NATURAL SELECTION 285 



those from the base of the spit, and the latter were longer 

 than those from the ' mainland ' habitat. Lutz believes that 

 ' where the soil is loose ' — as on the sand-spit, especially at 

 the apex — those eggs which are not deeply buried will almost 

 certainly perish. In this way selection acts against the off- 

 spring of females having short ovipositors in a habitat where 

 the soil is loose. 



Confirmatory evidence is found in G. arenaceus, which 

 lives regularly on sand and has a long ovipositor, though it 

 is not stated whether the correlation is found throughout the 

 genus. Differences of unknown significance in the tegmina 

 and wings accompany the lengthening of the ovipositor. The 

 difference in average length between the ovipositors of the 

 crickets at the apex and those on the mainland is only 2 mm., 

 which is scarcely likely to provide sufficient extra depth to 

 be of much account. The range of variation overlaps very 

 considerably (see fig. 25). 



It is not easy to arrive at a decision concerning this case. 

 There is no evidence that eggs buried in the sand are uncovered 

 and destroyed more frequently than those of the mainland 

 animals. No exact expression of the density of the soil is 

 given. The differences in tegmina and wings, which are not 

 proposed as adaptive, might indicate a general ' colonial ' 

 divergence due to isolation between the three groups. This 

 is not much more than a fair theoretical case. 



4. Reciprocal modification of the head of the beetle Carabus mor- 

 billosus and the shell of the snail Otala tigri [Boettger, 1921, p. 321). 



Boettger states that in Morocco and Algeria, where the 

 snail develops a larger oral denticle than usual, which serves 

 to close the mouth of the shell, the Carabids, which prey on 

 the snail, have narrower heads. He gives certain facts con- 

 cerning the geographical variation which tend to confirm 

 his hypothesis ; but many difficult questions are not met, 

 e.g. whether in areas from which the Carabid is absent the 

 snail has a less pronounced denticle. Boettger also (p. 325) 

 weakens his case by suggesting that the denticle may be a 

 ' Verdunstungsschutz,' and is evidently in two minds as to 

 what its origin really may be. The subject is not treated 

 statistically and is scarcely evidential, though it is perhaps 

 suggestive. 



