NO-EB, 1902. MONTHLY WEATHER REVIEW. 623
tion and the construction of highways. Each locality has its
own peculiarities and possi1)ilities and must be studied incle-
pendently in order to arrive at a safe working knowledge upon
which to launch a new enterprise. General statements as to
climate, soil, flora, and hydrographic conditions are often mis-
leading and wholly inadequate for either business or scientific
purposes, and the sooner esact data are available the sooner
may: we expect to see the development of the island go for-
ward by the successful establishment of new enterprises.
In recognition of these peculiar conditions and in the hope of
securing for and presenting to the public fuller and more exact
information relative thereto, the Chief of the Weather Bureau
has just authorized some special work in the collection of all
information possible, both new and old, bearing on the climn-
tology of the island and its relation to crops, hygiene. water
power, commerce, and other matters of public interest. Intro-
ductory to this work, the results of which will be published
later, and more especially to that part relatiw to the water
resources of the island, it may be well to make a few observa-
tions in regard to its topography ancl rainfall.
In form, the island of Porto Rico is strikingly rectangular,
the greater asis lying almost exactly east and west. The nver-
age width of the island is about 35 miles, and the avernge
length about 105 miles. The area is, therefore, about 3,670
square miles. This, of course, is exclusive of the smaller cle-
pendent islands such as Vieques, Culebra, and others. The
orography of the i s l p d constitutes its most striking and im-
portant characteristic. The mountains are not very high, but
almost the entire surface of the island is covered with moun-
tain peaks of varying heights ancl shapes, apparently without
order or system as to arrangement. In places these mountains
meet the sea with such abruptness that there is not room for
even a driveway between mountain and sea, and in no place is
there a marginal coast of considera.ble breadth. Notwith-
standing this apparently confused and indiscriminate firrange-
ment of the mountaina, there is indeed a well-defined water-
shed dividing the island into two uiiequnl portions. This divide
extends from the southwest corner of the island eastward,
closely parallel to the southern coast line ancl from 10 to 15
miles from it until within the vicinity of Guayama it approaches
even nearer to the coast aacl then trends northeastward as does
the coast line finally culminating in the highest peak on the
island, El Yunque. This dividing ridge varies in height from
2,500 to 3,670 feet, and is made iip of several sections, each
having its own name. It will thus be evident that about one-
third of the island is situated to the south of this divide and
two-thirds to the north. The southern dope is necessarily very
abrupt while the northern is inore gradual. Both are broken
up into innumerable hills and niountnins with as iiiany streailis
between, giving to the island a truly rugged aspect. It may
help to a proper appreciation of the mountainous nature of the
interior to know that there are in Porto Rico betxeen 1,900
and 1,300 streams having names and of these about 30 are true
rivers, many of them having magnificent waterfalls, especially
those on the northern slope. Thus the island is divided hy-
drographically into two distinct sections: one lying to the enst
and south of the watershed, where the rivers are short and the
velocity of their waters very great. The rivers of this section
vary in length frcm 5 to 15 miles measured along their stream
beds. The other section lies to the north and west of the di-
vide and has longer rivers with larger basins. Here the rivers
vary in length from 35 to 40 miles. These rivers are often
very precipitous in their descent, especially for the first few
miles below their sources where they frequently descend from
1,000 to 2,000 feet within less than 10 miles. Naturally, there-
fore, one would expect to find not a few magnificent waterfalls
of great power and beauty. The possibilities of these falls as
sources of power ancl for other purposes certainly merit a more
thorough investigation than hasheretofore been accorded them.
Porto Rico lies well within the northeast trade winds, which
blow with great constancy throughout the year, and owing to
the peculiar relation of the topography of the island to these
winds, we find that the hydrographic divisions answer also for
climatological purposes, the essential and determining causes
being practically the same. Given, therefore, a thorough
knowledge of the topography, one is at once in possession of
a keg to the peculiar distribution of the rainfall over the
island. Under existing conditions, the east and north sides
of the divide must receive by far the greater rainfall and the
south and west sides the less. The records abundantly con-
firm the correctness of this statement. The heaviest rainfall
in Porto Rim occurs in the neighborhood of Luquillo in the
northeastern part of the island. There the annual fall amounts
to about 150 inches. A fall of 100 inches or more at stations
situated on or near the watershed is of common occurrence.
Along the northern slope and especially near the coast the fall
is much less, the average for a north coast station being about
55 inches. That portion of the island south and west of the
divide receives still less rainfall. Not only so, but the fall
seems to be more or less erratic, often resulting in severe
droughts and rendering irrigation indispensable. It is thus
seen that Porto Rico has wet and dry sides, more pronounced
than its so-called r r wet and dry seasons.” The effect of this
difference in the rainfall is at once seen by contrasting the
truly tropical luxuriance of the vegetation on the north side
with the barren hills and semiarid slopes on the south side.
The dry stony river beds of the south side are also in marked
contrast to the over~owing,bounding streams on the north side.
As above intimated, there exist in Porto Rico some inter-
esting a?d seemingly valuable waterfalls. Among the best
known and perhaps the most important, we may mention the
following, viz;
Salto de Rio de la Platn, near Comerio; Salto Rio Blanco;
Salto Morones, in Arecibo Rirer, near Utuado; Salto Maldo-
naclo, in Arecibo River; Salto Sanchez, in Arecibo River; Salto
Palmieri, in Arecibo river; mid Salto Paso Palma, in Yayuya.
A franchise has just been granted for the development of the
first-mentioned fall, the object being to utilize its power to
operate a railroad from Catano to Yonce. Promoters are also
investigating some of the other falls with a view to their de-
velopment for electrical plants and other purposes. We hope
soon to be able to present some cletails as to the size of the
nliove-mentioned falls, their available power, and other informa-
tion relative thereto of value to engineers ancl agiicultmist8.
NOTES AND EXTRACTS.
TROMHOLT’S C*TALOt3UE OF kop, :md 190 miles northwest of Soclan-Kyla, and observations
No one has labored more faithfully to eliiciclate the problem of the auroras were made at all three stations at pmmnged
of the aurora than Dr. Sophus Tromholt of RoRtock, Norway, lY0lnentS Of tillle for the express purpose of determining their
who died April 17, 189G. altitudes. The results of these special observations have as
Among his most important works was that clone when he, in Yet heen OnlS Partially Published.
1883-83, personally occupied a station a t Houtokeino, in Fin- In Petermann’s Blittheilungen, 1892, Vol. SSXVIII, Trom-
mark, Norway, in order to study the auroras observed by the liolt published a preliminary report on two important works
Norwegian party at Bossekop, Norway, and the Finnish party that he then had in hand, namely, a complete catalogue of
a t Sodan-Kyla. Tromholt was about 66 miles south of Bosse- the observations of auroras, as recorded by Norwegian ob-
624 MONTHLY WEATHER REVIEW. NOVEBIBEE, 1902
servers since the year 1594, with special descriptions of the
more important auroral displays ancl a general analysis of the
reports. As it seemed likely that there would be a long delay
in the publication of this important work,, the summaries
given by Tromholt in this preliminary communication were
extremely welcome. My personal interest in the matter even
led me to offer to supervise the publication in America of the
original work, since the author had failed to fincl a publisher in
Europe, but his early death closed our negotiations in this
respect and it appears that his manuscript was subsequently
given by his widow to his friend J. Fr. Schroeter for the purpose
of publication in Norway, if possible. This desire has now
been realized, and the publication a t the joint expense of the
Academy of Sciences a t Christiania and the Fricltjof-Nnnxen
Fund, as recommended by Professor Dr. Mohn, is a valuable
addition to our knowledge of the aurora borealis.
This volume is a quarto of $23 pages of test besides the pref-
ace. I n addition to the preface by Schroeter, there is also one
prepared by Tromholt, from which we learn that he began the
work of making this collection in 1879 and did not shrink from
the great labor of examining general historical and archwo-
logical works, public ancl private libraries, and the archives of
about a dozen public institutions, including the university
library a t Leipsic. Especial attention was given to Norwegian
newspapers. In a11 cases the complete records of details were
copied. The dates are always reckoned, as in astronomy, from
noon to noon; the four Rubdivisions of Norway are: I , north
of 68' 30'; 11, between 68' 30'and 65'; 111, between 65" and
61' 30'; IV, south of 61' 30'. The total number of stations
from which observations have been. gathered is 216 ancl the
total number of sources of information referred to in tlie cata-
logue is about 300.
Under the editorship of Schroeter, the material collected
by Tromholt has been judiciously sifted and apparently noth-
ing has been published that is not a distinct addition to exact
knowledge. Especially must we approve of Schroeter's good
judgment as an editor in restricting the publication of detailed
descriptions, partly because so many are given in the pul~lica-
tions of the International Polar Research, 1882-83. by es-
pert physicists that nothing more would seem to be necessary;
partly because there is abundant reason to think that tlie same
aurora presents different aspects to observers a few miles
apart, so that the mass of details must be useless until we nn-
clerstancl more about the location ancl origin of the phenomena.
The list of auroras (Catalogue D) begins with September
27, 1594, and ends with April 22, 1878, covering 310 pages,
and giving about 9 000 observations on nearly 6 000 clifferent
auroral dates. For each date is given the age of the nioon,
the place and time of observation, a few words of description,
and a reference to the source of information. I u the second
section, special descriptions are given of suspected auroras,
from 1550 to 1569, and of auroras proper from 1739 to 1878.
I n the third section, Table A gives the monthly and annual
sunis of the days on which auroras were observed somewhere
in Norway. The auroral year, like the snow year, is reckoned
from July to June, inclusive. The annual sums are then trans-
formed into smoothed numbers by the formula 1/10 (n + 311 + 4s + 2d + e). The aununl sums are also given for each of the
four sections into which the author subdivides the area of
Norway. On page 353 these annual sums, from 1'761 to
1878, are combinecl into one total, representing the monthly
distribution of 5 891 days with auroras. This computation is
given for each of the four subdivisions, RS well as for all Nor-
way. The monthly sums are also converted into comparable
percentages by reduction to a uniform summation of 1 000,
whence it appears that in the northernmost division of Nor-
way the annual periodicity shows a maximum a t the winter
solstice, whereas in the southernmost section two maxima occur
corresponding to the equinoxes. Therefore, in the northern
portion of Norway the annual periodicity is of the same char-
acter as in the arctic regions generally. The annual period
prevailing in the southern division of Norway is similar to that
founcl in all middle latitudes. A combination of all the records
of Tromholt for Norway, and Rubenson for Sweden, is given on
pages 353-414, for the years 1733, March 4, to 1878, April 22.
The study of the geographical distribution of these observa-
tions shows again a maximum in January in the northern re-
gions, and niasima about September 21 and March 21 in the
southern regions.
I n the third part of the volume Schroeter has presented us
with a new and complete recomputation of the preliminary
discussion founcl among the papers of Tromholt, so that the
figures differ somewhat from those published by the latter in
Petermann's iUittheilungen. Especially has he omitted the
attempt to investigate the apparent periodicity of the aurora
clue to the obscuration by the varying brightness of the moon,
since such an investigation can not lead to any satisfactory re-
sult. unless the cloucliness of the sky ha8 been recorded uni-
formly throughout many lunar months. For the same reason
the study of the annual periodicity must be omitted, but we
do not a t first see why a similar argument would not hold
good against .the investigrLtion of the sun-spot period which
Schroeter has undertaken. Notwithstanding the imperfections
of his data, Schroeter coaclucles that the results for Norway
are in general quit.e parallel to those deduced by Rubenson
from the observations in Sweden.
It is not likely that this great work of Tromholt's, taken in
conjmction with its predecessors, exhausts all accessible Euro-
pean records, but it goes a long way toward preparing for a
renewal of the comprehensive studies of the geographical and
chronological distribution of auroras that were initiated by
Fritz fifty years ago.-C. A.
MARYLAND ULIMATOLOGY.
Dr. 0. L. Fassig communicates to the Library of the
Weather Bureau copies of the Hhort paper on the climate of
Cecil County, Md., just now published as a part of the volume
of the Maryland Geological Survey on the Geology of Cecil
County. A third pamphlet on the climate of Garrett County
is now in press. A similar report on Allegheny County was
published in 1900. The plan is to take up each county of the
State in turn and tabulate and discuss all the meteorological
observations that have been macle at any time in the history
of t.he county, but especially those made under the auspices of
the Smithsonian Institution, 1848-1873, the Weather Bureau,
1870-1903, and the Maryland State Weather Service, 1896-
1903. The espense of publication is met by the Maryland
Geological Survey. Each climatic sketch forms a chapter in
the volume on the geology of the respective counties, In this
way the local peculiarities of each portion of the State will
be thoroughly presented.
Cecil County is in the extreme northeast section of the State
of Maryland. Its surface is most gently undulating with a
general elevat8ion of about 300 feet, but occasionally rising to
500. The tide waters of Chesapeake Bay penetrate every por-
tion of the southern part of the county. The principal records
discussed in this pamphlet are those of Woocllawn from 1865
to 1875, inclusive, a t which station the maximum temperature
of 100' occurred once, in July, 1868, and the minimum tem-
perature of -loo, also ouce, in January, 1873. The maximum
monthly precipitation was 11.81, in August, 1874, and minimum
monthly was 0.36, in October, 1875.-C. A.
. . .-
WEATHER BUREAU MEN AS INSTRUCTORS.
Mr. J. R. Weeks, Observer, United States Weather Bureau,
Macon, (fa., reports that he gave an informal lecture on the