Inside Fern Cave

by George Kline

Entrance to Fern CaveFive years of archaeological surveying on the land of the Gumbatwas and Kokiwas bands of the Modoc Indians (Ray, 1963 p. 205), has made a lasting impression on my life and inspired me to devote many hours of free time perusing all the published information I could assemble, to aid in the understanding of the art, artifacts, archetecture, and land features found in the area. The Modoc Indians lived in what is now the area on both sides of the California-Oregon border, between the Cascades and the Warner Mountains. The area is to this day very sparsely populated and most of the evidence of the past still remains intact. The rock art is mostly untouched except by the elements, and vandalism is rare. One of the reasons it has remained so is its remoteness. Many of the early white settlers who "attempted" to settle here and work the land by farming or ranching, failed miserably. Many historical homesteads only exist as ruins today.

The land is extremely rocky. The topsoil is thin to nonexistant most places. Basalt bedrock is exposed nearly everywhere. The soil is hard in the dry summer months as fresh water becomes scarce. Wells dry up, as do the natural springs. In the spring, as the snow melts and the rains fall, there is often widespread flooding and the soil becomes mucky, slippery, and gooey. It becomes difficult simply to walk.

Presently, this land is managed by the Modoc National Forest and is primarily used as summer cattle grazing, and some small scale timber harvests. Other lands within the area are managed by the National Park Service, caring for the Lava Beds National Monument, and the National Wildlife Service managing the lands of the Clear Lake, Tulelake, and Lower Klamath Wildlife Refuges.


This area is of volcanic origin. Mostly dacitic and vesicular basalts, forming the relatively flat Modoc Plateau. Various volcanic features, some dormant or potentially active, adorn the landscape. Shield and composite volcanoes, buttes and cindercones, chimneys and lava tubes, obsidian and pumice, are all abundant.

Throughout the Modoc Plateau, faulting has occured over the millenia, mostly in a north-south orientation. The rivers and creeks draining the land tend to follow these faults, and in turn, people followed these waterways. Many of these faults form vertical "walls" rising to an average of 30 to 50 feet. These fault scarps are facing to the east or west, hence allowing the sun to set or rise on these natural "canvasses", which the native peoples found convenient for their rock art. These formations are extremely numerous throughout the plateau as are the petroglyphs adorning them.

Lava flows from the Medicine Lake Highlands, a shield volcano of vast proportions, has exuded "rivers" of molten basalt along its flanks producing hundreds of "lava tube" caves. Within these lava tubes the temperature and humidity remain fairly constant throughout the year. This environment is ideal for the preservation of pictographs which were placed on the cave walls long ago. These lands are protected and managed by the National Park Service, and are referred to as the Lava Beds National Monument.

About the Pictographs and Petroglyphs

figure12smRock art is categorized into two distinct types. Pictographs are "painted" onto the surface of the rock utilizing pigment and a base on which the pigment will adhere. The pigments used here are either black and/or white. The black pigment has been found to be charcole. The white is primarily calcite which was scraped off the rocks of exposed basalt. This calcite was deposited on the rock overhangs and exposed ceilings from water seepage over the ages. These pigments were mixed with animal fats obtained from a variety of local sources such as waterfowl, deer, antelope, or small mammals.

figure2smPetroglyphs are either "pecked", scratched, or otherwise abraded into the natural patina of the rock. The petroglyphs are far more numerous and widespread here on the Modoc Pleateau, certainly unhampered by the geological morphology. A pictograph exposed to the weather and direct sun would soon be erased where a petroglyph remains for millenia until the rock is actually eroded. The petroglyphs found on the Modoc Plateau are mostly pecked into the rock utilizing a slightly harder rock in a tapping or hammering motion. Very few were made by scratching into the rock face.

figure5smDeciphering or extracting a particular "meaning" from these artworks is mostly guesswork. One pictograph (fig. 5 left) in Fern Cave, was for years, believed to be a depiction of a supernova (The Crab Nebula Supernova) that occurred in the year 1054 AD. Likenesses are found in several other rockart examples elsewhere on the planet. However, recent Carbon 14 tests of the charcole pigment in Fern Cave now prove that this is not the case.

figure7smAlternatively, a site located on Horse Mountain has been found to be a seasonal equinox marker. Sunrise, on the days of the spring or fall equinox, sunlight shines through a natural hole that has been enlarged and casts a spot of light in the opposite wall where it aligns with petroglyphs of lines and figures.

figure17smThe meaning behind most petroglyphs remains a mystery. Only the artist truely knows the meaning. One could asemble a dozen "experts," be they cryptologists, linguists, or whatever, and you'll probably get a dozen different translations. Like true art, they mean something different to everyone.

figure6smPetroglyphs are truely numerous in northern Modoc County. They are found along the rivers and creeks which follow the fault lines, as well as the vertical blocklike faults adjacent to mesas and ephemeral pools. One popular site (and easy to access) is called (naturally) Petroglyph Point (see #6 on Map). It is an extremely old eroded volcano that was once an island in the middle of Tulelake. Waterline marks on the soft tuff and ash still show the alternating past lake levels. Although there has been some vandalism, there are still many very old prehistoric, as well as historic examples of petroglyphs. Since the rock material is somewhat soft, wind and weather do continue to erode the rocks.

figure15smBoles Creek drains the area to the southeast into Clear Lake. Numerous petroglyph panels appear all along its course. On one panel along Boles Creek, the petroglyphs were found to continue below the level of the sediments (see #7 on Map). Upon excavation, it was discovered that the strata covering the petroglyphs contained a layer of volcanic ash deposited from the eruption of Mount Montezuma, (known today as Crater Lake, Oregon), which erupted over 5,000 years ago (Howe 2), so it is known that this particular panel was created more than 5,000 years ago.

Fern Cave

figure8smFern Cave is located within only a few minutes walk from Captain Jack's Stronghold, where a band of 70 Modoc warriors, men women and children, held off hundreds of U. S Army troops for five months (Murray 220). Fern Cave was formed approximately 25,000 years ago from lava flows exuded from Mammoth Crater in the north slope of the Medicine Lake shield volcano (see #1 on Map). The figure11smentrance to the cave is a hole in the ground measuring about 8 x 10 feet. The entrance was formed when the roof of the cave, at its widest point, collapsed, forming the rocky deposit now covered with a lush green carpet of ferns (see Entrance to Fern Cave). Even though the outside temperature varies from winter lows of -40F to summer highs of over 100F, the temperature inside the cave figure10smremains fairly constant all year, from about 45F to 55F, with a constant humidity of 80%. Upon entering the cave, one climps straight down an aluminum ladder ti the top of the collapsed pile. As soon as your head passes below ground level, you instantly feel the damp, cool air. My visits to the cave have always been in the heat of summer, and the feeling uon entering was of instant relief. A shrt figure1smpath leads one down the collapsed rubble to the cave floor at the west facing wall. On the large boulder to the right, just as you reach the floor. you see a mortar and pestal, about waist high. The path continues to the left, between the rubble pile and the cave wall. On the walls, from knee high to eye level, are pictographs in black, of many circular shapes, some with sun-like rays (see fig. 1 left).

figure5smFurther down along this wall, we are now getting further from the natural light at the cave entrance, and must use artificial light for viewing the features, and for safety. The next grouping of features show the likenesses of two adult, shadowlike figures, with a small juvinile between, at about belly height (see fig. 5 left). Above their heads one sees other--what appear to be--human figures, with circles and other unknown forms. In the upper left corner is what was formerly believed to represent the 1054 Crab Nebula supernova event, seen as a crescent with points facing down, a circle directly below and another directly above and to theleft. Continue now across the cave to the opposite wall, following the collapsed rubble on around, still to the left. On the cave floor, within a few feet of the rubble is a "basket hopper mortar." This artifact resembles a very shallow bowl, on top of a somewhat flat basalt slab. It was used in conjunction with a woven basket with a small hole left in the bottom. The basket was placed over the mortar, and material was pounded to a pulp inside the basket, with a maul stone.

figure3smMoving on to the far wall (see fig. 3 left) your flashlight beam lights up a fantastic array of black and white contrasting figures , which defy description (but I'll try). Hard to miss here, is the scorpion-like figure amongst all the lines of dots surrounding many other mysterious forms. Further on to the right, is another equally inspiring figure9smdisplay of contrast (see fig. 9 left)--shadowy humanlike forms, bright white concentric circles with radii, black "spiders" and "waterbugs." It is difficult to resist an attempt at deciphering the figures, but here I am merely trying to describe "shapes" and likenesses in which we can collectively understand in our context.

Continuing down the wall, still farther from the collapse, and getting progressively darkes, we find that the pictographs are less densely displayed, and the use of white pigment wanes.

Sage Lake

This lake is a small, shallow, seasonal accumulation of winter snowmelt and spring rainfall that dries up completely every summer. It is located between two block fault scarps climbing up to a height of a least 60 feet to a mesa on either side to the east and west (#4 on map). Petroglyphs were only found to be on the western side, facing the east (fig 2-1, fig 2-2). To the west the rock walls were found to be "less desirable canvasses" possibly because they were not quite smooth on the surface as well as only a few feet in height at this point.

Mud Lake

Here is a similar formation to Sage Lake and located about a mile to the south (#3 on map). The walls to the west are as high or higher and may petroglyphs were found here, and also are facing east (fig 3-1, fig 3-2, fig 3-3, fig 3-4). From this vantage point one can look to the east and Clear Lake comes into view. Also notable are many observed flocks of Canada geese flying low over the mesa ridge above, as this was the lowest point to cross to and from Clear Lake to Tulelake. When surveying the top of the mesa, several hunting blinds of stacked rocks forming an arc were discovered scattered along tis "flyway." When the flocks flew over one morning they were merely 30 or 40 feet above our heads.

To the east of Mud Lake the land rises gently until it is found to drop off steeply to the east, creating another fault scarp wall of basalt facing Clear Lake (#5 on map). More petroglyphs are found here (fig 4-1,fig 4-2).


These locations contained many more petroglyphs than were reported here. To do a complete job would require writing a book. Only the best photographs were used, and the clearest examples were photographed. Changes in lighting and shade either bring them out to clarity, or wash them into obscurity. I firmly believe I could spend several years recording and photographing in detail only the petroglyphs in the vicinity of clear Lake, and the creeks draining into it from the east. I hope to someday get the opportunity to do so.


Howe, Carrol B.
1979  Ancient Modocs of California and Oregon. Portland: Binford and Mort Publishing.

Murray, Keith A.
1965  The Modocs and Their War. Norman, Oklahoma: University of Oklahoma Press.

Ray, Verne F.
1965  Primative Pragmatists. Seattle: University of Washington Press.

Lavabeds National Monument