The adverse criticisms of RTPE centred on two aspects – its obscurity and its claim to calculate physical constants. I shall deal with these two aspects in turn.

Obscurity

One reason that RTPE was found obscure by its readers, I have argued, was that they expected something different in style. But there was a more important reason; they also expected something different in content. For, though they had read MTR with approval, they had not appreciated the novel philosophical positions Eddington had taken up already by 1923. These positions arose out of general relativity but the main lesson of Eddington's book was, as its name implies, the mathematics of the theory. This could be read and the philosophical standpoints ignored.

I distinguished eight novel ideas that Eddington exhibited in MTR. All are relevant to RTPE and I discuss their relevance here in three groups of increasing importance:

1. (i) The notions of selective subjectivism and of falsifiability are neither of much importance, though for different reasons. It is hardly surprising that selective subjectivism fits well, for it is consciously followed by Eddington and expresses his Kantian preoccupation. Falsifiability is not important because it has taken a back seat. Eddington allows himself such freedom to change to a new model half-way through an argument that the ostensible falsifiability of an exact numerical prediction is a hoax.

2. (ii) Intermediate in importance are non-redundancy, descriptive tolerance and structure. Non-redundancy is taken for granted and used without mention from time to time. For example, the fact that just four Ea give rise to a fifth E5 which anti-commutes with all the four gives rise to various attempts at interpretation. It is taken for granted that there must be an interpretation and that it will be physically important.

3. […]