NATURAL HISTORY AND PHYSIOGRAPHY OF NEW BRUNSWICK 
423 
dancing movements increasing the illusion; fifth , its origin is 
probably electrical, and it is very likely a phase of the phenomenon 
known to sailors as St. Elmo’ s Fire. 
I have, of course made efforts to ascertain if any such phen- 
omenon is known elsewhere in the world. Professor R. De C. 
Ward, Assistant Professor of Climatology in Harvard Uni- 
versity, writes me that he knows of no record of a similar 
phenomenon, and no development of St. Elmo’s Fire so great 
that it could be mistaken for a burning ship. Professor A. H. 
Pierce, my companion in my visit to this region last summer, 
has, however, called my attention to references to an allied sub- 
ject in the Journal of the Society for Psychical Research, XII, 
1905, 108, and again in (the Proceedings of the same Society, XIX, 
1905, 80, where an account is given of lights claimed to have been 
seen around Tremadoc Bay in Wales ; but the conclusion is reached 
that in all probability they have only a subjective basis, though the 
statement is also made that lights of unexplained origin were 
reported as common on the Welsh Coast over two hundred years 
ago. It is also of interest to note that Schmitt’s newly-published 
Monographic de l’lsle d’ Anticosti (57) mentions manifestations 
of St. Elmo’s Fire observed at that Island. 
It is plain that in this phenomenon we have a subject which 
invites accurate investigation. It can best be studied by a 
scientifically-trained person, a physician or other student accus- 
tomed to scientific evidence, resident at Caraquet or Grande Anse. 
93. The Origin of the Northumbrian System of Rivers 
Read in abstract May 2, 190s. 
The most striking physical feature of New Brunswick con- 
sists in its abundant great rivers. These are, however, so diverse 
in direction, and interlock so complexly as to make the elucidation 
of their origin seem well-nigh hopeless. But prolonged study : s 
showing that in reality they are derived from three primal sys- 
tems, a Fundian system which I have described in an earlier note 
(No. 75), a Northumbrian System, which is here considered, 
