perhaps misleading, poetical clothing, which, from this presentation of
its appearance, its source, its utility, etc., shall be recognized by
the intelligence, i.e., can and shall be guessed. 'Catch-questions,'
on the contrary, are not to be guessed, the questioner intending
himself to give the solution; at their best they are intended to trick
the hearer, and since their solution is impossible to the uninitiated
are not 'true riddles' but false ones. Since I propose to divide the
total riddle material of each single nation between these two great
chief groups, may I not somewhat extend the scope of the latter,
including some things which are rejected from most collections as
having little to do with actual riddles--those questions which are
generally insoluble and such tests of wisdom as appeal not to wit
and understanding, but to knowledge--which are certainly not true
riddles. Thus, in the group here characterized as 'false' different
classes of things are brought together, the characteristics of which
I shall investigate later." It would be interesting to quote the
author's discussion further. We can, however, only state that he
recognizes three classes of "false riddles," to which he gives the
names "wisdom tests," "life-ransoming riddles," and "catch-questions."

Of "true riddles" there is a vast variety of form and content. Most
typical is the descriptive riddle of a single object to be guessed. In
its complete and normal form Petsch claims that such a riddle
consists of five elements or parts. 1 Introduction; 2 denominative;
3 descriptive; 4 restraint or contrast; 5 conclusion. 1 and 5 are
merely formal, trimmings; 2 and 3 are inherent and essential; 4
is common and adds vigor and interest. Such complete and "normal"
riddles are rare in any language. Usually one or more of the five
elements are lacking. It is only by such an analysis of riddle forms
that a comparative study of riddles can be made. Any single riddle is