104 
Grande, and whose summit is only 50 feet 
above the lake and 154 feet above sea level. 
This gap, which occupies so important a rela- 
tion to the proposed canal, is the product of 
the familiar process of stream capture. Ow- 
ing to the decided advantages possessed by 
the streams flowing directly to the Pacific 
over those flowing eastward, at first to the 
bay indenting the Pacific coast and after- 
wards to the lake, the former were able to cut 
backward through the divide into the drain- 
age area of the latter and to divert their 
headwaters. In this way an eastward- 
flowing stream originally occupying the 
position of the Tola, the upper Rio Grande, 
the Guiscoyol and the Lajas was beheaded 
and the drainage of a large part of its basin 
was diverted to the Pacific. The deserted 
valley of this stream forms the low gap 
through which the canal route is located. 
It is so broad and level that accurate in- 
strumental work is required to determine 
the actual summit of the continental divide. 
Considering the origin of Lake Nicaragua, 
it is manifest that it must originally have 
extended entirely down to the point where 
its waters escaped through the gap in the 
continental divide—that is, to the present 
Castillo Rapids. This point, however, is now 
more than 30 miles down the San Juan River 
from the lake. The upper portion of that river 
meanders through an alluvial plain which 
becomes narrower down streams and has evi- 
dently been reclaimed from the waters of 
the lake by sedimentation. It is well 
recognized that lakes are ephemeral features, 
and the commonest way in which they are 
obliterated is by the filling at their upper 
ends with sediment deposited at the mouths 
of tributaries. In this case, however, the 
process is reversed. The area of the lake 
is being contracted chiefly by filling at its 
lower end. The filling is being accom- 
plished by the tributaries entering this 
lower portion of the lake, many of which 
have been converted into tributaries of the 
SCIENCE. 
[N. 8. Von. X. No. 239. 
San Juan. The present river channel does 
not coincide with the position of the river 
which formerly occupied this basin before 
it was drowned by the waters of the lake. 
Its position is dependent on the relative 
amounts of sediment delivered by the tribu- 
taries on either side, and it has been pushed 
toward the northern side of the old basin 
by the larger tributaries from the south, the 
Frio and Poco Sol. This portion of the 
San Juan may best be described as a resid- 
ual river channel—that is, a broad arm of the 
lake has been gradually constricted by the 
deposition of sediment on its margin, and all 
that remains is the narrow river channel 
kept open by the current of water flowing 
from the lake. This hypothesis, verified 
by borings made in the river channel, has 
been of material service in so locating the 
canal line that all rock excavation in this 
portion between the lake and the Castillo 
Rapids should be avoided. 
While the writer has no intention of 
touching upon the engineering features of 
the canal problem, it may be stated that 
the geologic examination of the route, in- 
cluding the boring, has resulted, in nearly 
every case, in showing that conditions are 
more favorable than they had previously 
been assumed. In the few cases in which 
less favorable conditions were found modi- 
fications in the plans suggested themselves 
by which the unfavorable conditions are 
avoided. 
Thus the project, which has repeatedly 
been pronounced feasible by eminent engi- 
neers, is placed in a still stronger position 
by the most exacting scientific tests. 
C. WitLarp Haves. 
U.S. GEOLOGICAL SURVEY, July, 1899. 
TRANSPARENCY AND OPACITY.* 
One kind of opacity is due to absorption ; 
but the lecture dealt rather with that de- 
* Abstract of a lecture given by Lord Rayleigh be- 
fore the Royal Institution of Great Britain. 
