In Magicians' Tricks: How They Are Done, 1910, p. 46, Hatton and Plate describe doing color changes while the deck is bound with a rubber band. A few cards were prepared by drawing two crossed black lines on their faces, which from a distance appeared to be the rubber band that was wrapped around the deck.
In the Aug. 1911 issue of The Sphinx, Vol. 10 No. 9, p. 109, J. Edward Stewart added more verisimilitude to the gimmick by gluing lengths of real rubber band to the face of one card. This was loaded onto the bottom of a deck bound in rubber bands. The face of the deck was displayed and the feke card palmed off to effect the color change.
The next year, Lionel T. Scott marketed “The Rubber Cross Color Change” (The Sphinx, Vol. 11 No. 9, Nov. 1912, p. 177), and after a year of unknown success with sales, published a method for the effect in the 1913 issue of the same journal (The Sphinx, Vol. 12 No. 10, Dec. 1913, p. 197). The name of Scott's trick, “The 'All Rubber' Color Change”, suggests its inspiration was the “part rubber” method given by Hatton and Plate. The method Scott contributed produced only one color change, although his previous ad promised two. A solution for doing two consecutive changes with a banded deck (which may well have been the same one Scott marketed) appeared the next month: R. W. Hoel suggested a modification to the construction of Scott's feke card and handling that made a second color change possible with the bound deck (The Sphinx, Vol. 12 No. 11, Jan. 1914, p. 214).
(This entry is based on research done by Reinhard Müller.)