538 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1907. 
uncovered now, it is because the tumulus has been removed by climatic 
influences or by the hand of man. 
The simple stone chambers, or stone cists—that is, three perpendicu- 
lar slabs of stone covered with a very large cap stone—are not found 
in Japan. This is the more remarkable from the fact that in Korea, 
mostly in the northern part, I have seen a great number of these, 
while megalithic dolmens appear to be lacking there. 
The stones of the Japanese dolmens, particularly the roof stones, 
are often very large, but regularly hewn stones are the exception. 
There are, according to Gowland, only four dolmens of the last 
sort, and they belong to a comparatively recent period. Generally 
they are put together without any mortar (which, however, was 
doubtlessly known then) and the interstices filled with small stones. 
Occasionally true rock graves are found, graves of regular shape 
hewn out of the rock. Judging from the finds in them they belong 
to the dolmen age. They differ in their whole execution from the 
primitive caves mentioned before. 
Gowland differentiates four forms of the dolmens: First, the sim- 
-ple covered passage (allée couverte); second, the covered passage 
broadening out on one side at the inner end into a chamber; third, 
the same form with a symmetrical widening out on both sides (this 
is the most usual form), and fourth, dolmens with two separated 
chambers one lying behind the other. 
Tt is likely that the last form always represents a later stage of 
development; perhaps also the social position of the deceased in- 
fluenced the form. 
The chambers are rectangular in shape. The length varies from 
14 to 8 meters; the gallery leading in is often longer. The breadth 
of the chambers is generally less than 3 meters and the average height 
about 2 meters, although it may rise to 5 meters. Some of them are 
vaultlike. The tumulus over the grave is sometimes as large as 30 
meters in length and 10 in height, but usually only half that size. 
The entrance is almost always from the south, though frequently a 
little toward the east or west. Deviations amounting to 40°, which 
are observed in the large Japanese dolmen as well as in the small 
dolmens of Korea, can perhaps be explained from the time of year 
of the burial. East and west are easier to determine than true south, 
on account of the rising and setting of the sun. In midsummer the 
sun rises toward the north, in winter toward the south. If the people 
founded their orientation on the rising of the sun, as they probably 
did, south would be too far to the east in summer and too far to the 
west in winter. 
Whether the peculiar position of the dolmen entrances toward the 
south is to be attributed solely to the sun and its worship, or whether 
it is based on some other religious or astronomical idea, it is difficult 
