(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "History and construction of the aqueduct leading to McMillan Park, Washington, D.C."



The lifasaington Aqueduct was the first attempt in 
the city of Washington to obtain the city's water supply 
from one source. Previous to its installation the water 
supply had been obtained from nearby springs and wells. 
As the city grew in ovulation and spred out, this source 
became ./holly inadequate. Several locations were in- 
vestigated as sources of water for this new system. 

)tomao at Great Palls was finally decided upon as 
the best point from which to brii • .-or to the city. A 
dam was erected there to regulate the /rater f" : into 
the aque&uot. The details of th itory and. construction 
of the dam at this point will be found in Mr. Lozupone's 
thesis . 

is aqueduct itself was 9 feet in diameter 
lined with three rings of brick. In brin^in- the 
aqueduct from Great Falls to the Seor • :-voir, 

six bridg ;re necessary. Of these six , two are out- 
standing are the Rook Creek and the Cabin John 
bridges. r, he former is no Ion *er standing, but t 
latter is Btill in service supporting bot > conduit 
and a roadv/ay. It was for the fir t 40 years, the long- 
est single span masonry arc rorld, In ? . how- 
ever, it .. repl ced in this respect by the Luxembur ; 
bridge in "Cure bh a span of £77.7 feet. 



./i "tli the completion of the aqueduct in 1863, l, ir 
was supplied directly to the city with no filtering. In 
188£, Congress deemed it advisable to increase the water 

>ly and to accomplish this, the ./ashin.^ton Aqueduct 
Tunnel was constructed to carry Jaior from the Georgetown 
reservoir to a new reservoir to be constructed called the 
i c. iillan ar'-r reservoir, It was here that the first 
filtration plant, one of the slow-filter type, an 
located. Ilth the completion of this project, the City 
of Washington found themselves euuiped with a very ex- 
cellent (rater supply system. 



HISTORY AND CONSTRUCTION OF ' JDUCT 
LEADING TO MoMILLAB PARK, SASH IN GO? ON, D. 0. 

PART I 
HISTORY OF THE AQUEDUCT 

The aqueduct ultimately leading into LJcuillan 
Park, Washington, D« 0. begins at a dam across the Potomac 
River about one mile above Great Falls, L'aryland. The dam 
is located about eleven miles north of Washington. The 
reasons for constructing this new water supply system may 
be easily seen in studying the early history of water 
supply in the Capital. 

The City of Washington was founded in 1791. Its 
source of water supply was the usual source of tho-'*e days, 
namely, springs and wells. The city was located on a rela- 
tively low level with a result that at all times there was 
a generous supply of ground water. The movement toward 
better water supply is in several books and papers attri- 
buted to President Washington and L' Enfant. It is stated 
that these two spent considerable time in investigating such 
systems of water supply. However, in the records at Dale- 
carlia it states that such is not the case. The President 
at that time was completely absorbed in affairs cf state and 

i.iajor L 1 Enfant was interested in a water supply in connection 
with a large waterfall which he proposed for its scenic 
effect in front of the Capitol. There were several springs 



-2- 



and wells from which the population of the city obtained 
its water. The four largest and best known of these 
springs are as follows : 

(1) Old City Sprii 3rth side of G Street 

between 4-1 2 Street and 6th Street. 
(£) Gaffreys Spring — North side of P Street 
between 9th and 10th Streets. 

(3) Franklin Square Spring — Between I and 

Streets and 15th and 14th Streets. 

(4) Smith Spring— 

The water of Old City Spring was the first water 
to be piped in the City of Washington. It was piped in 
1802. The first pipes used when compared with the rubble 
and brick aqueduct were crude conveyors. They were mac; e 
of seasoned logs hollowed out in the center and joined 
together with connections made of wrought iron, ./ater was 
only piped in this manner short distances because of leak- 
age. The Government Buildings were the principal receivers 
of this piped water, although several houses on Pennsylvania 
Avenue received water in this manner. The Franklin Square 
Springs were the source of supply for the Hfhlte House and 
adjacent executive buildings. The Smith Spring supplied 
the Capitol and Pennsylvania Avenue. The pipes were layed 
by the people who expected to use the water. 

A relatively generous supply of water was 
available from the springs and wells around the Capitol 



(3) 



for domestic use for the first fifty years of its 
existence (1790-1840), and the system of supplying water 
was entirely satisfactory to the public. It was the 
government who first began to look for a better source 
of water supply. It did not think that the water piped 
from Smith's Spring in 1832 to the Capitol to protect the 
valuable records stored there from fire was a satisfactory 
means of protection although it had served its purpose for 
many years. There were several plans proposed for obtain- 
ing this water supply. One of the plans promised was 
"Skinner's Scheme", a description of "which may be found in 
Government Document Dumber 281, Inarch 8, 18 30. It is suf- 
ficient to say that it was entirely impractical but it was 
however the first plan to consider the Potomac River as 
supplying the entire City of .Washington. Two more plans 
were proposed by Robert i.iills who like Skinner was a C.E. 
a complete description of these projects are found in 
Government Document Dumber 544, April 3, 1830. His pro- 
jects were entirely logical but were not adopted, however. 
In 1850, Congress appropriated i;500.00 to enable 
the .<ar Department to make an examination of the most 
available mode of supplying the City of Washington with 
pure water. It had been brought sharply to the attention 
of the people that even the larger springs am 7 wells could 
still be only used locally. It became increasingly evident 

that the city must find a more generous supply of water. 
The appropriation of 1850 was followed in 1852 by an 



(4) 



appropriation of v5,000.00 by Congress for "determining the 
best means of affording the City of . Washington and George- 
town with an unfailing and abundant supply of good whole- 
some water." The first appropriation of v 500. 00 was used 
to make a survey of the water requirements of the city. 
Both Great Palls and Rock Greek were given consideration 
as potential sources of supply. The survey was conducted 
under the leadership of Gol. George ./. Hughes of Corps of 
Topographic Engineering. The second report, following the 
i. 5,000.00 appropriation was made by Lt. Montgomery C. Meigs . 
The complete report may be found in Senate document Number 
48, February BE, 1853. This report contained the Meigs plan 

which consisted of circular brick cavity or conduit from 
Great Falls to Georgetown, a receiving reservoir formed by 
damming Little Falls Branch, a distribution reservoir, and 
a series of high service reservoirs to serve the points of 
higher elevation in the city. The State of Maryland granted 
its consent to this project by a special law on Llay 3, 1053. 
The consent was to take effect only when the approval of the 
Ghesapeake and Ohio Ganal Company had been obtained. This 
approval was obtained in June, 1853. The approval of 
President franklin Pierce was obtained on June £8, 1853. 
It was President Pierce who made the decision to use a nine 
foot conduit rather than a seven foot one. The main reasons 

for this decision was that the nine foot conduit gave a 
cross section area 1.6 times the area of a seven foot one. 
Congress in ^arch, 1853, appropriated ,.lu^Q00.00 for 






v . «#*£ »«^-**» 




This map shc%g zhe arainage areu. of Dig I otomac 
Gr lis. 



(5) 



commencing the work under Captain Meigs. Purveys were 
begun at once to secure the necessary property rights. 
The task of obtaining the necessary property 
rights before Captain Lleigs was longer, more difficult, 
and more expensive than he had anticipated. In a few 
cases the owners would not sell under any conditions, 
and all of the property had to be condemned and appraised 
by a special jury. One of the most difficult casos and one 
which caused a revision of the original plans was that of 
the Great Falls Manufacturing Company. This company main- 
tained that they were the "owners of the Great Palls of 
the Potomac with the tract of land thereunto belonging." 
This prevented the government from obtaining title to 
that tract of land, but it finally arranged to buy a small 
plot of land just above Lock Ho. £0. The result of this 
ohange in plans was that the conduit was shortened some 
2,985 feet. 



-6- 



PART II 
CONSTRUCTION OF THE AQUEDUCT 



A force of laborers and mechanics were collected 
at Great Falls in the early part of November, 1853. This 
was about eight months after Congress had given its approv- 
al to the project. The first operations were directed parti- 
cularly toward the construction of the feeder. It was that 
part of the acueduct which took the water from the Potomac, 
carried it under the Chesapeake and Ohio Canal, and into the 
control house. The feeder is only a few hundred feet long. 
Operations were also directed toward the opening of the 
first mile of work. 




! 








The Dam From The Maryland 
Side 



another view of the 
jam at tr aat Falls 




Louth of the 
Feeder 



-7- 




The Control House at Great Palls 
The force of three hundred to four hundred men 
continued working through that winter and the following 
spring. By June, 1854, practically all of the initial 
appropriation of $100,000 had been used and Congress had 
adjourned without making a second appropriation. ,/ork 
was suspended. However, by this time all lands needed in 
Liar y land had been condemned, a sandstone quarry had been 
bought in Seneca, seven miles above Great H'alls, and a 
small portion of the brick conduit had been constructed. 
The temporary suspension of construction was a 
very fortunate occurance for Captain Ileigs. The prelim- 
inary plans and estimates had been made in a very hasty 
manner. Construction could not have proceeded in an 
orderly manner in the shape they were. The year delay 
gave Captain Meigs ample time to get his plans in better 
shape . 

,/ork was resumed in Liar oh, 1855, immediately 
following an appropriation of ^250,000 by Congress, and 
contracts for grading and culverts were advertised. The 
law at that tine required a period of 60 days before the 



-8- 



bids could be opened. This held back that phase of con- 
struation considerably. Construction was again suspended 
in the spring of 1856. This time it was not due to the 
lack of appropriations by Congress because it had just 
appropriated another $250,000. The suspension was due to 
the fact that the law made the money available for exist- 
ing liabilities. This discontinuance was a serious handi- 
cap since contracts were suspended, men laid off, and 
construction subjected to deterioration. 

This state of affairs was brought to the attention 
of Congress and as a result the first large appropriation 
of Oi, 000, 000 was obtained in 1857. This appropriation was 
followed by another in 1858, for ..;800,000. As a result of 
these two large appropriations, a period of activity last- 
ing two years followo . „t the beginning of this period, 

88 feet of the conduit had been built and work on the 
bridges and reservoirs had not even been started. .Vhen this 
period of activity ceased in June, 1859, due to lack of 
funds again, 98% of the conduit was finished and Cabin John 
Bridge was the only large item of the work unfinished. 

..hen work was closed down in 1859, it remained 
that way until June, 1860, when Congress appropriated 

'00,000 to complete the project. It was at this time 
that iieigs was removed by the Secretary of i¥ar until Gept ember, 
1861. Iieigs was replaced by Captain iBonham. i.Iany 
interruptions occurred in the year 1861. i i the first 
r of the Civil ./ar. Lost of the laborers working on the 






Aqueduct were transfer ed to the lingineering Division of the 
Army and were employed erecting entrenchments on Virginian 
shores. 

In June, 1862, the affairs of the Washington 
Aqueduct mrejB turned over to the Interior Department, and 
Lleigs was transfered to the regular army as Quartermaster 
General at Washington. At this time the conduit was 
practically finished, Cabin John Bridge was half completed, 
and the dam at Great Falls was not even started. 

jater from the j-'otomac was first introduced into 
the conduit on December 3; 1863, and on December 5 allowed 
to flow into the receiving reservoir. It showed remarkable 
water- tightness in most places. The conduit was again 
closed on January 14, but following that it was not closed 
again for twenty-seven years {September, 1891). The con- 
duit has been drained and inspected at frequent intervals 
though since 1896. 

The conduit is a circular brick ring with a 
0.18, j grade or a fall of 9 inches per 5000 feet. A CO' 
of the specifications of conduit and materials ii ven 
below to explain the construction of the aqueduct. 

Specifications for Conduit: The conduit will be 
circular and generally 9 feet interior diameter; it will 
when of brick generally be built of three separate 4 l/£ 
inch rings of hard brick, ./hen it passes through the ground 
risings as high as the intrados of the arch, the inner r 
will sometimes be omitted, and the brick work reduced to 
nine inches. 



• 10- 



In rook cuts and other deep cuts where suitable 
stone Is on hand, concrete or rubble stone masonry will be 
substituted for brick, in whole or in part, particularly in 
the lower or reversed arch. 

The centur of the conduit at any point is the 
grade of the conduit at that point. The excavation and 
embankment will be made to the level of the grade at each 
place of operation before the contractor for the conduit 
will be allowed to commence his work there. The excavation 
for the lower semi- circle or reverse arch, however, will 
generally be made by the contractor for the conduit r>.nd 
it will be trimmed out In advance of the lay in the 
masonry. 

Brick IJasonry in Conduit : The bricks, at the 
time of laying will be thoroughly wet; every brick must 
be laid and pressed down into a full bed of mortar, which 
shall cover its bed and joint; and this bed shall be 
done at one operation for each brick so that no mortar need 
be washed in after the brick is placed. The inner edge of 
the joint of each course will be the least possible to 
admit of mortar between the bricks. 'he joint of mortar 
between each two rings will not be less than 3 8 inch in 
thickness. 

.. jne katsom.-^ of Conduit : This will be made of 
small rubble stone, none of which will exceed 14 inches in 
width or depth. Each stone will be laid in a full bed of 
mortar and hammered until the mortar is pressed out at the 



-11- 



front and the joints are completely filled. Its inner 
surface will be well plastered with a coat of cement mortar 
and be floated smooth and even. 

Concrete : One barrel of cement (300 lb. net) with 
two and one-half barrels of sand and 30 cubic feet of stone 
will make a batch of concrete. The mortar having been spread 
on a bed of plan!:, the broken stone will be spread evenly 
over it and the whole mass turned over twice and thoroughly 
mixed with a hoe or shovel. Hien nixed ana laid it should 
be rammed into a compact and water-tight mass. (Hot : 

r3e proportions are roughly 1: 2 1/fe : 10 which is much 
weaker than used at present) . 

i,. or tar : The mortar for masonry will be mare of 
two and one- half parts of sand to one part of cement. The 
sand and cement after bein.R measured, will be mixed dry 
and small quantities only taken from the heap, will be mixed 
with water as required. (Note: The mortar for concrete 
was made first and the amount of 'water was kept lovv.}. 

Cement : All cement must pass the test of settiD 
hard under water ano not breaking into lumps. This was of 
course natural cement as Portland cement was not made at 
that time. It came from Maryland, New Jersey and Pennsylvania, 
and was delivered on the dock in Georgetown. The inspectors 
tested small specimens which were considered good if the set 
occurred in 30 minutes or less. 

Sand : 1: me but good clean shar .'lint sand will 
be received. Proposals will state the number of bushels 
the bidder ./ill undertake to deliver. (Uote: Contract price 



-12- 



was 6 cents a bushel, delivered on the bank of the canal 
in prepared bins. Lower prices were later o competi- 
tion. ) . 

Br ick : None but well-made, hard, burnt brick:, 
entirely acceptable to the engineer, will be received. They 
must be made in moulds conforming in size to the municipal 
regulations of .Washington City. 

„1 materials such us cement, brick, sa: . orms, 
centers, etc., were furnished by the United States for the 
purpose of economy. Also to aid in economy the amount of 
materials for each unit of work were computed and if the 
contractors exceeded that amount, the value of the wasted 
materials were deducted from his monthly statements. Con- 
struction under contract was done on the unit basis cost 
throughout. 

rth Excavation ,0.15 per cu. yd. 

Hard Hocl: Excavation 1.25 

Soft Hoc]: Excavation — 0.75 

Lay in ■ rick 2.25 

Laying Stone Masonry o.OO 

rnishing Seneca ttoneO. 

Concrete in place 3.S5 

Finishing Brick 8.25 

The only machine for construction at that time 
3 the steam driven derrick. As a result tunnels were 
found to be cheaper than deep, open roc]: cuts, ;re were 
eleven tunnels from Great ^alls to Dalecarlia ranging from 
11 feet to 14o7 feet in length for a total length of 
feet. The tunnels were more or less irregular in shape and 
bained a circle 11 feet in diameter. It was the ori?*inal 
plan of Captain Meigs to line all tunnels with a 2 foot 



it It 


it 


rt n 


it 


II II 


ii 


n n 


•i 


ft. 

yd. 

lousand 



-13- 



lining of hrick. However he later changed his mind and 
decided that one foot was sufficient. ./hen he found that 
the cost of the entire aqueduct was much greater than his 
estimate, he decided to omit the lining entirely in the 
longer tunnels. As a result, five of the short tunnels 
lined with bric>: for a total length of 10L8 feet. The re- 
maining six tunnels were not lined at all. These tunnels 
held up in this condition fairly well for the first fifty 
years, but in 1891 when the first inspection was made, Col. 
Elliot noticed that several small rock; slips had occurred 
and blocked the passage to a depth of 4 feet. 

Lining of these tunnels was begun in 1911 under 
Col- Longfitt, Because the reservoirs could only supply 
the city for four days and since it would then take two 
weeks to fill them, the job of lining was the most difficult 
and expensive one connected with the aqueduct. fork con- 
tinued over a period of seven years at a cost of -.50,000 
and only 1,000 feet had been lined. k% this rate it would 
have taken 28 years to finish the ;'ob. This was not 
necessary however since a movement was in progress for a 
new aqueduct which would then enable them to drain the old 
conduit and make the necessary repairs. 






:',.'*' 










Dalecarlia 



-14- 



Ci IN JOHN BRIDGE 



The Washington Aqueduct had six hridges alto- 
gether. Two of the bridges were far greater in size and 
importance than the others. These were the Cabin John 
Bridge and the Rock Greek Bridge. The picture of Bridge 
No. 3 is an example of the smaller bridges. The Rock Creek 
Bridge was the only one in which arch ribs were utilized 
to convey the water supply of a city and at Lhe same time 
support a roadway. How, ] owever, it no longer exists. 
It carried its actual load for over half a century and 
probably now would still be in service if it had been wide 
enough. 



--»<_?* 



Bridge No. 3 






,/ork was begun on the Cabin John Bridge early in 
1857. All soil and rock formation was cleared or blasted 
away until a solid ledge of the hill was exposed. Here a 
proper face was made for the reception of the arch by a 
hydraulic bed of cement and broken stone. 

.s the superstructure was to be very massive, it 
required a. correspondingly heavy center or trestle to sup- 
port it during construction. The trestle rested in a series 
of stone piers which piers still remain. In addition to this 



-15- 



main trestle, a furhter framework was constructed to carry 
a system of traveling cranes by which means the stones were 
transported to various portions of the structure as it was 
used. All the granite and sandstone was transported on the 
Chesapeake and Ohio Canal, which was only 1,000 feet away 
from the bridge. The materials were carried from the canal 
to the bridge by constructing a dam across the creek near 
the canal and a lock to permit boats to pass from the canal 
up to the pool under the bridge. All of the abutment stone 
was secured from a uiiarry a few hundred feet up the valley. 

The bridge is a single arch bridge. In the first 
designs, Captain Meiga planned to have it a series of small 
arches. As it was finally constructed, a single masonry 
arch, it was the longest of its kind in the world for 40 
years or until 1903, when the Luxemburg bridge in Europe sras 
constructed. Its span was 277.7 feet or 57.7 feet longer 
than the Cabin John Arch which was 280 feet. 

The single arch is constructed of dressed granite 
from Quiney, Massachusetts, 4 foot 3 inches thick at the 
crown and 6 foot £ inches thick at the spring line. It is 
backed with a secondary arch or rubble arch composed of 
Seneca sandstone. The spandrel arches are rock faced ashlar 
with rubble backing. There are five spandrel arches at the 
west end and four at the east end. These spandrels are 
hidden by the vertical side walls for the sake of appearance. 
All of the remaining face work is Seneca sandstone. 

./hen the bridge was first placed in service in 



CABIN JOHN BRIDGE. 




a 



^^% 




LOCATION 
West Abutment 


MASONRY 

Co. yds 
2543/6 


CONCRE 

cu.yds. 
41343 


cu-yds. 


East Abutment 


308?./? 


45?7S 




Granite Arch 


933.63 






bubble Arch 


IS59J9 






Load 


300265 






Under Sprina Line 
Parapet Walk 


257.40 






Conduit thru Bridge 
Totat 


if ?/*./& 


25*66 


31G.00 

J 16. 00 



Radius of \nlfados ofGmmte Rina - 134 285*' 
Radius of Lxbrados of Granite Rw \45.z69s ' 



-16- 



December, 1863, several leaks were observed. The water was 

shut off and the lower half plastered. ;/hen the water was 

turned on again the depth was kept below the middle lev si 

and no leaks were observed. However, in 1892, when an 

inspection was made of the entire aqueduct, this leak was 

found to be in the invert of the conduit about 40 feet west 

of the bridge, but the water was flowing under the conduit 

and washing out the abutments. This leakage was again 

repaired by plastering but this was not permanent and before 

long conditions became dangerous. The danger was eliminated 
i 
by placing a metal lining of 501 feet of cast iron plated 

i - inch thick cast in the form of arcs of an 8 foot circle 

3 inches wide 'within the conduit. In this condition it 

■ 

exists today. 



m 





The Parapet of Cabin 
John Bri re 



The Arch ( Granite j 
of Cabin John Bridge 



-17- 



. fASHIEGTOH ' OTCT TUIT1JEL 
OR 

EL 



The Washington Aqueduct Tunnel is that part of 
the Washington Aqueduct which connects the Georgetown 
Reservoir and the McMillan Park Reservoir. In 1882, Uajor 
Lydecker was instructed to make such preliminary examina- 
tion and surveys as were necessary in preparing a project 
for increasing the water supply of Washington. The result 
of this report was that liajor lydecker recommended either 
a tunnel through solid rock in a direct line between the 
points or a surface conduit. He stated that the advan- 
tages of the tunnel far outweighed the advantages of the 
surface conduit. 

A trial shaft was driven at the eastern end of 
the tunnel because it was believed that here the rock bed 
was lowest. The shaft was 116 feet deep when it reached 
rook. As a result of this survey, Con-;res3 appropriated 
50, 904.55 to start work. Construction was begun in 18 
Several shafts were lowered and work begun in both direc- 
tions from these shafts. A picture of the present vie./ of 
the Rock Greek shaft is shown below. 



r' 


■f 


< 


- 







CROSS-SECTION of TUNNEL 



DRY STONE F*Cft\m A 
FfUBBU MASON*? FILL 




SCALE -l"^' 



NO.i 



CROSS-SECTIONS of TUNNEL 




SCALE' 1-4* 




uo.z 



uaa 



-IB- 



Operations were suspended in February, 1806, be- 
cause appropriations were exhausted. At that time the 
total length of excavation was 18,540 feet of which 14 C JQ 
feet were lined with masonry. The lining of all work be- 
fore February, 1886, consisted of a facing of brick masonry 
laid in hydraulic cement mortars with a backing of concrete 
for the invert and side walls and of dry stone for the 
space between the arch and the rock through whioh the 
tunnel was excavated. 

In August, 1886, Conress appropriated $395,000 
to continue the work but only after an investigation by 
the Board of Engineers as to the methods of doing the work 
especially the character of the lining. This board con- 
cluded in their report that the exclusive use of concrete 
for lining was inpracticable, and they indorsed the use of 
the brick arch but recommended that the filling between the 
arch and the rock be done "with rubble laid in cement 
mortar." This of course increased the cost of filling 
from $2.59 per cu. yd. to $4.76 per cm. yd. Under these 
conditions work continued until September, 1887, when a halt 
was called because of lack of funds. 

In March, 1888, Congress appropriated $355,000 
and stated that all work was to be completed by November, 
1, 1888. This was not accomplished, however, because in 
September, 1888, charges were made that the work was being 
done in an ear mariner by the contractors, ./ork ■.vas 
suspended and an immediate investigation followed. The 
its as reported show that systematic frauds had been in 



-19- 



progress in the construction and that numerous and extensive 
voids existed in the lining of the tunnel. 

On December 5, 1895, an expert co: on con- 
sisting of Major J. L. Liar shall, Captain James L. lusk, Mr. 
Alphonse F. Telay, Mr. Desmond " n itz Gerald, and Lt. D. 
Gaillard met in Washington to decide on the feasibility of 
completing the tunnel. They decided favorably and work was 
again begun on July 15, 1898 and the tunnel was completed 
in 1905. 

The length of the tunnel was £1,400 feet and the 
cross-section of it varies with the conditions as is she 
in the acoompanin^ sketches. 



The construction of the Washington was 
a major factor in the development of that city. Before its 

construction the water sup ly was so low that it was 
difficult to keep the people of the oity from tapping pj 
lines to government buildings. But since its construction 

bh the number of improvements made upon the sy -tern, 
it has been able to afford to the city of n/ashington an 
ample quantity of water. 




-cLlillan Park Reservoir 



[I 



The Military Engineer 

Annual Reports of the Office of bhe /ashinrton Aqueduct 

{ 188£ - 1905 ) 
Records of the Columbian Historical Sooiety 

i records at Dalecarlia which v/ere kindly furnished 

Br, Hardy. 
The City of Washington by Hewell