OF PLANORBIS AT STEINHEIM. 25 



They arise as differences and remain differences, and have at no time any value as 

 representative characteristics. They, however, appear in the same forms as the repre- 

 sentative or parallel characteristics ; and the question now arises, can they also be reduced 

 to the control of the same law of succession in the series and in the individual ? 



Evidently there is no such succession, for if we take an individual out of any of the 

 series, and attempt to show this, we are met at once by the insuperable difficulty that 

 there is nothing to compare it with in other series. These characteristics, therefore, 

 present themselves with remarkable clearness, as increasing by heredity throughout 

 the series, or as in the case of the Fourth Series increasing only in a certain number of 

 species, and being then partly superseded by an evidently prepotent tendency in the 

 remainder of the species to revert to the rounded form of the whorl, or they are absent as 

 in the geratologous series, the second sub-series. Here, I think we have the key. The 

 geratologous series, if the comparisons made above are approximately correct, owes its 

 purely retrogressive character to the disadvantageous nature of the surroundings, and in 

 these the differential characteristics are not developed, while ui all others they are devel- 

 oped in precise proportion to their rank as progressive series. Thus in the first sub- 

 series, the fine crowded costse or the enlarged widely separated cost« or size, form alone 

 the progressive characters of the series. In the third sub-series, the development of the 

 third carination is decidedly progressive, as is also the serial difference of the Second 

 and Third Series, and even the Fourth as far as it goes. 



Not only do these differential characteristics as a whole progress, or increase, but they 

 are progressive in the differences, which they present within each series. That is to say, 

 that the gibbous underside of the whorl in PI. supremus is only a little more gibbous in 

 PI. supremus than it is in PI. "^^'/Z °^"* . The whorl of PI. ^Sr is more angular in PI. 

 crescens, the flattened lower side of PI. •s««"^'^^^«""'' stm flatter in PI. tenuis and PI. discoid- 

 eus, the costse of PI. S^Sus ^re largest in var. major of the largest size. There is then 

 uniformity in the way in which these differences of the series act, they are all 

 progressive, but their progress in each series consists alone in their increase in 

 intensity of expression or size in each series. Thus their uniformities are in the 

 strongest possible contrast to the uniformities of the representative characteristics. 

 These do not agree with each other in the same series, but have their uniform- 

 ities in the representative forms of different series, whereas the differential charac- 

 teristics have aU their uniformities in the same series, and do not agree with each 

 other in different series. The increase of intensity in each series is, as above stated, 

 directly proportional to the more or less progressive character of the series, being 

 nothing in the exclusively geratologous series, and most intense in the most progressive 

 or Fourth Series. 



If, therefore, the absence of the differential characteristics can be accounted for by 

 the action of disadvantageous surroundings in the second sub-series, it becomes evident 

 that the existence and permanency of the same class of characteristics in other series must 

 be due to the selective action of the same surroundings. This conclusion can be 

 farther sustained by the great increase in size of the First Series, or most progressive 

 or differential one, and the gradual decrease in size of each series towards the left of 

 plate 9. Proportionate size and weight are acknowledged by physiologists as the most 



