R. M. C. Eagar writes: Harper & Ryan (1990) point out the increasing use now made of sophisticated statistical methods, including multivariate techniques and Principal Component Analysis, for the special needs of palaeon- tologists. They also illustrate the use of the computer-based PALSTAT package by two case histories. The first one (Harper & Ryan 1990, pp. 938-941), the use of the data for the description and separation of the Carboniferous non-marine bivalve species Anthraconaia salteri by Leitch (1940) is a most unfortunate choice. They did not refer to any subsequent work that has contributed to or affected the interpretation of this species, and as a result their new conclusions require a step-by-step revision and several refutations.
The generic name Anthracomya, which they use, is incorrect. As recognized by Trueman & Weir in 1946, Anthracomyu, which was ‘founded by Salter in 1861 with A. adamsii Salter as type, was already preoccupied, the name having been used for a genus of Diptera in 1856’. Anthruconaia, the name substituted to meet the rules of nomenclature, has been in use for the last 45 years.
The horizon of the A. salteri assemblage, which Harper & Ryan reinvestigate, is wrongly given by them. Leitch (1940, p. 30) concluded that this assemblage, overlying the Musselband Coal of the Scottish Central Coalfield, should be placed at the base of the Similis-Pulchra Zone (now Chronozone) and his opinion was repeated without comment by Harper & Ryan. In fact Lumsden & Calver (1958) showed that the Musselband Coal must