20 The Church of All Saints, Martin, Wilts. 



chapel (not central with either of the two easternmost bays, nor quite 

 opposite the pier coming between them) is a very remarkable recessed 

 five-light bay window of quite a domestic type, but coeval with the 

 enlargement of the chapel, and like the east window there is no 

 cusping in the head ; it projects on the outside and is roofed 

 ti-ansversely with the rest, the recess is carried to the floor inside 

 (not like the somewhat similar specimen at North Bradley, where it 

 stops at the sill level, forming the mensa of a tomb) and is separated 

 from the chapel by an archway of the same type as the two opening 

 into the nave and chancel, respectively. These arches of two orders 

 of chamfers spring from pier shafts with moulded capitals of debased 

 type, and the centre from which they are described is below the cap 

 level. There are two small crosses cut on the abaci of the caps to 

 the bay. A squint was formed at this time between the chapel and 

 the chancel, directed towards the high altar, and a large piscina 

 with square sunk bowl (without projection beyond the wall) was 

 inserted in the south waU of the chancel. 



At about the same time important alterations were made in the 

 nave. The walls were raised to theii- present level (the com-sed flint 

 and stonework clearly distinguishes this fi-om the Norman work on 

 the south side), and the waggon-head roof of four bays with tie- 

 beams and plaster panels, which now remains, was put on. The 

 westernmost window on the north side, without cusping, label, or 

 inner arch, was also inserted ; it has since lost its mullions. The 

 other window in this wall is an eai-lier insertion (circa 1430) and has 

 an outside label mould and inside arch, but it, too, is now without 

 mullions or tracery. [The easternmost window in the south wall of 

 the nave is a modern insertion.] 



In spite of the tower having already shown serious signs of 

 settlements, the builders in the first half of the fifteenth century 

 did not hesitate to raise it by one stage, and upon this to erect a 

 stone spii-e, but before doing so they proceeded to strengthen the 

 thiiioenth century substnietui'e, the foundations of which were very 

 defective. Underpinning of existing walls does not seem to have 

 been practised in the mediaeval period, but instead of it one fre- 

 quently meets with immense buttresses and ties, which must have 



