On 'The bones of the insane' (letter received 23 December 2013) I would like to make some critical comments on Jennifer Wallis's paper 'The bones of the insane' in History of Psychiatry 24(2): 196-211 (2013). They are not directed at her argument but at correcting her representation of the work of Alfred Walter Campbell. Notable in Campbell's work are his use of an objective standard provided by Mercier's instrument, his use of a proper comparison group, and his microscopic study of the anatomical structure of affected ribs. In my view, those qualities are not adequately conveyed by Wallis. Campbell published two papers, the first on the ribs of 58 patients (35 male, 23 female) from Rainhill Asylum (Campbell, 1895, presented November 1894), and the second on 58 'sane' patients who had died in the Royal Southern Hospital (RSH), Liverpool (Campbell, 1897, presented July 1895). From the first he concluded: 'the influence of chronic wasting diseases in the production of bone weakness is probably equivalent to if not greater than that of mental deterioration'. Although he found lowered strength in General Paralysis of the Insane, he made it clear there were not enough data to establish a relation between rib frangibility and it or other types of mental disease. Campbell (1895) provided only a 'provisional standard' based on six 'sane' patients. His later RSH measures provided a more adequate one, and he found their ribs were of almost the same strength as those from Rainhill. Here are my comments. First, nothing in either paper suggests Campbell was 'confident' about his figures for the 'insane' in the first but 'hesitant' in the second. Rather than expressing confidence, he said he was not prepared to offer 'any definite answer' on the relation of rib strength to the form of 'insanity' (Campbell, 1895: 256). It would take 'some thousands of cases' to establish a positive conclusion. In the second paper he did not mention the form of disorder but quite strongly rejected as 'far from being correct' the 'established doctrine' that the ribs of the 'insane' were inherently weaker (Campbell, 1897: 205). Second, Campbell did not use another Rainhill sample in the second study: the Rainhill averages in the first are the same, and to the second decimal point, as those tabulated in the second (compare 1895, Point 3, p. 256, and 1897, p. 205). Third, there were 58 subjects in each of the Rainhill and RSH groups, and not 50 in the latter. In a private record written just before his death, Campbell (1937: 3) described his work as 'A study in conjunction with the late Dr. Charles Mercier to prove that the supposed fragility of bones in the insane was a myth'. Having previously discussed the soundness and possible uniqueness of the methods used by Campbell in establishing that conclusion (Macmillan, 2011-2012, 2012), I can hardly agree that his results were 'inconclusive,' as Wallis states.