16 Einar Lexow. 
been in common use for centuries in the Balkan countries. 
At Butmir in Bosnia they have been found in an early stratum 
of the neolithic age, dating from perhaps 2500 B. C. Other 
early examples from Petreny in Bessarabia and from Sesklo 
and Dimini in Thessalia are well known. Most of the modern 
archeologists, who have dealt with this subject, have therefore 
maintained that the system of spiral decoration with which 
we are here concerned, has originated somewhere in this part 
of Europe, and thence spread to the adjoining countries. From 
Creta where it is found in the Middle Minoan period II, it 
may have found its way to Egypt, a hypothesis which accords 
well with the fully established synchronism of this period with 
the XIIth dynasty. 
But the solution of this problem cannot fail to be influen- 
ced by the discoveries recently made at Hal Tarxien, Malta, 
by T. Zammit. A system of subterranean buildings have here * 
been unearthed, and their sculptured altar stones show very 
interesting spiral patterns. As remains of interments belonging 
to the bronze age were found in a stratum about 1 ft. above 
the altar stones, the latter must be of an earlier date and 
probably belong to the eneolithic period, which if contemporary 
with the same period in Crete can be dated to the centuries 
just before 2200 B. C. ‘ 
The decorative system of the Maltese stones falls into two 
different classes, one in which the spirals are the dominating 
feature, and one in which a system of furcation plays the 
leading part. T. Zammit suggests that the former class repre- 
sents the earlier phase, as such stones are generally found 
in that part of the building, which he considers to have been 
constructed before the part, in which the more complicated 
branch decorations are found. There are, nevertheless, many 
indications that the reversed order is more probable. The 
rooms, in which the stones with branch decorations have been 
found, adjoin the main entrance, and had to be traversed be- 
fore the other part of the building could be reached; it is 
therefore reasonable to suppose that they were first completed. 
Neither does the history of ornamentation offer any parallel 
to a development from geometrical spiral motives to a compli- 
cated system of branches. But that a phylomorphic design 
can degenerate into a system of spirals is not seldom found, 
