290 
INFERIOR OOLITE AMMONITES. 
As will be seen from the foot-note below, it is possible to infer from the 
phenomena of progressive developmental variation that the prior stages of the 
Sonnininse and Amaltheinae were, in backward order, a costate stage, a striate stage, 
a smooth evolute stage, and a globose stage. There is in the inner whorls of 
Sonninia and Witchellia evidence of a globose stage remaining, and in the inner 
whorls of ZurcJieria a certain amount of evidence of a striate and a smooth stage 
succeeding the globose stage ; but the other stages have been omitted on account 
of " the unequal action of earlier inheritance" (p. 224). 
I am unable to suggest any species which might represent the ancestors of the 
Amaltlieidsd in the costate, &c., stages. Prof. Hyatt points out to me the likeness 
of the inner whorls of the Amaltheiclse to the inner whorls of Coroniceras ; but this, 
I think, is too long a gap and I fancy that the resemblance of these inner 
whorls in these respects is merely homoplastic. 
The majority of the Sonnininse do not progress beyond the unispinous stage, 
but exhibit the various phases of retrogression to the costate and striate stages of 
decline. Haplopleuroceras, however, is an exception, for it attains to thebispinous 
stage. 
The annexed genealogical table shows the relationship existing between the 
Sonnininse and the Amaltheinse. The reasons why I have found it necessary to 
modify to a certain extent the surmises put forward in my paper on " Descent of 
Sonninia, &c." (' Quart. Journ. Geol. Soc.,' vol. xlv), as regards Sonninia, Haplo- 
pleuTOceras, &c., will be found in the articles on those genera. 
phylogenetic series passed through them just the same. Ontogeny does not always repeat phylogeny i 
— certain stages of development may be omitted from the ontogenetic record — because of — 
1st. The "unequal action of earlier inheritance " (p. 224). 
2nd. The " elimination o£ dissimilar stages" (p. 159). 
3rd. The "partial or modifying action of earlier inheritance." 
The developmental variations mentioned above progress independently of changes in the ventral ; 
area. A carina may appear at any stage, or may not appear at all. 
Retrogressive developmental variation may commence at any time, and the phylogenetic series, I 
or part thereof, may acquire the reversed series of changes. If it has reached the costate stage it \ 
retrogresses to the striate (Calocems, Dumortierid) ; if the spinous, it goes back to the costate, , 
then to the striate, and so on. There are, however, instances, but they are not many, where a new | 
phase of progressive developmental variation succeeds a period of retrogression, and then the same j 
progressive sequence is again maintained. | 
In retrogression " the elimination of dissimilar stages " (p. 159) plays a very important part in j 
obscuring the ontogenetic record. The reversing of the changes frequently results in bringing a j 
dissimilar stage between two approximately similar stages ; as, for instance, in the aged form ot I 
siphus," Zieten, a unispinous stage, comes between the progressive costate stage and the retro- 
gressive costate stage. 
It is thought undesirable to give a more detailed account of the important biological phenomena 
of progressive and retrogressive developmental variation at this stage of the Monograph. A. more 
elaborate treatise is in MS., and vvill probably be published at another time. 
