614 Mr. P. R. Lowe on the 



sliowiij the palatal plates of the Ruff will be seen to be in 

 respect of all these characters obviously Eroliine. 



(4) The Maxillo-palatine process in tlie Eroliinse is seen 

 to have completely fused with the pre-palatal portion of the 

 palatal plate of either side as a thin elongate plate with 

 parallel internal borders devoid of sculpturing. 



In the Tringinre the maxillo-palatine process is observed 

 to extend backwards on either side of the vomer as a very 

 attenuated pear-shaped or sac-like process, which is free 

 throughout the greater part of its course. 



In the Ruff the maxillo-palatine process is slightly 

 specialised and, curiously euough, its inner margin is slightly 

 crenated, but, as will be evident from the figure, it is 

 undoubtedly Eroliine rather than Tringine. 



(5) In the Eroliinse the inwardly-projecting plate of the 

 maxillary is fenestrated (cf. text-fig. 11 B, /e.) as it is in 

 the Ruft'. In the Tringinse this fenestration is conspicuous 

 by its absence, although there are indications of it. 



The above characters would appear to be sufficient not 

 only to differentiate between the Eroliinse and the Tringinse, 

 constant as such characters are in all the genera and species 

 of either subfamily which I have examined {cf. list below), 

 but also to demonstrate the fact that there is no longer 

 any possible excuse for grouping the Ruff with the 

 Tringinse. 



There are, however, other characters serving to distinguish 

 the Eroliinse from the Tringinse, and incidentally to clear up 

 the misconception as to the position of the Ruff. These, 

 owing to want of space, we can only refer to very briefly. 

 They are as follows: — 



(6) In the Eroliinse the line of the culmen of the bill 

 nLakes an obtuse angle (roughly 140°) with the basi- 

 sphenoidal rostrum. In the Tringinse the angle made by 

 these two lines is much more obtuse or nearly non-existent. 

 It thus comes about that in the Eroliinse the line of the 

 culmen forms a slope which is all but identical and con- 

 tinuous with the slope of the line representing the inter- 

 orbital depression. In the Tringinse the slope of this 



