6 THE SALAMANDER 



year. The amplexus thus normally occurs while there are seg- 

 menting eggs in the oviduct, but whether this is always the case is 

 uncertain, neither is there any evidence as to whether the sperms are 

 able to fertilize the eggs immediately they leave the male, or whether 

 the period of rest in the receptaculum seminis is essential, and if so 

 what changes, if any, they then undergo. The embryos developing 

 from the eggs fertilized in June have attained full maturity by the 

 autumn, but they are normally carried over in the 'uterus' until the 

 following spring, i.e. for a total period of ten to twelve months. 

 Under favourable circumstances, however, two batches of larvae may 

 be deposited in the same year, and both copulation and deposition of 

 larvae may occur at almost any period of the year except during the 

 winter. 



2. Deposition of Larvae. 



The female deposits the larvae in water, usually in the springs and 

 small streams which abound on the lower slopes of the mountains of 

 central Europe. She does not enter the water bodily, since the adult 

 Salamander is unable to swim and would speedily drown if totally 

 immersed in water beyond its depth, but merely remains at the 

 water's edge with her cloaca and tail submerged. Deposition may 

 occur at intervals over a period as long as two days. The total num- 

 ber of larvae deposited by any female during this period varies from 

 twelve up to as many as seventy-two (Fatio, 1872). Such high 

 numbers are, however, rare and the average may be taken in the 

 region of thirty, and they occur in approximately equal numbers 

 in each oviduct. Dead larvae in various stages of development, and 

 of degeneration, are by no means uncommon and are deposited 

 together with the living ones. The embryos remain within the 

 vitelline membrane until the very moment of deposition, and may 

 in fact actually be 'laid' within it, but normally the act of extrusion 

 ruptures the membrane and allows the larva to swim free. The 

 empty membrane is then pushed out from the cloaca by the suc- 

 ceeding larva. The Salamander is thus ovoviviparous rather than 

 strictly viviparous, that is to say, it lays eggs but the eggs have pre- 

 viously developed within its body to hatching-point before their 

 deposition. The embryo feeds exclusively on ^^o^ yolk. Kaufman 

 (19 1 3) finds that the number of degenerating larvae is proportional 

 to the length of the 'uterus' and the number of embryos, and she 

 therefore concludes that the death of these larvae is due to pressure. 

 She also finds that the weight of a full-grown embryo is i -6 times the 

 weight of the egg from which it developed, and since it is independent 



