1910] Johnson: Quantitativt Study of Salpa Chain. Ill 



the salpa chain are in general accord with previous observations 

 mi repetitive scries in plants and thai the chain furnishes an 

 excellent and hitherto unnoticed example of periodicity in growth. 



8. GRAPHS AS SPECIES CHARACTERS. 



A comparison of the graphs of the chains for the three species 

 shows slight differences and suggests that there may he a char- 

 acteristic graph for each species. It is possible that the course 

 of the graph may he affected by many things such as the form 

 of the chain— straight or curved, — the number of zooids in the 

 block, the size of the parent, rate of growth of the chain, and 

 possibly by other less evident conditions. 



If several chains of each species were measured at different 

 times of the year it might lie found that the differences suggested 

 by these preliminary graphs or other differences are constant for 

 the species. Any such investigation would have to take into 

 account the season of the year and the size of the parent ; thus 

 showing for a given species, the size relation of the zooids to each 

 other at different stages of their growth and any possible seasonal 

 variation as well as variation dependant upon the age or size of 

 the plant. These sets of observations could then be compared to 

 see if the graphs are in any way constant for the species. 



9. EQUISETUM EOBUSTUM. 



Since this relation is found to exist in such widely differing 

 structures as the zooids of a salpa chain and the leaves of a 

 branch, it is advisable to extend the investigation farther to 

 include repetitive series of as many kinds as possible in both the 

 plant and animal world. While it is not within my province to 

 go extensively into the matter so far as plants are concerned, I 

 will include a figure showing graphically the measurements made 

 upon a fertile shoot of Equisetum robustum. (Fig. 15.) 



I first measured the length of each internode and then the 

 diameter of each internode at each end and at certain inter- 

 mediate points. The measurements showed that the longest 

 internode is near, hut not at the base of the stalk and that in 

 each internode the maximum diameter occurs in the middle and 



