JUNE 20, 1884.] 
Mr. J. B. Marcou has about completed the sorting 
and arranging of the type specimens in the National 
museum, and will soon take the field. He will devote 
his time to the study of the mesozoic and tertiary 
formations along the Atlantic coast, especially in 
_ New Jersey, Maryland, and Virginia, and possibly in 
North Carolina, beginning in New Jersey. 
Prof. L. C. Johnson has assorted and labelled all 
of his collections made last season, so as to show 
the localities from which they were obtained, and the 
geological horizons which they represent. He has 
now left for Mississippi, where he will begin to col- 
lect in the cretaceous and in the older tertiary. 
Prof. William M. Fontaine is still engaged in the 
study, classification, and description of the fossil 
plants from the younger mesozoic strata, and in the 
preparation of drawings in illustration of his work. 
Prof. H. S. Williams reports progress in his work 
of elahorating the material collected by him, and in 
writing his report upon the comparative study of the 
Devonian faunas of western New York. 
U. &. signal-office, 
The progress of tornado investigation.1—In the 
study of tornadoes it has become necessary to under- 
take something more than a simple record of their 
occurrence, or an occasional investigation of those 
that are attended with unusual destruction to life 
and property. A practical knowledge of the nature 
of these destructive storms is a matter of the utmost 
importance to the inhabitants of certain sections of 
the country; and not least among the objects aimed 
at by the chief signal-officer, in directing the continu- 
ance of tornado investigation, is to allay any need- 
less anxiety or fear on the part of those people living 
in the regions most frequented by these storms. 
Methods of observation based on reports from sta- 
tions situated from one hundred to two hundred 
miles apart, as in the case of cyclones and hurri- 
canes, are inadequate to develop the mysteries of the 
funnel-shaped tornado-cloud. 
As a consequence, therefore, a new plan was de- 
vised, based on the result of special investigations in 
1882 and 1883, by means of which it is now sought 
to study more intimately the origin, character, fre- 
quency, and geographical distribution of tornadoes. 
To inaugurate the details of the proposed system of 
work, it became necessary to establish a corps of ob- 
servers, whose duty should be to report the occur- 
rence of tornadoes, and make examinations of their 
paths and various phenomema; for which purpose 
special definite instructions are issued. 
The observers are called tornado reporters, and 
now number about eight hundred. Their stations 
are mostly located in the states of Alabama, Georgia, 
South Carolina, North Carolina, Missouri, Arkansas, 
Kansas, Lilinois, Indiana, lowa, Nebraska, Wisconsin, 
and Minnesota. These are the states in which tor- 
nadoes are of most frequent occurrence: and this dis- 
tribution limits our study to certain states, and even 
certain portions of a single state; for there are por- 
1 Communicated by permission of the chief signal-ofiicer, 
U,S. army. 
SCIENCE. 
767 
tions of some states that are frequently visited, and 
other portions seldom, if ever, visited by tornadoes. 
In the regions of greatest frequency the stations num- 
ber from one to three in each county, depending 
upon its size. 
Tornado reporters, in return for their voluntary 
contributions, are supplied with the tornado publica- 
tions of the signal-service; they are also furnished 
with the material necessary for the proper record and 
mailing of observations and reports. 
Reports are forwarded to the chief signal-officer 
as soon as possible after the occurrence of a tornado, 
and consist of detailed descriptions, instrumental ob- 
servations, photographs, diagrams, charts, and illus- 
trations. ' 
While attention is mainly given to the examination 
and report of tonadoes for the current year, each re- 
porter is instructed to work up the past history of 
these storms in his state, making careful search after 
any facts relating to windfalls, or other traces of past 
tornadoes. Some of the results sought to be attained 
by the above method of investigation may be briefly 
given as follows: 1°, to determine the origin of 
tornadoes, and their relation to other atmospheric 
phenomena; 2°, to determine the geographical dis- 
tribution of tornadoes, and their relative frequency of 
occurrence in different states, and in different parts 
of the same state; 8°, to determine the conditions 
of formation with a view to the prediction of tor- 
nadoes; 4°, to determine the means of protection 
for life and property; 5°, to determine the periodi- 
city of the occurrence of tornadoes, and their relative 
frequency by seasons, months, parts of month, and 
time of day; 6°, to determine the prevailing char- 
acteristics of tornadoes; 7°, to determine the relation 
of tornado regions to areas of barometric minimum. 
A review of the past year gives the following as 
some of the principal results: — 
1°. That there is a definite portion of an area of low 
pressure within which the conditions for the develop- 
ment of tornadoes is most favorable; and this has 
been called the ‘dangerous octant.’ 
2°. That there is a definite relation between the 
position of tornado regions and the region of high 
contrasts in temperature, the former lying to the 
south and east. 
3°. That there is a similar definite relation of 
position of tornado regions and the region of high 
contrasts in dew-point; the former being, as before, to 
the south and east. 
4°, That the position of tornado regions is to the 
south and east of the region of high contrasts of cool 
northerly and warm southerly winds,—a rule that 
seems to follow from the preceding, and is of use when 
observations of temperature and dew-point are not 
accessible. 
5°. The relation of tornado regions to the move- 
ment of upper and lower clouds has been studied, 
and good results are still hoped for. 
6°. The study of the relation of tornado regions to 
the form of barometric depressions seems to show 
that tornadoes are more frequent when the major 
axis of the barometric troughs trend north and south, 
