THE AMEBIC AN NATURALIST [Vol. XLIII 



all of Kellicott's curves except that for brain weight clearly 

 show) to be widely divergent from mediocrity. But the "free- 

 hand" method of smoothing assumes, actually if not intention- 

 ally, that the average based on two individuals is just as "right" 

 (*. e., expressive of the true relationship which it is the purpose 

 of the investigation to discover) as is that based on 25. This 

 obvious error any adequate method of curve fitting will avoid. 



The bearing of these remarks further appears clearly in the 

 ease of some of the relative curves, wherein the percentage which 

 the particular organ weight is of the total body weight is plotted. 

 As was to be expected from what is known of growth in man 

 the general trend of these percentage curves is downward. 

 While this is the general trend, several of the curves (r. </., the 

 heart curve, plate 2) show at the very beginning a slight rise to 

 a maximum and then the downward curve. It is plain from the 

 discussion that Kellicott considers the rise at the beginning of 

 these percentage curves to be a real and significant phenomenon 

 of growth. It is very doubtful, however, whether the data war- 

 rant such a conclusion. Before accepting it one would like to 

 see the measurements of a much larger number of very young 

 (i. < .. just hatched) individuals added to the curves, and then 

 have a curve fitted by some adequate method to the observations. 



The general result of this interesting and careful piece of 

 work is to show that the regression line of organ weight on body 

 weight in dogfish of different sizes is not of the some form for all 

 organs. Some organs (e. g., rectal gland, pancreas) show a 

 nearly linear increase in weight as the body increases in size; 

 others (e. g., the brain) show the logarithmic like curve which 

 one associates with growth curves. The author gives an inter- 

 esting discussion of the significance of the fact that the muscular 

 and skeletal tissues tend to "outgrow" their visceral accompani- 

 ments in forms of indeterminate growth like the dogfish. He 

 regards the condition of determinate growth seen in higher verte- 

 brates as "an adaptation on the part of the organism, such that 

 muscles and supporting tissues cease their growth at such a 

 point that brain and viscera remain -competent to maintain a 

 physiological balance. ' ' 



In passing it may he noted that Kellicott's work, while itself 

 strictly morphological, suggests on every page problems for 

 experimental work on the physiology of the growth process. In 



