I forgot to say that I also have a therspit c.1250 (in Barking, Sf), and this is a recorded compound:
Ondlang žęs sices innon žone žyrs pyt (http://www.esawyer.org.uk/charter/222.html). This would all seem to confirm the interpretation.
Keith
PS: the recorded cases of žyrs in the OE corpus are:
sing pl
N žyrs žyrsas
A
D žyrse
G žyrses šyrsa
Dear Keith, Dear All,
There’s the unlocated fieldname Thursmare 1227 in PN E Riding p328. Also perhaps relevant are Thrusmyre C15 in Wighill and Thursemyer 1553 in Broughton, PN W Riding 4 p244 and 6 p44; and le Thursemyre in Windermere, PN Westmorland 1 p198. There are also compounds of grendel with mere, and Grendel the literary character is described as a žyrs.
Jeremy Harte
From: The English Place-Name List [mailto:[log in to unmask]]
On Behalf Of Keith Briggs
Sent: 03 January 2018 09:38
To: [log in to unmask]
Subject: Thersmere
I have found four separate places called Thersmere in Suffolk, in documents from the 13th to the 16th century. It seems that the first element must derive from OE žyrs, a giant or demon. This would be the same etymology as proposed by Ekwall (DEPN) for Tusmore in Oxfordshire, although Gelling was lukewarm about the idea in PN O i.216. Are there other examples of the compound elsewhere?
Keith