(A) There is substantial variation among the typeforms found on the slab-serif mosaic signage
(B) A possible precursor of the NYC mosaic type: Cushing Medium, designed in 1897. Note angled serifs on the crossbars, e.g. C, E, F, G. L, and T, and the curved lower crossbars of E and L. It is more condensed than the NYC mosaic font, but it shares many other characteristics.
Dorsey, another possible source for the Dual Contract type fonts, found in the 1911 ATF catalog.
(C) This section from the IRT Lexington Ave. – 86th St. Station is a good example of the mosaic techniques used in the NYC subway signage. Mosaic signage was either prepared in sections and mounted on the walls, or applied using an indirect method of placing the tiles to a sheet of paper in reverse, mounting the tiles on the walls with a quick-drying glue, and applying grout in situ. For the most part the tiles are placed in an interlocking, offset pattern, a technique used by the Guastavino brothers in the overhead vaults of the earliest NYC subway stations, Astor Place and City Hall stations. Note that very rarely do four tile corners come together at any one point; but rather, a vertical joint abuts a horizontal one, or vice versa. Technically this produces a more stable sheet of tiles, less apt to develop cracks along the joints. Aesthetically, it works to keep the tiles from "gatoring," and unifies the mosaic field. The letter "S" above is a good example of this. Also note the care and placement of the outer layer of outline mosaics on the letter forms.
Decorative Wall Borders
This page focuses on the design and typography of the post-1908 rectangular station-name signs of the NYC subway system. The common elements of the signage are:
1) Wide aspect ratio, somewhere between 3:1 to 6:1, depending on the number of characters
2) Contrasting color inset borders in varying widths and patterns, but in a style used consistently across the system.
3) Multi-color mosaic background, usually in analogous colors (adjacent on the color wheel), sometimes complementary (opposite sides of the wheel), e.g. IRT Lexington's 59th St., or triadic. You name it, they tried it.
4) Mosaic letterforms with an outer layer of outline mosaics.
5) A slab-serif font, an Arts & Crafts font similar to "Egyptian" fonts of the late 19th century, but with characteristics specific to the subway signs.
First of all, any discussion of the typefonts of the NYC subways, even limited as described above, must take into account the wide variation of the hand-set typeforms. A quick glance at (A), left, shows a considerable range of typeforms even among the same letter of the same typeface. To describe the type found on the signs as a single typeface is a bit misleading, given the wide range of forms, even within a single sign. That said, there are some general characteristics which lend themselves to investigation.
So what typeface is it? The contemporaneous font closest to the subway style is the medium weight of Cushing (see B, at left), designed in 1897 by J. Stearns for American Type Founders, although there are substantive differences between it and the Dual Contract serif faces. While it shares the slab-serif form of the subway fonts, and the angled crossbar serifs, Cushing is more condensed, has beefier serifs, and lacks certain letter-specific characteristics, like the curving lower crossbar of the capital "E" and "L." Cushing's cap R and its descender, usually a distinctive signifier of a type design, is substantially different. Best guess is that Squire Vickers and his staff started with something very like Cushing and then created their own design.
Another candidate, although unlikely, is Dorsey, found in the 1911 American Type Founders catalog, along with Cushing, though no longer available. It has the same slanted serifs found in Cushing, without the compression of the condensed style of Cushing, and thinner serifs, similar to those found on the Vickers' font. That said, there are sufficient differences to rule it out. The fact that there are two type fonts available at the time Vickers was designing the Dual Contract signs and similar to what he came up with, tells me that he was using a commonly-accepted type design format of the time.
(Contrary to what Wikipedia says, the serif typefaces used on the "Dual Contract" signage is not Times New Roman, which was not created until 1931, well after the Dual Contract subway stations were built.)
What are "slab serif" fonts? Heavy, blockish fonts with massive, square serifs. They're derived from the work of Vincent Figgins (who coined the phrase "sans serif") in the early 19th century, soon popularized as "Egyptian" fonts, that have nothing to do with Egypt. By the time of the Arts & Crafts movement of the late 19th century, slab serifs were considered "workmen's" fonts because their sturdy clarity was thought to promote readability. Popular slab-serifs include Clarendon, Beton, and Memphis.
So the question remains: who designed these elegant mosaic fonts? They were created during the aegis of Squire Vickers, the architect of the subway system from 1906 to 1943, with likely help from his assistant architect, W. Herbert Dole (see the History section). There is some debate as to whether the distinctive "slab serif" font used in the signs were of Vickers' design or incorporated from the earlier work of Heins & LaFarge, his predecessors at the subway system. Paul Shaw, who teaches lettering at Parsons in New York and author of "Helvetica and the New York Subway System," says that Vickers simply carried forward the type styles created by H&L. On the other hand, the authors of "Subway Style" say that Vickers broke with the H&L type formats and created the new type face. [See the Resources page for info on both books] Only thing I can add to the debate is that the earliest example of the new font is the Essex Street station, albeit in a variant form, which opened in 1908, two years after Vickers started. I can find no examples of it prior to that.
Brackets. The NYC mosaic font serifs are "bracketed," meaning that the joint between the serif and the main stroke of the glyph is fileted with a wedge-shaped or curved shoulder. Look at the bottom serifs of "T" and "R" in (D), e.g. The subway mosaic serifs exhibit both kinds of brackets, curved and wedge.
Random Tile Placement. Look at the background of example (C) at left. There are probably a half-dozen or more different tonal values represented. The trick that makes it work is that the various colors are placed randomly. This is a lot easier said than done. I can vouch for this. Random is hard – I've spent an ungodly number of hours trying to imitate the random tile placement of the NYC subway mosaic tiles. But when it's done well, the effect is magical. Even more so when the color values are contrasting, viz., Christopher St./Sheridan Sq. or Rector Street.