n8 



The Scottish Naturalist. 



being passed by, and remaining normal, or apparently normal. 

 All the flowers in an inflorescence are not doubled, but a few or 

 many may be so, varying in different plants. The bracts frequently 

 become abnormal ; in the species of Valeriana there sometimes 

 occur forms intermediate between bracts and rays of the pappus, 

 or some become petaloid, or even leaf-like. The latter is seen 

 only in severe attacks. In those cases in which the flowers are 

 doubled, I could recognise the doubled flowers with certainty as 

 such, only several weeks after infection was completed. The 

 doubled flowers appeared to be later in arriving at full opening 

 than were healthy flowers on axes of the same importance, and, of 

 course, they required a longer time to develope. 



If the plants under observation were less suited for the nutrition 

 of Phytoptus the number of abnormal organs was small. The 

 abnormalities occurred at a definite time. The leaves or flowers 

 that afterwards unfolded themselves remained normal, unless the 

 same plant were subjected at various intervals of time to serious 

 infection. The appearance of individual abnormal leaves or 

 flowers gave the impression of spontaneous variation. I am con- 

 vinced that many cases that have been explained as spontaneous 

 variation have been caused by the action of parasites, though a 

 Phytoptus must not be regarded as in all cases the agent. 



To those plants that are less fitted to give nourishment to the 

 Phytoptus belong several species of Valeriana, especially V. saxatilis, 

 celtica, saliunca, tuberosa, the species of Valerianella, viz. Szovit- 

 siana, vesicaria, and, as it appears, olitoria, and the Cruciferce. 



Other species of Valeriana, e.g. V. Phu and montana stand in 

 an intermediate position. V. saxatilis was repeatedly infected ; 

 but I obtained only metaschematic single flowers (none being 

 double), and in only one case the characteristic abnormal leaves 

 with the pectinate teeth. In V. celtica also I obtained the toothed 

 leaves ; but the teeth stood not in pectinate arrangement but 

 singly. V. saliunca presented obstacles to successful cultivation 

 but the infected specimens were soon abandoned by the Phytoptus ; 

 and in V. tuberosa I obtained only abnormally formed leaves. 

 The period at which infection was attempted in these plants was 

 probably not rightly hit off. 



In Valerianella vesicaria in fourteen days after infection there 

 appeared rolls along the margins of the leaves, but not affecting 

 the entire margin; these afterwards became broader, as in the 



