pines, they were present between roughly 5800 and 7000 feet, while on the Pacific slope they inhabited oak woodland from 1600 feet to about 7000 feet elevation.

In Live Oak Canyon in December of 1950, tracks indicated that a bobcat had killed a gray squirrel in a small draw beneath the oaks. In Evey Canyon on March 6, 1951, while watching for bats at late twilight, I observed a gray squirrel traveling through the branches of a nearby oak. A great horned owl glided into the oak in an attempt to catch the squirrel, which leaped quickly into a dense mass of foliage and escaped. For roughly ten minutes the owl perched in the oak watching its intended prey, then flew off down the canyon amid frantic scolding by the squirrel.

On March 17, 1951, a female gray squirrel taken at about 3500 feet elevation in San Antonio Canyon contained two embryos, each roughly 40 millimeters long.

Spermophilus beecheyi beecheyi (Richardson)

Beechey Ground Squirrel

From the coastal sage belt, into the yellow pine forest of the Pacific slope, this species is common on land cleared by man or disturbed in the course of construction, or on severely eroded slopes where the original climax vegetation is partly or completely absent. Thus in the sage belt, ground squirrels live along dirt roads through the brush, on the heavily eroded banks often found in the foothills, on land grazed closely by sheep, and in those parts of major washes such as San Antonio and Cucamonga washes where scatterings of huge boulders offer prominent vantage points. In San Antonio Canyon Spermophilus was restricted to the vicinity of roads and firebreaks, and an especially large colony of at least forty individuals lived at a dump one mile southwest of Camp Baldy at about 4500 feet elevation. Ground squirrels used burned stems of large laurel sumac as observation posts. Because of a preference for open areas offering unobstructed outlooks, ground squirrels originally probably did not penetrate the main belt of heavy chaparral on the Pacific slope of the range except in some of the large washes.

In the spring of 1951 and the preceding summer there was a marked increase in the ground squirrel population near Padua Hills as a result of sheep grazing on approximately one-half square mile of sage land. Grasses and smaller shrubs were eaten down to the ground, and in some places coastal sagebrush and Haplopappus were killed by browsing and trampling. The area formerly had a sparse growth of bushes with intervening growths of tall grasses and one colony of perhaps 20 ground squirrels; but after the sheep grazing the area was open brushland with large clear spaces on which the herbage was trimmed to the ground, and had at least four colonies of ground squirrels as large as the first. Also there were other ground squirrels established in various parts of the area. Probably the dry weather in the winter of 1950-51 with consequent retardation of the vegetation aided the spread of the squirrels in this area.

In the sage belt, most ground squirrels are dormant by December. In 1951, after a mild winter, squirrels were noted on January 25 near Padua Hills. On February 8, 1951, males in breeding condition were collected, and on March 16, a female taken near San Antonio Wash carried three small embryos. In early March of 1951, ground squirrels were active at 4500 feet elevation in San Antonio Canyon.

Specimen examined.—Los Angeles County: 1 mi. S and 2 mi. E Big Pines, 8000 ft., 1.

Spermophilus beecheyi fisheri (Merriam)

California Ground Squirrel

This ground squirrel inhabited the desert slope of the mountains up to 5000 feet elevation, and was most common in the juniper belt; burrows often were made under large junipers. In May, 1949, ground squirrels were common in the rocks adjacent to Mescal Wash at an elevation of 4500 feet. In an apple orchard near Valyermo, squirrels fed on the fallen fruit in early November of 1951.

No squirrel was seen in December, January, and February, indicating that all were below ground in winter.

Specimen examined.—San Bernardino County: Desert Springs, 4000 ft., 1 (PC).

Ammospermophilus leucurus leucurus (Merriam)

Antelope Ground Squirrel

Antelope ground squirrels were common in the Joshua tree woodland where they were noted up to 4500 feet elevation in Graham Canyon. None was found on the pinyon slopes, possibly because of the competition offered there by Eutamias merriami, or because the rocky nature of the soil there rendered burrowing difficult.

Although observed less often in winter than in summer, this species is active all year. On February 6, 1949, in Mescal Wash, an antelope ground squirrel was foraging over the snow which was at least six inches deep. These squirrels were attracted to the carcasses of rodents used as bait for carnivore sets, and caused a good deal of trouble by disturbing the traps.

Antelope ground squirrels used the topmost twigs of box-thorn bushes extensively as lookout posts, and many of their burrows were at the bases of these thorny bushes. This habit of regularly using observation posts is well developed in each species of ground squirrel found in the San Gabriels.

Specimens examined.—Los Angeles County: 6 mi. E and 1 mi. S Llano, 3500 ft., 2.

This chipmunk was characteristic of the most boreal parts of the San Gabriel Mountains. It was recorded from 6800 feet elevation at Big Pines, to an altitude of approximately 9800 feet near Mt. San Antonio, and was common where coniferous timber was interspersed with snowbrush chaparral. In upper Icehouse Canyon and near Telegraph Peak these chipmunks were associated with lodgepole pines and chinquapin, and one mile east of Mt. San Antonio individuals were often observed in thickets of manzanita. This chipmunk usually shunned pure stands of coniferous timber except as temporary forage ground.

On Blue Ridge these chipmunks used the uppermost stems of snowbrush as vantage points, and when disturbed ran nimbly over thorny surfaces of the brush in seeking refuge in the tangled growth.

In early November of 1951, these animals were not yet in hibernation on Blue Ridge. They were noted on November 6, after the season's first snows had melted; on November 13, however, a cold wind with drifting fog kept most of them under cover, and only two were noted in the course of the day.

The lower limit of the range of this species, on the coastal face of the range, is roughly coincident with that of manzanita—that is to say, it begins in the main belt of chaparral above the lower foothills. E. merriami seems to reach maximum abundance amid the granite talus, and scrub oak and Pseudotsuga growth at the upper edge of the chaparral association. It was absent, however, from all but the lower fringe of the yellow pine forest association.

On the desert slope merriami was partial to rocky areas in the pinyon-juniper association but was also in the black oak woods on the Ball Flat fire road near Jackson Lake. Nowhere was Eutamias merriami and E. speciosus observed on common ground.

No specimens of this species were taken in the field work in the San Gabriels, nor did I find any rangers or residents of the mountains who had seen flying squirrels in the area. Nevertheless sign found in the white fir forests in the Big Pines area indicated that flying squirrels may occur there. On a number of occasions dissected pine cones were noted on the horizontal limbs and bent trunks of white firs. These cones were too large to have been carried there by chipmunks, and gray squirrels were often completely absent from the areas. I suspect that extensive trapping in the coniferous forests of the higher parts of the mountains would produce specimens of flying squirrels. Willett (1944:19) mentions that flying squirrels probably occur in the San Gabriel Mountains.

This gopher was found below about 5000 feet elevation in disturbed or open areas from Cajon Wash at Devore westward all along the coastal base of the San Gabriel Range. In the lower part of the chaparral belt the gopher evidently was absent from the chaparral-covered slopes, but was common along roads and on fire trails.

Burt (1932) and von Bloeker (1932) discuss the distribution of the three subspecies of this species, pallescens, neglecta, and mohavensis, which are in the San Gabriel Mountains area, and Burt indicates that pallescens grades toward mohavensis in the southern part of Antelope Valley.

In the forests of yellow pine and white fir of the higher parts of the San Gabriel Mountains the workings of this gopher were common, and sign of its presence was found above 4500 feet on both slopes of the mountain range. The rocky character of the coastal slope seems to limit the occurrence of gophers, for they are not continuously distributed there. On the desert slope they occur locally down into the pinyon-juniper belt.

In the vicinity of Big Pines, on the interior slope, these gophers preferred broken forest where snow brush or other brush occurred; their workings, however, were also found beneath groves of conifers and black oaks. The abundance of earth cores resting on the duff indicated that this species is active in the snow in winter.

One specimen of this subspecies was taken on December 31, 1951, in the Joshua tree belt, eight miles east of Llano, 3700 feet elevation.

This pocket mouse is restricted to the coastal sage scrub association, and was recorded from Cajon Wash west to Live Oak Canyon. The mouse does not inhabit even the lower edge of the chaparral belt, but in the coastal sage flats is usually the most abundant rodent. In disturbed parts of the coastal sage belt fallax is less common, and was never trapped in channels of rocky washes. Trap lines in the eroded adobe banks of the foothills, where white sage and coastal sagebrush are the dominant plants, took mostly these pocket mice. Although the soil of such slopes is compact and seemingly is unsuitable for burrowing by heteromyids, fallax is the most common rodent. Because few burrows of pocket mice were noted there, it is possible that the many old unused burrows of Spermophilus and Dipodomys which honeycomb certain parts of adobe banks are used also by fallax; some of these burrows shelter Peromyscus eremicus and Peromyscus californicus.

These mice are inactive above ground in cold weather. In the sage belt near Thompson Canyon, where this subspecies had been found to be the most common rodent, none was trapped on the sub-freezing night of December 3, 1948, although other rodents were found in usual numbers. Individuals have been taken on nights of intermittent rain, yet none has been trapped on freezing nights.

This species is characteristically heavily infested by a large species of mite. Usually these mites congregate around the base of the tail.

On October 11, 1949, one lactating female and two carrying embryos were taken.

On the desert slope of the mountains this species is found in the part of the pinyon-juniper association that is between elevations of 4000 and 5200 feet. The mouse is absent from the higher chaparral and pinyon-covered slopes, but is present on south slopes in the pinyon belt where more open growths of pinyons and scrub oaks are interspersed with yucca. I recorded this pocket mouse from the vicinity of Cajon Pass west to Valyermo.

The local distribution of pallidus is striking because of its close positive correlation with the distribution of yucca. On benches around 5000 feet, where yuccas are scattered in their occurrence, pallidus is nearly always taken near (often right at the base of) this plant. Lower in the juniper belt the dry rocky south slopes supporting yucca plants are well populated by pallidus, while adjacent flats, and north slopes grown to antelope brush and scrub oak, are completely uninhabited. Near the mouth of Grandview Canyon, on steep rocky southern exposures grown sparsely to burro weed and yucca, one hundred traps produced in one night eight pallidus and no other rodents. Here many of these pocket mice were trapped on large fractured rock outcroppings, where most or all of the mice probably lived in the daytime in the deep cracks; in any event no burrows were noted near these rocks.

This species prefers barren slopes supporting yucca plants. These plants produce large seeds which are staple food items for P. f. pallidus and other rodents during the lean part of the year, that is to say, late summer and autumn. Many of the dry capsules of the yucca plants were examined in October, 1951, and these generally still contained a few seeds. Pocket mice taken in October usually carried in their cheek pouches seeds of yucca together with some other material, and often they carried only the seeds of yucca. Probably the wind shakes only a few seeds out of the capsules at a time, thus tending to drop the seeds over a fairly long period.

Trapping in winter in the juniper belt revealed that these pocket mice were not active above ground on nights colder than about 40° F. On nights when the temperature was about 36° F. none was taken, but on the one night in late December, 1948, when the minimum was 44° F., several specimens were taken. In this same area in May 1949, pocket mice were the most numerous rodents. Because of their evident sensitivity to cold weather, these mice must remain below ground for weeks at a time during the cold weather of December and January.

Specimens of pallidus from the desert slope of the San Gabriels are grayer (less brown) than specimens taken farther southeast in the Mojave and Colorado deserts. Further sampling of populations of Perognathus fallax from areas adjacent to the San Gabriels might demonstrate differences of sufficient magnitude to warrant subspecific distinction of the San Gabriel population. Possibly, however, the San Gabriel series manifests only local variation in the race pallidus. Grinnell (1933:54) characterizes the ecological niche of the race pallidus as being "open, sandy ground, often ... surrounded by rocky slopes," whereas these pocket mice in the San Gabriels inhabited gravelly or rocky juniper-dotted benches.

Mice of this subspecies were recorded from the lower chaparral association below about 4000 feet elevation along the coastal face of the San Gabriel Range. They were trapped on greasewood-covered slopes, in mixed growths of white sage and buckwheat, and beneath scrub oak and lilac chaparral; however none was taken in the heavy chaparral of the upper parts of the chaparral association.

One small juvenile in gray pelage was taken in San Antonio Canyon on October 1, 1951.

On Blue Ridge these mice were recorded between 7100 and 8000 feet elevation. Here they were restricted to dense tracts of snowbrush and sagebrush, often where these tracts were interspersed with, or beneath, open groves of conifers. These mice seemed to favor areas where this thick brush was broken by patches of open, grass-covered ground. Benson (1930:450) records this subspecies from Swarthout Valley, near Big Pines, at 6860 feet elevation.

While setting traps for pocket gophers one mile southwest of Big Pines, in September of 1951, I frightened a pocket mouse from its burrow. The animal jumped into the tangle of interlacing twigs of a nearby clump of snowbrush, and with great dexterity climbed into the center of the bush, where it was lost to view. I was surprised at the facility with which this saltatorial rodent traveled through the network of small branches.

In winter, in areas inhabited by this mouse, snow covers the ground for long periods during which these mice are probably forced to remain below ground.

This rat is common in the Joshua tree and juniper belts, and locally penetrates the pinyon belt at about 5000 feet elevation. It occurs regularly along the entire desert slope of the San Gabriel Mountains.

The upper limit of the range of this species roughly coincides with the upper limit of the juniper belt, and within this range it was found to inhabit areas having widely different soil types. It occurred on the sandy ground of desert washes, the gravelly soil of the juniper-clad benches, and the mixed sandy and rocky ground of washes in canyons. A preference is shown by panamintinus for fairly level ground. Rough terrain or steep slopes are generally avoided, whereas rather large colonies of these kangaroo rats are found in small flats of the desert foothills.

Below about 4500 elevation on the interior slope this species was the most numerous rodent, and seemed to reach maximum abundance in the Joshua tree association. About 500 trap-nights in the juniper belt near Graham Canyon yielded 31 specimens, whereas about 300 trap-nights in Joshua tree flats took 34 individuals.

The cheek pouches of many specimens taken in early winter contained green shoots of grass and little dry material. On many occasions rat traps set next to wood rat nests beneath large junipers produced panamintinus, and many of these animals had their cheek pouches crammed full of juniper berries.

In December, 1948, panamintinus was trapped consistently on nights when the temperature dropped to below 20° F. On December 27, 1948, after a three inch snowfall, tracks of this species were noted in the snow at the mouth of Mescal Canyon.

Parts of the skulls of this species were found in many coyote feces from the desert slope.

This kangaroo rat barely enters the area under consideration and is almost restricted to the Joshua tree association, for only a few individuals were taken at the lower edge of the juniper benches. This species inhabits the Joshua tree belt all along the desert base of the San Gabriels.

As mentioned in the description of the Joshua tree association, the relative numbers of Dipodomys merriami and D. panamintinus shifted from 1948 to 1951, possibly concurrent with the seasons of low rainfall in this period. Whereas in 1948 merriami was decidedly less abundant than panamintinus in the Joshua tree belt, in 1951 the numbers were reversed.

In December, 1951, it was found by tending the traps in the early evening that merriami foraged fairly early before the ground had frozen solidly.

One specimen of this subspecies was trapped on November 26, 1951, in a sandy channel of Cajon Wash near Devore beneath a clump of scale-broom.

This species was found below about 4000 feet elevation all along the coastal face of the range and reached maximum abundance in the level tracts of coastal sage. It was one of the most abundant rodents there, usually being second to Perognathus fallax in point of numbers. Large colonies of kangaroo rats occurred locally on sandy ground adjacent to large washes. The rats were found sparingly on the foothill adobe banks and in the greasewood chaparral of the lower foothills, but in heavy chaparral where a layer of plant debris covered the ground, such as on north slopes grown to scrub oak and lilac, kangaroo rats were completely absent. Thus, in the lower chaparral belt, this rodent had a discontinuous distribution.

The coyote probably is one of the major predators of these kangaroo rats; remains of this rodent were often found in coyote feces, and coyotes excavated many burrow systems in large kangaroo rat colonies in the sandy ground near San Antonio Wash. The soil there is so soft that coyotes probably were often successful in digging out their prey. The shed skin of a large Pacific rattlesnake (Crotalus viridis helleri) was found four feet inside the mouth of a kangaroo rat burrow; probably this reptile preys on agilis. Great horned owls (Bubo virginianus pacificus) come down nightly from the chaparral to hunt in the sage flats. Beneath the perches of these owls I have found pellets containing bones of agilis.

All the specimens of this species from the desert slope of the San Gabriel Range are referred to the subspecies perplexus. They were taken in brushy habitats between the elevations of 4500 and 7400 feet. Throughout much of this area perplexus was found only in certain restricted areas more or less surrounded by inhospitable ground. For example, at 7400 feet on Blue Ridge, they were found occasionally in the strips of sagebrush and lilac brush which locally capped this ridge. Often these patches of chaparral on Blue Ridge were surrounded by areas unsuitable for kangaroo rats: on the Pacific slope, talus, oaks, and yellow pines prevailed; on the ridge scattered yellow pine groves were present; and on the steep desert slope there were yellow pines and white firs. In Swarthout Valley perplexus was found in flats that supported basin sagebrush and Haploppus, while the coniferous forests to the south, and pinyon-covered slopes to the north were uninhabited. On flats supporting antelope brush and juniper, perplexus was often common, but it did not penetrate the chaparral of adjacent slopes grown to scrub oak and mountain-mahogany. In general then, perplexus was found in fairly open brushy flats or slopes, even where these were surrounded by unsuitable habitats.

Specimens of D. agilis from the desert slope two miles east of Valyermo are referrable to the subspecies perplexus. A series taken in Cajon Wash at Devore, on the Pacific slope, is intermediate between agilis, of the coastal slope of the San Gabriels, and perplexus of the desert slope, but approaches more nearly the later subspecies. Thus, different subspecies of D. agilis occur on opposite slopes of the San Gabriel Mountains, with intergradation taking place in the Cajon Pass area and probably also at the west end of the Mountains.

Both scrub oak acorns and juniper berries were found in the cheek pouches of this subspecies, and one immature individual taken in Swarthout Valley had its cheek pouches stuffed with approximately 550 seeds of brome grass.

On November 13, 1951, at 7500 feet on Blue Ridge, a small juvenile was taken; it must have been born not earlier than September.

This species inhabited grassy areas of the coastal sage belt, and reached maximum abundance on cleared land grown thickly to weeds and scattered brush. The mouse was only locally abundant—being scarce throughout much of the sage belt—but was found under contrasting conditions. In San Antonio Wash the species was taken among rocks and sparse weeds, at Palmer Canyon specimens were trapped on a barren ridge sparsely clothed with greasewood and white sage, and also one mile E of Big Pines in flats supporting basin sagebrush and a fairly dense growth of grasses. The western harvest mouse was recorded from 1500 feet elevation to 3200 feet on the Pacific slope, and at 6600 feet near Big Pines on the desert slope.

Those specimens of harvest mice from near Big Pines may be grading toward the desert race megalotis; my series of specimens from this locality, however, is too small for clear indications on this point.

Individuals in juvenal pelage were taken on November 26, 1951, near Devore.

In Mescal Wash on the desert slope of the San Gabriels, this mouse was one of the most abundant mammals and was the only rodent other than Peromyscus maniculatus regularly trapped in the barren channels of washes. In Mescal Wash, at an altitude of 4000 feet, eremicus occurred along with the chaparral-inhabiting Peromyscus boylii and Peromyscus californicus. The two species last mentioned were associated with the occasional large patches of manzanita, antelope brush, and other brush of the wash, whereas eremicus was trapped in the rocky and sandy channels among scattered bushes of scale-broom. No specimens of eremicus were taken on the juniper-clad benches adjacent to the wash.

This mouse was recorded from 1900 feet elevation, one mile south of the mouth of San Antonio Canyon, to 3200 feet elevation in Cajon Canyon. This subspecies is characteristic of the sage belt and shows a strong preference for the rough rocky areas found in dry washes. Although in many areas the channels of the washes are immediately adjacent to sandy sagebrush-covered flats, eremicus is not common in the latter areas. Rocks seem to be essential to eremicus, for sandy areas in the sageland which were devoid of rocks yielded only an occasional specimen. For example, 100 trap-nights in the main channel of San Antonio Wash yielded 23 eremicus and only six other rodents; while in the sandy sage areas nearby 200 trap-nights yielded only one eremicus and 32 other rodents.

In lower San Antonio Canyon eremicus seemed restricted to the rocky canyon bottom, none having been trapped on the steep slopes nearby. This subspecies occurs commonly, however, on the adobe banks grown to white sage at the base of the foothills. There eremicus occurred on common ground with Perognathus fallax fallax, and was often the only Peromyscus taken.

This species may be restricted by temperature; washes above 4000 feet elevation, which seemed suitable were uninhabited by these mice.

On December 1, 1949, two females taken at the mouth of Palmer Canyon had well advanced embryos. A female trapped in San Antonio Canyon on September 19, 1951, was lactating. Juveniles were caught in the sage belt in October, 1951.

This mouse inhabits areas supporting chaparral on the coastal slope of the San Gabriels below 5000 feet. In the chaparral it is usually the most plentiful rodent, being dominant on slopes which have been burned over and on which greasewood chaparral has taken over. On one such slope at the head of Cow Canyon, at 4500 feet, this was the only rodent trapped, although an occasional wood rat house was noted. Trapping records gave the impression that this form was the most ubiquitous rodent in the entire chaparral belt. Nearly every trap line, even in such non-productive areas as oak woodland, took the California mouse; and in many areas, as in thick lilac brush, this mouse was by far the most abundant rodent. Specimens were taken on the damp ground next to San Antonio Creek, and in the riparian growth. In San Antonio Wash the California mouse was found in thickets of laurel sumac and lemonade berry, or other large shrubs, but were absent from most of the adjacent sageland. The one place where they were found away from heavy brush was on a series of barren adobe banks, near Palmer Canyon, clothed mostly with white sage. Here they found shelter in the unused burrows of kangaroo rats and ground squirrels.

The only place on the desert slope where this species was taken was in Mescal Wash. There it was taken occasionally near the large clumps of antelope-brush and manzanita which grew in the main channels of the wash.

Lactating females of this species were taken in October, 1949, and February, 1950. Two pregnant females were trapped on February 25, 1950, at the mouth of Palmer Canyon.

This species occurs from 1000 feet elevation to above 9000 feet elevation on the Pacific slope of the Mountains, but although probably the most widespread rodent in the area it is absent from many habitats. This mouse reaches maximum abundance in the coastal sage scrub association, particularly where the soil is sandy with scattered vegetation—usually coastal sagebrush and black sage. On the foothill adobe slopes none was trapped, nor have any been taken in most of the chaparral habitats. A few gambeli were trapped amid the talus beneath growths of scrub oak and bay trees in San Antonio Canyon, at 4300 feet elevation. On Blue Ridge, at elevations of from 7200 feet to 8300 feet, this mouse inhabited areas clothed with snowbush, basin sagebrush, currant, and scattered conifers, and was found sparingly in the coniferous forests. Thus this species lives on contrasting soil types in association with many different vegetational assemblages, from the coastal base to the crest of the range.

There is a rather wide variation in color in gambeli from the San Gabriels. Certain individuals taken in open, sandy coastal sage areas are pale, some being indistinguishable from examples of sonoriensis taken in the pinyon-juniper association on the desert slope. Specimens from San Antonio Canyon have somewhat darker pelage than those from the sage belt, and than individuals taken on Blue Ridge. Possibly a large series of Peromyscus maniculatus from the San Gabriel Mountains would show definite local trends in color of pelage.

This species is active on sub-freezing and rainy nights as evidenced by trapping results, and at Big Pines there were tracks around the bases of conifers after a heavy snowfall in December, 1951. Several females taken in the sage belt in October, 1948, carried embryos, and a lactating female was recorded from Blue Ridge on November 13, 1951. Juveniles have been taken in September, October, November, and December.

This subspecies is associated with contrasting types of soil and vegetation. It is seemingly absent from the upper pinyon-juniper sage flats and areas grown to chaparral, but is fairly common on the gravelly benches dotted with junipers, and in the washes issuing from the canyons on the desert slope. It is present in small numbers in the Joshua tree association.

In 1951 the numbers of sonoriensis were noticeably less than in 1948; probably this was correlated with the series of dry winters in this period. In December, 1948, this animal was one of the most common rodents in Mescal Wash, 200 trap-nights yielding thirteen specimens; but in November, 1951, none was taken. In parts of the juniper belt, where an average of about six sonoriensis was taken per 100 trap-nights in 1948, the average had dropped to one per 100 trap-nights in 1951.

Specimens of this species from the desert slope of the mountains have been assigned to the subspecies sonoriensis. Those from Blue Ridge tend toward sonoriensis in color, and may be considered as intergrades between this subspecies and gambeli.

This species was active on nights when the temperature was as low as 10° F., and individuals were trapped in the juniper belt in December, 1948, when four inches of snow lay on the ground.

Gray-pelaged juveniles were taken on the desert slope in December, 1948, and a female taken in Mescal Canyon on December 22 of this year carried four embryos near term.

The main range of this mouse in the San Gabriel Mountains lies between 1600 and 6000 feet elevation on the Pacific slope of the Mountains, thus encompassing much of the chaparral and oak woodland associations. It was the most common mammal in the oak woodland association in the lower foothills and often was trapped there on leaf mold beneath the oaks. While trapping for shrews I regularly took this species in riparian growth right down to the edge of the water. In San Antonio Canyon many boylii were trapped beneath logs and dense vegetation, and on wet seepage slopes adjacent to the creek.

This species shows a definite predilection for rocky habitats where these occur in the chaparral. In heavy lilac brush near Camp Baldy Peromyscus boylii was outnumbered by P. californicus, yet where talus slopes or boulder piles occurred boylii was more numerous. At the head of Cow Canyon amid boulders beneath scrub oak, bay, and big cone-spruce, this species was especially abundant and no other Peromyscus was taken.

Of special interest is the occurrence of this mouse on the desert slope of the mountains; there it was taken beneath scrub oaks in the pinyon-juniper association at the mouth of Mescal Canyon, and amid boulder and debris piles in Mescal Wash at 4000 feet elevation. While manzanita and scrub oak grew in the wash at the points of capture, the animals were actually surrounded by the desert conditions of the Joshua woodland, and associated with such desert forms as Onychomys torridus pulcher and Peromyscus eremicus eremicus.

Immature individuals were taken in October, November, February, and March, and a female with two large embryos was taken near Icehouse Canyon on November 8, 1951.

Only once was this mouse found outside the pinyon-juniper association of the desert slope; in November, 1949, several were collected near Cajon in mixed manzanita, scrub oak, and greasewood chaparral. This was the only Peromyscus of regular occurrence in the pinyon-juniper area, and was recorded from the upper limit of this association, near Jackson Lake, at 6000 feet, to the lower limit of the association at the mouth of Graham Canyon at roughly 4000 feet elevation.

Although in the juniper belt truei often occurs on common ground with Peromyscus maniculatus sonoriensis, the habitat preferences of these animals are generally complementary. Where the mice occur together, traps set in a variety of locations caught Peromyscus maniculatus, but typically traps set amid the brush or on the open ground away from the junipers were productive. On the contrary truei was invariably trapped quite near the junipers and often in association with the large nests of Neotoma fuscipes simplex. In fact traps set right on the beds of litter beneath the junipers were most likely to catch truei. Records kept of trapping localities show that truei was without exception trapped within twenty feet of some treelike shelter such as junipers, pinyons, Joshua tree or scrub oaks. Thus Peromyscus maniculatus occupies the open stretches between the trees, while truei inhabits the ground beneath and immediately adjacent to the trees. In Nevada the piÑon mouse prefers rocky areas (Hall, 1946:520). In the San Gabriel Mountains this mouse does not seem to have this predilection.

In the juniper belt truei was second to Dipodomys panamintinus in point of numbers. In the course of 500 trap-nights in the juniper belt twenty-two truei were taken with thirty-six Dipodomys.

I consider my series of Peromyscus truei from the desert slope of the San Gabriels to represent the subspecies montipinoris. The series is closely comparable to specimens of the subspecies montipinoris in the California Museum of Vertebrate Zoology from the Mount Pinos area, but differs from specimens of the race chlorus from the San Bernardino Mountains in certain diagnostic characteristics. In his recent paper on Peromyscus truei, Hoffmeister (1951) considered the populations of this species in the San Gabriels to be of the race chlorus. Hoffmeister had only one specimen available from the San Gabriel Mountains (Lytle Creek, on the Pacific slope) which was intermediate between montipinoris and chlorus, but on the basis of cranial measurements it was referred to the race chlorus. Specimens of Peromyscus truei from the eastern end of the desert slope of the San Gabriel Mountains and the Cajon Pass area would probably demonstrate that the race montipinoris, which occupies the desert slope of the San Gabriels, intergrades with the race chlorus, which occurs in the San Bernardino Range immediately to the east, in the Cajon Pass area. Although montipinoris occurs on the desert slope of the San Gabriels, chlorus may occur on the Pacific slope. I took no specimens of the piÑon mouse on the Pacific slope of the San Gabriel Mountains.

In December, 1948, many small juveniles were taken in the juniper belt, and on October 15, 1951, two females trapped at the head of Grandview Canyon had embryos: one three and the other four. On November 13, 1951, a partially gray-pelaged subadult female was trapped which had recently suckled young.

Grasshopper mice seemed to be partial to the more sandy parts of the Joshua tree flats where the mice were trapped regularly but not abundantly. This mouse inhabited the barren sandy channels of Mescal Wash but was rare on the adjacent juniper-clad benches. In the arid, sandy washes this typical desert rodent penetrated the high pinyon-juniper association.

Wherever grasshopper mice occurred they were outnumbered by most of the other rodent species. For example, on November 26, 1949, below Graham Canyon, 100 snap traps yielded 10 Dipodomys panamintinus mohavensis, 2 Dipodomys merriami merriami, 4 Peromyscus maniculatus sonoriensis, and 3 Onychomys torridus pulcher.

Where abandoned kangaroo rat burrows were common in the Joshua tree belt these burrows were used as retreats by Onychomys. Some traps set at the entrances to old burrows caught grasshopper mice.

This species was on the Pacific face of the Mountains from 1600 feet elevation in the coastal sage belt, to 4800 feet elevation in open groves of big cone-spruce and scrub oak of the chaparral association.

The local distribution of this woodrat is determined by suitable nesting sites. Although taken in different types of vegetation, lepida, without exception, was associated with rocky areas or areas supporting patches of prickly-pear cactus. In the channels of San Antonio Wash, lepida was commonly associated with jumbles of boulders and boulder-dotted cut banks. There the vegetation is sparse, and the rats dwell among the rocks; only their droppings and faint trails indicate their presence. Among boulders lepida builds only small houses of sticks and debris, and even these only occasionally. The effect of the prickly-pear cactus on the distribution of lepida in the sageland is striking; trap lines there yielded no woodrats where extensive rock piles and patches of prickly-pear were absent, but many rats were taken where patches of prickly-pear are plentiful. On an acre supporting coastal sagebrush at the mouth of San Antonio Canyon, at 1800 feet elevation, there were fourteen patches of prickly-pear, each covering at least thirty square feet. In these patches there were thirteen occupied woodrat nests. Only one patch lacked an occupied nest, and this one contained the remains of an old nest. On this acre there were at least thirteen individuals. In the sagebrush belt only an occasional large patch of cactus lacks a woodrat house occupied by lepida. Seemingly Neotoma fuscipes does not build houses in patches of prickly-pear.

Most of the houses built by Neotoma lepida are small and simple as compared to those of Neotoma fuscipes, and often in rocky areas no nests are in evidence. The most elaborate nests are built among the pads and spines of the prickly-pear and under laurel sumac or other large shrubs growing near washes. One of three houses examined at the mouth of San Antonio Canyon was on sandy ground in a patch of Opuntia measuring approximately 11 x 14 feet. The house was 14 inches high and 41 x 37 inches at the base. It was built around the main stem of the prickly-pear and a rock about 10 inches in diameter. The house was constructed of sticks of coastal sagebrush and buckwheat, and was dotted with dissected fruits and flowers of the prickly-pear. The main chamber was arched over by the main stem of the prickly-pear and was roughly 12 x 19 inches, inside dimensions, being reached through two three-inch openings, one on the east side of the chamber and one on the north side of the chamber. Two cup-shaped nests were inside the chamber, these being constructed mostly of grasses, and each resembling a well constructed bird nest 4 inches in diameter. The grass nests were free of feces, but feces were piled up against the west side of the chamber with many snail shells and dissected fruits and flowers of prickly-pear. Thirty-five inches from the main chamber was a third grass nest on the ground beneath a cluster of cactus pads. Next to this there was a blind burrow about eight inches long, and one and three-quarters inches in diameter. No burrow led to the main chamber, in this or in either of the other houses, but all had at least one short blind burrow beneath the house.

At many houses there were one to three grass nests outside the house on the ground, within four feet of the house. From each nest a well worn path lead to the house. Traps set in these nests invariably caught woodrats.

The many prickly-pear fruits and snail shells in and around the houses of lepida probably were remnants of food. So many of the rodents caught in traps near woodrat nests were partly eaten—usually the brains were taken—that I suspect the woodrats of eating their relatives. The heads of many composite annuals were piled near woodrat nests.

Immature individuals were taken in September, October, and early November, and on September 26, 1951, a lactating female was trapped near Palmer Canyon.

An old female bobcat trapped in Thompson Canyon had masses of cactus thorns beneath her skin, especially about the forelegs. These thorns were probably received while she was foraging in growths of prickly-pear for woodrats. The other bobcats from San Antonio Wash also had accumulations of thorns under the skin of the forelegs. Fragments of the skulls of Neotoma lepida were recovered from horned owl pellets and coyote feces.

Specimens examined.—Total, 7, distributed as follows: Los Angeles County: San Antonio Canyon, 4500 ft., 2; San Antonio Wash, 1800 ft., 5 (2 PC).