gleefully added to my memory of the other instances which have now been related in this paper. The second example of a coincidence in research work is almost uncanny. I was at work upon my ‘Making of London’ (1912) and was behindhand with my proofs. On a Friday morning I received a special request from the publishers to send off the revised proof of a sheet not returned in its proper order. A difficulty had arisen. This part of the proof dealt with the relationship of the Tower of London to the city, and my story was incomplete without the evidence of the Tudor period. I could find nothing, and my morning’s work of research ended fruitlessly. We expected a distinguished visitor immediately after luncheon, and I knew there would be no chance of my finishing the proof after his arrival. I was singularly vexed at my want of success, and suddenly I said to myself ‘I can’t get any information by ordinary means, I will try extraordinary.’ I have no idea what led me to this decision. I was not in a credulous mood, but frankly annoyed with myself. Taking down a volume, selected by mere chance, of the folio edition of the ‘Report of the Historical Manuscripts Commission,’ printed in double columns and containing some hundreds of pages, I stuck my paper-knife into the top of the volume, on the principle of the old practice of Sortes Virgilianae, and turned to discover the results. On the right-hand page of the volume thus arrived at I found the record of a legal case drawn up by Lord Coke dealing with the very subject upon which I was interested. Such a coincidence is surely quite remarkable, even in a lifetime of literary research.