In evaluating boundaries I have relied most heavily on the information of Merriam (map 3) and Kroeber (map 1). Merriam's data are contained in a 1:500,000 map of California, together with a descriptive text. The map and the description were made up by Dr. Merriam's daughter, Mrs. Zenaida Merriam Talbot, during the years 1939 to 1946, from information in Merriam's notes and journals, the latter of which are not accessible to this writer. Often, where Merriam's boundaries disagree with those of Kroeber or other authors, Merriam's line will follow a stream, whereas the alternative follows a ridge or drainage diversion. When the evidence is inconclusive, I have usually followed Kroeber's method and chosen the ridge rather than the stream as the boundary. In this area the streams are small and easily crossed during most of the year and therefore would not constitute a barrier sufficient for the divergence of dialects. On the other hand, the hills were visited only briefly for hunting and gathering; the population depended to a great extent on the products of streams for its subsistence, and consequently all the permanent villages were in the lowlands and canyons. For this reason, the ridges rather than the streams would tend to be boundaries. Kroeber has discussed this point more generally (1939, p. 216) and also in greater detail (1925a, p. 160).

EXTERIOR BOUNDARIES

The southern boundary of the Athabascans begins at Usal Creek on the coast and goes eastward for a few miles before swinging south to include the drainages of Hollow Tree Creek and the South Fork of the Eel in Kato territory. It turns north to enclose the headwaters of South Fork and proceeds along the ridge dividing Ten Mile Creek from the main Eel until it reaches the drainage of Blue Rock Creek; it then passes around north of the creek and crosses the Eel near the mouth of the creek. From this point it runs in an easterly direction around the drainage of Hulls Creek.

Kroeber's map in the Handbook shows the southern boundary beginning a few miles south of Usal Creek, but Merriam and Nomland both maintain that the creek itself is the boundary and Gifford (1939, p. 304) says that both Sinkyone and Yuki were spoken in the village situated at the mouth of the creek. The information of all four authors came from either Sally or Tom Bell, wife and husband, who are respectively Shelter Cove Sinkyone and Coast Yuki. I have accepted Merriam's boundary, since it agrees with Nomland's.

Merriam maintains that the western boundary of the Kato runs along the South Fork of the Eel and he is partly supported in this by Barrett (1908, map), whose boundary includes the drainage of South Fork but not the drainage of Hollow Tree Creek. Barrett, however, disavows any certainty on this particular boundary. Kroeber's line, which does include the drainage of Hollow Tree Creek in Kato territory, is supported by a specific statement from Gifford (1939, p. 296) that "Hollow Tree Creek did not belong to the Coast Yuki although they fished there." I have therefore accepted Kroeber's version.

All authorities agree on the southern and eastern boundaries of the Kato as far north as the drainage of Blue Rock Creek. Merriam claims this drainage for the Wailaki, whereas both Kroeber and Foster claim it for the ta'no'm tribelet of the Yuki. It is evident that this territory was disputed, for it was the scene of several of the wars involving the Wailaki, the Kato, and the Yuki (Kroeber, 1925a, p. 165; 1925b). Kroeber obtained a detailed list of place names in this area from a ta'no'm Yuki, whereas Merriam's Wailaki information is only of a most general nature. For this reason I have given the territory to the Yuki.

All the authorities, except Foster, agree on the rest of the southern boundary of the Athabascans. Foster has the Yuki-Wailaki line cross Hulls Creek about five miles from its mouth instead of passing south of its drainage. Both Kroeber and Merriam favor the more southern line, and Goddard (1924, p. 224) says that the Wailaki claimed a fishing spot in the disputed area, so I have accepted this version.

The eastern boundary of the Athabascans runs north along the ridge separating the drainages of the North Fork and Middle Fork of the Eel until it reaches the headwaters of the Mad River. Thence it runs in a northern direction along the ridge that separates the drainage of the Mad River from that of the South Fork of the Trinity until it reaches Grouse Creek, where it turns eastward to cross the South Fork of the Trinity at the mouth of the creek. It continues north on the east side of South Fork, following the crest until it crosses the main Trinity about five miles above its confluence with South Fork, and then follows around the headwaters of Horse Linto Creek and Mill Creek.

Merriam's eastern Athabascan boundary conflicts with the one drawn by Kroeber, Foster, and Goddard in assigning the northern part of the drainage of the Middle Fork of the Eel to the Pitch Wailaki instead of to the Yuki. Merriam is almost certainly wrong here, for Goddard (1924) definitely does not include this area within Wailaki territory and his information in this region appears to have been especially reliable. Moreover, Merriam got his information from natives of the main Eel River, who were evidently not on good terms with their relatives to the east and knew little about them. I have therefore accepted the Kroeber boundary.

The next conflict is to the north of this, where Kroeber's boundary runs up the ridge separating the Mad River from the South Fork of the Trinity, whereas Merriam's runs along South Fork itself in the twenty miles from Yolla Bolly Mountain northwest to Ruth. Essene (1942) agrees with Merriam on this point, but his data add nothing to the argument, since he worked with the same Lassik informant as Merriam. I have accepted Kroeber's version because it is corroborated by both Goddard (1907) and Du Bois (1935, map 1), who agree in assigning the valley of the South Fork of the Trinity to the Wintun.

Kroeber and Merriam agree on the line running north of Ruth as far as a point about fifteen miles south of Grouse Creek, where Merriam's line drifts westward to follow the north-south channel of Grouse Creek for a short distance, whereas Kroeber's line follows due north along the drainage pattern. Essene supports Kroeber, but his informant did not come from this region so her testimony perhaps cannot be relied on heavily. I have accepted Kroeber's line because it follows the drainage pattern.

Kroeber's boundary also conflicts with Merriam's on the east side of South Fork. Kroeber's line runs along the ridge separating South Fork from the main Trinity whereas Merriam's runs along the Trinity itself. The testimony of Dixon on the Chimariko (1910, pp. 295-296) supports Kroeber, so I have accepted the latter's line.

The northern boundary of the Athabascans runs west, parallel to Mill Creek, crossing the Trinity a few miles south of its confluence with the Klamath, and then continues west until it reaches Bald Hills Ridge, which separates Redwood Creek drainage from Klamath River drainage. It continues north along this ridge and then turns east to cross Redwood Creek about ten miles southeast of Orick.

Goddard (1914a, pl. 38) indicates three Athabascan summer camps on the Yurok side of the dividing ridge. This may mean that some Athabascan territory was included in the Klamath drainage, but if so, it would contradict the testimony of the Yurok (Kroeber, 1925a, fig. 1; Waterman, 1920, map 2). However, the land away from the Klamath was little used by the Yurok (Kroeber, 1925a, p. 8), so it may be that this territory was claimed by both groups. I have accepted Kroeber's boundary here. Otherwise there are no conflicts on the northern boundary.

The western boundary of the Athabascans runs due south from Redwood Creek, following the 124th Meridian, crossing the North Fork of the Mad River at Blue Lake and crossing the main Mad River a few miles above the mouth of North Fork. From here the line follows south around the drainage of Humboldt Bay until it crosses the Eel River at the mouth of the Van Duzen, whence it runs south to Bear River Ridge, which it follows west to the ocean.

A major conflict in the western boundary of the Athabascans involves the drainage of the North Fork of the Mad River. Kroeber and Loud both assign this area to the Wiyot, whereas Merriam assigns it to the Athabascans. Neither Kroeber nor Loud gives specific data in support of his contention; thus Merriam's specific local information quoted below, renders his line preferable.

Sunday, August 11, 1918.... I found two old men of the same tribe, who were born and reared at the Blue Lake rancheria 'Ko-tin-net—the westernmost village of the Ha-whil-kut-ka tribe.

I have therefore accepted Merriam's boundary.

From the Mad River south to the Eel there is general agreement except that, as usual, Merriam's lines tend to follow the streams, whereas those of Kroeber and Loud follow the ridges. Another conflict comes at the crossing of the Eel River. Curtis (1924, 13:67) says the line crosses at the mouth of the Van Duzen. Nomland (1938, map 1), Loud, and Merriam all agree with this. Powers (1877, p. 101) and Kroeber both locate the line a few miles up the river from this point at Eagle Prairie, while Nomland's Wiyot informant (Nomland and Kroeber, 1936, map 1) places the line even farther south at the mouth of Larabee Creek. The weight of evidence indicates that the line was probably near the mouth of the Van Duzen; Goddard (1929, p. 292) states that there was a Bear River village near there.

There is also some disagreement on the northern boundary of the Bear River group. Nomland says that it is at Fleener Creek, about five miles north of Bear River Ridge, whereas Kroeber indicates a line about two miles north of Bear River Ridge. Loud, Merriam, and Goddard, on the other hand, all indicate that the boundary is Bear River Ridge itself. Nomland's boundary is almost certainly in error, since Loud gives Wiyot villages occurring south of that line. Most of the evidence points to Bear River Ridge as the line, and this version has been accepted.

INTERIOR BOUNDARIES

There is no disagreement on the western boundary of the Hupa. It runs north and south along Bald Hills Ridge, dividing the drainages of Redwood Creek and the Trinity River. Merriam gives the Hupa two divisions—the Tin-nung-hen-na-o, or Hupa proper, and the Ts´a?-nung-wha?, or Southern Hupa. The line dividing these two groups lies just north of the main Trinity to the east of South Fork and along Madden Creek to the west of South Fork. Kroeber (1925a, p. 129) and Goddard (1903a, p. 7) do not give any support for a linguistic division, as indicated by Merriam, but there does seem to have been some cultural difference.

In the division of the territory west of the Hupa Merriam differs radically from Kroeber and Goddard, although all three scholars divide the area between two groups. Kroeber and Goddard call the northernmost group Chilula, an anglicization of the Yurok word tsulu-la meaning "Bald Hills people," and the southern, Whilkut, from the Hupa word hoilkut-hoi meaning "Redwood Creek people" or "upper Redwood Creek people."

Merriam calls the first of his two divisions Hoilkut and says that they lived on Redwood Creek and on the North Fork of the Mad. This group he further subdivides into three parts: one, living on lower Redwood Creek, corresponds to the Chilula of Kroeber and Goddard; another, on upper Redwood Creek, corresponds to part of Kroeber's Whilkut; and a third, on the North Fork of the Mad River, corresponds to a part of Loud's Wiyot.

Merriam calls his second division Ma-we-nok. They live in the drainage of the main Mad River and correspond to a part of Kroeber's Whilkut.

It would appear that, except for Goddard's Chilula information (Goddard, 1914a), Merriam's data are the most detailed and therefore preferable. He had informants from lower Redwood Creek, from the North Fork of the Mad River, and from the main Mad River. For this reason I have accepted his boundaries. I therefore propose that all the peoples previously included under the terms Whilkut or Chilula be called Whilkut. This seems justified by Merriam's statements, on the one hand, that the Mad River Ma-we-nok differed but little in speach from their Whilkut neighbors, and, on the other hand, that the other groups in the area called themselves hoilkut or terms related to this.

If this proposal is accepted, the Whilkut may then be divided into four subgroups—the Chilula Whilkut, the Kloki Whilkut, the Mad River Whilkut, and the North Fork Whilkut. The Chilula Whilkut would occupy essentially the territory assigned to the Chilula by Goddard and Kroeber—the drainage of Redwood Creek from about ten miles southeast of Orick to about a mile above the mouth of Minor Creek. Above them are the Kloki Whilkut, occupying the upper drainage of Redwood Creek. The name Kloki Whilkut means "prairie" Whilkut, a name used by these people for themselves, according to Merriam, and derived from the prairies that occur on upper Redwood Creek. The Mad River Whilkut would be the group in the drainage of Mad River from the mouth of North Fork as far up as Bug Creek above Iaqua Buttes. The North Fork Whilkut would then be the group in the entire drainage of the North Fork of the Mad River.

The northern boundary of the Nongatl begins in the west near Kneeland at the Wiyot boundary and runs southeast around Iaqua Buttes and the drainage of the Mad River, then northeast to Grouse Creek. Kroeber and Merriam agree on this boundary east of Iaqua Buttes, but west of that landmark Merriam's line takes a northeast-southwest direction whereas Kroeber's line runs due east-west. I have accepted Merriam's line here because he has more detailed information than Kroeber on the neighboring Whilkut. Neither has much information on the Nongatl themselves.

One of the main interior lines of the Athabascans is the one which, running north and south along the South Fork of the Eel, divides the coastal groups on the west from the interior peoples to the east. It begins at the mouth of the Van Duzen on the main Eel and runs south along the Eel as far as Scotia, dividing the Nongatl from the Bear River group. At Scotia it coincides with the Sinkyone-Nongatl boundary and then continues in a southerly direction but, instead of lying immediately on the river, it drifts slightly to the east to include also the land adjacent to the stream. It continues thus near to, but off, the main Eel until it crosses the river at about McCann, a few miles above the mouth of South Fork. After crossing the main Eel, the line goes south, including the immediate river valley of the South Fork of the Eel in Sinkyone territory, until it turns west to cross South Fork at the mouth of Hollow Tree Creek, continuing to the coast at Usal Creek.

This section of the Athabascan boundary has been much disputed. It seems certain that the western side of the Eel from the mouth of the Van Duzen to Scotia was Bear River territory. This distribution is attested by Powers (1877, p. 107), who says that the Bear River group owned as far south as the mouth of South Fork, by Nomland's Bear River informant (1938, map 1), by Kroeber, and by Goddard, who says (1929, p. 291), "There was, however, one village at the mouth of Van Duzen creek which was allied to Bear River both in its dialect and politically." This evidence is fully in accordance with that of Merriam.

The eastern side of the river along this stretch goes to the Nongatl by default. Kroeber claims it for the Bear River people and Nomland's Wiyot informant claimed it for the Wiyot (Nomland and Kroeber, 1936, map 1) but except for these sources possession is denied by Wiyot, Bear River, and Sinkyone alike.

South of Scotia the area is also in dispute. Nomland and Kroeber claim that the eastern side of the Eel from Scotia to the mouth of South Fork is Nongatl. They say (1936, p. 40):

In any event, Eel river from Scotia to Larrabee was not Mattole, as Kroeber has it in map 1 of his Handbook, nor was it Sinkyone. Nomland's Bear River, Mattole, and Sinkyone informants were positive on the point. If Athabascan, the stretch in question belonged to the Nongatl (Saia). Otherwise it was Wiyot.

Merriam, on the contrary claims that this territory was definitely Sinkyone.

We must evaluate the statements of the informants involved before reaching a decision on this point. Nomland's Bear River informant was evidently not particularly accurate on boundaries, for she placed the northern boundary of the Bear River group at Fleener Creek when it was almost certainly at Bear River Ridge (see p. 163). Therefore her testimony may be questioned on the present point also. Nomland's Sinkyone informants were from the Shelter Cove Sinkyone of the Briceland area to the south, and furthermore only one of them was said to be reliable. Merriam, however, presents detailed evidence in the form of place names obtained from George Burt, a very good informant who was born and raised among the northern Sinkyone at Bull Creek. I have therefore accepted the evidence of George Burt via Merriam, even though several of Nomland's informants deny it.

Actually, I have accepted Merriam's line as far south as Phillipsville on the South Fork of the Eel, even though it conflicts somewhat with the lines of Nomland and Kroeber. Merriam's information for this stretch of South Fork is supported in detail by Goddard's village lists. South of Phillipsville, Merriam's line runs along South Fork itself instead of lying slightly east of it. This line is contradicted by Goddard, whose informant, a native of the region, gave Goddard village names on both sides of the river as far south as Garberville. I have accepted the line indicated by Goddard's information along this stretch.

South of Garberville I have relied heavily on Nomland. She had three informants from the Shelter Cove Sinkyone—Sally Bell, Tom Bell, and Jack Woodman, of whom she considered only the last reliable. Merriam seems to have relied entirely on Sally Bell for information about this group and his information should therefore be somewhat discounted.

The Bear River-Mattole boundary is not disputed. Merriam and Nomland agree that it begins on the coast at Davis Creek and then follows the ridge east to the headwaters of Bear River. The two authors do not agree on the Bear River-Sinkyone line. Nomland's boundary goes due east from Bear River headwaters to strike the South Fork of the Eel a few miles above its mouth. Merriam's line instead goes north to intercept the main Eel at Scotia. I have accepted Merriam's version on the basis of George Burt's evidence, even though Kroeber agrees with Nomland.

The Mattole-Sinkyone boundary begins at Spanish Flat on the coast and goes northeast from there, crossing the Mattole River just above the mouth of Upper North Fork, Mattole River, and continuing in that direction to the headwaters of the Bear River. I have altered Merriam's map on this point. It shows the Mattole-Sinkyone line reaching the coast at Big Flat, a point about six miles down the coast from Spanish Flat. Merriam's notes say, however, that the line ends at Spanish Flat. Merriam's line crosses the Mattole River near the town of Upper Mattole about five miles below the mouth of Upper North Fork, but Goddard's Mattole informant gave him villages as far up as the mouth of Upper North Fork and I have considered this fact to be decisive. Nomland's Mattole-Sinkyone line reaches the coast at Four Mile Creek, about five miles up the coast from Merriam's line at Spanish Flat. This line of Nomland's is probably a tribelet boundary, which Merriam and Goddard give as occurring at about that point (see Mattole Tribelets). Otherwise Nomland's boundary agrees with that of Merriam.

Merriam's line dividing the northern or Lolangkok Sinkyone from the southern or Shelter Cove Sinkyone begins in the east on South Fork Eel about a mile or two above the mouth of Salmon Creek, runs west from there through Kings Peak, and crosses the Mattole River just north of Ettersberg, intersecting the Mattole line a few miles from the coast. This line as given is the same as Merriam's, except that his begins in the east at Redwood Creek instead of at Salmon Creek. The change here is based on Goddard's village list, which indicates the present line.

The Lassik-Nongatl line begins in the east just below Ruth on the Mad River. It goes west from there around the headwaters of the Van Duzen River until it crosses the Eel at the mouth of Dobbyn Creek and thence west to the Sinkyone line. Kroeber and Merriam agree on the eastern part of this line but Essene disagrees with them, including a much larger portion of the drainage of the Mad and Van Duzen rivers in Lassik territory. I am at a loss to explain this version, since Essene's informant from the Lassik was the same one consulted by Merriam. It is not clear that Essene's boundaries were obtained from his informants, and this fact may explain the discrepancy. I have accepted the Kroeber-Merriam line here. To the west of this, Kroeber's line, instead of crossing the Eel, follows the river toward the northwest, so none of the main Eel River valley falls in Nongatl territory. Goddard gives villages on the main Eel which are said to be allied with others in the Blocksburg region, so the Nongatl must have claimed at least a small section of the Eel. I have therefore accepted the Merriam version.

The Wailaki-Lassik boundary begins in the east at the head of the Mad River and runs west to the North Fork of the Eel, which it crosses at the mouth of Salt Creek. It follows Salt Creek for a short way and then goes west to Kekawaka Creek, which it follows to its mouth on the main Eel. It crosses the Eel here and then goes west to intersect the Sinkyone boundary at the East Branch of the South Fork of the Eel. The boundary as given here is identical with the one given by Merriam, except that he includes part of the drainage of the Mad within Wailaki territory whereas Kroeber does not. I have accepted Kroeber's version, because it is supported in a negative way by Goddard (1924), who fails to include any Mad River drainage in Pitch Wailaki territory.

West of this area, Kroeber's boundary runs considerably north of Merriam's and of the boundary I have accepted. Merriam's line seems preferable because it is supported by Goddard and because Merriam's information is more specific than Kroeber's.

According to the information of Merriam and Goddard, the Wailaki may be divided into three groups—the Eel River Wailaki, the North Fork Wailaki, and the Pitch Wailaki. The eastern group, the Pitch Wailaki, occupy the drainage of North Fork Eel River above Asbill Creek, Hulls Creek, and Casoose Creek. Their western boundary begins in the north on Salt Creek near its confluence with North Fork Eel. It runs south from this point along Salt Creek and beyond it, crossing the North Fork of the Eel just above the mouth of Asbill Creek and intersecting the Yuki-Wailaki line near Summit Valley. The northern border of the North Fork Wailaki begins in the west on the main Eel River at the mouth of Cottonwood Creek, about three miles north of the mouth of North Fork Eel, and runs from there eastward for about six miles, where it hits the western boundary of the Pitch Wailaki. The western boundary of the North Fork Wailaki is the main Eel River from the mouth of Cottonwood Creek south to the Yuki line near Bell Springs Railroad Station.

The Kato-Wailaki line runs from the head of Blue Rock Creek in the east to the mouth of Hollow Tree Creek on the South Fork of the Eel in the west. This is Kroeber's version of the boundary. Merriam's version places the line somewhat south of this, beginning at Rattlesnake Creek in the west and going eastward south of Blue Rock Creek. Since I have ceded the drainage of Blue Rock Creek to the Yuki (see p. 160) in accordance with the views of Kroeber, I must, as a corollary, accept the northern boundary of the Kato as given by him.

The net result of the foregoing discussion is that the line surrounding the Athabascan peoples of Northwestern California remains much the same as Kroeber showed it in 1925, whereas the tribal boundaries are considerably changed. In the north, the Chilula and Whilkut occupy almost entirely different areas and the Hupa have been divided into two subgroups. On the coast, the Bear River and Mattole are divided, but this division had been shown by Goddard and Nomland previously. The Sinkyone have been divided into two subgroups and the Wailaki into three.

A really major difference is the accretion of territory by the Nongatl. This group is one about which least is known and this may be the reason why the map shows their territory as so extensive. It is very likely that data from a few good informants would show that the Nongatl actually comprise several distinct groups. There is a hint of this in Essene's account of Lassik war stories (1942, p. 91). He notes that the Nai'aitci, centering near the town of Bridgeville, were distinct from the Blocksburg people. Both of these groups are placed within the Nongatl area. No doubt more detailed information than we possess would show that the area which we have labeled Nongatl was actually occupied by two, three, or even more distinct groups.