{330} a folio book on Orders of Knighthood, and, giving to many of them an antiquity of several centuries,—­often either fabulous or greatly exaggerated,—­provided them all with imaginary collars, of which he exhibits engravings.  M. Favyn’s book was republished in English, and his collars have been handed down from that time to this, in all our heraldic picture-books.  This is one important warning which it is necessary to give any one who undertakes to investigate this question.  From my own experience of the difficulty with which the mind is gradually disengaged from preconceived and prevailing notions on such points, which it has originally adopted as admitting of no question, I know it is necessary to provide that others should not view my arguments through a different medium to myself.  And I cannot state too distinctly, even if I incur more than one repetition, that the Collar of Esses was not a badge of knighthood nor a badge of personal merit; but it was a collar of livery; and the idea typified by livery was feudal dependence, or what we now call party.  The earliest livery collar I have traced is the French order of cosses de geneste, or broomcods:  and the term “order”, I beg to explain, is in its primary sense exactly equivalent to “livery:”  it was used in France in that sense before it came to be applied to orders of knighthood.  Whether there was any other collar of livery in France, or in other countries of Europe, I have not hitherto ascertained; but I think it highly probable that there was.  In England we have some slight glimpses of various collars, on which it would be too long here to enter; and it is enough to say, that there were only two of the king’s livery, the Collar of Esses and the Collar of Roses and Suns.  The former was the collar of our Lancastrian kings, the latter of those of the house of York.  The Collar of Roses and Suns had appendages of the heraldic design which was then called “the king’s beast,” which with Edward IV. was the white lion of March, and with Richard III. the white boar.  When Henry VII. resumed the Lancastrian Collar of Esses, he added to it the portcullis of Beaufort.  In the former Lancastrian regions it had no pendant, except a plain or jewelled ring, usually of the trefoil form.  All the pendant badges which I have enumerated belong to secular heraldry, as do the roses and suns which form the Yorkist collar.  The letter S is an emblem of a somewhat different kind; and, as it proves, more difficult to bring to a satisfactory solution than the symbols of heraldic blazon.  As an initial it will bear many interpretations—­it may be said, an indefinite number, for every new Oedipus has some fresh conjecture to propose.  And this brings me to render the account required by Dr. Rock of the reasons which led me to conclude that the letter S originated with the office of Seneschallus or Steward.  I must still refer to the Gentleman’s Magazine for 1842, or to the republication of my essays which