BOTANICAL GAZETTE. 41 



in which a great part of the spikelets in the lower portion of the spike 

 are double, or in pairs, contrary to the generic character.'" The 

 above remarks were sent me by Mr. Boott, on seeing my lists of 1878 

 and 1879, which contain some remarkable forms of Triticum repens. 

 No. 1578 (from Colorado) is characterized generally by involute, nar- 

 row, rigid, faindy nerved leaves, glumes and palets but slightly nerv- 

 ed, short awned. Some specimens have 9 joints to the rachis of 

 spike, lowest joint with single spikelet (as in normal specimens of T. 

 repens), the upper three the same, but all the rest with double spike- 

 lets at each joint. Another has 11 joints, lowest one double, all the 

 rest single. From the same root is another stalk with 15 joints, low- 

 est one single, next three double, rest sintzle. Others have 13 joints, 



7 double (always the lowest double unless stated otherwise); 18 joints, 



8 double; 19 joints, 12 double; 22 joints, 15 double. From the 

 same root as last is one with 22 joints, lowest single, next nine dou- 

 ble. From Utah, under Nos. 1004a and 15 16, are two very distinct 

 forms, (i)"the mountain form with broad, flat, green, nerved leaves; 

 broad, acute, green, conspicuously nerved glumes and palets; (2) 

 the form of the dry and heated valleys, with very long and narrow 

 spikelets (over one inch long), whole plant glaucous ; leaves rigid, 

 involute, rather short, narrow ; glumes and palets almost horny, 

 smooth, scarcely nerved. Under (i), I have two specimens with 13 

 joints, lowest three triple spikeleted, rest double. Others have 10 to 

 15 joints, lowest one double, all the rest single. Others have 25 to 

 30 joints (very lone; spikes), lowest five double. Under (2), are some 

 with about 18 joints, lower half all double. Others have 18 jomts, 

 all single but the third from the bottom, which is double. One other 

 has 18 joints, lowest one double, next two single, next three double, 

 the rest single. 



I have many specimens of the most remarkable forms (besides 

 those already sent out) as well as very many of these forms of T. re- 

 pens with only single spikelets. I have shown the most remarkable 

 forms to Dr. Vasey, who considers them remarkable forms of T. fe- 

 pens. These forms with double spikelets are not uncommon, for I 

 have found them in many places in Colorado and Utah. 



That the distinction of double spikelets in Elymus tends to be 

 broken is shown in forms of E. condensatiis and triticoides which often 

 have the lowest joint single spikeleted, and the upper five also. Two 

 or more single spikelets often occur in both the above and in E. are- 

 narius, Sibiricus and occasionally in E. Canadensis. 



In most of the eastern species, the glumes of Elymus form an ap- 

 parent involucre, and to the amateur do not seem to be glumes, but 

 in Elymus arenarius, L. and condensatus, Presl. , in all the spikelets of 

 the former and the upper ones of the latter, they are attached to the 

 spikelets as closely as in Tritiaan repens, and more closely than in T. 

 ■violaceum, especially the long awned form. A question asked by one 

 of our leading botanists will find a good place here : "What is there 

 to distinguish Elymus Sibiricus from Triticum violaceum but the double 

 and single spikelets," and if these fail, what then ?- Marcus E, 



JONES. 



