268 COLOURATION OF REPTILES AXD AM I'll IIUANS. 



tainly not have seen it. Another female found on the roadside 

 at the same phice was greyish brown with yellow patches on the 

 dorsal line and on the sides. No two creatures could differ 

 more completely in colour than these two females. A male 

 A. hispida observed on a hill at Somerset West the same year was 

 a varied green and olive ; this specimen would have escaped ob- 

 servation also had it not been in my path. When it did move, 

 as in the case of the two females, it was but for a few feet, and 

 then it became motionless. I noticed that if my attention were 

 taken from them for a moment I had some difficulty in locating 

 them again. 



I have never known A. aculcata to turn green ; probably the 

 aridness of its terrestrial habitat has something to do with it. 



A rapid and very localized change of coloiu% not found in 

 other lizards, is exhibited by the females of Agama aciileata, but 

 is unknown in the males. \Mien teased she sometimes exhibits 

 scarlet blotches on the back at each side of the vertebral line; in 

 these blotches the colour extends even to the tips of the scales. I 

 noticed that these appear when the creature meets with a violent 

 death, especially if the head be crushed, and that bearing females 

 produce them very easily. They are perhaps calculated to have 

 a terrifying effect on attacking foes, as presumably is the case 

 when chameleons suddenly change colour on alarm. 



The male Agaiiia atra can make itself perfectly black dor- 

 sally with or without a white vertebral streak, to suit the rock on 

 which it sits. It may also assume a light pink dotted with ocelli. 

 The range of colour possessed by the female is similar, and in 

 addition she is sometimes covered with red blotches, as in the 

 female A. aculcata, though of a deeper red; the ground colour of 

 the body on such occasions is a lemon doited over with ocelli. 



I have frequently watched specimens of the frog Rana 

 fuscigiila, which were light green adorned with a light vertebral 

 line, leaving the water on the edge of a dam. The pale green 

 ground broken by the dorsal streak harmonized so effectually 

 with the soft green shadows, the reflections of the trees, and 

 with the faded willow leaves on the side, that if I had not actually 

 seen them leave the water it would be almost impossible to 

 find them. They seem to fade awa}- and become lost in the 

 colour of the surrotmdings, and if the attention is distracted for 

 a moment it would be some time before their exact position could 

 be located again. These same specimens, when put in a box, 

 became a uniform brown or sepia. Others, again, are often 

 spotted with dark brown according to the light and shade of the 

 surroundings. 



Another striking example of variable adaptive colouration is 

 found in Xciiopus lu-x'is. Specimens taken from dark, deep wells 

 are uniform black dorsally ; those in shallow pools, where the 

 surroundings are bright, are usually a light sand colour; while 

 those in dams overhung by trees where the water is fairl}- clear 

 are spotted and marbled in sympathy whh the light and shade. 



The snake Psaiiiniophis fiircafiis furnishes a good example of 



