---

PART II
Main Thoroughfares

In considering the economical widths for the main thoroughfares of a city, so many complex factors are involved that no exact and indisputable conclusions can be reached; but there are certain facts and principles that ought to remove such decisions from the realm of purely arbitrary whim and custom by which they are now generally settled. Practically every normal main thoroughfare, even of the most compact type, must provide for car tracks in the middle. On straight runs, according to the present practice and with the new cars in Pittsburgh, the width occupied from the outside of one car to the outside of the other is 17 feet 8½ inches. At that, the cars are narrower than the modern standard in some other American cities, and the clearance between the cars is reduced to less than a reasonable requirement for safety. On curving roads, such as the Pittsburgh topography often imposes, the space occupied is greater. Without allowing any clearance on the outside, a space not less than 18 feet, and preferably more, should be allowed for the actual cars on straight runs.

In Pittsburgh, the gauge of the car tracks was originally made to conform to the prevailing local gauge of other vehicles, on the mistaken theory that it was desirable to have the smooth tracks used by wagons; and this has resulted in the almost invariable conformity of the wagon gauge to that of the tracks, regardless of the size or character of the vehicle. With the added fact that Pittsburgh pavements are prevailingly bad, and that the form of rail is such that it is very difficult for a wagon to turn out when it has once got into the track, the teamsters in Pittsburgh are more inveterate in the habit of driving in the car tracks, and less ready to turn aside for cars or other vehicles, than in most cities. The severe and constantly repeated strain of the horses, which is required to wrench heavily-loaded wagons free from the tracks, is, in the aggregate, a serious economic loss; and the delays not only to the street cars but to all forms of wheeled traffic, caused by the conditions described, are incalculably great. But even good pavements and the use of a grooved rail would not cure the trouble in Pittsburgh streets as now laid out, because, almost universally, there is not sufficient room for a vehicle to pass between the cars and another vehicle standing or slowly moving next the curb.

In every street, vehicles must be free to stop for loading and unloading, and on a busy thoroughfare the space next the curb is so much used in this manner as to become merely a series of sidings into which slow-moving vehicles can turn from time to time in order to clear the main passageway. The result of the conditions above described is that practically the whole wheeled traffic in Pittsburgh streets is inevitably concentrated on the eighteen-foot width where the cars run. The extent to which this reduces the average speed of travel and the total capacity of the thoroughfare has been strikingly illustrated for Pittsburghers by the contrast of the former sluggish congestion of traffic on Smithfield Street with the sparse appearance and rapid movement of the same traffic since the "one-way" regulations have made it possible to get one free line in each direction for moving vehicles separate from the cars. The same striking increase in capacity is to be secured, without the grave inconveniences and drawbacks of the "one-way street" regulations, where the space between the cars and the curb can be made wide enough for two lines of vehicles, instead of just enough for one or for one and a half, as is usual in Pittsburgh.

It is very difficult to determine just what is the most economical allowance of width. There is much variation in the widths of the vehicles themselves, and the necessary amount of clearance varies with the average skill of the drivers and with the effectiveness of the police control. The width of the line is plainly determined by the widest vehicles in it rather than by the narrowest. In Pittsburgh the customary width of the heavier and wider wagons is now controlled by the practical necessity of fitting the wheels to the railway gauge of 5 feet 2½ inches, and the widths are considerably less than prevail in New York, Boston, and other reasonably well-paved cities where the wagons are not fitted to the car tracks. About 7 feet over all is now the ordinary maximum in Pittsburgh, a few auto trucks and delivery vans exceeding that figure slightly, and an occasional three-horse team occupying over 8 feet. In New York and Boston, wagons measuring from 7 to 8 feet from hub to hub are common, and they sometimes considerably exceed 8 feet.

Just as in the case of steam and electric railway equipment, the tendency is constantly in the direction of heavier, longer, wider vehicles, for the sake of the operating economy due to large units; and, with the steady increase in the use of motor vehicles for business purposes, this tendency is likely to be greatly accelerated. There is every reason to expect that motor trucks will gradually increase in size until a limit is fixed by the public authorities in order to protect the pavements, and for the sake of standardizing the lines of travel in relation to the street widths. But, in the interests of economy of operation, the limit should be as high as practicable, probably not less than 8 feet.

If 8 feet be allowed for each vehicle, plus only a foot of clearance, the cars and one row of vehicles on each side, between them and the vehicles standing at the curb, would require 54 feet between curbs. A wagon backed up to the curb on a busy street will seriously discommode travel at that, and the clearance allowed is very small. A width of 54 to 60 feet between curbs is, therefore, highly desirable in the main thoroughfares.

As a matter of fact, with the widths of vehicles which now prevail in Pittsburgh, if standing and slow-moving vehicles are compelled to keep in contact with the curb, it is possible to keep open a line of travel on each side of the car tracks, with only occasional blockades, where the width between curbs is 50 feet, or, at a pinch, even 48 or 47 feet. That is to say, the difference in traffic capacity between a thoroughfare 50 feet from curb to curb and one 45 feet is enormous; while the difference between 45 feet and 40 feet is very slight.

Since a main thoroughfare is apt in time to become a retail trading street, wide sidewalk space is important. It is a common rule to make the distance of the curb from the property line one-third the width of the roadway.

A total width of 90 feet, with a 54-foot roadway and 18-foot sidewalks, is a satisfactory minimum for meeting the practical requirements of an ordinary main traffic street; a width of 100 feet is preferable, and 80 feet may be regarded as a rather niggardly irreducible minimum.

In this connection it is interesting to note the standard widths adopted in European cities. The standard in London is 48 feet[5] between curbs and 80 feet between buildings for secondary avenues, and 100 feet over all for principal arteries; and 140 feet over all is proposed for two great main arteries, the cutting of which, through the midst of the city, is being considered. In German cities of the second size, such as Leipzig, Frankfort and Hanover, the standards are as follows: for strictly local streets, 33 to 47 feet; for secondary thoroughfares, 50 to 80 feet, and for main thoroughfares, 85 to 118 feet. A Prussian law, in force since 1875, and apparently drawn up to meet the requirements of Berlin with its heavier traffic, requires the following dimensions for the laying out of new streets and for the alteration of old ones: local streets, 40 to 65 feet; secondary thoroughfares, 65 to 95 feet; main thoroughfares, over 95 feet.[6]

The above considerations apply only to the ordinary main thoroughfares of normal character. In most of the great cities of the world, there has been a considerable development of special thoroughfares of much greater width, including, for example, locations for transportation lines (surface or elevated), on separate rights of way decorated with trees; and including tree-shaded promenades and garden strips. These have usually been laid out in suburban sections before they were much built up; or, if within the built-up districts, on the sites of old fortifications, canals, or other abandoned engineering works. The latter opportunities are lacking at Pittsburgh, except in connection with the river banks. In the suburban localities of Pittsburgh, so much of the available building land is topographically divided into narrow strips that it would be cut to pieces in an exceptionally uneconomical manner by any boulevards, of the type usual in flatter cities, where a substantially uniform width of 150, or 200, or 300 feet is not infrequently carried through for considerable distances. As a general rule, any width to be secured for esthetic purposes in connection with Pittsburgh suburban thoroughfares, over and above that needed for handling the expectable future street traffic, must not be in the form of a general and continuous widening. But occasional pieces here and there may be taken for park purposes, as, for instance, a steep sidehill adjacent to the line and unavailable or difficult for building. Or a narrow ridge, on which the thoroughfare runs, may have at some point so little available building land fronting upon it that the whole can reasonably be parked for a short distance, thus keeping open the distant views.

There are two special forms of street, developed here and there in hilly cities all over the world, of which Pittsburgh needs to take account in its suburban development. In many instances, and for long distances, existing suburban thoroughfares that must be enlarged and improved, and others that must be laid out, are compelled to run along the face of hills so steep that a street of level cross section, even though limited to 80 feet, would leave the land on one or both sides so far above or below the grade as to destroy its value for building purposes. In such cases it is often practicable to make use of one-sided streets or two-level streets. The former are designed to give accessible frontage on one side only, usually the uphill side. The property on the opposite side is reached by the next street, which is laid out correspondingly nearer in order not to make the lots too deep. The width of such a one-sided street may be curtailed without reducing its thoroughfare capacity because it is freed from local business all along one side. Bluff Street, though not a thoroughfare, is an excellent Pittsburgh example of the one-sided street, and illustrates the great attractiveness which such streets often possess. In a two-level street a longitudinal bank, or retaining wall, is introduced in the middle so as to adapt it to the topography and bring each half of it nearer to the natural surface where the abutting property fronts upon it. Such a street must normally be wider than a single thoroughfare of the same capacity, the saving in construction and in the development of abutting land more than counterbalancing the cost of extra width.

Widths for outlying thoroughfares in a district like Pittsburgh, therefore, cannot be determined by any general rule. Each must be laid out as a problem by itself, the principal objects in each problem being to select a tolerably direct line on reasonable gradient, and so to fix the side lines of the location that it shall be possible to meet the immediate needs by constructing an economical suburban road, where it does not already exist, and ultimately to convert it into an ample urban thoroughfare with the minimum of cost and inconvenience.

Whatever radical changes may be made to improve the present or safeguard the future condition of the thoroughfare system in regions that are now rural, there remains a huge problem within the district where the street system has already crystallized into substantially its final form. Here increased capacity can, for the most part, be secured only by local improvements and widenings of existing thoroughfares.

Fortunately, the building up of the street frontage with solid blocks of stores, apartments, and business structures, has at most points followed rather slowly after the earlier wave of detached dwelling houses, and a large proportion of the streets which are destined to be the main arteries of the huge future city are still lined by buildings which are set back at various distances from the street, leaving front dooryards between them and the sidewalk. Outside of the down town district, and a limited area in East Liberty, it is possible, therefore, to provide for the ultimate widening of these streets without the destruction of many valuable structures, provided the preliminary steps are promptly taken.

As traffic increases and the lots come to be used for business purposes, such a set-back becomes inconvenient and undesirable, and one by one the buildings are either extended to the sidewalk by new additions, or new buildings are erected on the sidewalk line. The reason for this change is not usually that additional lot depth is required, for often considerable yards are left unoccupied at the rear, but is simply that on a commercial street the buildings need to be as close to the stream of traffic as possible; and since the individual lot owner cannot move the street as a whole up to his building, he has to extend or move his building to the street. His immediate purpose is thus served, and ultimately the whole row of buildings is similarly advanced in response to changed conditions. But at just about the time when this process is fully completed, the volume of traffic flowing over the street is apt to have become so great that everybody recognizes the street to be too narrow for the increased traffic it has now to carry. If the case is a bad one, the inconvenience due to overcrowding the traveled way will in time reach a point where, in spite of the great cost of such an operation, the buildings all along one or both sides of the street have to be destroyed and a new building line established—it may be on the very line where most of the original buildings stood before increasing traffic began to offer inducements to move them forward to the sidewalk. Indeed, it may be said as a general rule that on any street where the buildings are set back from the sidewalk line the very advancement of a few buildings to the sidewalk line is a sign which points directly to the growth of travel and indicates that ample width will soon be needed in that thoroughfare.

As soon as these conditions appear, it is time to act. As already noted, it is not, in most cases, the desire to utilize a greater depth of lot which leads to the change, but the desire to get next to the sidewalk and to do away with a front yard which has served its purpose and is not wanted under the new conditions. If the street is one likely to have a considerable amount of through travel, it would be reasonable at once to lay it out wide enough to handle such travel; and the cost of the land taken for the widening would be charged, at least in part, to the abutters, for they get, by the change, what many of them already want and what the rest will soon be wanting—direct frontage on a busy sidewalk.

A still wiser course of procedure would be to determine on the widening of these future main thoroughfares before any buildings have been advanced to the sidewalk line, and to establish building lines far enough apart to leave room for all probable future requirements; but to make no physical widening of the street until the growth of travel—or the demands of the abutters—call for shifting the sidewalks over to the established building line and enlarging the roadway to correspond. This is the invariable practice in Washington and in most well-conducted European cities. It is the plan to some extent in New York, where just recently the sidewalks of Fifth Avenue have been moved back against the building line on the space formerly occupied by stoops, areaways, and dooryards. Pennsylvania Avenue and Sixteenth Street, in Washington, are both laid out 160 feet wide from building line to building line, although Pennsylvania Avenue is an important business artery and Sixteenth Street is a residence street without heavy traffic and with no commercial business. On the former, the wide sidewalks are in immediate contact with the fronts of the buildings, as is proper for a business street, and the roadway, with car tracks in the middle, is more than wide enough to carry all traffic that can ever be concentrated upon it. Whereas, on Sixteenth Street, the traveled portion of the street, including sidewalks and the space for sidewalk trees, is only 80 feet wide; and the remainder is occupied by front dooryards 40 feet deep, which the householders are at liberty to fence and use almost as freely as if they owned them in fee simple. At the same time all the householders are protected against the premature action of any individual lot owner who might see a possible advantage in being among the first to bid for a commercial business by building a flat-house with stores under it out upon the sidewalk line 40 feet in advance of the other houses. This is the sort of thing that is happening every now and then in Pittsburgh on streets where the great majority of the owners would prefer to have the set-back continued for some years longer. In Washington this crowding forward cannot be done; but when a reasonably large proportion of the owners on any street, or any block of a street, are ready for the change, the front yards are abolished and the sidewalk is moved over into contact with the buildings. If a single owner wants to put in a store long before his neighbors are ready to give up their front yards and long before the City is ready to widen the street to increase its traffic capacity, he is of course at liberty to do so; but he must not move forward of the general building line. What he usually does is to abolish his own front dooryard and substitute an extra wide piece of sidewalk paving in place of it, sometimes using the space for outdoor stands, or show cases, to attract trade. He may even be permitted to erect light temporary structures, such as awnings, on the space between his main building and the present sidewalk line, under which, in good weather, he can do a very good business.

There is, then, one course of action which overshadows, in permanent importance and in urgency, all other things that Pittsburgh could do at the present time for the improvement of its main thoroughfare system. That is to establish new building lines, at a suitable distance apart, along all of its present and prospective main thoroughfares which there is any prospect of being able to widen.

Pittsburgh, in common with other cities in Pennsylvania, has a remarkable power, which is of the utmost importance in connection with the intelligent control of its street development, but of which it has not hitherto taken adequate advantage; a power that appears to be denied to the cities of every other state in the Union, although effectively used in some other countries. Pittsburgh may legally lay out a street in anticipation of a future need, and yet postpone entering upon the land for construction or for opening it to the public. Until the city legally enters on the street, the owner of the land has the free use thereof, and he receives payment only when the opening takes place; but if, in the interim, he shall have erected any structure within the limits of the proposed street, he will receive no compensation therefor when the street is opened. Although similar laws have been declared unconstitutional in other states, this provision has been sustained in Pennsylvania, and the power has been effectively exercised in numberless cases since the middle of the last century.

Philadelphia has applied the same principle to street widenings, as for example in the case of Chestnut Street. The procedure is to define a building line, set back a certain distance from the street line, and to permit no new buildings to be erected in front of that line, but to pay damages only when the power to prevent the erection of a new building is actually exercised.

The Chestnut Street widening was authorized by legislation which provided merely that the street should be widened ten feet, without specifying the procedure or method of awarding damages.[7]

The procedure used in the widening, as above described, had apparently no other authority than the general acts under which Pittsburgh has proceeded in laying out new streets.[8] This application of those acts has been sustained by the courts. If it is held that a specific extension of the principle of the Act of 1871 to the widening of Chestnut Street was implied in the ordinance of 1874, under authority of the Act of April, 1870, and that it is not generally applicable to widenings, a general act so intending ought to be secured from the legislature.

In the Chestnut Street case existing buildings covered most of the space between the building line and the street line, and the exercise of the power, with the consequent accruement of damages, occurred in each case only when the original building was torn down by the owner and he was required to set the new building back to the new building line.

The same principle is equally applicable to those cases where the existing buildings are mostly or wholly back of the new building line; the damages becoming due in such a case only when a building permit for the erection of a new structure encroaching on the designated open space is actually withheld.

The advantages of such a method of procedure, in the case of those highways where all, or nearly all, of the buildings are now set back from the street and where a widening will ultimately be needed, are obvious and very great. In a large percentage of cases, where the street is still mainly residential, the majority of the abutters would welcome the establishment of a building line for their own protection from inconsiderate neighbors; just as the majority of people will pay higher prices for lots in a neighborhood that is protected by properly drawn restrictions for setbacks, etc., imposed by a land company. In a great many such cases abutters could be induced to waive any claims for damages on condition that the building line should be applied to the whole street. Furthermore, the actual net damages to be paid would be distributed over a long period, and a considerable proportion of them, in many cases, could properly be assessed on adjacent benefited property owners.

When the actual physical widening of the street takes place, through absorbing the restricted zones on each side of it, the damages for land taking will be comparatively small, because at that time most of the abutters will want nothing so much as that very widening, if only to bring the sidewalks in contact with the fronts of their buildings. But regardless of its clear financial advantages to the City, in reducing its total payments for street widening and especially in distributing the burden of that cost over a long period without running up a large bonded indebtedness and interest charges, the fundamental argument for this method of procedure is that it avoids the absolute dead loss to the whole community resulting from the destruction of valuable buildings. It is not practicable to avoid this in any other way and still accomplish the result of widened thoroughfares. Theoretically, it could be done by a direct widening of all the highways in the ordinary manner, if it were to be done promptly; but there are comparatively few cases in which there would be enough immediate advantage in the increased width to make the proposition attractive; and it is obvious that any such wholesale immediate action would involve a sudden and enormous financial burden which it is utterly impracticable for the City to assume.

If, after the gradual piecemeal process of widening at moderate and distributed expense has been begun, the City thinks it would prefer to have the process over and done with promptly, it is just as able to complete the widening immediately, by wholesale condemnation, as if the gradual process had never been entered upon. If the City begins on the gradual process, it can always change to the other when it feels rich enough, or when the buildings on the old lines have become few enough; and in the meantime the erection of new and costly buildings, obstructive to the proposed widening, has been prevented at comparatively slight expense. If the City does nothing, pending such time as it can afford to make the widening at a single operation, the cost of the operation is liable to mount at least as fast as the City's ability to pay for it.

While the method proposed is peculiarly adapted to handling the problem of a thoroughfare along which the majority of the frontage is not yet occupied by buildings standing on the street line, it may be objected that it is not suitable for widening one that is built up, like Forbes Street. It is true that the patchwork appearance of such a street during the process of gradual reconstruction is somewhat unsightly,—with here and there a wide place where new buildings have gone up, and between them narrow parts, thus exposing the blank side walls of old buildings projecting beyond the new ones. Yet in cities where the sense of civic beauty is far more acute than it generally is in America, this temporarily ragged condition is accepted as a small price to pay for the economical and certain accomplishment of a great permanent improvement.

It is obvious that the flow of traffic moves regardless of the artificial boundaries of the city and the surrounding boroughs, and that if an efficient system of thoroughfares is to be involved for the Pittsburgh Industrial District it will be necessary to disregard those boundaries in planning it. This has been done in the preliminary studies which have resulted in this report, and the necessity for it must control the form of any permanent organization for preparing final plans and executing them. If these duties are to be entrusted to officers of the City, and the city boundaries remain unchanged, those officers must have authority from the legislature to deal with territory beyond the boundaries of the city, as is the case in a limited way in Wisconsin cities.[9]

The simplest and most logical procedure, if the boundaries of the city and of the boroughs are to remain substantially unchanged, would be to establish a common agency for dealing with the general problems of city planning for all of the municipalities and the related parts of the country outside of them. The Constitution of Pennsylvania apparently prevents the formation of a special metropolitan board for the Pittsburgh Industrial District, but general authority might be obtained under which the County could establish such a board. If the difficulty should be met simply by extending the boundaries of the city, it is important that the new boundaries should include not merely those areas which are now seen to have close physical relations with the city, but a great extent of territory within which the beginnings of urban or suburban growth have started, or are likely to start, during the next generation.

Whether the duty of planning and providing for the main transportation lines is made a city affair or a county affair, those who are charged with it should be free to go as far in any given direction as the demands of the traffic lead them. They should neither be limited by arbitrary boundaries in those directions where scattering but connected urban development may reach out furthest from the center, nor compelled to extend their operations to an arbitrary boundary in those directions where such development falls short.

SPECIFIC RECOMMENDATIONS
MAIN ARTERIES

As noted earlier in this report, one of the two main eastward thoroughfare routes, from the Point District, must lie along the flat land between the Allegheny River and the bluff southeast of the Pennsylvania tracks. Through this bottle-neck must pass the trunk line (or lines) of one of the largest thoroughfare systems leading from the down town district of Pittsburgh. At the foot of the Lawrenceville hill the system branches into two main lines of extension. On the one hand are Penn and Liberty Avenues, extending, by different routes, through the Garfield, Bloomfield, Friendship and Shadyside Districts to East Liberty; and from there connecting directly to Squirrel Hill, Highland Park, Homewood, Brushton, Wilkinsburg and all points further east. On the other hand is Butler Street, following the low land along the river through Lawrenceville to Morningside and Highland Park. Via the Forty-third Street bridge, this line reaches Millvale and the country north thereof; via the Sharpsburg and Aspinwall bridges it reaches Etna, Sharpsburg, Aspinwall, and Shaler and O'Hara townships, and connects directly with the Freeport Road, the only thoroughfare leading up the Allegheny River. The trunk line of this system is composed of two narrow streets, Penn Avenue and Liberty Avenue, the one 60 and the other 50 feet in width. Even now this accommodation is inadequate, and, considering the extent of territory served and the increase of through traffic to be expected as the city grows and the outlying lands develop, a much greater capacity for general traffic through this throat will very soon be needed.

There are four different ways in which this greater capacity might be realized.

In the first place, a new street might be cut through north of Penn Avenue. Smallman Street, from Twenty-first to Thirty-sixth Street, already forms a good sized piece of such a thoroughfare. Pike Street would be its normal extension in town to Eleventh Street, but, like Try Street near Second Avenue, it has been surrendered to the Pennsylvania Railroad for a connecting line and spur tracks. Furthermore, it is very narrow (not over 40 feet) and is difficult to widen on account of the many industrial plants abutting thereon. The connections from such a thoroughfare with Penn Avenue, Liberty Avenue and Butler Street at one end, and with the down town thoroughfares at the other, are quite indirect; and they could be improved only at great expense.

The only other place for a new thoroughfare is along Spring Alley, between Penn and Liberty Avenues. As this whole block is only 220 feet wide, including the alley, it is obvious that a broad avenue through the middle of it would leave the abutting property in very uneconomical shape.

As a modification of this plan, the widening of Spring Alley entirely on the south side was considered. As this would leave lots 40 feet or less in depth between the new street and Liberty Avenue, it would mean the practical destruction of the half-block from Spring Alley to Liberty Avenue. The remaining strip could be taken as a central parking space in a wide boulevard thoroughfare, extending from Spring Alley to the railroad; or Liberty Avenue could be abandoned, and the space, left between the new street and the railroad, could be used for warehouses or for business wanting direct railroad connections; or it might be sold in whole or in part to the Pennsylvania Railroad, for additional track space. It is obvious that each of these plans cuts up the property undesirably: the first is not only costly but is extravagantly wasteful of land in a region where available land is strictly limited and should therefore be put to its most efficient use: and the other plans both involve an entire redistribution of the land south of the new street. They could hardly be executed without powers of "excess condemnation" for which constitutional authority is lacking.

A third plan would be to widen Liberty Avenue on the north side. There is no special difficulty in the way of this scheme, and it could certainly be more easily carried out, and at less cost, than any of the Spring Alley plans. Merely as a traffic way between two points, Liberty Avenue widened would be perfectly satisfactory, but several incidental considerations must be borne in mind. First, the lots on the north side of the street would be cut at least to 70 and probably to 50 feet, neither of which is a desirable depth for lots on a main thoroughfare; and second, the street would have business frontage on one side only. The latter is an uneconomic arrangement from the point of view both of the real estate owner and of the City, and the street would be much less agreeable than if it were separated from the railroad.

The fourth plan would be to widen Penn Avenue. This street is now 60 feet in width, and most of the lots on each side are 100 feet deep, except for several blocks on the north side where they are about 120 feet. The street is built up solidly on both sides, but scarcely any of the buildings are new or costly. The property values are almost uniformly a little higher than on Liberty Avenue. If Penn Avenue were widened 10 feet on each side, making an 80-foot thoroughfare, the abutting lots would still be 90 feet or over in depth; and if the street were made 100 feet wide, the lots would still be 80 feet deep. Though it might cost somewhat more to widen Penn Avenue than Liberty Avenue, it is evident that the abutting property would be left in far better shape, and the benefit to be had from increased frontage value would be much greater.

After due consideration of each of the above plans, bearing in mind the cost, the difficulty of carrying it out, and the value of the result, both as an important main thoroughfare artery and as a local improvement, it is recommended that Penn Avenue be widened to 100 feet. If the widening is to be accomplished by the gradual process,[10] that is by merely establishing the new building lines at the present time, and by paying damages only when new buildings are set back to this line, the widening should probably be made on both sides: for in this way the minimum set-back will be required for individual new developments and the lots will be left of a good depth on both sides of the street. But if the widening is all to be made at once, it will be less costly to make it entirely on the south side. In either case, the lots remaining will be none too deep, and it is suggested that ultimately Spring Alley may be abandoned and the opportunity furnished for deep lots for warehouses and similar purposes, fronting on a large thoroughfare and having direct railroad connections over Liberty Avenue in the rear.

The other eastward thoroughfare system lies south of the Hill District. From Soho eastward there are two main branches to the system: on the one hand are Forbes Street and Fifth Avenue, leading through Oakland to Bellefield, Shadyside, East Liberty, Squirrel Hill, and all points east; on the other hand is a possible and much-needed thoroughfare reaching Greenfield, Hazelwood, Glenwood, and Hays, and from there, by branches and extensions, connecting to Homestead, Duquesne, McKeesport, and points up the Monongahela and Youghiogheny Rivers, as well as to the country south in Baldwin, Mifflin, Snowden, and Jefferson townships.

At present the trunk-line of this system (from the Point District past Soho hill) is composed of three narrow streets, Second Avenue, Forbes Street, and Fifth Avenue, which all together are no more than adequate to accommodate the present surface travel. Future developments in the East End, up the Monongahela, and in the country south of Homestead, and improved thoroughfare connections with the two latter regions, will undoubtedly increase the through traffic on these streets to such an extent that their capacity will soon be taxed beyond its limit. There can be no doubt that more accommodations will be needed in the near future.

At first thought it was hoped that Second Avenue might be improved to accommodate a reasonable increase in east and west traffic; but the Baltimore & Ohio Railroad on one side, and several large industrial plants on the other, present serious obstacles to widening it. A plan to exchange locations with the railroad was considered, but it did not appear to offer sufficient advantage to the railroad to tempt them to cooperate in the matter.

Incidentally, Second Avenue can and should be widened to 80 feet, from Ross Street to the Tenth Street bridge, thus making a good connection between the Point District and the South Side.

It remained then to secure the desired street capacity, in some way, through the valley now occupied by Forbes Street and Fifth Avenue. To avoid the higher land values on these streets, various schemes were tried to get a third thoroughfare in this valley, first on the south and then on the north side, but without success. The indirectness of line and the seriousness of grade difficulties, coupled with cost of cutting new connections at either end, more than outweighed the advantages offered by the cheaper land.

One proposition, however, is worthy of special remark. That was to cut a new street from Fifth Avenue, near Sixth Avenue, to the end of Colwell Street, widen the latter, carry it over the Moultrie Street valley on a high viaduct, skirt around Soho hill, partly above and partly below Beelen Street, and either join Fifth Avenue at Robinson Street, or, going over this street, follow along the hillside and meet the southerly end of Bayard Street. The cost of constructing this line, the complication of grades with cross-streets (owing to the width of the new street), and the difficulty of getting good connections with any thoroughfares leading up the Monongahela, practically put it out of the question as a solution of the main problem in hand. But it offers many advantages as a specialized thoroughfare for fast-moving automobiles for the East End. It is well up on the hill, furnishing, at times, fine outlooks over the river; the gradient need nowhere be over 4 per cent, and the line could be easily laid out so as to have very few grade crossings with other important streets. It is urged that this route be borne in mind when the demand is felt for another "Grant Boulevard," south of the Hill.

It remained, then, to consider adequate widenings of Fifth Avenue or Forbes Street. The former is now 60 feet wide throughout; it is by far the more important thoroughfare at present, land values are much higher than on Forbes Street, and new and somewhat costly buildings are already crowding out the cheap houses of an older generation. Forbes Street is also 60 feet in width, except near its westerly end where it is only 50 feet, but the buildings, on the whole, are much less valuable than those on Fifth Avenue. Lot depths are practically the same, and so are the street gradients. It is evident, therefore, that the widening of Forbes Street should be a far less costly undertaking than the widening of Fifth Avenue.

A thorough study of the possible eastward extensions of Forbes Street has developed no obstacles to using it as the principal thoroughfare. By referring to "Outlying Thoroughfare Improvements" below (Sections 11, 14, 15, 16, 17 and 19), it will be seen that a cross connection can easily be secured at Brady Street to Fifth Avenue—the latter being the easier street to widen beyond this point, as well as offering somewhat better gradients. It will also be seen that the thoroughfare extension to Greenfield, Hazelwood, Glenwood, and eastward, can branch from Forbes Street (just east of Brady) more easily and cheaply than it could from Fifth Avenue. Forbes Street, moreover, enters the down town district at a slightly more central, and, considering the proposed improvements in the down town district, a more advantageous point.

It is recommended, therefore, that Forbes Street be made the main artery of this eastbound thoroughfare system, and that it be widened to 100 feet. As in the case of Penn Avenue, the widening should be made on both sides if done by the gradual process; but if done all at one time, it should be made entirely on the south side.

Between the Point District and the South Hills there is now urgent need for a thoroughfare connection of adequate capacity and on reasonable gradients. At present the only access for surface traffic—except electric cars—is via the Brownsville Road, or South Eighteenth Street, or the inclines. The two roads are steep, from 7 to 8 per cent, and the inclines are expensive and of very limited capacity. The South Hills country is sparsely developed as yet, but, being comparatively free from smoke and very near to the business district, it offers unusually desirable opportunities for homes, and it must soon be thickly settled. The need for a good thoroughfare to this region will then be of far greater importance even than now.

Only two reasonable ways of securing such a thoroughfare appear. One is by a new slanting road up the hillside south of the river, much longer, and so on an easier gradient, than Brownsville Road; the other is by some high-level bridge and tunnel scheme, such as that proposed by residents of the South Hills.

The opportunities for a hillside road have been studied with some care, but the excessive length required to get a reasonable gradient, and the difficulties and high cost of constructing a wide thoroughfare on the steep hillside, have proved to be serious drawbacks to all possible plans for such a street.

In any thoroughfare scheme to the South Hills, it is reasonably clear that the end to be attained is the most direct access possible on easy gradients to the higher levels of the South Hills country. For it is on the upper levels, the hilltops and the upper slopes, that most of the present development has taken place; and there can be little doubt that in the future, even when building space is at a much higher premium than it is now, the overwhelming majority of the population will be found on the hills rather than in the narrow valleys.

There are certain general tendencies which are observable, both in America and in Europe, in cities which have a large area of hilltop land separated by deep valleys. The hills are generally preferred for residential purposes, and the earliest roads or trails often follow the ridges, plunging down and climbing up again steeply to get from one ridge to another. The main roads in the second stage of development are apt to seek the valleys for the sake of good gradients, with a corresponding development of the most active urban growth in the valleys and on the lower slopes; the hilltop development being retarded by lack of transportation facilities. Nevertheless the continued attractiveness of the uplands slowly builds them up, and as the wealth of the community grows there is an inevitable tendency to reduce the obstacles to ready connection between one hill district and another by raising the levels of the bridges which cross the intervening valleys. Bolder and bolder viaducts are built, until finally there is a complete and more or less independent highway system on the upper levels, and the major part of the residential district comes to be there too.

Obviously, therefore, every opportunity should be utilized to gain grade, in the approach to the South Hills District, by starting at a high elevation and wasting no distance in level stretches, if the most efficient thoroughfare artery to this district is to be secured.