350 THE FRESH-WATER LOCHS OF SCOTLAND 
On 15th July 1903, in Grass Water, the fourth spine was absent, 
the second and third spines were parallel, the second being long. 
On 17th July 1903, in Clings Water, the surface temperature being 
54° "8 Fahr., there was much the same kind of Ceratium as that in 
Grass Water (Plate XV. fig. 23). 
On 24th July 1903, in Loch Burraland, the fourth spine was 
absent and all the other spines were comparatively short ; the body 
was relatively large and coarsely reticulate. 
On 4th August 1903, in Loch Cliff, the surface temperature being 
56° '3 Fahr., the fourth spine was absent, and the second and third 
spines were small. 
On 6th August 1903, in Loch Snarravoe, the fourth spine was 
always present, though small ; all the other spines were somewhat 
reduced, though not much so ; on the whole, however, it can be 
described as a small and stumpy form. 
The interpretation of the extreme variability of Ceratium hirundi- 
nella from different lochs is not verv obvious, though doubtless the 
temperature is an important causative factor. That this variation is 
not erratic, but has some relation to the physical conditions of the 
environment, is shown by the fact that in the Shetland Islands, where 
the temperature and other conditions are fairly uniform, the Ceratium 
is all of one type ; moreover, generally speaking, all the individuals 
of any loch in any of the areas are either identical or they differ but 
slightly amongst themselves, and we do not find the two extremes in 
any one loch, though they do sometimes occur in difi'erent lochs of 
the same neighbourhood. 
We have seen that in the Shetland Islands the organism was 
always small, stumpy, and coarsely reticulated, the spines short, and 
the fourth spine absent or small ; in the Lewis lochs, on the other 
hand, we found a larger and more slender form, the spines were long, 
the fourth spine being sometimes as well developed as the third. 
And considering only the North Uist and Benbecula lochs, it was 
usual to find in the lakes of late spring a small short-spined form 
with no fourth spine or only a very small one, whilst in summer, when 
the waters were warmer, the Ceratium was of a large and slender type. 
And as I have already mentioned, the Shetland lochs were on the 
whole colder than those of Lewis, so that quite possibly the differ- 
ences between the extreme forms are consequent on the difference in 
thermal conditions. 
Nevertheless, I cannot think that this is the whole explanation of 
the difference, for in a number of cases — e.g. Loch Langavat (Lewis), 
17th July 1903, surface temperature 56° '1 ; Loch a' Ghlinne-Dorcha 
(North Uist), 25th June 1904, surface temperature 55°'0 Fahr.; Loch 
Crogavat (North Uist), 25th June 1904, surface temperature 55° "2 
