200 SCIENCE AND FRUIT GROWING 



is usually found that those trees which are attacked most freely, 

 are the ones which are the most stunted ; and it seems probable 

 that in many cases the scale is the consequence, and not the 

 cause, of this stunting. Thus, from a plantation where a certain 

 amount of scale existed, some 200 trees were moved, at an age 

 when removal checked their growth for several years : all the 

 moved trees were found to become covered with scale to a very 

 much greater extent than their former unmoved neighbours, 

 and the only possible conclusion was that stunting was the cause, 

 of this increase. Nor is it difficult to see the reason of this, 

 for a tree which is in active growth must, even during the winter, 

 experience such expansion and contraction of its bark as would 

 loosen the attachment of the scales to it, and render the removal 

 of these by heavy rain or other agents a simple matter. In the 

 case of the drier bark of a stunted tree, the scales would remain 

 unloosened. 1 Nevertheless, when the scale has increased to such 

 an extent as to cover the greater part of the bark, it cannot be 

 doubted that the amount of injury done by it will be very serious, 

 if not fatal. 



It may be laid down as a general rule that the examination 

 of the action of insecticides or fungicides should consist of two 

 parts : laboratory experiments, in which the action on the 

 insect or fungus to be destroyed is studied under the most exact 

 conditions possible ; and field experiments, in which the insecti- 

 cide is applied to the trees in the plantation itself, under the 

 variable conditions existing in practice. Laboratory experiments 

 are necessary, not only as forming the groundwork for trials in 

 the field, but they are essential in affording a check on the 

 results of the latter; for field trials may often fail owing to 

 a variety of accidental circumstances, generally dependent on 

 the weather, and, in the absence of more precise evidence to the 

 contrary, any failure of this sort may lead to very erroneous 

 conclusions. The investigations at Woburn afford striking 

 examples of cases where experiments, had they been confined 

 to trees growing in the open, would have led to nothing but 

 conflicting and misleading results. 



Laboratory experiments, however, should be rendered as nearly 

 similar as possible in all essential features to field trials. In 

 the case of those on mussel scale, they were made on the scales 

 while these were still attached to infested branches which had 

 been severed from the trees,' and they were kept in the open 



1 See Marlatt, Year Book of U.S. Dep. of Agric., 1900, p. 248. 



