Contains three letters, largely unrelated. In the first letter, dated 15 Jan 1938 and numbered 8a (38), Thomson informs Ledermann that his [Thomson's] manuscript regarding univariate selection has been sent back to him by F C Bartlett, as the publisher requires it typed, and advises him that his own manuscript has likely been sent back for the same reason.

In the second letter, dated 17 Jan 1938 and numbered 8b (38), Thomson asks Ledermann to check the Psychometric offprints for any misprints. He refers to work by Dodd he had originally enclosed on loan from J Neyman. This has not survived in the collection.

In the third letter, dated 12 Feb 1938 and numbered 8 (38), Thomson continues from 7 (38) (Coll-1064/40). He reproduces matrices of raw marks, and refers to partial correlations between the tests for constant average marks for each child. He provides several pages of examples and working.