Email from Larry Poppe, February 2011:
Sorry it has taken a while for me to get back to you, but I wanted to refresh my memory on the relevant history of the Shepard/Schlee ternary diagrams.
1) The original program (Schlee and Webster, 1967, Sedimentology) did not differentiate within the gravel fraction. When appropriate, its classification output only notified the user "gravel >10% ". An example of the original program's output, which coincidently was used to generate the data for Schlee (1973, USGS Professional Paper 529-L), can be viewed in Hathaway (1971, WHOI Technical Report 71-15). To get his Gravel (gravel >75%), Sandy Gravel (75%>gravel >50%, mud<20%), Gravelly Sand (sand>50%, gravel>10%, mud<20%) and Gravelly Sediment (10% <gravel<75%, mud >20%) fractions, Schlee (1973) had to manually calculate these gravel fractions from the Gravel>10% data.
2) During the early 1970s, Schlee produced a few slightly different coarse-fraction ternary diagrams. For example, see the diagram in Schlee, Folger, and O'hara, 1973, Bottom sediments on the Continental Shelf off the Northeastern United States - Cape Cod to cape Ann: USGS Misc. Geol. Invest. Map I-746.
3) Our first rewrite/recompilation (punch cards to fortran) of the Schlee and Webster (1967) software was Poppe, Eliason, and Fredericks (1985, APSAS - An Automated Particle Size Analysis System: USGS Circ. 963), but no change was made to the 'Gravel>10%' part of the code.
4) The first use of the classification system shown by the graphic in SEDCLASS was by the ConMap series digital and paper maps. For example, see Poppe, Schlee, Butman, and Lane, 1989, The distribution of surficial sediment textures on Georges Bank and in the Gulf of Maine, U.S. Atlantic Continental Margin: USGS Misc. Invest. Series Map I-1986-A. Although different, as a co-author Schlee wanted the ternary diagrams cited as 'based on Shepard (1954) as modified by Schlee (1973)' and, although not entirely accurate, this request made sense.
5) As operating systems and computer hardware continued to change, the software was re-written in 'C' and ran under the GNU operating system, a predecessor of LINUX (see Poppe and Polloni, 2000, USGS OFR 00-358). This software, with input from Schlee, based classification of gravel contents on the diagrams shown in the ConMap series.
6) Again, as operating systems and computer hardware continued to change, the software was re-written in Visual Basic. In an effort to make the software useful to a wider range sedimentologists, the programs designed to run under Windows and be laboratory non-specific. The program SEDCLASS and its eventual upgrade to SEDPLOT were written to offer both the Folk and the 'based on Shepard (1954) as modified by Schlee(1973)' classification systems. Although the diagrams have not changed from those used in the ConMap series, the classification has been altered. Gravelly Sediment is now 10% <gravel<50%; gravel is now simply >50%. Although to some extent unintentionally misleading, Schlee (1973) is now cited mainly to credit the original modification.
I hope that this history clears some things up and gives you a feel for the evolution that occurred over the last 30+ years.