BY W. J. RAINBOW. 541 



and when she is free from maternal cares ; (second) the ncM 

 wherein the animal moults, or in which the cocoon is established; 

 and (third) the cocoon containing the eggs. 



In the construction of the retreat varying degrees of perfection 

 are displayed, and these are influenced by circumstances, such as 

 the length of time a spider occupies a nest, for it not infrequently 

 happens that after establishing a home in one locality it will 

 reject it for another, probably out of consideration for its larder, 

 the position taken up in the first instance not being sufliciently 

 productive; again, another reason for changing may arise from 

 exposure to disturbance or attack. It may, therefore, be laid 

 down as a law that the higher or lower the degree of perfection 

 attained is due entirely to the length or brevity of the spider's 

 sojourn in the retreat. This fact has been recorded by Wagner,* 

 who, in explanation, observes that " the shorter its sojourn, the 

 simpler the structure of the retreat; the longer the sojourn of the 

 animal the greater is the quantity of silk used, and the more 

 careful the workmanship in the construction of the retreat" [j^lus 

 il est court — plus simple est la sti'ucture de la retraite, et inverse- 

 ment : plus longtemps y reste 1' animal — plus grande est la quantite 

 de fils, qui s' accumulent, et plus soigneusement se fait la retraite.] 

 When the retreat is constructed on the spur of a shrub, such as 

 the ti-tree or Leptospermum, as in the case of Chtbiona moc/esta, 

 Koch, it is secured by threads to the stem and its leaves, but 

 when on the other hand a broad leaf is selected, the spider takes 

 up its position on the reverse side and draws the edges over, 

 binding them in position by a series of threads. In describing 

 the retreat of C. pallidula, Clerck, Wagner has given a series of 

 figures! elucidating the manner in which the edges of the leaf are 

 brought together, and as the modus operandi is the same through- 

 out the species, they may be briefly explained. The first figure 

 (fig. 161 on the author's plate) represents a transverse section of 



* Loc cif. p. 119. 

 t Loc. cit., pp. 119-122, PI. II. figs. 161-164. 



