In the 1930s and subsequently, Lucas supplied a car voltage regulator of the twin bobbin type.
A detailed description of how it operated can be found here, but in summary, one solenoid operated a cut-out to prevent reverse current from the battery discharging through the dynamo, and the other controlled the voltage available to the dynamo field windings via a pair of vibrating contacts.
A pair of so called "series windings" were wound as secondaries on the primary windings controlling the above functions. The winding on the cut-out, which carried the dynamo output and operated through the above mentioned cut-out contacts, was 2mm o/d round enamelled wire of about 14 turns. This was connected to the series winding on the voltage regulator solenoid by means of a soldered joint.
This second winding was made of rectangular section enameled wire of approx 3mm x 0.5 mm. It carried the same current as the winding on the cut-out but had only half the cross-sectional area of the round wire winding.
It is noticeable on old units that the enamel on the cut-out series winding is often perfect whereas the enamel on the rectangular section winding may be dark or burnt. Clearly the power was being preferentially dissipated in this higher resistance winding because of its smaller cross-section.
What I would like to know is why Lucas used a rectangular section wire at all, and having done so, why didn't they select a cross-section the same as the round wire?
There must have been a reason for using the rectangular section wire that outweighed the cost of the extra operation of making the soldered joint and having two different types of wire on hand.
Since the fields produced by the series windings are only dependent on Amp-turns, it is understandable that modern repros of these units use round wire of identical size for both the series windings. Why did Lucas not do this?