Thanks for that information.
I go along with Gadger's remarks. I think it likely that the arms are original & the balls later or repro. (one had a slight chamfer & the other an unbroken face where it entered the cup!)
I will grovel under the car with a sharp 3/8 drill....I'm not even thinking about taking the cups out...and clean everything.
I will use the 3/8 BSF bolt I have with a normal spring washer and lock nut it for good measure....the bolt is long enough.
It will interest me to see if this stops the front brakes coming on when applying steering lock....my original problem, before I found the radius arms to be disconnected!

Is the patent number someone else’s used, or an Austin one? Hard to believe the idea new.

As the assembly does not bolt up solid, the later nut and split pin (or locked nuts) seems more appropriate than a spring washer.

Davids description has me lost.

The bolt must be a very free fit thru the ball so it can tilt. Expect a normal fit thru cups and radius arms, but does not really matter if loose, as bolt only prevents the arms parting.

Normally, faces of radius arms should be flat and perpendicualr throughout length. May have become twisted, or been twisted to fiddle castor.

Location: Auckland, NZ

sorry to confuse you Bob!
The bolt is a free fit in the cups and the ball, but the hole diameter in the arms will not accept the 3/8 bolt.
I think the o/s arm may have been "adjusted" for some reason....the arm was under some tension, but could be persuaded to fit onto the ball. the N/S lined up ok. Both now have vertical faces.
I note your comment about the tightness of the joint & may forgo a spring washer
Been getting rusted & broken steel screws out of the hood front rail today, so maybe get underneath tomorrow!
David