44.0 REPORT—1899. 
It is necessary to change about two gallons of the water every day. 
At the end of 43 to 45 days some specimens were found on the bottom of - 
the jar metamorphosed. For convenience of observation the remainder 
were transferred to a number of half-gallon jars which were immersed in 
one of the Laboratory tanks to keep them cool. About three larvee were 
placed in each jar, but development did not go on as well or as quickly as 
in the large jar. 
The reflection will occur to most people that the method I have 
described is a roundabout and cumbersome one. Why not, it may be 
urged, transfer at once 200 blastulz to a ten-gallon jar? The answer to 
this is that I have tried this experiment, and that it failed. From my 
experience it seems as if it were necessary to allow Nature to select the 
healthiest larve ; the experimenter cannot pick out the blastule which 
are fit to survive. At the end of 14 days a larva has a much greater hold 
on life, and if unhealthy conditions supervene, such as insufficient change 
in the water, it will often continue to grow without developing the organs 
of the adult. 
Plenty of room is a cardinal condition for the success of all rearing 
experiments. When I had removed all the best larve to the half-gallon 
jars, I left behind in the ten-gallon jar a few of what I thought unhealthy 
larve. To my surprise and delight on the last day of my stay in Plymouth, 
I found practically all these either metamorphosed or just about to complete 
this process. To sum up, the necessary conditions for success are : (1) Selec- 
tion of perfectly ripe full-grown males and females for fertilisation ; (2) use 
of outside water ; (3) action of natural selection for the first week ; (4) the © 
use of the plunger ; (5) frequent change of water ; (6) shading from exces- 
sive light ; and (7) plenty of room. 
This is not the place to enter into a discussion of the internal changes 
which go on in development, a subject which I reserve for a later paper, 
but it seemed to me that these observations of the conditions of successful 
rearing might be of interest to a wider circle than specialists in echino- 
derms. It was some time ago usual to regard the larve of echinoderms 
as the most difficult objects to rear. So far from this being the case, I 
believe that of all trwe larve they are really the easiest. I use the term 
‘true larve ’ advisedly, for comparative embryology has too long confined 
itself to the study of cases such as those of Astacus amongst Crustacea, 
Pisidium and Cyclas amongst Mollusca, and Asterina gibbosa amongst 
Echinodermata, where the word ‘larva’ is only applicable to the young by 
a scarcely justifiable stretch of its meaning. 
A larva is exposed to the struggle for existence with the environ- 
ment, and depends on its own exertions for food, but this is not the case, 
as reflection will show, with most of the life-histories which, so to speak, 
have served as paradigms for comparative embryology. And yet, when 
one or two cases of true larval development have been successfully in- 
vestigated, how full of meaning and interest they have shown themselves 
to be! I need only mention the instances of Lucifer and Penzus 
amongst Crustacea to prove this. The difficulty in such studies has 
always been the question of rearing. 
I hope that the observations recorded above may be of assistance to 
any investigator who is attempting to make advances in this field, which 
seems to me to be one of the most hopeful for the future development of 
comparative embryology. 
