368 OVIPOSITION AND HABITS OF CERTAIN BATRACHIANS, 



yellowish ; and of R. phyllochroa yellow, the lower pole slightly 

 tinged with yellow. Exceptionally one may see this kind of 

 spawn floating free, or attached otherwise than as described; 

 but this is probably accidental or due to the absence of grass- 

 stalks or twigs. The spawn of this section may be met with 

 under favourable conditions at almost any time of the year ; even 

 in mid-winter. 



The third section includes the spawn of Pseudophryne australis 

 and P. bibronii, already referred to, which may be found during 

 summer and autumn. (For further details see p. 376.) 



During copulation the males in some species clasp under the 

 arms, in others round the waist ; thus the embrace is axillary in 

 Mixophyes fasciolatus, H. aurea^ H. citropus, H. ewingii var. 

 calliscelis, H. p)hyllochroa, and H. ccerulea ; it is inguinal in Lim- 

 nodynastes tasmaniensis, Hyperolia marmorata^ Grinia signifera, 

 Pseudojohryne australis, and P. bibronii. In a footnote to his 

 description of Limnodynastes ornatiis Mr. Boulenger says (I.e. 

 p. 262) "that one of the females of this species has on the breast 

 two cicatrices which are evidently caused by the thumbs of the 

 male ; this proves that the male seizes the female under the axillae 

 and not round the waist." In L. tasmaniensis, as I have had 

 ample opportunity of observing, the embrace is inguinal as stated 

 above. I have had only a single and early opportunity of observ- 

 ing the coitus in L. dorsalis and L. 2^e'^onii, when I was not suffi- 

 ciently alive to the desirability of noting the mode of embrace, 

 and I regret to say that I have never had the chance since of 

 repeating the observation ; hence I am unable to speak with con- 

 fidence in the case of these two species. 



The tadpoles of the different species in their earlier stages offer 

 few characters sufficient for their identification — unless perhaps a 

 study of those of the mouth with its horny fringes would yield such. 

 As the completion of the metamorphosis approaches, however, the 

 determination becomes less difficult, the characters of the webbing, 

 or of the metatarsal tubercles, or of the disks when the hind limbs 

 have developed, being among the earliest satisfactorily recognisable 



