20 THE AMERICAN NATURALIST. [VoL. XXXIV. 
different individuals. On June 29 none of the stars meas- 
ured more than 1 mm.;! on July 18 one of the larger speci- 
mens measured 5 mm.; July 26, 9 mm.; August 18, 18 mm. ; 
September 26, 35 mm. ; October 25, 54 mm. (over 2 inches). 
Curve 1, Fig. 1, is a graphic representation of the growth of 
these stars and is constructed from measurements of the 
largest specimens which could be found in the car at each 
observation. At any time after the first few weeks some of 
the specimens were three times as large as others, and between 
these all intermediate stages could be found. On August 18, 
for instance, eight specimens, arranged according to their sizes, 
measured 7%, 8%, 10%, 12, I3, 14%, 16, and 18 mm. A 
similar series — or a more perfect one — might at any time 
be arranged from specimens taken along the shores. 
From the fact that in starfish of the same age such great 
differences in size cxist, with all gradations between the ex- 
tremes, it is evident that there is danger in applying to them 
Agassiz's method of ascertaining age. Indeed, the specimens 
which were calculated by him to be three years old were not 
one-fifth as large — linear measurement — as those known by 
actual observation to be only three months old. 
Coming now to the consideration of the conditions which 
determine the difference in the rate of growth in these animals, 
we may exclude certain factors at once ; for example, the influ- 
ence of light, temperature and density of the water, and size 
of containing vessel, inasmuch as these conditions are practi- 
cally the same for all the individuals under observation. 
It appears from some other experiments, however, that dif- 
ference in the amount of food is the principal, if not the only, 
factor. The starfish differs from the higher animals in this 
important respect; it can eat and assimilate many times as 
much food as is necessary merely to maintain a healthy condi- 
tion. When food is accessible, the starfish eats voraciously 
and grows with great rapidity ; but, on the other hand, it will 
live for months almost without food and apparently remain 
healthy, though it does not grow. As one watches a large 
1 All the measurements given below as “length” are from the mouth to tip of 
arm. 
