THE BEHAVIOR OF LIMPETS 391 



homing ability, and second, to ascertain something as to the 

 nature of such ability should it be present. 



All of the limpets under observation were found to move 

 about, at periods of high tide only. The movement began with 

 the first wetting from the incoming tide, and proceeded more 

 or less continuously, till the retreating water left the animals 

 high and dry once more. 



About 30 limpets, representing four species of Acmea and 

 the one species of Lottia, were marked by filing Roman numerals 

 into their shells. This method made it necessary to mark the 

 animals but once, for the grooves could be filed quite deep into 

 the shells without injuring the animals in the least. 



The limpets were chosen, so that all the possible situations 

 were represented. Some were on horizontal rocks, some on 

 vertical ledges; some were exposed, some were not, etc. The 

 spot on which the limpet was resting, on the first day of obser- 

 vation, was enclosed in a small rectangle scratched into the 

 rock and alongside this rectangle the roman number which the 

 limpet carried was also filed into the rock. From day to day 

 the location of each limpet was determined by referring it back 

 to this numbered rectangle. The daily positions were plotted 

 on squared paper and the resulting graph, together with the 

 field notes, constitutes the permanent record of the limpet's 

 activities. The graphs that follow will furnish an idea of the 

 behavior of the different species of Acmea, during the period 

 of observation. 



Figure 1 indicates the wanderings of an individual of the 

 species Acmea spectrum, during a period of 27 days. It is at 

 once evident that this animal did not habitually return to any 

 given spot on the rock, when the tide retreated. It will be 

 noted further that there was even a tendency to change locali- 

 ties. This individual was situated on a flat rock, where the 

 waves washed it much of the time. 



The readings represented in Figure 1 were taken but once 

 a day and it soon became obvious that there was considerable 

 to be learned by observations made at more frequent intervals. 

 To check up this point a number of days was spent taking hourly 

 observations upon individual limpets. Figure 1A shows the 

 result of hourly readings upon the individual referred to in 

 Figure 1. Figure 1A, from its appearance, might represent a 

 series of 5 consecutive readings taken from some part of Fig. 1, 



