No one undertaking the study of this problem is likely to underrate the difficulty of the task. To find a solution is obviously a matter of great importance; but it is also a matter in which small success may reasonably be expected. The door of the secret has remained unopened for so long. The door is not one which can be forced, and the key is missing. It would be worse than sanguine to hope for its discovery after a light search. Nevertheless there is some hope. “When a door-key is missing,” says Dr. Gardiner, “the householder does not lose time in deploring the intricacy of the lock; he tries every key at his disposal to see whether it will fit the wards, and only sends for the locksmith when he finds that his own keys are useless. So it is with historical inquiry.... Try, if need be, one hypothesis after another.... Apply them to the evidence, and when one fails to unlock the secret, try another. Only when all imaginable keys have failed, have you a right to call the public to witness your avowal of incompetence to solve the riddle.”[139] In the case of the Gunpowder Plot Dr. Gardiner tried the key afforded by the traditional story and found that it fitted the lock. With the secret of Sir Edmund Godfrey’s death the method must be different. There is no traditional story to test. What seems more remarkable is that no determined attempt has been made to construct a consistent theory to fill the empty place. Contemporaries who approached the question answered it according to their prejudice, and selected only such evidence as would support their preconceptions. Later historians who have answered definitely have arrived at their conclusions by considering the balance of general probability in the matter, and have supported them from the contemporaries whose evidence lies on the side to which the balance seems to them to fall. No one has formed a hypothesis to explain the facts and tested it by all the evidence, in whatever direction it seems to point. The following study is an attempt to accomplish this. It would be impertinent to suppose that it offers a perfect key to fit the lock. But it offers a key with which trial may be made, and which may not be found altogether of the wrong size and shape. There is at least a hypothesis to be tested. If it jars with established fact, this will be detected. If the test reveals assumptions which are beyond the scope of legitimate imagination, it must be discarded. At least its abandonment will be because it has been shown to be inconsistent with the facts of the case. It will then leave the way clear for the same test to be applied to another theory.