SONNININtE. 
289 
coronate ancestor with sulcate-carinate ventral area. I fancy, however, that we 
cannot derive the Sonnininse from this source on account of Ziircheria — a genus in 
which the sulcate-carinate ventral area is not developed, and in which, therefore, 
we cannot suppose the absence of a sulcate-carinate ventral area to be due to 
"modifying action." We must suppose an earlier keelless coronate form 
corresponding to the inner whorls of Haplople^iroceras subspinatum (PI. XLIX, 
figs. 5, 5 fl) ; and by the laws of Ammonite-variation such a form preceded the 
coronate sulcate-carinate form of the snh-iamiij Amaltheijise. Practically speaking, 
then, the earliest state in which we meet with the Soimininse is the " unispinous 
stage." ^ 
Soc.,' vol. xlv, p. 653) ; and it is certainly more probable to explain them by " the modifying action " 
than to derive this form from a separate branch, as I did in that paper on account of the ontogenetic 
diflference. The smooth, globose, inner whorls o£ Liparoceras striatum (descendant of " capricorniim " 
through " heterogenum ") are explainable by the same theory — that all available material has been 
used to acquire the globose shape as the shortest method to attain the adult form of " striatum" and 
consequently correlating ornaments (shown by adult " heterogenum ") are not reproduced in youthful 
"striatum." The exigencies of space compel me to withhold fuller details of this most important 
biological phenomenon. 
1 It is necessary to briefly explain this and other terms to be introduced. The mutations by 
which various Ammonite-species arose may be referred to as "normal or developmental variation," in 
contradistinction to " abnormal " or " spontaneous " variation, such as " sports," or the pathological 
cases shown in PI. ix, figs. 8 — 10 ; PI. xlvii, figs. 15 — 17 ; PL li, figs. 13—15. Developmental 
variations follow each other in regular order, and appear to be gradual functional modifications 
adapted to meet the requirements of the individual ; but abnormal variation appears to be due to 
disturbance in the economy of the parent (" sports ") or of the individual (pathological cases). 
Developmental variations may be classed as progressive {anagenesis, Hyatt) and retrogressive 
{catagenesis, Hyatt). At any time in the history of a phylogenetic series the former may be arrested 
and the latter be substituted, or the series may divide into two groups, one continuing in progression, 
the other declining in retrogression. Retrogression is, shortly, the reversal of progression with certain 
new features (broader whorls, more compressed sides, &c.) introduced. 
Among Ammonites proper,* ontogenetical investigations have shown me that the progressive 
development variations may be referred to the following stages, which follow each other in regular 
order as here set down : — Grlobose (example — Am. gloiosus) ; smooth evolute {Am. miserahilis, poly- 
morphus) ; striate {Am. sir iaries, poly morphus lineattis) ; costate — the enlargement of strije into ribs 
{Caloceras, Am. planicosta, &c.) ; unispinous — excessive growth of ribbing at one point, producing 
a spine {Coroniceras, Am. ziphus, &c.) ; bispinous — additional growth of the rib at another point, 
making two spines on the side {Am. Birchi, hifer, Valdani, &c.) ; multispinous — a breaking up into 
several rows of spines {Am. TiawsJcerensis to a certain extent, and several Cretaceous species). 
The examples here given are those Ammonites which have arrived at the stage of development 
mentioned. In many cases {Birchi, hifer, &c.) their inner whorls clearly show the various previous 
stages following one another in the order above detailed. In other cases some of the stages are 
omitted from the ontogenetic record ; but it is certain — in many cases is demonstrable — that the 
* Equal to Ammonitinse, Hyatt, and excluding the Lytuceratince, Hyatt, which belong to a different 
radical stock. 
37 
