A CRITICAL EE-STATEMENT OF THE BIOGENETIC LAW. 93 
luside this rapidly growing bouse (the roof of earlier whorls being periodically- 
destroyed to make room for their successors), the larva secretes a second 
flatter shell (the rudiment of the adult shell) which fits his visceral hump 
and adheres to it, but is temporarily fixed in the cavity of the outer shell 
until he shifts his position, when it is carried forward — as though he were 
trying to improve on an Ammonite's arrangements by the device of a 
portable septum *. Now, the point of the analogy is this : that the spiral rows- 
of tubercles on the outside of the shell are variable in different species, and 
that in the same individual they may go through a cycle of changes exactly 
like the progressive changes of an Ammonite. The shell may pass through 
a smooth stage, a unituberculate stage (the outer row of tubercles), a di- 
tuberculate stage (with both rows developed), a spiny stage, and lastly a 
ribbed stage, in which cross-ribs join the tubercles of the two rows together. 
Unfortunately for the completeness of the analogy, Ecliinospira does not (so 
far as I have yet seen) present a gerontic stage, for, being only a larva, and 
usually very lively and vigorous, he quits his cage before old age comes 
over him, and transEorms himself into a torpid Ascidian-eating Lamellaria. 
I hope to publish shortly some figures of the remarkable process of meta- 
morphosis, of which I was lucky enough to be an eye-witness last year at 
Plymouth, as well as some further details of the growth of the larval shell ; 
but for my present purpose I refer to Simroth (1885, text-fig. 5 ; Taf. xvi. 
figs. 1 (fe 2 ; xviii. figs. 1-2, 6-8), whose excellent figures sufiiciently 
illustrate my immediate points. Meanwhile I submit (1) that the characters 
of the larval shells of Lamellaria and its allies are purely cenogenetic, with 
no relationship to the characters of any adult ancestors ; (2) that gradual and 
progressive changes in the shell of the same individual, from one type of 
"ornament" to another, occur regularly, and are apparently determined by 
the constitution^ size, and vigour of the larva under tli* particular conditions- 
of its existence ; and (3) that different degrees of the power of tubercle- and 
spine-formation characterise the larvas under different conditions of existence. 
In this case, from which all specific influence of adult ancestry is excluded, 
there is no escape from the conclusion that the power to perform these 
variously graduated operations, and the extent of that power, are essentially 
functions of the zygotic constitution, though there is a considerable margin 
for the direct influence of conditions. I conclude that, if this is so for the 
"cenogenetic^' larvte of Lamellaria, it is not likely to have been different 
for the " palingenetic " stages of Ammonites. 
15. I return to the keynote with a direct comparison between the ontogeny 
and phylogeny of an animal in which the skeleton has been an important 
index of racial structure thi-pughout geological time — the Grinoid Antedon, 
* The two shells correspond to the two layers of an ordinary Molluscan shell, dislocated 
from their original union. The outer or larval shell, corresponding to the prismatic layer,, 
ia formed by mantle-edge alone ; the inner or adult shell, correspondiug to the nacreous, 
layer, is formed by the visceral surface of the mantle alone. 
