Xll. LIVERPOOL BIOLOGICAL SOCIETY. 



narrowest places the channel is crossed by cause- 

 ways, the northern one carrying the main road 

 and the L. and N.W.R., and the other being 

 merely a roadway. In each there is an opening 

 which partially admits the tide to so wide an 

 expanse of lagoon that the water never reaches 

 flood tide level nor falls to low-water mark. One 

 of the bays, or inlets, of this lake stretches west- 

 wards towards the sea at Tre-arrdur Bay, from 

 which it is only separated by about a quarter of a 

 mile of salt marsh and a range of sand dunes. At 

 high tide, therefore, there is nothing but the sand 

 dunes, backed by some sandy fields, which may 

 cover rock, to keep the sea from cutting Holy 

 Island in two and entering the inlet of the lake. 

 There is, however, no record that it has ever done 

 this. But in Tre-arrdur Bay are stumps of trees 

 in situ on the shore from about half-tide down- 

 wards ; as far as I could judge, below the level 

 of the lake and its marshes. A legend tells that 

 St. Bridget of old crossed from Ireland on a green 

 sod, which became a hillock, on which she built a 

 chapel. This chapel was, as it seems, in existence 

 a century ago, surrounded by oak trees, said to be 

 the same whose stumps remain. But the sea 

 made an inroad, and destroyed the chapel and the 

 trees. Now the northern causeway is compara- 

 tively new, and, consequently, the level of the 

 water in the lagoon must have formerly been the 

 same as that of the sea. Had the sea merely 

 forced back the sand dunes, under present con- 

 ditions the trees might have been growing 

 below high- water mark, though it is unlikely. 

 But, as they stand, the only conclusion seems to 



