PREFACE. iii 



We at once notice that tliey all have perfect flowers, and, 

 therefore, consult § 4 of this key. Directed from there 

 to § 5, we separate our dozen into two heaps, on account of 

 the number of their stamens. Six of them have more 

 than 20 stamens, and six less. 



Let us attack first the half dozen with over 20 stamens. 

 In § G we ascertain that four of them have their stamens 

 distinct, while the other two have them united. To learn 

 the names of the four plants with distinct stamens, we con- 

 sult § 7, and there are led to observe that three of them have 

 their stamens inserted on the receptacle, while one has 

 his attached to the calyx. The plant with the stamens 

 inserted into the calyx must be looked for in 48 ; and 

 from there, as it has 3-12 ovaries, we proceed to 57. 

 The number of the perianth-pieces being definite, namely, 

 only five, we are directed to 58, where, ascertaining that 

 the fruit is a follicle, etc., we establish that the plant un- 

 der consideration is a Sjnrc^a. 



As to the three other plants wdth their distinct stamens 

 inserted on the receptacle, we are in § 8 advised to ascer- 

 tain the number of pistils. We find that two of the 

 plants have one pistil only, while one of them has five 

 pistils. As regards the latter, we are directed to 25, and 

 verifying there that the pistils are distinct, we proceed to 

 26 ; and as the pistils are not sunk in hollows of the re- 

 ceptacle, but free on its surface, to 27. The leaves being 

 not peltate (or shield-shaped), we are shown to 28, and 

 from there, as the plant is an lierb, to 29. The ovary be- 

 ing not one-ovuled, etc., we pass on to 3S, and from tliere, 

 the petals being present, to 41. The petals being spurred, 

 we consult § 42, where, the petals proving to be all alike, 

 it becomes evident that our phint is Aqwilegia Cana- 

 densi6\ L. 



