THE initial reaction of hremostasis1 and arterial thrombosis•-• is adhesion and aggregation of blood platelets to the wound surface and to the site of intimal damage. The elucidation of the mechanism of this reaction is of great importance not only for diagnosis and management of hremorrhagic disorders, but also for the problem of thrombosis. Several factors are involved:(1) a property of the platelets which enables them to become adhesive when exposed to the physiological adhesiveness-producing mechanism (thrombasthenia represents by definition the congenital lack of this property);(2) a tissue substrate for the platelets to adhere to, probably collagen and elastic fibres•·•;(3) a plasma factor, which is lacking in von Willebrand's disease• and (4) calcium. A fifth factor, possibly triggering the adhesionaggregation reaction, was reported by Hellem as being an acidic, dialysable and heat-stable factor which was found in red blood cells8• This factor produced strong agglutination of platelets in citrated or oxalated plasma and converted the non-adhesive circulating platelets to adhesive platelets showing immediate adherence to foreign surfaces. This article describes the identification of this factor as adenosine diphosphate. The effect of a number of other nucleotides on the adhesiveness of human platelets has also been examined. The semi-quantitative test for the adhesiveness of human blood platelets described by Hellem•·• was used throughout this work. 0· 05 ml. of saline containing the material to be tested was added to 0· 5 mi. of platelet-rich plasma, and the reaction considered positive if clumping of the platelets occurred. An extract obtained by treating human red blood cells with boiling water8 was passed through a'Dowex lx 8'column in the formate form. The column was eluted stepwise with increasing concentration of formic acid and ammonium formate-formic acid mixtures. A typical elution diagram of ultra-violet absorbing material recorded at 260 m [L is seen in Fig. l. The eluted fractions (I-VI) were concentrated to dryness in vacuum in the frozen state, and then dissolved in a volume of saline equal to that of the original extract. Only fractions V and VI were positive in the platelet test. Occasionally the elution diagram would contain a large inactive peak at the site of fraction IV.

Fractions V and VI both yielded adenine on hydrolysis with N hydrochloric acid for 1 hr. at 100 C. as shown by paper chromatography in isopropanolfconcentratedhydrochloricacid/water (68: 17· 6: 14-4). In electrophoresis at pH 5· 7 in 0· 2 M ammonium acetate at 900 V. for 2 hr., fraction V gave two ultraviolet absorbing zones, one of which had the same mobility and ultra-violet absorption spectrum in 0· 1 N hydrochloric acid and in 0· 1 N sodium hydroxide as adenosine diphosphate. By chromatography of