8 Donald Reddick and V. B. Stewart 



was due to space limitations. (2j The varieties ' employed in'^ the test 

 were so nmnerous and the growth habits so different that comparisons 

 between varieties would lead to error. ' On the other hand, the number of 

 trees of any one variety was scarcely large enough to allow for satisfactory 

 comparisons. (3) The j-ield of the trees was expected to give quantitative 

 data. The trees had only just come mto bearing, and so these data were 

 not obtainable at the time when, of necessity, the experiment was ter- 

 minated. It is not hkely that the data would have been of sufficient 



FiGLTJE 2. BALDWIN TREE C-3 



The galls had increased somewhat in sizs. but there was no indication 

 that they had interfered with the development of the tree 



extent to be very reliable. (4) The apple-tree borer (Saperda Candida) 

 proved a great pest. Sometimes as many as six larvae were foimd in a 

 single tree. If these larvae are removed, httle injury vnR foUow; but 

 if one or two are missed — and this apparently is almost ine^Ttable — great 

 damage may be wrought within a few weeks. Most of the fluctuations 

 in general appearance of trees from j^ear to year, without question, are 

 attributable to this borer. (5) Frost, rabbits, mice, and other fortuitous 

 circumstances reduced the size of some trees so that, with the small 

 numbers involved, it did not seem worth while at the end even^'to weigh 

 the trunks and the crowns, although this could have been done very 

 easilv. 



