638 



SCIENCE. 



[N. S. Vol. XXII. No. 568. 



correction. In my paper^ I concluded as fol- 

 lows, concerning the observed variations in 

 the number and arrangement of the scutes of 

 the loggerhead, Thalassochelys caretta: 



These variations fi'om the normal type all lie 

 in the direct line of descent, and the more serious 

 the variation, the further back it points. More- 

 over the changes necessary to turn any given 

 variation into another one less abnormal until 

 ultimately the normal condition is reached, are not 

 erratic, but stand in strict correlation with each 

 other and proceed strictly on definite lines. I, 

 therefore, call this kind of atavistic variation 

 orthogenetic. 



Further, I had suggested that the indi- 

 vidual, abnormal turtles ' grow out of these 

 irregularities by the reduction or squeezing 

 out of certain of the scutes,' and I concluded 

 as follows: 



Of course there is no proof of what I have tried 

 to explain. Comparative anatomy and common 

 sense tell us it is so. But common sense is not 

 evidence in a sceptical court. The only way of 

 proving the correctness of the view explained in 

 this paper would be to take a number of abnormal 

 turtlets and to watch, while they are growing up, 

 if and how they mend their irregular shells and 

 become normal. 



I think I had stated the case fairly. It, left 

 no doubt about the definition of at least one 

 kind of orthogenetic -variation. Of course, it 

 can not be expected that a turtlet with, say, 

 twenty-four dorsal scutes can, during its 

 growth, reduce them to the normal number of 

 sixteen. I had said as much. Many a turtlet 

 probably sticks fast during the mending proc- 

 ess. Otherwise there could be no abnormal 

 adult specimens, if my surmise is correct. 

 But if they do amend the number of scutes, 

 these cases of orthogenetic variation are 

 simply ontogenetic stages, passing reminis- 

 cences of earlier, phylogenetic conditions. We 

 ourselves begin with an embryonic tail of 

 numerous metameres, and their ultimate re- 

 duction to five, four, three or two free caudal 

 vertebrae falls under the same category of 

 variation. 



Orthogenetic variation implies something 



^ ' Orthogenetic Variation in the Shells of 

 Chelonia,' Willey's Zoological Results, Pt. III., 

 May, 1899, pp. 207-222, pis. XXIV.-XXV. 



progressive, no matter whether it means in- 

 crease or decrease in numbers. A case of in- 

 creasing tendency is, for instance, the stria- 

 tion of certain species of the lizard genus 

 Cnemidophorus, of which the small G. deppei 

 varies in the possession of seven to eleven 

 white stripes, the eleven-striped individuals 

 and the six-striped G. sexlineatus representing 

 the two extremes. A study of these lizards 

 has convinced me that in many cases a limited 

 increase in the number of stripes takes place 

 during individual growth. For detail I must 

 be permitted to refer to my paper.^ 



Mr. Coker does not believe that orthogenetic 

 variation (or, as he says, ' determinate') exists 

 in the individual turtle. He holds that my 

 assumption can not be proved by percentages. 

 Whilst I rejected the notion that the irregular 

 number of scutes has any serious influence 

 upon the successful life of the turtles, he 

 thinks it is quite conceivable. But ' at least 

 the more extreme cases may be but one of the 

 evidences of a congenital weakness,' so ' that 

 a greater proportion of abnormal than of nor- 

 mal turtles fail in the struggle.' Of course, 

 this may be the case. But how are we to find 

 it out, unless we watch a number of them 

 growing up? The fact remains that there 

 are adults with very abnormal scutes which 

 do very well. Mr. Coker is indignant at my 

 saying that no less than fourteen per cent, of 

 large turtles were wrong, when I had only one 

 such abnormal specimen out of seven large 

 specimens. I grant that this was the wrong 

 way of putting the case. Let me, therefore, 

 mend the sentence as follows: 



No less than twenty-two per cent, of turtles of 

 the considerable size of two feet and upwards, are 

 wrong in their scutes, and do, or did very well, 

 for all we know to the contrary. 



On the other hand, he found that of 28 

 embiyos of T. caretta, of one nest, only 9 had 

 more than the normal number of 16 scutes, 

 with totals of 17, 18, 19 and 20; whilst two 

 had each one costal too few, making a total of 



^ ' Evolution of the Colour-pattern and Ortho- 

 genetic Variation in Certain Mexican Species of 

 Lizards with Adaptation to their Surroundings,' 

 Proc. Roy. 8oc. London, Vol. 72, May, 1903, pp. 

 109-125, pis. 3-5. 



