C. Staugaard, Skakbladet 11/1907, Mate In 2 Moves
But I am not here to tallk about Pam-Krabbe castling if you read the title more closely. Instead, I am here to talk about ANOTHER type of castling with a promoted rook. Sounds impossible? Well it's not!
The earliest I have managaed to trace this heavily obscure loophole in chess history as far back as 1918, which is its probable origin.
The problem was given on August 1, 1918 in The Brooklyn Daily Eagle on page in text: https://bklyn.newspapers.com/image/55625880/
It reads: "William M. Russell, best known in local chess circles and as player and organizer, has turned his hand to composition of problems and contributes the following three-mover for the benefit of The Eagle solvers.
White (5 pieces)-K on K, B on QKt3 and QB3: P on QKt2 and KR6.
Black (1 pieces)- K on QR8
"I'm not going to give the solution," writes Mr. Russell, "but I assure you there is one, and that it does not start with a capture or a check. The idea was given to me by a position my brother showed me of one of Blackburne's games. The solution obeys every one of the laws of the British Chess Code. Sam Loyd would have liked to have been the composer of this problem."
Checkmate In 3 Moves
The problem was brough up again the next week, on August 8, to give a hint to the solution on page 16: https://bklyn.newspapers.com/image/55626989/. Here is the relevant text in full.
"In giving the special problem in three moves by William M. Russell last week, the request to of the author to print the motto "Castles" was overlooked. A number of telephones calls and letters had to be answered in this connection, but Orin Frink Jr. accepted the proposition in the proper spirit. Mr. Frink writes:
"I always thought that Christmas was the time for bringing up problems of which we must be assured there really is a solution, and obey every one of the laws in the British Chess Code." At any rate, any problem which obeys ALL laws at one and the same time is surely worth of the authorship of our late friend Elizabeth. In regard to the problem itself, it seems that the old "Castles, QR" gag doesn't work. The pawn on KR6 must be there to kill time, or to bite on, as it cannot queen in time to do any damage. I object to Mr. Russell's statement to the effect that the solution does not start with a capture or a check. Such hints are SO apt to give away the key move, especially in a problem of this sort.
Perhaps it might have been well to place a particular emphasis on Mr. Russell's references to the British Chess Code and in view of the motto, now supplied, it may not to be amiss to the quote from the British Chess Code the first condition of castling, which is: "(1) Neither the king or the rook have moved in the game." The solution and a last work from Mr. Russell will be printed in a week hence."
And on the week hence, on August 15, the solution finally appeared under the banner "Russell's Castling Problem" on page 16: https://bklyn.newspapers.com/image/55627909/. Here are relevant snippets..
"Coming back to the 'Castling" problem of last week, which, as explained last week, hinges upon the literal reading of the British Chess Code, already quoted, little remains to be added to the exhaustive treatment of the subject by the authot herewith submittted:
The solution is supported by both ancient and modern authority. On page 8 of Alain C. White's book on "The Theory Of Pawn Promotion," and on page 14 of hia "Running The Gauntlet," you will observe that promoted Fers (the ancient queen, that could leap two squares on its first move, but on its subsequent moves could only move one square diagonal) had on being promoted had on being promoted the same privlege of leaping two squares on its first move, as in the beginning of the game. On the same theory, an unmoved king and a promoted rook can castle.
(Skipped text.) In those days, castling wasn't as widely accepted for problems as it is now days, and this part of the article tries to justify having a castling solution.
My problem has only a two move theme, Pawn Rooks and Castles, mate. By placing the Rook's Pawn on the 7th and by exchanging the White pawn on QKt2 for a Black pawn, you would have a two-mover."
Mate In 2 Moves
The August 15 Article was reprinted on pages 209-210 in the book "American Chess Bulletin Volumes 13-15": https://books.google.com/books?id=CqJJAAAAYAAJ&newbks
Thus was born the idea of promoting with a castling rook, but doing so on the first rank. The king moves two spaces on the first rank to the left or right, and the rook jumps of the top of the world to soar next to the king. This idea would live on, but seemingly unknown and largely unexplored, being something that you can remember but can never find the original source for.
