No matter what came up, I could not expect to find anyone at the staff headquarters of Semenoff before eleven o’clock. There might be a report at daylight that a train had been attacked by Bolshevists twenty miles away; it could not be verified before noon. If I got a message in cipher from Vladivostok at nine o’clock in the morning that Russians had attacked Japanese troops near Chita, and asking me to get the details, in fifteen minutes I could have the Japanese version of the affair, but it would take forty-eight hours to get the Russian statement of what had happened. If my telegram came over the regular Russian wire, it might be delivered to the hotel proprietor, who would call my attention to it the following week. In the meantime, messages for the Czecho-Slovak commandant, or a French correspondent of whom I had never heard, or some British officer, would be brought to my room at all hours of the night. My room was a receiving station for all telegrams which the Russian operators did not know where to deliver--and when one came for me, it was mis-delivered, generally. My cipher messages were transmitted as numbers, and when Russian operators received them, they always left out thirty or forty numerals in the middle of the message so that when it was being deciphered, the last half of it did not decipher at all, but when transformed into letters looked like an alphabet that had run _amok_. I generally asked for it to be repeated to my own signal corps operator over our private wire, but that wire was open for us but an hour in the forenoon and an hour in the afternoon. So if I got a defective message over the Russian wire five minutes after my wire was closed in the morning, I had to wait until afternoon before I could with any surety of getting the message through tell Vladivostok to repeat. And the repeated message would not reach me till the next morning. I suspect that there was a system of sabotage being practiced on us. There was every evidence that Semenoff’s officers in charge of the commercial telegraph office had held up our cipher messages until they had made an effort to decipher them. Or knowing that they had done something which they did not want reported or contemplated doing something, they held up the message as long as possible. For instance, if Semenoff had his armored train in the station and the engines with steam up ready to move up the line for some purpose, any message I sent through at that time was suspected to contain the information that his armored train was about to move in a certain direction. So the message was not put on the wire till the movement was completed.