Let me begin with a small sermon:
Contemplating the architectural nuances at Penn Station is a very hazardous undertaking as most of it must be visualized (gone) versus what little can be derived from that which survives, but there are still a wide variety of attributes that can either be reconstructed or actually witnessed, if one studies the terminal's building plans and is willing to take a comprehensive tour, plus has some knowledge about how the facility is utilized, which has actually changed very little, with or without the building itself.
You know, one annoying fact of life today is that our society tends to rush to spot judgments in history and everything else--as for Penn Station, that means many school-age children are being led to believe that Penn Station was closed and torn down many years ago. If that were the case, then exactly where are all those trains going today? Of course, this is a matter of the press and the public being largely unable (or unwilling) to distinguish between the characteristics of form and function. The truth is, after 101 years, the functions of Pennsylvania Station ("Hey uncle, does the title name mean that all of its trains go to Pennsylvania?") survives quite nicely, thank you; as for form, THAT is quite different!
In any case, if not too late or of little interest, the phenomenon of Platform J is one result of the somewhat circumstantial placement of the terminal within Manhattan. Expanding out a bit, the final location of the terminal's basic "plot" was mostly determined by what point the longest, straightest surface elevation could be obtained between the two river crossings. There were a number of contributing factors to this equation, but in sum when you add up the under-river grade requirements (as mild as possible in deference to haulage capabilities), minimal curvatures and the points of access that were available on either side of Manhattan (one existing, the other fabricated through land acquisition), the longest, straightest and most level point of convergence for all "axis" became (as we know) the Tenderloin District: Seventh to Ninth Avenues and West 31st to West 33rd Streets.
For its part, the LIRR's "brass" (in the form of then-President Baldwin) was a bit chagrined by this foreknowledge. Nevertheless, it was to be that or nothing at all as far as the LIRR was concerned, at least at the parent PRR's expense. Were we able to go back in time to the company's Long Island City offices, I'm sure we'd appreciate their position as a collective, corporate "nose holding" when it came to the long-awaited terminal's coming location in the scheme of things. Unfortunately, the natural underlying (i.e. "rolling") topography of Manhattan's East Side through Midtown (now long lost and only slightly evident thanks to decades of real estate development) and the inherently uneven state of terrain below the East River at the projected location of the railway tunnel crossings prevented any sort of workable grade from being achieved east of Fifth Avenue. Even then (1902) the hierarchy regrettably had to concede there would be no way to reach the Grand Central Terminal area, which was where the LIRR originally desired to go with its own tunnel. Plainly, this was a case of the economic realities taking precedence over operational desirabilities. Nor was this factor lost on the PRR and Cassatt; unfortunately the mitigating factors were just not possible to change.
Even at that, localized variances in topography required that the terminal and its immediate approaches be contained in a giant, subterranean concrete box ("bath tub" if you will) of which the size and scope not only had to be carefully considered and designed relative to cost and location, but also to utilization, as once it was complete there would be little opportunity for significant expansion or alteration in the foreseeable future. This was the biggest point of contention between the "railroad" and "real estate" factions connected to the project. In point of fact, PRR President Cassatt personally directed his Engineering department to factor the potential of adding a second set of tubes beneath the "North" (Hudson) River into the final specifications of this mammoth basement. Even to his death this was an issue over which he agonized, but whatever limited chance there was of adding a second set of tracks to the "Meadows Division" and/or under the Hudson were eventually rent asunder by the financial Panic of 1907, which created the situation we ALL still agonize over (perhaps no more so than a certain New Jersey governor!) more than a century later.
In any case, the foundation of Penn Station proceeded through the construction phase with that and a number of uncertainties attached. By informal agreement between corporate parent and its progeny, a proportion of Penn Station's facilities and operational accommodations (i.e. "slots") were to be directed toward the Long Island Rail Road. The exact nature of these requirements weren't as yet known when the terminal's specifications were finalized in 1902 before construction began, which made the task of visualizing it all the more daunting. The upshot is that those tracks still utilized by MTA LIRR in today (generally 13-21) are basically the same as what was foreseen in PRR's "committee meetings" but not yet fact in 1902. What was greatly different at that stage was the means of "suburban" operation. The "MU" as such hadn't yet evolved and was not yet understood--trains powered by some type of "motor" car (think pre-MU elevated operations such as Brooklyn or Chicago) were expected to utilize a loop that would have occupied the area which was actually built as "Yard C" in the northwest corner of the property, at Ninth Ave. and West 33rd Street. It was the 1904-1905 adaptation of emerging rapid transit technologies for electrification of the LIRR's lines in Brooklyn and Queens (an aspect prodigiously supervised by George Gibbs) that finally framed the means in which LIRR would be best suited to utilize its "piece" of Pennsylvania Station.
The PRR itself was in no better a position--their original vision had called for the through haulage of traditional steam-powered consists by yet another "motor" of some indeterminate type, similar to the only known method employed to that time by the B&O in Baltimore. As such, the AA-1's (10001 and 10002) were gradually developed to fill this void in then-available motive power, being of sufficient promise in meeting the PRR's desired (and required) speeds and haulage capabilities as to force a complete revision of the terminal's operating plan in 1905. As we know, the AA-1 prototypes paved the way for the DD-1's, which in turn were what initially made operations at Penn Station a more viable undertaking than originally projected. Heavy terminal switching was also foreseen for the PRR's locals on Tracks 1-4 (and gave rise to what became Yard E), and the PRR actually did deploy some of its "New York Terminal" fleet for this purpose during the terminal's first 20 years, a task which was gradually subsumed (though not completely so) by the PRR's availability of electric MU's after the 1932 insertion of AC catenary. Nowadays, of course, NJTransit (and everyone else) is able to universally overcome this limitation through the latter-day employment of "push-pull" operation.
Given the size of the concrete box needed to enable Penn Station to meet its operational requirements (enumerated above) and the actual availability of real estate in relation to the existing street pattern of Midtown and the former Tenderloin, the terminal building would be smaller in relative size. By necessity this produced an "underhang" along the terminal's north and south sides, beneath the north side of West 31st Street and the south side of West 33rd Street (containing Tracks 1 and 21, respectively). This presented an interesting challenge in placement of the building itself on top of the concrete box, having to be somewhat suspended in the middle of the site but so as not to require that underlying trackage be significantly diverted around its underpinnings.
As such, two oddities resulted:
On the southerly side (West 31st Street), an extensive horizontal system of heavy steel beams was extended out from beneath the building's south edifice to be perched atop the south wall of the concrete box, thus providing for a transfer of the weight burden to an artificial yet immovable base. This deck of beams then became the underground "box" containing Tracks 1 and 2, plus Platform A and likewise supported the restoration of West 31st Street above. This was a simpler method of structural design, which in this particular location was able to be used because of the stub-ended nature of trackage (1-4) in the southeast corner of the terminal foundation.
On the northerly side of Penn Station, however, the same "box" had to accommodate the escape of both "C" and "JO" interlocking complexes into the "subways" beneath West 32nd and West 33rd Streets, thence beneath the East River, AND yield sufficient track space for both platform berthing and the navigation of an interlocking at either end (in this case "C" and "KN"). As a result the horizontal deck of steel beams could not be employed, but rather a substantial system of VERTICAL steel support pilings was driven along the edge of the building's north edifice, thus bearing most of its weight directly into the terminal's bed rock with horizontal supporting beams placed between the vertical supports, only above the complex of track and switch work. In conformance with the trackage profile as it concurrently passed between "C" and "KN," Platform 10 (Tracks 18 and 19) was thus "spread" around this vertical support system of beams and pilings, which to this day is highly evident on that particular platform, even though the building for which it was intended above is long gone. Conversely, this forced Platform I (Tracks 17 and 18) to be a little bit "squeezed" and somewhat shorter, which also threw its alignment out of sorts with the Main Concourse toward Eighth Avenue.
To be precise, while every other platform at Penn Station conforms to a more or less strict standard of about 30 feet across (between tracks), Platform J is a whopping 51 feet across, placing an approximate breadth value of 20 feet(!) on the original edifice of Pennsylvania Station. Each platform at Penn Station is unique in overall measurement, yet all, with one small exception, remain precisely as they were when it opened in September 1910. THAT is quite a testament to the prowess of the Pennsylvania's engineering forces. As acquired in a first-hand, unofficial survey conducted in the spring of 2011 on site, the derived dimensions in this area were thus:
Platform I=780 feet by 26 feet.
Platform J=810 feet by 51 feet.
Platform K=720 by 31.
As point of information, platforms F (tracks 11 and 12) is the longest at Penn Station, at 1,470 feet in length and 31 feet in width. Similarly, platforms E and G were of considerable length as well, platform E sized at 1100 by 31 feet and platform G at 1,100 by 29, which on opening were closely matched by those at Manhattan Transfer: 1,100 long feet by 28 feet across. I believe but for the "continuous platforms" engineered into the two subways beneath the Loop in Chicago these remain the longest high-level railway platforms in North America, if not beyond.
Hope this helps to explain (if not over-explain) this particular uniqueness.
George Chiasson Jr.