Manufacturing variations
Two standard Mills castings shown, the one on the left a guaranteed July 1917 No.23 MkII made by REVO, the one on the right a late 1916 No.5 or No.23 MkI/II (base plug not original to the body). The difference in diameter measured across identical sections is a very noticeable 3.5mm, and the REVO body is much more bulky.
This is purely down to manufacturing variation, and although the dimensions of the casing were specified, they were non-critical. The size could and did vary appreciably from the mean. If it was happening in late 1916 to mid 1917 on the the well established 23MkII, there is more than a slight probability that it was happening on the early transverse castings.
In July 1915 (and throughout the rest of his days) Mills complained bitterly that he was having to show competing contractors how to make and tool-up to produce the bomb. Part of that show-and-tell was helping competitors' pattern makers to produce their mould patterns. The pattern makers were skilled people, but whether for the early transverse castings, or for the later common longitudinal castings, every hand-carved wooden pattern would be slightly different - in size and shape. Hence we see today a small sample of the boundless variety that was produced in the 56 million or so of the No.5 family produced.
Mills' first style of body - the labour intensive transverse casting with machined horizontal grooves - was introduced to the first batch of 20 or so contractors in May-June 1915. However, on inspection drawings dated July 1915 the common longitudinal type casting is shown as an acceptable alternative. Though we will never know for certain, it is quite probable all these contractors had a go at the transverse casting, before adopting the less labour intensive alternative. Unfortunately it's very rare to find any No.5, yet alone transverse cast ones, with a named base plug that is proven to belong to it.
Tom.