EXTERNAL FACTORS ON THE SPAWN DATE I49 



smell that guides a frog to its pond, then it must surely be the smell of 

 the algae, for it is these that are very well known to be the cause of the 

 smell of normal water. 



It will almost certainly be difficult to unravel this complicated sub- 

 ject by more field observations. The obvious thing to do is to try to 

 fmd some direct relation between the algae and the frogs. Work 

 on this is proceeding, but it may be of interest to see why it is difficult. 

 In a small pond there may be as many as two hundred species of algae. 

 Since -R. tetnporaria spawns in a considerable variety of habitats, it is 

 unlikely that the frogs are comiected with only one species. Probably, 

 on the analogy of most herbivorous animals, they are confined to a 

 large but not unliinited number of algal species. We shall have to 

 identify a group, any one of which may serve for the frogs. Algae 

 have very complicated life histories, and the water undertakings some- 

 times find that tastes and smells are at their worst when the algae are 

 dying away. But the critical period could equally well be a maximum 

 period of abundance, or the beginning of a cycle. Or it might be 

 some phase in the life of the plants, such as the liberation of swarmers. 

 Worse still, it could be that the frogs avoid certain species of algae. 

 The fishing industry now uses the presence of certain species of 

 planktonic algae that occur in patches of sea to show them where not to 

 cast their nets, for the fishes are not to be found in those areas. But 

 perhaps this is unlikely, in view of the migration observations, which 

 seemed to show definite attractions. Methods of sampling the sessile 

 species of algae, the most likely group, are difficult, particularly from 

 surfaces of mud. I know of none that can be used quantitatively on 

 floating weed mats, often used by frogs, and often encrusted with 

 algae. Finally, our knowledge of the events in small ponds is far 

 behind that available for lakes. Ponds are very comphcated places, 

 liable to violent changes in conditions. 



The direct experimental study of the weather factors is out of the 

 question. We can control the temperature of small laboratory vessels 

 quite easily, and light is not a difficult factor, if we are working imder 

 sterile conditions and can be sure that we are not dealing with indirect 

 effects. Humidity can be controlled by more elaborate instrumen- 

 tation. There is not sufficient knowledge available to imitate under 

 controlled conditions the immense complexity of rain falling on soil 

 and running through it into water. Even if we did know how to do 

 this, we should have to set up a large number of planned experiments 



