这个有点奇怪

有人写了一个我认为不起作用的sql，但它确实可以，而且还返回了正确的结果。我已经写了一个简化的示例，但我认为它说明了这一点。

drop table #client;
drop table #transactions;
drop table #history;

create table #client (
    clientId int,
    name varchar(50)
);
create table #transactions (
    transid int,
    clientId int,
    Amount int
);
create table #history (
    transid int,
    Amount int
);

insert into #client values (1, 'User 1');
insert into #client values (2, 'User 2');
insert into #client values (3, 'User 3');

insert into #transactions values (1, 1, 50);
insert into #transactions values (2, 1, 35);
insert into #transactions values (3, 1, 25);

insert into #transactions values (4, 2, 10);
insert into #transactions values (5, 2, 50);
insert into #transactions values (6, 1, 35);
insert into #transactions values (7, 3, 25);
insert into #transactions values (8, 3, 10);

insert into #history values (1, 50);
insert into #history values (2, 35);
insert into #history values (3, 25);
insert into #history values (4, 10);
insert into #history values (5, 50);
insert into #history values (6, 35);
insert into #history values (7, 25);
insert into #history values (8, 10);

select * from #history
join #transactions on #history.transid = #transactions.transid
join #client on #transactions.clientId = #client.clientId and #history.transid = #transactions.transid

最后一个联接将3个表联接在一起，这在我对编写SQL的了解中是行不通的。我错了吗（可能是这样）