439 



labial enters the orbit, in one the sixth and seventh on both 

 sides do so (fig. 4), while in the remaining two examples 

 the labials are out off from the orbital margin, each by two 

 suboculars (fig. 5). 



Though apparently of fundamental importance, the 

 conditions here detailed may be produced by slight superficial 

 changes. If we turn to figs. 3 and 4 of my paper (1894), it will 

 be apparent that if the lower preocular and the seventh labial 

 of fig. 3 were fused, the condition would be similar to that 

 in fig. 4, apart from the number of the labial involved. Then 

 as regards the presence or absence of suboculars, the relative 

 conditions are really slight, depending upon the degree with 

 which the apex of the labial is pushed upwards towards the 

 orbit, tresulting in a separation of the orbiitals or otherwise. 

 Similar variability exists in regard to the posterior pre- 

 frontals : in one of two specimens of A . ramsayi taken 

 together, the shields are in contact for more than half their 

 length (fig. 6), whereas in the other they are quite separated 

 (fig. 7). The loreals are very unstable, ranging from one to 

 three, while in one example there are six on each side. 



In addition to the larger number of ventral and sub- 

 caudal plates found in A. melanocephalus, we may note the 

 smaller number of the lower labials, 14 to 17 ; they range 

 from 17 to 20 in A. ramsayi. The number of scales round 

 the body varies from 53 to 63, but the differences are not 

 correlated with the species. In most snakes the scales 

 are normally arranged in an odd number of rows, the ventral 

 plate connecting the two even sides. Having no enlarged 

 ventral plates, the Typhlopidae furnish an exception, bearing 

 an even number of scales round the body. Krefft counted 52 in 

 the type of A. melanocephalus, but on examination Boulenger 

 corrected the number to 53 ; similarly I am allowing 63 scales 

 for the sipecimen credited with 64 by Longman. 



The "entire" anal is probably a character of the genus, 

 though Macleay stated that the type of A. ramsayi had two 

 anal plates. ' 



As far as known also, the distribution of the two species 

 is different, A. melanocephalus being a northern form 

 recorded from eastern and northern Queensland, and A . 

 ramsayi transcontinental, occurring in northern New South 

 Wales, Southern Queensland, across the central districts to 

 Western Australia south of lat S. 23°. 



The following table shows the characters of the several 

 examples known to me by descriptions or specimens. The 

 details of ^4. melanocephalus, the type excepted, are kindly 

 supplied by Mr. Longman from specimens in the Queensland 

 Museum : — 



