FGDYTHJNJKGTRD

THE CONTROL OF RIVER FLOODS WITH SPECIAL REFERENCE TO THE MIAMI CONSERVANCY DISTRICT OF OHIO ." BY CHARLES H . PAUL, C .E .

Chief Engineer, The Miami Conservancy District, Dayton . Ohio .

EARLY HISTORY .

PROBLEMS of protection from floods were encountered with the first occupation of river valleys in prehistoric days .

The ancients exercised a measure of control over the Euphrates, the Tigris, and the Nile, by levees and by deflecting parts of the flood waters into depressions in the desert .

These early works were tied up with irrigation works to a large extent, and it is hard to say which, in those days, was considered most important .

As civilization developed and property increased in value, the problem of flood protection has assumed more and more impor- tance .

European engineers, in the early days, studied methods of controlling the periodic overflow of their rivers, especially in the rich broad valleys of central France and Germany, and as early as the year 171 1 retarding basins were used for flood control purposes .

At that time, two rubble masonry dams were con- structed across the valley of the Loire River in central France .

The upper one, at Pinay, has its crest about fifty-six feet above low water .

At river bed elevation there is an opening or vertical slot about sixty-four feet wide, reaching the full height of the dam . About four miles downstream is another similar dam built to supplement the controlling action of the first .

These two dams, which have been in use for more than two hundred years, are still in operation, and records show that they have justified their existence during many large floods on the Loire River, particularly those of 1790, 1846, t8g6, 1866 and ¶907 .

Many other dams forming retarding basins for flood control have been built in France, Germany and Austria .

Some of these are for flood con- trol only, and some for the combined purpose of flood control and * Presented at a joint meeting of the Institute and the Philadelphia Sec\ tion, American Society of Civil Engineers, held Thursday, March 13, 1924 . 162 CHARLES II .

PAUL .

IJ . F . I .

storage .

A list t has been published, giving some of the better- known retarding basin projects for flood control, in actual use, up to the year 1920 . That list shows forty-five in Europe, one in India, and five in the United States, not including the works of the Miami Conservancy District .

The attempts at flood control by levees in China, dating back to ancient times, are too well known to require more than passing mention . Levee systems have been used in many places in Europe, particularly along the Seine, the Loire and the Rhone rivers in France .

There are 322 miles of levees along the Po River in Italy, which was the first river in Europe to be leveed . There is no doubt that along with the early levee systems for flood control, a certain amount of channel enlargement was carried on to accom- plish the same purpose . Here again, some of the channel improve- ment and levee systems were built exclusively for flood control, and some for the combined purposes of flood control, and channel regulation, or improvement for navigation .

In our own country the work of the Mississippi River Com- mission is the outstanding example of flood control on a large scale by levees . A combination of levees and dredging is a com- mon solution of flood problems in many parts of the United States, and while retarding basin control had not been practised extensively in this country before the works of the Miami Con- servancy District were built, still there were several relatively small retarding basin projects which had been in operation for a number of years .

THE :913 FLOOD .

The flood of March, 1913, in the Miami valley, was not only the most severe of which there is any record in that valley, but, as regards damage, was the greatest that has occurred in the eastern half of the United States since the days of first settlement .

A complete description of this flood and the damage which it wrought has been published in one of the Technical Reports of the Miami Conservancy District .

2 The flood was caused princi- pally by hard rains which commenced on March 23rd, and contin- ued with scarcely any interruption until the 27th . As a result of 'The Miami Conservancy District Technical Report, Part 7, "Hydraulics of Miami Flood Control Project," p . 49 .

'Technical Report, Part I, "The Miami Valley and the 1913 Flood ." Aug ., 1924 .1 CONTROL OF RiVER FLOODS .

1 6 3 previous rains the soil was saturated, and this increased the runoff to the extent that during the latter part of the storm the runoff was too per cent . of the rainfall .

The drainage area of the Miami River system is about 3600 square miles, of such size and shape that a heavy storm of intense rainfall may centre over it, and tinder certain conditions may result in a runoff practically equal to the rainfall, as was the case in 1913 .

In the building of cities, railroads, and bridges, and in locating improvements near rivers, the usual high-water stages are taken into account .

It has frequently been assumed that past floods, with perhaps a little estimated increase, are a reliable criterion of what may happen in the future .

The fallacy of this reasoning for districts where records are available only over short periods, was well illustrated in the Miami valley during the flood of 1913 .

Levees and bridges had been built there to accommodate the largest flood that had occurred during the forty years or so that records were available .

The water in Dayton during the crest of the 1913 flood stood about six feet higher than the tops of those levees .

Not a bridge across the river was passable, many of them were washed out entirely . Nearly four times as much water came down the river as the leveed channel could carry .

Large parts of the business and residential districts of the city were overflowed to depths up to twelve feet .

Similar conditions existed in the other cities and towns throughout the valley .

The loss of life is not definitely known, but has been esti- mated at about 400 .

Property loss has been estimated at about $too,ooo,ooo .

It is not easy to make a reliable estimate of loss in such cases .

The indirect losses, from a disaster of this sort, are often fully as important as the direct losses .

Subsequent deaths result from exposure and shock, or in other cases health is permanently broken . Depreciation of property values, inter- ruption of business, loss of customers, loss of time and energy in cleaning out mud and water, are hard to translate into money values .

Many buildings were entirely destroyed ; in Hamilton 200 residences were washed away and carried down the river ; in both Dayton and Hamilton the flood was accompanied by fire, and many buildings were burned .

Large areas of asphalt street sur- faces were peeled off and carried away ; tiled floors inside store 164 CHARLES H . PAUL .

[J .F .1 .

buildings were torn up and destroyed ; sewers, gas mains and water pipes were filled with stud ; in some places where the current was swift, streets were washed out several feet deep ; in other places gravel and debris were piled up on the streets and in the yards to a considerable depth . Every railroad bridge, on a Too- mile stretch of river, was wholly or partly destroyed, and more than half of the highway bridges were also taken out .

The water of the river, during the flood, carried a heavy load of silt .

As the flood waters broke in windows and doors, and flowed through buildings, the current was checked inside the rooms, and much of the silt settled to the floors or in the cracks .

This process continued as long as the flood lasted . As a result, every crack and opening in floors, walls, or furniture, was filled with this sticky, slimy mud . On the floors of the houses, it lay from a few inches to a foot or more deep .

ORGANIZATION OF THE MIAMI CONSERVANCY DISTRICT .

Relief committees were organized in the different cities of the valley immediately after the flood, and plans for the prevention of future floods began to be formulated . Within sixty days after the flood had passed, and while people were still overwhelmed with the problems of their own personal losses, a fund of more than $2,ooo,ooo was subscribed, in the City of Dayton alone, for the purpose of making investigations and studying the problem of flood prevention for the city . Each community worked inde- pendently at first, the thought being that the desired results could be accomplished by channel enlargement or relocation, and levee improvement . It was not long after the beginning of a systematic Study of the problem, however, that it became apparent that channel improvement alone, to the extent required, was not prac- ticable . The next thought was for retarding basins . Fortunately, there were suitable sites, along the upper reaches of the rivers, where retarding basins could be formed at not too great expense .

It was apparent, however, that while channel enlargement by itself would not be sufficient, still a large measure of relief could be obtained at moderate expense by cleaning out bars and islands, strengthening levees and raising them a limited amount, improv- ing channel alignment, and paving slopes where required . A cer- tain amount of this work would give better returns for the money Aug ., 19241 CONTROL Or RIVER FLOODS .

1 6 5 spent than if retarding basin control were used exclusively .

The final solution of the problem, therefore, was a combination of channel improvement and retarding basins, so adjusted as to capacities as to give the necessary control at the lowest cost .

This naturally pointed to the fact that one city alone could do little along this line by itself, and that it was a job for the people of the whole valley to undertake as a unit .

Then followed prob- lems of organization, securing public support, questions of legis- lation, adjustment of conflicting interests-problems in human engineering, which were no small part of the whole big job of securing the necessary flood protection .

Under the Ohio laws there was no practical way for the people of the valley to organize for this particular purpose, and it became necessary therefore to enact legislation which would provide for such organization . The Conservancy Act of Ohio, passed in 1914, was the result of this effort, and paved the way for the organization of the Miami Conservancy District, which was effected soon thereafter .

ENGINEERING STUDIES .

In the meantime, while the necessary legislation was being prepared, the engineering investigations were going on and data were being collected and recorded, leading up to the studies for this comprehensive plan .

Fortunately, the engineers were called in only a short time after the flood had passed, when high-water marks and certain other precise records were still obtainable . The vast quantity of data collected included typical cross-sections of old channels ; levee profiles ; cross-sections of the valleys at con- trolling points ; surface slopes (luring peak of flood ; size of bridge openings ; time of flood crest at different stations ; velocity of flood flow ; runoff data ; rainfall data ; information on past floods ; topo- graphic surveys . The main object, of course, was to determine the peak flow, total discharge and high-water marks of the 1913 flood, and, by a study of past floods as related to rainfall and runoff conditions in the Miami valley and elsewhere, to determine what maximum could be set for future possible floods in this locality .

Corresponding studies and computations were started as soon as enough data had been obtained .

These included studies of flow in open channels ; determination of the value of " n " in the Kutter's formula, tinder various conditions ; computations of peak 1 6 6 CHARLES H .

PAUL .

[J .

h . L flow and total discharge, at various points, during the 1913 flood ; study of shape, size and character of drainage area ; determination of maximum possible flood ; study and investigation of various methods of flood control ; estimates of costs of various flood control plans .

As soon as the retarding basin scheme became a possibility, other questions presented themselves, among the more important of which were the determination of the location, size and number of dams ; tentative channel capacities through the cities ; determination of height of dams and size of conduits ; back- water effects and curves ; spillway capacities ; balancing of retard- ing basin capacities, outlet capacities, and channel capacities ; and finally, a determination of the amount of public and private prop- erty affected ; valuations of properties as to damages and benefits ; preliminary estimates of cost .

PREPARATION OF PLAN, AND APPRAISALS .

There is no place, in a paper of this sort, for a discussion of the details of handling these various features of preliminary investigations .

They are mentioned simply as a suggestion for those who have similar problems to undertake in the future .

By the time the necessary legislation had been passed and the perma- nent organization of the District effected, the engineering studies had proceeded to the point where a complete plan could be outlined, indicating exact methods to be used in securing the necessary flood protection, together with volume and character of work required, types of structures, and estimates of cost . This so-called " Official Plan " was presented to the Conservancy Court by the Board of Directors of the District, as required by law, and was approved by the Court during the fall of tgi6 .

The next step was an appraisement of damages or benefits to property affected by the proposed construction .

A Board of Appraisers, appointed by the Conservancy Court, prepared an appraisal roll of damages and benefits which was presented to the Court for approval and set for hearing during the summer of 1917 .

There were about 70,000 pieces of property affected .

In assessing benefits, the value of the property and the extent of protection required were taken into consideration .

For example, a property subject to flooding ten feet deep was assessed a larger percentage of its value than one subject to only three-foot flood- Aug,I924 .1 CONTROL OF RIVER FLOODS .

Ib 7 ing .

That is to say, each property was assessed in accordance with the degree of benefit to be received .

The appraisal roll was approved with so few exceptions that the security for the bonds, as represented by uncontested appraisals, was ample to justify proceeding with the financing of the project and preparing for the construction of the works .

DESCRIPTION OF PROJECT .

Fig . r, a map of the District, shows the river system of the ;Miami valley, and on this map are indicated the principal features of the flood control project .

It will be seen that the three main branches of the river, the Stillwater, Miami and Mad rivers, and another less important tributary, Wolf Creek, all come together within the limits of the City of Dayton .

Loramie Creek joins the Miami at the upper end of the valley just above Piqua ; Twin Creek conies into the Miami River between Franklin and Middle- town ; Four Mile Creek joins the river just above Hamilton . The drainage area above Hamilton is 3600 square miles ; above Dayton it is 2600 square miles .

The flood control plan provides for channel improvement, through all the cities and towns effected by floods, to the extent that it is economically feasible, supplemented by retarding basins which will hold back the crest of the flood, by restricting the flow through the darns, and backing up the surplus water temporarily in the basins .

The outlets through the dams are not controlled by gates, but are designed of such size that their combined discharge under maximum head (full basin) will not overtax the capacity of the improved river channels through the cities below .

Pro- vision is made for a maximum flood 40 per cent . greater than the 1913 flood . Channel improvement is provided at Piqua, Troy, Dayton, West Carrollton . Miamisburg, Franklin, Middletown and Hamilton .

Dams forming retarding basins are located at Lock- ington on Loramie Creek ; at l nglewood on Stillwater River ; at Taylorsville on the Miami ; at Huffman on Mad River ; and at Germantown on Twin Creek, In addition to building the darns and the channel improvement w orks . it was necessary to relocate about fifty miles of railroad lines, in order to get them out of the way of the dams and retard- ing basins .

New locations satisfactory to the railroad companies were made at the expense of the District, under the general direr- 168 CHARLES H . PAUL .

Fro . r .

STATE C{ CH10 TFIE MIAMI CIXISER/PNC( DISTRCT GENERAL MAP ~ •S MOWINO .- FLOOD CONTROL WORK AND STEAM AND ELECTRIC RAILWAY RELOCA1U4S SCALE CE MILES 2 1 0 E 4 8 8 m .,vm+a LEGEND 996d - RETARDINO BASIN MIL DAM, IMMMO RIVER IMRROVEMENT - RAILROAD RELOOATU'I [J . F . i, Aug ., 1924 .] CONTROL OF RIVEIt FLOODS .

tion of the railroad companies .

The old lines then became the property of the District and the rails and ties were salvaged .

Most of the railroad relocations were done by contract .

The other work was done by a construction organization built tip and equipped by the District for that purpose .

CHANNEL IMPROVEMENT .

The combination of channel improvement and retarding basin control permitted the channel enlargement to be confined to moder- ate limits through most of the cities . At Hamilton, however, the channel had been so severely encroached upon by industrial plants FIG .

2 .

I 6o Chom p ,on Coated Pope, Co .

1973 high water line, elevE070 lewewo 0101 IMBE 1973 flood, protection works built -el 588 .5 _i R :

8 a sting wo17 top of lcvoe'l proposed""l .

059 I flexible cent, revetment 8ft wide I surface I blech e1 .595Z 150 f t --~I ~--- - 540 ft '~ Cross-section of channel at Hamilton, showing original conditions and pr\ oposed improve- ment . This view shows how the river channels were encroached upon in places \ by industrial ptsars . It shows the standard cross-section of improved channel, with a standa\ rd levee at the left and a retaining wall at the right, where space for a levee is not a\ vailable . The two heavy dotted lines show the net effect, at Hamilton, of the flood control work\ .

that a considerable widening was unavoidable .

This work at Hamilton involved considerable property damage .

This is illus- trated by Fig .

2, which shows a restriction of the channel by two paper mills on opposite sides of the river less than 450 feet apart .

The peak flow at Hamilton, during the 1913 flood, was 350,000 second feet, and at this point the water was backed up to a depth of about forty-seven feet and stood above second floor level in both buildings . In providing for the improved channel it was necessary to purchase one of the mills outright, and increase the width of the channel materially, in order to provide the neces- sary relief . Fortunately there were not many places as bad as this .

Fig .

2 also shows the standard cross-section of the improved channel .

It has a low-water channel about i5o feet wide, with flat slopes or beaches on either side extending out to the toes of Vos . .

JA No . 1184 - 13 Sterling Paper Co . 17o CHARLES H .

PAUL .

[J. F . I .

the levees .

The low-water channel is located in the centre where the river is straight, and near the outside of the bends where the channel is curved .

The side slopes of the levees are 2 to 1, and where necessary because of high velocities, the slopes are paved with concrete .

Tn places where there is not room for the levee slopes, concrete retaining walls are substituted .

Dragline excavators were used on all the larger features of the work, not only in making channel excavation through the Flc 3 .

Floating equipment on channel work at Dayton . The dragline machine, mounted on a spud scow, loads the large scows which are towed to an unloading point n\ ear the waste bank, which in this case is the point of land at the junction of the two river\ s . The other machine, on shore near the end of the concrete bridge, unloads the scows and plac\ es the material in the waste bank . This view shows also a section of completed levee at the right .

cities, but also in excavating for outlet structures for the dams, and in the borrow pit excavation for the embankment material at the dams .

The District used twenty-one of these machines vary- ing in size from the small machine having a 30-foot boom and Y 4 -yard bucket, to the larger machines with ioo-foot boom and 5-yard buckets .

These largest machines, by using a smaller bucket, can handle a boom as long as 135 feet .

In many places the dragline machine placed the material excavated from the chan- nel directly into levee .

Where this could not be done with one throw, it was economical to move the material two or three times, with a dragline machine, between excavation and levee .

At Dayton and Hamilton, however, the channel excavation Aug ., 1 Q24 .1 CONTROL OF RIVER FLOODS .

[71 was so much in excess of levee requirements that the large part of the excavated material had to be wasted .

The nature of the channel at Hamilton was such that a construction track could be placed within the river channel, but out of danger of moderate floods .

The waste material was loaded into t2-yard dump cars and hauled to waste banks by 40-ton standard gauge dinkey loco- motives .

At many points on the Dayton work, it was imprac- ticable to place tracks within the river channel, as even a small rise would wash them out . It was possible, however, to use scows for transporting a considerable part of the waste material there .

A part of the work was excavated by one of the large dragline machines mounted on a spud scow, and in other places the scows were loaded by a machine on shore .

The loaded scows were towed to an unloading point near the waste bank, and were unloaded by another dragline machine which placed the material directly into the spoil bank .

On the channel improvement work at the various cities and towns, three general methods were required . First, that in which channel excavation is the essential or most prominent feature, as %as the case at Dayton and Hamilton .

Second, that in which the work is confined almost entirely to levee construction, as was the case at West Carrollton, Miamisburg, Franklin and Middletown .

Third, a combination of the two, as at Troy and Piqua . At places where levee construction was the principal item, and where the channel excavation did not yield enough material to build the levees, the levees were usually built by dragline machines, the necessary material being taken from borrow pits alongside .

In a few cases it was more economical to build sections of the levees by teams .

THE DAMS .

At each of the dams it was necessary to build a construction camp to take care of the required number of employees and their families, as there were no towns close enough to any of the dams to provide the necessary accommodation .

These construction camps consisted of a number of cottages for the accommodation of employees and their families ; bunk houses and dormitories for the laborers and mechanics ; a mess house to accommodate several hundred men ; a general store ; club house ; shops and warehouses ; water supply and sewerage connections to each building ; electric lights, and other conveniences of a well-managed up-to-date camp . 172 CHARLES H . PAUL . I 1 .

F . 1 .

Preliminary investigations and examinations of foundation conditions dictated the selection of earth dams at all of the sites .

Borrow pit investigations indicated that hydraulic fill dams would be practicable and economical, therefore hydraulic fill dams with massive concrete outlet works and spillways were determined upon at each of the sites .

The first important item of construction at each dam was the building of the outlet structure, or at least enough of it so that FIG . 4 .

Outlet and spillway structure at Taylorsville . The side walls and floors were built first, leaving the gap unobstructed, to carry flood flows dung construction . When the dam embank meet had been nearly completed, the cross-dam was built between the wall\ s leaving conduit openings of proper size at the bottom . 'rhe top of the cross-dam, left lower than the work on either side, forms the spillway . These conduit openings, at Taylorsville, are each fourteen feet wide and twenty feet high .

the river could be diverted from its old channel to permit the construction of the dam .

' Suitable rock foundations for the outlet and spillway structures were found at all of the dam sites . Two types of outlet structure were used . One type is illustrated by Fig . 4, which is an opening through the dam formed by two heavy concrete retaining walls facing each other with concrete floor between .

That part of the structure, being built first, gave an ample waterway for floods during construction periods . After the embankment had been built up to the point where there was no further danger of overtopping in case of a flood, a cross-dam was Aug ., tgzd .1 CONTROL OF RIVER FLOODS .

1i 3 built between the two retaining walls and through the bottom of this cross-dam, at river level, the conduits for permanent flood control were constructed .

The top of the cross-clam was placed twelve to fifteen feet below the top of the dams, thus forming the spillway .

Two of the dams, however, were so high that the construction of retaining walls to sustain the hydraulic fill embankment was prohibitive .

At these places, therefore, twin conduits were used Flu . 5 .

4 Cross-section of outlet conduits at Germantown dam- The opening at the r\ ight shows the original construction for handling flood flows while the dam was being b\ uilt . When the dam embankment had been nearly completed, the bottoms of the conduits were f\ illed in and floored over, as shown at the left, leaving the openings of proper size for perm\ anent flood control . This also shows the location of construction joints in the conduit arches .

extending through the base of the clam at river-bed elevation .

In order to secure additional capacity to care for floods during con- struction, these conduits were at first made deeper than was required, and as soon as the embankment had been completed, the bottoms of these conduits were filled in and floored over, leav- ing the opening of the size required for permanent flood control .

Fig . 5 shows the details of this type of outlet structure, and illus- trates how the additional capacity was obtained temporarily for stream control during construction .

This illustration shows also the construction joints in the conduit arches, the interesting feat- 1i 4 CHARLES H . PAUL .

ure of which is the pair of joints near the crown of the arch .

The "keystone" section between these two joints was poured after the concrete on either side had set, and this method proved effective in preventing cracks along the crown of the arch, which are so common with this kind of construction .

At the two lop of hmd wall FIG . 6 .

elee7500 -- -2,3'-0"- ---a - 0" ° tap oF~ainlno rvoil-ele¢ i , 7- a0 1 .

tan ofportiiiar wall-elev 1290, j .t-47230 ' 7/8/0 -` elee722 lea e -v . 7220 clev 716 Stilling pool at outlet end of conduits .

A deep pool, upstream from the main weir, forms water cushion . which causes the hydraulic jump to form on the inclined floor . The main weir also helps to spread out the stream, and to block the formation of eddies and back cur\ rents .

The second weir completes the spreading of the stream . Destructive velocities are effectively reduced by this means .

dams where these conduits were used, the spillways were sepa- rate structures .

At the downstream end of each of the outlet structures a stilling pool was built for the purpose of dissipating the energy of this swiftly flowing water (Fig .

6) .

This structure is so designed FIG . 7 .

[J . F . L cu of rich "- pervious core The hydraulic fill process . The discharge pipe from the dredge pumps lays down the material along the outside slope of the dam . The coarse material remains in the outside . The fine sand and silt flow down the beach to the pool, where they settle th\ rough to form the impervious core . The damis built up in layers varying in thickness from two to four feet at the different dams .

that the hydraulic jump or standing wave is formed within the limits of the concrete structure, and the floor and walls are made of sufficient thickness to withstand the action .

The velocity is thus reduced, before it leaves the structure, to an extent that it will not erode the unlined channel below .s t' o" - `The theory of the hydraulic jump and back-water curves is discussed in Vol . 3 of the Technical Reports of the Miami Conservancy District . Aug ., 1924 .

1 CONTROL OE RIVER FLOODS .

At each of the dams a gravel washing and screening plant was erected for furnishing material for the concrete work . These plants were standardized and all parts were interchangeable . Suit- able sand and gravel for concrete was found near each site .

The material was hauled to the plant in dump cars and dumped into a hopper at track level, from which a belt conveyor took it to the top of the plant and delivered it to the screens .

Wash water was delivered to the screens by a small centrifugal pump . The material FIG . H .

Placing hydraulic fill . The discharge line from the dredge pump is extended, length by length, as the layer of fill is brought forward . The coarser of the material remains near the outside slope, the finest particles being carried with the water to the \ core pool, where it slowly settles end consolidates to form the impervious core .

a was separated into three sizes, sand, fine gravel up to f % inches, nd coarse gravel from r% to 3 inches in size .

Each of these sizes passed by gravity into separate storage bins .

Oversize material could be wasted or diverted into the coarse gravel bins .

In the heavy walls it was the practice to use oversize up to about 6-inch cobbles .

A one cubic yard mixer was placed in such position that the material from each of the three bins could be fed to the charging hopper by gravity through adjustable measuring boxes .

Cement was stored in a shed near by and was brought to the mixing platform on small warehouse trucks as needed .

The mixer 1 7 6 CHARLES H . PAUL .

[J . F . 1 .

discharged into side dump cars, or bottom dump buckets on trucks, which were hauled to the work by 3-ton narrow gauge gasoline locomotives .

As none of the borrow pit material lay high enough to be sluiced direct into the dams, it was necessary at each of the sites to lift the material into the embankment by means of dredge pumps . In some cases, it was hauled to the dredge pumps in dump cars which were loaded in the borrow pits by means of dragline Fin .

y .

Ta loville dam under construction . The discharge pipe, in the foreground . takes the material from rs the dredge pump to its place in the embankment of the dam . Friction loss in these pipes vanes from six to fourteen feet per hundred, depending on th\ e quantity and character of the material handled .

A booster pump was necessary to put the material out to the far section . The lower section of the dam a the closure section .

machines ; in other cases it was broken down by hydraulic giants in the borrow pits and sluiced direct to the dredge pumps .

Regardless of how the material was brought to the dredge pumps, the process from there on was practically the same at all of the dams .

The hydraulic fill process automatically breaks up the borrow pit material into its component parts, and distributes the material in the dam in such manner that the coarse material is left in the outer portion of the dam, forming stable outer slopes, and the fine material is concentrated in the centre where it settles through a central pool or " core pool " to form an impervious core .

The discharge lines from the dredge pumps are laid close to the Aug ., 14)24 .1 CONTROL OF RIVER FLOODS .

1 7 7 outer slope and parallel with it, and as the material is discharged from the end of the pipe, the gravel and coarse material drop there and remain, while the fines are carried on, with the water, to the central pool, where the core section is formed by slow deposition of this fine silt through water .

Fig .

7 illustrates the process and also shows a typical cross-section of the conservancy dams .

Fifteen-inch dredge pumps were used at all of the dams . These pumps would handle single stones twelve to fourteen inches in diameter, but it was found that a number of these stones coming FIG . 10 .

Germantown dam completed, showing hydraulic jump in action during small \ flood .

Velocity reduced from twenty-five feet per second at conduit outlet' to SIX feet per second at outset of stilling pool .

along together were likely to cause trouble, so it was arranged to reject, at the sump, all stones that would not pass through a seven-inch hole .

This oversize was set aside for use as paving or riprap material .

PROGRESS .

Construction work began during the season of 1918 .

The intention was to push the work with all possible speed consistent with the economy, because flood protection for the valley was of vital importance .

All the larger jobs were started at once, and as far as practicable all work was carried on by two 1o-hour shifts .

The quantities handled were very large .

Not counting the 178 CHARLES H .

PAUL .

IJ . F . t .

railroad relocations, there were about 19,ooo,ooo cubic yards of earth and rock to he moved for dams and channel improvements, and about 250,000 cubic yards of concrete work in structures at the dams, and along the improved channels . The one item of hydraulic fill at the dams amounted to more than 7,000,000 cubic yards .

Some very high progress records were made, especially in connection with hydraulic fill . At one dam, with one dragline machine in the borrow pit and one dredge pump in operation, an average month's work was about 6o,ooo to 70,000 cubic yards and in one month 91,500 cubic yards were put into the dam . At another dam with four giants in the borrow pit and two dredge pumps, the highest monthly output was 107,000 cubic yards . At another dam with three dragline machines feeding two dredge pumps, the highest monthly record was i8o,ooo cubic yards ; a record of i 5o,ooo cubic yards per month was not uncommon, and during each of three consecutive years more than i,ooo,ooo cubic yards were placed in that dam .

Every effort was made to so arrange the work that partial protection from flood would be secured at the earliest possible date, and that protection against a maximum flood would be assured as soon as could be . During the season of 1921 the work was in such shape that a repetition of the 1913 flood would cause only a limited amount of damage, and by the end of 1922, the work was ready to handle any flood that could possibly occur .

On the tenth anniversary of the flood of 1913, the people of the valley were secure in the knowledge that flood danger for them was a thing of the past .

FIRST TEST OF COMPLETED WORKS .

Soon after the work had been brought to such stage of com- pletion that a large flood could be handled, a storm occurred which gave the first real test . On April II, 1922, a severe rainstorm began . which culminated on the 14th with very heavy showers .

The rain was general throughout the valley ; 1% inches fell in less than two hours in many places ; the total rainfall for the last twenty-four hours was 3% inches, and this fell on satu- rated ground .

Under the old conditions, as they were in 1913, a flood stage of eighteen feet in Dayton, the old danger mark, would have been Aug ., J024-1 CONTROL or • RIVER Fr .ooos .

1 79 reached or exceeded, and a bad flood scare would have resulted .

Actually, however, with the flood control works in operation, the river at Dayton only reached a stage of 9 .6 feet . This reduc- tion in gauge height was due to both the river channel enlargement and the backing up of water behind the dams . All of the basins stored water, the maximum depth at two of the dams being approximately forty feet, with lesser amounts at the other three dams .

At one darn the water in the basin was backed up for a distance of nearly five miles .

Everything worked out according to plan . The water was carried through the improved channels smoothly and swiftly, the familiar turbulent appearance in former floods being entirely absent .

At all of the dams the hydraulic jump worked just as expected, and only a short distance below each of the outlets the water was flowing smoothly, and with moderate velocity .

There was such intense interest by the public, in the working of the system, that it was necessary to place traffic men at one of the dams to keep the crowds moving . The storms of April, 1922, were widespread throughout the country, and many cony munities suffered loss of life and property due to floods, but the citizens of the Miami valley viewed their own situation with com- placency, feeling confident that future floods in their valley were absolutely under control . While it was apparent that the flood of April, 1922, was small compared with the maximum that the project was designed to handle, still it afforded a clear demon- stration of the effectiveness of the works, and it was apparent that a much larger flood would lie similarly handled with equal success .

The Number of Alpha-particles Emitted by Radium .

H . G rcER and A .

WERNEER .

(Zeit . f . Phys ., Vol .

21, No . 3 .) -When Geiger was working with Rutherford they calculated the number of alpha-particles emitted in a second by a gram of radium to be 3 .57 x 10 10 .

Hess and Lawson in 1918 published 3 .72 x to 70 as their determination of the same quantity . There is a difference of 4 per cent . between these values of a fundamental radio-active constant and in physics so great an uncertainty is intolerable . Geiger, now asso- ciated with Werner, turns again to the same problem, but he has a long experience in radio-active matters behind him and the resources of the Reichsanstalt at his command . In this investigation an alpha- particle is detected by the scintillation it produces on a zinc sulphide\ screen .

Two possible sources of error present themselves . Does every particle assuredly cause a flash of light when it strikes the