TRANSACTIONS OF SECTION G. 505 
will be formed by rock-fill dams, which are the best adapted for the natural 
conditions of the country, where labour is scarce and inefficient, good materials 
rare, and earth tremors rather frequent. 
Rock-fill Dams.—The rock-fill dams can be built almost entirely by mechanical 
implements served by a few good workmen, and if the slopes do not exceed about 
30 degrees the dam can resist any shock of earthquake without great damage, 
and in any case repairs are easy. This type of dam is very common in the 
Western States of America, and has been adopted with success also in Italy in 
the valleys of Cenischia, Biaschina, and Devero. 
A special feature of the new reservoirs will be the ‘ automatic syphon-spill- 
ways’ applied by Mr. Gregotti to many dams and canals both in Italy and else- 
where. It consists of a tube in the shape of a syphon, generally of square 
section, made of ferro-concrete, and capable of discharging from 1 to 15 cubic 
metres (525 cubie feet) per second, according to section and head, and if larger 
discharges are necessary then several syphons are distributed side by side up to 
ten. As soon as the water in the reservoir exceeds by 2 or 3 inches the normal 
level, the syphon is automatically primed, owing to the special conformation of 
its top-lip, and begins to act with full discharge till the level in the reservoir is 
lower than the lip of the syphon, when air gets in and the flow ceases. 
The velocity of the water in the syphon is that due to the difference of level 
from the reservoir and the outlet of the syphon, and thus is greater, and the 
discharge is correspondingly greater, than in an ordinary overfall-spillway, where 
the velocity is due to the more limited head between the level in the reservoir 
and the sill of the spillway. 
At equal discharge the syphon-spillway works better and costs much less than 
the usual overfall-spillway. 
For example, in the Lagolungo reservoir near Genoa, the old overflow-spill- 
way required a head of 1:40m. of water (4’ 8/) and a volume of 440,000 cubic 
metres of water (more than 15 million cubic feet) ran to waste over the spill- 
way after a storm had passed. A battery of ten syphons, each with an internal 
square section of 1:90m. (6/ 4/ by 6/ 4/) and a working head of 5-75m, 
(18/ 8”), was substituted to the old spillway, capable of discharging 150 cubic 
metres per second (5,250 cubic feet); at the same time allowing the sill to be 
raised by 0°95 (3 feet), by which an extra volume of 300,000 cubic metres of 
water (104 million cubic feet) are impounded. 
Besides, it was proved that the syphons give the discharge of 150 c.m. with 
only 4 inches of super-elevation of the water over the lip, or with an elevation 
of 14 inches less than before, when the overfall-spillway was used, thus reducing 
correspondingly the water-pressure against the dam. 
The greater volume of 300,000 cubic metres of water thus impounded repre- 
sents an income of at least 15,000 francs (600/.), a year, and, capitalised at 5 per 
cent., a total of 300,000 francs (12,000/.), equal to at least six times the cost of 
the new syphon-spillway ; which, moreover, is more compact and more efficient. 
Of these automatic syphon-spillways there are more than 100 on Italian dams 
and canals, with heads varying from 38 to 20 feet, and all give complete 
satisfaction. 
The author exnibited several views of ruins of Roman works, and a cinema- 
film showing the method of constructing the railway and the striking difference 
in the zones cultivated with and without irrigation. 
3. Investigalian of Nile River Flood Record from a.p. 641 to a.p. 1451 
for Traces of Periodicity. By T. W. Krexe, M.Inst.C.F. 
The author was led into this investigation when examining records of rain- 
fall at various places, particularly those of the Commonwealth of Australia, 
with a view to ascertaining whether periodicity exists. 
The results of previous study of this meteorological question were given by 
the author in a paper entitled ‘The Great Weather Cycle,’ read before the Royal 
Society of New South Wales on July 1, 1910, and published in their Proceedings 
(see volume xliv., pages 25 to 76), in which he sought to show that the period 
must be a long one. He was of opinion that the defects which were apparent in 
