SEPTEMBER 28, 1916] 
The first and larger portion of the book deals 
with the archeology, ethnology, linguistic divi- 
sions, habits, and beliefs of the ancient inhabitants 
of the present States of Nicaragua, Costa Rica, 
and Panama. The people belong to the same stock, 
with modifications, which stretched from the 
Sonoran region through the whole of Central 
America far into the western—let us say Andine— 
portion of the southern continent, leaving the 
bulk of the latter to what may be called the typical 
neo-tropical races, notably Guarani, Tupi, Arawak, 
and Caribs. In conformity with the configuration 
of the isthmus, the influence from the Inca centre 
was insignificant in comparison with that from the 
Maya-Mexican. It is the reverse with the fauna 
and flora, which naturally date from much earlier 
epochs with broader contact. 
The last hundred pages are devoted to the West 
Indies. These fairest and most fertile islands 
being subject to earthquakes, volcanic eruptions, 
and hurricanes, it is, according to our author, not 
surprising that their inhabitants developed a reli- 
gious system which consisted in the main of a pro- 
pitiation of the powers of Nature. Nevertheless, 
according to the testimony of C. and F. Columbus, 
Be ee 
Fic. 2.—Porto Rico : Stone seat. (Scale, one-third.) From ‘‘Central American and 
West Indian Archzology.” 
father and son, they seem to have been a kindly, 
honest, and generous folk, to whom contact with 
the Spaniards meant speedy extermination. 
The original stock of the whole archipelago 
were Arawaks from the’ southern continent, and 
were in turn followed by the more bellicose Caribs, 
who at the time of their discovery were in posses- 
sion of the Lesser Antilles and fast extending into 
some of the greater islands. They had a patriar- 
chal system, with caciques or small chiefs. A tie 
of friendship was the mutual exchange of their 
names. They practised the couvade. The 
Arawaks were armed with sword-clubs and 
javelins, hurled by means of ornamented spear- 
throwers. The chief weapon of the Caribs was 
the bow. The male’ prisoners were eaten. 
There is no evidence that stone weapons were 
used, the blades of stone found being tools. 
Charms were made of wood and stone, images of 
man and ‘animals. Interesting are the stone 
collars which are said to be the translation into 
stone of originally a wooden hoop, a tree-fork 
bent and fixed by bandages into this shape. The 
enclosing of a spirit in such a circle is connected 
NO. 2448, VOL. 98] 
NATURE 71 
with tree-worship. Beautiful stone collars are 
found also in the Maya-Mexican countries. 
But there is very little known about such and 
similar curiosities beyond the often gratuitous 
accounts of the old chronists, and the same applies 
to the beliefs and habits of these vanished people. 
Some are no doubt of genuine indigenous origin; 
others point to the western or to the southern 
mainland, just as one would expect. Since these 
islanders used dug-outs, large enough for even 
long voyages, it is not surprising that their kind 
of civilisation—as shown by their weapons, orna- 
ments, stone masks, figurines, pottery, etc.—was - 
more or less alike all over the archipelago. 
THE ABNORMAL PROPAGATION OF 
SOUND BY THE ATMOSPHERE. 
M® S. FUJIWHARA has recently published 
a second valuable memoir on the abnormal 
propagation of sound-waves in the atmosphere 
(Bull. Centr. Meteor. Observatory of Japan, vol. 
il., 1916, pp. 1-82). As in his previous paper 
(Nature, vol. xcii., 1914, p. 592), he ascribes the 
peculiarities investigated to variations of the air- 
temperature and of the velocity and 
direction of the wind, and he concludes 
that the structure of the upper atmosphere 
may be inferred from the form of the 
region of audibility. 
The present memoir consists of two 
chapters. The first is theoretical, and 
deals with the modes of propagation of 
sound-waves through the atmosphere, the 
structure of which may be one of the five 
types observed by Capt. C. J. P. Cave 
at Ditcham Park, ,Petersfield, Hants. 
He shows that the region of audibility, in- 
cluding the sound-source (or proximate 
region), may in many cases be of trian- 
gular, or rather fish-tail, form, the axial 
direction of which may be not only in the 
direction of the wind then prevailing, but also in 
any other direction, at right angles or even oppo- 
site to that of the wind; for it depends chiefly on 
the direction of the wind in the lower stratum of 
the atmosphere with respect to that at the earth’s 
surface. In other cases the proximate region may 
assume a spiral form, with the vertex at the source 
of sound and extending in the sense of the veering 
of the wind in the upper atmosphere. The regions 
of audibility due to winds at different heights may 
overlap one another. Detached regions of audi- 
bility may appear in a zone subtending an angle 
of less than two right angles at the sound-source, 
and they may be of any form. If the wind 
remains steady in direction and velocity in the 
upper layers, or if there is a decrease in velocity 
in the upper atmosphere, detached regions should 
not occur.. With an increase of velocity in the 
upper atmosphere, detached regions may occur 
in the same direction as the proximate region. 
But if there should be a reversal or great 
change of direction in the upper layers, or if the 
upper wind should blow from a distant low-pres- 
sure centre with frequent reversal in the lower 
