I have found the source of the picture. 
This deadfall is called Haida deadfall trap (from Swanton 1908)
The discription reads as flows:
"In  deadfall  traps,  a  trigger  is  disrupted when the bait is taken, and a heavy object such as a  log falls on the bear.   Swanton (1905b:68-69) described a Haida deadfall trap, illustrated in Figure 1:
The  hadjigā’ñwa-i  (a)  are  four posts,  two on  each side  of a  bear trail.  These are fastened together in pair by the kiut!a’sk!î (b). 
Between them  lies  a  timber  called  the q!atA’nlanu (c), while the deadfall proper consists  of a  timber called sî’txasq!a’gida (d) hung above this at  one  end  and  weighted  at  the other  end,  which  rests  upon  the ground.  
The suspended end is held by a loop (lq!ō’ya-i), which passes over  a  short stick,  the x.ā’ña (e), which  is  supported  in  its  turn  by one  of  the kiut!a’sk!î.   A  rope is fastened  to  the  inner  end  of  this x.ā’ña  and  carried  down  to  the notch  in  another  stick  called sqaolg.ai’wa-  î  (f), which  is  fastened to  a stake  at one side  of the bear trail.
Other cords, qa-ĭ’tu (g), are then fastened between the two front posts and carried down to this loop.  The bear, coming against the latter,  in  its  endeavors  to  get through pulls the loop (h) out of the notch in the sqaolg.ai’wa- i.  This in turn releases the x.ā’ña allowing the sî’txasq!a’gida to fall upon the animal’s back."
Source: 
https://www.researchgate.net/public...ern_Northwest_Coast_of_North_America#pf9I thought it might be kind of a toggle trigger. 
But this design seems to be made for higher release force on purpose, likely to avoid small game triggering the trap. 
It would be interseting to know details how the notch in stick f was cut and how the cords where attached.