LOCAL VARIATIONS AND VIBRATIONS OF EARTH’S SURFACE. 57 
that did not seem n A self-recording aneroid works 
with the lake register to show any variation in barometric pres- 
sure, and there is nothing in its record on July 28th to justify 
this ‘change. The barometer was 0200 inch below the average 
to the mean. Its changes, therefore, do not accord with those in 
the lake on that day, and I am unable to offer any solution of the 
difficulty, except a change in the vertical. The recording machine 
is fixed on six piles driven into the bed of the lake, and then 
braced together so as to make the support as firm as possible. 
No change can take place in the relation of the pencil and the 
paper except by a change in level of the water, or a change in the 
length of 6 feet of brass chain connecting the float and the wheel, 
ny chan 
inappreciable. I mention this to show that the change in the 
level of the water cannot have been due to instrumental changes, 
and there was no wind to produce it, nor any sufficient change in 
the barometer to account for it. 
The diurnal change can be seen in the photoliths ere on 
July 30, 31, fat 1 and 2, and still more on March 14 and 15. 
of registration. Mare and 15 also show this curve as super- 
added on to other curves or waves, and as the observer expressly 
states that there was no wind on these dates, the curve cannot be 
of recording such complicated phenomena. An observer watching 
such changes in a level or pendulum would naturally take the 
observations at intervals, and these photoliths show clearly that he 
might determine to do it at such oP say every half-hour, as 
would give him a very imperfect representation of what was going 
on, whereas the ert SRE — ens clearly all the changes. 
On the record ies July 28 a good example will be seen of the 
periodic waves, whose vod is about two hours and eleven 
minutes ; the greatest a of level yet recorded in these waves 
is 13 inches, The first wave of a set is nearly always the largest, 
and it often happens fabs a wave is left out, or its period altered 
by a small or secondary wave. At A, for instance, the interval 
between the crests is nearly double that of subsequent waves, and 
at B a small wave seems to cut off the crest of the larger, one. 
These are common features, both in calm and windy weather, and 
are very perplexing unless they represent changes in the vertical. 
As I pointed out just now, two hours after the northerly wind 
ceased on _ July the water began to fall at the south end of 
the lake, and fell 54 inches, when there was not a breath of wind - 
