In a s e r i e s of recen t papers , 1-3 the ways in which hydrogen t rapping by lat t ice defects can affect and cont ro l the na ture , extent and k ine t ics of hydrogen embr i t t l emen t , have been success fu l ly analyzed and modeled. Most impor tant ly , it was shown that a m i c r o s t r uc tu r e cons is t ing of a un i form d i s t r ibu t ion of fine, s t rong, hydrogen t raps should maximize an a l l oy ' s r e s i s tance to hydrogen. There fo re , if one is to des ign al loys exhibit ing such a m i c r o s t r u c t u r e , the need a r i s e s to know what kind of t r aps should be used. The a i m of this c lass i f i ca t ion is to answer that need. When a hydrogen atom jumps f rom a n o r m a l lat t ice s i te into a t rap , the probabi l i ty of an eventual r e t u r n to the f o r m e r si te is dec reased . The re a re two main r ea sons why, in a c rys t a l lat t ice, jump probab i l i t i e s should be modified: There exis t s a force that pushes the hydrogen a tom in a p r e f e r r e d d i rec t ion , F ig . 1 (a). In that case, the lat t ice is not m o d i f i e d i . e , the average jump height is u n c h a n g e d b u t it is ea s i e r , and thus more probable , for an a tom in si te B to jump in si te A than to jump in s i te C. Moreover , the backward jump will be more difficult than the forward jump. The a tom can then be sa id to be a t t rac ted to si te A. The lat t ice is d i s tor ted and may even be comple te ly modified, F ig . l(b). This t ime, the average jump height is changing f rom si te to s i te . These r ea sons a re suff icient to a r b i t r a r i l y d i s t in guish between two ex t reme types of t r aps . T r a p s where both c h a r a c t e r s a t t r a c t i v e and p h y s i c a l c o e x i s t wil l be designated as mixed t r aps . At t rac t ive T r a p s . Such a t r ap is i l l u s t r a t ed in Fig . l (c) . It is a region of the la t t ice , of d imens ion Di, where hydrogen atoms are subjected to an a t t rac t ive force . F o r c e s that could act on a diffusing a tom in a c ry s t a l la t t ice a re of four types, 4 and a re due e i ther to e l ec t r i c a l f ields, s t r e s s f ields, t e m p e r a t u r e g r a dients , or to the nonideal par t of a chemica l potent ia l g rad ien t . Because a hydrogen atom d i s so lves in t r a n s i t ion meta ls by giving up i ts excess e l ec t ron to the col lect ive e l ec t ron gas of the metal , 5 any defect that in t roduces an e lec t ron vacancy wil l a t t r ac t hydrogen to achieve local neu t ra l i ty . This wil l be the case for impur i t i e s on the left of the ma t r ix in the per iodic table.6 This is a typical example of an " e l e c t r i c a l f o r c e " . The second type of force is well documented in the l i t e r a tu r e , 5,7 and owes its o r ig in to t ens i l e " s t r e s s f i e ld s " induced by defects such as d i s loca t ions , coherent , and s emicohe ren t g ra in boundar ies and pa r t i c l e s , c r ack t ips , and so for th . Since the so lub i l i ty of hydrogen in s tee l i nc r ea se s with t e m p e r a t u r e , a t t rac t ive forces will be genera ted by " t h e r m a l g r a d i e n t s " whenever heterogeneous t e m p e r a t u r e d i s t r i b u -