FOOD FROM THE SEA. 391 
water transferred to the artificial sea-water when the latter was being 
inoculated was not always quite the same, and that when a culture was 
inoculated two or three times a much larger quantity of natural sea- 
water was introduced than when it was inoculated only once. Was it 
the addition of this increased quantity of natural sea-water that enabled 
the diatoms to grow in the purely artificial solutions ? Definite experi- 
ments were made to determine this point, and it soon appeared that the 
previous irregularity of the results could be accounted for in this way. 
It was found that if the inoculations were always made by adding only 
a very small quantity of the liquid from which the living culture was 
taken, say, by adding just one or two drops of culture to a flask contain- 
ing 75 cubic centimetres of artificial sea-water, only a very slight growth, 
if any at all, took place. Uf, however, before the artificial water was 
inoculated in this way, and also before it was sterilized by boiling, as 
little as 1 part in 100 of natural sea-water was added to it, the diatoms 
grew well and excellent cultures were obtained. With 4 per cent of 
natural sea-water added to the purely artificial solution the cultures 
were quite as good as, if not better than the best that were got when 
natural sea-water was used entirely as the basis of the culture solution. 
This result is somewhat remarkable because it seems to show the 
absolutely essential importance to the growth of the diatoms of an 
extraordinarily minute trace of some substance which exists in the 
natural sea-water. In the artificial water all the salts were included that 
occur in natural sea-water in any quantity above a mere trace —a trace 
so small that it is hardly capable of accurate measurement. Yet when 
this trace of substance is diluted down by adding | part of it to 100 parts 
of artificial sea-water, there is still enough of it introduced to make a 
vigorous and abundant growth possible where no growth at all was 
possible without it. This substance can hardly itself be regarded as a 
food substance. Its real action can only be conjectured, but we may 
think of it as a growth stimulant without the aid of which the plant is 
unable to build up its structure out of the real food substances. 
In the course of further experiments it was shown that, provided this 
small quantity of natural sea-water were present, the composition of 
the artificial sea-water could be altered to an extraordinary extent 
without much if any effect being produced on the culture. — In the 
first place the density of the artificial water could be changed within very 
wide limits without affecting the cultures appreciably. The solution 
might be diluted to one-half the normal strength or even below, or it 
might be concentrated so as to increase the density by as much as 50 
per cent. 
All the Potassium chloride might be omitted or the amount of Potas- 
